pre-release

The changes to remove LCK_NONBLOCK from the LVM locks broke clvmd because the
code was clearly wrong but working anyway! The constant was being masked rather
than the variable that was supposed to match against it.

INSTALL

@@ -1,44 +1,30 @@
-LVM2 installation
-=================
+Installation
+============
 
-1) Install device-mapper
-
-   Ensure the device-mapper has been installed on the machine.
-
-   The device-mapper should be in the kernel (look for 'device-mapper'
-   messages in the kernel logs) and /usr/include/libdevmapper.h 
-   and libdevmapper.so should be present.
-
-   The device-mapper is available from:
-     ftp://ftp.sistina.com/pub/LVM2/device-mapper/
-
-
-2) Generate custom makefiles.
+1) Generate custom makefiles.
 
    Run the 'configure' script from the top directory.
 
-   If you wish to use the built-in LVM2 shell and have GNU readline 
-   installed (http://www.gnu.org/directory/readline.html) use:
-     ./configure --enable-readline
-
    If you don't want to include the LVM1 backwards-compatibility code use:
      ./configure --with-lvm1=none 
 
    To separate the LVM1 support into a shared library loaded by lvm.conf use:
      ./configure --with-lvm1=shared
 
+   Use ./configure --help to see other options.
 
-3) Build and install LVM2.
+2) Build and install.
 
-   Run 'make install' from the top directory.
+   Run 'make' from the top directory to build everything you configured.
+   Run 'make install' to build and install everything you configured.
 
+   If you only want the device-mapper libraries and tools use
+   'make device-mapper' or 'make install_device-mapper'.
 
-4) Create a configuration file
+3) If using LVM2, create a configuration file.
 
    The tools will work fine without a configuration file being
    present, but you ought to review the example file in doc/example.conf.
-   For example, specifying the devices that LVM2 is to use can
-   make the tools run more efficiently - and avoid scanning /dev/cdrom!
 
 Please also refer to the WHATS_NEW file and the manual pages for the 
 individual commands.

Makefile.in

@@ -1,6 +1,6 @@
 #
 # Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
-# Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+# Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
 #
 # This file is part of LVM2.
 #
@@ -18,43 +18,55 @@ VPATH = @srcdir@
 
 SUBDIRS = doc include man scripts
 
+ifeq ("@UDEV_RULES@", "yes")
+  SUBDIRS += udev
+endif
+
 ifeq ("@INTL@", "yes")
   SUBDIRS += po
 endif
 
-SUBDIRS += lib tools daemons
+SUBDIRS += lib tools daemons libdm
 
-ifeq ("@DMEVENTD@", "yes")
-  SUBDIRS += dmeventd
+ifeq ("@APPLIB@", "yes")
+  SUBDIRS += liblvm
 endif
 
 ifeq ($(MAKECMDGOALS),distclean)
   SUBDIRS += daemons/clvmd \
-	     dmeventd \
+	     daemons/dmeventd/plugins \
+	     daemons/dmeventd \
 	     lib/format1 \
 	     lib/format_pool \
 	     lib/locking \
 	     lib/mirror \
 	     lib/snapshot \
-	     po \
-	     test/mm test/device test/format1 test/regex test/filters
-  DISTCLEAN_TARGETS += lib/misc/configure.h
+	     test/api \
+	     test \
+	     po
+  DISTCLEAN_TARGETS += lib/misc/configure.h lib/misc/lvm-version.h
+  DISTCLEAN_DIRS += lcov_reports*
 endif
 
 include make.tmpl
 
-daemons: lib
-lib: include
-tools: lib
-dmeventd: tools
-po: tools daemons dmeventd
+libdm: include
+lib: libdm
+liblvm: lib
+daemons: lib tools
+tools: lib device-mapper
+po: tools daemons
+
+libdm.device-mapper: include.device-mapper
+daemons.device-mapper: libdm.device-mapper
+tools.device-mapper: libdm.device-mapper
+device-mapper: tools.device-mapper daemons.device-mapper man.device-mapper
 
 ifeq ("@INTL@", "yes")
 lib.pofile: include.pofile
 tools.pofile: lib.pofile
 daemons.pofile: lib.pofile
-dmeventd.pofile: tools.pofile
-po.pofile: tools.pofile daemons.pofile dmeventd.pofile
+po.pofile: tools.pofile daemons.pofile
 pofile: po.pofile
 endif
 
@@ -71,3 +83,39 @@ endif
 
 check: all
 	$(MAKE) -C test all
+
+ifneq ("@LCOV@", "")
+.PHONY: lcov-reset lcov lcov-dated
+
+ifeq ($(MAKECMDGOALS),lcov-dated)
+LCOV_REPORTS_DIR=$(top_srcdir)/lcov_reports-$(shell date +%Y%m%d%k%M%S)
+else
+LCOV_REPORTS_DIR=$(top_srcdir)/lcov_reports
+endif
+
+lcov-reset:
+	$(LCOV) -d $(top_srcdir)/dmeventd --zerocounters
+	$(LCOV) -d $(top_srcdir)/libdm --zerocounters
+	$(LCOV) -d $(top_srcdir)/lib --zerocounters
+	$(LCOV) -d $(top_srcdir)/tools --zerocounters
+
+lcov: all
+	$(RM) -rf $(LCOV_REPORTS_DIR)
+	$(MKDIR_P) $(LCOV_REPORTS_DIR)
+	$(LCOV) -b ${top_srcdir}/libdm -d $(top_srcdir)/libdm -c -o $(LCOV_REPORTS_DIR)/libdm.info
+	$(LCOV) -b $(top_srcdir)/lib -d $(top_srcdir)/lib -c -o $(LCOV_REPORTS_DIR)/lib.info
+	$(LCOV) -b $(top_srcdir)/tools -d $(top_srcdir)/tools -c -o $(LCOV_REPORTS_DIR)/tools.info
+	DMEVENTD_INFO="$(LCOV_REPORTS_DIR)/dmeventd.info" ;\
+	DMEVENTD_INFO_A="-a $$DMEVENTDINFO" ;\
+	$(LCOV) -b $(top_srcdir)/dmeventd -d $(top_srcdir)/dmeventd -c -o $$DMEVENTD_INFO || DMEVENTD_INFO_A="" ;\
+	$(LCOV) $$DMEVENTD_INFO_A -a $(LCOV_REPORTS_DIR)/lib.info \
+		-a $(LCOV_REPORTS_DIR)/libdm.info \
+		-a $(LCOV_REPORTS_DIR)/tools.info \
+		-o $(LCOV_REPORTS_DIR)/lvm.info
+ifneq ("@GENHTML@", "")
+	$(GENHTML) -o $(LCOV_REPORTS_DIR) -p $(top_srcdir) $(LCOV_REPORTS_DIR)/lvm.info
+endif
+
+lcov-dated: lcov
+
+endif

README

@@ -1,23 +1,27 @@
-This directory contains LVM2, the new version of the userland LVM
-tools designed for the new device-mapper for the Linux kernel.
+This tree contains the LVM2 and device-mapper tools and libraries.
 
-The device-mapper needs to be installed before compiling these LVM2 tools.
-
-For more information about LVM2 read the WHATS_NEW file.
+For more information about LVM2 read the changelog in the WHATS_NEW file.
 Installation instructions are in INSTALL.
 
 There is no warranty - see COPYING and COPYING.LIB.
 
 Tarballs are available from:
   ftp://sources.redhat.com/pub/lvm2/
-  ftp://sources.redhat.com/pub/dm/
 
 To access the CVS tree use:
   cvs -d :pserver:cvs@sources.redhat.com:/cvs/lvm2 login
   CVS password: cvs
   cvs -d :pserver:cvs@sources.redhat.com:/cvs/lvm2 co LVM2
 
-Mailing list for discussion/bug reports etc.
+Mailing list for general discussion related to LVM2:
   linux-lvm@redhat.com
   Subscribe from https://www.redhat.com/mailman/listinfo/linux-lvm
 
+Mailing list for LVM2 development, patches and commits:
+  lvm-devel@redhat.com
+  Subscribe from https://www.redhat.com/mailman/listinfo/linux-lvm
+
+Mailing list for device-mapper development, including kernel patches
+and multipath-tools:
+  dm-devel@redhat.com
+  Subscribe from https://www.redhat.com/mailman/listinfo/dm-devel

VERSION

@@ -1 +1 @@
-2.02.29-cvs (2007-08-24)
+2.02.51(1)-cvs (2009-08-06)

VERSION_DM

@@ -0,0 +1 @@
+1.02.36-cvs (2009-08-06)

WHATS_NEW

@@ -1,5 +1,580 @@
-Version 2.02.29 -
-==================================
+Version 2.02.51 - 6th August 2009
+=================================
+  Fix locking in clvmd (2.02.50).
+  Add --noudevsync option for relevant LVM tools.
+  Add activation/udev_sync to lvm.conf.
+  Only change LV symlinks on ACTIVATE not PRELOAD.
+  Make lvconvert honour log mirror options combined with downconversion.
+  Allow LV suspend while --ignorelockingfailure is in force.
+  Update synopsis in lvconvert manpage to mention --repair.
+  Set cookies in activation code and wait for udev to complete processing.
+  Added configure --enable-udev_rules --enable-udev_sync.
+  Added configure --with-udev-prefix --with-udevdir.
+  Added udev dir to hold udev rules.
+  Add devices/data_alignment_detection to lvm.conf.
+  Add devices/data_alignment_offset_detection to lvm.conf.
+  Add --dataalignmentoffset to pvcreate to shift start of aligned data area.
+  Fix _mda_setup() to not check first mda's size before pe_align rounding.
+  Document -I option of clvmd in the man page.
+  Fix configure script to handle multiple clvmd selections.
+  Fix lvm2app.pc installation filename.
+  Remove pv_t, vg_t & lv_t handles from lib.  Only liblvm uses them.
+  Rename lvm.h to lvm2app.h for now.
+
+Version 2.02.50 - 28th July 2009
+================================
+  Change test/api/test.c prompt so it's not confused with the main lvm prompt.
+  Update liblvm unit tests in test/api to cover latest liblvm changes.
+  Add unimplemented lvm_lv_resize and lvm_pv_resize skeletons to liblvm.
+  Add lvm_library_get_version to liblvm.
+  Add lvm_config_override to liblvm to allow caller to override LVM config.
+  Add lvm_lv_is_active and lvm_lv_is_suspended to liblvm.
+  Add lvm_lv_activate and lvm_lv_deactivate to liblvm.
+  Add lvm_scan, lvm_vg_reduce and lvm_vg_remove_lv to liblvm.
+  Add functions to get numeric properties to liblvm.
+  Add lvm_{pv|vg|lv}_get_{name|uuid} to liblvm.
+  Add lvm_vg_list_pvs and lvm_vg_list_lvs to liblvm.
+  Add lvm_vg_open and lvm_vg_create_lv_linear to liblvm.
+  Add lvm_list_vg_names/uuids to liblvm.
+  Add lvm_errno and lvm_errmsg to liblvm to obtain failure information.
+  Rename lvm_create/destroy to lvm_init/quit.
+  Rename lvm_reload_config to lvm_config_reload.
+  Refactor _override_settings to use new override_config_tree_from_string.
+  Add vg_reduce to metadata.c and metadata-exported.h.
+  Update lvm.h to clarify API behavior and return codes.
+  Update lvm_vg_extend to do an implicit pvcreate on the device.
+  Update display.c to use vg_free(vg) instead of duplicating the calculation.
+  Refactor vg_size, vg_free, and pv_mda_count field calculations for liblvm.
+  Refactor pvcreate and lvcreate for liblvm.
+  Add global/wait_for_locks to lvm.conf so blocking for locks can be disabled.
+  All LV locks are non-blocking so remove LCK_NONBLOCK from separate macros.
+  Fix race condition with vgcreate and vgextend on same device (2.02.49).
+  Remove redundant validate_name call from vgreduce.
+  Remove unused handles lvseg, pvseg inside liblvm/lvm.h.
+  Add liblvm2app Makefile installation targets.
+  Add liblvm pkgconfig file.
+  Use newly-independent LVM_LIBAPI in liblvm soname.  E.g. liblvm2app.so.2.1.
+  Add an API version number, LVM_LIBAPI, to the VERSION string for liblvm.
+  Pass a pointer to struct cmd_context to init_multiple_segtypes
+  Return EINVALID_CMD_LINE not success when invalid VG name format is used.
+  Remove unnecessary messages after vgcreate/vgsplit refactor (2.02.49).
+  Add log_errno to set a specific errno and replace log_error in due course.
+  Change create_toolcontext to still return an object if it fails part-way.
+  Add EUNCLASSIFIED (-1) as the default LVM errno code.
+  Store any errno and error messages issued while processing each command.
+  Use log_error macro consistently throughout in place of log_err.
+
+Version 2.02.49 - 15th July 2009
+================================
+  Exclude VG_GLOBAL from vg_write_lock_held so scans open devs read-only again.
+  Add unit test case for liblvm VG create/delete APIs.
+  Add liblvm APIs to implement creation and deletion of VGs.
+  Initialize cmd->cmd_line to "liblvm" in new liblvm library.
+  Place handles to liblvm objects for pv, vg, lv, lvseg, pvseg inside lvm.h.
+  Refactor vgsplit and vgextend to remove READ_REQUIRE_RESIZEABLE flag.
+  Use _exit() not exit() after forking to avoid flushing libc buffers twice.
+  Add cast to log_info arg in _find_labeller to avoid Sparc64 warning.
+  Make cmd->cmd_line const.
+  Fix dev name mismatch in vgcreate man page example.
+  Refactor vg_remove_single for use in liblvm.
+  Make all tools use consistent lock ordering obtaining VG_ORPHAN lock second.
+  Check md devices for a partition table during device scan.
+  Add extended device (blkext) and md partition (mdp) types to filters.
+  Make text metadata read errors for segment areas more precise.
+  Fix text segment metadata read errors to mention correct segment name.
+  Include segment and LV names in text segment import error messages.
+  Add parent node to config_node structure.
+  Update vgsplit and vgcreate to call new vg_create and 'set' functions.
+  Change vg_create to take minimal parameters, obtain a lock, and return vg_t.
+  Refactor vgchange extent_size, max_lv, max_pv, and alloc_policy for liblvm.
+  Update t-vgcreate-usage.sh to test for default vg properties.
+  Fix memory leak in vgsplit when re-reading the vg.
+  Make various exit/cleanup paths more robust after lvm init failures.
+  Use LCK_NONBLOCK implicitly instead of explicit vg_read() flag.
+  Remove unnecessary locking and existence tests from new vg_read() interface.
+  Permit several segment types to be registered by a single shared object.
+  Update the man pages to document size units uniformly.
+  Allow commandline sizes to be specified in terms of bytes and sectors.
+  Update 'md_chunk_alignment' to use stripe-width to align PV data area.
+  Update test/t-inconsistent-metadata.sh to match new vg_read interface.
+  Add lvmcache_init() to polldaemon initialization.
+  Convert tools to use new vg_read / vg_read_for_update.
+  Fix segfault in vg_release when vg->cmd is NULL.
+
+Version 2.02.48 - 30th June 2009
+================================
+  Abort if automatic metadata correction fails when reading VG to update it.
+  Explicitly request fallback to default major number in device mapper.
+  Ignore suspended devices during repair.
+  Call vgreduce --removemissing automatically to fix missing PVs in dmeventd.
+  Suggest using lvchange --resync when adding leg to not-yet-synced mirror.
+  Destroy toolcontext on clvmd exit to avoid memory pool leaks.
+  Fix lvconvert not to poll mirror if no conversion in progress.
+  Fix memory leaks in toolcontext error path.
+  Reinstate partial activation support in clustered mode. (2.02.40)
+  Allow metadata correction even when PVs are missing.
+  Use 'lvm lvresize' instead of 'lvresize' in fsadm.
+  Do not use '-n' realine option in fsadm for busybox compatiblity.
+  Add vg_lock_newname() library function for vgrename, vgsplit and vgcreate.
+  Round up requested readahead to at least one page and print warning.
+  Try to repair vg before actual vgremove when force flag provided.
+  Fix possible double release of VG after recovery.
+  Add parameter to process_each_vg specifying what to do with inconsistent VG.
+  Unify error messages when processing inconsistent volume group.
+  Use lvconvert --repair instead of vgreduce in mirror dmeventd DSO.
+  Introduce lvconvert --use_policies (repair policy according to lvm.conf).
+  Update clvmd-corosync to match new corosync API.
+  Fix lib Makefile to include any shared libraries in default target.
+  Fix rename of active snapshot with virtual origin.
+  Fix convert polling to ignore LV with different UUID.
+  Cache underlying device readahead only before activation calls.
+  Fix segfault when calculating readahead on missing device in vgreduce.
+  Remove verbose 'visited' messages.
+  Handle multi-extent mirror log allocation when smallest PV has only 1 extent.
+  Add LSB standard headers and functions (incl. reload) to clvmd initscript.
+  When creating new LV, double-check that name is not already in use.
+  Remove /dev/vgname/lvname symlink automatically if LV is no longer visible.
+  Rename internal vorigin LV to match visible LV.
+  Suppress 'removed' messages displayed when internal LVs are removed.
+  Fix lvchange -a and -p for sparse LVs.
+  Fix lvcreate --virtualsize to activate the new device immediately.
+  Make --snapshot optional with lvcreate --virtualsize.
+  Generalise --virtualoriginsize to --virtualsize.
+  Skip virtual origins in process_each_lv_in_vg() without --all.
+  Fix counting of virtual origin LVs in vg_validate.
+  Attempt to load dm-zero module if zero target needed but not present.
+
+Version 2.02.47 - 22nd May 2009
+===============================
+  Rename liblvm.so to liblvm2app.so and use configure --enable-applib.
+  Reinstate version in liblvm2cmd.so soname. (2.02.44)
+
+Version 2.02.46 - 21st May 2009
+===============================
+  Inherit readahead setting from underlying devices during activation.
+  Detect LVs active on remote nodes by querying locks if supported.
+  Enable online resizing of mirrors.
+  Use suspend with flush when device size was changed during table preload.
+  Implement query_resource_fn for cluster_locking.
+  Support query_resource_fn in locking modules.
+  Introduce CLVMD_CMD_LOCK_QUERY command for clvmd.
+  Fix pvmove to revert operation if temporary mirror creation fails.
+  Fix metadata export for VG with missing PVs.
+  Add vgimportclone and install it and the man page by default.
+  Force max_lv restriction only for newly created LV.
+  Remove unneeded import parameter from lv_create_empty.
+  Merge lv_is_displayable and lv_is_visible functions.
+  Introduce lv_set_visible & lv_set_hidden functions.
+  Fix lv_is_visible to handle virtual origin.
+  Introduce link_lv_to_vg and unlink_lv_from_vg functions.
+  Remove lv_count from VG and use counter function instead.
+  Fix snapshot segment import to not use duplicate segments & replace.
+  Do not query nonexistent devices for readahead.
+  Remove NON_BLOCKING lock flag from tools and set a policy to auto-set.
+  Remove snapshot_count from VG and use function instead.
+  Fix first_seg() call for empty segment list.
+  Add install_lvm2 makefile target to install only the LVM2 components.
+  Reject missing PVs from allocation in toollib.
+  Fix PV datalignment for values starting prior to MDA area. (2.02.45)
+  Add sparse devices: lvcreate -s --virtualoriginsize (hidden zero origin).
+  Fix minimum width of devices column in reports.
+  Add lvs origin_size field.
+  Fix linux configure --enable-debug to exclude -O2.
+  Implement lvconvert --repair for repairing partially-failed mirrors.
+  Fix vgreduce --removemissing failure exit code.
+  Fix remote metadata backup for clvmd.
+  Introduce unlock_and_release_vg macro.
+  Introduce vg_release() to be called to free every struct volume_group.
+  Alloc PV internal structure from VG mempool if possible.
+  Fix metadata backup to run after vg_commit always.
+  Tidy clvmd volume lock cache functions.
+  Fix pvs report for orphan PVs when segment attributes are requested.
+  Fix pvs -a output to not read volume groups from non-PV devices.
+  Add MMC (mmcblk) device type to filters.
+  Introduce memory pools per volume group (to reduce memory for large VGs).
+  Use copy of PV structure when manipulating global PV lists.
+  Always return exit error status when locking of volume group fails.
+  Fix mirror log convert validation question.
+  Avoid referencing files from DESTDIR during build process.
+  Avoid creating some static libraries unless configured --enable-static_link.
+  Enable use of cached metadata for pvs and pvdisplay commands.
+  Add missing 'device-mapper' internal subdir build dependency.
+  Fix memory leak in mirror allocation code.
+  Save and restore the previous logging level when log level is changed.
+  Fix error message when archive initialization fails.
+  Make sure clvmd-corosync releases the lockspace when it exits.
+  Fix segfault for vgcfgrestore on VG with missing PVs.
+  Block SIGTERM & SIGINT in clvmd subthreads.
+  Detect and conditionally wipe swapspace signatures in pvcreate.
+  Fix maximal volume count check for snapshots if max_lv set for volume group.
+  Fix lvcreate to remove unused cow volume if the snapshot creation fails.
+  Fix error messages when PV uuid or pe_start reading fails.
+  Build new liblvm application-level library.
+  Rename liblvm.a to liblvm-internal.a.
+  Flush memory pool and fix locking in clvmd refresh and backup command.
+  Fix unlocks in clvmd-corosync. (2.02.45)
+  Fix error message when adding metadata directory to internal list fails.
+  Fix size and error message of memory allocation at backup initialization.
+  Remove old metadata backup file after renaming VG.
+  Restore log_suppress state when metadata backup file is up-to-date.
+
+Version 2.02.45 - 3rd March 2009
+================================
+  Avoid scanning empty metadata areas for VG names.
+  Attempt proper clean up in child before executing new binary in exec_cmd().
+  Do not scan devices if reporting only attributes from PV label.
+  Use pkgconfig to obtain corosync library details during configuration.
+  Fix error returns in clvmd-corosync interface to DLM.
+  Add --refresh to vgchange and vgmknodes man pages.
+  Pass --test from lvresize to fsadm as --dry-run.
+  Supply argv[] list to exec_cmd() to allow for variable number of parameters.
+  Prevent fsadm from checking mounted filesystems.
+  No longer treats any other key as 'no' when prompting in fsadm.
+  Tidy fsadm command line processing.
+  Add lib/lvm.h and lib/lvm_base.c for the new library interface.
+  Move tools/version.h to lib/misc/lvm-version.h.
+  Split LVM_VERSION into MAJOR, MINOR, PATCHLEVEL, RELEASE and RELEASE_DATE.
+  Add system_dir parameter to create_toolcontext().
+  Add --dataalignment to pvcreate to specify alignment of data area.
+  Exclude LCK_CACHE locks from _vg_lock_count, fixing interrupt unblocking.
+  Provide da and mda locations in debug message when writing text format label.
+  Mention the restriction on file descriptors at invocation on the lvm man page.
+  Index cached vgmetadata by vgid not vgname to cope with duplicate vgnames.
+  No longer require kernel and metadata major numbers to match.
+  Add a fully-functional get_cluster_name() to clvmd corosync interface.
+  Remove duplicate cpg_initialize from clvmd startup.
+  Add option to /etc/sysconfig/cluster to select cluster type for clvmd.
+  Allow clvmd to start up if its lockspace already exists.
+  Separate PV label attributes which do not need parse metadata when reporting.
+  Remove external dependency on the 'cut' command from fsadm.
+  Fix pvs segfault when pv mda attributes requested for not available PV.
+  Add fsadm support for reszing ext4 filesysystems.
+  Move locking_type reading inside init_locking().
+  Rename get_vgs() to get_vgnames() and clarify related error messages.
+  Allow clvmd to be built with all cluster managers & select one on cmdline.
+  Mention --with-clvmd=corosync in ./configure.
+  Replace internal vg_check_status() implementation.
+  Rename vg_read() to vg_read_internal().
+
+Version 2.02.44 - 26th January 2009
+===================================
+  Fix --enable-static_link after the recent repository changes.
+  Add corosync/DLM cluster interface to clvmd.
+  Add --nameprefixes, --unquoted, --rows to pvs, vgs, lvs man pages.
+  Fix lvresize size conversion for fsadm when block size is not 1K.
+  Fix pvs segfault when run with orphan PV and some VG fields.
+  Display a 'dev_size' of zero for missing devices in reports.
+  Add pv_mda_size to pvs and vg_mda_size to vgs.
+  Fix lvmdump /sys listing to include virtual devices directory.
+  Add "--refresh" functionality to vgchange and vgmknodes.
+  Avoid exceeding LV size when wiping device.
+  Calculate mirror log size instead of using 1 extent.
+  Ensure requested device number is available before activating with it.
+  Fix incorrect exit status from 'help <command>'.
+  Fix vgrename using UUID if there are VGs with identical names.
+  Fix segfault when invalid field given in reporting commands.
+  Move is_static from cmd to global is_static().
+  Refactor init_lvm() for lvmcmdline and clvmd.
+  Add liblvm interactive test infrastructure to build.
+  Add skeleton lvm2.h file in preparation for a shared library interface.
+  Use better random seed value in temp file creation.
+  Add read_urandom to read /dev/urandom. Use in uuid calculation.
+  Use displayable_lvs_in_vg and lv_is_displayable for consistency throughout.
+  Fix race in vgcreate that would result in second caller overwriting first.
+  Fix uninitialised lv_count in vgdisplay -c.
+  Don't skip updating pvid hash when lvmcache_info struct got swapped.
+  Add tinfo to termcap search path for pld-linux.
+  Fix startup race in clvmd.
+  Generate Red Hat clvmd startup script at config time with correct paths.
+  Fix clvmd & dmeventd builds after tree restructuring.
+  Cope with snapshot dependencies when removing a whole VG with lvremove.
+  Make man pages and tool help text consistent using | for alternative options.
+
+Version 2.02.43 - 10th November 2008
+====================================
+  Merge device-mapper into the lvm2 tree.
+  Correct prototype for --permission on lvchange and lvcreate man pages.
+  Exit with non-zero status from vgdisplay if couldn't show any requested VG.
+  Move list.c into libdevmapper and rename functions.
+  Rename a couple of variables that matched function names.
+  Use simplified x.y.z version number in libdevmapper.pc.
+  Remove ancient debian directory.
+  Split out lvm-logging.h from log.h and lvm-globals.[ch] from log.[ch].
+
+Version 2.02.42 - 26th October 2008
+===================================
+  Accept locking fallback_to_* options in the global section as documented.
+  Fix temp table activation in mirror conversions not to happen in other cmds.
+  Fix temp table in mirror conversions to use always-present error not zero.
+
+Version 2.02.41 - 17th October 2008
+===================================
+  Use temp table to set device size when converting mirrors.
+  In resume_mirror_images replace activate_lv with resume_lv as workaround.
+  Avoid overwriting in-use on-disk text metadata by forgetting MDA_HEADER_SIZE.
+  Fix snapshot monitoring library to not cancel monitoring invalid snapshot.
+  Generate man pages from templates and include version.
+  Add usrlibdir and usrsbindir to configure.
+  Fix conversion of md chunk size into sectors.
+  Free text metadata buffer after a failure writing it.
+  Fix misleading error message when there are no allocatable extents in VG.
+  Fix handling of PVs which reappeared with old metadata version.
+  Fix mirror DSO to call vgreduce with proper parameters.
+  Fix validation of --minor and --major in lvcreate to require -My always.
+  Fix release: clvmd build, vgreduce consolidate & tests, /dev/ioerror warning.
+
+Version 2.02.40 - 19th September 2008
+=====================================
+  Allow lvremove to remove LVs from VGs with missing PVs.
+  In VG with PVs missing, by default allow activation of LVs that are complete.
+  Track PARTIAL_LV and MISSING_PV flags internally.
+  Require --force with --removemissing in vgreduce to remove partial LVs.
+  No longer write out PARTIAL flag into metadata backups.
+  Treat new default activation/missing_stripe_filler "error" as an error target.
+  Remove internal partial_mode.
+  Add devices/md_chunk_alignment to lvm.conf.
+  Pass struct physical_volume to pe_align and adjust for md chunk size.
+  Store sysfs location in struct cmd_context.
+  Avoid shuffling remaining mirror images when removing one, retaining primary.
+  Add missing LV error target activation in _remove_mirror_images.
+  Prevent resizing an LV while lvconvert is using it.
+  Avoid repeatedly wiping cache while VG_GLOBAL is held in vgscan & pvscan.
+  Fix pvresize to not allow resize if PV has two metadata areas.
+  Fix setting of volume limit count if converting to lvm1 format.
+  Fix vgconvert logical volume id metadata validation.
+  Fix lvmdump metadata gather option (-m) to work correctly.
+  Fix allocation bug in text metadata format write error path.
+  Fix vgcfgbackup to properly check filename if template is used.
+  configure aborts if lcov or genhtml are missing with --enable-profiling
+  vgremove tries to remove lv snapshot first.
+  Added function lv_remove_with_dependencies().
+  Improve file descriptor leak detection to display likely culprit and filename.
+  Change clustered mirror kernel module name from cmirror to dm-log-clustered.
+  Avoid looping forever in _pv_analyze_mda_raw used by pvck.
+  Change lvchange exit status to indicate if any part of the operation failed.
+  Fix pvchange and pvremove to handle PVs without mdas.
+  Refactor _text_pv_read and always return mda list if requested.
+  Fix configure to work w/o readline unless --enable-readline used. (2.02.39)
+  Remove is_lvm_partition template which has not yet been coded.
+  Refactor pvcreate to separate parameter parsing from validation logic.
+  Check for label_write() failure in _text_pv_write().
+  Add pvcreate tests and update vgsplit tests to handle lvm1 and lvm2 metadata.
+  Fix pvchange -M1 -u to preserve existing extent locations when there's a VG.
+  Cease recognising snapshot-in-use percentages returned by early devt kernels.
+  Add backward-compatible flags field to on-disk format_text metadata.
+  Fix dmeventd monitoring libraries to link against liblvm2cmd again. (2.02.39)
+
+Version 2.02.39 - 27th June 2008
+================================
+  Enable readline by default if available.
+  Update autoconf to 2008-01-16.
+  Add $DISTCLEAN_DIRS to make.tmpl.in.
+  Create coverage reports with --enable-profiling and make lcov or lcov-dated.
+  Fix up cache for PVs without mdas after consistent VG metadata is processed.
+  Update validation of safe mirror log type conversions in lvconvert.
+  Fix lvconvert to disallow snapshot and mirror combinations.
+  Fix reporting of LV fields alongside unallocated PV segments.
+  Add --unquoted and --rows to reporting tools.
+  Add and use uninitialized_var() macro to suppress invalid compiler warnings.
+  Introduce enum for md minor sb version to suppress compiler warning.
+  Avoid undefined return value after _memlock manipulation in lvm2_run.
+  Avoid link failure if configured without --enable-cmdlib or --enable-readline.
+  Make clvmd return at once if other nodes down in a gulm or openais cluster.
+  Fix and improve readahead 'auto' calculation for stripe_size.
+  Fix lvchange output for -r auto setting if auto is already set.
+  Add test case for readahead.
+  Avoid ambiguous use of identifier error_message_produced.
+  Begin syncing configure.in for merge/unification with device-mapper.
+  Fix add_mirror_images not to dereference uninitialized log_lv upon failure.
+  Don't call openlog for every debug line output by clvmd.
+  Add --force to lvextend and lvresize.
+  Fix vgchange not to activate component mirror volumes directly.
+  Fix test directory clean up in make distclean.
+
+Version 2.02.38 - 11th June 2008
+================================
+  Fix tracking of validity of PVs with no mdas in lvmcache.
+  Fix return values for reporting commands when run with no PVs, LVs, or VGs.
+  Add omitted unlock_vg() call when sigint_caught() during vg processing.
+  Fix free_count when reading pool metadata.
+  Fix segfault when using pvcreate on a device containing pool metadata.
+  Fix segfault after _free_vginfo by remembering to remove vginfo from list.
+  Tweak detection of invalid fid after changes to PVs in VG in _vg_read.
+  Revert assuming precommitted metadata is live when activating (unnecessary).
+  Drop cached metadata for disappearing VG in vgmerge.
+  In script-processing mode, stop if any command fails.
+  Warn if command exits with non-zero status code without a prior log_error.
+  Check lv_count in vg_validate.
+  Add --nameprefixes to reporting tools for field name prefix output format.
+
+Version 2.02.37 - 6th June 2008
+===============================
+  Make clvmd-cman use a hash rather than an array for node updown info.
+  Correct config file line numbers in messages when parsing comments.
+  Drop cached metadata when renaming a VG.
+  Allow for vginfo changing during _vg_read.
+  Decode numbers in clvmd debugging output.
+  Add missing deactivation after activation failure in lvcreate -Zy.
+  When activating, if precommitted metadata is still cached, assume it's live.
+  When removing LV symlinks, skip any where the VG name is not determined.
+  Drop metadata cache if update fails in vg_revert or vg_commit.
+  Avoid spurious duplicate VG messages referring to VGs that are gone.
+  Drop dev_name_confirmed error message to debug level.
+  Fix setpriority error message to signed int.
+  Temporarily disable dmeventd mirror monitoring during lvchange --resync.
+  Refactor some vginfo manipulation code.
+  Add assertions to trap deprecated P_ and V_ lock usage.
+  Add missing mutex around clvmd lvmcache_drop_metadata library call.
+  Fix uninitialised mutex in clvmd if all daemons are not running at startup.
+  Avoid using DLM locks with LCK_CACHE type P_ lock requests.
+  When asked to drop cached committed VG metadata, invalidate cached PV labels.
+  Drop metadata cache before writing precommitted metadata instead of after.
+  Don't touch /dev in vgrename if activation is disabled.
+
+Version 2.02.36 - 29th April 2008
+=================================
+  Fix fsadm.sh to work with older blockdev, blkid & readlink binaries.
+  Fix lvresize to pass new size to fsadm when extending device.
+  Remove unused struct in clvmd-openais, and use correct node count.
+  Fix nodes list in clvmd-openais, and allow for broadcast messages.
+  Exclude VG_GLOBAL from internal concurrent VG lock counter.
+  Fix vgsplit internal counting of snapshot LVs.
+  Fix vgmerge snapshot_count when source VG contains snapshots.
+  Simplify clvmd-openais by using non-async saLckResourceLock.
+  Fix internal LV counter when a snapshot is removed.
+  Fix metadata corruption writing lvm1-formatted metadata with snapshots.
+  Fix lvconvert -m0 allocatable space check.
+
+Version 2.02.35 - 15th April 2008
+=================================
+  Drop cached VG metadata before and after committing changes to it.
+  Rename P_global to P_#global.
+  Don't attempt remote metadata backups of non-clustered VGs. (2.02.29)
+  Don't store fid in VG metadata cache to avoid clvmd segfault. (2.02.34)
+  Update vgsplit tests to verify loosening of active LV restriction.
+  Update vgsplit to only restrict split with active LVs involved in split.
+  Add lv_is_active() to determine whether an lv is active.
+
+Version 2.02.34 - 10th April 2008
+=================================
+  Improve preferred_names lvm.conf example.
+  Fix vgdisplay 'Cur LV' field to match lvdisplay output.
+  Fix lv_count report field to exclude hidden LVs.
+  Add vg_is_clustered() helper function.
+  Fix vgsplit to only move hidden 'snapshotN' LVs when necessary.
+  Update vgsplit tests for lvnames on the cmdline.
+  Update vgsplit man page to reflect lvnames on the cmdline.
+  Update vgsplit to take "-n LogicalVolumeName" on the cmdline.
+  Use clustered mirror log with pvmove in clustered VGs, if available.
+  Fix some pvmove error status codes.
+  Fix vgsplit error paths to release vg_to lock.
+  Indicate whether or not VG is clustered in vgcreate log message.
+  Mention default --clustered setting in vgcreate man page.
+  Add config file overrides to clvmd when it reads the active LVs list.
+  Fix vgreduce to use vg_split_mdas to check sufficient mdas remain.
+  Add (empty) orphan VGs to lvmcache during initialisation.
+  Fix orphan VG name used for format_pool.
+  Create a fid for internal orphan VGs.
+  Update lvmcache VG lock state for all locking types now.
+  Fix output if overriding command_names on cmdline.
+  Add detection of clustered mirror log capability.
+  Add check to vg_commit() ensuring VG lock held before writing new VG metadata.
+  Add validation of LV name to pvmove -n.
+  Make clvmd refresh the context correctly when lvm.conf is updated.
+  Add some basic internal VG lock validation.
+  Add per-command flags to control which commands use the VG metadata cache.
+  Fix vgsplit locking of new VG (2.02.30).
+  Avoid erroneous vgsplit error message for new VG. (2.02.29)
+  Suppress duplicate message when lvresize fails because of invalid vgname.
+  Cache VG metadata internally while VG lock is held.
+  Fix redundant lvresize message if vg doesn't exist.
+  Fix another allocation bug with clvmd and large node IDs.
+  Add find_lv_in_lv_list() and find_pv_in_pv_list().
+  Fix uninitialised variable in clvmd that could cause odd hangs.
+  Add vgmerge tests.
+  Add pvseg_is_allocated() for identifying a PV segment allocated to a LV.
+  Add list_move() for moving elements from one list to another.
+  Add 'is_reserved_lvname()' for identifying hidden LVs.
+  Correct command name in lvmdiskscan man page.
+  clvmd no longer crashes if it sees nodeids over 50.
+  Fix potential deadlock in clvmd thread handling.
+  Refactor text format initialisation into _init_text_import.
+  Escape double quotes and backslashes in external metadata and config data.
+  Add functions for escaping double quotes in strings.
+  Rename count_chars_len to count_chars.
+  Use return_0 in a couple more places.
+  Correct a function name typo in _line_append error message.
+  Include limits.h in clvmd so it compiles with newer headers.
+  Add VirtIO disks (virtblk) to filters.
+  Fix resetting of MIRROR_IMAGE and VISIBLE_LV after removal of LV. (2.02.30)
+  Fix remove_layer_from_lv to empty the LV before removing it. (2.02.30)
+  Add missing no-longer-used segs_using_this_lv test to check_lv_segments.
+  Remove redundant non-NULL tests before calling free in clvmd.c.
+  Avoid a compiler warning: make is_orphan's parameter const.
+  Fix lvconvert detection of mirror conversion in progress. (2.02.30)
+  Avoid automatic lvconvert polldaemon invocation when -R specified. (2.02.30)
+  Fix 'pvs -a' to detect VGs of PVs without metadata areas.
+  Divide up internal orphan volume group by format type.
+  Update usage message for clvmd.
+  Fix clvmd man page not to print <br> and clarified debug options.
+  Fix lvresize to support /dev/mapper prefix in the LV name.
+  Fix unfilled parameter passed to fsadm from lvresize.
+  Update fsadm to call lvresize if the partition size differs (with option -l).
+  Fix fsadm to support VG/LV names.
+
+Version 2.02.33 - 31st January 2008
+===================================
+  Fix mirror log name construction during lvconvert. (2.02.30)
+  Make monitor_dev_for_events recurse through the stack of LVs.
+  Clean up some more compiler warnings.
+  Some whitespace tidy-ups.
+  Use stack return macros throughout.
+  Rely upon internally-cached PV labels while corresponding VG lock is held.
+
+Version 2.02.32 - 29th January 2008
+===================================
+  Fix two check_lv_segments error messages to show whole segment.
+  Refactor mirror log attachment code.
+  Fix internal metadata corruption in lvchange --resync. (2.02.30)
+  Fix new parameter validation in vgsplit and test mode. (2.02.30)
+  Remove redundant cnxman-socket.h file from clvmd directory.
+  Fix pvs, vgs, lvs error exit status on some error paths.
+
+Version 2.02.31 - 19th January 2008
+===================================
+  Fix lvcreate --nosync not to wait for non-happening sync. (2.02.30)
+  Add very_verbose lvconvert messages.
+  Avoid readahead error message with default setting of lvcreate -M1. (2.02.29)
+
+Version 2.02.30 - 17th January 2008
+===================================
+  Set default readahead to twice maximium stripe size.
+  Reinstate VG extent size and stripe size defaults (halved). (2.02.29)
+  Add lists of stacked LV segments using each LV to the internal metadata.
+  Change vgsplit -l (for unimplemented --list) into --maxlogicalvolumes.
+  Fix process_all_pvs to detect non-orphans with no MDAs correctly.
+  Don't use block_on_error with mirror targets version 1.12 and above.
+  Update vgsplit to accept vgcreate options when new VG is destination.
+  Update vgsplit to accept existing VG as destination.
+  lvconvert waits for completion of initial sync by default.
+  Refactor vgcreate for parameter validation and add tests.
+  Add new convert_lv field to lvs output.
+  Print warning when lvm tools are running as non-root.
+  Add snapshot dmeventd library (enables dmeventd snapshot monitoring).
+  Prevent pvcreate from overwriting MDA-less PVs belonging to active VGs.
+  Fix a segfault if using pvs with --all argument. (2.02.29)
+  Update --uuid argument description in man pages.
+  Fix vgreduce PV list processing not to process every PV in the VG. (2.02.29)
+  Extend lvconvert to use polldaemon.
+  Add support for stacked mirrors.
+  Major restructuring of pvmove and lvconvert layer manipulation code.
+  Replace tools/fsadm with scripts/fsadm.sh.
+  Append fields to report/pvsegs_cols_verbose.
+  Permit LV segment fields with PV segment reports.
+  Add seg_start_pe and seg_pe_ranges to reports.
+
+Version 2.02.29 - 5th December 2007
+===================================
   Make clvmd backup vg metadata on remote nodes.
   Refactor pvmove allocation code.
   Decode cluster locking state in log message.
@@ -36,14 +611,16 @@ Version 2.02.29 -
   Add const attributes to pv accessor functions.
   Refactor vg_add_snapshot() and lv_create_empty().
   Handle new sysfs subsystem/block/devices directory structure.
-  Tests are run with LVM_SYSTEM_DIR pointing to private root and /dev dirs.
+  Run test with LVM_SYSTEM_DIR pointing to private root and /dev dirs.
   Fix a bug in lvm_dump.sh checks for lvm/dmsetup binaries.
   Fix underquotations in lvm_dump.sh.
   Refactor lvcreate stripe and mirror parameter validation.
-  All tools: print --help output to stdout, not stderr.
-  After a diagnostic, suggest --help, rather than printing all --help output.
+  Print --help output to stdout, not stderr.
+  After a cmdline processing error, don't print help text but suggest --help.
   Add %PVS extents option to lvresize, lvextend, and lvcreate.
+  Add 'make check' to run tests in new subdirectory 'test'.
   Moved the obsolete test subdirectory to old-tests.
+  Cope with relative paths in configure --with-dmdir.
   Remove no-longer-correct restrictions on PV arg count with stripes/mirrors.
   Fix strdup memory leak in str_list_dup().
   Link with -lpthread when static SELinux libraries require that.

WHATS_NEW_DM

@@ -1,10 +1,104 @@
+Version 1.02.36 - 6th August 2009
+=================================
+  Add udevcookies, udevcomplete, udevcomplete_all and --noudevwait to dmsetup.
+  Add libdevmapper functions to support synchronisation with udev.
+
+Version 1.02.35 - 28th July 2009
+================================
+  Add LOG_LINE_WITH_ERRNO macro.
+  Use log_error macro consistently throughout in place of log_err.
+
+Version 1.02.34 - 15th July 2009
+================================
+  Use _exit() not exit() after forking to avoid flushing libc buffers twice.
+  Rename plog macro to LOG_LINE & add LOG_MESG variant for dm_dump_memory_debug.
+  Change plog to use dm_log_with_errno unless deprecated dm_log_init was used.
+  Add dm_log_with_errno and dm_log_with_errno_init, deprecating the old fns.
+  Fix whitespace in linear target line to fix identical table line detection.
+  Add device number to more log messages during activation.
+
+Version 1.02.33 - 30th June 2009
+================================
+  Don't fallback to default major number: use dm_task_set_major_minor. (1.02.31)
+  Do not fork daemon when dmeventd cannot be found.
+  Add crypt target handling to libdevmapper tree nodes.
+  Add splitname command to dmsetup.
+  Add subsystem, vg_name, lv_name, lv_layer fields to dmsetup reports.
+  Make mempool optional in dm_split_lvm_name().
+
+Version 1.02.32 - 21st May 2009
+===============================
+  Only generate libdevmapper.a when configured to link statically.
+  Export dm_tree_node_size_changed() from libdevmapper.
+  Propagate the table size_changed property up the dm device tree.
+  Detect failure to free memory pools when releasing the library.
+  Fix segfault when getopt processes dmsetup -U, -G and -M options.
+
+Version 1.02.31 - 3rd March 2009
+================================
+  If kernel supports only one dm major number, use in place of any supplied.
+
+Version 1.02.30 - 26th January 2009
+====================================
+  Add "all" field to reports expanding to all fields of report type.
+  Enforce device name length and character limitations in libdm.
+  Replace _dm_snprintf with EMIT_PARAMS macro for creating target lines.
+
+Version 1.02.29 - 10th November 2008
+====================================
+  Merge device-mapper into the LVM2 tree.
+  Split out dm-logging.h from log.h.
+  Use lvm-types.h.
+  Add usrsbindir to configure.
+
+Version 1.02.28 - 18th September 2008
+=====================================
+  Only resume devices in dm_tree_preload_children if size changes.
+  Extend deptree buffers so the largest possible device numbers fit.
+  Generate versioned libdevmapper-event.so.
+  Underline longer report help text headings.
+
+Version 1.02.27 - 25th June 2008
+================================
+  Align struct memblock in dbg_malloc for sparc.
+  Add --unquoted and --rows to dmsetup.
+  Avoid compiler warning about cast in dmsetup.c's OFFSET_OF macro.
+  Fix inverted no_flush debug message.
+  Remove --enable-jobs from configure. (Set at runtime instead.)
+  Bring configure.in and list.h into line with the lvm2 versions.
+
+Version 1.02.26 - 6th June 2008
+===============================
+  Initialise params buffer to empty string in _emit_segment.
+  Skip add_dev_node when ioctls disabled.
+  Make dm_hash_iter safe against deletion.
+  Accept a NULL pointer to dm_free silently.
+  Add tables_loaded, readonly and suspended columns to reports.
+  Add --nameprefixes to dmsetup.
+  Add field name prefix option to reporting functions.
+  Calculate string size within dm_pool_grow_object.
+
+Version 1.02.25 - 10th April 2008
+=================================
+  Remove redundant if-before-free tests.
+  Use log_warn for reporting field help text instead of log_print.
+  Change cluster mirror log type name (s/clustered_/clustered-/)
+
+Version 1.02.24 - 20th December 2007
+====================================
+  Fix deptree to pass new name to _resume_node after a rename.
+  Suppress other node operations if node is deleted.
+  Add node operation stack debug messages.
+  Report error when empty device name passed to readahead functions.
+  Fix minimum readahead debug message.
+
 Version 1.02.23 - 5th December 2007
 ===================================
   Update dm-ioctl.h after removal of compat code.
   Add readahead support to libdevmapper and dmsetup.
   Fix double free in a libdevmapper-event error path.
   Fix configure --with-dmeventd-path substitution.
-  Allow a DM_DEV_DIR environment variable to override /dev.
+  Allow a DM_DEV_DIR environment variable to override /dev in dmsetup.
   Create a libdevmapper.so.$LIB_VERSION symlink within the build tree.
   Avoid static link failure with some SELinux libraries that require libpthread.
   Remove obsolete dmfs code from tree and update INSTALL.
@@ -22,7 +116,7 @@ Version 1.02.22 - 21st August 2007
 Version 1.02.21 - 13th July 2007
 ================================
   Introduce _LOG_STDERR to send log_warn() messages to stderr not stdout.
-  Fix dmsetup -o devno string termination. (1.02.20) 
+  Fix dmsetup -o devno string termination. (1.02.20)
 
 Version 1.02.20 - 15th June 2007
 ================================
@@ -310,4 +404,3 @@ Version 1.00.08 - 27 Feb 2004
   Fixed DESTDIR for make install/install_static_lib.
   Updated README/INSTALL to reflect move to sources.redhat.com.
   Updated autoconf files to 2003-06-17.
-

autoconf/config.guess

@@ -1,10 +1,10 @@
 #! /bin/sh
 # Attempt to guess a canonical system name.
 #   Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002, 2003, 2004, 2005, 2006 Free Software Foundation,
-#   Inc.
+#   2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
+#   Free Software Foundation, Inc.
 
-timestamp='2006-07-02'
+timestamp='2008-01-23'
 
 # This file is free software; you can redistribute it and/or modify it
 # under the terms of the GNU General Public License as published by
@@ -56,8 +56,8 @@ version="\
 GNU config.guess ($timestamp)
 
 Originally written by Per Bothner.
-Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005
-Free Software Foundation, Inc.
+Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001,
+2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
 
 This is free software; see the source for copying conditions.  There is NO
 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
@@ -161,6 +161,7 @@ case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
 	    arm*) machine=arm-unknown ;;
 	    sh3el) machine=shl-unknown ;;
 	    sh3eb) machine=sh-unknown ;;
+	    sh5el) machine=sh5le-unknown ;;
 	    *) machine=${UNAME_MACHINE_ARCH}-unknown ;;
 	esac
 	# The Operating System including object format, if it has switched
@@ -329,7 +330,7 @@ case "${UNAME_MACHINE}:${UNAME_SYSTEM}:${UNAME_RELEASE}:${UNAME_VERSION}" in
     sun4*:SunOS:5.*:* | tadpole*:SunOS:5.*:*)
 	echo sparc-sun-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
 	exit ;;
-    i86pc:SunOS:5.*:*)
+    i86pc:SunOS:5.*:* | i86xen:SunOS:5.*:*)
 	echo i386-pc-solaris2`echo ${UNAME_RELEASE}|sed -e 's/[^.]*//'`
 	exit ;;
     sun4*:SunOS:6*:*)
@@ -531,7 +532,7 @@ EOF
 		echo rs6000-ibm-aix3.2
 	fi
 	exit ;;
-    *:AIX:*:[45])
+    *:AIX:*:[456])
 	IBM_CPU_ID=`/usr/sbin/lsdev -C -c processor -S available | sed 1q | awk '{ print $1 }'`
 	if /usr/sbin/lsattr -El ${IBM_CPU_ID} | grep ' POWER' >/dev/null 2>&1; then
 		IBM_ARCH=rs6000
@@ -780,7 +781,7 @@ EOF
     i*:CYGWIN*:*)
 	echo ${UNAME_MACHINE}-pc-cygwin
 	exit ;;
-    i*:MINGW*:*)
+    *:MINGW*:*)
 	echo ${UNAME_MACHINE}-pc-mingw32
 	exit ;;
     i*:windows32*:*)
@@ -790,12 +791,18 @@ EOF
     i*:PW*:*)
 	echo ${UNAME_MACHINE}-pc-pw32
 	exit ;;
-    x86:Interix*:[3456]*)
-	echo i586-pc-interix${UNAME_RELEASE}
-	exit ;;
-    EM64T:Interix*:[3456]*)
-	echo x86_64-unknown-interix${UNAME_RELEASE}
-	exit ;;
+    *:Interix*:[3456]*)
+    	case ${UNAME_MACHINE} in
+	    x86)
+		echo i586-pc-interix${UNAME_RELEASE}
+		exit ;;
+	    EM64T | authenticamd)
+		echo x86_64-unknown-interix${UNAME_RELEASE}
+		exit ;;
+	    IA64)
+		echo ia64-unknown-interix${UNAME_RELEASE}
+		exit ;;
+	esac ;;
     [345]86:Windows_95:* | [345]86:Windows_98:* | [345]86:Windows_NT:*)
 	echo i${UNAME_MACHINE}-pc-mks
 	exit ;;
@@ -829,7 +836,14 @@ EOF
 	echo ${UNAME_MACHINE}-pc-minix
 	exit ;;
     arm*:Linux:*:*)
-	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	eval $set_cc_for_build
+	if echo __ARM_EABI__ | $CC_FOR_BUILD -E - 2>/dev/null \
+	    | grep -q __ARM_EABI__
+	then
+	    echo ${UNAME_MACHINE}-unknown-linux-gnu
+	else
+	    echo ${UNAME_MACHINE}-unknown-linux-gnueabi
+	fi
 	exit ;;
     avr32*:Linux:*:*)
 	echo ${UNAME_MACHINE}-unknown-linux-gnu
@@ -950,6 +964,9 @@ EOF
     x86_64:Linux:*:*)
 	echo x86_64-unknown-linux-gnu
 	exit ;;
+    xtensa*:Linux:*:*)
+    	echo ${UNAME_MACHINE}-unknown-linux-gnu
+	exit ;;
     i*86:Linux:*:*)
 	# The BFD linker knows what the default object file format is, so
 	# first see if it will tell us. cd to the root directory to prevent
@@ -1208,6 +1225,15 @@ EOF
     SX-6:SUPER-UX:*:*)
 	echo sx6-nec-superux${UNAME_RELEASE}
 	exit ;;
+    SX-7:SUPER-UX:*:*)
+	echo sx7-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-8:SUPER-UX:*:*)
+	echo sx8-nec-superux${UNAME_RELEASE}
+	exit ;;
+    SX-8R:SUPER-UX:*:*)
+	echo sx8r-nec-superux${UNAME_RELEASE}
+	exit ;;
     Power*:Rhapsody:*:*)
 	echo powerpc-apple-rhapsody${UNAME_RELEASE}
 	exit ;;
@@ -1458,9 +1484,9 @@ This script, last modified $timestamp, has failed to recognize
 the operating system you are using. It is advised that you
 download the most up to date version of the config scripts from
 
-  http://savannah.gnu.org/cgi-bin/viewcvs/*checkout*/config/config/config.guess
+  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.guess;hb=HEAD
 and
-  http://savannah.gnu.org/cgi-bin/viewcvs/*checkout*/config/config/config.sub
+  http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob_plain;f=config.sub;hb=HEAD
 
 If the version you run ($0) is already up to date, please
 send the following data and any information you think might be

autoconf/config.sub

@@ -1,10 +1,10 @@
 #! /bin/sh
 # Configuration validation subroutine script.
 #   Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
-#   2000, 2001, 2002, 2003, 2004, 2005, 2006 Free Software Foundation,
-#   Inc.
+#   2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
+#   Free Software Foundation, Inc.
 
-timestamp='2006-09-20'
+timestamp='2008-01-16'
 
 # This file is (in principle) common to ALL GNU software.
 # The presence of a machine in this file suggests that SOME GNU software
@@ -72,8 +72,8 @@ Report bugs and patches to <config-patches@gnu.org>."
 version="\
 GNU config.sub ($timestamp)
 
-Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005
-Free Software Foundation, Inc.
+Copyright (C) 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001,
+2002, 2003, 2004, 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
 
 This is free software; see the source for copying conditions.  There is NO
 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE."
@@ -245,12 +245,12 @@ case $basic_machine in
 	| bfin \
 	| c4x | clipper \
 	| d10v | d30v | dlx | dsp16xx \
-	| fr30 | frv \
+	| fido | fr30 | frv \
 	| h8300 | h8500 | hppa | hppa1.[01] | hppa2.0 | hppa2.0[nw] | hppa64 \
 	| i370 | i860 | i960 | ia64 \
 	| ip2k | iq2000 \
 	| m32c | m32r | m32rle | m68000 | m68k | m88k \
-	| maxq | mb | microblaze | mcore \
+	| maxq | mb | microblaze | mcore | mep \
 	| mips | mipsbe | mipseb | mipsel | mipsle \
 	| mips16 \
 	| mips64 | mips64el \
@@ -324,7 +324,7 @@ case $basic_machine in
 	| clipper-* | craynv-* | cydra-* \
 	| d10v-* | d30v-* | dlx-* \
 	| elxsi-* \
-	| f30[01]-* | f700-* | fr30-* | frv-* | fx80-* \
+	| f30[01]-* | f700-* | fido-* | fr30-* | frv-* | fx80-* \
 	| h8300-* | h8500-* \
 	| hppa-* | hppa1.[01]-* | hppa2.0-* | hppa2.0[nw]-* | hppa64-* \
 	| i*86-* | i860-* | i960-* | ia64-* \
@@ -369,10 +369,14 @@ case $basic_machine in
 	| v850-* | v850e-* | vax-* \
 	| we32k-* \
 	| x86-* | x86_64-* | xc16x-* | xps100-* | xscale-* | xscalee[bl]-* \
-	| xstormy16-* | xtensa-* \
+	| xstormy16-* | xtensa*-* \
 	| ymp-* \
 	| z8k-*)
 		;;
+	# Recognize the basic CPU types without company name, with glob match.
+	xtensa*)
+		basic_machine=$basic_machine-unknown
+		;;
 	# Recognize the various machine names and aliases which stand
 	# for a CPU type and a company and sometimes even an OS.
 	386bsd)
@@ -443,6 +447,14 @@ case $basic_machine in
 		basic_machine=ns32k-sequent
 		os=-dynix
 		;;
+	blackfin)
+		basic_machine=bfin-unknown
+		os=-linux
+		;;
+	blackfin-*)
+		basic_machine=bfin-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
 	c90)
 		basic_machine=c90-cray
 		os=-unicos
@@ -475,8 +487,8 @@ case $basic_machine in
 		basic_machine=craynv-cray
 		os=-unicosmp
 		;;
-	cr16c)
-		basic_machine=cr16c-unknown
+	cr16)
+		basic_machine=cr16-unknown
 		os=-elf
 		;;
 	crds | unos)
@@ -668,6 +680,14 @@ case $basic_machine in
 		basic_machine=m68k-isi
 		os=-sysv
 		;;
+	m68knommu)
+		basic_machine=m68k-unknown
+		os=-linux
+		;;
+	m68knommu-*)
+		basic_machine=m68k-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
 	m88k-omron*)
 		basic_machine=m88k-omron
 		;;
@@ -683,6 +703,10 @@ case $basic_machine in
 		basic_machine=i386-pc
 		os=-mingw32
 		;;
+	mingw32ce)
+		basic_machine=arm-unknown
+		os=-mingw32ce
+		;;
 	miniframe)
 		basic_machine=m68000-convergent
 		;;
@@ -809,6 +833,14 @@ case $basic_machine in
 		basic_machine=i860-intel
 		os=-osf
 		;;
+	parisc)
+		basic_machine=hppa-unknown
+		os=-linux
+		;;
+	parisc-*)
+		basic_machine=hppa-`echo $basic_machine | sed 's/^[^-]*-//'`
+		os=-linux
+		;;
 	pbd)
 		basic_machine=sparc-tti
 		;;
@@ -925,6 +957,9 @@ case $basic_machine in
 		basic_machine=sh-hitachi
 		os=-hms
 		;;
+	sh5el)
+		basic_machine=sh5le-unknown
+		;;
 	sh64)
 		basic_machine=sh64-unknown
 		;;
@@ -1014,6 +1049,10 @@ case $basic_machine in
 		basic_machine=tic6x-unknown
 		os=-coff
 		;;
+	tile*)
+		basic_machine=tile-unknown
+		os=-linux-gnu
+		;;
 	tx39)
 		basic_machine=mipstx39-unknown
 		;;
@@ -1219,7 +1258,7 @@ case $os in
 	      | -os2* | -vos* | -palmos* | -uclinux* | -nucleus* \
 	      | -morphos* | -superux* | -rtmk* | -rtmk-nova* | -windiss* \
 	      | -powermax* | -dnix* | -nx6 | -nx7 | -sei* | -dragonfly* \
-	      | -skyos* | -haiku* | -rdos* | -toppers*)
+	      | -skyos* | -haiku* | -rdos* | -toppers* | -drops*)
 	# Remember, each alternative MUST END IN *, to match a version number.
 		;;
 	-qnx*)
@@ -1414,6 +1453,9 @@ case $basic_machine in
 	m68*-cisco)
 		os=-aout
 		;;
+        mep-*)
+		os=-elf
+		;;
 	mips*-cisco)
 		os=-elf
 		;;

configure.in

@@ -1,8 +1,6 @@
-##
+###############################################################################
 ## Copyright (C) 2000-2004 Sistina Software, Inc. All rights reserved.
-## Copyright (C) 2004, 2007 Red Hat, Inc. All rights reserved.
-##
-## This file is part of LVM2.
+## Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
 ##
 ## This copyrighted material is made available to anyone wishing to use,
 ## modify, copy, or redistribute it subject to the terms and conditions
@@ -18,9 +16,7 @@ AC_PREREQ(2.57)
 dnl -- Process this file with autoconf to produce a configure script.
 AC_INIT
 AC_CONFIG_SRCDIR([lib/device/dev-cache.h])
-
-################################################################################
-AC_CONFIG_HEADERS(lib/misc/configure.h)
+AC_CONFIG_HEADERS([lib/misc/configure.h])
 
 ################################################################################
 dnl -- Setup the directory where autoconf has auxilary files
@@ -39,50 +35,58 @@ case "$host_os" in
 		CLDNOWHOLEARCHIVE="-Wl,-no-whole-archive"
 		LDDEPS="$LDDEPS .export.sym"
 		LDFLAGS="$LDFLAGS -Wl,--export-dynamic"
-		LIB_SUFFIX="so"
+		LIB_SUFFIX=so
 		DEVMAPPER=yes
 		ODIRECT=yes
+		DM_IOCTLS=yes
 		SELINUX=yes
 		REALTIME=yes
 		CLUSTER=internal
-		FSADM=no ;;
+		FSADM=yes
+		;;
 	darwin*)
 		CFLAGS="$CFLAGS -no-cpp-precomp -fno-common"
 		COPTIMISE_FLAG="-O2"
 		CLDFLAGS="$CLDFLAGS"
 		CLDWHOLEARCHIVE="-all_load"
 		CLDNOWHOLEARCHIVE=
-		LDDEPS="$LDDEPS"
-		LDFLAGS="$LDFLAGS"
-		LIB_SUFFIX="dylib"
+		LIB_SUFFIX=dylib
 		DEVMAPPER=yes
 		ODIRECT=no
+		DM_IOCTLS=no
 		SELINUX=no
 		REALTIME=no
 		CLUSTER=none
-		FSADM=no ;;
+		FSADM=no
+		;;
 esac
 
 ################################################################################
 dnl -- Checks for programs.
+AC_PROG_SED
 AC_PROG_AWK
 AC_PROG_CC
+
+dnl probably no longer needed in 2008, but...
+AC_PROG_GCC_TRADITIONAL
 AC_PROG_INSTALL
 AC_PROG_LN_S
 AC_PROG_MAKE_SET
+AC_PROG_MKDIR_P
 AC_PROG_RANLIB
 AC_PATH_PROG(CFLOW_CMD, cflow)
 AC_PATH_PROG(CSCOPE_CMD, cscope)
 
 ################################################################################
-dnl -- Checks for header files.
+dnl -- Check for header files.
 AC_HEADER_DIRENT
 AC_HEADER_STDC
 AC_HEADER_SYS_WAIT
 AC_HEADER_TIME
 
-AC_CHECK_HEADERS(fcntl.h limits.h locale.h stddef.h syslog.h sys/file.h sys/ioctl.h sys/param.h sys/time.h,,AC_MSG_ERROR(bailing out))
-AC_CHECK_HEADERS(assert.h ctype.h libgen.h signal.h stdio.h sys/mman.h sys/resource.h sys/stat.h sys/types.h sys/utsname.h sys/wait.h time.h,,AC_MSG_ERROR(bailing out))
+AC_CHECK_HEADERS([locale.h stddef.h syslog.h sys/file.h sys/time.h assert.h \
+  libgen.h signal.h sys/mman.h sys/resource.h sys/utsname.h sys/wait.h time.h], ,
+  [AC_MSG_ERROR(bailing out)])
 
 case "$host_os" in
 	linux*)
@@ -91,12 +95,19 @@ case "$host_os" in
 		AC_CHECK_HEADERS(machine/endian.h sys/disk.h,,AC_MSG_ERROR(bailing out)) ;;
 esac
 
+AC_CHECK_HEADERS([ctype.h dirent.h errno.h fcntl.h getopt.h inttypes.h limits.h \
+  stdarg.h stdio.h stdlib.h string.h sys/ioctl.h sys/param.h sys/stat.h \
+  sys/types.h unistd.h], , [AC_MSG_ERROR(bailing out)])
+AC_CHECK_HEADERS(termios.h sys/statvfs.h)
+
 ################################################################################
-dnl -- Checks for typedefs, structures, and compiler characteristics.
+dnl -- Check for typedefs, structures, and compiler characteristics.
 AC_C_CONST
 AC_C_INLINE
+AC_CHECK_MEMBERS([struct stat.st_rdev])
 AC_TYPE_OFF_T
 AC_TYPE_PID_T
+AC_TYPE_SIGNAL
 AC_TYPE_SIZE_T
 AC_TYPE_MODE_T
 AC_CHECK_MEMBERS([struct stat.st_rdev])
@@ -104,7 +115,9 @@ AC_STRUCT_TM
 
 ################################################################################
 dnl -- Check for functions
-AC_CHECK_FUNCS(gethostname getpagesize memset munmap setlocale strcasecmp strchr strdup strncasecmp strerror strrchr strstr strtol strtoul,,AC_MSG_ERROR(bailing out))
+AC_CHECK_FUNCS([gethostname getpagesize memset mkdir rmdir munmap setlocale \
+  strcasecmp strchr strdup strncasecmp strerror strrchr strstr strtol strtoul \
+  uname], , [AC_MSG_ERROR(bailing out)])
 AC_FUNC_ALLOCA
 AC_FUNC_CLOSEDIR_VOID
 AC_FUNC_FORK
@@ -114,22 +127,28 @@ AC_FUNC_MEMCMP
 AC_FUNC_MMAP
 AC_FUNC_STAT
 AC_FUNC_STRTOD
+AC_FUNC_VPRINTF
+
+################################################################################
+dnl -- Enables statically-linked tools
+AC_MSG_CHECKING(whether to use static linking)
+AC_ARG_ENABLE(static_link,
+  [  --enable-static_link    Use this to link the tools to their libraries
+                          statically.  Default is dynamic linking],
+  STATIC_LINK=$enableval, STATIC_LINK=no)
+AC_MSG_RESULT($STATIC_LINK)
 
 ################################################################################
 dnl -- Prefix is /usr by default, the exec_prefix default is setup later
 AC_PREFIX_DEFAULT(/usr)
 
-################################################################################
-dnl -- Parallel make jobs?
-AC_ARG_ENABLE(jobs, [  --enable-jobs=NUM       Number of jobs to run simultaneously], JOBS=-j$enableval, JOBS=-j2)
-
 ################################################################################
 dnl -- Setup the ownership of the files
 AC_MSG_CHECKING(file owner)
 OWNER="root"
 
 AC_ARG_WITH(user,
-  [  --with-user=USER        Set the owner of installed files ],
+  [  --with-user=USER        Set the owner of installed files [[USER=root]] ],
   [ OWNER="$withval" ])
 AC_MSG_RESULT($OWNER)
 
@@ -142,7 +161,7 @@ dnl -- Setup the group ownership of the files
 AC_MSG_CHECKING(group owner)
 GROUP="root"
 AC_ARG_WITH(group,
-  [  --with-group=GROUP      Set the group owner of installed files ],
+  [  --with-group=GROUP      Set the group owner of installed files [[GROUP=root]] ],
   [ GROUP="$withval" ])
 AC_MSG_RESULT($GROUP)
 
@@ -150,6 +169,33 @@ if test x$GROUP != x; then
 	GROUP="-g $GROUP"
 fi
 
+################################################################################
+dnl -- Setup device node ownership
+AC_MSG_CHECKING(device node uid)
+
+AC_ARG_WITH(device-uid,
+  [  --with-device-uid=UID   Set the owner used for new device nodes [[UID=0]] ],
+  [ DM_DEVICE_UID="$withval" ], [ DM_DEVICE_UID="0" ] )
+AC_MSG_RESULT($DM_DEVICE_UID)
+
+################################################################################
+dnl -- Setup device group ownership
+AC_MSG_CHECKING(device node gid)
+
+AC_ARG_WITH(device-gid,
+  [  --with-device-gid=UID   Set the group used for new device nodes [[GID=0]] ],
+  [ DM_DEVICE_GID="$withval" ], [ DM_DEVICE_GID="0" ] )
+AC_MSG_RESULT($DM_DEVICE_GID)
+
+################################################################################
+dnl -- Setup device mode
+AC_MSG_CHECKING(device node mode)
+
+AC_ARG_WITH(device-mode,
+  [  --with-device-mode=MODE Set the mode used for new device nodes [[MODE=0600]] ],
+  [ DM_DEVICE_MODE="$withval" ], [ DM_DEVICE_MODE="0600" ] )
+AC_MSG_RESULT($DM_DEVICE_MODE)
+
 ################################################################################
 dnl -- LVM1 tool fallback option
 AC_MSG_CHECKING(whether to enable lvm1 fallback)
@@ -261,30 +307,13 @@ if test x$MIRRORS = xinternal; then
 fi
 
 ################################################################################
-dnl -- Enables staticly-linked tools
-AC_MSG_CHECKING(whether to use static linking)
-AC_ARG_ENABLE(static_link, [  --enable-static_link    Use this to link the tools to their libraries
-                          statically.  Default is dynamic linking],  STATIC_LINK=$enableval, STATIC_LINK=no)
-AC_MSG_RESULT($STATIC_LINK)
-
-################################################################################
-dnl -- Enable readline
+dnl -- Disable readline
 AC_MSG_CHECKING(whether to enable readline)
-AC_ARG_ENABLE(readline, [  --enable-readline       Enable readline support],
-READLINE=$enableval, READLINE=no)
+AC_ARG_ENABLE([readline],
+  [  --disable-readline      Disable readline support],
+  [READLINE=$enableval], [READLINE=maybe])
 AC_MSG_RESULT($READLINE)
 
-if test x$READLINE = xyes; then
-	AC_DEFINE([READLINE_SUPPORT], 1, [Define to 1 to include the LVM readline shell.])
-fi
-
-################################################################################
-dnl -- Disable selinux
-AC_MSG_CHECKING(whether to enable selinux support)
-AC_ARG_ENABLE(selinux, [  --disable-selinux       Disable selinux support],
-SELINUX=$enableval)
-AC_MSG_RESULT($SELINUX)
-
 ################################################################################
 dnl -- Disable realtime clock support
 AC_MSG_CHECKING(whether to enable realtime support)
@@ -296,7 +325,7 @@ AC_MSG_RESULT($REALTIME)
 dnl -- Build cluster LVM daemon
 AC_MSG_CHECKING(whether to build cluster LVM daemon)
 AC_ARG_WITH(clvmd,
-  [  --with-clvmd=TYPE       Build cluster LVM Daemon: cman/gulm/none/all
+  [  --with-clvmd=TYPE       Build cluster LVM Daemon: cman/gulm/corosync/none/all
                           [TYPE=none] ],
   [ CLVMD="$withval" ],
   [ CLVMD="none" ])
@@ -310,11 +339,24 @@ if  test x$CLVMD != xnone && test x$CLUSTER = xnone; then
 	CLUSTER=internal
 fi
 
+dnl -- Look for corosync libraries if required.
+if [[ "x$CLVMD" = xall -o `expr x"$CLVMD" : '.*corosync.*'` != 0 ]]; then
+	PKG_CHECK_MODULES(QUORUM, libquorum, [],
+		[AC_MSG_RESULT([no pkg for quorum library, using -lquorum]);
+		QUORUM_LIBS="-lquorum"])
+	PKG_CHECK_MODULES(CONFDB, libconfdb, [],
+		[AC_MSG_RESULT([no pkg for confdb library, using -lconfdb]);
+		CONFDB_LIBS="-lconfdb"])
+	PKG_CHECK_MODULES(CPG, libcpg, [],
+		[AC_MSG_RESULT([no pkg for libcpg library, using -lcpg]);
+		CPG_LIBS="-lcpg"])
+fi
+
 ################################################################################
 dnl -- Enable debugging
 AC_MSG_CHECKING(whether to enable debugging)
 AC_ARG_ENABLE(debug,    [  --enable-debug          Enable debugging],
-DEBUG=$enableval, DEBUG=no)
+  DEBUG=$enableval, DEBUG=no)
 AC_MSG_RESULT($DEBUG)
 
 dnl -- Normally turn off optimisation for debug builds
@@ -328,21 +370,65 @@ fi
 dnl -- Override optimisation
 AC_MSG_CHECKING(for C optimisation flag)
 AC_ARG_WITH(optimisation,
-  [  --with-optimisation=OPT C optimisation flag [OPT=-O2] ],
+  [  --with-optimisation=OPT C optimisation flag [[OPT=-O2]] ],
   [ COPTIMISE_FLAG="$withval" ])
 AC_MSG_RESULT($COPTIMISE_FLAG)
 
+################################################################################
+dnl -- Enable profiling
+AC_MSG_CHECKING(whether to gather gcov profiling data)
+AC_ARG_ENABLE(profiling,
+  AC_HELP_STRING(--enable-profiling, [Gather gcov profiling data]),
+  PROFILING=$enableval, PROFILING=no)
+AC_MSG_RESULT($PROFILING)
+
+if test "x$PROFILING" = xyes; then
+  COPTIMISE_FLAG="$COPTIMISE_FLAG -fprofile-arcs -ftest-coverage"
+  AC_PATH_PROG(LCOV, lcov, no)
+  AC_PATH_PROG(GENHTML, genhtml, no)
+  if test "$LCOV" = no -o "$GENHTML" = no ; then
+    AC_MSG_ERROR([lcov and genhtml are required for profiling])
+  fi
+fi
+
 ################################################################################
 dnl -- Disable devmapper
 AC_MSG_CHECKING(whether to use device-mapper)
-AC_ARG_ENABLE(devmapper, [  --disable-devmapper     Disable device-mapper interaction],
+AC_ARG_ENABLE(devmapper, [  --disable-devmapper     Disable LVM2 device-mapper interaction],
 DEVMAPPER=$enableval)
 AC_MSG_RESULT($DEVMAPPER)
 
 if test x$DEVMAPPER = xyes; then
-	AC_DEFINE([DEVMAPPER_SUPPORT], 1, [Define to 1 to enable device-mapper interaction.])
+	AC_DEFINE([DEVMAPPER_SUPPORT], 1, [Define to 1 to enable LVM2 device-mapper interaction.])
 fi
 
+################################################################################
+dnl -- Enable udev synchronisation
+AC_MSG_CHECKING(whether to enable synchronisation with udev processing)
+AC_ARG_ENABLE(udev_sync, [  --enable-udev_sync      Enable synchronisation with udev processing],
+UDEV_SYNC=$enableval, UDEV_SYNC=no)
+AC_MSG_RESULT($UDEV_SYNC)
+
+if test x$UDEV_SYNC = xyes; then
+	AC_DEFINE([UDEV_SYNC_SUPPORT], 1, [Define to 1 to enable synchronisation with udev processing.])
+fi
+
+dnl -- Enable udev rules
+AC_MSG_CHECKING(whether to enable installation of udev rules required for synchronisation)
+AC_ARG_ENABLE(udev_rules, [  --enable-udev_rules     Install rule files needed for udev synchronisation],
+UDEV_RULES=$enableval, UDEV_RULES=$UDEV_SYNC)
+AC_MSG_RESULT($UDEV_RULES)
+
+################################################################################
+dnl -- Compatibility mode
+AC_ARG_ENABLE(compat,   [  --enable-compat         Enable support for old device-mapper versions],
+  DM_COMPAT=$enableval, DM_COMPAT=no)
+
+################################################################################
+dnl -- Disable ioctl
+AC_ARG_ENABLE(ioctl,   [  --disable-driver        Disable calls to device-mapper in the kernel],
+  DM_IOCTLS=$enableval)
+
 ################################################################################
 dnl -- Disable O_DIRECT
 AC_MSG_CHECKING(whether to enable O_DIRECT)
@@ -354,16 +440,37 @@ if test x$ODIRECT = xyes; then
 	AC_DEFINE([O_DIRECT_SUPPORT], 1, [Define to 1 to enable O_DIRECT support.])
 fi
 
+################################################################################
+dnl -- Enable liblvm2app.so
+AC_MSG_CHECKING(whether to build liblvm2app.so application library)
+AC_ARG_ENABLE(applib,
+  [  --enable-applib         Build application library],
+  APPLIB=$enableval, APPLIB=no)
+AC_MSG_RESULT($APPLIB)
+AC_SUBST([LVM2APP_LIB])
+test x$APPLIB = xyes \
+  && LVM2APP_LIB=-llvm2app \
+  || LVM2APP_LIB=
+
 ################################################################################
 dnl -- Enable cmdlib
 AC_MSG_CHECKING(whether to compile liblvm2cmd.so)
 AC_ARG_ENABLE(cmdlib, [  --enable-cmdlib         Build shared command library],
 CMDLIB=$enableval, CMDLIB=no)
 AC_MSG_RESULT($CMDLIB)
+AC_SUBST([LVM2CMD_LIB])
+test x$CMDLIB = xyes \
+  && LVM2CMD_LIB=-llvm2cmd \
+  || LVM2CMD_LIB=
+
+################################################################################
+dnl -- Enable pkg-config
+AC_ARG_ENABLE(pkgconfig,   [  --enable-pkgconfig      Install pkgconfig support],
+  PKGCONFIG=$enableval, PKGCONFIG=no)
 
 ################################################################################
 dnl -- Enable fsadm
-AC_MSG_CHECKING(whether to build fsadm)
+AC_MSG_CHECKING(whether to install fsadm)
 AC_ARG_ENABLE(fsadm, [  --enable-fsadm          Enable fsadm],
 FSADM=$enableval)
 AC_MSG_RESULT($FSADM)
@@ -375,34 +482,52 @@ AC_ARG_ENABLE(dmeventd, [  --enable-dmeventd       Enable the device-mapper even
 DMEVENTD=$enableval)
 AC_MSG_RESULT($DMEVENTD)
 
+BUILD_DMEVENTD=$DMEVENTD
+
 dnl -- dmeventd currently requires internal mirror support
-if test x$DMEVENTD = xyes && test x$MIRRORS != xinternal; then
-AC_MSG_ERROR(
---enable-dmeventd currently requires --with-mirrors=internal
-)
+if test x$DMEVENTD = xyes; then
+   if test x$MIRRORS != xinternal; then
+      AC_MSG_ERROR(
+	--enable-dmeventd currently requires --with-mirrors=internal
+      )
+   fi
+   if test x$CMDLIB = xno; then
+      AC_MSG_ERROR(
+        --enable-dmeventd requires --enable-cmdlib to be used as well
+      )
+   fi
 fi
 
 if test x$DMEVENTD = xyes; then
 	AC_DEFINE([DMEVENTD], 1, [Define to 1 to enable the device-mapper event daemon.])
 fi
+
 ################################################################################
-dnl -- Mess with default exec_prefix
+dnl -- getline included in recent libc
+
+AC_CHECK_LIB(c, getline, AC_DEFINE([HAVE_GETLINE], 1,
+  [Define to 1 if getline is available.]))
+
+################################################################################
+dnl -- canonicalize_file_name included in recent libc
+
+AC_CHECK_LIB(c, canonicalize_file_name,
+  AC_DEFINE([HAVE_CANONICALIZE_FILE_NAME], 1,
+    [Define to 1 if canonicalize_file_name is available.]))
+
+################################################################################
+dnl -- Clear default exec_prefix - install into /sbin rather than /usr/sbin
 if [[ "x$exec_prefix" = xNONE -a "x$prefix" = xNONE ]];
  then  exec_prefix="";
 fi;
 
-################################################################################
-dnl -- Checks for library functions.
-AC_PROG_GCC_TRADITIONAL
-AC_TYPE_SIGNAL
-AC_FUNC_VPRINTF
-AC_CHECK_FUNCS(mkdir rmdir uname,,AC_MSG_ERROR(bailing out))
-
 ################################################################################
 dnl -- Check for termcap (Shamelessly copied from parted 1.4.17)
-if test x$READLINE = xyes; then
-	AC_SEARCH_LIBS(tgetent, ncurses curses termcap termlib, ,
-		AC_MSG_ERROR(
+if test x$READLINE != xno; then
+	AC_SEARCH_LIBS([tgetent], [tinfo ncurses curses termcap termlib],
+	  [tg_found=yes], [tg_found=no])
+	test x$READLINE:$tg_found = xyes:no &&
+	  AC_MSG_ERROR(
 termcap could not be found which is required for the
 --enable-readline option (which is enabled by default).  Either disable readline
 support with --disable-readline or download and install termcap from:
@@ -412,7 +537,6 @@ Note: if you are using precompiled packages you will also need the development
 Note: (n)curses also seems to work as a substitute for termcap.  This was
   not found either - but you could try installing that as well.
 )
-	)
 fi
 
 ################################################################################
@@ -436,12 +560,21 @@ Features cannot be 'shared' when building statically
 )
 fi
 
+################################################################################
+dnl -- Disable selinux
+AC_MSG_CHECKING(whether to enable selinux support)
+AC_ARG_ENABLE(selinux, [  --disable-selinux       Disable selinux support],
+  SELINUX=$enableval)
+AC_MSG_RESULT($SELINUX)
+
 ################################################################################
 dnl -- Check for selinux
 if test x$SELINUX = xyes; then
 	AC_CHECK_LIB(sepol, sepol_check_context, HAVE_SEPOL=yes, HAVE_SEPOL=no)
 
 	if test x$HAVE_SEPOL = xyes; then
+		AC_DEFINE([HAVE_SEPOL], 1,
+		  [Define to 1 if sepol_check_context is available.])
 		LIBS="-lsepol $LIBS"
 	fi
 
@@ -454,7 +587,7 @@ if test x$SELINUX = xyes; then
 		AC_MSG_WARN(Disabling selinux)
 	fi
 
-	# With --enable-static_link and selinux enabled, linking lvm.static
+	# With --enable-static_link and selinux enabled, linking
 	# fails on at least Debian unstable due to unsatisfied references
 	# to pthread_mutex_lock and _unlock.  See if we need -lpthread.
 	if test "$STATIC_LINK-$HAVE_SELINUX" = yes-yes; then
@@ -482,13 +615,15 @@ fi
 
 ################################################################################
 dnl -- Check for getopt
-AC_CHECK_HEADERS(getopt.h, AC_DEFINE([HAVE_GETOPTLONG], 1, [Define to 1 to if getopt_long is available.]))
+AC_CHECK_HEADERS(getopt.h, AC_DEFINE([HAVE_GETOPTLONG], 1, [Define to 1 if getopt_long is available.]))
 
 ################################################################################
 dnl -- Check for readline (Shamelessly copied from parted 1.4.17)
-if test x$READLINE = xyes; then
-	AC_CHECK_LIB(readline, readline, ,
-		AC_MSG_ERROR(
+if test x$READLINE != xno; then
+	rl_found=yes
+	AC_CHECK_LIB([readline], [readline], , [rl_found=no])
+	test x$READLINE:$rl_found = xyes:no &&
+	  AC_MSG_ERROR(
 GNU Readline could not be found which is required for the
 --enable-readline option (which is enabled by default).  Either disable readline
 support with --disable-readline or download and install readline from:
@@ -496,8 +631,11 @@ support with --disable-readline or download and install readline from:
 Note: if you are using precompiled packages you will also need the development
 package as well (which may be called readline-devel or something similar).
 )
-	)
-	AC_CHECK_FUNC(rl_completion_matches, AC_DEFINE([HAVE_RL_COMPLETION_MATCHES], 1, [Define to 1 if rl_completion_matches() is available.]))
+	if test $rl_found = yes; then
+		AC_CHECK_FUNCS([rl_completion_matches])
+		AC_DEFINE([READLINE_SUPPORT], 1,
+		  [Define to 1 to include the LVM readline shell.])
+	fi
 fi
 
 ################################################################################
@@ -508,6 +646,7 @@ AC_ARG_ENABLE(nls, [  --enable-nls            Enable Native Language Support],
 AC_MSG_RESULT($INTL)
 
 if test x$INTL = xyes; then
+# FIXME - Move this - can be device-mapper too
 	INTL_PACKAGE="lvm2"
 	AC_PATH_PROG(MSGFMT, msgfmt)
 	if [[ "x$MSGFMT" == x ]];
@@ -517,32 +656,42 @@ if test x$INTL = xyes; then
 	fi;
 
 	AC_ARG_WITH(localedir,
-  		    [  --with-localedir=DIR    Translation files in DIR [PREFIX/share/locale]],
-  		    [ LOCALEDIR="$withval" ],
-  		    [ LOCALEDIR='${prefix}/share/locale' ])
+		    [  --with-localedir=DIR    Translation files in DIR [[PREFIX/share/locale]] ],
+		    [ LOCALEDIR="$withval" ],
+		    [ LOCALEDIR='${prefix}/share/locale' ])
 fi
 
 ################################################################################
 AC_ARG_WITH(confdir,
-	    [  --with-confdir=DIR      Configuration files in DIR [/etc]],
-  	    [ CONFDIR="$withval" ],
- 	    [ CONFDIR='/etc' ])
+	    [  --with-confdir=DIR      Configuration files in DIR [[/etc]]],
+	    [ CONFDIR="$withval" ],
+	    [ CONFDIR='/etc' ])
 
 AC_ARG_WITH(staticdir,
-	    [  --with-staticdir=DIR    Static binary in DIR [EXEC_PREFIX/sbin]],
-  	    [ STATICDIR="$withval" ],
- 	    [ STATICDIR='${exec_prefix}/sbin' ])
+	    [  --with-staticdir=DIR    Static binary in DIR [[EPREFIX/sbin]]],
+	    [ STATICDIR="$withval" ],
+	    [ STATICDIR='${exec_prefix}/sbin' ])
 
-AC_ARG_WITH(dmdir,
-	    [  --with-dmdir=DIR        Build against device-mapper source tree in DIR],
-  	    [ DMDIR="$withval" CPPFLAGS="$CPPFLAGS -I$DMDIR/include"],
-	    [ DMDIR= ])
+AC_ARG_WITH(usrlibdir,
+	    [  --with-usrlibdir=DIR],
+	    [ usrlibdir="$withval"],
+	    [ usrlibdir='${prefix}/lib' ])
 
-# Convert a relative dir name to absolute.
-case $DMDIR in
-  /*) ;;
-  *) DMDIR="`pwd`/$DMDIR" ;;
-esac
+AC_ARG_WITH(usrsbindir,
+	    [  --with-usrsbindir=DIR],
+	    [ usrsbindir="$withval"],
+	    [ usrsbindir='${prefix}/sbin' ])
+
+################################################################################
+AC_ARG_WITH(udev_prefix,
+	    [  --with-udev-prefix=UPREFIX      Install udev rule files in UPREFIX [[EPREFIX]]],
+            [ udev_prefix="$withval"],
+            [ udev_prefix='${exec_prefix}' ])
+
+AC_ARG_WITH(udevdir,
+	    [  --with-udevdir=DIR      udev rules in DIR [[UPREFIX/lib/udev/rules.d]]],
+	    [ udevdir="$withval"],
+	    [ udevdir='${udev_prefix}/lib/udev/rules.d' ])
 
 ################################################################################
 dnl -- Ensure additional headers required
@@ -558,11 +707,6 @@ if test x$CLVMD != xnone; then
 	AC_FUNC_SELECT_ARGTYPES
 fi
 
-if test x$FSADM = xyes; then
-	AC_CHECK_HEADERS(fstab.h sys/mount.h sys/vfs.h,,AC_MSG_ERROR(bailing out))
-	AC_CHECK_FUNCS(memmove,,AC_MSG_ERROR(bailing out))
-fi
-
 if test x$CLUSTER != xnone; then
 	AC_CHECK_HEADERS(sys/socket.h sys/un.h,,AC_MSG_ERROR(bailing out))
 	AC_CHECK_FUNCS(socket,,AC_MSG_ERROR(bailing out))
@@ -576,12 +720,14 @@ if test x$INTL = xyes; then
 	AC_CHECK_HEADERS(libintl.h,,AC_MSG_ERROR(bailing out))
 fi
 
-AC_CHECK_HEADERS(libdevmapper.h,,AC_MSG_ERROR(bailing out))
-
 if test x$HAVE_SELINUX = xyes; then
 	AC_CHECK_HEADERS(selinux/selinux.h,,AC_MSG_ERROR(bailing out))
 fi
 
+if test x$UDEV_SYNC = xyes; then
+	AC_CHECK_HEADERS(sys/ipc.h sys/sem.h,,AC_MSG_ERROR(bailing out))
+fi
+
 ################################################################################
 AC_PATH_PROG(MODPROBE_CMD, modprobe)
 
@@ -590,52 +736,131 @@ if test x$MODPROBE_CMD != x; then
 fi
 
 ################################################################################
-if test "-f VERSION"; then
-  LVM_VERSION="\"`cat VERSION`\""
-else
-  LVM_VERSION="Unknown"
+dnl -- dmeventd pidfile and executable path
+AH_TEMPLATE(DMEVENTD_PIDFILE, [Path to dmeventd pidfile.])
+if test "$BUILD_DMEVENTD" = yes; then
+	AC_ARG_WITH(dmeventd-pidfile,
+		    [  --with-dmeventd-pidfile=PATH    dmeventd pidfile [[/var/run/dmeventd.pid]] ],
+		    [ AC_DEFINE_UNQUOTED(DMEVENTD_PIDFILE,"$withval") ],
+		    [ AC_DEFINE_UNQUOTED(DMEVENTD_PIDFILE,"/var/run/dmeventd.pid") ])
+fi
+
+AH_TEMPLATE(DMEVENTD_PATH, [Path to dmeventd binary.])
+if test "$BUILD_DMEVENTD" = yes; then
+        dmeventd_prefix="$exec_prefix"
+        if test "x$dmeventd_prefix" = "xNONE"; then
+                dmeventd_prefix="$prefix"
+        fi
+        if test "x$dmeventd_prefix" = "xNONE"; then
+                dmeventd_prefix=""
+        fi
+	AC_ARG_WITH(dmeventd-path,
+		    [  --with-dmeventd-path=PATH       dmeventd path [[EPREFIX/sbin/dmeventd]] ],
+		    [ AC_DEFINE_UNQUOTED(DMEVENTD_PATH,"$withval") ],
+		    [ AC_DEFINE_UNQUOTED(DMEVENTD_PATH,"$dmeventd_prefix/sbin/dmeventd") ])
 fi
 
 ################################################################################
-AC_SUBST(JOBS)
-AC_SUBST(STATIC_LINK)
-AC_SUBST(LVM1)
-AC_SUBST(POOL)
-AC_SUBST(SNAPSHOTS)
-AC_SUBST(MIRRORS)
-AC_SUBST(OWNER)
-AC_SUBST(GROUP)
+dnl -- which kernel interface to use (ioctl only)
+AC_MSG_CHECKING(for kernel interface choice)
+AC_ARG_WITH(interface,
+  [  --with-interface=IFACE  Choose kernel interface (ioctl) [[ioctl]] ],
+  [ interface="$withval" ],
+  [ interface=ioctl ])
+if [[ "x$interface" != xioctl ]];
+then
+  AC_MSG_ERROR(--with-interface=ioctl required. fs no longer supported.)
+fi
+AC_MSG_RESULT($interface)
+
+################################################################################
+DM_LIB_VERSION="\"`cat VERSION_DM 2>/dev/null || echo Unknown`\""
+AC_DEFINE_UNQUOTED(DM_LIB_VERSION, $DM_LIB_VERSION, [Library version])
+
+DM_LIB_PATCHLEVEL=`cat VERSION_DM | $AWK -F '[[-. ]]' '{printf "%s.%s.%s",$1,$2,$3}'`
+
+LVM_VERSION="\"`cat VERSION 2>/dev/null || echo Unknown`\""
+
+VER=`cat VERSION`
+LVM_RELEASE_DATE="\"`echo $VER | $SED 's/.* (//;s/).*//'`\""
+VER=`echo "$VER" | $AWK '{print $1}'`
+LVM_RELEASE="\"`echo "$VER" | $AWK -F '-' '{print $2}'`\""
+VER=`echo "$VER" | $AWK -F '-' '{print $1}'`
+LVM_MAJOR=`echo "$VER" | $AWK -F '.' '{print $1}'`
+LVM_MINOR=`echo "$VER" | $AWK -F '.' '{print $2}'`
+LVM_PATCHLEVEL=`echo "$VER" | $AWK -F '[[(.]]' '{print $3}'`
+LVM_LIBAPI=`echo "$VER" | $AWK -F '[[()]]' '{print $2}'`
+
+################################################################################
+AC_SUBST(APPLIB)
+AC_SUBST(BUILD_DMEVENTD)
 AC_SUBST(CFLAGS)
-AC_SUBST(COPTIMISE_FLAG)
+AC_SUBST(CFLOW_CMD)
 AC_SUBST(CLDFLAGS)
-AC_SUBST(CLDWHOLEARCHIVE)
 AC_SUBST(CLDNOWHOLEARCHIVE)
-AC_SUBST(LDDEPS)
-AC_SUBST(LDFLAGS)
-AC_SUBST(LIB_SUFFIX)
-AC_SUBST(LIBS)
-AC_SUBST(LVM_VERSION)
-AC_SUBST(LVM1_FALLBACK)
+AC_SUBST(CLDWHOLEARCHIVE)
+AC_SUBST(CLUSTER)
+AC_SUBST(CLVMD)
+AC_SUBST(CMDLIB)
+AC_SUBST(CONFDB_CFLAGS)
+AC_SUBST(CONFDB_LIBS)
+AC_SUBST(CONFDIR)
+AC_SUBST(COPTIMISE_FLAG)
+AC_SUBST(CPG_CFLAGS)
+AC_SUBST(CPG_LIBS)
+AC_SUBST(CSCOPE_CMD)
 AC_SUBST(DEBUG)
 AC_SUBST(DEVMAPPER)
-AC_SUBST(HAVE_LIBDL)
-AC_SUBST(HAVE_SELINUX)
-AC_SUBST(HAVE_REALTIME)
-AC_SUBST(CMDLIB)
-AC_SUBST(MSGFMT)
-AC_SUBST(LOCALEDIR)
-AC_SUBST(CONFDIR)
-AC_SUBST(STATICDIR)
-AC_SUBST(DMDIR)
-AC_SUBST(INTL_PACKAGE)
-AC_SUBST(INTL)
-AC_SUBST(CLVMD)
-AC_SUBST(CLUSTER)
-AC_SUBST(FSADM)
 AC_SUBST(DMEVENTD)
-AC_SUBST(CFLOW_CMD)
-AC_SUBST(CSCOPE_CMD)
+AC_SUBST(DM_COMPAT)
+AC_SUBST(DM_DEVICE_GID)
+AC_SUBST(DM_DEVICE_MODE)
+AC_SUBST(DM_DEVICE_UID)
+AC_SUBST(DM_IOCTLS)
+AC_SUBST(DM_LIB_VERSION)
+AC_SUBST(DM_LIB_PATCHLEVEL)
+AC_SUBST(FSADM)
+AC_SUBST(GROUP)
+AC_SUBST(HAVE_LIBDL)
+AC_SUBST(HAVE_REALTIME)
+AC_SUBST(HAVE_SELINUX)
+AC_SUBST(INTL)
+AC_SUBST(INTL_PACKAGE)
+AC_SUBST(JOBS)
+AC_SUBST(LDDEPS)
+AC_SUBST(LIBS)
+AC_SUBST(LIB_SUFFIX)
+AC_SUBST(LOCALEDIR)
+AC_SUBST(LVM1)
+AC_SUBST(LVM1_FALLBACK)
+AC_SUBST(LVM_VERSION)
+AC_SUBST(LVM_LIBAPI)
+AC_SUBST(LVM_MAJOR)
+AC_SUBST(LVM_MINOR)
+AC_SUBST(LVM_PATCHLEVEL)
+AC_SUBST(LVM_RELEASE)
+AC_SUBST(LVM_RELEASE_DATE)
+AC_SUBST(MIRRORS)
+AC_SUBST(MSGFMT)
+AC_SUBST(OWNER)
+AC_SUBST(PKGCONFIG)
+AC_SUBST(POOL)
+AC_SUBST(QUORUM_CFLAGS)
+AC_SUBST(QUORUM_LIBS)
+AC_SUBST(SNAPSHOTS)
+AC_SUBST(STATICDIR)
+AC_SUBST(STATIC_LINK)
+AC_SUBST(UDEV_RULES)
 AC_SUBST([LIB_PTHREAD])
+AC_SUBST(interface)
+AC_SUBST(kerneldir)
+AC_SUBST(missingkernel)
+AC_SUBST(kernelvsn)
+AC_SUBST(tmpdir)
+AC_SUBST(udev_prefix)
+AC_SUBST(udevdir)
+AC_SUBST(usrlibdir)
+AC_SUBST(usrsbindir)
 
 ################################################################################
 dnl -- First and last lines should not contain files to generate in order to
@@ -645,8 +870,11 @@ Makefile
 make.tmpl
 daemons/Makefile
 daemons/clvmd/Makefile
-dmeventd/Makefile
-dmeventd/mirror/Makefile
+daemons/dmeventd/Makefile
+daemons/dmeventd/libdevmapper-event.pc
+daemons/dmeventd/plugins/Makefile
+daemons/dmeventd/plugins/mirror/Makefile
+daemons/dmeventd/plugins/snapshot/Makefile
 doc/Makefile
 include/Makefile
 lib/Makefile
@@ -654,25 +882,23 @@ lib/format1/Makefile
 lib/format_pool/Makefile
 lib/locking/Makefile
 lib/mirror/Makefile
+lib/misc/lvm-version.h
 lib/snapshot/Makefile
-test/Makefile
+libdm/Makefile
+libdm/libdevmapper.pc
+liblvm/Makefile
+liblvm/liblvm2app.pc
 man/Makefile
 po/Makefile
+scripts/clvmd_init_red_hat
 scripts/Makefile
+test/Makefile
+test/api/Makefile
 tools/Makefile
-tools/version.h
-tools/fsadm/Makefile
+udev/Makefile
 ])
 AC_OUTPUT
 
 if test x$ODIRECT != xyes; then
   AC_MSG_WARN(Warning: O_DIRECT disabled: low-memory pvmove may lock up)
 fi
-
-if test x$FSADM == xyes; then
-  AC_MSG_WARN(fsadm support is untested)
-fi
-
-if test x$DMEVENTD == xyes; then
-  AC_MSG_WARN(dmeventd support is untested)
-fi

daemons/Makefile.in

@@ -15,9 +15,18 @@ srcdir = @srcdir@
 top_srcdir = @top_srcdir@
 VPATH = @srcdir@
 
+.PHONY: dmeventd clvmd
+
 ifneq ("@CLVMD@", "none")
   SUBDIRS = clvmd
 endif
 
+ifeq ("@DMEVENTD@", "yes")
+  SUBDIRS += dmeventd
+endif
+
 include $(top_srcdir)/make.tmpl
 
+ifeq ("@DMEVENTD@", "yes")
+device-mapper: dmeventd.device-mapper
+endif

daemons/clogd/Makefile

@@ -0,0 +1,77 @@
+###############################################################################
+###############################################################################
+##
+##  Copyright (C) 2009 Red Hat, Inc.  All rights reserved.
+##
+##  This copyrighted material is made available to anyone wishing to use,
+##  modify, copy, or redistribute it subject to the terms and conditions
+##  of the GNU General Public License v.2.
+##
+###############################################################################
+###############################################################################
+
+SOURCES = clogd.c cluster.c functions.c link_mon.c local.c logging.c
+
+TARGET = $(shell if [ ! -e /usr/include/linux/dm-log-userspace.h ]; then \
+		echo 'no_clogd_kernel_headers'; \
+	elif [ ! -e /usr/include/linux/ext2_fs.h ]; then \
+		echo 'no_e2fsprogs_devel'; \
+	elif [ ! -e /usr/include/openais/saCkpt.h ]; then \
+		echo 'no_openais_devel'; \
+	else \
+		echo 'clogd'; \
+	fi)
+
+ifneq ($(DEBUG), )
+CFLAGS += -DDEBUG
+endif
+
+ifneq ($(MEMB), )
+CFLAGS += -DMEMB
+endif
+
+ifneq ($(CKPT), )
+CFLAGS += -DCKPT
+endif
+
+ifneq ($(RESEND), )
+CFLAGS += -DRESEND
+endif
+
+CFLAGS += -g
+
+LDFLAGS += $(shell if [ -e /usr/lib64/openais ]; then \
+		echo '-L/usr/lib64/openais -L/usr/lib64 -L/lib64'; \
+	else \
+		echo '-L/usr/lib/openais -L/usr/lib -L/lib'; \
+	fi)
+LDFLAGS += -lcpg -lSaCkpt -lext2fs -ldevmapper
+
+all: ${TARGET}
+
+clogd: ${SOURCES}
+	${CC} ${CFLAGS} -o $@ $^ ${LDFLAGS}
+
+no_clogd_kernel_headers:
+	echo "Unable to find clogd kernel headers"
+	exit 1
+
+no_e2fsprogs_devel:
+	echo "Unable to find ext2fs kernel headers."
+	echo "Install 'e2fsprogs-devel'?"
+	exit 1
+
+no_openais_devel:
+	echo "Unable to find openAIS headers."
+	echo "http://sources.redhat.com/cluster/wiki/"
+	exit 1
+
+install: clogd
+	install -d /usr/sbin
+	install clogd /usr/sbin
+
+uninstall:
+	rm /usr/sbin/clogd
+
+clean:
+	rm -f *.o clogd *~

daemons/clogd/clogd.c

@@ -0,0 +1,265 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <string.h>
+#include <errno.h>
+#include <sched.h>
+#include <sys/types.h>
+#include <sys/wait.h>
+#include <sys/stat.h>
+#include <signal.h>
+#include <unistd.h>
+#include <fcntl.h>
+#include <linux/types.h>
+#include <sys/socket.h>
+#include <linux/netlink.h>
+#include <linux/dm-log-userspace.h>
+#include <linux/dm-ioctl.h>
+
+#include "functions.h"
+#include "local.h"
+#include "cluster.h"
+#include "common.h"
+#include "logging.h"
+#include "link_mon.h"
+
+static int exit_now = 0;
+static sigset_t signal_mask;
+static int signal_received;
+
+static void process_signals(void);
+static void daemonize(void);
+static void init_all(void);
+static void cleanup_all(void);
+
+int main(int argc, char *argv[])
+{
+	daemonize();
+
+	init_all();
+
+	/* Parent can now exit, we're ready to handle requests */
+	kill(getppid(), SIGTERM);
+
+	LOG_PRINT("Starting clogd:");
+	LOG_PRINT(" Built: "__DATE__" "__TIME__"\n");
+	LOG_DBG(" Compiled with debugging.");
+
+	while (!exit_now) {
+		links_monitor();
+
+		links_issue_callbacks();
+
+		process_signals();
+	}
+	exit(EXIT_SUCCESS);
+}
+
+/*
+ * parent_exit_handler: exit the parent
+ * @sig: the signal
+ *
+ */
+static void parent_exit_handler(int sig)
+{
+	exit_now = 1;
+}
+
+/*
+ * create_lockfile - create and lock a lock file
+ * @lockfile: location of lock file
+ *
+ * Returns: 0 on success, -1 otherwise
+ */
+static int create_lockfile(char *lockfile)
+{
+	int fd;
+	struct flock lock;
+	char buffer[50];
+
+	if((fd = open(lockfile, O_CREAT | O_WRONLY,
+		      (S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH))) < 0)
+		return -errno;
+
+	lock.l_type = F_WRLCK;
+	lock.l_start = 0;
+	lock.l_whence = SEEK_SET;
+	lock.l_len = 0;
+
+	if (fcntl(fd, F_SETLK, &lock) < 0) {
+		close(fd);
+		return -errno;
+	}
+
+	if (ftruncate(fd, 0) < 0) {
+		close(fd);
+		return -errno;
+	}
+
+	sprintf(buffer, "%d\n", getpid());
+
+	if(write(fd, buffer, strlen(buffer)) < strlen(buffer)){
+		close(fd);
+		unlink(lockfile);
+		return -errno;
+	}
+
+	return 0;
+}
+
+static void sig_handler(int sig)
+{
+	sigaddset(&signal_mask, sig);
+	++signal_received;
+}
+
+static void process_signal(int sig){
+	int r = 0;
+
+	switch(sig) {
+	case SIGINT:
+	case SIGQUIT:
+	case SIGTERM:
+	case SIGHUP:
+		r += log_status();
+		break;
+	case SIGUSR1:
+	case SIGUSR2:
+		log_debug();
+		/*local_debug();*/
+		cluster_debug();
+		return;
+	default:
+		LOG_PRINT("Unknown signal received... ignoring");
+		return;
+	}
+
+	if (!r) {
+		LOG_DBG("No current cluster logs... safe to exit.");
+		cleanup_all();
+		exit(EXIT_SUCCESS);
+	}
+
+	LOG_ERROR("Cluster logs exist.  Refusing to exit.");
+}
+
+static void process_signals(void)
+{
+	int x;
+
+	if (!signal_received)
+		return;
+
+	signal_received = 0;
+
+	for (x = 1; x < _NSIG; x++) {
+		if (sigismember(&signal_mask, x)) {
+			sigdelset(&signal_mask, x);
+			process_signal(x);
+		}
+	}
+}
+
+/*
+ * daemonize
+ *
+ * Performs the steps necessary to become a daemon.
+ */
+static void daemonize(void)
+{
+	int pid;
+	int status;
+
+	signal(SIGTERM, &parent_exit_handler);
+
+	pid = fork();
+
+	if (pid < 0) {
+		LOG_ERROR("Unable to fork()");
+		exit(EXIT_FAILURE);
+	}
+
+	if (pid) {
+		/* Parent waits here for child to get going */
+		while (!waitpid(pid, &status, WNOHANG) && !exit_now);
+		if (exit_now)
+			exit(EXIT_SUCCESS);
+
+		switch (WEXITSTATUS(status)) {
+		case EXIT_LOCKFILE:
+			LOG_ERROR("Failed to create lockfile");
+			LOG_ERROR("Process already running?");
+			break;
+		case EXIT_KERNEL_SOCKET:
+			LOG_ERROR("Unable to create netlink socket");
+			break;
+		case EXIT_KERNEL_BIND:
+			LOG_ERROR("Unable to bind to netlink socket");
+			break;
+		case EXIT_KERNEL_SETSOCKOPT:
+			LOG_ERROR("Unable to setsockopt on netlink socket");
+			break;
+		case EXIT_CLUSTER_CKPT_INIT:
+			LOG_ERROR("Unable to initialize checkpoint service");
+			LOG_ERROR("Has the cluster infrastructure been started?");
+			break;
+		case EXIT_FAILURE:
+			LOG_ERROR("Failed to start: Generic error");
+			break;
+		default:
+			LOG_ERROR("Failed to start: Unknown error");
+			break;
+		}
+		exit(EXIT_FAILURE);
+	}
+
+	setsid();
+	chdir("/");
+	umask(0);
+
+	close(0); close(1); close(2);
+	open("/dev/null", O_RDONLY); /* reopen stdin */
+	open("/dev/null", O_WRONLY); /* reopen stdout */
+	open("/dev/null", O_WRONLY); /* reopen stderr */
+
+	LOG_OPEN("clogd", LOG_PID, LOG_DAEMON);
+
+	if (create_lockfile("/var/run/clogd.pid"))
+		exit(EXIT_LOCKFILE);
+
+	signal(SIGINT, &sig_handler);
+	signal(SIGQUIT, &sig_handler);
+	signal(SIGTERM, &sig_handler);
+	signal(SIGHUP, &sig_handler);
+	signal(SIGPIPE, SIG_IGN);
+	signal(SIGUSR1, &sig_handler);
+	signal(SIGUSR2, &sig_handler);
+	sigemptyset(&signal_mask);
+	signal_received = 0;
+}
+
+/*
+ * init_all
+ *
+ * Initialize modules.  Exit on failure.
+ */
+static void init_all(void)
+{
+	int r;
+
+	if ((r = init_local()) ||
+	    (r = init_cluster())) {
+		exit(r);
+	}
+}
+
+/*
+ * cleanup_all
+ *
+ * Clean up before exiting
+ */
+static void cleanup_all(void)
+{
+	cleanup_local();
+	cleanup_cluster();
+}

daemons/clogd/cluster.h

@@ -0,0 +1,46 @@
+#ifndef __CLUSTER_LOG_CLUSTER_DOT_H__
+#define __CLUSTER_LOG_CLUSTER_DOT_H__
+
+#include <libdevmapper.h>
+#include <linux/dm-log-userspace.h>
+
+/*
+ * There is other information in addition to what can
+ * be found in the dm_ulog_request structure that we
+ * need for processing.  'clog_request' is the wrapping
+ * structure we use to make the additional fields
+ * available.
+ */
+struct clog_request {
+	struct dm_list list;
+
+	/*
+	 * 'originator' is the machine from which the requests
+	 * was made.
+	 */
+	uint32_t originator;
+
+	/*
+	 * 'pit_server' is the "point-in-time" server for the
+	 * request.  (I.e.  The machine that was the server at
+	 * the time the request was issued - only important during
+	 * startup.
+	 */
+	uint32_t pit_server;
+
+	/*
+	 * The request from the kernel that is being processed
+	 */
+	struct dm_ulog_request u_rq;
+};
+
+int init_cluster(void);
+void cleanup_cluster(void);
+void cluster_debug(void);
+
+int create_cluster_cpg(char *uuid, uint64_t luid);
+int destroy_cluster_cpg(char *uuid);
+
+int cluster_send(struct clog_request *rq);
+
+#endif /* __CLUSTER_LOG_CLUSTER_DOT_H__ */

daemons/clogd/common.h

@@ -0,0 +1,22 @@
+#ifndef __CLUSTER_LOG_COMMON_DOT_H__
+#define __CLUSTER_LOG_COMMON_DOT_H__
+
+/*
+#define EXIT_SUCCESS 0
+#define EXIT_FAILURE 1
+*/
+
+#define EXIT_LOCKFILE              2
+
+#define EXIT_KERNEL_SOCKET         3 /* Failed netlink socket create */
+#define EXIT_KERNEL_BIND           4
+#define EXIT_KERNEL_SETSOCKOPT     5
+
+#define EXIT_CLUSTER_CKPT_INIT     6 /* Failed to init checkpoint */
+
+#define EXIT_QUEUE_NOMEM           7
+
+
+#define DM_ULOG_REQUEST_SIZE 1024
+
+#endif /* __CLUSTER_LOG_COMMON_DOT_H__ */

daemons/clogd/functions.h

@@ -0,0 +1,23 @@
+#ifndef __CLOG_FUNCTIONS_DOT_H__
+#define __CLOG_FUNCTIONS_DOT_H__
+
+#include <linux/dm-log-userspace.h>
+#include "cluster.h"
+
+#define LOG_RESUMED   1
+#define LOG_SUSPENDED 2
+
+int local_resume(struct dm_ulog_request *rq);
+int cluster_postsuspend(char *, uint64_t);
+
+int do_request(struct clog_request *rq, int server);
+int push_state(const char *uuid, uint64_t luid,
+	       const char *which, char **buf, uint32_t debug_who);
+int pull_state(const char *uuid, uint64_t luid,
+	       const char *which, char *buf, int size);
+
+int log_get_state(struct dm_ulog_request *rq);
+int log_status(void);
+void log_debug(void);
+
+#endif /* __CLOG_FUNCTIONS_DOT_H__ */

daemons/clogd/link_mon.c

@@ -0,0 +1,138 @@
+#include <stdlib.h>
+#include <errno.h>
+#include <poll.h>
+
+#include "logging.h"
+
+struct link_callback {
+	int fd;
+	char *name;
+	void *data;
+	int (*callback)(void *data);
+
+	struct link_callback *next;
+};
+
+static int used_pfds = 0;
+static int free_pfds = 0;
+static struct pollfd *pfds = NULL;
+static struct link_callback *callbacks = NULL;
+
+int links_register(int fd, char *name, int (*callback)(void *data), void *data)
+{
+	int i;
+	struct link_callback *lc;
+
+	for (i = 0; i < used_pfds; i++) {
+		if (fd == pfds[i].fd) {
+			LOG_ERROR("links_register: Duplicate file descriptor");
+			return -EINVAL;
+		}
+	}
+
+	lc = malloc(sizeof(*lc));
+	if (!lc)
+		return -ENOMEM;
+
+	lc->fd = fd;
+	lc->name = name;
+	lc->data = data;
+	lc->callback = callback;
+
+	if (!free_pfds) {
+		struct pollfd *tmp;
+		tmp = realloc(pfds, sizeof(struct pollfd) * ((used_pfds*2) + 1));
+		if (!tmp) {
+			free(lc);
+			return -ENOMEM;
+		}
+		
+		pfds = tmp;
+		free_pfds = used_pfds + 1;
+	}
+
+	free_pfds--;
+	pfds[used_pfds].fd = fd;
+	pfds[used_pfds].events = POLLIN;
+	pfds[used_pfds].revents = 0;
+	used_pfds++;
+
+	lc->next = callbacks;
+	callbacks = lc;
+	LOG_DBG("Adding %s/%d", lc->name, lc->fd);
+	LOG_DBG(" used_pfds = %d, free_pfds = %d",
+		used_pfds, free_pfds);
+
+	return 0;
+}
+
+int links_unregister(int fd)
+{
+	int i;
+	struct link_callback *p, *c;
+
+	for (i = 0; i < used_pfds; i++)
+		if (fd == pfds[i].fd) {
+			/* entire struct is copied (overwritten) */
+			pfds[i] = pfds[used_pfds - 1];
+			used_pfds--;
+			free_pfds++;
+		}
+
+	for (p = NULL, c = callbacks; c; p = c, c = c->next)
+		if (fd == c->fd) {
+			LOG_DBG("Freeing up %s/%d", c->name, c->fd);
+			LOG_DBG(" used_pfds = %d, free_pfds = %d",
+				used_pfds, free_pfds);
+			if (p)
+				p->next = c->next;
+			else
+				callbacks = c->next;
+			free(c);
+			break;
+		}
+
+	return 0;
+}
+
+int links_monitor(void)
+{
+	int i, r;
+
+	for (i = 0; i < used_pfds; i++) {
+		pfds[i].revents = 0;
+	}
+
+	r = poll(pfds, used_pfds, -1);
+	if (r <= 0)
+		return r;
+
+	r = 0;
+	/* FIXME: handle POLLHUP */
+	for (i = 0; i < used_pfds; i++)
+		if (pfds[i].revents & POLLIN) {
+			LOG_DBG("Data ready on %d", pfds[i].fd);
+				
+			/* FIXME: Add this back return 1;*/
+			r++;
+		}
+
+	return r;
+}
+
+int links_issue_callbacks(void)
+{
+	int i;
+	struct link_callback *lc;
+
+	for (i = 0; i < used_pfds; i++)
+		if (pfds[i].revents & POLLIN)
+			for (lc = callbacks; lc; lc = lc->next)
+				if (pfds[i].fd == lc->fd) {
+					LOG_DBG("Issuing callback on %s/%d",
+						lc->name, lc->fd);
+					lc->callback(lc->data);
+					break;
+				}
+	return 0;
+}

daemons/clogd/link_mon.h

@@ -0,0 +1,9 @@
+#ifndef __LINK_MON_DOT_H__
+#define __LINK_MON_DOT_H__
+
+int links_register(int fd, char *name, int (*callback)(void *data), void *data);
+int links_unregister(int fd);
+int links_monitor(void);
+int links_issue_callbacks(void);
+
+#endif /* __LINK_MON_DOT_H__ */

daemons/clogd/local.c

@@ -0,0 +1,403 @@
+#include <unistd.h>
+#include <errno.h>
+#include <string.h>
+#include <stdint.h>
+#include <sys/types.h>
+#include <sys/socket.h>
+#include <sys/poll.h>
+#include <linux/connector.h>
+#include <linux/netlink.h>
+
+#include "linux/dm-log-userspace.h"
+#include "functions.h"
+#include "cluster.h"
+#include "common.h"
+#include "logging.h"
+#include "link_mon.h"
+#include "local.h"
+
+static int cn_fd;  /* Connector (netlink) socket fd */
+static char recv_buf[2048];
+static char send_buf[2048];
+
+
+/* FIXME: merge this function with kernel_send_helper */
+static int kernel_ack(uint32_t seq, int error)
+{
+	int r;
+	struct nlmsghdr *nlh = (struct nlmsghdr *)send_buf;
+	struct cn_msg *msg = NLMSG_DATA(nlh);
+
+	if (error < 0) {
+		LOG_ERROR("Programmer error: error codes must be positive");
+		return -EINVAL;
+	}
+
+	memset(send_buf, 0, sizeof(send_buf));
+
+	nlh->nlmsg_seq = 0;
+	nlh->nlmsg_pid = getpid();
+	nlh->nlmsg_type = NLMSG_DONE;
+	nlh->nlmsg_len = NLMSG_LENGTH(sizeof(struct cn_msg));
+	nlh->nlmsg_flags = 0;
+
+	msg->len = 0;
+	msg->id.idx = CN_IDX_DM;
+	msg->id.val = CN_VAL_DM_USERSPACE_LOG;
+	msg->seq = seq;
+	msg->ack = error;
+
+	r = send(cn_fd, nlh, NLMSG_LENGTH(sizeof(struct cn_msg)), 0);
+	/* FIXME: do better error processing */
+	if (r <= 0)
+		return -EBADE;
+
+	return 0;
+}
+
+
+/*
+ * kernel_recv
+ * @rq: the newly allocated request from kernel
+ *
+ * Read requests from the kernel and allocate space for the new request.
+ * If there is no request from the kernel, *rq is NULL.
+ *
+ * This function is not thread safe due to returned stack pointer.  In fact,
+ * the returned pointer must not be in-use when this function is called again.
+ *
+ * Returns: 0 on success, -EXXX on error
+ */
+static int kernel_recv(struct clog_request **rq)
+{
+	int r = 0;
+	int len;
+	struct cn_msg *msg;
+	struct dm_ulog_request *u_rq;
+
+	*rq = NULL;
+	memset(recv_buf, 0, sizeof(recv_buf));
+
+	len = recv(cn_fd, recv_buf, sizeof(recv_buf), 0);
+	if (len < 0) {
+		LOG_ERROR("Failed to recv message from kernel");
+		r = -errno;
+		goto fail;
+	}
+
+	switch (((struct nlmsghdr *)recv_buf)->nlmsg_type) {
+	case NLMSG_ERROR:
+		LOG_ERROR("Unable to recv message from kernel: NLMSG_ERROR");
+		r = -EBADE;
+		goto fail;
+	case NLMSG_DONE:
+		msg = (struct cn_msg *)NLMSG_DATA((struct nlmsghdr *)recv_buf);
+		len -= sizeof(struct nlmsghdr);
+
+		if (len < sizeof(struct cn_msg)) {
+			LOG_ERROR("Incomplete request from kernel received");
+			r = -EBADE;
+			goto fail;
+		}
+
+		if (msg->len > DM_ULOG_REQUEST_SIZE) {
+			LOG_ERROR("Not enough space to receive kernel request (%d/%d)",
+				  msg->len, DM_ULOG_REQUEST_SIZE);
+			r = -EBADE;
+			goto fail;
+		}
+
+		if (!msg->len)
+			LOG_ERROR("Zero length message received");
+
+		len -= sizeof(struct cn_msg);
+
+		if (len < msg->len)
+			LOG_ERROR("len = %d, msg->len = %d", len, msg->len);
+
+		msg->data[msg->len] = '\0'; /* Cleaner way to ensure this? */
+		u_rq = (struct dm_ulog_request *)msg->data;
+
+		if (!u_rq->request_type) {
+			LOG_DBG("Bad transmission, requesting resend [%u]",
+				msg->seq);
+			r = -EAGAIN;
+
+			if (kernel_ack(msg->seq, EAGAIN)) {
+				LOG_ERROR("Failed to NACK kernel transmission [%u]",
+					  msg->seq);
+				r = -EBADE;
+			}
+		}
+
+		/*
+		 * Now we've got sizeof(struct cn_msg) + sizeof(struct nlmsghdr)
+		 * worth of space that precede the request structure from the
+		 * kernel.  Since that space isn't going to be used again, we
+		 * can take it for our purposes; rather than allocating a whole
+		 * new structure and doing a memcpy.
+		 *
+		 * We should really make sure 'clog_request' doesn't grow
+		 * beyond what is available to us, but we need only check it
+		 * once... perhaps at compile time?
+		 */
+//		*rq = container_of(u_rq, struct clog_request, u_rq);
+		*rq = (void *)u_rq -
+			(sizeof(struct clog_request) -
+			 sizeof(struct dm_ulog_request));
+
+		/* Clear the wrapper container fields */
+		memset(*rq, 0, (void *)u_rq - (void *)(*rq));
+		break;
+	default:
+		LOG_ERROR("Unknown nlmsg_type");
+		r = -EBADE;
+	}
+
+fail:
+	if (r)
+		*rq = NULL;
+
+	return (r == -EAGAIN) ? 0 : r;
+}
+
+static int kernel_send_helper(void *data, int out_size)
+{
+	int r;
+	struct nlmsghdr *nlh;
+	struct cn_msg *msg;
+
+	memset(send_buf, 0, sizeof(send_buf));
+
+	nlh = (struct nlmsghdr *)send_buf;
+	nlh->nlmsg_seq = 0;  /* FIXME: Is this used? */
+	nlh->nlmsg_pid = getpid();
+	nlh->nlmsg_type = NLMSG_DONE;
+	nlh->nlmsg_len = NLMSG_LENGTH(out_size + sizeof(struct cn_msg));
+	nlh->nlmsg_flags = 0;
+
+	msg = NLMSG_DATA(nlh);
+	memcpy(msg->data, data, out_size);
+	msg->len = out_size;
+	msg->id.idx = CN_IDX_DM;
+	msg->id.val = CN_VAL_DM_USERSPACE_LOG;
+	msg->seq = 0;
+
+	r = send(cn_fd, nlh, NLMSG_LENGTH(out_size + sizeof(struct cn_msg)), 0);
+	/* FIXME: do better error processing */
+	if (r <= 0)
+		return -EBADE;
+
+	return 0;
+}
+
+/*
+ * do_local_work
+ *
+ * Any processing errors are placed in the 'rq'
+ * structure to be reported back to the kernel.
+ * It may be pointless for this function to
+ * return an int.
+ *
+ * Returns: 0 on success, -EXXX on failure
+ */
+static int do_local_work(void *data)
+{
+	int r;
+	struct clog_request *rq;
+	struct dm_ulog_request *u_rq = NULL;
+
+	r = kernel_recv(&rq);
+	if (r)
+		return r;
+
+	if (!rq)
+		return 0;
+
+	u_rq = &rq->u_rq;
+	LOG_DBG("[%s]  Request from kernel received: [%s/%u]",
+		SHORT_UUID(u_rq->uuid), RQ_TYPE(u_rq->request_type),
+		u_rq->seq);
+	switch (u_rq->request_type) {
+	case DM_ULOG_CTR:
+	case DM_ULOG_DTR:
+	case DM_ULOG_GET_REGION_SIZE:
+	case DM_ULOG_IN_SYNC:
+	case DM_ULOG_GET_SYNC_COUNT:
+	case DM_ULOG_STATUS_INFO:
+	case DM_ULOG_STATUS_TABLE:
+	case DM_ULOG_PRESUSPEND:
+		/* We do not specify ourselves as server here */
+		r = do_request(rq, 0);
+		if (r)
+			LOG_DBG("Returning failed request to kernel [%s]",
+				RQ_TYPE(u_rq->request_type));
+		r = kernel_send(u_rq);
+		if (r)
+			LOG_ERROR("Failed to respond to kernel [%s]",
+				  RQ_TYPE(u_rq->request_type));
+			
+		break;
+	case DM_ULOG_RESUME:
+		/*
+		 * Resume is a special case that requires a local
+		 * component to join the CPG, and a cluster component
+		 * to handle the request.
+		 */
+		r = local_resume(u_rq);
+		if (r) {
+			LOG_DBG("Returning failed request to kernel [%s]",
+				RQ_TYPE(u_rq->request_type));
+			r = kernel_send(u_rq);
+			if (r)
+				LOG_ERROR("Failed to respond to kernel [%s]",
+					  RQ_TYPE(u_rq->request_type));
+			break;
+		}
+		/* ELSE, fall through */
+	case DM_ULOG_IS_CLEAN:
+	case DM_ULOG_FLUSH:
+	case DM_ULOG_MARK_REGION:
+	case DM_ULOG_GET_RESYNC_WORK:
+	case DM_ULOG_SET_REGION_SYNC:
+	case DM_ULOG_IS_REMOTE_RECOVERING:
+	case DM_ULOG_POSTSUSPEND:
+		r = cluster_send(rq);
+		if (r) {
+			u_rq->data_size = 0;
+			u_rq->error = r;
+			kernel_send(u_rq);
+		}
+
+		break;
+	case DM_ULOG_CLEAR_REGION:
+		r = kernel_ack(u_rq->seq, 0);
+
+		r = cluster_send(rq);
+		if (r) {
+			/*
+			 * FIXME: store error for delivery on flush
+			 *        This would allow us to optimize MARK_REGION
+			 *        too.
+			 */
+		}
+
+		break;
+	default:
+		LOG_ERROR("Invalid log request received (%u), ignoring.",
+			  u_rq->request_type);
+
+		return 0;
+	}
+
+	if (r && !u_rq->error)
+		u_rq->error = r;
+
+	return r;
+}
+
+/*
+ * kernel_send
+ * @u_rq: result to pass back to kernel
+ *
+ * This function returns the u_rq structure
+ * (containing the results) to the kernel.
+ * It then frees the structure.
+ *
+ * WARNING: should the structure be freed if
+ * there is an error?  I vote 'yes'.  If the
+ * kernel doesn't get the response, it should
+ * resend the request.
+ *
+ * Returns: 0 on success, -EXXX on failure
+ */
+int kernel_send(struct dm_ulog_request *u_rq)
+{
+	int r;
+	int size;
+
+	if (!u_rq)
+		return -EINVAL;
+
+	size = sizeof(struct dm_ulog_request) + u_rq->data_size;
+
+	if (!u_rq->data_size && !u_rq->error) {
+		/* An ACK is all that is needed */
+
+		/* FIXME: add ACK code */
+	} else if (size > DM_ULOG_REQUEST_SIZE) {
+		/*
+		 * If we gotten here, we've already overrun
+		 * our allotted space somewhere.
+		 *
+		 * We must do something, because the kernel
+		 * is waiting for a response.
+		 */
+		LOG_ERROR("Not enough space to respond to server");
+		u_rq->error = -ENOSPC;
+		size = sizeof(struct dm_ulog_request);
+	}
+
+	r = kernel_send_helper(u_rq, size);
+	if (r)
+		LOG_ERROR("Failed to send msg to kernel.");
+
+	return r;
+}
+
+/*
+ * init_local
+ *
+ * Initialize kernel communication socket (netlink)
+ *
+ * Returns: 0 on success, values from common.h on failure
+ */
+int init_local(void)
+{
+	int r = 0;
+	int opt;
+	struct sockaddr_nl addr;
+
+	cn_fd = socket(PF_NETLINK, SOCK_DGRAM, NETLINK_CONNECTOR);
+	if (cn_fd < 0)
+		return EXIT_KERNEL_SOCKET;
+
+	/* memset to fix valgrind complaint */
+	memset(&addr, 0, sizeof(struct sockaddr_nl));
+
+	addr.nl_family = AF_NETLINK;
+	addr.nl_groups = CN_IDX_DM;
+	addr.nl_pid = 0;
+
+	r = bind(cn_fd, (struct sockaddr *) &addr, sizeof(addr));
+	if (r < 0) {
+		close(cn_fd);
+		return EXIT_KERNEL_BIND;
+	}
+
+	opt = addr.nl_groups;
+	r = setsockopt(cn_fd, 270, NETLINK_ADD_MEMBERSHIP, &opt, sizeof(opt));
+	if (r) {
+		close(cn_fd);
+		return EXIT_KERNEL_SETSOCKOPT;
+	}
+
+	/*
+	r = fcntl(cn_fd, F_SETFL, FNDELAY);
+	*/
+
+	links_register(cn_fd, "local", do_local_work, NULL);
+
+	return 0;
+}
+
+/*
+ * cleanup_local
+ *
+ * Clean up before exiting
+ */
+void cleanup_local(void)
+{
+	links_unregister(cn_fd);
+	close(cn_fd);
+}

daemons/clogd/local.h

@@ -0,0 +1,9 @@
+#ifndef __CLUSTER_LOG_LOCAL_DOT_H__
+#define __CLUSTER_LOG_LOCAL_DOT_H__
+
+int init_local(void);
+void cleanup_local(void);
+
+int kernel_send(struct dm_ulog_request *rq);
+
+#endif /* __CLUSTER_LOG_LOCAL_DOT_H__ */

daemons/clogd/logging.c

@@ -0,0 +1,47 @@
+#include <stdio.h>
+#include <syslog.h>
+
+char *__rq_types_off_by_one[] = {
+	"DM_ULOG_CTR",
+	"DM_ULOG_DTR",
+	"DM_ULOG_PRESUSPEND",
+	"DM_ULOG_POSTSUSPEND",
+	"DM_ULOG_RESUME",
+	"DM_ULOG_GET_REGION_SIZE",
+	"DM_ULOG_IS_CLEAN",
+	"DM_ULOG_IN_SYNC",
+	"DM_ULOG_FLUSH",
+	"DM_ULOG_MARK_REGION",
+	"DM_ULOG_CLEAR_REGION",
+	"DM_ULOG_GET_RESYNC_WORK",
+	"DM_ULOG_SET_REGION_SYNC",
+	"DM_ULOG_GET_SYNC_COUNT",
+	"DM_ULOG_STATUS_INFO",
+	"DM_ULOG_STATUS_TABLE",
+	"DM_ULOG_IS_REMOTE_RECOVERING",
+	NULL
+};
+
+int log_tabbing = 0;
+int log_is_open = 0;
+
+/*
+ * Variables for various conditional logging
+ */
+#ifdef MEMB
+int log_membership_change = 1;
+#else
+int log_membership_change = 0;
+#endif
+
+#ifdef CKPT
+int log_checkpoint = 1;
+#else
+int log_checkpoint = 0;
+#endif
+
+#ifdef RESEND
+int log_resend_requests = 1;
+#else
+int log_resend_requests = 0;
+#endif

daemons/clogd/logging.h

@@ -0,0 +1,60 @@
+#ifndef __CLUSTER_LOG_LOGGING_DOT_H__
+#define __CLUSTER_LOG_LOGGING_DOT_H__
+
+#include <stdio.h>
+#include <syslog.h>
+
+/* SHORT_UUID - print last 8 chars of a string */
+#define SHORT_UUID(x) (strlen(x) > 8) ? ((x) + (strlen(x) - 8)) : (x)
+
+extern char *__rq_types_off_by_one[];
+#define RQ_TYPE(x) __rq_types_off_by_one[(x) - 1]
+
+extern int log_tabbing;
+extern int log_is_open;
+extern int log_membership_change;
+extern int log_checkpoint;
+extern int log_resend_requests;
+
+#define LOG_OPEN(ident, option, facility) do { \
+		openlog(ident, option, facility); \
+		log_is_open = 1;		  \
+	} while (0)
+
+#define LOG_CLOSE(void) do { \
+		log_is_open = 0; \
+		closelog();	 \
+	} while (0)
+
+#define LOG_OUTPUT(level, f, arg...) do {				\
+		int __i;						\
+		char __buffer[16];					\
+		FILE *fp = (level > LOG_NOTICE) ? stderr : stdout;	\
+		if (log_is_open) {					\
+			for (__i = 0; (__i < log_tabbing) && (__i < 15); __i++) \
+				__buffer[__i] = '\t';			\
+			__buffer[__i] = '\0';				\
+			syslog(level, "%s" f "\n", __buffer, ## arg);	\
+		} else {						\
+			for (__i = 0; __i < log_tabbing; __i++)		\
+				fprintf(fp, "\t");			\
+			fprintf(fp, f "\n", ## arg);			\
+		}							\
+	} while (0)
+
+
+#ifdef DEBUG
+#define LOG_DBG(f, arg...) LOG_OUTPUT(LOG_DEBUG, f, ## arg)
+#else /* DEBUG */
+#define LOG_DBG(f, arg...)
+#endif /* DEBUG */
+
+#define LOG_COND(__X, f, arg...) do {\
+		if (__X) { 	     \
+			LOG_OUTPUT(LOG_NOTICE, f, ## arg); \
+		} \
+	} while (0)
+#define LOG_PRINT(f, arg...) LOG_OUTPUT(LOG_NOTICE, f, ## arg)
+#define LOG_ERROR(f, arg...) LOG_OUTPUT(LOG_ERR, f, ## arg)
+
+#endif /* __CLUSTER_LOG_LOGGING_DOT_H__ */

daemons/clvmd/Makefile.in

@@ -15,30 +15,40 @@ srcdir = @srcdir@
 top_srcdir = @top_srcdir@
 VPATH = @srcdir@
 
+QUORUM_LIBS = @QUORUM_LIBS@
+QUORUM_CFLAGS = @QUORUM_CFLAGS@
+CONFDB_LIBS = @CONFDB_LIBS@
+CONFDB_CFLAGS = @CONFDB_CFLAGS@
+CPG_LIBS = @CPG_LIBS@
+CPG_CFLAGS = @CPG_CFLAGS@
+
 SOURCES = \
 	clvmd-command.c  \
 	clvmd.c          \
 	lvm-functions.c  \
 	refresh_clvmd.c
 
-ifeq ("@CLVMD@", "gulm")
+ifneq (,$(findstring gulm,, "@CLVMD@,"))
 	GULM = yes
 endif
 
-ifeq ("@CLVMD@", "cman")
+ifneq (,$(findstring cman,, "@CLVMD@,"))
 	CMAN = yes
 endif
 
-ifeq ("@CLVMD@", "openais")
+ifneq (,$(findstring openais,, "@CLVMD@,"))
 	OPENAIS = yes
-	GULM = no
-	CMAN = no
 endif
 
-ifeq ("@CLVMD@", "all")
+ifneq (,$(findstring corosync,, "@CLVMD@,"))
+	COROSYNC = yes
+endif
+
+ifneq (,$(findstring all,, "@CLVMD@,"))
 	GULM = yes
 	CMAN = yes
-	OPENAIS = no
+	OPENAIS = yes
+	COROSYNC = yes
 endif
 
 ifeq ("@DEBUG@", "yes")
@@ -63,10 +73,18 @@ ifeq ("$(OPENAIS)", "yes")
 	DEFS += -DUSE_OPENAIS
 endif
 
+ifeq ("$(COROSYNC)", "yes")
+	SOURCES += clvmd-corosync.c
+	LMLIBS += $(QUORUM_LIBS) $(CONFDB_LIBS) $(CPG_LIBS) -ldlm
+	CFLAGS += $(QUORUM_CFLAGS) $(CONFDB_CFLAGS) $(CPG_CFLAGS)
+	DEFS += -DUSE_COROSYNC
+endif
+
+
 TARGETS = \
 	clvmd
 
-LVMLIBS = -llvm -lpthread
+LVMLIBS = -llvm-internal -lpthread
 
 ifeq ("@DMEVENTD@", "yes")
 	LVMLIBS += -ldevmapper-event
@@ -82,7 +100,7 @@ include $(top_srcdir)/make.tmpl
 INSTALL_TARGETS = \
 	install_clvmd
 
-clvmd: $(OBJECTS) $(top_srcdir)/lib/liblvm.a
+clvmd: $(OBJECTS) $(top_srcdir)/lib/liblvm-internal.a
 	$(CC) -o clvmd $(OBJECTS) $(CFLAGS) $(LDFLAGS) \
 		$(LVMLIBS) $(LMLIBS) $(LIBS)
 
@@ -90,7 +108,7 @@ clvmd: $(OBJECTS) $(top_srcdir)/lib/liblvm.a
 
 install_clvmd: $(TARGETS)
 	$(INSTALL) -D $(OWNER) $(GROUP) -m 555 $(STRIP) clvmd \
-		$(sbindir)/clvmd
+		$(usrsbindir)/clvmd
 
 install: $(INSTALL_TARGETS)
 

daemons/clvmd/clvm.h

@@ -62,6 +62,7 @@ static const char CLVMD_SOCKNAME[] = "\0clvmd";
 /* Lock/Unlock commands */
 #define CLVMD_CMD_LOCK_LV           50
 #define CLVMD_CMD_LOCK_VG           51
+#define CLVMD_CMD_LOCK_QUERY	    52
 
 /* Misc functions */
 #define CLVMD_CMD_REFRESH	    40

daemons/clvmd/clvmd-cman.c

@@ -17,6 +17,10 @@
  * CMAN communication layer for clvmd.
  */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -36,23 +40,28 @@
 #include <fcntl.h>
 #include <getopt.h>
 #include <errno.h>
+#include <libdevmapper.h>
 #include <libdlm.h>
 
 #include "clvmd-comms.h"
 #include "clvm.h"
-#include "log.h"
+#include "lvm-logging.h"
 #include "clvmd.h"
 #include "lvm-functions.h"
 
 #define LOCKSPACE_NAME "clvmd"
 
+struct clvmd_node
+{
+	struct cman_node *node;
+	int clvmd_up;
+};
+
 static int num_nodes;
 static struct cman_node *nodes = NULL;
 static struct cman_node this_node;
 static int count_nodes; /* size of allocated nodes array */
-static int max_updown_nodes = 50;	/* Current size of the allocated array */
-/* Node up/down status, indexed by nodeid */
-static int *node_updown = NULL;
+static struct dm_hash_table *node_updown_hash;
 static dlm_lshandle_t *lockspace;
 static cman_handle_t c_handle;
 
@@ -72,6 +81,8 @@ struct lock_wait {
 
 static int _init_cluster(void)
 {
+	node_updown_hash = dm_hash_create(100);
+
 	/* Open the cluster communication socket */
 	c_handle = cman_init(NULL);
 	if (!c_handle) {
@@ -99,8 +110,13 @@ static int _init_cluster(void)
 	/* Create a lockspace for LV & VG locks to live in */
 	lockspace = dlm_create_lockspace(LOCKSPACE_NAME, 0600);
 	if (!lockspace) {
-		syslog(LOG_ERR, "Unable to create lockspace for CLVM: %m");
-		return -1;
+		if (errno == EEXIST) {
+			lockspace = dlm_open_lockspace(LOCKSPACE_NAME);
+		}
+		if (!lockspace) {
+			syslog(LOG_ERR, "Unable to create lockspace for CLVM: %m");
+			return -1;
+		}
 	}
 	dlm_ls_pthread_init(lockspace);
 	DEBUGLOG("DLM initialisation complete\n");
@@ -165,8 +181,10 @@ static int _cluster_do_node_callback(struct local_client *client,
 
 	for (i = 0; i < _get_num_nodes(); i++) {
 		if (nodes[i].cn_member && nodes[i].cn_nodeid) {
-			callback(client, (char *)&nodes[i].cn_nodeid, node_updown[nodes[i].cn_nodeid]);
-			if (!node_updown[nodes[i].cn_nodeid])
+			int up = (int)(long)dm_hash_lookup_binary(node_updown_hash, (char *)&nodes[i].cn_nodeid, sizeof(int));
+
+			callback(client, (char *)&nodes[i].cn_nodeid, up);
+			if (!up)
 				somedown = -1;
 		}
 	}
@@ -184,7 +202,7 @@ static void event_callback(cman_handle_t handle, void *private, int reason, int
 		log_notice("clvmd on node %s has died\n", namebuf);
 		DEBUGLOG("Got port closed message, removing node %s\n", namebuf);
 
-		node_updown[arg] = 0;
+		dm_hash_insert_binary(node_updown_hash, (char *)&arg, sizeof(int), (void *)0);
 		break;
 
 	case CMAN_REASON_STATECHANGE:
@@ -239,28 +257,13 @@ static void _add_up_node(const char *csid)
 	/* It's up ! */
 	int nodeid = nodeid_from_csid(csid);
 
-	if (nodeid >= max_updown_nodes) {
-	        int new_size = nodeid + 10;
-		int *new_updown = realloc(node_updown, new_size);
-
-		if (new_updown) {
-			node_updown = new_updown;
-			max_updown_nodes = new_size;
-			DEBUGLOG("realloced more space for nodes. now %d\n",
-				 max_updown_nodes);
-		} else {
-			log_error
-				("Realloc failed. Node status for clvmd will be wrong. quitting\n");
-			exit(999);
-		}
-	}
-	node_updown[nodeid] = 1;
+	dm_hash_insert_binary(node_updown_hash, (char *)&nodeid, sizeof(int), (void *)1);
 	DEBUGLOG("Added new node %d to updown list\n", nodeid);
 }
 
 static void _cluster_closedown()
 {
-	unlock_all();
+	destroy_lvhash();
 	dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
 	cman_finish(c_handle);
 }
@@ -288,7 +291,12 @@ static void count_clvmds_running(void)
 	int i;
 
 	for (i = 0; i < num_nodes; i++) {
-		node_updown[nodes[i].cn_nodeid] = is_listening(nodes[i].cn_nodeid);
+		int nodeid = nodes[i].cn_nodeid;
+
+		if (is_listening(nodeid) == 1)
+			dm_hash_insert_binary(node_updown_hash, (void *)&nodeid, sizeof(int), (void*)1);
+		else
+			dm_hash_insert_binary(node_updown_hash, (void *)&nodeid, sizeof(int), (void*)0);
 	}
 }
 
@@ -297,6 +305,8 @@ static void get_members()
 {
 	int retnodes;
 	int status;
+	int i;
+	int high_nodeid = 0;
 
 	num_nodes = cman_get_node_count(c_handle);
 	if (num_nodes == -1) {
@@ -325,11 +335,10 @@ static void get_members()
 		exit(6);
 	}
 
-	if (node_updown == NULL) {
-		size_t buf_len = sizeof(int) * max(num_nodes, max_updown_nodes);
-		node_updown = malloc(buf_len);
-		if (node_updown)
-			memset(node_updown, 0, buf_len);
+	/* Get the highest nodeid */
+	for (i=0; i<retnodes; i++) {
+		if (nodes[i].cn_nodeid > high_nodeid)
+			high_nodeid = nodes[i].cn_nodeid;
 	}
 }
 

daemons/clvmd/clvmd-command.c

@@ -50,6 +50,10 @@
 
 */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/utsname.h>
@@ -67,9 +71,8 @@
 #include <libdevmapper.h>
 #include <libdlm.h>
 
-#include "list.h"
 #include "locking.h"
-#include "log.h"
+#include "lvm-logging.h"
 #include "lvm-functions.h"
 #include "clvmd-comms.h"
 #include "clvm.h"
@@ -87,6 +90,7 @@ int do_command(struct local_client *client, struct clvm_header *msg, int msglen,
 	int arglen = msglen - sizeof(struct clvm_header) - strlen(msg->node);
 	int status = 0;
 	char *lockname;
+	const char *locktype;
 	struct utsname nodeinfo;
 	unsigned char lock_cmd;
 	unsigned char lock_flags;
@@ -118,9 +122,11 @@ int do_command(struct local_client *client, struct clvm_header *msg, int msglen,
 		lockname = &args[2];
 		/* Check to see if the VG is in use by LVM1 */
 		status = do_check_lvm1(lockname);
-		/* P_global causes a cache refresh */
-		if (strcmp(lockname, "P_global") == 0)
-		      do_refresh_cache();
+		/* P_#global causes a full cache refresh */
+		if (!strcmp(lockname, "P_" VG_GLOBAL))
+			do_refresh_cache();
+		else
+			drop_metadata(lockname + 2);
 
 		break;
 
@@ -139,6 +145,14 @@ int do_command(struct local_client *client, struct clvm_header *msg, int msglen,
 		}
 		break;
 
+	case CLVMD_CMD_LOCK_QUERY:
+		lockname = &args[2];
+		if (buflen < 3)
+			return EIO;
+		if ((locktype = do_lock_query(lockname)))
+			*retlen = 1 + snprintf(*buf, buflen, "%s", locktype);
+		break;
+
 	case CLVMD_CMD_REFRESH:
 		do_refresh_cache();
 		break;
@@ -154,7 +168,11 @@ int do_command(struct local_client *client, struct clvm_header *msg, int msglen,
 		break;
 
 	case CLVMD_CMD_VG_BACKUP:
-		lvm_do_backup(&args[2]);
+		/*
+		 * Do not run backup on local node, caller should do that.
+		 */
+		if (!client)
+			lvm_do_backup(&args[2]);
 		break;
 
 	default:
@@ -251,7 +269,11 @@ int do_pre_command(struct local_client *client)
 		break;
 
 	case CLVMD_CMD_LOCK_VG:
-       	        status = lock_vg(client);
+		lockname = &args[2];
+		/* We take out a real lock unless LCK_CACHE was set */
+		if (!strncmp(lockname, "V_", 2) ||
+		    !strncmp(lockname, "P_#", 3))
+			status = lock_vg(client);
 		break;
 
 	case CLVMD_CMD_LOCK_LV:
@@ -265,6 +287,7 @@ int do_pre_command(struct local_client *client)
 	case CLVMD_CMD_GET_CLUSTERNAME:
 	case CLVMD_CMD_SET_DEBUG:
 	case CLVMD_CMD_VG_BACKUP:
+	case CLVMD_CMD_LOCK_QUERY:
 		break;
 
 	default:
@@ -295,6 +318,7 @@ int do_post_command(struct local_client *client)
 
 	case CLVMD_CMD_LOCK_VG:
 	case CLVMD_CMD_VG_BACKUP:
+	case CLVMD_CMD_LOCK_QUERY:
 		/* Nothing to do here */
 		break;
 

daemons/clvmd/clvmd-comms.h

@@ -93,5 +93,22 @@ struct cluster_ops *init_cman_cluster(void);
 struct cluster_ops *init_openais_cluster(void);
 #endif
 
+#ifdef USE_COROSYNC
+#  include <corosync/corotypes.h>
+#  define COROSYNC_CSID_LEN (sizeof(int))
+#  define COROSYNC_MAX_CLUSTER_MESSAGE         65535
+#  define COROSYNC_MAX_CLUSTER_MEMBER_NAME_LEN CS_MAX_NAME_LENGTH
+#  ifndef MAX_CLUSTER_MEMBER_NAME_LEN
+#    define MAX_CLUSTER_MEMBER_NAME_LEN       CS_MAX_NAME_LENGTH
+#  endif
+#  ifndef CMAN_MAX_CLUSTER_MESSAGE
+#    define CMAN_MAX_CLUSTER_MESSAGE          65535
+#  endif
+#  ifndef MAX_CSID_LEN
+#    define MAX_CSID_LEN sizeof(int)
+#  endif
+struct cluster_ops *init_corosync_cluster(void);
+#endif
+
 
 #endif

daemons/clvmd/clvmd-corosync.c

@@ -0,0 +1,643 @@
+/******************************************************************************
+*******************************************************************************
+**
+**  Copyright (C) 2009 Red Hat, Inc. All rights reserved.
+**
+*******************************************************************************
+******************************************************************************/
+
+/* This provides the interface between clvmd and corosync/DLM as the cluster
+ * and lock manager.
+ *
+ */
+
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
+#include <pthread.h>
+#include <sys/types.h>
+#include <sys/utsname.h>
+#include <sys/ioctl.h>
+#include <sys/socket.h>
+#include <sys/stat.h>
+#include <sys/file.h>
+#include <sys/socket.h>
+#include <netinet/in.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include <signal.h>
+#include <fcntl.h>
+#include <string.h>
+#include <stddef.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <errno.h>
+#include <utmpx.h>
+#include <syslog.h>
+#include <assert.h>
+#include <libdevmapper.h>
+
+#include <corosync/corotypes.h>
+#include <corosync/cpg.h>
+#include <corosync/quorum.h>
+#include <corosync/confdb.h>
+#include <libdlm.h>
+
+#include "locking.h"
+#include "lvm-logging.h"
+#include "clvm.h"
+#include "clvmd-comms.h"
+#include "lvm-functions.h"
+#include "clvmd.h"
+
+/* Timeout value for several corosync calls */
+#define LOCKSPACE_NAME "clvmd"
+
+static void cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len);
+static void cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries);
+static void _cluster_closedown(void);
+
+/* Hash list of nodes in the cluster */
+static struct dm_hash_table *node_hash;
+
+/* Number of active nodes */
+static int num_nodes;
+static unsigned int our_nodeid;
+
+static struct local_client *cluster_client;
+
+/* Corosync handles */
+static cpg_handle_t cpg_handle;
+static quorum_handle_t quorum_handle;
+
+/* DLM Handle */
+static dlm_lshandle_t *lockspace;
+
+static struct cpg_name cpg_group_name;
+
+/* Corosync callback structs */
+cpg_callbacks_t cpg_callbacks = {
+	.cpg_deliver_fn =            cpg_deliver_callback,
+	.cpg_confchg_fn =            cpg_confchg_callback,
+};
+
+quorum_callbacks_t quorum_callbacks = {
+	.quorum_notify_fn = NULL,
+};
+
+struct node_info
+{
+	enum {NODE_UNKNOWN, NODE_DOWN, NODE_UP, NODE_CLVMD} state;
+	int nodeid;
+};
+
+
+/* Set errno to something approximating the right value and return 0 or -1 */
+static int cs_to_errno(cs_error_t err)
+{
+	switch(err)
+	{
+	case CS_OK:
+		return 0;
+        case CS_ERR_LIBRARY:
+		errno = EINVAL;
+		break;
+        case CS_ERR_VERSION:
+		errno = EINVAL;
+		break;
+        case CS_ERR_INIT:
+		errno = EINVAL;
+		break;
+        case CS_ERR_TIMEOUT:
+		errno = ETIME;
+		break;
+        case CS_ERR_TRY_AGAIN:
+		errno = EAGAIN;
+		break;
+        case CS_ERR_INVALID_PARAM:
+		errno = EINVAL;
+		break;
+        case CS_ERR_NO_MEMORY:
+		errno = ENOMEM;
+		break;
+        case CS_ERR_BAD_HANDLE:
+		errno = EINVAL;
+		break;
+        case CS_ERR_BUSY:
+		errno = EBUSY;
+		break;
+        case CS_ERR_ACCESS:
+		errno = EPERM;
+		break;
+        case CS_ERR_NOT_EXIST:
+		errno = ENOENT;
+		break;
+        case CS_ERR_NAME_TOO_LONG:
+		errno = ENAMETOOLONG;
+		break;
+        case CS_ERR_EXIST:
+		errno = EEXIST;
+		break;
+        case CS_ERR_NO_SPACE:
+		errno = ENOSPC;
+		break;
+        case CS_ERR_INTERRUPT:
+		errno = EINTR;
+		break;
+	case CS_ERR_NAME_NOT_FOUND:
+		errno = ENOENT;
+		break;
+        case CS_ERR_NO_RESOURCES:
+		errno = ENOMEM;
+		break;
+        case CS_ERR_NOT_SUPPORTED:
+		errno = EOPNOTSUPP;
+		break;
+        case CS_ERR_BAD_OPERATION:
+		errno = EINVAL;
+		break;
+        case CS_ERR_FAILED_OPERATION:
+		errno = EIO;
+		break;
+        case CS_ERR_MESSAGE_ERROR:
+		errno = EIO;
+		break;
+        case CS_ERR_QUEUE_FULL:
+		errno = EXFULL;
+		break;
+        case CS_ERR_QUEUE_NOT_AVAILABLE:
+		errno = EINVAL;
+		break;
+        case CS_ERR_BAD_FLAGS:
+		errno = EINVAL;
+		break;
+        case CS_ERR_TOO_BIG:
+		errno = E2BIG;
+		break;
+        case CS_ERR_NO_SECTIONS:
+		errno = ENOMEM;
+		break;
+	default:
+		errno = EINVAL;
+		break;
+	}
+	return -1;
+}
+
+static char *print_corosync_csid(const char *csid)
+{
+	static char buf[128];
+	int id;
+
+	memcpy(&id, csid, sizeof(int));
+	sprintf(buf, "%d", id);
+	return buf;
+}
+
+static void cpg_deliver_callback (cpg_handle_t handle,
+				  const struct cpg_name *groupName,
+				  uint32_t nodeid,
+				  uint32_t pid,
+				  void *msg,
+				  size_t msg_len)
+{
+	int target_nodeid;
+
+	memcpy(&target_nodeid, msg, COROSYNC_CSID_LEN);
+
+	DEBUGLOG("%u got message from nodeid %d for %d. len %zd\n",
+		 our_nodeid, nodeid, target_nodeid, msg_len-4);
+
+	if (nodeid != our_nodeid)
+		if (target_nodeid == our_nodeid || target_nodeid == 0)
+			process_message(cluster_client, (char *)msg+COROSYNC_CSID_LEN,
+					msg_len-COROSYNC_CSID_LEN, (char*)&nodeid);
+}
+
+static void cpg_confchg_callback(cpg_handle_t handle,
+				 const struct cpg_name *groupName,
+				 const struct cpg_address *member_list, size_t member_list_entries,
+				 const struct cpg_address *left_list, size_t left_list_entries,
+				 const struct cpg_address *joined_list, size_t joined_list_entries)
+{
+	int i;
+	struct node_info *ninfo;
+
+	DEBUGLOG("confchg callback. %zd joined, %zd left, %zd members\n",
+		 joined_list_entries, left_list_entries, member_list_entries);
+
+	for (i=0; i<joined_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&joined_list[i].nodeid,
+					      COROSYNC_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = joined_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						      (char *)&ninfo->nodeid,
+						      COROSYNC_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	for (i=0; i<left_list_entries; i++) {
+		ninfo = dm_hash_lookup_binary(node_hash,
+					      (char *)&left_list[i].nodeid,
+					      COROSYNC_CSID_LEN);
+		if (ninfo)
+			ninfo->state = NODE_DOWN;
+	}
+
+	for (i=0; i<member_list_entries; i++) {
+		if (member_list[i].nodeid == 0) continue;
+		ninfo = dm_hash_lookup_binary(node_hash,
+				(char *)&member_list[i].nodeid,
+				COROSYNC_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = member_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						(char *)&ninfo->nodeid,
+						COROSYNC_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
+
+	num_nodes = member_list_entries;
+}
+
+static int _init_cluster(void)
+{
+	cs_error_t err;
+
+	node_hash = dm_hash_create(100);
+
+	err = cpg_initialize(&cpg_handle,
+			     &cpg_callbacks);
+	if (err != CS_OK) {
+		syslog(LOG_ERR, "Cannot initialise Corosync CPG service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise Corosync CPG service: %d", err);
+		return cs_to_errno(err);
+	}
+
+	err = quorum_initialize(&quorum_handle,
+				&quorum_callbacks);
+	if (err != CS_OK) {
+		syslog(LOG_ERR, "Cannot initialise Corosync quorum service: %d",
+		       err);
+		DEBUGLOG("Cannot initialise Corosync quorum service: %d", err);
+		return cs_to_errno(err);
+	}
+
+
+	/* Create a lockspace for LV & VG locks to live in */
+	lockspace = dlm_create_lockspace(LOCKSPACE_NAME, 0600);
+	if (!lockspace) {
+		if (errno == EEXIST) {
+			lockspace = dlm_open_lockspace(LOCKSPACE_NAME);
+		}
+		if (!lockspace) {
+			syslog(LOG_ERR, "Unable to create lockspace for CLVM: %m");
+			quorum_finalize(quorum_handle);
+			return -1;
+		}
+	}
+	dlm_ls_pthread_init(lockspace);
+	DEBUGLOG("DLM initialisation complete\n");
+
+	/* Connect to the clvmd group */
+	strcpy((char *)cpg_group_name.value, "clvmd");
+	cpg_group_name.length = strlen((char *)cpg_group_name.value);
+	err = cpg_join(cpg_handle, &cpg_group_name);
+	if (err != CS_OK) {
+		cpg_finalize(cpg_handle);
+		quorum_finalize(quorum_handle);
+		dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+		syslog(LOG_ERR, "Cannot join clvmd process group");
+		DEBUGLOG("Cannot join clvmd process group: %d\n", err);
+		return cs_to_errno(err);
+	}
+
+	err = cpg_local_get(cpg_handle,
+			    &our_nodeid);
+	if (err != CS_OK) {
+		cpg_finalize(cpg_handle);
+		quorum_finalize(quorum_handle);
+		dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+		syslog(LOG_ERR, "Cannot get local node id\n");
+		return cs_to_errno(err);
+	}
+	DEBUGLOG("Our local node id is %d\n", our_nodeid);
+
+	DEBUGLOG("Connected to Corosync\n");
+
+	return 0;
+}
+
+static void _cluster_closedown(void)
+{
+	DEBUGLOG("cluster_closedown\n");
+	destroy_lvhash();
+
+	dlm_release_lockspace(LOCKSPACE_NAME, lockspace, 1);
+	cpg_finalize(cpg_handle);
+	quorum_finalize(quorum_handle);
+}
+
+static void _get_our_csid(char *csid)
+{
+	memcpy(csid, &our_nodeid, sizeof(int));
+}
+
+/* Corosync doesn't really have nmode names so we
+   just use the node ID in hex instead */
+static int _csid_from_name(char *csid, const char *name)
+{
+	int nodeid;
+	struct node_info *ninfo;
+
+	if (sscanf(name, "%x", &nodeid) == 1) {
+		ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+		if (ninfo)
+			return nodeid;
+	}
+	return -1;
+}
+
+static int _name_from_csid(const char *csid, char *name)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+	if (!ninfo)
+	{
+		sprintf(name, "UNKNOWN %s", print_corosync_csid(csid));
+		return -1;
+	}
+
+	sprintf(name, "%x", ninfo->nodeid);
+	return 0;
+}
+
+static int _get_num_nodes()
+{
+	DEBUGLOG("num_nodes = %d\n", num_nodes);
+	return num_nodes;
+}
+
+/* Node is now known to be running a clvmd */
+static void _add_up_node(const char *csid)
+{
+	struct node_info *ninfo;
+
+	ninfo = dm_hash_lookup_binary(node_hash, csid, COROSYNC_CSID_LEN);
+	if (!ninfo) {
+		DEBUGLOG("corosync_add_up_node no node_hash entry for csid %s\n",
+			 print_corosync_csid(csid));
+		return;
+	}
+
+	DEBUGLOG("corosync_add_up_node %d\n", ninfo->nodeid);
+
+	ninfo->state = NODE_CLVMD;
+
+	return;
+}
+
+/* Call a callback for each node, so the caller knows whether it's up or down */
+static int _cluster_do_node_callback(struct local_client *master_client,
+				     void (*callback)(struct local_client *,
+						      const char *csid, int node_up))
+{
+	struct dm_hash_node *hn;
+	struct node_info *ninfo;
+	int somedown = 0;
+
+	dm_hash_iterate(hn, node_hash)
+	{
+		char csid[COROSYNC_CSID_LEN];
+
+		ninfo = dm_hash_get_data(node_hash, hn);
+		memcpy(csid, dm_hash_get_key(node_hash, hn), COROSYNC_CSID_LEN);
+
+		DEBUGLOG("down_callback. node %d, state = %d\n", ninfo->nodeid,
+			 ninfo->state);
+
+		if (ninfo->state != NODE_DOWN)
+			callback(master_client, csid, ninfo->state == NODE_CLVMD);
+		if (ninfo->state != NODE_CLVMD)
+			somedown = -1;
+	}
+	return somedown;
+}
+
+/* Real locking */
+static int _lock_resource(const char *resource, int mode, int flags, int *lockid)
+{
+	struct dlm_lksb lksb;
+	int err;
+
+	DEBUGLOG("lock_resource '%s', flags=%d, mode=%d\n", resource, flags, mode);
+
+	if (flags & LKF_CONVERT)
+		lksb.sb_lkid = *lockid;
+
+	err = dlm_ls_lock_wait(lockspace,
+			       mode,
+			       &lksb,
+			       flags,
+			       resource,
+			       strlen(resource),
+			       0,
+			       NULL, NULL, NULL);
+
+	if (err != 0)
+	{
+		DEBUGLOG("dlm_ls_lock returned %d\n", errno);
+		return err;
+	}
+	if (lksb.sb_status != 0)
+	{
+		DEBUGLOG("dlm_ls_lock returns lksb.sb_status %d\n", lksb.sb_status);
+		errno = lksb.sb_status;
+		return -1;
+	}
+
+	DEBUGLOG("lock_resource returning %d, lock_id=%x\n", err, lksb.sb_lkid);
+
+	*lockid = lksb.sb_lkid;
+
+	return 0;
+}
+
+
+static int _unlock_resource(const char *resource, int lockid)
+{
+	struct dlm_lksb lksb;
+	int err;
+
+	DEBUGLOG("unlock_resource: %s lockid: %x\n", resource, lockid);
+	lksb.sb_lkid = lockid;
+
+	err = dlm_ls_unlock_wait(lockspace,
+				 lockid,
+				 0,
+				 &lksb);
+	if (err != 0)
+	{
+		DEBUGLOG("Unlock returned %d\n", err);
+		return err;
+	}
+	if (lksb.sb_status != EUNLOCK)
+	{
+		DEBUGLOG("dlm_ls_unlock_wait returns lksb.sb_status: %d\n", lksb.sb_status);
+		errno = lksb.sb_status;
+		return -1;
+	}   
+
+
+	return 0;
+}
+
+static int _is_quorate()
+{
+	int quorate;
+	if (quorum_getquorate(quorum_handle, &quorate) == CS_OK)
+		return quorate;
+	else
+		return 0;
+}
+
+static int _get_main_cluster_fd(void)
+{
+	int select_fd;
+
+	cpg_fd_get(cpg_handle, &select_fd);
+	return select_fd;
+}
+
+static int _cluster_fd_callback(struct local_client *fd, char *buf, int len,
+				const char *csid,
+				struct local_client **new_client)
+{
+	cluster_client = fd;
+	*new_client = NULL;
+	cpg_dispatch(cpg_handle, CS_DISPATCH_ONE);
+	return 1;
+}
+
+static int _cluster_send_message(const void *buf, int msglen, const char *csid,
+				 const char *errtext)
+{
+	struct iovec iov[2];
+	cs_error_t err;
+	int target_node;
+
+	if (csid)
+		memcpy(&target_node, csid, COROSYNC_CSID_LEN);
+	else
+		target_node = 0;
+
+	iov[0].iov_base = &target_node;
+	iov[0].iov_len = sizeof(int);
+	iov[1].iov_base = (char *)buf;
+	iov[1].iov_len = msglen;
+
+	err = cpg_mcast_joined(cpg_handle, CPG_TYPE_AGREED, iov, 2);
+	return cs_to_errno(err);
+}
+
+/*
+ * We are not necessarily connected to a Red Hat Cluster system,
+ * but if we are, this returns the cluster name from cluster.conf.
+ * I've used confdb rather than ccs to reduce the inter-package
+ * dependancies as well as to allow people to set a cluster name
+ * for themselves even if they are not running on RH cluster.
+ */
+static int _get_cluster_name(char *buf, int buflen)
+{
+	confdb_handle_t handle;
+	int result;
+	size_t namelen = buflen;
+	hdb_handle_t cluster_handle;
+	confdb_callbacks_t callbacks = {
+		.confdb_key_change_notify_fn = NULL,
+		.confdb_object_create_change_notify_fn = NULL,
+		.confdb_object_delete_change_notify_fn = NULL
+	};
+
+	/* This is a default in case everything else fails */
+	strncpy(buf, "Corosync", buflen);
+
+	/* Look for a cluster name in confdb */
+	result = confdb_initialize (&handle, &callbacks);
+        if (result != CS_OK)
+		return 0;
+
+        result = confdb_object_find_start(handle, OBJECT_PARENT_HANDLE);
+	if (result != CS_OK)
+		goto out;
+
+        result = confdb_object_find(handle, OBJECT_PARENT_HANDLE, (void *)"cluster", strlen("cluster"), &cluster_handle);
+        if (result != CS_OK)
+		goto out;
+
+        result = confdb_key_get(handle, cluster_handle, (void *)"name", strlen("name"), buf, &namelen);
+        if (result != CS_OK)
+		goto out;
+
+	buf[namelen] = '\0';
+
+out:
+	confdb_finalize(handle);
+	return 0;
+}
+
+static struct cluster_ops _cluster_corosync_ops = {
+	.cluster_init_completed   = NULL,
+	.cluster_send_message     = _cluster_send_message,
+	.name_from_csid           = _name_from_csid,
+	.csid_from_name           = _csid_from_name,
+	.get_num_nodes            = _get_num_nodes,
+	.cluster_fd_callback      = _cluster_fd_callback,
+	.get_main_cluster_fd      = _get_main_cluster_fd,
+	.cluster_do_node_callback = _cluster_do_node_callback,
+	.is_quorate               = _is_quorate,
+	.get_our_csid             = _get_our_csid,
+	.add_up_node              = _add_up_node,
+	.reread_config            = NULL,
+	.cluster_closedown        = _cluster_closedown,
+	.get_cluster_name         = _get_cluster_name,
+	.sync_lock                = _lock_resource,
+	.sync_unlock              = _unlock_resource,
+};
+
+struct cluster_ops *init_corosync_cluster(void)
+{
+	if (!_init_cluster())
+		return &_cluster_corosync_ops;
+	else
+		return NULL;
+}

daemons/clvmd/clvmd-gulm.c

@@ -44,9 +44,8 @@
 #include <ccs.h>
 #include <libgulm.h>
 
-#include "list.h"
 #include "locking.h"
-#include "log.h"
+#include "lvm-logging.h"
 #include "clvm.h"
 #include "clvmd-comms.h"
 #include "lvm-functions.h"
@@ -249,7 +248,7 @@ static void _cluster_closedown(void)
 {
     DEBUGLOG("cluster_closedown\n");
     in_shutdown = 1;
-    unlock_all();
+    destroy_lvhash();
     lg_lock_logout(gulm_if);
     lg_core_logout(gulm_if);
     lg_release(gulm_if);
@@ -665,6 +664,7 @@ static int _cluster_do_node_callback(struct local_client *master_client,
 {
     struct dm_hash_node *hn;
     struct node_info *ninfo;
+    int somedown = 0;
 
     dm_hash_iterate(hn, node_hash)
     {
@@ -686,12 +686,14 @@ static int _cluster_do_node_callback(struct local_client *master_client,
 	    client = dm_hash_lookup_binary(sock_hash, csid, GULM_MAX_CSID_LEN);
 
 	}
+	DEBUGLOG("down_callback2. node %s, state = %d\n", ninfo->name, ninfo->state);
 	if (ninfo->state != NODE_DOWN)
 		callback(master_client, csid, ninfo->state == NODE_CLVMD);
 
-
+	if (ninfo->state != NODE_CLVMD)
+		somedown = -1;
     }
-    return 0;
+    return somedown;
 }
 
 /* Convert gulm error codes to unix errno numbers */
@@ -948,7 +950,7 @@ static int get_all_cluster_nodes()
 		}
 		else {
 			DEBUGLOG("Cannot resolve host name %s\n", nodename);
-			log_err("Cannot resolve host name %s\n", nodename);
+			log_error("Cannot resolve host name %s\n", nodename);
 		}
 	}
 	free(nodename);

daemons/clvmd/clvmd-openais.c

@@ -11,6 +11,10 @@
  *
  */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/utsname.h>
@@ -39,9 +43,8 @@
 #include <openais/saLck.h>
 #include <openais/cpg.h>
 
-#include "list.h"
 #include "locking.h"
-#include "log.h"
+#include "lvm-logging.h"
 #include "clvm.h"
 #include "clvmd-comms.h"
 #include "lvm-functions.h"
@@ -50,11 +53,6 @@
 /* Timeout value for several openais calls */
 #define TIMEOUT 10
 
-static void lck_lock_callback(SaInvocationT invocation,
-			      SaLckLockStatusT lockStatus,
-			      SaAisErrorT error);
-static void lck_unlock_callback(SaInvocationT invocation,
-				SaAisErrorT error);
 static void cpg_deliver_callback (cpg_handle_t handle,
 				  struct cpg_name *groupName,
 				  uint32_t nodeid,
@@ -92,11 +90,6 @@ cpg_callbacks_t cpg_callbacks = {
 	.cpg_confchg_fn =            cpg_confchg_callback,
 };
 
-SaLckCallbacksT lck_callbacks = {
-        .saLckLockGrantCallback      = lck_lock_callback,
-        .saLckResourceUnlockCallback = lck_unlock_callback
-};
-
 struct node_info
 {
 	enum {NODE_UNKNOWN, NODE_DOWN, NODE_UP, NODE_CLVMD} state;
@@ -110,13 +103,6 @@ struct lock_info
 	SaNameT              lock_name;
 };
 
-struct lock_wait
-{
-	pthread_cond_t cond;
-	pthread_mutex_t mutex;
-	int status;
-};
-
 /* Set errno to something approximating the right value and return 0 or -1 */
 static int ais_to_errno(SaAisErrorT err)
 {
@@ -209,7 +195,7 @@ static int ais_to_errno(SaAisErrorT err)
 	return -1;
 }
 
-static char *print_csid(const char *csid)
+static char *print_openais_csid(const char *csid)
 {
 	static char buf[128];
 	int id;
@@ -255,12 +241,13 @@ static void cpg_deliver_callback (cpg_handle_t handle,
 
 	memcpy(&target_nodeid, msg, OPENAIS_CSID_LEN);
 
-	DEBUGLOG("Got message from nodeid %d for %d. len %d\n",
-		 nodeid, target_nodeid, msg_len-4);
+	DEBUGLOG("%u got message from nodeid %d for %d. len %d\n",
+		 our_nodeid, nodeid, target_nodeid, msg_len-4);
 
-	if (target_nodeid == our_nodeid)
-		process_message(cluster_client, (char *)msg+OPENAIS_CSID_LEN,
-				msg_len-OPENAIS_CSID_LEN, (char*)&nodeid);
+	if (nodeid != our_nodeid)
+		if (target_nodeid == our_nodeid || target_nodeid == 0)
+			process_message(cluster_client, (char *)msg+OPENAIS_CSID_LEN,
+					msg_len-OPENAIS_CSID_LEN, (char*)&nodeid);
 }
 
 static void cpg_confchg_callback(cpg_handle_t handle,
@@ -302,34 +289,27 @@ static void cpg_confchg_callback(cpg_handle_t handle,
 			ninfo->state = NODE_DOWN;
 	}
 
-	num_nodes = joined_list_entries;
-}
+	for (i=0; i<member_list_entries; i++) {
+		if (member_list[i].nodeid == 0) continue;
+		ninfo = dm_hash_lookup_binary(node_hash,
+				(char *)&member_list[i].nodeid,
+				OPENAIS_CSID_LEN);
+		if (!ninfo) {
+			ninfo = malloc(sizeof(struct node_info));
+			if (!ninfo) {
+				break;
+			}
+			else {
+				ninfo->nodeid = member_list[i].nodeid;
+				dm_hash_insert_binary(node_hash,
+						(char *)&ninfo->nodeid,
+						OPENAIS_CSID_LEN, ninfo);
+			}
+		}
+		ninfo->state = NODE_CLVMD;
+	}
 
-static void lck_lock_callback(SaInvocationT invocation,
-			      SaLckLockStatusT lockStatus,
-			      SaAisErrorT error)
-{
-	struct lock_wait *lwait = (struct lock_wait *)(long)invocation;
-
-	DEBUGLOG("lck_lock_callback, error = %d\n", error);
-
-	lwait->status = error;
-	pthread_mutex_lock(&lwait->mutex);
-	pthread_cond_signal(&lwait->cond);
-	pthread_mutex_unlock(&lwait->mutex);
-}
-
-static void lck_unlock_callback(SaInvocationT invocation,
-				SaAisErrorT error)
-{
-	struct lock_wait *lwait = (struct lock_wait *)(long)invocation;
-
-	DEBUGLOG("lck_unlock_callback\n");
-
-	lwait->status = SA_AIS_OK;
-	pthread_mutex_lock(&lwait->mutex);
-	pthread_cond_signal(&lwait->cond);
-	pthread_mutex_unlock(&lwait->mutex);
+	num_nodes = member_list_entries;
 }
 
 static int lck_dispatch(struct local_client *client, char *buf, int len,
@@ -359,7 +339,7 @@ static int _init_cluster(void)
 	}
 
 	err = saLckInitialize(&lck_handle,
-			      &lck_callbacks,
+					NULL,
 			      &ver);
 	if (err != SA_AIS_OK) {
 		cpg_initialize(&cpg_handle, &cpg_callbacks);
@@ -402,7 +382,7 @@ static int _init_cluster(void)
 static void _cluster_closedown(void)
 {
 	DEBUGLOG("cluster_closedown\n");
-	unlock_all();
+	destroy_lvhash();
 
 	saLckFinalize(lck_handle);
 	cpg_finalize(cpg_handle);
@@ -435,7 +415,7 @@ static int _name_from_csid(const char *csid, char *name)
 	ninfo = dm_hash_lookup_binary(node_hash, csid, OPENAIS_CSID_LEN);
 	if (!ninfo)
 	{
-		sprintf(name, "UNKNOWN %s", print_csid(csid));
+		sprintf(name, "UNKNOWN %s", print_openais_csid(csid));
 		return -1;
 	}
 
@@ -457,7 +437,7 @@ static void _add_up_node(const char *csid)
 	ninfo = dm_hash_lookup_binary(node_hash, csid, OPENAIS_CSID_LEN);
 	if (!ninfo) {
 		DEBUGLOG("openais_add_up_node no node_hash entry for csid %s\n",
-			 print_csid(csid));
+			 print_openais_csid(csid));
 		return;
 	}
 
@@ -475,6 +455,7 @@ static int _cluster_do_node_callback(struct local_client *master_client,
 {
 	struct dm_hash_node *hn;
 	struct node_info *ninfo;
+	int somedown = 0;
 
 	dm_hash_iterate(hn, node_hash)
 	{
@@ -488,22 +469,20 @@ static int _cluster_do_node_callback(struct local_client *master_client,
 
 		if (ninfo->state != NODE_DOWN)
 			callback(master_client, csid, ninfo->state == NODE_CLVMD);
+		if (ninfo->state != NODE_CLVMD)
+			somedown = -1;
 	}
-	return 0;
+	return somedown;
 }
 
 /* Real locking */
 static int _lock_resource(char *resource, int mode, int flags, int *lockid)
 {
-	struct lock_wait lwait;
 	struct lock_info *linfo;
 	SaLckResourceHandleT res_handle;
 	SaAisErrorT err;
 	SaLckLockIdT lock_id;
-
-	pthread_cond_init(&lwait.cond, NULL);
-	pthread_mutex_init(&lwait.mutex, NULL);
-	pthread_mutex_lock(&lwait.mutex);
+	SaLckLockStatusT lockStatus;
 
 	/* This needs to be converted from DLM/LVM2 value for OpenAIS LCK */
 	if (flags & LCK_NONBLOCK) flags = SA_LCK_LOCK_NO_QUEUE;
@@ -526,24 +505,24 @@ static int _lock_resource(char *resource, int mode, int flags, int *lockid)
 		return ais_to_errno(err);
 	}
 
-	err = saLckResourceLockAsync(res_handle,
-				     (SaInvocationT)(long)&lwait,
-				     &lock_id,
-				     mode,
-				     flags,
-				     0);
-	if (err != SA_AIS_OK)
+	err = saLckResourceLock(
+			res_handle,
+			&lock_id,
+			mode,
+			flags,
+			0,
+			SA_TIME_END,
+			&lockStatus);
+	if (err != SA_AIS_OK && lockStatus != SA_LCK_LOCK_GRANTED)
 	{
 		free(linfo);
 		saLckResourceClose(res_handle);
 		return ais_to_errno(err);
 	}
-
+			
 	/* Wait for it to complete */
-	pthread_cond_wait(&lwait.cond, &lwait.mutex);
-	pthread_mutex_unlock(&lwait.mutex);
 
-	DEBUGLOG("lock_resource returning %d, lock_id=%llx\n", lwait.status,
+	DEBUGLOG("lock_resource returning %d, lock_id=%llx\n", err,
 		 lock_id);
 
 	linfo->lock_id = lock_id;
@@ -551,43 +530,34 @@ static int _lock_resource(char *resource, int mode, int flags, int *lockid)
 
 	dm_hash_insert(lock_hash, resource, linfo);
 
-	return ais_to_errno(lwait.status);
+	return ais_to_errno(err);
 }
 
 
 static int _unlock_resource(char *resource, int lockid)
 {
-	struct lock_wait lwait;
 	SaAisErrorT err;
 	struct lock_info *linfo;
 
-	pthread_cond_init(&lwait.cond, NULL);
-	pthread_mutex_init(&lwait.mutex, NULL);
-	pthread_mutex_lock(&lwait.mutex);
-
 	DEBUGLOG("unlock_resource %s\n", resource);
 	linfo = dm_hash_lookup(lock_hash, resource);
 	if (!linfo)
 		return 0;
 
 	DEBUGLOG("unlock_resource: lockid: %llx\n", linfo->lock_id);
-	err = saLckResourceUnlockAsync((SaInvocationT)(long)&lwait, linfo->lock_id);
+	err = saLckResourceUnlock(linfo->lock_id, SA_TIME_END);
 	if (err != SA_AIS_OK)
 	{
 		DEBUGLOG("Unlock returned %d\n", err);
 		return ais_to_errno(err);
 	}
 
-	/* Wait for it to complete */
-	pthread_cond_wait(&lwait.cond, &lwait.mutex);
-	pthread_mutex_unlock(&lwait.mutex);
-
 	/* Release the resource */
 	dm_hash_remove(lock_hash, resource);
 	saLckResourceClose(linfo->res_handle);
 	free(linfo);
 
-	return ais_to_errno(lwait.status);
+	return ais_to_errno(err);
 }
 
 static int _sync_lock(const char *resource, int mode, int flags, int *lockid)

daemons/clvmd/clvmd.c

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -17,6 +17,12 @@
  * CLVMD: Cluster LVM daemon
  */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
+#include <libdevmapper.h>
+
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -37,16 +43,16 @@
 #include <getopt.h>
 #include <syslog.h>
 #include <errno.h>
+#include <limits.h>
 #include <libdlm.h>
 
 #include "clvmd-comms.h"
 #include "lvm-functions.h"
 #include "clvm.h"
-#include "version.h"
+#include "lvm-version.h"
 #include "clvmd.h"
 #include "refresh_clvmd.h"
-#include "list.h"
-#include "log.h"
+#include "lvm-logging.h"
 
 #ifndef TRUE
 #define TRUE 1
@@ -74,7 +80,7 @@ static unsigned max_cluster_member_name_len;
 
 /* Structure of items on the LVM thread list */
 struct lvm_thread_cmd {
-	struct list list;
+	struct dm_list list;
 
 	struct local_client *client;
 	struct clvm_header *msg;
@@ -89,7 +95,7 @@ static pthread_t lvm_thread;
 static pthread_mutex_t lvm_thread_mutex;
 static pthread_cond_t lvm_thread_cond;
 static pthread_mutex_t lvm_start_mutex;
-static struct list lvm_cmd_head;
+static struct dm_list lvm_cmd_head;
 static volatile sig_atomic_t quit = 0;
 static volatile sig_atomic_t reread_config = 0;
 static int child_pipe[2];
@@ -102,6 +108,8 @@ static int child_pipe[2];
 #define DFAIL_TIMEOUT    5
 #define SUCCESS          0
 
+typedef enum {IF_AUTO, IF_CMAN, IF_GULM, IF_OPENAIS, IF_COROSYNC} if_type_t;
+
 /* Prototypes for code further down */
 static void sigusr2_handler(int sig);
 static void sighup_handler(int sig);
@@ -138,6 +146,7 @@ static void hton_clvm(struct clvm_header *hdr);
 static void ntoh_clvm(struct clvm_header *hdr);
 static void add_reply_to_list(struct local_client *client, int status,
 			      const char *csid, const char *buf, int len);
+static if_type_t parse_cluster_interface(char *ifname);
 
 static void usage(char *prog, FILE *file)
 {
@@ -146,10 +155,26 @@ static void usage(char *prog, FILE *file)
 	fprintf(file, "\n");
 	fprintf(file, "   -V       Show version of clvmd\n");
 	fprintf(file, "   -h       Show this help information\n");
-	fprintf(file, "   -d       Don't fork, run in the foreground\n");
+	fprintf(file, "   -d       Set debug level\n");
+	fprintf(file, "            If starting clvmd then don't fork, run in the foreground\n");
 	fprintf(file, "   -R       Tell all running clvmds in the cluster to reload their device cache\n");
+	fprintf(file, "   -C       Sets debug level (from -d) on all clvmd instances clusterwide\n");
 	fprintf(file, "   -t<secs> Command timeout (default 60 seconds)\n");
 	fprintf(file, "   -T<secs> Startup timeout (default none)\n");
+	fprintf(file, "   -I<cmgr> Cluster manager (default: auto)\n");
+	fprintf(file, "            Available cluster managers: ");
+#ifdef USE_COROSYNC
+	fprintf(file, "corosync ");
+#endif
+#ifdef USE_CMAN
+	fprintf(file, "cman ");
+#endif
+#ifdef USE_OPENAIS
+	fprintf(file, "openais ");
+#endif
+#ifdef USE_GULM
+	fprintf(file, "gulm ");
+#endif
 	fprintf(file, "\n");
 }
 
@@ -179,8 +204,10 @@ void debuglog(const char *fmt, ...)
 		va_end(ap);
 	}
 	if (debug == DEBUG_SYSLOG) {
-		if (!syslog_init)
+		if (!syslog_init) {
 			openlog("clvmd", LOG_PID, LOG_DAEMON);
+			syslog_init = 1;
+		}
 
 		va_start(ap,fmt);
 		vsyslog(LOG_DEBUG, fmt, ap);
@@ -188,6 +215,61 @@ void debuglog(const char *fmt, ...)
 	}
 }
 
+static const char *decode_cmd(unsigned char cmdl)
+{
+	static char buf[128];
+	const char *command;
+
+	switch (cmdl) {
+	case CLVMD_CMD_TEST:
+		command = "TEST";
+		break;
+	case CLVMD_CMD_LOCK_VG:
+		command = "LOCK_VG";
+		break;
+	case CLVMD_CMD_LOCK_LV:
+		command = "LOCK_LV";
+		break;
+	case CLVMD_CMD_REFRESH:
+		command = "REFRESH";
+		break;
+	case CLVMD_CMD_SET_DEBUG:
+		command = "SET_DEBUG";
+		break;
+	case CLVMD_CMD_GET_CLUSTERNAME:
+		command = "GET_CLUSTERNAME";
+		break;
+	case CLVMD_CMD_VG_BACKUP:
+		command = "VG_BACKUP";
+		break;
+	case CLVMD_CMD_REPLY:
+		command = "REPLY";
+		break;
+	case CLVMD_CMD_VERSION:
+		command = "VERSION";
+		break;
+	case CLVMD_CMD_GOAWAY:
+		command = "GOAWAY";
+		break;
+	case CLVMD_CMD_LOCK:
+		command = "LOCK";
+		break;
+	case CLVMD_CMD_UNLOCK:
+		command = "UNLOCK";
+		break;
+	case CLVMD_CMD_LOCK_QUERY:
+		command = "LOCK_QUERY";
+		break;
+	default:
+		command = "unknown";
+		break;
+	}
+
+	sprintf(buf, "%s (0x%x)", command, cmdl);
+
+	return buf;
+}
+
 int main(int argc, char *argv[])
 {
 	int local_sock;
@@ -196,6 +278,7 @@ int main(int argc, char *argv[])
 	signed char opt;
 	int cmd_timeout = DEFAULT_CMD_TIMEOUT;
 	int start_timeout = 0;
+	if_type_t cluster_iface = IF_AUTO;
 	sigset_t ss;
 	int using_gulm = 0;
 	int debug_opt = 0;
@@ -204,7 +287,7 @@ int main(int argc, char *argv[])
 	/* Deal with command-line arguments */
 	opterr = 0;
 	optind = 0;
-	while ((opt = getopt(argc, argv, "?vVhd::t:RT:C")) != EOF) {
+	while ((opt = getopt(argc, argv, "?vVhd::t:RT:CI:")) != EOF) {
 		switch (opt) {
 		case 'h':
 			usage(argv[0], stdout);
@@ -237,6 +320,9 @@ int main(int argc, char *argv[])
 				exit(1);
 			}
 			break;
+		case 'I':
+			cluster_iface = parse_cluster_interface(optarg);
+			break;
 		case 'T':
 			start_timeout = atoi(optarg);
 			if (start_timeout <= 0) {
@@ -289,13 +375,15 @@ int main(int argc, char *argv[])
 	signal(SIGHUP,  sighup_handler);
 	signal(SIGPIPE, SIG_IGN);
 
-	/* Block SIGUSR2 in the main process */
+	/* Block SIGUSR2/SIGINT/SIGTERM in process */
 	sigemptyset(&ss);
 	sigaddset(&ss, SIGUSR2);
+	sigaddset(&ss, SIGINT);
+	sigaddset(&ss, SIGTERM);
 	sigprocmask(SIG_BLOCK, &ss, NULL);
 
 	/* Initialise the LVM thread variables */
-	list_init(&lvm_cmd_head);
+	dm_list_init(&lvm_cmd_head);
 	pthread_mutex_init(&lvm_thread_mutex, NULL);
 	pthread_cond_init(&lvm_thread_cond, NULL);
 	pthread_mutex_init(&lvm_start_mutex, NULL);
@@ -303,7 +391,7 @@ int main(int argc, char *argv[])
 
 	/* Start the cluster interface */
 #ifdef USE_CMAN
-	if ((clops = init_cman_cluster())) {
+	if ((cluster_iface == IF_AUTO || cluster_iface == IF_CMAN) && (clops = init_cman_cluster())) {
 		max_csid_len = CMAN_MAX_CSID_LEN;
 		max_cluster_message = CMAN_MAX_CLUSTER_MESSAGE;
 		max_cluster_member_name_len = CMAN_MAX_NODENAME_LEN;
@@ -312,7 +400,7 @@ int main(int argc, char *argv[])
 #endif
 #ifdef USE_GULM
 	if (!clops)
-		if ((clops = init_gulm_cluster())) {
+		if ((cluster_iface == IF_AUTO || cluster_iface == IF_GULM) && (clops = init_gulm_cluster())) {
 			max_csid_len = GULM_MAX_CSID_LEN;
 			max_cluster_message = GULM_MAX_CLUSTER_MESSAGE;
 			max_cluster_member_name_len = GULM_MAX_CLUSTER_MEMBER_NAME_LEN;
@@ -320,9 +408,18 @@ int main(int argc, char *argv[])
 			syslog(LOG_NOTICE, "Cluster LVM daemon started - connected to GULM");
 		}
 #endif
+#ifdef USE_COROSYNC
+	if (!clops)
+		if (((cluster_iface == IF_AUTO || cluster_iface == IF_COROSYNC) && (clops = init_corosync_cluster()))) {
+			max_csid_len = COROSYNC_CSID_LEN;
+			max_cluster_message = COROSYNC_MAX_CLUSTER_MESSAGE;
+			max_cluster_member_name_len = COROSYNC_MAX_CLUSTER_MEMBER_NAME_LEN;
+			syslog(LOG_NOTICE, "Cluster LVM daemon started - connected to Corosync");
+		}
+#endif
 #ifdef USE_OPENAIS
 	if (!clops)
-		if ((clops = init_openais_cluster())) {
+		if ((cluster_iface == IF_AUTO || cluster_iface == IF_OPENAIS) && (clops = init_openais_cluster())) {
 			max_csid_len = OPENAIS_CSID_LEN;
 			max_cluster_message = OPENAIS_MAX_CLUSTER_MESSAGE;
 			max_cluster_member_name_len = OPENAIS_MAX_CLUSTER_MEMBER_NAME_LEN;
@@ -361,6 +458,9 @@ int main(int argc, char *argv[])
 	/* This needs to be started after cluster initialisation
 	   as it may need to take out locks */
 	DEBUGLOG("starting LVM thread\n");
+
+	/* Don't let anyone else to do work until we are started */
+	pthread_mutex_lock(&lvm_start_mutex);
 	pthread_create(&lvm_thread, NULL, lvm_thread_fn,
 			(void *)(long)using_gulm);
 
@@ -381,6 +481,8 @@ int main(int argc, char *argv[])
 	/* Do some work */
 	main_loop(local_sock, cmd_timeout);
 
+	destroy_lvm();
+
 	return 0;
 }
 
@@ -527,7 +629,7 @@ static void timedout_callback(struct local_client *client, const char *csid,
 		char nodename[max_cluster_member_name_len];
 
 		clops->name_from_csid(csid, nodename);
-		DEBUGLOG("PJC: checking for a reply from %s\n", nodename);
+		DEBUGLOG("Checking for a reply from %s\n", nodename);
 		pthread_mutex_lock(&client->bits.localsock.reply_mutex);
 
 		reply = client->bits.localsock.replies;
@@ -538,7 +640,7 @@ static void timedout_callback(struct local_client *client, const char *csid,
 		pthread_mutex_unlock(&client->bits.localsock.reply_mutex);
 
 		if (!reply) {
-			DEBUGLOG("PJC: node %s timed-out\n", nodename);
+			DEBUGLOG("Node %s timed-out\n", nodename);
 			add_reply_to_list(client, ETIMEDOUT, csid,
 					  "Command timed out", 18);
 		}
@@ -574,6 +676,11 @@ static void main_loop(int local_sock, int cmd_timeout)
 {
 	DEBUGLOG("Using timeout of %d seconds\n", cmd_timeout);
 
+	sigset_t ss;
+	sigemptyset(&ss);
+	sigaddset(&ss, SIGINT);
+	sigaddset(&ss, SIGTERM);
+	pthread_sigmask(SIG_UNBLOCK, &ss, NULL);
 	/* Main loop */
 	while (!quit) {
 		fd_set in;
@@ -630,7 +737,7 @@ static void main_loop(int local_sock, int cmd_timeout)
 				}
 
 				if (FD_ISSET(thisfd->fd, &in)) {
-					struct local_client *newfd;
+					struct local_client *newfd = NULL;
 					int ret;
 
 					/* Do callback */
@@ -901,8 +1008,7 @@ static int read_from_local_sock(struct local_client *thisfd)
 		}
 
 		/* Free the command buffer */
-		if (thisfd->bits.localsock.cmd)
-			free(thisfd->bits.localsock.cmd);
+		free(thisfd->bits.localsock.cmd);
 
 		/* Clear out the cross-link */
 		if (thisfd->bits.localsock.pipe_client != NULL)
@@ -937,8 +1043,7 @@ static int read_from_local_sock(struct local_client *thisfd)
 		}
 
 		/* Free any old buffer space */
-		if (thisfd->bits.localsock.cmd)
-			free(thisfd->bits.localsock.cmd);
+		free(thisfd->bits.localsock.cmd);
 
 		/* See if we have the whole message */
 		argslen =
@@ -984,6 +1089,14 @@ static int read_from_local_sock(struct local_client *thisfd)
 			}
 		}
 
+		/* Initialise and lock the mutex so the subthread will wait after
+		   finishing the PRE routine */
+		if (!thisfd->bits.localsock.threadid) {
+			pthread_mutex_init(&thisfd->bits.localsock.mutex, NULL);
+			pthread_cond_init(&thisfd->bits.localsock.cond, NULL);
+			pthread_mutex_init(&thisfd->bits.localsock.reply_mutex, NULL);
+		}
+
 		/* Only run the command if all the cluster nodes are running CLVMD */
 		if (((inheader->flags & CLVMD_FLAG_LOCAL) == 0) &&
 		    (check_all_clvmds_running(thisfd) == -1)) {
@@ -1055,12 +1168,6 @@ static int read_from_local_sock(struct local_client *thisfd)
 
 		thisfd->bits.localsock.pipe = comms_pipe[1];
 
-		/* Initialise and lock the mutex so the subthread will wait after
-		   finishing the PRE routine */
-		pthread_mutex_init(&thisfd->bits.localsock.mutex, NULL);
-		pthread_cond_init(&thisfd->bits.localsock.cond, NULL);
-		pthread_mutex_init(&thisfd->bits.localsock.reply_mutex, NULL);
-
 		/* Make sure the thread has a copy of it's own ID */
 		newfd->bits.pipe.threadid = thisfd->bits.localsock.threadid;
 
@@ -1166,8 +1273,8 @@ static void process_remote_command(struct clvm_header *msg, int msglen, int fd,
 	/* Get the node name as we /may/ need it later */
 	clops->name_from_csid(csid, nodename);
 
-	DEBUGLOG("process_remote_command %d for clientid 0x%x XID %d on node %s\n",
-		 msg->cmd, msg->clientid, msg->xid, nodename);
+	DEBUGLOG("process_remote_command %s for clientid 0x%x XID %d on node %s\n",
+		 decode_cmd(msg->cmd), msg->clientid, msg->xid, nodename);
 
 	/* Check for GOAWAY and sulk */
 	if (msg->cmd == CLVMD_CMD_GOAWAY) {
@@ -1415,7 +1522,8 @@ next_pre:
 		DEBUGLOG("Waiting for next pre command\n");
 
 		pthread_mutex_lock(&client->bits.localsock.mutex);
-		if (client->bits.localsock.state != PRE_COMMAND) {
+		if (client->bits.localsock.state != PRE_COMMAND &&
+		    !client->bits.localsock.finished) {
 			pthread_cond_wait(&client->bits.localsock.cond,
 					  &client->bits.localsock.mutex);
 		}
@@ -1437,8 +1545,9 @@ static int process_local_command(struct clvm_header *msg, int msglen,
 	int replylen = 0;
 	int status;
 
-	DEBUGLOG("process_local_command: msg=%p, msglen =%d, client=%p\n", msg,
-		 msglen, client);
+	DEBUGLOG("process_local_command: %s msg=%p, msglen =%d, client=%p\n",
+		 decode_cmd(msg->cmd), msg, msglen, client);
+
 	if (replybuf == NULL)
 		return -1;
 
@@ -1545,8 +1654,7 @@ static void send_local_reply(struct local_client *client, int status, int fd)
 		}
 		thisreply = thisreply->next;
 
-		if (tempreply->replymsg)
-			free(tempreply->replymsg);
+		free(tempreply->replymsg);
 		free(tempreply);
 	}
 
@@ -1577,8 +1685,7 @@ static void free_reply(struct local_client *client)
 
 		thisreply = thisreply->next;
 
-		if (tempreply->replymsg)
-			free(tempreply->replymsg);
+		free(tempreply->replymsg);
 		free(tempreply);
 	}
 	client->bits.localsock.replies = NULL;
@@ -1613,7 +1720,7 @@ static void send_version_message()
 static int send_message(void *buf, int msglen, const char *csid, int fd,
 			const char *errtext)
 {
-	int len;
+	int len = 0;
 	int saved_errno = 0;
 	struct timespec delay;
 	struct timespec remtime;
@@ -1692,13 +1799,10 @@ static int process_work_item(struct lvm_thread_cmd *cmd)
  */
 static __attribute__ ((noreturn)) void *lvm_thread_fn(void *arg)
 {
-	struct list *cmdl, *tmp;
+	struct dm_list *cmdl, *tmp;
 	sigset_t ss;
 	int using_gulm = (int)(long)arg;
 
-	/* Don't let anyone else to do work until we are started */
-	pthread_mutex_lock(&lvm_start_mutex);
-
 	DEBUGLOG("LVM thread function started\n");
 
 	/* Ignore SIGUSR1 & 2 */
@@ -1718,20 +1822,19 @@ static __attribute__ ((noreturn)) void *lvm_thread_fn(void *arg)
 		DEBUGLOG("LVM thread waiting for work\n");
 
 		pthread_mutex_lock(&lvm_thread_mutex);
-		if (list_empty(&lvm_cmd_head))
+		if (dm_list_empty(&lvm_cmd_head))
 			pthread_cond_wait(&lvm_thread_cond, &lvm_thread_mutex);
 
-		list_iterate_safe(cmdl, tmp, &lvm_cmd_head) {
+		dm_list_iterate_safe(cmdl, tmp, &lvm_cmd_head) {
 			struct lvm_thread_cmd *cmd;
 
 			cmd =
-			    list_struct_base(cmdl, struct lvm_thread_cmd, list);
-			list_del(&cmd->list);
+			    dm_list_struct_base(cmdl, struct lvm_thread_cmd, list);
+			dm_list_del(&cmd->list);
 			pthread_mutex_unlock(&lvm_thread_mutex);
 
 			process_work_item(cmd);
-			if (cmd->msg)
-				free(cmd->msg);
+			free(cmd->msg);
 			free(cmd);
 
 			pthread_mutex_lock(&lvm_thread_mutex);
@@ -1777,7 +1880,7 @@ static int add_to_lvmqueue(struct local_client *client, struct clvm_header *msg,
 	    ("add_to_lvmqueue: cmd=%p. client=%p, msg=%p, len=%d, csid=%p, xid=%d\n",
 	     cmd, client, msg, msglen, csid, cmd->xid);
 	pthread_mutex_lock(&lvm_thread_mutex);
-	list_add(&lvm_cmd_head, &cmd->list);
+	dm_list_add(&lvm_cmd_head, &cmd->list);
 	pthread_cond_signal(&lvm_thread_cond);
 	pthread_mutex_unlock(&lvm_thread_mutex);
 
@@ -1824,8 +1927,9 @@ static int open_local_sock()
 		log_error("Can't create local socket: %m");
 		return -1;
 	}
-	/* Set Close-on-exec */
+	/* Set Close-on-exec & non-blocking */
 	fcntl(local_socket, F_SETFD, 1);
+	fcntl(local_socket, F_SETFL, fcntl(local_socket, F_GETFL, 0) | O_NONBLOCK);
 
 	memset(&sockaddr, 0, sizeof(sockaddr));
 	memcpy(sockaddr.sun_path, CLVMD_SOCKNAME, sizeof(CLVMD_SOCKNAME));
@@ -1937,3 +2041,20 @@ int sync_unlock(const char *resource, int lockid)
 	return clops->sync_unlock(resource, lockid);
 }
 
+static if_type_t parse_cluster_interface(char *ifname)
+{
+	if_type_t iface = IF_AUTO;
+
+	if (!strcmp(ifname, "auto"))
+		iface = IF_AUTO;
+	if (!strcmp(ifname, "cman"))
+		iface = IF_CMAN;
+	if (!strcmp(ifname, "gulm"))
+		iface = IF_GULM;
+	if (!strcmp(ifname, "openais"))
+		iface = IF_OPENAIS;
+	if (!strcmp(ifname, "corosync"))
+		iface = IF_COROSYNC;
+
+	return iface;
+}

daemons/clvmd/cnxman-socket.h

@@ -1,226 +0,0 @@
-/******************************************************************************
-*******************************************************************************
-**
-**  Copyright (C) Sistina Software, Inc.  1997-2003  All rights reserved.
-**  Copyright (C) 2004 Red Hat, Inc.  All rights reserved.
-**
-**  This copyrighted material is made available to anyone wishing to use,
-**  modify, copy, or redistribute it subject to the terms and conditions
-**  of the GNU General Public License v.2.
-**
-*******************************************************************************
-******************************************************************************/
-
-/* CMAN socket interface header,
-   may be include by user or kernel code */
-
-#ifndef __CNXMAN_SOCKET_H
-#define __CNXMAN_SOCKET_H
-
-/* A currently unused number. TIPC also uses this number and you're unlikely
-   to be using both.
- */
-#define AF_CLUSTER 30
-#define PF_CLUSTER AF_CLUSTER
-
-/* Protocol(socket) types */
-#define CLPROTO_MASTER 2
-#define CLPROTO_CLIENT 3
-
-/* ioctls -- should register these properly */
-#define SIOCCLUSTER_NOTIFY            _IOW('x', 0x01, int)
-#define SIOCCLUSTER_REMOVENOTIFY      _IO( 'x', 0x02)
-#define SIOCCLUSTER_GETMEMBERS        _IOR('x', 0x03, struct cl_cluster_nodelist)
-#define SIOCCLUSTER_SETEXPECTED_VOTES _IOW('x', 0x04, int)
-#define SIOCCLUSTER_ISQUORATE         _IO( 'x', 0x05)
-#define SIOCCLUSTER_ISLISTENING       _IOW('x', 0x06, struct cl_listen_request)
-#define SIOCCLUSTER_GETALLMEMBERS     _IOR('x', 0x07, struct cl_cluster_nodelist)
-#define SIOCCLUSTER_SET_VOTES         _IOW('x', 0x08, int)
-#define SIOCCLUSTER_GET_VERSION       _IOR('x', 0x09, struct cl_version)
-#define SIOCCLUSTER_SET_VERSION       _IOW('x', 0x0a, struct cl_version)
-#define SIOCCLUSTER_ISACTIVE          _IO( 'x', 0x0b)
-#define SIOCCLUSTER_KILLNODE          _IOW('x', 0x0c, int)
-#define SIOCCLUSTER_GET_JOINCOUNT     _IO( 'x', 0x0d)
-#define SIOCCLUSTER_SERVICE_REGISTER  _IOW('x', 0x0e, char)
-#define SIOCCLUSTER_SERVICE_UNREGISTER _IO('x', 0x0f)
-#define SIOCCLUSTER_SERVICE_JOIN      _IO( 'x', 0x10)
-#define SIOCCLUSTER_SERVICE_LEAVE     _IO( 'x', 0x20)
-#define SIOCCLUSTER_SERVICE_SETSIGNAL _IOW('x', 0x30, int)
-#define SIOCCLUSTER_SERVICE_STARTDONE _IOW('x', 0x40, unsigned int)
-#define SIOCCLUSTER_SERVICE_GETEVENT  _IOR('x', 0x50, struct cl_service_event)
-#define SIOCCLUSTER_SERVICE_GETMEMBERS _IOR('x', 0x60, struct cl_cluster_nodelist)
-#define SIOCCLUSTER_SERVICE_GLOBALID  _IOR('x', 0x70, uint32_t)
-#define SIOCCLUSTER_SERVICE_SETLEVEL  _IOR('x', 0x80, int)
-#define SIOCCLUSTER_GETNODE	      _IOWR('x', 0x90, struct cl_cluster_node)
-#define SIOCCLUSTER_BARRIER           _IOW('x', 0x0a0, struct cl_barrier_info)
-
-/* These were setsockopts */
-#define SIOCCLUSTER_PASS_SOCKET       _IOW('x', 0x0b0, struct cl_passed_sock)
-#define SIOCCLUSTER_SET_NODENAME      _IOW('x', 0x0b1, char *)
-#define SIOCCLUSTER_SET_NODEID        _IOW('x', 0x0b2, int)
-#define SIOCCLUSTER_JOIN_CLUSTER      _IOW('x', 0x0b3, struct cl_join_cluster_info)
-#define SIOCCLUSTER_LEAVE_CLUSTER     _IOW('x', 0x0b4, int)
-
-
-/* Maximum size of a cluster message */
-#define CMAN_MAX_CLUSTER_MESSAGE          1500
-#define CMAN_MAX_CLUSTER_MEMBER_NAME_LEN   255
-#define MAX_BARRIER_NAME_LEN           33
-#define MAX_SA_ADDR_LEN                12
-#define MAX_CLUSTER_NAME_LEN           16
-
-/* Well-known cluster port numbers */
-#define CLUSTER_PORT_MEMBERSHIP  1	/* Mustn't block during cluster
-					 * transitions! */
-#define CLUSTER_PORT_SERVICES    2
-#define CLUSTER_PORT_SYSMAN      10	/* Remote execution daemon */
-#define CLUSTER_PORT_CLVMD       11	/* Cluster LVM daemon */
-#define CLUSTER_PORT_SLM         12	/* LVM SLM (simple lock manager) */
-
-/* Port numbers above this will be blocked when the cluster is inquorate or in
- * transition */
-#define HIGH_PROTECTED_PORT      9
-
-/* Reasons for leaving the cluster */
-#define CLUSTER_LEAVEFLAG_DOWN     0	/* Normal shutdown */
-#define CLUSTER_LEAVEFLAG_KILLED   1
-#define CLUSTER_LEAVEFLAG_PANIC    2
-#define CLUSTER_LEAVEFLAG_REMOVED  3	/* This one can reduce quorum */
-#define CLUSTER_LEAVEFLAG_REJECTED 4	/* Not allowed into the cluster in the
-					 * first place */
-#define CLUSTER_LEAVEFLAG_INCONSISTENT 5	/* Our view of the cluster is
-						 * in a minority */
-#define CLUSTER_LEAVEFLAG_DEAD         6	/* Discovered to be dead */
-#define CLUSTER_LEAVEFLAG_FORCE     0x10	/* Forced by command-line */
-
-/* OOB messages sent to a local socket */
-#define CLUSTER_OOB_MSG_PORTCLOSED  1
-#define CLUSTER_OOB_MSG_STATECHANGE 2
-#define CLUSTER_OOB_MSG_SERVICEEVENT 3
-
-/* Sendmsg flags, these are above the normal sendmsg flags so they don't
- * interfere */
-#define MSG_NOACK     0x010000	/* Don't need an ACK for this message */
-#define MSG_QUEUE     0x020000	/* Queue the message for sending later */
-#define MSG_MULTICAST 0x080000	/* Message was sent to all nodes in the cluster
-				 */
-#define MSG_ALLINT    0x100000	/* Send out of all interfaces */
-#define MSG_REPLYEXP  0x200000	/* Reply is expected */
-
-typedef enum { NODESTATE_REMOTEMEMBER, NODESTATE_JOINING, NODESTATE_MEMBER,
-	       NODESTATE_DEAD } nodestate_t;
-
-
-struct sockaddr_cl {
-	unsigned short scl_family;
-	unsigned char scl_flags;
-	unsigned char scl_port;
-	int           scl_nodeid;
-};
-
-/*
- * This is how we pass the multicast & receive sockets into kernel space.
- */
-struct cl_passed_sock {
-	int fd;			/* FD of master socket to do multicast on */
-	int number;		/* Socket number, to match up recvonly & bcast
-				 * sockets */
-        int multicast;          /* Is it multicast or receive ? */
-};
-
-/* Cluster configuration info passed when we join the cluster */
-struct cl_join_cluster_info {
-	unsigned char votes;
-	unsigned int expected_votes;
-	unsigned int two_node;
-	unsigned int config_version;
-
-        char cluster_name[17];
-};
-
-
-/* This is the structure, per node, returned from the membership ioctl */
-struct cl_cluster_node {
-	unsigned int size;
-	unsigned int node_id;
-	unsigned int us;
-	unsigned int leave_reason;
-	unsigned int incarnation;
-	nodestate_t state;
-	char name[CMAN_MAX_CLUSTER_MEMBER_NAME_LEN];
-	unsigned char votes;
-};
-
-/* The struct passed to the membership ioctls */
-struct cl_cluster_nodelist {
-        uint32_t max_members;
-        struct cl_cluster_node *nodes;
-};
-
-/* Structure passed to SIOCCLUSTER_ISLISTENING */
-struct cl_listen_request {
-	unsigned char port;
-        int           nodeid;
-};
-
-/* A Cluster PORTCLOSED message - received by a local user as an OOB message */
-struct cl_portclosed_oob {
-	unsigned char cmd;	/* CLUSTER_OOB_MSG_PORTCLOSED */
-	unsigned char port;
-};
-
-/* Get all version numbers or set the config version */
-struct cl_version {
-	unsigned int major;
-	unsigned int minor;
-	unsigned int patch;
-	unsigned int config;
-};
-
-/* structure passed to barrier ioctls */
-struct cl_barrier_info {
-	char cmd;
-	char name[MAX_BARRIER_NAME_LEN];
-	unsigned int flags;
-	unsigned long arg;
-};
-
-typedef enum { SERVICE_EVENT_STOP, SERVICE_EVENT_START, SERVICE_EVENT_FINISH,
-		SERVICE_EVENT_LEAVEDONE } service_event_t;
-
-typedef enum { SERVICE_START_FAILED, SERVICE_START_JOIN, SERVICE_START_LEAVE }
-		service_start_t;
-
-struct cl_service_event {
-	service_event_t type;
-	service_start_t start_type;
-	unsigned int event_id;
-	unsigned int last_stop;
-	unsigned int last_start;
-	unsigned int last_finish;
-	unsigned int node_count;
-};
-
-
-/* Commands to the barrier ioctl */
-#define BARRIER_IOCTL_REGISTER 1
-#define BARRIER_IOCTL_CHANGE   2
-#define BARRIER_IOCTL_DELETE   3
-#define BARRIER_IOCTL_WAIT     4
-
-/* Attributes of a barrier - bitmask */
-#define BARRIER_ATTR_AUTODELETE 1
-#define BARRIER_ATTR_MULTISTEP  2
-#define BARRIER_ATTR_MANUAL     4
-#define BARRIER_ATTR_ENABLED    8
-#define BARRIER_ATTR_CALLBACK  16
-
-/* Attribute setting commands */
-#define BARRIER_SETATTR_AUTODELETE 1
-#define BARRIER_SETATTR_MULTISTEP  2
-#define BARRIER_SETATTR_ENABLED    3
-#define BARRIER_SETATTR_NODES      4
-#define BARRIER_SETATTR_CALLBACK   5
-#define BARRIER_SETATTR_TIMEOUT    6
-
-#endif

daemons/clvmd/lvm-functions.c

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -13,6 +13,10 @@
  * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/utsname.h>
@@ -33,7 +37,6 @@
 #include <libdevmapper.h>
 #include <libdlm.h>
 
-#include "list.h"
 #include "lvm-types.h"
 #include "clvm.h"
 #include "clvmd-comms.h"
@@ -43,7 +46,8 @@
 /* LVM2 headers */
 #include "toolcontext.h"
 #include "lvmcache.h"
-#include "log.h"
+#include "lvm-logging.h"
+#include "lvm-globals.h"
 #include "activate.h"
 #include "locking.h"
 #include "archiver.h"
@@ -61,37 +65,170 @@ struct lv_info {
 	int lock_mode;
 };
 
+static const char *decode_locking_cmd(unsigned char cmdl)
+{
+	static char buf[128];
+	const char *type;
+	const char *scope;
+	const char *command;
+
+	switch (cmdl & LCK_TYPE_MASK) {
+	case LCK_NULL:   
+		type = "NULL";   
+		break;
+	case LCK_READ:   
+		type = "READ";   
+		break;
+	case LCK_PREAD:  
+		type = "PREAD";  
+		break;
+	case LCK_WRITE:  
+		type = "WRITE";  
+		break;
+	case LCK_EXCL:   
+		type = "EXCL";   
+		break;
+	case LCK_UNLOCK: 
+		type = "UNLOCK"; 
+		break;
+	default:
+		type = "unknown";
+		break;
+	}
+
+	switch (cmdl & LCK_SCOPE_MASK) {
+	case LCK_VG: 
+		scope = "VG"; 
+		break;
+	case LCK_LV: 
+		scope = "LV"; 
+		break;
+	default:
+		scope = "unknown";
+		break;
+	}
+
+	switch (cmdl & LCK_MASK) {
+	case LCK_LV_EXCLUSIVE & LCK_MASK:
+		command = "LCK_LV_EXCLUSIVE";  
+		break;
+	case LCK_LV_SUSPEND & LCK_MASK:    
+		command = "LCK_LV_SUSPEND";    
+		break;
+	case LCK_LV_RESUME & LCK_MASK:     
+		command = "LCK_LV_RESUME";     
+		break;
+	case LCK_LV_ACTIVATE & LCK_MASK:   
+		command = "LCK_LV_ACTIVATE";   
+		break;
+	case LCK_LV_DEACTIVATE & LCK_MASK: 
+		command = "LCK_LV_DEACTIVATE"; 
+		break;
+	default:
+		command = "unknown";
+		break;
+	}
+
+	sprintf(buf, "0x%x %s (%s|%s%s%s%s%s%s)", cmdl, command, type, scope,
+		cmdl & LCK_NONBLOCK   ? "|NONBLOCK" : "",
+		cmdl & LCK_HOLD       ? "|HOLD" : "",
+		cmdl & LCK_LOCAL      ? "|LOCAL" : "",
+		cmdl & LCK_CLUSTER_VG ? "|CLUSTER_VG" : "",
+		cmdl & LCK_CACHE      ? "|CACHE" : "");
+
+	return buf;
+}
+
+static const char *decode_flags(unsigned char flags)
+{
+	static char buf[128];
+
+	sprintf(buf, "0x%x (%s%s%s)", flags,
+		flags & LCK_PARTIAL_MODE	  ? "PARTIAL_MODE " : "",
+		flags & LCK_MIRROR_NOSYNC_MODE	  ? "MIRROR_NOSYNC " : "",
+		flags & LCK_DMEVENTD_MONITOR_MODE ? "DMEVENTD_MONITOR " : "");
+
+	return buf;
+}
+
 char *get_last_lvm_error()
 {
 	return last_error;
 }
 
-/* Return the mode a lock is currently held at (or -1 if not held) */
-static int get_current_lock(char *resource)
+/*
+ * Hash lock info helpers
+ */
+static struct lv_info *lookup_info(const char *resource)
 {
 	struct lv_info *lvi;
 
 	pthread_mutex_lock(&lv_hash_lock);
 	lvi = dm_hash_lookup(lv_hash, resource);
 	pthread_mutex_unlock(&lv_hash_lock);
-	if (lvi) {
+
+	return lvi;
+}
+
+static void insert_info(const char *resource, struct lv_info *lvi)
+{
+	pthread_mutex_lock(&lv_hash_lock);
+	dm_hash_insert(lv_hash, resource, lvi);
+	pthread_mutex_unlock(&lv_hash_lock);
+}
+
+static void remove_info(const char *resource)
+{
+	pthread_mutex_lock(&lv_hash_lock);
+	dm_hash_remove(lv_hash, resource);
+	pthread_mutex_unlock(&lv_hash_lock);
+}
+
+/*
+ * Return the mode a lock is currently held at (or -1 if not held)
+ */
+static int get_current_lock(char *resource)
+{
+	struct lv_info *lvi;
+
+	if ((lvi = lookup_info(resource)))
 		return lvi->lock_mode;
-	} else {
-		return -1;
-	}
+
+	return -1;
+}
+
+
+void init_lvhash()
+{
+	/* Create hash table for keeping LV locks & status */
+	lv_hash = dm_hash_create(100);
+	pthread_mutex_init(&lv_hash_lock, NULL);
+	pthread_mutex_init(&lvm_lock, NULL);
 }
 
 /* Called at shutdown to tidy the lockspace */
-void unlock_all()
+void destroy_lvhash()
 {
 	struct dm_hash_node *v;
+	struct lv_info *lvi;
+	char *resource;
+	int status;
 
 	pthread_mutex_lock(&lv_hash_lock);
-	dm_hash_iterate(v, lv_hash) {
-		struct lv_info *lvi = dm_hash_get_data(lv_hash, v);
 
-		sync_unlock(dm_hash_get_key(lv_hash, v), lvi->lock_id);
+	dm_hash_iterate(v, lv_hash) {
+		lvi = dm_hash_get_data(lv_hash, v);
+		resource = dm_hash_get_key(lv_hash, v);
+
+		if ((status = sync_unlock(resource, lvi->lock_id)))
+			DEBUGLOG("unlock_all. unlock failed(%d): %s\n",
+				 status,  strerror(errno));
+		free(lvi);
 	}
+
+	dm_hash_destroy(lv_hash);
+	lv_hash = NULL;
+
 	pthread_mutex_unlock(&lv_hash_lock);
 }
 
@@ -104,10 +241,7 @@ int hold_lock(char *resource, int mode, int flags)
 
 	flags &= LKF_NOQUEUE;	/* Only LKF_NOQUEUE is valid here */
 
-	pthread_mutex_lock(&lv_hash_lock);
-	lvi = dm_hash_lookup(lv_hash, resource);
-	pthread_mutex_unlock(&lv_hash_lock);
-	if (lvi) {
+	if ((lvi = lookup_info(resource))) {
 		/* Already exists - convert it */
 		status =
 		    sync_lock(resource, mode, LKF_CONVERT | flags,
@@ -133,11 +267,9 @@ int hold_lock(char *resource, int mode, int flags)
 			free(lvi);
 			DEBUGLOG("hold_lock. lock at %d failed: %s\n", mode,
 				 strerror(errno));
-		} else {
-		        pthread_mutex_lock(&lv_hash_lock);
-			dm_hash_insert(lv_hash, resource, lvi);
-			pthread_mutex_unlock(&lv_hash_lock);
-		}
+		} else
+			insert_info(resource, lvi);
+
 		errno = saved_errno;
 	}
 	return status;
@@ -150,10 +282,7 @@ int hold_unlock(char *resource)
 	int status;
 	int saved_errno;
 
-	pthread_mutex_lock(&lv_hash_lock);
-	lvi = dm_hash_lookup(lv_hash, resource);
-	pthread_mutex_unlock(&lv_hash_lock);
-	if (!lvi) {
+	if (!(lvi = lookup_info(resource))) {
 		DEBUGLOG("hold_unlock, lock not already held\n");
 		return 0;
 	}
@@ -161,9 +290,7 @@ int hold_unlock(char *resource)
 	status = sync_unlock(resource, lvi->lock_id);
 	saved_errno = errno;
 	if (!status) {
-	    	pthread_mutex_lock(&lv_hash_lock);
-		dm_hash_remove(lv_hash, resource);
-		pthread_mutex_unlock(&lv_hash_lock);
+		remove_info(resource);
 		free(lvi);
 	} else {
 		DEBUGLOG("hold_unlock. unlock failed(%d): %s\n", status,
@@ -306,35 +433,53 @@ static int do_deactivate_lv(char *resource, unsigned char lock_flags)
 	return 0;
 }
 
+const char *do_lock_query(char *resource)
+{
+	int mode;
+	const char *type = NULL;
+
+	mode = get_current_lock(resource);
+	switch (mode) {
+		case LKM_NLMODE: type = "NL"; break;
+		case LKM_CRMODE: type = "CR"; break;
+		case LKM_CWMODE: type = "CW"; break;
+		case LKM_PRMODE: type = "PR"; break;
+		case LKM_PWMODE: type = "PW"; break;
+		case LKM_EXMODE: type = "EX"; break;
+	}
+
+	DEBUGLOG("do_lock_query: resource '%s', mode %i (%s)\n", resource, mode, type ?: "?");
+
+	return type;
+}
+
 /* This is the LOCK_LV part that happens on all nodes in the cluster -
    it is responsible for the interaction with device-mapper and LVM */
 int do_lock_lv(unsigned char command, unsigned char lock_flags, char *resource)
 {
 	int status = 0;
 
-	DEBUGLOG("do_lock_lv: resource '%s', cmd = 0x%x, flags = %x\n",
-		 resource, command, lock_flags);
+	DEBUGLOG("do_lock_lv: resource '%s', cmd = %s, flags = %s\n",
+		 resource, decode_locking_cmd(command), decode_flags(lock_flags));
 
-	pthread_mutex_lock(&lvm_lock);
 	if (!cmd->config_valid || config_files_changed(cmd)) {
 		/* Reinitialise various settings inc. logging, filters */
-		if (!refresh_toolcontext(cmd)) {
+		if (do_refresh_cache()) {
 			log_error("Updated config file invalid. Aborting.");
-			pthread_mutex_unlock(&lvm_lock);
 			return EINVAL;
 		}
 	}
 
-	if (lock_flags & LCK_PARTIAL_MODE)
-		init_partial(1);
-
+	pthread_mutex_lock(&lvm_lock);
 	if (lock_flags & LCK_MIRROR_NOSYNC_MODE)
 		init_mirror_in_sync(1);
 
 	if (!(lock_flags & LCK_DMEVENTD_MONITOR_MODE))
 		init_dmeventd_monitor(0);
 
-	switch (command) {
+	cmd->partial_activation = (lock_flags & LCK_PARTIAL_MODE) ? 1 : 0;
+
+	switch (command & LCK_MASK) {
 	case LCK_LV_EXCLUSIVE:
 		status = do_activate_lv(resource, lock_flags, LKM_EXMODE);
 		break;
@@ -366,15 +511,14 @@ int do_lock_lv(unsigned char command, unsigned char lock_flags, char *resource)
 		break;
 	}
 
-	if (lock_flags & LCK_PARTIAL_MODE)
-		init_partial(0);
-
 	if (lock_flags & LCK_MIRROR_NOSYNC_MODE)
 		init_mirror_in_sync(0);
 
 	if (!(lock_flags & LCK_DMEVENTD_MONITOR_MODE))
 		init_dmeventd_monitor(DEFAULT_DMEVENTD_MONITOR);
 
+	cmd->partial_activation = 0;
+
 	/* clean the pool for another command */
 	dm_pool_empty(cmd->mem);
 	pthread_mutex_unlock(&lvm_lock);
@@ -391,8 +535,8 @@ int pre_lock_lv(unsigned char command, unsigned char lock_flags, char *resource)
 	   before suspending cluster-wide.
 	 */
 	if (command == LCK_LV_SUSPEND) {
-		DEBUGLOG("pre_lock_lv: resource '%s', cmd = 0x%x, flags = %d\n",
-			 resource, command, lock_flags);
+		DEBUGLOG("pre_lock_lv: resource '%s', cmd = %s, flags = %s\n",
+			 resource, decode_locking_cmd(command), decode_flags(lock_flags));
 
 		if (hold_lock(resource, LKM_PWMODE, LKF_NOQUEUE))
 			return errno;
@@ -411,8 +555,8 @@ int post_lock_lv(unsigned char command, unsigned char lock_flags,
 		int oldmode;
 
 		DEBUGLOG
-		    ("post_lock_lv: resource '%s', cmd = 0x%x, flags = %d\n",
-		     resource, command, lock_flags);
+		    ("post_lock_lv: resource '%s', cmd = %s, flags = %s\n",
+		     resource, decode_locking_cmd(command), decode_flags(lock_flags));
 
 		/* If the lock state is PW then restore it to what it was */
 		oldmode = get_current_lock(resource);
@@ -437,7 +581,7 @@ int post_lock_lv(unsigned char command, unsigned char lock_flags,
 	return 0;
 }
 
-/* Check if a VG is un use by LVM1 so we don't stomp on it */
+/* Check if a VG is in use by LVM1 so we don't stomp on it */
 int do_check_lvm1(const char *vgname)
 {
 	int status;
@@ -453,9 +597,14 @@ int do_refresh_cache()
 	DEBUGLOG("Refreshing context\n");
 	log_notice("Refreshing context");
 
+	pthread_mutex_lock(&lvm_lock);
+
 	ret = refresh_toolcontext(cmd);
 	init_full_scan_done(0);
 	lvmcache_label_scan(cmd, 2);
+	dm_pool_empty(cmd->mem);
+
+	pthread_mutex_unlock(&lvm_lock);
 
 	return ret==1?0:-1;
 }
@@ -470,9 +619,10 @@ static void drop_vg_locks()
 	char line[255];
 	FILE *vgs =
 	    popen
-	    ("lvm pvs --nolocking --noheadings -o vg_name", "r");
+	    ("lvm pvs  --config 'log{command_names=0 prefix=\"\"}' --nolocking --noheadings -o vg_name", "r");
 
-	sync_unlock("P_orphans", LCK_EXCL);
+	sync_unlock("P_" VG_ORPHANS, LCK_EXCL);
+	sync_unlock("P_" VG_GLOBAL, LCK_EXCL);
 
 	if (!vgs)
 		return;
@@ -499,6 +649,17 @@ static void drop_vg_locks()
 		DEBUGLOG("vgs fclose failed: %s\n", strerror(errno));
 }
 
+/*
+ * Drop lvmcache metadata
+ */
+void drop_metadata(const char *vgname)
+{
+	DEBUGLOG("Dropping metadata for VG %s\n", vgname);
+	pthread_mutex_lock(&lvm_lock);
+	lvmcache_drop_metadata(vgname);
+	pthread_mutex_unlock(&lvm_lock);
+}
+
 /*
  * Ideally, clvmd should be started before any LVs are active
  * but this may not be the case...
@@ -511,7 +672,7 @@ static void *get_initial_state()
 	char line[255];
 	FILE *lvs =
 	    popen
-	    ("lvm lvs --nolocking --noheadings -o vg_uuid,lv_uuid,lv_attr,vg_attr",
+	    ("lvm lvs  --config 'log{command_names=0 prefix=\"\"}' --nolocking --noheadings -o vg_uuid,lv_uuid,lv_attr,vg_attr",
 	     "r");
 
 	if (!lvs)
@@ -551,7 +712,7 @@ static void *get_initial_state()
 	return NULL;
 }
 
-static void lvm2_log_fn(int level, const char *file, int line,
+static void lvm2_log_fn(int level, const char *file, int line, int dm_errno,
 			const char *message)
 {
 
@@ -560,7 +721,7 @@ static void lvm2_log_fn(int level, const char *file, int line,
  	   We need to NULL the function ptr otherwise it will just call
 	   back into here! */
 	init_log_fn(NULL);
-	print_log(level, file, line, "%s", message);
+	print_log(level, file, line, dm_errno, "%s", message);
 	init_log_fn(lvm2_log_fn);
 
 	/*
@@ -598,14 +759,6 @@ static void check_config()
 	log_error("locking_type not set correctly in lvm.conf, cluster operations will not work.");
 }
 
-void init_lvhash()
-{
-	/* Create hash table for keeping LV locks & status */
-	lv_hash = dm_hash_create(100);
-	pthread_mutex_init(&lv_hash_lock, NULL);
-	pthread_mutex_init(&lvm_lock, NULL);
-}
-
 /* Backups up the LVM metadata if it's changed */
 void lvm_do_backup(const char *vgname)
 {
@@ -614,33 +767,38 @@ void lvm_do_backup(const char *vgname)
 
 	DEBUGLOG("Triggering backup of VG metadata for %s. suspended=%d\n", vgname, suspended);
 
-	vg = vg_read(cmd, vgname, NULL /*vgid*/, &consistent);
-	if (vg) {
-		if (consistent)
-			check_current_backup(vg);
-	}
-	else {
+	pthread_mutex_lock(&lvm_lock);
+
+	vg = vg_read_internal(cmd, vgname, NULL /*vgid*/, &consistent);
+
+	if (vg && consistent)
+		check_current_backup(vg);
+	else
 		log_error("Error backing up metadata, can't find VG for group %s", vgname);
-	}
+
+	vg_release(vg);
+	dm_pool_empty(cmd->mem);
+
+	pthread_mutex_unlock(&lvm_lock);
 }
 
 /* Called to initialise the LVM context of the daemon */
 int init_lvm(int using_gulm)
 {
-	if (!(cmd = create_toolcontext(NULL, 0, 1))) {
+	if (!(cmd = create_toolcontext(1, NULL))) {
 		log_error("Failed to allocate command context");
 		return 0;
 	}
 
+	if (stored_errno()) {
+		destroy_toolcontext(cmd);
+		return 0;
+	}
+
 	/* Use LOG_DAEMON for syslog messages instead of LOG_USER */
 	init_syslog(LOG_DAEMON);
 	openlog("clvmd", LOG_PID, LOG_DAEMON);
-	init_debug(cmd->current_settings.debug);
-	init_verbose(cmd->current_settings.verbose + VERBOSE_BASE_LEVEL);
-	set_activation(cmd->current_settings.activation);
-	archive_enable(cmd, cmd->current_settings.archive);
-	backup_enable(cmd, cmd->current_settings.backup);
-	cmd->cmd_line = (char *)"clvmd";
+	cmd->cmd_line = "clvmd";
 
 	/* Check lvm.conf is setup for cluster-LVM */
 	check_config();
@@ -656,3 +814,10 @@ int init_lvm(int using_gulm)
 
 	return 1;
 }
+
+void destroy_lvm(void)
+{
+	if (cmd)
+		destroy_toolcontext(cmd);
+	cmd = NULL;
+}

daemons/clvmd/lvm-functions.h

@@ -22,16 +22,19 @@ extern int pre_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
 		       char *resource);
 extern int do_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
 		      char *resource);
+extern const char *do_lock_query(char *resource);
 extern int post_lock_lv(unsigned char lock_cmd, unsigned char lock_flags,
 			char *resource);
 extern int do_check_lvm1(const char *vgname);
 extern int do_refresh_cache(void);
 extern int init_lvm(int using_gulm);
+extern void destroy_lvm(void);
 extern void init_lvhash(void);
+extern void destroy_lvhash(void);
 extern void lvm_do_backup(const char *vgname);
 extern int hold_unlock(char *resource);
 extern int hold_lock(char *resource, int mode, int flags);
-extern void unlock_all(void);
 extern char *get_last_lvm_error(void);
+extern void drop_metadata(const char *vgname);
 
 #endif

daemons/clvmd/refresh_clvmd.c

@@ -18,6 +18,10 @@
  *
  */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
+
+#include <configure.h>
 #include <stddef.h>
 #include <sys/socket.h>
 #include <sys/un.h>
@@ -26,6 +30,7 @@
 #include <libdevmapper.h>
 #include <stdint.h>
 #include <stdio.h>
+#include <limits.h>
 
 #include "clvm.h"
 #include "refresh_clvmd.h"

daemons/clvmd/tcp-comms.c

@@ -13,7 +13,10 @@
    It can also make outgoing connnections to the other clvmd nodes.
 */
 
+#define _GNU_SOURCE
+#define _FILE_OFFSET_BITS 64
 
+#include <configure.h>
 #include <pthread.h>
 #include <sys/types.h>
 #include <sys/utsname.h>

daemons/dmeventd/Makefile.in

@@ -17,23 +17,34 @@ VPATH = @srcdir@
 
 SOURCES = libdevmapper-event.c
 
+ifeq ("@STATIC_LINK@", "yes")
 LIB_STATIC = libdevmapper-event.a
+endif
+LIB_VERSION = $(LIB_VERSION_DM)
 
 ifeq ("@LIB_SUFFIX@","dylib")
   LIB_SHARED = libdevmapper-event.dylib
 else
   LIB_SHARED = libdevmapper-event.so
+  VERSIONED_SHLIB = $(LIB_SHARED).$(LIB_VERSION)
 endif
 
 TARGETS = dmeventd
 CLEAN_TARGETS = dmeventd.o
 
-include ../make.tmpl
+ifneq ($(MAKECMDGOALS),device-mapper)
+  SUBDIRS+=plugins
+endif
+
+include $(top_srcdir)/make.tmpl
+
+all: dmeventd
+device-mapper: dmeventd $(LIB_STATIC)
 
 LDFLAGS += -ldl -ldevmapper -lpthread
 CLDFLAGS += -ldl -ldevmapper -lpthread
 
-dmeventd: $(LIB_SHARED) dmeventd.o
+dmeventd: $(LIB_SHARED) $(VERSIONED_SHLIB) dmeventd.o
 	$(CC) -o $@ dmeventd.o $(CFLAGS) $(LDFLAGS) \
 	-L. -ldevmapper-event $(LIBS) -rdynamic
 
@@ -52,6 +63,8 @@ endif
 
 install: $(INSTALL_TYPE) install_include install_dmeventd
 
+install_device-mapper: $(INSTALL_TYPE) install_include install_dmeventd
+
 install_include:
 	$(INSTALL) -D $(OWNER) $(GROUP) -m 444 libdevmapper-event.h \
 		$(includedir)/libdevmapper-event.h
@@ -74,6 +87,10 @@ install_static: libdevmapper-event.a
 		$(libdir)/libdevmapper-event.a.$(LIB_VERSION)
 	$(LN_S) -f libdevmapper-event.a.$(LIB_VERSION) $(libdir)/libdevmapper-event.a
 
+$(VERSIONED_SHLIB): $(LIB_SHARED)
+	$(RM) -f $@
+	$(LN_S) $(LIB_SHARED) $@
+
 .PHONY: distclean_lib distclean
 
 distclean_lib: 

daemons/dmeventd/dmeventd.c

@@ -22,10 +22,9 @@
 #include "configure.h"
 #include "libdevmapper.h"
 #include "libdevmapper-event.h"
-#include "list.h"
 #include "dmeventd.h"
 //#include "libmultilog.h"
-#include "log.h"
+#include "dm-logging.h"
 
 #include <dlfcn.h>
 #include <errno.h>
@@ -36,6 +35,7 @@
 #include <sys/time.h>
 #include <sys/resource.h>
 #include <unistd.h>
+#include <signal.h>
 #include <arpa/inet.h>		/* for htonl, ntohl */
 
 #ifdef linux
@@ -57,11 +57,11 @@ static volatile sig_atomic_t _thread_registries_empty = 1;	/* registries are emp
 static int _debug = 0;
 
 /* List (un)link macros. */
-#define	LINK(x, head)		list_add(head, &(x)->list)
+#define	LINK(x, head)		dm_list_add(head, &(x)->list)
 #define	LINK_DSO(dso)		LINK(dso, &_dso_registry)
 #define	LINK_THREAD(thread)	LINK(thread, &_thread_registry)
 
-#define	UNLINK(x)		list_del(&(x)->list)
+#define	UNLINK(x)		dm_list_del(&(x)->list)
 #define	UNLINK_DSO(x)		UNLINK(x)
 #define	UNLINK_THREAD(x)	UNLINK(x)
 
@@ -100,7 +100,7 @@ static pthread_mutex_t _global_mutex;
 
 /* Data kept about a DSO. */
 struct dso_data {
-	struct list list;
+	struct dm_list list;
 
 	char *dso_name;		/* DSO name (eg, "evms", "dmraid", "lvm2"). */
 
@@ -142,7 +142,7 @@ struct dso_data {
 	int (*unregister_device)(const char *device, const char *uuid,
 				 int major, int minor, void **user);
 };
-static LIST_INIT(_dso_registry);
+static DM_LIST_INIT(_dso_registry);
 
 /* Structure to keep parsed register variables from client message. */
 struct message_data {
@@ -167,7 +167,7 @@ struct message_data {
  * occurs and the event processing function of the DSO gets called.
  */
 struct thread_status {
-	struct list list;
+	struct dm_list list;
 
 	pthread_t thread;
 
@@ -188,14 +188,14 @@ struct thread_status {
 	struct dm_task *current_task;
 	time_t next_time;
 	uint32_t timeout;
-	struct list timeout_list;
+	struct dm_list timeout_list;
 	void *dso_private; /* dso per-thread status variable */
 };
-static LIST_INIT(_thread_registry);
-static LIST_INIT(_thread_registry_unused);
+static DM_LIST_INIT(_thread_registry);
+static DM_LIST_INIT(_thread_registry_unused);
 
 static int _timeout_running;
-static LIST_INIT(_timeout_registry);
+static DM_LIST_INIT(_timeout_registry);
 static pthread_mutex_t _timeout_mutex = PTHREAD_MUTEX_INITIALIZER;
 static pthread_cond_t _timeout_cond = PTHREAD_COND_INITIALIZER;
 
@@ -238,7 +238,7 @@ static struct thread_status *_alloc_thread_status(struct message_data *data,
 	ret->dso_data = dso_data;
 	ret->events = data->events.field;
 	ret->timeout = data->timeout.secs;
-	list_init(&ret->timeout_list);
+	dm_list_init(&ret->timeout_list);
 
 	return ret;
 }
@@ -458,7 +458,7 @@ static struct thread_status *_lookup_thread_status(struct message_data *data)
 {
 	struct thread_status *thread;
 
-	list_iterate_items(thread, &_thread_registry)
+	dm_list_iterate_items(thread, &_thread_registry)
 	    if (!strcmp(data->device_uuid, thread->device.uuid))
 		return thread;
 
@@ -488,13 +488,13 @@ static void *_timeout_thread(void *unused __attribute((unused)))
 	pthread_cleanup_push(_exit_timeout, NULL);
 	pthread_mutex_lock(&_timeout_mutex);
 
-	while (!list_empty(&_timeout_registry)) {
+	while (!dm_list_empty(&_timeout_registry)) {
 		struct thread_status *thread;
 
 		timeout.tv_sec = 0;
 		curr_time = time(NULL);
 
-		list_iterate_items_gen(thread, &_timeout_registry, timeout_list) {
+		dm_list_iterate_items_gen(thread, &_timeout_registry, timeout_list) {
 			if (thread->next_time <= curr_time) {
 				thread->next_time = curr_time + thread->timeout;
 				pthread_kill(thread->thread, SIGALRM);
@@ -521,8 +521,8 @@ static int _register_for_timeout(struct thread_status *thread)
 
 	thread->next_time = time(NULL) + thread->timeout;
 
-	if (list_empty(&thread->timeout_list)) {
-		list_add(&_timeout_registry, &thread->timeout_list);
+	if (dm_list_empty(&thread->timeout_list)) {
+		dm_list_add(&_timeout_registry, &thread->timeout_list);
 		if (_timeout_running)
 			pthread_cond_signal(&_timeout_cond);
 	}
@@ -542,9 +542,9 @@ static int _register_for_timeout(struct thread_status *thread)
 static void _unregister_for_timeout(struct thread_status *thread)
 {
 	pthread_mutex_lock(&_timeout_mutex);
-	if (!list_empty(&thread->timeout_list)) {
-		list_del(&thread->timeout_list);
-		list_init(&thread->timeout_list);
+	if (!dm_list_empty(&thread->timeout_list)) {
+		dm_list_del(&thread->timeout_list);
+		dm_list_init(&thread->timeout_list);
 	}
 	pthread_mutex_unlock(&_timeout_mutex);
 }
@@ -697,7 +697,7 @@ static void _monitor_unregister(void *arg)
 	}
 	/* we may have been relinked to unused registry since we were
 	   called, so check that */
-	list_iterate_items(thread_iter, &_thread_registry_unused)
+	dm_list_iterate_items(thread_iter, &_thread_registry_unused)
 		if (thread_iter == thread) {
 			thread->status = DM_THREAD_DONE;
 			_unlock_mutex();
@@ -838,7 +838,7 @@ static struct dso_data *_lookup_dso(struct message_data *data)
 {
 	struct dso_data *dso_data, *ret = NULL;
 
-	list_iterate_items(dso_data, &_dso_registry)
+	dm_list_iterate_items(dso_data, &_dso_registry)
 	    if (!strcmp(data->dso_name, dso_data->dso_name)) {
 		_lib_get(dso_data);
 		ret = dso_data;
@@ -1109,7 +1109,7 @@ static int _get_registered_dev(struct message_data *message_data, int next)
 	_lock_mutex();
 
 	/* Iterate list of threads checking if we want a particular one. */
-	list_iterate_items(thread, &_thread_registry)
+	dm_list_iterate_items(thread, &_thread_registry)
 		if (_want_registered_device(message_data->dso_name,
 					    message_data->device_uuid,
 					    thread)) {
@@ -1132,10 +1132,10 @@ static int _get_registered_dev(struct message_data *message_data, int next)
 	thread = hit;
 
 	while (1) {
-		if (list_end(&_thread_registry, &thread->list))
+		if (dm_list_end(&_thread_registry, &thread->list))
 			goto out;
 
-		thread = list_item(thread->list.n, struct thread_status);
+		thread = dm_list_item(thread->list.n, struct thread_status);
 		if (_want_registered_device(message_data->dso_name, NULL, thread)) {
 			hit = thread;
 			break;
@@ -1439,12 +1439,12 @@ static void _process_request(struct dm_event_fifos *fifos)
 static void _cleanup_unused_threads(void)
 {
 	int ret;
-	struct list *l;
+	struct dm_list *l;
 	struct thread_status *thread;
 
 	_lock_mutex();
-	while ((l = list_first(&_thread_registry_unused))) {
-		thread = list_item(l, struct thread_status);
+	while ((l = dm_list_first(&_thread_registry_unused))) {
+		thread = dm_list_item(l, struct thread_status);
 		if (thread->processing)
 			break;	/* cleanup on the next round */
 
@@ -1469,7 +1469,7 @@ static void _cleanup_unused_threads(void)
 				break;
 			}
 
-			list_del(l);
+			dm_list_del(l);
 			syslog(LOG_ERR,
 			       "thread can't be on unused list unless !thread->events");
 			thread->status = DM_THREAD_RUNNING;
@@ -1479,7 +1479,7 @@ static void _cleanup_unused_threads(void)
 		}
 
 		if (thread->status == DM_THREAD_DONE) {
-			list_del(l);
+			dm_list_del(l);
 			pthread_join(thread->thread, NULL);
 			_lib_put(thread->dso_data);
 			_free_thread_status(thread);
@@ -1740,8 +1740,8 @@ int main(int argc, char *argv[])
 	while (!_exit_now) {
 		_process_request(&fifos);
 		_cleanup_unused_threads();
-		if (!list_empty(&_thread_registry)
-		    || !list_empty(&_thread_registry_unused))
+		if (!dm_list_empty(&_thread_registry)
+		    || !dm_list_empty(&_thread_registry_unused))
 			_thread_registries_empty = 0;
 		else
 			_thread_registries_empty = 1;

daemons/dmeventd/libdevmapper-event.c

@@ -12,7 +12,7 @@
  * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
 
-#include "lib.h"
+#include "dmlib.h"
 #include "libdevmapper-event.h"
 //#include "libmultilog.h"
 #include "dmeventd.h"
@@ -425,6 +425,12 @@ static int _start_daemon(struct dm_event_fifos *fifos)
 
       start_server:
 	/* server is not running */
+
+	if (!strncmp(DMEVENTD_PATH, "/", 1) && stat(DMEVENTD_PATH, &statbuf)) {
+		log_error("Unable to find dmeventd.");
+		return_0;
+	}
+
 	pid = fork();
 
 	if (pid < 0)
@@ -432,7 +438,7 @@ static int _start_daemon(struct dm_event_fifos *fifos)
 
 	else if (!pid) {
 		execvp(DMEVENTD_PATH, NULL);
-		exit(EXIT_FAILURE);
+		_exit(EXIT_FAILURE);
 	} else {
 		if (waitpid(pid, &status, 0) < 0)
 			log_error("Unable to start dmeventd: %s",

daemons/dmeventd/plugins/Makefile.in

@@ -16,7 +16,7 @@ srcdir = @srcdir@
 top_srcdir = @top_srcdir@
 VPATH = @srcdir@
 
-SUBDIRS += mirror
+SUBDIRS += mirror snapshot
 
 include $(top_srcdir)/make.tmpl
 

daemons/dmeventd/plugins/mirror/Makefile.in

@@ -1,6 +1,6 @@
 #
 # Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
-# Copyright (C) 2004-2005 Red Hat, Inc. All rights reserved.
+# Copyright (C) 2004-2005, 2008 Red Hat, Inc. All rights reserved.
 #
 # This file is part of LVM2.
 #
@@ -17,7 +17,7 @@ top_srcdir = @top_srcdir@
 VPATH = @srcdir@
 
 INCLUDES += -I${top_srcdir}/tools
-CLDFLAGS += -L${top_srcdir}/tools -ldevmapper -llvm2cmd
+CLDFLAGS += -L${top_srcdir}/tools -ldevmapper @LVM2CMD_LIB@
 
 SOURCES = dmeventd_mirror.c
 
@@ -27,10 +27,13 @@ else
   LIB_SHARED = libdevmapper-event-lvm2mirror.so
 endif
 
+LIB_VERSION = $(LIB_VERSION_LVM)
+
 include $(top_srcdir)/make.tmpl
 
-install: libdevmapper-event-lvm2mirror.$(LIB_SUFFIX)
+install_lvm2: libdevmapper-event-lvm2mirror.$(LIB_SUFFIX)
 	$(INSTALL) -D $(OWNER) $(GROUP) -m 555 $(STRIP) $< \
 		$(libdir)/$<.$(LIB_VERSION)
 	$(LN_S) -f $<.$(LIB_VERSION) $(libdir)/$<
 
+install: install_lvm2

daemons/dmeventd/plugins/mirror/dmeventd_mirror.c

@@ -13,6 +13,7 @@
  */
 
 #include "lvm2cmd.h"
+#include "errors.h"
 
 #include <libdevmapper.h>
 #include <libdevmapper-event.h>
@@ -125,8 +126,9 @@ out_parse:
 	return ME_IGNORE;
 }
 
-static void _temporary_log_fn(int level, const char *file,
-			      int line, const char *format)
+static void _temporary_log_fn(int level, const char *file __attribute((unused)),
+			      int line __attribute((unused)),
+			      const char *format)
 {
 	if (!strncmp(format, "WARNING: ", 9) && (level < 5))
 		syslog(LOG_CRIT, "%s", format);
@@ -151,7 +153,7 @@ static int _remove_failed_devices(const char *device)
 	}
 
 	/* FIXME Is any sanity-checking required on %s? */
-	if (CMD_SIZE <= snprintf(cmd_str, CMD_SIZE, "vgreduce --config devices{ignore_suspended_devices=1} --removemissing %s", vg)) {
+	if (CMD_SIZE <= snprintf(cmd_str, CMD_SIZE, "lvconvert --config devices{ignore_suspended_devices=1} --repair --use-policies %s/%s", vg, lv)) {
 		/* this error should be caught above, but doesn't hurt to check again */
 		syslog(LOG_ERR, "Unable to form LVM command: Device name too long");
 		dm_pool_empty(_mem_pool);  /* FIXME: not safe with multiple threads */
@@ -160,11 +162,18 @@ static int _remove_failed_devices(const char *device)
 
 	r = lvm2_run(_lvm_handle, cmd_str);
 
+	if (r == ECMD_PROCESSED) {
+		snprintf(cmd_str, CMD_SIZE, "vgreduce --removemissing %s", vg);
+		if (lvm2_run(_lvm_handle, cmd_str) != 1)
+			syslog(LOG_ERR, "Unable to remove failed PVs from VG %s", vg);
+	}
+
 	dm_pool_empty(_mem_pool);  /* FIXME: not safe with multiple threads */
-	return (r == 1) ? 0 : -1;
+	return (r == ECMD_PROCESSED) ? 0 : -1;
 }
 
-void process_event(struct dm_task *dmt, enum dm_event_mask event,
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute((unused)),
 		   void **unused __attribute((unused)))
 {
 	void *next = NULL;
@@ -222,8 +231,11 @@ void process_event(struct dm_task *dmt, enum dm_event_mask event,
 	pthread_mutex_unlock(&_event_mutex);
 }
 
-int register_device(const char *device, const char *uuid, int major, int minor,
-		   void **unused __attribute((unused)))
+int register_device(const char *device,
+		    const char *uuid __attribute((unused)),
+		    int major __attribute((unused)),
+		    int minor __attribute((unused)),
+		    void **unused __attribute((unused)))
 {
 	int r = 0;
 
@@ -259,8 +271,11 @@ out:
 	return r;
 }
 
-int unregister_device(const char *device, const char *uuid, int major, int minor,
-		   void **unused __attribute((unused)))
+int unregister_device(const char *device,
+		      const char *uuid __attribute((unused)),
+		      int major __attribute((unused)),
+		      int minor __attribute((unused)),
+		      void **unused __attribute((unused)))
 {
 	pthread_mutex_lock(&_register_mutex);
 

daemons/dmeventd/plugins/snapshot/Makefile.in

@@ -1,8 +1,8 @@
 #
 # Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
-# Copyright (C) 2004-2005 Red Hat, Inc. All rights reserved.
+# Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
 #
-# This file is part of LVM2.
+# This file is part of the LVM2.
 #
 # This copyrighted material is made available to anyone wishing to use,
 # modify, copy, or redistribute it subject to the terms and conditions
@@ -17,20 +17,23 @@ top_srcdir = @top_srcdir@
 VPATH = @srcdir@
 
 INCLUDES += -I${top_srcdir}/tools
-CLDFLAGS += -L${top_srcdir}/tools -ldevmapper -llvm2cmd
+CLDFLAGS += -L${top_srcdir}/tools -ldevmapper @LVM2CMD_LIB@
 
-SOURCES = dmeventd_mirror.c
+SOURCES = dmeventd_snapshot.c
 
 ifeq ("@LIB_SUFFIX@","dylib")
-  LIB_SHARED = libdevmapper-event-lvm2mirror.dylib
+  LIB_SHARED = libdevmapper-event-lvm2snapshot.dylib
 else
-  LIB_SHARED = libdevmapper-event-lvm2mirror.so
+  LIB_SHARED = libdevmapper-event-lvm2snapshot.so
 endif
 
+LIB_VERSION = $(LIB_VERSION_LVM)
+
 include $(top_srcdir)/make.tmpl
 
-install: libdevmapper-event-lvm2mirror.$(LIB_SUFFIX)
+install_lvm2: libdevmapper-event-lvm2snapshot.$(LIB_SUFFIX)
 	$(INSTALL) -D $(OWNER) $(GROUP) -m 555 $(STRIP) $< \
 		$(libdir)/$<.$(LIB_VERSION)
 	$(LN_S) -f $<.$(LIB_VERSION) $(libdir)/$<
 
+install: install_lvm2

daemons/dmeventd/plugins/snapshot/dmeventd_snapshot.c

@@ -0,0 +1,213 @@
+/*
+ * Copyright (C) 2007-2008 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "libdevmapper.h"
+#include "libdevmapper-event.h"
+#include "lvm2cmd.h"
+#include "lvm-string.h"
+
+#include <errno.h>
+#include <signal.h>
+#include <string.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <pthread.h>
+#include <unistd.h>
+
+#include <syslog.h> /* FIXME Replace syslog with multilog */
+/* FIXME Missing openlog? */
+
+/* First warning when snapshot is 80% full. */
+#define WARNING_THRESH 80
+/* Further warnings at 85%, 90% and 95% fullness. */
+#define WARNING_STEP 5
+
+static pthread_mutex_t _register_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+/*
+ * Number of active registrations.
+ */
+static int _register_count = 0;
+
+static struct dm_pool *_mem_pool = NULL;
+static void *_lvm_handle = NULL;
+
+struct snap_status {
+	int invalid;
+	int used;
+	int max;
+};
+
+/*
+ * Currently only one event can be processed at a time.
+ */
+static pthread_mutex_t _event_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+static void _temporary_log_fn(int level,
+			      const char *file __attribute((unused)),
+			      int line __attribute((unused)),
+			      const char *format)
+{
+	if (!strncmp(format, "WARNING: ", 9) && (level < 5))
+		syslog(LOG_CRIT, "%s", format);
+	else
+		syslog(LOG_DEBUG, "%s", format);
+}
+
+/* FIXME possibly reconcile this with target_percent when we gain
+   access to regular LVM library here. */
+static void _parse_snapshot_params(char *params, struct snap_status *stat)
+{
+	char *p;
+	/*
+	 * xx/xx	-- fractions used/max
+	 * Invalid	-- snapshot invalidated
+	 * Unknown	-- status unknown
+	 */
+	stat->used = stat->max = 0;
+
+	if (!strncmp(params, "Invalid", 7)) {
+		stat->invalid = 1;
+		return;
+	}
+
+	/*
+	 * When we return without setting non-zero max, the parent is
+	 * responsible for reporting errors.
+	 */
+	if (!strncmp(params, "Unknown", 7))
+		return;
+
+	if (!(p = strstr(params, "/")))
+		return;
+
+	*p = '\0';
+	p++;
+
+	stat->used = atoi(params);
+	stat->max = atoi(p);
+}
+
+void process_event(struct dm_task *dmt,
+		   enum dm_event_mask event __attribute((unused)),
+		   void **private)
+{
+	void *next = NULL;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
+	struct snap_status stat = { 0 };
+	const char *device = dm_task_get_name(dmt);
+	int percent, *percent_warning = (int*)private;
+
+	/* No longer monitoring, waiting for remove */
+	if (!*percent_warning)
+		return;
+
+	if (pthread_mutex_trylock(&_event_mutex)) {
+		syslog(LOG_NOTICE, "Another thread is handling an event.  Waiting...");
+		pthread_mutex_lock(&_event_mutex);
+	}
+
+	dm_get_next_target(dmt, next, &start, &length, &target_type, &params);
+	if (!target_type)
+		goto out;
+
+	_parse_snapshot_params(params, &stat);
+	/*
+	 * If the snapshot has been invalidated or we failed to parse
+	 * the status string. Report the full status string to syslog.
+	 */
+	if (stat.invalid || !stat.max) {
+		syslog(LOG_ERR, "Snapshot %s changed state to: %s\n", device, params);
+		*percent_warning = 0;
+		goto out;
+	}
+
+	percent = 100 * stat.used / stat.max;
+	if (percent >= *percent_warning) {
+		syslog(LOG_WARNING, "Snapshot %s is now %i%% full.\n", device, percent);
+		/* Print warning on the next multiple of WARNING_STEP. */
+		*percent_warning = (percent / WARNING_STEP) * WARNING_STEP + WARNING_STEP;
+	}
+out:
+	pthread_mutex_unlock(&_event_mutex);
+}
+
+int register_device(const char *device,
+		    const char *uuid __attribute((unused)),
+		    int major __attribute((unused)),
+		    int minor __attribute((unused)),
+		    void **private)
+{
+	int r = 0;
+	int *percent_warning = (int*)private;
+
+	pthread_mutex_lock(&_register_mutex);
+
+	/*
+	 * Need some space for allocations.  1024 should be more
+	 * than enough for what we need (device mapper name splitting)
+	 */
+	if (!_mem_pool && !(_mem_pool = dm_pool_create("snapshot_dso", 1024)))
+		goto out;
+
+	*percent_warning = WARNING_THRESH; /* Print warning if snapshot is full */
+
+	if (!_lvm_handle) {
+		lvm2_log_fn(_temporary_log_fn);
+		if (!(_lvm_handle = lvm2_init())) {
+			dm_pool_destroy(_mem_pool);
+			_mem_pool = NULL;
+			goto out;
+		}
+		lvm2_log_level(_lvm_handle, LVM2_LOG_SUPPRESS);
+		/* FIXME Temporary: move to dmeventd core */
+		lvm2_run(_lvm_handle, "_memlock_inc");
+	}
+
+	syslog(LOG_INFO, "Monitoring snapshot %s\n", device);
+
+	_register_count++;
+	r = 1;
+
+out:
+	pthread_mutex_unlock(&_register_mutex);
+
+	return r;
+}
+
+int unregister_device(const char *device,
+		      const char *uuid __attribute((unused)),
+		      int major __attribute((unused)),
+		      int minor __attribute((unused)),
+		      void **unused __attribute((unused)))
+{
+	pthread_mutex_lock(&_register_mutex);
+
+	syslog(LOG_INFO, "No longer monitoring snapshot %s\n",
+	       device);
+
+	if (!--_register_count) {
+		dm_pool_destroy(_mem_pool);
+		_mem_pool = NULL;
+		lvm2_run(_lvm_handle, "_memlock_dec");
+		lvm2_exit(_lvm_handle);
+		_lvm_handle = NULL;
+	}
+
+	pthread_mutex_unlock(&_register_mutex);
+
+	return 1;
+}

debian/README.Debian

@@ -1,6 +0,0 @@
-LVM2 requires the device-mapper kernel module (dm-mod).  This is
-available as a kernel patch for 2.4 (in kernel-patch-device-mapper), and
-is distributed with linux 2.5 and above.  The LVM1 kernel module (lvm-mod)
-will not work with lvm2 packages.  dm-mod and lvm-mod may both be loaded
-in the kernel at the same time with no problems.  Without dm-mod, this
-package is pretty useless.

debian/changelog

@@ -1,82 +0,0 @@
-lvm2 (1.95.15-1) unstable; urgency=low
-  
-  * New upstream release.
-  * Remove undocumented manpage symlinks.
-  * Update description to be more informative.  (Closes: #173499)
-  * Add kernel-patch-device-mapper suggestion.
-  * Update standards version.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Sun, 16 Feb 2002 04:21:26 -0400
-
-lvm2 (1.95.11-1) unstable; urgency=low
-
-  * New upstream release.  (Closes: #171436)
-  * Removed TODO and INTRO from debian/docs; added WHATS_NEW.
-  * Remove vgcfgrestore.8 undocumented symlink.
-  * Added a README.Debian, mentioning the device-mapper kernel module
-    requirement that lvm2 has.  (Closes: #171674, #163020)
-  * Get rid of debian/conffiles (debhelper's smart enough to figure that out).
-  * debian/copyright fix to appease lintian.
-  * Fix typo in tools/commands.h that caused /usr/sbin/; to be created.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Mon,  9 Dec 2002 02:51:02 -0400
-
-lvm2 (1.95.10-2) unstable; urgency=low
-
-  * Fix software raid problems by ensuring lvm init script runs after
-    raidtools init script.  (Closes: #152569)
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Tue,  3 Sep 2002 04:05:43 -0400
-
-lvm2 (1.95.10-1) unstable; urgency=low
-
-  * New upstream release (Beta 3.2).
-  * Change all references to /dev/device-mapper/control to
-    /dev/mapper/control.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Sun,  1 Sep 2002 18:55:12 -0400
-
-lvm2 (0.95.05-3) unstable; urgency=low
-
-  * Get rid of awk dependency in init script.  (Closes: #146257)
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Sun, 12 May 2002 04:39:06 -0500
-
-lvm2 (0.95.05-2) unstable; urgency=low
-
-  * Use ${shlibs:Depends} in Depends.
-  * Get rid of postinst/postrm scripts, use debhelper's init script instead.
-  * Add Conflicts against lvm10, lvm-common.
-  * Fix endian issues on big-endian machines.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Thu,  2 May 2002 23:53:53 -0500
-
-lvm2 (0.95.05-1) unstable; urgency=low
-
-  * New release (Beta2).
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Thu, 25 Apr 2002 00:37:41 -0500
-
-lvm2 (0.95.04cvs20020306-1) unstable; urgency=low
-
-  * CVS updated.
-  * Convert from debian native package.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Wed,  6 Mar 2002 00:43:21 -0500
-
-lvm2 (0.95.04cvs20020304) unstable; urgency=low
-
-  * CVS updated.
-  * Enhance init script; create devmapper control device, etc.
-  * Add dmsetup as a suggestion.
-  * Add /etc/lvm/lvm.conf conffile.
-  * Add undocumented(7) for the commands missing manpages.
-  
- -- Andres Salomon <dilinger@mp3revolution.net>  Mon,  4 Mar 2002 04:51:26 -0500
-
-lvm2 (0.95.02cvs20020220) unstable; urgency=low
-
-  * Initial Release.
-
- -- Andres Salomon <dilinger@mp3revolution.net>  Wed, 20 Feb 2002 03:17:25 -0500
-

debian/control

@@ -1,24 +0,0 @@
-Source: lvm2
-Section: admin
-Priority: optional
-Maintainer: Andres Salomon <dilinger@mp3revolution.net>
-Build-Depends: debhelper (>> 3.0.0), libdevmapper-dev (>= 0.96.04), libreadline4-dev
-Standards-Version: 3.5.8.0
-
-Package: lvm2
-Architecture: any
-Depends: ${shlibs:Depends}
-Conflicts: lvm10, lvm-common
-Replaces: lvm10, lvm-common
-Provides: lvm-binaries
-Suggests: dmsetup, kernel-patch-device-mapper
-Description: The Linux Logical Volume Manager
- This is LVM2, the rewrite of The Linux Logical Volume Manager.  LVM
- supports enterprise level volume management of disk and disk subsystems
- by grouping arbitrary disks into volume groups. The total capacity of
- volume groups can be allocated to logical volumes, which are accessed as
- regular block devices.
- .
- LVM2 is currently stable, but has some unimplemented features (most notably,
- pvmove and e2fsadm).  It is not yet recommended for production use.  It is
- backwards-compatible with LVM1 (lvm10), and requires Linux kernel 2.4.

debian/copyright

@@ -1,25 +0,0 @@
-This package was debianized by Andres Salomon <dilinger@mp3revolution.net> on
-Wed, 20 Feb 2002 03:17:25 -0500.
-
-It was downloaded from http://www.sistina.com/products_lvm.htm
-
-Upstream Author: LVM Development Team
-
-Copyright (c) 2001-2002 LVM Development Team
-
-LVM2 is free software; you can redistribute it and/or modify
-it under the terms of the GNU General Public License as published by
-the Free Software Foundation; either version 2 of the License, or
-(at your option) any later version.
-
-LVM2 is distributed in the hope that it will be useful,
-but WITHOUT ANY WARRANTY; without even the implied warranty of
-MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-GNU General Public License for more details.
-
-You should have received a copy of the GNU General Public License
-along with this program; if not, write to the Free Software
-Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA
-
-On Debian systems, the full text of the GPL can be found in
-/usr/share/common-licenses/GPL

debian/dirs

@@ -1,4 +0,0 @@
-etc/lvm
-usr/share/man/man5
-usr/share/man/man8
-sbin

debian/docs

@@ -1,5 +0,0 @@
-BUGS
-README
-VERSION
-WHATS_NEW
-doc/*

debian/init.d

@@ -1,64 +0,0 @@
-#! /bin/sh
-#
-# lvm2		This script handles LVM2 initialization/shutdown.
-#
-#		Written by Andres Salomon <dilinger@mp3revolution.net>.
-#
-
-PATH=/sbin:/bin:/usr/sbin:/usr/bin
-NAME=lvm2
-DESC=LVM
-
-test -x /sbin/vgchange || exit 0
-modprobe dm-mod >/dev/null 2>&1
-
-# Create necessary files in /dev for device-mapper
-create_devfiles() {
-	DIR="/dev/mapper"
-	FILE="$DIR/control"
-	major=$(grep "[0-9] misc$" /proc/devices | sed 's/[ ]\+misc//')
-	minor=$(grep "[0-9] device-mapper$" /proc/misc | sed 's/[ ]\+device-mapper//')
-
-	if test ! -d $DIR; then
-		mkdir --mode=755 $DIR >/dev/null 2>&1
-	fi
-
-	if test ! -c $FILE -a ! -z "$minor"; then
-		mknod --mode=600 $FILE c $major $minor >/dev/null 2>&1
-	fi
-}
-
-case "$1" in
-  start)
-	echo -n "Initializing $DESC: "
-	create_devfiles
-	vgchange -a y
-
-#	# Mount all LVM devices
-#	for vg in $( vgchange -a y 2>/dev/null | grep active | awk -F\" '{print $2}' ); do
-#		MTPT=$( grep $vg /etc/fstab | awk '{print $2}' )
-#		mount $MTPT
-#	done
-	echo "$NAME."
-	;;
-  stop)
-	echo -n "Shutting down $DESC: "
-	# We don't really try all that hard to shut it down; far too many
-	# things that can keep it from successfully shutting down.
-	vgchange -a n
-	echo "$NAME."
-	;;
-  restart|force-reload)
-	echo -n "Restarting $DESC: "
-	vgchange -a n
-	sleep 1
-	vgchange -a y
-	echo "$NAME."
-	;;
-  *)
-	echo "Usage: /etc/init.d/$NAME {start|stop|restart|force-reload}" >&2
-	exit 1
-	;;
-esac
-
-exit 0

debian/manpages

@@ -1,26 +0,0 @@
-debian/lvm2/usr/share/man/man5/lvm.conf.5
-debian/lvm2/usr/share/man/man8/lvchange.8
-debian/lvm2/usr/share/man/man8/lvcreate.8
-debian/lvm2/usr/share/man/man8/lvdisplay.8
-debian/lvm2/usr/share/man/man8/lvextend.8
-debian/lvm2/usr/share/man/man8/lvm.8
-debian/lvm2/usr/share/man/man8/lvmchange.8
-debian/lvm2/usr/share/man/man8/lvreduce.8
-debian/lvm2/usr/share/man/man8/lvremove.8
-debian/lvm2/usr/share/man/man8/lvrename.8
-debian/lvm2/usr/share/man/man8/lvscan.8
-debian/lvm2/usr/share/man/man8/pvchange.8
-debian/lvm2/usr/share/man/man8/pvcreate.8
-debian/lvm2/usr/share/man/man8/pvdisplay.8
-debian/lvm2/usr/share/man/man8/pvscan.8
-debian/lvm2/usr/share/man/man8/vgcfgbackup.8
-debian/lvm2/usr/share/man/man8/vgchange.8
-debian/lvm2/usr/share/man/man8/vgck.8
-debian/lvm2/usr/share/man/man8/vgcreate.8
-debian/lvm2/usr/share/man/man8/vgdisplay.8
-debian/lvm2/usr/share/man/man8/vgextend.8
-debian/lvm2/usr/share/man/man8/vgmerge.8
-debian/lvm2/usr/share/man/man8/vgreduce.8
-debian/lvm2/usr/share/man/man8/vgremove.8
-debian/lvm2/usr/share/man/man8/vgrename.8
-debian/lvm2/usr/share/man/man8/vgscan.8

debian/rules

@@ -1,119 +0,0 @@
-#!/usr/bin/make -f
-# Sample debian/rules that uses debhelper. 
-# GNU copyright 1997 by Joey Hess.
-#
-# This version is for a hypothetical package that builds an
-# architecture-dependant package, as well as an architecture-independent
-# package.
-
-# Uncomment this to turn on verbose mode. 
-#export DH_VERBOSE=1
-
-# This is the debhelper compatibility version to use.
-export DH_COMPAT=3
-
-# These are used for cross-compiling and for saving the configure script
-# from having to guess our platform (since we know it already)
-DEB_HOST_GNU_TYPE   ?= $(shell dpkg-architecture -qDEB_HOST_GNU_TYPE)
-DEB_BUILD_GNU_TYPE  ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_TYPE)
-
-ifneq (,$(findstring debug,$(DEB_BUILD_OPTIONS)))
-	CFLAGS += -g
-endif
-ifeq (,$(findstring nostrip,$(DEB_BUILD_OPTIONS)))
-	INSTALL_PROGRAM += -s
-endif
-
-configure: configure-stamp
-configure-stamp:
-	dh_testdir
-	# Add here commands to configure the package.
-	./configure --host=$(DEB_HOST_GNU_TYPE) --build=$(DEB_BUILD_GNU_TYPE) --prefix=/ --mandir=\$${prefix}/usr/share/man --infodir=\$${prefix}/usr/share/info
-
-	touch configure-stamp
-
-build-arch: configure-stamp build-arch-stamp
-build-arch-stamp:
-	dh_testdir
-
-	# Add here command to compile/build the package.
-	$(MAKE)
-
-	touch build-arch-stamp
-
-build-indep: configure-stamp build-indep-stamp
-build-indep-stamp:
-	dh_testdir
-
-	# Add here command to compile/build the arch indep package.
-	# It's ok not to do anything here, if you don't need to build
-	#  anything for this package.
-	#/usr/bin/docbook-to-man debian/lvm2.sgml > lvm2.1
-
-	touch build-indep-stamp
-
-build: build-arch build-indep
-
-clean:
-	dh_testdir
-	dh_testroot
-	rm -f build-stamp configure-stamp
-
-	# Add here commands to clean up after the build process.
-	-$(MAKE) distclean
-	-test -r /usr/share/misc/config.sub && \
-	  cp -f /usr/share/misc/config.sub config.sub
-	-test -r /usr/share/misc/config.guess && \
-	  cp -f /usr/share/misc/config.guess config.guess
-
-
-	dh_clean
-
-install: DH_OPTIONS=
-install: build
-	dh_testdir
-	dh_testroot
-	dh_clean -k
-	dh_installdirs
-
-	# Add here commands to install the package into debian/lvm2.
-	$(MAKE) install prefix=$(CURDIR)/debian/lvm2
-	install -m 0644 doc/example.conf debian/lvm2/etc/lvm/lvm.conf
-
-
-# Build architecture-independent files here.
-# Pass -i to all debhelper commands in this target to reduce clutter.
-binary-indep: build install
-# nada.
-
-# Build architecture-dependent files here.
-binary-arch: build install
-	dh_testdir
-	dh_testroot
-
-#	dh_installdebconf
-	dh_installdocs
-	dh_installexamples
-#	dh_installlogrotate -a
-#	dh_installemacsen -a
-#	dh_installpam -a
-#	dh_installmime -a
-	dh_installinit --update-rcd-params="start 26 S . start 50 0 6 ."
-	dh_installcron
-	dh_installman
-	dh_installinfo 
-	dh_installchangelogs
-	dh_strip
-	dh_link
-	dh_compress
-	dh_fixperms
-	dh_makeshlibs
-	dh_installdeb
-#	dh_perl -a
-	dh_shlibdeps
-	dh_gencontrol
-	dh_md5sums
-	dh_builddeb
-
-binary: binary-indep binary-arch
-.PHONY: build clean binary-indep binary-arch binary install configure

dmeventd/mirror/dmeventd_mirror.c

@@ -1,281 +0,0 @@
-/*
- * Copyright (C) 2005 Red Hat, Inc. All rights reserved.
- *
- * This file is part of LVM2.
- *
- * This copyrighted material is made available to anyone wishing to use,
- * modify, copy, or redistribute it subject to the terms and conditions
- * of the GNU Lesser General Public License v.2.1.
- *
- * You should have received a copy of the GNU Lesser General Public License
- * along with this program; if not, write to the Free Software Foundation,
- * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
- */
-
-#include "lvm2cmd.h"
-
-#include <libdevmapper.h>
-#include <libdevmapper-event.h>
-#include <errno.h>
-#include <signal.h>
-#include <string.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <pthread.h>
-#include <unistd.h>
-
-#include <syslog.h> /* FIXME Replace syslog with multilog */
-/* FIXME Missing openlog? */
-
-#define ME_IGNORE    0
-#define ME_INSYNC    1
-#define ME_FAILURE   2
-
-/*
- * register_device() is called first and performs initialisation.
- * Only one device may be registered or unregistered at a time.
- */
-static pthread_mutex_t _register_mutex = PTHREAD_MUTEX_INITIALIZER;
-
-/*
- * Number of active registrations.
- */
-static int _register_count = 0;
-
-static struct dm_pool *_mem_pool = NULL;
-static void *_lvm_handle = NULL;
-
-/*
- * Currently only one event can be processed at a time.
- */
-static pthread_mutex_t _event_mutex = PTHREAD_MUTEX_INITIALIZER;
-
-static int _get_mirror_event(char *params)
-{
-	int i, r = ME_INSYNC;
-	char **args = NULL;
-	char *dev_status_str;
-	char *log_status_str;
-	char *sync_str;
-	char *p = NULL;
-	int log_argc, num_devs;
-
-	/*
-	 * dm core parms:	     0 409600 mirror
-	 * Mirror core parms:	     2 253:4 253:5 400/400
-	 * New-style failure params: 1 AA
-	 * New-style log params:     3 cluster 253:3 A
-	 *			 or  3 disk 253:3 A
-	 *			 or  1 core
-	 */
-
-	/* number of devices */
-	if (!dm_split_words(params, 1, 0, &p))
-		goto out_parse;
-
-	if (!(num_devs = atoi(p)))
-		goto out_parse;
-	p += strlen(p) + 1;
-
-	/* devices names + "400/400" + "1 AA" + 1 or 3 log parms + NULL */
-	args = dm_malloc((num_devs + 7) * sizeof(char *));
-	if (!args || dm_split_words(p, num_devs + 7, 0, args) < num_devs + 5)
-		goto out_parse;
-
-	dev_status_str = args[2 + num_devs];
-	log_argc = atoi(args[3 + num_devs]);
-	log_status_str = args[3 + num_devs + log_argc];
-	sync_str = args[num_devs];
-
-	/* Check for bad mirror devices */
-	for (i = 0; i < num_devs; i++)
-		if (dev_status_str[i] == 'D') {
-			syslog(LOG_ERR, "Mirror device, %s, has failed.\n", args[i]);
-			r = ME_FAILURE;
-		}
-
-	/* Check for bad disk log device */
-	if (log_argc > 1 && log_status_str[0] == 'D') {
-		syslog(LOG_ERR, "Log device, %s, has failed.\n",
-		       args[2 + num_devs + log_argc]);
-		r = ME_FAILURE;
-	}
-
-	if (r == ME_FAILURE)
-		goto out;
-
-	p = strstr(sync_str, "/");
-	if (p) {
-		p[0] = '\0';
-		if (strcmp(sync_str, p+1))
-			r = ME_IGNORE;
-		p[0] = '/';
-	} else
-		goto out_parse;
-
-out:
-	if (args)
-		dm_free(args);
-	return r;
-	
-out_parse:
-	if (args)
-		dm_free(args);
-	syslog(LOG_ERR, "Unable to parse mirror status string.");
-	return ME_IGNORE;
-}
-
-static void _temporary_log_fn(int level, const char *file,
-			      int line, const char *format)
-{
-	if (!strncmp(format, "WARNING: ", 9) && (level < 5))
-		syslog(LOG_CRIT, "%s", format);
-	else
-		syslog(LOG_DEBUG, "%s", format);
-}
-
-static int _remove_failed_devices(const char *device)
-{
-	int r;
-#define CMD_SIZE 256	/* FIXME Use system restriction */
-	char cmd_str[CMD_SIZE];
-	char *vg = NULL, *lv = NULL, *layer = NULL;
-
-	if (strlen(device) > 200)  /* FIXME Use real restriction */
-		return -ENAMETOOLONG;	/* FIXME These return code distinctions are not used so remove them! */
-
-	if (!dm_split_lvm_name(_mem_pool, device, &vg, &lv, &layer)) {
-		syslog(LOG_ERR, "Unable to determine VG name from %s",
-		       device);
-		return -ENOMEM;	/* FIXME Replace with generic error return - reason for failure has already got logged */
-	}
-
-	/* FIXME Is any sanity-checking required on %s? */
-	if (CMD_SIZE <= snprintf(cmd_str, CMD_SIZE, "vgreduce --config devices{ignore_suspended_devices=1} --removemissing %s", vg)) {
-		/* this error should be caught above, but doesn't hurt to check again */
-		syslog(LOG_ERR, "Unable to form LVM command: Device name too long");
-		dm_pool_empty(_mem_pool);  /* FIXME: not safe with multiple threads */
-		return -ENAMETOOLONG; /* FIXME Replace with generic error return - reason for failure has already got logged */
-	}
-
-	r = lvm2_run(_lvm_handle, cmd_str);
-
-	dm_pool_empty(_mem_pool);  /* FIXME: not safe with multiple threads */
-	return (r == 1) ? 0 : -1;
-}
-
-void process_event(struct dm_task *dmt, enum dm_event_mask event,
-		   void **unused __attribute((unused)))
-{
-	void *next = NULL;
-	uint64_t start, length;
-	char *target_type = NULL;
-	char *params;
-	const char *device = dm_task_get_name(dmt);
-
-	if (pthread_mutex_trylock(&_event_mutex)) {
-		syslog(LOG_NOTICE, "Another thread is handling an event.  Waiting...");
-		pthread_mutex_lock(&_event_mutex);
-	}
-	do {
-		next = dm_get_next_target(dmt, next, &start, &length,
-					  &target_type, &params);
-
-		if (!target_type) {
-			syslog(LOG_INFO, "%s mapping lost.\n", device);
-			continue;
-		}
-
-		if (strcmp(target_type, "mirror")) {
-			syslog(LOG_INFO, "%s has unmirrored portion.\n", device);
-			continue;
-		}
-
-		switch(_get_mirror_event(params)) {
-		case ME_INSYNC:
-			/* FIXME: all we really know is that this
-			   _part_ of the device is in sync
-			   Also, this is not an error
-			*/
-			syslog(LOG_NOTICE, "%s is now in-sync\n", device);
-			break;
-		case ME_FAILURE:
-			syslog(LOG_ERR, "Device failure in %s\n", device);
-			if (_remove_failed_devices(device))
-				/* FIXME Why are all the error return codes unused? Get rid of them? */
-				syslog(LOG_ERR, "Failed to remove faulty devices in %s\n",
-				       device);
-			/* Should check before warning user that device is now linear
-			else
-				syslog(LOG_NOTICE, "%s is now a linear device.\n",
-					device);
-			*/
-			break;
-		case ME_IGNORE:
-			break;
-		default:
-			/* FIXME Provide value then! */
-			syslog(LOG_INFO, "Unknown event received.\n");
-		}
-	} while (next);
-
-	pthread_mutex_unlock(&_event_mutex);
-}
-
-int register_device(const char *device, const char *uuid, int major, int minor,
-		   void **unused __attribute((unused)))
-{
-	int r = 0;
-
-	pthread_mutex_lock(&_register_mutex);
-
-	syslog(LOG_INFO, "Monitoring mirror device %s for events\n", device);
-
-	/*
-	 * Need some space for allocations.  1024 should be more
-	 * than enough for what we need (device mapper name splitting)
-	 */
-	if (!_mem_pool && !(_mem_pool = dm_pool_create("mirror_dso", 1024)))
-		goto out;
-
-	if (!_lvm_handle) {
-		lvm2_log_fn(_temporary_log_fn);
-		if (!(_lvm_handle = lvm2_init())) {
-			dm_pool_destroy(_mem_pool);
-			_mem_pool = NULL;
-			goto out;
-		}
-		lvm2_log_level(_lvm_handle, LVM2_LOG_SUPPRESS);
-		/* FIXME Temporary: move to dmeventd core */
-		lvm2_run(_lvm_handle, "_memlock_inc");
-	}
-
-	_register_count++;
-	r = 1;
-
-out:
-	pthread_mutex_unlock(&_register_mutex);
-
-	return r;
-}
-
-int unregister_device(const char *device, const char *uuid, int major, int minor,
-		   void **unused __attribute((unused)))
-{
-	pthread_mutex_lock(&_register_mutex);
-
-	syslog(LOG_INFO, "No longer monitoring mirror device %s for events\n",
-	       device);
-
-	if (!--_register_count) {
-		dm_pool_destroy(_mem_pool);
-		_mem_pool = NULL;
-		lvm2_run(_lvm_handle, "_memlock_dec");
-		lvm2_exit(_lvm_handle);
-		_lvm_handle = NULL;
-	}
-
-	pthread_mutex_unlock(&_register_mutex);
-
-	return 1;
-}

doc/Makefile.in

@@ -20,10 +20,11 @@ CONFDEST=lvm.conf
 
 include $(top_srcdir)/make.tmpl
 
-install:
+install_lvm2:
 	@if [ ! -e $(confdir)/$(CONFDEST) ]; then \
 		echo "Installing $(CONFSRC) as $(confdir)/$(CONFDEST)"; \
 		@INSTALL@ -D $(OWNER) $(GROUP) -m 644 $(CONFSRC) \
 			$(confdir)/$(CONFDEST); \
 	fi
 
+install: install_lvm2

doc/example.conf

@@ -25,7 +25,8 @@ devices {
     # list of regular expressions in turn and the first match is used.
     preferred_names = [ ]
 
-    # preferred_names = [ "^/dev/mpath/", "^/dev/[hs]d" ]
+    # Try to avoid using undescriptive /dev/dm-N names, if present.
+    # preferred_names = [ "^/dev/mpath/", "^/dev/mapper/mpath", "^/dev/[hs]d" ]
 
     # A filter that tells LVM2 to only use a restricted set of devices.
     # The filter consists of an array of regular expressions.  These
@@ -85,13 +86,45 @@ devices {
     # If sysfs is mounted (2.6 kernels) restrict device scanning to 
     # the block devices it believes are valid.
     # 1 enables; 0 disables.
-    sysfs_scan = 1	
+    sysfs_scan = 1
 
     # By default, LVM2 will ignore devices used as components of
     # software RAID (md) devices by looking for md superblocks.
     # 1 enables; 0 disables.
     md_component_detection = 1
 
+    # By default, if a PV is placed directly upon an md device, LVM2
+    # will align its data blocks with the md device's stripe-width.
+    # 1 enables; 0 disables.
+    md_chunk_alignment = 1
+
+    # By default, the start of a PV's data area will be a multiple of
+    # the 'minimum_io_size' or 'optimal_io_size' exposed in sysfs.
+    # - minimum_io_size - the smallest request the device can perform
+    #   w/o incurring a read-modify-write penalty (e.g. MD's chunk size)
+    # - optimal_io_size - the device's preferred unit of receiving I/O
+    #   (e.g. MD's stripe width)
+    # minimum_io_size is used if optimal_io_size is undefined (0).
+    # If md_chunk_alignment is enabled, that detects the optimal_io_size.
+    # This setting takes precedence over md_chunk_alignment.
+    # 1 enables; 0 disables.
+    data_alignment_detection = 1
+
+    # Alignment (in KB) of start of data area when creating a new PV.
+    # If a PV is placed directly upon an md device and md_chunk_alignment or
+    # data_alignment_detection is enabled this parameter is ignored.
+    # Set to 0 for the default alignment of 64KB or page size, if larger.
+    data_alignment = 0
+
+    # By default, the start of the PV's aligned data area will be shifted by
+    # the 'alignment_offset' exposed in sysfs.  This offset is often 0 but
+    # may be non-zero; e.g.: certain 4KB sector drives that compensate for
+    # windows partitioning will have an alignment_offset of 3584 bytes
+    # (sector 7 is the lowest aligned logical block, the 4KB sectors start
+    # at LBA -1, and consequently sector 63 is aligned on a 4KB boundary).
+    # 1 enables; 0 disables.
+    data_alignment_offset_detection = 1
+
     # If, while scanning the system for PVs, LVM2 encounters a device-mapper
     # device that has its I/O suspended, it waits for it to become accessible.
     # Set this to 1 to skip such devices.  This should only be needed
@@ -123,7 +156,7 @@ log {
     # There are 6 syslog-like log levels currently in use - 2 to 7 inclusive.
     # 7 is the most verbose (LOG_DEBUG).
     level = 0
-    
+
     # Format of output messages
     # Whether or not (1 or 0) to indent messages according to their severity
     indent = 1
@@ -169,7 +202,7 @@ backup {
     # Where should archived files go ?
     # Remember to back up this directory regularly!
     archive_dir = "/etc/lvm/archive"
-    
+
     # What is the minimum number of archive files you wish to keep ?
     retain_min = 10
 
@@ -187,7 +220,7 @@ shell {
 
 # Miscellaneous global LVM2 settings
 global {
-    
+
     # The file creation mask for any files and directories created.
     # Interpreted as octal if the first digit is zero.
     umask = 077
@@ -235,6 +268,9 @@ global {
     # Type 3 uses built-in clustered locking.
     locking_type = 1
 
+    # Set to 0 to fail when a lock request cannot be satisfied immediately.
+    wait_for_locks = 1
+
     # If using external locking (type 2) and initialisation fails,
     # with this set to 1 an attempt will be made to use the built-in
     # clustered locking.
@@ -265,11 +301,23 @@ global {
 }
 
 activation {
-    # Device used in place of missing stripes if activating incomplete volume.
-    # For now, you need to set this up yourself first (e.g. with 'dmsetup')
-    # For example, you could make it return I/O errors using the 'error' 
-    # target or make it return zeros.
-    missing_stripe_filler = "/dev/ioerror"
+    # Set to 0 to disable udev syncronisation (if compiled into the binaries).
+    # Processes will not wait for notification from udev.
+    # They will continue irrespective of any possible udev processing
+    # in the background.  You should only use this if udev is not running
+    # or has rules that ignore the devices LVM2 creates.
+    # The command line argument --nodevsync takes precedence over this setting.
+    # If set to 1 when udev is not running, and there are LVM2 processes
+    # waiting for udev, run 'dmsetup udevcomplete' manually to wake them up.
+    udev_sync = 1
+
+    # How to fill in missing stripes if activating an incomplete volume.
+    # Using "error" will make inaccessible parts of the device return
+    # I/O errors on access.  You can instead use a device path, in which 
+    # case, that device will be used to in place of missing stripes.
+    # But note that using anything other than "error" with mirrored 
+    # or snapshotted volumes is likely to result in data corruption.
+    missing_stripe_filler = "error"
 
     # How much stack (in KB) to reserve for use while devices suspended
     reserved_stack = 256
@@ -303,8 +351,10 @@ activation {
     # A disk log ensures that a mirror does not need to be re-synced
     # (all copies made the same) every time a machine reboots or crashes.
     #
-    # In the event of a failure, the specified policy will be used to
-    # determine what happens:
+    # In the event of a failure, the specified policy will be used to determine
+    # what happens. This applies to automatic repairs (when the mirror is being
+    # monitored by dmeventd) and to manual lvconvert --repair when
+    # --use-policies is given.
     #
     # "remove" - Simply remove the faulty device and run without it.  If
     #            the log device fails, the mirror would convert to using
@@ -324,20 +374,13 @@ activation {
     #            will preserve the mirror characteristic of the device.
     #            This policy acts like "remove" if no suitable device and
     #            space can be allocated for the replacement.
-    #            Currently this is not implemented properly and behaves
-    #            similarly to:
     #
-    # "allocate_anywhere" - Operates like "allocate", but it does not
-    #            require that the new space being allocated be on a
-    #            device is not part of the mirror.  For a log device
-    #            failure, this could mean that the log is allocated on
-    #            the same device as a mirror device.  For a mirror
-    #            device, this could mean that the mirror device is
-    #            allocated on the same device as another mirror device.
-    #            This policy would not be wise for mirror devices
-    #            because it would break the redundant nature of the
-    #            mirror.  This policy acts like "remove" if no suitable
-    #            device and space can be allocated for the replacement.
+    # "allocate_anywhere" - Not yet implemented. Useful to place the log device
+    #            temporarily on same physical volume as one of the mirror
+    #            images. This policy is not recommended for mirror devices
+    #            since it would break the redundant nature of the mirror. This
+    #            policy acts like "remove" if no suitable device and space can
+    #            be allocated for the replacement.
 
     mirror_log_fault_policy = "allocate"
     mirror_device_fault_policy = "remove"
@@ -384,10 +427,20 @@ activation {
 # dmeventd {
     # mirror_library is the library used when monitoring a mirror device.
     #
-    # "libdevmapper-event-lvm2mirror.so" attempts to recover from failures.
-    # It removes failed devices from a volume group and reconfigures a
-    # mirror as necessary.
-    #
+    # "libdevmapper-event-lvm2mirror.so" attempts to recover from
+    # failures.  It removes failed devices from a volume group and
+    # reconfigures a mirror as necessary. If no mirror library is
+    # provided, mirrors are not monitored through dmeventd.
+
     # mirror_library = "libdevmapper-event-lvm2mirror.so"
+
+    # snapshot_library is the library used when monitoring a snapshot device.
+    #
+    # "libdevmapper-event-lvm2snapshot.so" monitors the filling of
+    # snapshots and emits a warning through syslog, when the use of
+    # snapshot exceedes 80%. The warning is repeated when 85%, 90% and
+    # 95% of the snapshot are filled.
+
+    # snapshot_library = "libdevmapper-event-lvm2snapshot.so"
 #}
 

doc/example_cmdlib.c

@@ -15,7 +15,8 @@
 #include "lvm2cmd.h"
 
 /* All output gets passed to this function line-by-line */
-void test_log_fn(int level, const char *file, int line, const char *format)
+void test_log_fn(int level, int dm_errno, const char *file, int line,
+		 const char *format)
 {
 	/* Extract and process output here rather than printing it */
 

include/.symlinks

@@ -1,4 +1,6 @@
 ../daemons/clvmd/clvm.h
+../daemons/dmeventd/libdevmapper-event.h
+../liblvm/lvm2app.h
 ../lib/activate/activate.h
 ../lib/activate/targets.h
 ../lib/cache/lvmcache.h
@@ -7,7 +9,6 @@
 ../lib/config/config.h
 ../lib/config/defaults.h
 ../lib/datastruct/btree.h
-../lib/datastruct/list.h
 ../lib/datastruct/lvm-types.h
 ../lib/datastruct/str_list.h
 ../lib/device/dev-cache.h
@@ -28,6 +29,7 @@
 ../lib/label/label.h
 ../lib/locking/locking.h
 ../lib/log/log.h
+../lib/log/lvm-logging.h
 ../lib/metadata/lv_alloc.h
 ../lib/metadata/metadata.h
 ../lib/metadata/metadata-exported.h
@@ -43,10 +45,16 @@
 ../lib/misc/lib.h
 ../lib/misc/lvm-exec.h
 ../lib/misc/lvm-file.h
+../lib/misc/lvm-globals.h
 ../lib/misc/lvm-string.h
+../lib/misc/lvm-version.h
 ../lib/misc/lvm-wrappers.h
 ../lib/misc/sharedlib.h
 ../lib/report/report.h
 ../lib/uuid/uuid.h
+../libdm/libdevmapper.h
+../libdm/misc/dm-ioctl.h
+../libdm/misc/dm-logging.h
+../libdm/misc/dmlib.h
+../libdm/misc/kdev_t.h
 ../po/pogen.h
-../tools/version.h

include/Makefile.in

@@ -20,7 +20,7 @@ VPATH = @srcdir@
 
 LN_S = @LN_S@
 
-.PHONY: clean distclean all install pofile install_cluster
+.PHONY: clean distclean all install pofile install_cluster install_device-mapper
 
 all: .symlinks_created
 
@@ -35,11 +35,17 @@ distclean:
 
 pofile: all
 
+device-mapper: all
+
 clean:
 
 install:
 
 install_cluster:
 
+install_device-mapper:
+
+install_lvm2:
+
 cflow:
 

lib/Makefile.in

@@ -1,6 +1,6 @@
 #
 # Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
-# Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+# Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
 #
 # This file is part of LVM2.
 #
@@ -38,11 +38,11 @@ SOURCES =\
 	commands/toolcontext.c \
 	config/config.c \
 	datastruct/btree.c \
-	datastruct/list.c \
 	datastruct/str_list.c \
 	device/dev-cache.c \
 	device/dev-io.c \
 	device/dev-md.c \
+	device/dev-swap.c \
 	device/device.c \
 	display/display.c \
 	error/errseg.c \
@@ -61,6 +61,7 @@ SOURCES =\
 	format_text/import_vsn1.c \
 	format_text/tags.c \
 	format_text/text_label.c \
+	freeseg/freeseg.c \
 	label/label.c \
 	locking/file_locking.c \
 	locking/locking.c \
@@ -77,6 +78,7 @@ SOURCES =\
 	misc/crc.c \
 	misc/lvm-exec.c \
 	misc/lvm-file.c \
+	misc/lvm-globals.c \
 	misc/lvm-string.c \
 	misc/lvm-wrappers.c \
 	misc/timestamp.c \
@@ -138,18 +140,19 @@ ifeq ("@DMEVENTD@", "yes")
   CLDFLAGS += -ldevmapper-event
 endif
 
-LIB_STATIC = liblvm.a
+LIB_NAME = liblvm-internal
+LIB_STATIC = $(LIB_NAME).a
 
-$(SUBDIRS): $(LIB_STATIC)
-
-CLEAN_TARGETS += liblvm.cflow
+CLEAN_TARGETS += $(LIB_NAME).cflow
 
 include $(top_srcdir)/make.tmpl
 
-liblvm.cflow: $(SOURCES)
+$(SUBDIRS): $(LIB_STATIC)
+
+$(LIB_NAME).cflow: $(SOURCES)
 	set -e; (echo -n "SOURCES += "; \
 		 echo $(SOURCES) | \
 		 sed "s/^/ /;s/ / $(top_srcdir)\/lib\//g;s/$$//"; \
 		 ) > $@
 
-cflow: liblvm.cflow
+cflow: $(LIB_NAME).cflow

lib/activate/activate.c

@@ -1,6 +1,6 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -52,11 +52,11 @@ int lvm1_present(struct cmd_context *cmd)
 }
 
 int list_segment_modules(struct dm_pool *mem, const struct lv_segment *seg,
-			 struct list *modules)
+			 struct dm_list *modules)
 {
 	unsigned int s;
 	struct lv_segment *seg2, *snap_seg;
-	struct list *snh;
+	struct dm_list *snh;
 
 	if (seg->segtype->ops->modules_needed &&
 	    !seg->segtype->ops->modules_needed(mem, seg, modules)) {
@@ -65,9 +65,9 @@ int list_segment_modules(struct dm_pool *mem, const struct lv_segment *seg,
 	}
 
 	if (lv_is_origin(seg->lv))
-		list_iterate(snh, &seg->lv->snapshot_segs)
+		dm_list_iterate(snh, &seg->lv->snapshot_segs)
 			if (!list_lv_modules(mem,
-					     list_struct_base(snh,
+					     dm_list_struct_base(snh,
 							      struct lv_segment,
 							      origin_list)->cow,
 					     modules))
@@ -100,11 +100,11 @@ int list_segment_modules(struct dm_pool *mem, const struct lv_segment *seg,
 }
 
 int list_lv_modules(struct dm_pool *mem, const struct logical_volume *lv,
-		    struct list *modules)
+		    struct dm_list *modules)
 {
 	struct lv_segment *seg;
 
-	list_iterate_items(seg, &lv->segments)
+	dm_list_iterate_items(seg, &lv->segments)
 		if (!list_segment_modules(mem, seg, modules))
 			return_0;
 
@@ -141,7 +141,8 @@ int target_version(const char *target_name, uint32_t *maj,
 {
 	return 0;
 }
-int target_present(const char *target_name, int use_modprobe)
+int target_present(struct cmd_context *cmd, const char *target_name,
+		   int use_modprobe)
 {
 	return 0;
 }
@@ -264,7 +265,7 @@ static int _passes_activation_filter(struct cmd_context *cmd,
 
 	if (!(cn = find_config_tree_node(cmd, "activation/volume_list"))) {
 		/* If no host tags defined, activate */
-		if (list_empty(&cmd->tags))
+		if (dm_list_empty(&cmd->tags))
 			return 1;
 
 		/* If any host tag matches any LV or VG tag, activate */
@@ -391,12 +392,32 @@ int target_version(const char *target_name, uint32_t *maj,
 	return r;
 }
 
-int target_present(const char *target_name, int use_modprobe)
+int module_present(struct cmd_context *cmd, const char *target_name)
 {
-	uint32_t maj, min, patchlevel;
+	int ret = 0;
 #ifdef MODPROBE_CMD
 	char module[128];
+	const char *argv[3];
+
+	if (dm_snprintf(module, sizeof(module), "dm-%s", target_name) < 0) {
+		log_error("module_present module name too long: %s",
+			  target_name);
+		return 0;
+	}
+
+	argv[0] = MODPROBE_CMD;
+	argv[1] = module;
+	argv[2] = NULL;
+
+	ret = exec_cmd(cmd, argv);
 #endif
+	return ret;
+}
+
+int target_present(struct cmd_context *cmd, const char *target_name,
+		   int use_modprobe)
+{
+	uint32_t maj, min, patchlevel;
 
 	if (!activation())
 		return 0;
@@ -406,14 +427,7 @@ int target_present(const char *target_name, int use_modprobe)
 		if (target_version(target_name, &maj, &min, &patchlevel))
 			return 1;
 
-		if (dm_snprintf(module, sizeof(module), "dm-%s", target_name)
-		    < 0) {
-			log_error("target_present module name too long: %s",
-				  target_name);
-			return 0;
-		}
-
-		if (!exec_cmd(MODPROBE_CMD, module, "", ""))
+		if (!module_present(cmd, target_name))
 			return_0;
 	}
 #endif
@@ -508,6 +522,13 @@ int lv_mirror_percent(struct cmd_context *cmd, struct logical_volume *lv,
 	struct dev_manager *dm;
 	struct lvinfo info;
 
+	/* If mirrored LV is temporarily shrinked to 1 area (= linear),
+	 * it should be considered in-sync. */
+	if (dm_list_size(&lv->segments) == 1 && first_seg(lv)->area_count == 1) {
+		*percent = 100.0;
+		return 1;
+	}
+
 	if (!activation())
 		return 0;
 
@@ -568,7 +589,7 @@ static int _lv_activate_lv(struct logical_volume *lv)
 	return r;
 }
 
-static int _lv_preload(struct logical_volume *lv)
+static int _lv_preload(struct logical_volume *lv, int *flush_required)
 {
 	int r;
 	struct dev_manager *dm;
@@ -576,7 +597,7 @@ static int _lv_preload(struct logical_volume *lv)
 	if (!(dm = dev_manager_create(lv->vg->cmd, lv->vg->name)))
 		return_0;
 
-	if (!(r = dev_manager_preload(dm, lv)))
+	if (!(r = dev_manager_preload(dm, lv, flush_required)))
 		stack;
 
 	dev_manager_destroy(dm);
@@ -598,7 +619,7 @@ static int _lv_deactivate(struct logical_volume *lv)
 	return r;
 }
 
-static int _lv_suspend_lv(struct logical_volume *lv, int lockfs)
+static int _lv_suspend_lv(struct logical_volume *lv, int lockfs, int flush_required)
 {
 	int r;
 	struct dev_manager *dm;
@@ -606,7 +627,7 @@ static int _lv_suspend_lv(struct logical_volume *lv, int lockfs)
 	if (!(dm = dev_manager_create(lv->vg->cmd, lv->vg->name)))
 		return_0;
 
-	if (!(r = dev_manager_suspend(dm, lv, lockfs)))
+	if (!(r = dev_manager_suspend(dm, lv, lockfs, flush_required)))
 		stack;
 
 	dev_manager_destroy(dm);
@@ -625,8 +646,8 @@ static int _lvs_in_vg_activated(struct volume_group *vg, unsigned by_uuid_only)
 	if (!activation())
 		return 0;
 
-	list_iterate_items(lvl, &vg->lvs) {
-		if (lvl->lv->status & VISIBLE_LV)
+	dm_list_iterate_items(lvl, &vg->lvs) {
+		if (lv_is_visible(lvl->lv))
 			count += (_lv_active(vg->cmd, lvl->lv, by_uuid_only) == 1);
 	}
 
@@ -651,14 +672,49 @@ int lvs_in_vg_opened(const struct volume_group *vg)
 	if (!activation())
 		return 0;
 
-	list_iterate_items(lvl, &vg->lvs) {
-		if (lvl->lv->status & VISIBLE_LV)
+	dm_list_iterate_items(lvl, &vg->lvs) {
+		if (lv_is_visible(lvl->lv))
 			count += (_lv_open_count(vg->cmd, lvl->lv) > 0);
 	}
 
 	return count;
 }
 
+/*
+ * Determine whether an LV is active locally or in a cluster.
+ * Assumes vg lock held.
+ * Returns:
+ * 0 - not active locally or on any node in cluster
+ * 1 - active either locally or some node in the cluster
+ */
+int lv_is_active(struct logical_volume *lv)
+{
+	int ret;
+
+	if (_lv_active(lv->vg->cmd, lv, 0))
+		return 1;
+
+	if (!vg_is_clustered(lv->vg))
+		return 0;
+
+	if ((ret = remote_lock_held(lv->lvid.s)) >= 0)
+		return ret;
+
+	/*
+	 * Old compatibility code if locking doesn't support lock query
+	 * FIXME: check status to not deactivate already activate device
+	 */
+	if (activate_lv_excl(lv->vg->cmd, lv)) {
+		deactivate_lv(lv->vg->cmd, lv);
+		return 0;
+	}
+
+	/*
+	 * Exclusive local activation failed so assume it is active elsewhere.
+	 */
+	return 1;
+}
+
 /*
  * Returns 0 if an attempt to (un)monitor the device failed.
  * Returns 1 otherwise.
@@ -669,9 +725,10 @@ int monitor_dev_for_events(struct cmd_context *cmd,
 #ifdef DMEVENTD
 	int i, pending = 0, monitored;
 	int r = 1;
-	struct list *tmp;
+	struct dm_list *tmp, *snh, *snht;
 	struct lv_segment *seg;
 	int (*monitor_fn) (struct lv_segment *s, int e);
+	uint32_t s;
 
 	/* skip dmeventd code altogether */
 	if (dmeventd_monitor_mode() == DMEVENTD_MONITOR_IGNORE)
@@ -683,8 +740,43 @@ int monitor_dev_for_events(struct cmd_context *cmd,
 	if (monitor && !dmeventd_monitor_mode())
 		return 1;
 
-	list_iterate(tmp, &lv->segments) {
-		seg = list_item(tmp, struct lv_segment);
+	/*
+	 * In case of a snapshot device, we monitor lv->snapshot->lv,
+	 * not the actual LV itself.
+	 */
+	if (lv_is_cow(lv))
+		return monitor_dev_for_events(cmd, lv->snapshot->lv, monitor);
+
+	/*
+	 * In case this LV is a snapshot origin, we instead monitor
+	 * each of its respective snapshots (the origin itself does
+	 * not need to be monitored).
+	 *
+	 * TODO: This may change when snapshots of mirrors are allowed.
+	 */
+	if (lv_is_origin(lv)) {
+		dm_list_iterate_safe(snh, snht, &lv->snapshot_segs)
+			if (!monitor_dev_for_events(cmd, dm_list_struct_base(snh,
+				    struct lv_segment, origin_list)->cow, monitor))
+				r = 0;
+		return r;
+	}
+
+	dm_list_iterate(tmp, &lv->segments) {
+		seg = dm_list_item(tmp, struct lv_segment);
+
+		/* Recurse for AREA_LV */
+		for (s = 0; s < seg->area_count; s++) {
+			if (seg_type(seg, s) != AREA_LV)
+				continue;
+			if (!monitor_dev_for_events(cmd, seg_lv(seg, s),
+						    monitor)) {
+				log_error("Failed to %smonitor %s",
+					  monitor ? "" : "un",
+					  seg_lv(seg, s)->name);
+				r = 0;
+			}
+		}
 
 		if (!seg_monitored(seg) || (seg->status & PVMOVE))
 			continue;
@@ -754,36 +846,41 @@ int monitor_dev_for_events(struct cmd_context *cmd,
 static int _lv_suspend(struct cmd_context *cmd, const char *lvid_s,
 		       int error_if_not_suspended)
 {
-	struct logical_volume *lv, *lv_pre;
+	struct logical_volume *lv = NULL, *lv_pre = NULL;
 	struct lvinfo info;
-	int lockfs = 0;
+	int r = 0, lockfs = 0, flush_required = 0;
 
 	if (!activation())
 		return 1;
 
 	if (!(lv = lv_from_lvid(cmd, lvid_s, 0)))
-		return_0;
+		goto_out;
 
 	/* Use precommitted metadata if present */
 	if (!(lv_pre = lv_from_lvid(cmd, lvid_s, 1)))
-		return_0;
+		goto_out;
 
 	if (test_mode()) {
 		_skip("Suspending '%s'.", lv->name);
-		return 1;
+		r = 1;
+		goto out;
 	}
 
 	if (!lv_info(cmd, lv, &info, 0, 0))
-		return_0;
+		goto_out;
 
-	if (!info.exists || info.suspended)
-		return error_if_not_suspended ? 0 : 1;
+	if (!info.exists || info.suspended) {
+		r = error_if_not_suspended ? 0 : 1;
+		goto out;
+	}
+
+	lv_calculate_readahead(lv, NULL);
 
 	/* If VG was precommitted, preload devices for the LV */
 	if ((lv_pre->vg->status & PRECOMMITTED)) {
-		if (!_lv_preload(lv_pre)) {
+		if (!_lv_preload(lv_pre, &flush_required)) {
 			/* FIXME Revert preloading */
-			return_0;
+			goto_out;
 		}
 	}
 
@@ -796,13 +893,20 @@ static int _lv_suspend(struct cmd_context *cmd, const char *lvid_s,
 	if (lv_is_origin(lv_pre) || lv_is_cow(lv_pre))
 		lockfs = 1;
 
-	if (!_lv_suspend_lv(lv, lockfs)) {
+	if (!_lv_suspend_lv(lv, lockfs, flush_required)) {
 		memlock_dec();
 		fs_unlock();
-		return 0;
+		goto out;
 	}
 
-	return 1;
+	r = 1;
+out:
+	if (lv_pre)
+		vg_release(lv_pre->vg);
+	if (lv)
+		vg_release(lv->vg);
+
+	return r;
 }
 
 /* Returns success if the device is not active */
@@ -821,26 +925,30 @@ static int _lv_resume(struct cmd_context *cmd, const char *lvid_s,
 {
 	struct logical_volume *lv;
 	struct lvinfo info;
+	int r = 0;
 
 	if (!activation())
 		return 1;
 
 	if (!(lv = lv_from_lvid(cmd, lvid_s, 0)))
-		return 0;
+		goto out;
 
 	if (test_mode()) {
 		_skip("Resuming '%s'.", lv->name);
-		return 1;
+		r = 1;
+		goto out;
 	}
 
 	if (!lv_info(cmd, lv, &info, 0, 0))
-		return_0;
+		goto_out;
 
-	if (!info.exists || !info.suspended)
-		return error_if_not_active ? 0 : 1;
+	if (!info.exists || !info.suspended) {
+		r = error_if_not_active ? 0 : 1;
+		goto out;
+	}
 
 	if (!_lv_activate_lv(lv))
-		return 0;
+		goto out;
 
 	memlock_dec();
 	fs_unlock();
@@ -848,7 +956,12 @@ static int _lv_resume(struct cmd_context *cmd, const char *lvid_s,
 	if (!monitor_dev_for_events(cmd, lv, 1))
 		stack;
 
-	return 1;
+	r = 1;
+out:
+	if (lv)
+		vg_release(lv->vg);
+
+	return r;
 }
 
 /* Returns success if the device is not active */
@@ -866,31 +979,36 @@ int lv_deactivate(struct cmd_context *cmd, const char *lvid_s)
 {
 	struct logical_volume *lv;
 	struct lvinfo info;
-	int r;
+	int r = 0;
 
 	if (!activation())
 		return 1;
 
 	if (!(lv = lv_from_lvid(cmd, lvid_s, 0)))
-		return 0;
+		goto out;
 
 	if (test_mode()) {
 		_skip("Deactivating '%s'.", lv->name);
-		return 1;
+		r = 1;
+		goto out;
 	}
 
 	if (!lv_info(cmd, lv, &info, 1, 0))
-		return_0;
+		goto_out;
 
-	if (!info.exists)
-		return 1;
+	if (!info.exists) {
+		r = 1;
+		goto out;
+	}
 
-	if (info.open_count && (lv->status & VISIBLE_LV)) {
+	if (info.open_count && lv_is_visible(lv)) {
 		log_error("LV %s/%s in use: not deactivating", lv->vg->name,
 			  lv->name);
-		return 0;
+		goto out;
 	}
 
+	lv_calculate_readahead(lv, NULL);
+
 	if (!monitor_dev_for_events(cmd, lv, 0))
 		stack;
 
@@ -899,6 +1017,10 @@ int lv_deactivate(struct cmd_context *cmd, const char *lvid_s)
 	memlock_dec();
 	fs_unlock();
 
+out:
+	if (lv)
+		vg_release(lv->vg);
+
 	return r;
 }
 
@@ -907,23 +1029,28 @@ int lv_activation_filter(struct cmd_context *cmd, const char *lvid_s,
 			 int *activate_lv)
 {
 	struct logical_volume *lv;
+	int r = 0;
 
-	if (!activation())
-		goto activate;
+	if (!activation()) {
+		*activate_lv = 1;
+		return 1;
+	}
 
 	if (!(lv = lv_from_lvid(cmd, lvid_s, 0)))
-		return 0;
+		goto out;
 
 	if (!_passes_activation_filter(cmd, lv)) {
 		log_verbose("Not activating %s/%s due to config file settings",
 			    lv->vg->name, lv->name);
 		*activate_lv = 0;
-		return 1;
-	}
+	} else
+		*activate_lv = 1;
+	r = 1;
+out:
+	if (lv)
+		vg_release(lv->vg);
 
-      activate:
-	*activate_lv = 1;
-	return 1;
+	return r;
 }
 
 static int _lv_activate(struct cmd_context *cmd, const char *lvid_s,
@@ -931,30 +1058,41 @@ static int _lv_activate(struct cmd_context *cmd, const char *lvid_s,
 {
 	struct logical_volume *lv;
 	struct lvinfo info;
-	int r;
+	int r = 0;
 
 	if (!activation())
 		return 1;
 
 	if (!(lv = lv_from_lvid(cmd, lvid_s, 0)))
-		return 0;
+		goto out;
 
 	if (filter && !_passes_activation_filter(cmd, lv)) {
 		log_verbose("Not activating %s/%s due to config file settings",
 			    lv->vg->name, lv->name);
-		return 0;
+		goto out;
+	}
+
+	if ((!lv->vg->cmd->partial_activation) && (lv->status & PARTIAL_LV)) {
+		log_error("Refusing activation of partial LV %s. Use --partial to override.",
+			  lv->name);
+		goto_out;
 	}
 
 	if (test_mode()) {
 		_skip("Activating '%s'.", lv->name);
-		return 1;
+		r = 1;
+		goto out;
 	}
 
 	if (!lv_info(cmd, lv, &info, 0, 0))
-		return_0;
+		goto_out;
 
-	if (info.exists && !info.suspended && info.live_table)
-		return 1;
+	if (info.exists && !info.suspended && info.live_table) {
+		r = 1;
+		goto out;
+	}
+
+	lv_calculate_readahead(lv, NULL);
 
 	if (exclusive)
 		lv->status |= ACTIVATE_EXCL;
@@ -967,6 +1105,10 @@ static int _lv_activate(struct cmd_context *cmd, const char *lvid_s,
 	if (r && !monitor_dev_for_events(cmd, lv, 1))
 		stack;
 
+out:
+	if (lv)
+		vg_release(lv->vg);
+
 	return r;
 }
 
@@ -996,9 +1138,10 @@ int lv_mknodes(struct cmd_context *cmd, const struct logical_volume *lv)
 	if (!_lv_info(cmd, lv, 1, &info, 0, 0, 0))
 		return_0;
 
-	if (info.exists)
-		r = dev_manager_lv_mknodes(lv);
-	else
+	if (info.exists) {
+		if (lv_is_visible(lv))
+			r = dev_manager_lv_mknodes(lv);
+	} else
 		r = dev_manager_lv_rmnodes(lv);
 
 	fs_unlock();

lib/activate/activate.h

@@ -30,6 +30,9 @@ struct lvinfo {
 	uint32_t read_ahead;
 };
 
+/* target attribute flags */
+#define MIRROR_LOG_CLUSTERED	0x00000001U
+
 void set_activation(int activation);
 int activation(void);
 
@@ -37,13 +40,15 @@ int driver_version(char *version, size_t size);
 int library_version(char *version, size_t size);
 int lvm1_present(struct cmd_context *cmd);
 
-int target_present(const char *target_name, int use_modprobe);
+int module_present(struct cmd_context *cmd, const char *target_name);
+int target_present(struct cmd_context *cmd, const char *target_name,
+		   int use_modprobe);
 int target_version(const char *target_name, uint32_t *maj,
                    uint32_t *min, uint32_t *patchlevel);
 int list_segment_modules(struct dm_pool *mem, const struct lv_segment *seg,
-			 struct list *modules);
+			 struct dm_list *modules);
 int list_lv_modules(struct dm_pool *mem, const struct logical_volume *lv,
-		    struct list *modules);
+		    struct dm_list *modules);
 
 void activation_release(void);
 void activation_exit(void);
@@ -87,6 +92,7 @@ int lvs_in_vg_activated(struct volume_group *vg);
 int lvs_in_vg_activated_by_uuid_only(struct volume_group *vg);
 int lvs_in_vg_opened(const struct volume_group *vg);
 
+int lv_is_active(struct logical_volume *lv);
 
 int monitor_dev_for_events(struct cmd_context *cmd,
 			    struct logical_volume *lv, int do_reg);

lib/activate/dev_manager.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2002-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -47,9 +47,9 @@ struct dev_manager {
 
 	struct cmd_context *cmd;
 
-	const char *stripe_filler;
 	void *target_state;
 	uint32_t pvmove_mirror_count;
+	int flush_required;
 
 	char *vg_name;
 };
@@ -59,8 +59,6 @@ struct lv_layer {
 	const char *old_name;
 };
 
-static const char *stripe_filler = NULL;
-
 static char *_build_dlid(struct dm_pool *mem, const char *lvid, const char *layer)
 {
 	char *dlid;
@@ -96,14 +94,13 @@ static int _read_only_lv(struct logical_volume *lv)
  * Low level device-layer operations.
  */
 static struct dm_task *_setup_task(const char *name, const char *uuid,
-				   uint32_t *event_nr, int task)
+				   uint32_t *event_nr, int task,
+				   uint32_t major, uint32_t minor)
 {
 	struct dm_task *dmt;
 
-	if (!(dmt = dm_task_create(task))) {
-		stack;
-		return NULL;
-	}
+	if (!(dmt = dm_task_create(task)))
+		return_NULL;
 
 	if (name)
 		dm_task_set_name(dmt, name);
@@ -114,12 +111,15 @@ static struct dm_task *_setup_task(const char *name, const char *uuid,
 	if (event_nr)
 		dm_task_set_event_nr(dmt, *event_nr);
 
+	if (major)
+		dm_task_set_major_minor(dmt, major, minor, 1);
+
 	return dmt;
 }
 
 static int _info_run(const char *name, const char *dlid, struct dm_info *info,
 		     uint32_t *read_ahead, int mknodes, int with_open_count,
-		     int with_read_ahead)
+		     int with_read_ahead, uint32_t major, uint32_t minor)
 {
 	int r = 0;
 	struct dm_task *dmt;
@@ -127,10 +127,8 @@ static int _info_run(const char *name, const char *dlid, struct dm_info *info,
 
 	dmtask = mknodes ? DM_DEVICE_MKNODES : DM_DEVICE_INFO;
 
-	if (!(dmt = _setup_task(name, dlid, 0, dmtask))) {
-		stack;
-		return 0;
-	}
+	if (!(dmt = _setup_task(name, dlid, 0, dmtask, major, minor)))
+		return_0;
 
 	if (!with_open_count)
 		if (!dm_task_no_open_count(dmt))
@@ -142,7 +140,7 @@ static int _info_run(const char *name, const char *dlid, struct dm_info *info,
 	if (!dm_task_get_info(dmt, info))
 		goto_out;
 
-	if (with_read_ahead) {
+	if (with_read_ahead && info->exists) {
 		if (!dm_task_get_read_ahead(dmt, read_ahead))
 			goto_out;
 	} else if (read_ahead)
@@ -160,9 +158,9 @@ int device_is_usable(dev_t dev)
 	struct dm_task *dmt;
 	struct dm_info info;
 	const char *name;
-        uint64_t start, length;
-        char *target_type = NULL;
-        char *params;
+	uint64_t start, length;
+	char *target_type = NULL;
+	char *params;
 	void *next = NULL;
 	int r = 0;
 
@@ -171,7 +169,7 @@ int device_is_usable(dev_t dev)
 		return 0;
 	}
 
-	if (!dm_task_set_major(dmt, MAJOR(dev)) || !dm_task_set_minor(dmt, MINOR(dev)))
+	if (!dm_task_set_major_minor(dmt, MAJOR(dev), MINOR(dev), 1))
 		goto_out;
 
 	if (!dm_task_run(dmt)) {
@@ -190,13 +188,13 @@ int device_is_usable(dev_t dev)
 	/* FIXME Also check for mirror block_on_error and mpath no paths */
 	/* For now, we exclude all mirrors */
 
-        do {
-                next = dm_get_next_target(dmt, next, &start, &length,
-                                          &target_type, &params);
-                /* Skip if target type doesn't match */
-                if (target_type && !strcmp(target_type, "mirror"))
+	do {
+		next = dm_get_next_target(dmt, next, &start, &length,
+					  &target_type, &params);
+		/* Skip if target type doesn't match */
+		if (target_type && !strcmp(target_type, "mirror"))
 			goto out;
-        } while (next);
+	} while (next);
 
 	/* FIXME Also check dependencies? */
 
@@ -213,23 +211,28 @@ static int _info(const char *name, const char *dlid, int mknodes,
 {
 	if (!mknodes && dlid && *dlid) {
 		if (_info_run(NULL, dlid, info, read_ahead, 0, with_open_count,
-			      with_read_ahead) &&
+			      with_read_ahead, 0, 0) &&
 	    	    info->exists)
 			return 1;
 		else if (_info_run(NULL, dlid + sizeof(UUID_PREFIX) - 1, info,
 				   read_ahead, 0, with_open_count,
-				   with_read_ahead) &&
+				   with_read_ahead, 0, 0) &&
 			 info->exists)
 			return 1;
 	}
 
 	if (name)
 		return _info_run(name, NULL, info, read_ahead, mknodes,
-				 with_open_count, with_read_ahead);
+				 with_open_count, with_read_ahead, 0, 0);
 
 	return 0;
 }
 
+static int _info_by_dev(uint32_t major, uint32_t minor, struct dm_info *info)
+{
+	return _info_run(NULL, NULL, info, NULL, 0, 0, 0, major, minor);
+}
+
 int dev_manager_info(struct dm_pool *mem, const char *name,
 		     const struct logical_volume *lv, int with_mknodes,
 		     int with_open_count, int with_read_ahead,
@@ -259,10 +262,8 @@ static int _status_run(const char *name, const char *uuid,
 	char *type = NULL;
 	char *params = NULL;
 
-	if (!(dmt = _setup_task(name, uuid, 0, DM_DEVICE_STATUS))) {
-		stack;
-		return 0;
-	}
+	if (!(dmt = _setup_task(name, uuid, 0, DM_DEVICE_STATUS, 0, 0)))
+		return_0;
 
 	if (!dm_task_no_open_count(dmt))
 		log_error("Failed to disable open_count");
@@ -339,7 +340,7 @@ static int _percent_run(struct dev_manager *dm, const char *name,
 	uint64_t start, length;
 	char *type = NULL;
 	char *params = NULL;
-	struct list *segh = &lv->segments;
+	struct dm_list *segh = &lv->segments;
 	struct lv_segment *seg = NULL;
 	struct segment_type *segtype;
 
@@ -348,10 +349,8 @@ static int _percent_run(struct dev_manager *dm, const char *name,
 	*percent = -1;
 
 	if (!(dmt = _setup_task(name, dlid, event_nr,
-				wait ? DM_DEVICE_WAITEVENT : DM_DEVICE_STATUS))) {
-		stack;
-		return 0;
-	}
+				wait ? DM_DEVICE_WAITEVENT : DM_DEVICE_STATUS, 0, 0)))
+		return_0;
 
 	if (!dm_task_no_open_count(dmt))
 		log_error("Failed to disable open_count");
@@ -369,12 +368,12 @@ static int _percent_run(struct dev_manager *dm, const char *name,
 		next = dm_get_next_target(dmt, next, &start, &length, &type,
 					  &params);
 		if (lv) {
-			if (!(segh = list_next(&lv->segments, segh))) {
+			if (!(segh = dm_list_next(&lv->segments, segh))) {
 				log_error("Number of segments in active LV %s "
 					  "does not match metadata", lv->name);
 				goto out;
 			}
-			seg = list_item(segh, struct lv_segment);
+			seg = dm_list_item(segh, struct lv_segment);
 		}
 
 		if (!type || !params || strcmp(type, target_type))
@@ -387,13 +386,12 @@ static int _percent_run(struct dev_manager *dm, const char *name,
 		    !segtype->ops->target_percent(&dm->target_state, dm->mem,
 						  dm->cmd, seg, params,
 						  &total_numerator,
-						  &total_denominator,
-						  percent))
+						  &total_denominator))
 			goto_out;
 
 	} while (next);
 
-	if (lv && (segh = list_next(&lv->segments, segh))) {
+	if (lv && (segh = dm_list_next(&lv->segments, segh))) {
 		log_error("Number of segments in active LV %s does not "
 			  "match metadata", lv->name);
 		goto out;
@@ -401,7 +399,7 @@ static int _percent_run(struct dev_manager *dm, const char *name,
 
 	if (total_denominator)
 		*percent = (float) total_numerator *100 / total_denominator;
-	else if (*percent < 0)
+	else
 		*percent = 100;
 
 	log_debug("LV percent: %f", *percent);
@@ -443,29 +441,22 @@ struct dev_manager *dev_manager_create(struct cmd_context *cmd,
 	struct dm_pool *mem;
 	struct dev_manager *dm;
 
-	if (!(mem = dm_pool_create("dev_manager", 16 * 1024))) {
-		stack;
-		return NULL;
-	}
+	if (!(mem = dm_pool_create("dev_manager", 16 * 1024)))
+		return_NULL;
 
-	if (!(dm = dm_pool_alloc(mem, sizeof(*dm))))
+	if (!(dm = dm_pool_zalloc(mem, sizeof(*dm))))
 		goto_bad;
 
 	dm->cmd = cmd;
 	dm->mem = mem;
 
-	if (!stripe_filler) {
-		stripe_filler = find_config_tree_str(cmd,
-						"activation/missing_stripe_filler",
-						DEFAULT_STRIPE_FILLER);
-	}
-	dm->stripe_filler = stripe_filler;
-
 	if (!(dm->vg_name = dm_pool_strdup(dm->mem, vg_name)))
 		goto_bad;
 
 	dm->target_state = NULL;
 
+	dm_udev_set_sync_support(cmd->current_settings.udev_sync);
+
 	return dm;
 
       bad:
@@ -509,10 +500,8 @@ int dev_manager_snapshot_percent(struct dev_manager *dm,
 	 */
 	log_debug("Getting device status percentage for %s", name);
 	if (!(_percent(dm, name, dlid, "snapshot", 0, NULL, percent,
-		       NULL))) {
-		stack;
-		return 0;
-	}
+		       NULL)))
+		return_0;
 
 	/* FIXME dm_pool_free ? */
 
@@ -544,10 +533,8 @@ int dev_manager_mirror_percent(struct dev_manager *dm,
 
 	log_debug("Getting device mirror status percentage for %s", name);
 	if (!(_percent(dm, name, dlid, "mirror", wait, lv, percent,
-		       event_nr))) {
-		stack;
-		return 0;
-	}
+		       event_nr)))
+		return_0;
 
 	return 1;
 }
@@ -601,7 +588,7 @@ static int _belong_to_vg(const char *vgname, const char *name)
 	old_origin = snap_seg->origin;
 
 	/* Was this the last active snapshot with this origin? */
-	list_iterate_items(lvl, active_head) {
+	dm_list_iterate_items(lvl, active_head) {
 		active = lvl->lv;
 		if ((snap_seg = find_cow(active)) &&
 		    snap_seg->origin == old_origin) {
@@ -635,7 +622,7 @@ static int _add_dev_to_dtree(struct dev_manager *dm, struct dm_tree *dtree,
 			       struct logical_volume *lv, const char *layer)
 {
 	char *dlid, *name;
-	struct dm_info info;
+	struct dm_info info, info2;
 
 	if (!(name = build_dm_name(dm->mem, lv->vg->name, lv->name, layer)))
 		return_0;
@@ -643,11 +630,35 @@ static int _add_dev_to_dtree(struct dev_manager *dm, struct dm_tree *dtree,
 	if (!(dlid = build_dlid(dm, lv->lvid.s, layer)))
 		return_0;
 
-        log_debug("Getting device info for %s [%s]", name, dlid);
-        if (!_info(name, dlid, 0, 1, 0, &info, NULL)) {
-                log_error("Failed to get info for %s [%s].", name, dlid);
-                return 0;
-        }
+	log_debug("Getting device info for %s [%s]", name, dlid);
+	if (!_info(name, dlid, 0, 1, 0, &info, NULL)) {
+		log_error("Failed to get info for %s [%s].", name, dlid);
+		return 0;
+	}
+
+	/*
+	 * For top level volumes verify that existing device match
+	 * requested major/minor and that major/minor pair is available for use
+	 */
+	if (!layer && lv->major != -1 && lv->minor != -1) {
+		/*
+		 * FIXME compare info.major with lv->major if multiple major support
+		 */
+		if (info.exists && (info.minor != lv->minor)) {
+			log_error("Volume %s (%" PRIu32 ":%" PRIu32")"
+				  " differs from already active device "
+				  "(%" PRIu32 ":%" PRIu32")",
+				  lv->name, lv->major, lv->minor, info.major, info.minor);
+			return 0;
+		}
+		if (!info.exists && _info_by_dev(lv->major, lv->minor, &info2) &&
+		    info2.exists) {
+			log_error("The requested major:minor pair "
+				  "(%" PRIu32 ":%" PRIu32") is already used",
+				  lv->major, lv->minor);
+			return 0;
+		}
+	}
 
 	if (info.exists && !dm_tree_add_dev(dtree, info.major, info.minor)) {
 		log_error("Failed to add device (%" PRIu32 ":%" PRIu32") to dtree",
@@ -682,7 +693,7 @@ static int _add_lv_to_dtree(struct dev_manager *dm, struct dm_tree *dtree, struc
 static struct dm_tree *_create_partial_dtree(struct dev_manager *dm, struct logical_volume *lv)
 {
 	struct dm_tree *dtree;
-	struct list *snh, *snht;
+	struct dm_list *snh, *snht;
 	struct lv_segment *seg;
 	uint32_t s;
 
@@ -691,35 +702,91 @@ static struct dm_tree *_create_partial_dtree(struct dev_manager *dm, struct logi
 		return NULL;
 	}
 
-	if (!_add_lv_to_dtree(dm, dtree, lv)) {
-		stack;
-		goto fail;
-	}
+	if (!_add_lv_to_dtree(dm, dtree, lv))
+		goto_bad;
 
 	/* Add any snapshots of this LV */
-	list_iterate_safe(snh, snht, &lv->snapshot_segs)
-		if (!_add_lv_to_dtree(dm, dtree, list_struct_base(snh, struct lv_segment, origin_list)->cow)) {
-			stack;
-			goto fail;
-		}
+	dm_list_iterate_safe(snh, snht, &lv->snapshot_segs)
+		if (!_add_lv_to_dtree(dm, dtree, dm_list_struct_base(snh, struct lv_segment, origin_list)->cow))
+			goto_bad;
 
 	/* Add any LVs used by segments in this LV */
-	list_iterate_items(seg, &lv->segments)
+	dm_list_iterate_items(seg, &lv->segments)
 		for (s = 0; s < seg->area_count; s++)
 			if (seg_type(seg, s) == AREA_LV && seg_lv(seg, s)) {
-				if (!_add_lv_to_dtree(dm, dtree, seg_lv(seg, s))) {
-					stack;
-					goto fail;
-				}
+				if (!_add_lv_to_dtree(dm, dtree, seg_lv(seg, s)))
+					goto_bad;
 			}
 
 	return dtree;
 
-fail:
+bad:
 	dm_tree_free(dtree);
 	return NULL;
 }
 
+static char *_add_error_device(struct dev_manager *dm, struct dm_tree *dtree,
+			       struct lv_segment *seg, int s)
+{
+	char *id, *name;
+	char errid[32];
+	struct dm_tree_node *node;
+	struct lv_segment *seg_i;
+	int segno = -1, i = 0;;
+	uint64_t size = seg->len * seg->lv->vg->extent_size;
+
+	dm_list_iterate_items(seg_i, &seg->lv->segments) {
+		if (seg == seg_i)
+			segno = i;
+		++i;
+	}
+
+	if (segno < 0) {
+		log_error("_add_error_device called with bad segment");
+		return_NULL;
+	}
+
+	sprintf(errid, "missing_%d_%d", segno, s);
+
+	if (!(id = build_dlid(dm, seg->lv->lvid.s, errid))) 
+		return_NULL;
+
+	if (!(name = build_dm_name(dm->mem, seg->lv->vg->name,
+				   seg->lv->name, errid)))
+		return_NULL;
+	if (!(node = dm_tree_add_new_dev(dtree, name, id, 0, 0, 0, 0, 0)))
+		return_NULL;
+	if (!dm_tree_node_add_error_target(node, size))
+		return_NULL;
+
+	return id;
+}
+
+static int _add_error_area(struct dev_manager *dm, struct dm_tree_node *node,
+			   struct lv_segment *seg, int s)
+{
+	char *dlid;
+	uint64_t extent_size = seg->lv->vg->extent_size;
+
+	if (!strcmp(dm->cmd->stripe_filler, "error")) {
+		/*
+		 * FIXME, the tree pointer is first field of dm_tree_node, but
+		 * we don't have the struct definition available.
+		 */
+		struct dm_tree **tree = (struct dm_tree **) node;
+		dlid = _add_error_device(dm, *tree, seg, s);
+		if (!dlid)
+			return_0;
+		dm_tree_node_add_target_area(node, NULL, dlid,
+					     extent_size * seg_le(seg, s));
+	} else
+		dm_tree_node_add_target_area(node,
+					     dm->cmd->stripe_filler,
+					     NULL, UINT64_C(0));
+
+	return 1;
+}
+
 int add_areas_line(struct dev_manager *dm, struct lv_segment *seg,
 		   struct dm_tree_node *node, uint32_t start_area,
 		   uint32_t areas)
@@ -733,11 +800,10 @@ int add_areas_line(struct dev_manager *dm, struct lv_segment *seg,
 		     (!seg_pvseg(seg, s) ||
 		      !seg_pv(seg, s) ||
 		      !seg_dev(seg, s))) ||
-		    (seg_type(seg, s) == AREA_LV && !seg_lv(seg, s)))
-			dm_tree_node_add_target_area(node,
-							dm->stripe_filler,
-							NULL, UINT64_C(0));
-		else if (seg_type(seg, s) == AREA_PV)
+		    (seg_type(seg, s) == AREA_LV && !seg_lv(seg, s))) {
+			if (!_add_error_area(dm, node, seg, s))
+				return_0;
+		} else if (seg_type(seg, s) == AREA_PV)
 			dm_tree_node_add_target_area(node,
 							dev_name(seg_dev(seg, s)),
 							NULL,
@@ -832,7 +898,7 @@ static int _add_segment_to_dtree(struct dev_manager *dm,
 				   const char *layer)
 {
 	uint32_t s;
-	struct list *snh;
+	struct dm_list *snh;
 	struct lv_segment *seg_present;
 
 	/* Ensure required device-mapper targets are loaded */
@@ -843,7 +909,8 @@ static int _add_segment_to_dtree(struct dev_manager *dm,
 		  layer ? "-" : "", layer ? : "");
 
 	if (seg_present->segtype->ops->target_present &&
-	    !seg_present->segtype->ops->target_present(seg_present)) {
+	    !seg_present->segtype->ops->target_present(seg_present->lv->vg->cmd,
+						       seg_present, NULL)) {
 		log_error("Can't expand LV %s: %s target support missing "
 			  "from kernel?", seg->lv->name, seg_present->segtype->name);
 		return 0;
@@ -856,7 +923,7 @@ static int _add_segment_to_dtree(struct dev_manager *dm,
 
 	/* If this is a snapshot origin, add real LV */
 	if (lv_is_origin(seg->lv) && !layer) {
-		if (seg->lv->vg->status & CLUSTERED) {
+		if (vg_is_clustered(seg->lv->vg)) {
 			log_error("Clustered snapshots are not yet supported");
 			return 0;
 		}
@@ -885,8 +952,8 @@ static int _add_segment_to_dtree(struct dev_manager *dm,
 
 	if (lv_is_origin(seg->lv) && !layer)
 		/* Add any snapshots of this LV */
-		list_iterate(snh, &seg->lv->snapshot_segs)
-			if (!_add_new_lv_to_dtree(dm, dtree, list_struct_base(snh, struct lv_segment, origin_list)->cow, NULL))
+		dm_list_iterate(snh, &seg->lv->snapshot_segs)
+			if (!_add_new_lv_to_dtree(dm, dtree, dm_list_struct_base(snh, struct lv_segment, origin_list)->cow, NULL))
 				return_0;
 
 	return 1;
@@ -940,7 +1007,7 @@ static int _add_new_lv_to_dtree(struct dev_manager *dm, struct dm_tree *dtree,
 
 	/* Create table */
 	dm->pvmove_mirror_count = 0u;
-	list_iterate_items(seg, &lv->segments) {
+	dm_list_iterate_items(seg, &lv->segments) {
 		if (!_add_segment_to_dtree(dm, dtree, dnode, seg, layer))
 			return_0;
 		/* These aren't real segments in the LVM2 metadata */
@@ -948,14 +1015,17 @@ static int _add_new_lv_to_dtree(struct dev_manager *dm, struct dm_tree *dtree,
 			break;
 		if (lv_is_cow(lv) && !layer)
 			break;
-		if (max_stripe_size < seg->stripe_size)
-			max_stripe_size = seg->stripe_size;
+		if (max_stripe_size < seg->stripe_size * seg->area_count)
+			max_stripe_size = seg->stripe_size * seg->area_count;
 	}
 
-	if (read_ahead == DM_READ_AHEAD_AUTO)
-		read_ahead = max_stripe_size;
-	else
+	if (read_ahead == DM_READ_AHEAD_AUTO) {
+		/* we need RA at least twice a whole stripe - see the comment in md/raid0.c */
+		read_ahead = max_stripe_size * 2;
+		if (!read_ahead)
+			lv_calculate_readahead(lv, &read_ahead);
 		read_ahead_flags = DM_READ_AHEAD_MINIMUM_FLAG;
+	}
 
 	dm_tree_node_set_read_ahead(dnode, read_ahead, read_ahead_flags);
 
@@ -974,7 +1044,8 @@ static int _create_lv_symlinks(struct dev_manager *dm, struct dm_tree_node *root
 	void *handle = NULL;
 	struct dm_tree_node *child;
 	struct lv_layer *lvlayer;
-	char *vgname, *lvname, *layer;
+	char *old_vgname, *old_lvname, *old_layer;
+	char *new_vgname, *new_lvname, *new_layer;
 	const char *name;
 	int r = 1;
 
@@ -986,12 +1057,24 @@ static int _create_lv_symlinks(struct dev_manager *dm, struct dm_tree_node *root
 		name = dm_tree_node_get_name(child);
 
 		if (name && lvlayer->old_name && *lvlayer->old_name && strcmp(name, lvlayer->old_name)) {
-	        	if (!dm_split_lvm_name(dm->mem, lvlayer->old_name, &vgname, &lvname, &layer)) {
-                		log_error("_create_lv_symlinks: Couldn't split up old device name %s", lvlayer->old_name);
-                		return 0;
-        		}
-			fs_rename_lv(lvlayer->lv, name, lvname);
-		} else if (!dev_manager_lv_mknodes(lvlayer->lv))
+			if (!dm_split_lvm_name(dm->mem, lvlayer->old_name, &old_vgname, &old_lvname, &old_layer)) {
+				log_error("_create_lv_symlinks: Couldn't split up old device name %s", lvlayer->old_name);
+				return 0;
+			}
+			if (!dm_split_lvm_name(dm->mem, name, &new_vgname, &new_lvname, &new_layer)) {
+				log_error("_create_lv_symlinks: Couldn't split up new device name %s", name);
+				return 0;
+			}
+			if (!fs_rename_lv(lvlayer->lv, name, old_vgname, old_lvname))
+				r = 0;
+			continue;
+		}
+		if (lv_is_visible(lvlayer->lv)) {
+			if (!dev_manager_lv_mknodes(lvlayer->lv))
+				r = 0;
+			continue;
+		}
+		if (!dev_manager_lv_rmnodes(lvlayer->lv))
 			r = 0;
 	}
 
@@ -1009,11 +1092,14 @@ static int _remove_lv_symlinks(struct dev_manager *dm, struct dm_tree_node *root
 	int r = 1;
 
 	while ((child = dm_tree_next_child(&handle, root, 0))) {
-        	if (!dm_split_lvm_name(dm->mem, dm_tree_node_get_name(child), &vgname, &lvname, &layer)) {
+		if (!dm_split_lvm_name(dm->mem, dm_tree_node_get_name(child), &vgname, &lvname, &layer)) {
 			r = 0;
 			continue;
 		}
 
+		if (!*vgname)
+			continue;
+
 		/* only top level layer has symlinks */
 		if (*layer)
 			continue;
@@ -1030,6 +1116,7 @@ static int _clean_tree(struct dev_manager *dm, struct dm_tree_node *root)
 	struct dm_tree_node *child;
 	char *vgname, *lvname, *layer;
 	const char *name, *uuid;
+	int r;
 
 	while ((child = dm_tree_next_child(&handle, root, 0))) {
 		if (!(name = dm_tree_node_get_name(child)))
@@ -1038,16 +1125,21 @@ static int _clean_tree(struct dev_manager *dm, struct dm_tree_node *root)
 		if (!(uuid = dm_tree_node_get_uuid(child)))
 			continue;
 
-        	if (!dm_split_lvm_name(dm->mem, name, &vgname, &lvname, &layer)) {
-                	log_error("_clean_tree: Couldn't split up device name %s.", name);
-                	return 0;
-        	}
+		if (!dm_split_lvm_name(dm->mem, name, &vgname, &lvname, &layer)) {
+			log_error("_clean_tree: Couldn't split up device name %s.", name);
+			return 0;
+		}
 
 		/* Not meant to be top level? */
 		if (!*layer)
 			continue;
 
-		if (!dm_tree_deactivate_children(root, uuid, strlen(uuid)))
+		dm_tree_set_cookie(root, 0);
+		r = dm_tree_deactivate_children(root, uuid, strlen(uuid));
+		if (!dm_udev_wait(dm_tree_get_cookie(root)))
+			stack;
+
+		if (!r)
 			return_0;
 	}
 
@@ -1081,14 +1173,18 @@ static int _tree_action(struct dev_manager *dm, struct logical_volume *lv, actio
 		break;
 	case DEACTIVATE:
  		/* Deactivate LV and all devices it references that nothing else has open. */
-		if (!dm_tree_deactivate_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1))
+		dm_tree_set_cookie(root, 0);
+		r = dm_tree_deactivate_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1);
+		if (!dm_udev_wait(dm_tree_get_cookie(root)))
+			stack;
+		if (!r)
 			goto_out;
 		if (!_remove_lv_symlinks(dm, root))
 			log_error("Failed to remove all device symlinks associated with %s.", lv->name);
 		break;
 	case SUSPEND:
 		dm_tree_skip_lockfs(root);
-		if ((lv->status & MIRRORED) && !(lv->status & PVMOVE))
+		if (!dm->flush_required && (lv->status & MIRRORED) && !(lv->status & PVMOVE))
 			dm_tree_use_no_flush_suspend(root);
 	case SUSPEND_WITH_LOCKFS:
 		if (!dm_tree_suspend_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1))
@@ -1101,17 +1197,29 @@ static int _tree_action(struct dev_manager *dm, struct logical_volume *lv, actio
 			goto_out;
 
 		/* Preload any devices required before any suspensions */
-		if (!dm_tree_preload_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1))
+		dm_tree_set_cookie(root, 0);
+		r = dm_tree_preload_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1);
+		if (!dm_udev_wait(dm_tree_get_cookie(root)))
+			stack;
+		if (!r)
 			goto_out;
 
-		if ((action == ACTIVATE) &&
-		    !dm_tree_activate_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1))
-			goto_out;
+		if (dm_tree_node_size_changed(root))
+			dm->flush_required = 1;
 
-		if (!_create_lv_symlinks(dm, root)) {
-			log_error("Failed to create symlinks for %s.", lv->name);
-			goto out;
+		if (action == ACTIVATE) {
+			dm_tree_set_cookie(root, 0);
+			r = dm_tree_activate_children(root, dlid, ID_LEN + sizeof(UUID_PREFIX) - 1);
+			if (!dm_udev_wait(dm_tree_get_cookie(root)))
+				stack;
+			if (!r)
+				goto_out;
+			if (!_create_lv_symlinks(dm, root)) {
+				log_error("Failed to create symlinks for %s.", lv->name);
+				goto out;
+			}
 		}
+
 		break;
 	default:
 		log_error("_tree_action: Action %u not supported.", action);
@@ -1134,13 +1242,19 @@ int dev_manager_activate(struct dev_manager *dm, struct logical_volume *lv)
 	return _tree_action(dm, lv, CLEAN);
 }
 
-int dev_manager_preload(struct dev_manager *dm, struct logical_volume *lv)
+int dev_manager_preload(struct dev_manager *dm, struct logical_volume *lv,
+			int *flush_required)
 {
 	/* FIXME Update the pvmove implementation! */
 	if ((lv->status & PVMOVE) || (lv->status & LOCKED))
 		return 1;
 
-	return _tree_action(dm, lv, PRELOAD);
+	if (!_tree_action(dm, lv, PRELOAD))
+		return 0;
+
+	*flush_required = dm->flush_required;
+
+	return 1;
 }
 
 int dev_manager_deactivate(struct dev_manager *dm, struct logical_volume *lv)
@@ -1155,8 +1269,10 @@ int dev_manager_deactivate(struct dev_manager *dm, struct logical_volume *lv)
 }
 
 int dev_manager_suspend(struct dev_manager *dm, struct logical_volume *lv,
-			int lockfs)
+			int lockfs, int flush_required)
 {
+	dm->flush_required = flush_required;
+
 	return _tree_action(dm, lv, lockfs ? SUSPEND_WITH_LOCKFS : SUSPEND);
 }
 

lib/activate/dev_manager.h

@@ -49,9 +49,10 @@ int dev_manager_mirror_percent(struct dev_manager *dm,
 			       struct logical_volume *lv, int wait,
 			       float *percent, uint32_t *event_nr);
 int dev_manager_suspend(struct dev_manager *dm, struct logical_volume *lv,
-			int lockfs);
+			int lockfs, int flush_required);
 int dev_manager_activate(struct dev_manager *dm, struct logical_volume *lv);
-int dev_manager_preload(struct dev_manager *dm, struct logical_volume *lv);
+int dev_manager_preload(struct dev_manager *dm, struct logical_volume *lv,
+			int *flush_required);
 int dev_manager_deactivate(struct dev_manager *dm, struct logical_volume *lv);
 
 int dev_manager_lv_mknodes(const struct logical_volume *lv);

lib/activate/fs.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -107,7 +107,7 @@ static int _mk_link(const char *dev_dir, const char *vg_name,
 {
 	char lv_path[PATH_MAX], link_path[PATH_MAX], lvm1_group_path[PATH_MAX];
 	char vg_path[PATH_MAX];
-	struct stat buf;
+	struct stat buf, buf_lp;
 
 	if (dm_snprintf(vg_path, sizeof(vg_path), "%s%s",
 			 dev_dir, vg_name) == -1) {
@@ -138,7 +138,7 @@ static int _mk_link(const char *dev_dir, const char *vg_name,
 	}
 
 	/* To reach this point, the VG must have been locked.
-	 * As locking fails if the VG is active under LVM1, it's 
+	 * As locking fails if the VG is active under LVM1, it's
 	 * now safe to remove any LVM1 devices we find here
 	 * (as well as any existing LVM2 symlink). */
 	if (!lstat(lvm1_group_path, &buf)) {
@@ -161,12 +161,34 @@ static int _mk_link(const char *dev_dir, const char *vg_name,
 			return 0;
 		}
 
+		if (dm_udev_get_sync_support()) {
+			/* Check udev created the correct link. */
+			if (!stat(link_path, &buf_lp) &&
+			    !stat(lv_path, &buf)) {
+				if (buf_lp.st_rdev == buf.st_rdev)
+					return 1;
+				else
+					log_warn("Symlink %s that should have been "
+						 "created by udev does not have "
+						 "correct target. Falling back to "
+						 "direct link creation", lv_path);
+			} else
+				log_warn("Symlink %s that should have been "
+					 "created by udev could not be checked "
+					 "for its correctness. Falling back to "
+					 "direct link creation.", lv_path);
+
+		}
+
 		log_very_verbose("Removing %s", lv_path);
 		if (unlink(lv_path) < 0) {
 			log_sys_error("unlink", lv_path);
 			return 0;
 		}
-	}
+	} else if (dm_udev_get_sync_support())
+		log_warn("The link %s should had been created by udev "
+			  "but it was not found. Falling back to "
+			  "direct link creation.", lv_path);
 
 	log_very_verbose("Linking %s -> %s", lv_path, link_path);
 	if (symlink(link_path, lv_path) < 0) {
@@ -175,10 +197,8 @@ static int _mk_link(const char *dev_dir, const char *vg_name,
 	}
 
 #ifdef HAVE_SELINUX
-        if (!dm_set_selinux_context(lv_path, S_IFLNK)) {
-                stack;
-                return 0;
-        }
+	if (!dm_set_selinux_context(lv_path, S_IFLNK))
+		return_0;
 #endif
 
 	return 1;
@@ -196,9 +216,14 @@ static int _rm_link(const char *dev_dir, const char *vg_name,
 		return 0;
 	}
 
-	if (lstat(lv_path, &buf) || !S_ISLNK(buf.st_mode)) {
-		if (errno == ENOENT)
-			return 1;
+	if (lstat(lv_path, &buf) && errno == ENOENT)
+		return 1;
+	else if (dm_udev_get_sync_support())
+		log_warn("The link %s should have been removed by udev "
+			 "but it is still present. Falling back to "
+			 "direct link removal.", lv_path);
+
+	if (!S_ISLNK(buf.st_mode)) {
 		log_error("%s not symbolic link - not removing", lv_path);
 		return 0;
 	}
@@ -225,17 +250,13 @@ static int _do_fs_op(fs_op_t type, const char *dev_dir, const char *vg_name,
 	switch (type) {
 	case FS_ADD:
 		if (!_mk_dir(dev_dir, vg_name) ||
-		    !_mk_link(dev_dir, vg_name, lv_name, dev)) {
-			stack;
-			return 0;
-		}
+		    !_mk_link(dev_dir, vg_name, lv_name, dev))
+			return_0;
 		break;
 	case FS_DEL:
 		if (!_rm_link(dev_dir, vg_name, lv_name) ||
-		    !_rm_dir(dev_dir, vg_name)) {
-			stack;
-			return 0;
-		}
+		    !_rm_dir(dev_dir, vg_name))
+			return_0;
 		break;
 		/* FIXME Use rename() */
 	case FS_RENAME:
@@ -249,10 +270,10 @@ static int _do_fs_op(fs_op_t type, const char *dev_dir, const char *vg_name,
 	return 1;
 }
 
-static LIST_INIT(_fs_ops);
+static DM_LIST_INIT(_fs_ops);
 
 struct fs_op_parms {
-	struct list list;
+	struct dm_list list;
 	fs_op_t type;
 	char *dev_dir;
 	char *vg_name;
@@ -292,21 +313,21 @@ static int _stack_fs_op(fs_op_t type, const char *dev_dir, const char *vg_name,
 	_store_str(&pos, &fsp->dev, dev);
 	_store_str(&pos, &fsp->old_lv_name, old_lv_name);
 
-	list_add(&_fs_ops, &fsp->list);
+	dm_list_add(&_fs_ops, &fsp->list);
 
 	return 1;
 }
 
 static void _pop_fs_ops(void)
 {
-	struct list *fsph, *fspht;
+	struct dm_list *fsph, *fspht;
 	struct fs_op_parms *fsp;
 
-	list_iterate_safe(fsph, fspht, &_fs_ops) {
-		fsp = list_item(fsph, struct fs_op_parms);
+	dm_list_iterate_safe(fsph, fspht, &_fs_ops) {
+		fsp = dm_list_item(fsph, struct fs_op_parms);
 		_do_fs_op(fsp->type, fsp->dev_dir, fsp->vg_name, fsp->lv_name,
 			  fsp->dev, fsp->old_lv_name);
-		list_del(&fsp->list);
+		dm_list_del(&fsp->list);
 		dm_free(fsp);
 	}
 }
@@ -316,10 +337,8 @@ static int _fs_op(fs_op_t type, const char *dev_dir, const char *vg_name,
 {
 	if (memlock()) {
 		if (!_stack_fs_op(type, dev_dir, vg_name, lv_name, dev,
-				  old_lv_name)) {
-			stack;
-			return 0;
-		}
+				  old_lv_name))
+			return_0;
 		return 1;
 	}
 
@@ -343,11 +362,17 @@ int fs_del_lv_byname(const char *dev_dir, const char *vg_name, const char *lv_na
 	return _fs_op(FS_DEL, dev_dir, vg_name, lv_name, "", "");
 }
 
-int fs_rename_lv(struct logical_volume *lv,
-		 const char *dev, const char *old_name)
+int fs_rename_lv(struct logical_volume *lv, const char *dev, 
+		const char *old_vgname, const char *old_lvname)
 {
-	return _fs_op(FS_RENAME, lv->vg->cmd->dev_dir, lv->vg->name, lv->name,
-		      dev, old_name);
+	if (strcmp(old_vgname, lv->vg->name)) {
+		return
+			(_fs_op(FS_DEL, lv->vg->cmd->dev_dir, old_vgname, old_lvname, "", "") &&
+			 _fs_op(FS_ADD, lv->vg->cmd->dev_dir, lv->vg->name, lv->name, dev, ""));
+	}
+	else 
+		return _fs_op(FS_RENAME, lv->vg->cmd->dev_dir, lv->vg->name, lv->name,
+			      dev, old_lvname);
 }
 
 void fs_unlock(void)

lib/activate/fs.h

@@ -26,8 +26,8 @@
 int fs_add_lv(const struct logical_volume *lv, const char *dev);
 int fs_del_lv(const struct logical_volume *lv);
 int fs_del_lv_byname(const char *dev_dir, const char *vg_name, const char *lv_name);
-int fs_rename_lv(struct logical_volume *lv,
-		 const char *dev, const char *old_name);
+int fs_rename_lv(struct logical_volume *lv, const char *dev, 
+		const char *old_vgname, const char *old_lvname);
 void fs_unlock(void);
 
 #endif

lib/cache/lvmcache.h

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
- * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -20,7 +20,8 @@
 #include "uuid.h"
 #include "label.h"
 
-#define ORPHAN ""
+#define ORPHAN_PREFIX "#"
+#define ORPHAN_VG_NAME(fmt) ORPHAN_PREFIX "orphans_" fmt
 
 #define CACHE_INVALID	0x00000001
 #define CACHE_LOCKED	0x00000002
@@ -34,8 +35,8 @@ struct volume_group;
 
 /* One per VG */
 struct lvmcache_vginfo {
-	struct list list;	/* Join these vginfos together */
-	struct list infos;	/* List head for lvmcache_infos */
+	struct dm_list list;	/* Join these vginfos together */
+	struct dm_list infos;	/* List head for lvmcache_infos */
 	const struct format_type *fmt;
 	char *vgname;		/* "" == orphan */
 	uint32_t status;
@@ -43,13 +44,15 @@ struct lvmcache_vginfo {
 	char _padding[7];
 	struct lvmcache_vginfo *next; /* Another VG with same name? */
 	char *creation_host;
+	char *vgmetadata;	/* Copy of VG metadata as format_text string */
+	unsigned precommitted;	/* Is vgmetadata live or precommitted? */
 };
 
 /* One per device */
 struct lvmcache_info {
-	struct list list;	/* Join VG members together */
-	struct list mdas;	/* list head for metadata areas */
-	struct list das;	/* list head for data areas */
+	struct dm_list list;	/* Join VG members together */
+	struct dm_list mdas;	/* list head for metadata areas */
+	struct dm_list das;	/* list head for data areas */
 	struct lvmcache_vginfo *vginfo;	/* NULL == unknown */
 	struct label *label;
 	const struct format_type *fmt;
@@ -59,7 +62,7 @@ struct lvmcache_info {
 };
 
 int lvmcache_init(void);
-void lvmcache_destroy(void);
+void lvmcache_destroy(struct cmd_context *cmd, int retain_orphans);
 
 /* Set full_scan to 1 to reread every filtered device label or
  * 2 to rescan /dev for new devices */
@@ -70,13 +73,14 @@ struct lvmcache_info *lvmcache_add(struct labeller *labeller, const char *pvid,
 				   struct device *dev,
 				   const char *vgname, const char *vgid,
 				   uint32_t vgstatus);
+int lvmcache_add_orphan_vginfo(const char *vgname, struct format_type *fmt);
 void lvmcache_del(struct lvmcache_info *info);
 
 /* Update things */
 int lvmcache_update_vgname_and_id(struct lvmcache_info *info,
 				  const char *vgname, const char *vgid,
 				  uint32_t vgstatus, const char *hostname);
-int lvmcache_update_vg(struct volume_group *vg);
+int lvmcache_update_vg(struct volume_group *vg, unsigned precommitted);
 
 void lvmcache_lock_vgname(const char *vgname, int read_only);
 void lvmcache_unlock_vgname(const char *vgname);
@@ -86,7 +90,7 @@ const struct format_type *fmt_from_vgname(const char *vgname, const char *vgid);
 struct lvmcache_vginfo *vginfo_from_vgname(const char *vgname,
 					   const char *vgid);
 struct lvmcache_vginfo *vginfo_from_vgid(const char *vgid);
-struct lvmcache_info *info_from_pvid(const char *pvid);
+struct lvmcache_info *info_from_pvid(const char *pvid, int valid_only);
 const char *vgname_from_vgid(struct dm_pool *mem, const char *vgid);
 struct device *device_from_pvid(struct cmd_context *cmd, struct id *pvid);
 int vgs_locked(void);
@@ -94,14 +98,18 @@ int vgname_is_locked(const char *vgname);
 
 /* Returns list of struct str_lists containing pool-allocated copy of vgnames */
 /* Set full_scan to 1 to reread every filtered device label */
-struct list *lvmcache_get_vgnames(struct cmd_context *cmd, int full_scan);
+struct dm_list *lvmcache_get_vgnames(struct cmd_context *cmd, int full_scan);
 
 /* Returns list of struct str_lists containing pool-allocated copy of vgids */
 /* Set full_scan to 1 to reread every filtered device label */
-struct list *lvmcache_get_vgids(struct cmd_context *cmd, int full_scan);
+struct dm_list *lvmcache_get_vgids(struct cmd_context *cmd, int full_scan);
 
 /* Returns list of struct str_lists containing pool-allocated copy of pvids */
-struct list *lvmcache_get_pvids(struct cmd_context *cmd, const char *vgname,
+struct dm_list *lvmcache_get_pvids(struct cmd_context *cmd, const char *vgname,
 				const char *vgid);
 
+/* Returns cached volume group metadata. */
+struct volume_group *lvmcache_get_vg(const char *vgid, unsigned precommitted);
+void lvmcache_drop_metadata(const char *vgname);
+
 #endif

lib/commands/errors.h

@@ -21,4 +21,6 @@
 #define EINVALID_CMD_LINE	3
 #define ECMD_FAILED		5
 
+/* FIXME Also returned by cmdlib. */
+
 #endif

lib/commands/toolcontext.c

@@ -1,6 +1,6 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -64,7 +64,7 @@ static int _get_env_vars(struct cmd_context *cmd)
 
 	/* Set to "" to avoid using any system directory */
 	if ((e = getenv("LVM_SYSTEM_DIR"))) {
-		if (dm_snprintf(cmd->sys_dir, sizeof(cmd->sys_dir),
+		if (dm_snprintf(cmd->system_dir, sizeof(cmd->system_dir),
 				 "%s", e) < 0) {
 			log_error("LVM_SYSTEM_DIR environment variable "
 				  "is too long.");
@@ -75,6 +75,49 @@ static int _get_env_vars(struct cmd_context *cmd)
 	return 1;
 }
 
+static void _get_sysfs_dir(struct cmd_context *cmd)
+{
+	static char proc_mounts[PATH_MAX];
+	static char *split[4], buffer[PATH_MAX + 16];
+	FILE *fp;
+	char *sys_mnt = NULL;
+
+	cmd->sysfs_dir[0] = '\0';
+	if (!*cmd->proc_dir) {
+		log_debug("No proc filesystem found: skipping sysfs detection");
+		return;
+	}
+
+	if (dm_snprintf(proc_mounts, sizeof(proc_mounts),
+			 "%s/mounts", cmd->proc_dir) < 0) {
+		log_error("Failed to create /proc/mounts string for sysfs detection");
+		return;
+	}
+
+	if (!(fp = fopen(proc_mounts, "r"))) {
+		log_sys_error("_get_sysfs_dir: fopen %s", proc_mounts);
+		return;
+	}
+
+	while (fgets(buffer, sizeof(buffer), fp)) {
+		if (dm_split_words(buffer, 4, 0, split) == 4 &&
+		    !strcmp(split[2], "sysfs")) {
+			sys_mnt = split[1];
+			break;
+		}
+	}
+
+	if (fclose(fp))
+		log_sys_error("fclose", proc_mounts);
+
+	if (!sys_mnt) {
+		log_error("Failed to find sysfs mount point");
+		return;
+	}
+
+	strncpy(cmd->sysfs_dir, sys_mnt, sizeof(cmd->sysfs_dir));
+}
+
 static void _init_logging(struct cmd_context *cmd)
 {
 	int append = 1;
@@ -119,6 +162,7 @@ static void _init_logging(struct cmd_context *cmd)
 	/* Test mode */
 	cmd->default_settings.test =
 	    find_config_tree_int(cmd, "global/test", 0);
+	init_test(cmd->default_settings.test);
 
 	/* Settings for logging to file */
 	if (find_config_tree_int(cmd, "log/overwrite", DEFAULT_OVERWRITE))
@@ -146,7 +190,7 @@ static void _init_logging(struct cmd_context *cmd)
 
 	/* Tell device-mapper about our logging */
 #ifdef DEVMAPPER_SUPPORT
-	dm_log_init(print_log);
+	dm_log_with_errno_init(print_log);
 #endif
 }
 
@@ -154,6 +198,7 @@ static int _process_config(struct cmd_context *cmd)
 {
 	mode_t old_umask;
 	const char *read_ahead;
+	struct stat st;
 
 	/* umask */
 	cmd->default_settings.umask = find_config_tree_int(cmd,
@@ -189,6 +234,8 @@ static int _process_config(struct cmd_context *cmd)
 		cmd->proc_dir[0] = '\0';
 	}
 
+	_get_sysfs_dir(cmd);
+
 	/* activation? */
 	cmd->default_settings.activation = find_config_tree_int(cmd,
 							   "global/activation",
@@ -218,6 +265,34 @@ static int _process_config(struct cmd_context *cmd)
 		return 0;
 	}
 
+	cmd->default_settings.udev_sync = find_config_tree_int(cmd,
+								"activation/udev_sync",
+								DEFAULT_UDEV_SYNC);
+
+	cmd->stripe_filler = find_config_tree_str(cmd,
+						  "activation/missing_stripe_filler",
+						  DEFAULT_STRIPE_FILLER);
+
+	/* FIXME Missing error code checks from the stats, not log_warn?, notify if setting overridden, delay message/check till it is actually used (eg consider if lvm shell - file could appear later after this check)? */
+	if (!strcmp(cmd->stripe_filler, "/dev/ioerror") &&
+	    stat(cmd->stripe_filler, &st))
+		cmd->stripe_filler = "error";
+
+	if (strcmp(cmd->stripe_filler, "error")) {
+		if (stat(cmd->stripe_filler, &st)) {
+			log_warn("WARNING: activation/missing_stripe_filler = \"%s\" "
+				 "is invalid,", cmd->stripe_filler);
+			log_warn("         stat failed: %s", strerror(errno));
+			log_warn("Falling back to \"error\" missing_stripe_filler.");
+			cmd->stripe_filler = "error";
+		} else if (!S_ISBLK(st.st_mode)) {
+			log_warn("WARNING: activation/missing_stripe_filler = \"%s\" "
+				 "is not a block device.", cmd->stripe_filler);
+			log_warn("Falling back to \"error\" missing_stripe_filler.");
+			cmd->stripe_filler = "error";
+		}
+	}
+
 	return 1;
 }
 
@@ -282,10 +357,8 @@ static int _init_tags(struct cmd_context *cmd, struct config_tree *cft)
 	if (!cmd->hosttags && find_config_int(cft->root, "tags/hosttags",
 					      DEFAULT_HOSTTAGS)) {
 		/* FIXME Strip out invalid chars: only A-Za-z0-9_+.- */
-		if (!_set_tag(cmd, cmd->hostname)) {
-			stack;
-			return 0;
-		}
+		if (!_set_tag(cmd, cmd->hostname))
+			return_0;
 		cmd->hosttags = 1;
 	}
 
@@ -301,17 +374,13 @@ static int _init_tags(struct cmd_context *cmd, struct config_tree *cft)
 		}
 		if (cn->child) {
 			passes = 0;
-			if (!_check_host_filters(cmd, cn->child, &passes)) {
-				stack;
-				return 0;
-			}
+			if (!_check_host_filters(cmd, cn->child, &passes))
+				return_0;
 			if (!passes)
 				continue;
 		}
-		if (!_set_tag(cmd, tag)) {
-			stack;
-			return 0;
-		}
+		if (!_set_tag(cmd, tag))
+			return_0;
 	}
 
 	return 1;
@@ -328,7 +397,7 @@ static int _load_config_file(struct cmd_context *cmd, const char *tag)
 		filler = "_";
 
 	if (dm_snprintf(config_file, sizeof(config_file), "%s/lvm%s%s.conf",
-			 cmd->sys_dir, filler, tag) < 0) {
+			 cmd->system_dir, filler, tag) < 0) {
 		log_error("LVM_SYSTEM_DIR or tag was too long");
 		return 0;
 	}
@@ -346,7 +415,7 @@ static int _load_config_file(struct cmd_context *cmd, const char *tag)
 	/* Is there a config file? */
 	if (stat(config_file, &info) == -1) {
 		if (errno == ENOENT) {
-			list_add(&cmd->config_files, &cfl->list);
+			dm_list_add(&cmd->config_files, &cfl->list);
 			goto out;
 		}
 		log_sys_error("stat", config_file);
@@ -361,7 +430,7 @@ static int _load_config_file(struct cmd_context *cmd, const char *tag)
 		return 0;
 	}
 
-	list_add(&cmd->config_files, &cfl->list);
+	dm_list_add(&cmd->config_files, &cfl->list);
 
       out:
 	if (*tag)
@@ -377,7 +446,7 @@ static int _load_config_file(struct cmd_context *cmd, const char *tag)
 static int _init_lvm_conf(struct cmd_context *cmd)
 {
 	/* No config file if LVM_SYSTEM_DIR is empty */
-	if (!*cmd->sys_dir) {
+	if (!*cmd->system_dir) {
 		if (!(cmd->cft = create_config_tree(NULL, 0))) {
 			log_error("Failed to create config tree");
 			return 0;
@@ -385,10 +454,8 @@ static int _init_lvm_conf(struct cmd_context *cmd)
 		return 1;
 	}
 
-	if (!_load_config_file(cmd, "")) {
-		stack;
-		return 0;
-	}
+	if (!_load_config_file(cmd, ""))
+		return_0;
 
 	return 1;
 }
@@ -399,12 +466,9 @@ static int _init_tag_configs(struct cmd_context *cmd)
 	struct str_list *sl;
 
 	/* Tag list may grow while inside this loop */
-	list_iterate_items(sl, &cmd->tags) {
-		if (!_load_config_file(cmd, sl->str)) {
-			stack;
-			return 0;
-		}
-
+	dm_list_iterate_items(sl, &cmd->tags) {
+		if (!_load_config_file(cmd, sl->str))
+			return_0;
 	}
 
 	return 1;
@@ -422,12 +486,10 @@ static int _merge_config_files(struct cmd_context *cmd)
 		}
 	}
 
-	list_iterate_items(cfl, &cmd->config_files) {
+	dm_list_iterate_items(cfl, &cmd->config_files) {
 		/* Merge all config trees into cmd->cft using merge/tag rules */
-		if (!merge_config_tree(cmd, cmd->cft, cfl->cft)) {
-			stack;
-			return 0;
-		}
+		if (!merge_config_tree(cmd, cmd->cft, cfl->cft))
+			return_0;
 	}
 
 	return 1;
@@ -435,10 +497,10 @@ static int _merge_config_files(struct cmd_context *cmd)
 
 static void _destroy_tags(struct cmd_context *cmd)
 {
-	struct list *slh, *slht;
+	struct dm_list *slh, *slht;
 
-	list_iterate_safe(slh, slht, &cmd->tags) {
-		list_del(slh);
+	dm_list_iterate_safe(slh, slht, &cmd->tags) {
+		dm_list_del(slh);
 	}
 }
 
@@ -446,7 +508,7 @@ int config_files_changed(struct cmd_context *cmd)
 {
 	struct config_tree_list *cfl;
 
-	list_iterate_items(cfl, &cmd->config_files) {
+	dm_list_iterate_items(cfl, &cmd->config_files) {
 		if (config_file_changed(cfl->cft))
 			return 1;
 	}
@@ -458,16 +520,18 @@ static void _destroy_tag_configs(struct cmd_context *cmd)
 {
 	struct config_tree_list *cfl;
 
-	if (cmd->cft && cmd->cft->root) {
+	dm_list_iterate_items(cfl, &cmd->config_files) {
+		if (cfl->cft == cmd->cft)
+			cmd->cft = NULL;
+		destroy_config_tree(cfl->cft);
+	}
+
+	if (cmd->cft) {
 		destroy_config_tree(cmd->cft);
 		cmd->cft = NULL;
 	}
 
-	list_iterate_items(cfl, &cmd->config_files) {
-		destroy_config_tree(cfl->cft);
-	}
-
-	list_init(&cmd->config_files);
+	dm_list_init(&cmd->config_files);
 }
 
 static int _init_dev_cache(struct cmd_context *cmd)
@@ -542,12 +606,12 @@ static struct dev_filter *_init_filter_components(struct cmd_context *cmd)
 
 	/*
 	 * sysfs filter. Only available on 2.6 kernels.  Non-critical.
-	 * Listed first because it's very efficient at eliminating 
+	 * Listed first because it's very efficient at eliminating
 	 * unavailable devices.
 	 */
 	if (find_config_tree_bool(cmd, "devices/sysfs_scan",
 			     DEFAULT_SYSFS_SCAN)) {
-		if ((filters[nr_filt] = sysfs_filter_create(cmd->proc_dir)))
+		if ((filters[nr_filt] = sysfs_filter_create(cmd->sysfs_dir)))
 			nr_filt++;
 	}
 
@@ -605,7 +669,7 @@ static int _init_filters(struct cmd_context *cmd, unsigned load_persistent_cache
 	if (cache_dir || cache_file_prefix) {
 		if (dm_snprintf(cache_file, sizeof(cache_file),
 		    "%s%s%s/%s.cache",
-		    cache_dir ? "" : cmd->sys_dir,
+		    cache_dir ? "" : cmd->system_dir,
 		    cache_dir ? "" : "/",
 		    cache_dir ? : DEFAULT_CACHE_SUBDIR,
 		    cache_file_prefix ? : DEFAULT_CACHE_FILE_PREFIX) < 0) {
@@ -615,7 +679,7 @@ static int _init_filters(struct cmd_context *cmd, unsigned load_persistent_cache
 	} else if (!(dev_cache = find_config_tree_str(cmd, "devices/cache", NULL)) &&
 		   (dm_snprintf(cache_file, sizeof(cache_file),
 				"%s/%s/%s.cache",
-				cmd->sys_dir, DEFAULT_CACHE_SUBDIR,
+				cmd->system_dir, DEFAULT_CACHE_SUBDIR,
 				DEFAULT_CACHE_FILE_PREFIX) < 0)) {
 		log_error("Persistent cache filename too long.");
 		return 0;
@@ -633,7 +697,7 @@ static int _init_filters(struct cmd_context *cmd, unsigned load_persistent_cache
 	if (find_config_tree_int(cmd, "devices/write_cache_state", 1))
 		cmd->dump_filter = 1;
 
-	if (!*cmd->sys_dir)
+	if (!*cmd->system_dir)
 		cmd->dump_filter = 0;
 
 	/*
@@ -668,19 +732,19 @@ static int _init_formats(struct cmd_context *cmd)
 	if (!(fmt = init_lvm1_format(cmd)))
 		return 0;
 	fmt->library = NULL;
-	list_add(&cmd->formats, &fmt->list);
+	dm_list_add(&cmd->formats, &fmt->list);
 #endif
 
 #ifdef POOL_INTERNAL
 	if (!(fmt = init_pool_format(cmd)))
 		return 0;
 	fmt->library = NULL;
-	list_add(&cmd->formats, &fmt->list);
+	dm_list_add(&cmd->formats, &fmt->list);
 #endif
 
 #ifdef HAVE_LIBDL
 	/* Load any formats in shared libs if not static */
-	if (!cmd->is_static &&
+	if (!is_static() &&
 	    (cn = find_config_tree_node(cmd, "global/format_libraries"))) {
 
 		struct config_value *cv;
@@ -694,10 +758,8 @@ static int _init_formats(struct cmd_context *cmd)
 				return 0;
 			}
 			if (!(lib = load_shared_library(cmd, cv->v.str,
-							"format", 0))) {
-				stack;
-				return 0;
-			}
+							"format", 0)))
+				return_0;
 
 			if (!(init_format_fn = dlsym(lib, "init_format"))) {
 				log_error("Shared library %s does not contain "
@@ -709,7 +771,7 @@ static int _init_formats(struct cmd_context *cmd)
 			if (!(fmt = init_format_fn(cmd)))
 				return 0;
 			fmt->library = lib;
-			list_add(&cmd->formats, &fmt->list);
+			dm_list_add(&cmd->formats, &fmt->list);
 		}
 	}
 #endif
@@ -717,17 +779,18 @@ static int _init_formats(struct cmd_context *cmd)
 	if (!(fmt = create_text_format(cmd)))
 		return 0;
 	fmt->library = NULL;
-	list_add(&cmd->formats, &fmt->list);
+	dm_list_add(&cmd->formats, &fmt->list);
 
 	cmd->fmt_backup = fmt;
 
 	format = find_config_tree_str(cmd, "global/format",
 				 DEFAULT_FORMAT);
 
-	list_iterate_items(fmt, &cmd->formats) {
+	dm_list_iterate_items(fmt, &cmd->formats) {
 		if (!strcasecmp(fmt->name, format) ||
 		    (fmt->alias && !strcasecmp(fmt->alias, format))) {
 			cmd->default_settings.fmt = fmt;
+			cmd->fmt = cmd->default_settings.fmt;
 			return 1;
 		}
 	}
@@ -736,9 +799,68 @@ static int _init_formats(struct cmd_context *cmd)
 	return 0;
 }
 
+int init_lvmcache_orphans(struct cmd_context *cmd)
+{
+	struct format_type *fmt;
+
+	dm_list_iterate_items(fmt, &cmd->formats)
+		if (!lvmcache_add_orphan_vginfo(fmt->orphan_vg_name, fmt))
+			return_0;
+
+	return 1;
+}
+
+struct segtype_library {
+	struct cmd_context *cmd;
+	void *lib;
+	const char *libname;
+};
+
+int lvm_register_segtype(struct segtype_library *seglib,
+			 struct segment_type *segtype)
+{
+	struct segment_type *segtype2;
+
+	segtype->library = seglib->lib;
+	segtype->cmd = seglib->cmd;
+
+	dm_list_iterate_items(segtype2, &seglib->cmd->segtypes) {
+		if (strcmp(segtype2->name, segtype->name))
+			continue;
+		log_error("Duplicate segment type %s: "
+			  "unloading shared library %s",
+			  segtype->name, seglib->libname);
+		segtype->ops->destroy(segtype);
+		return 0;
+	}
+
+	dm_list_add(&seglib->cmd->segtypes, &segtype->list);
+
+	return 1;
+}
+
+static int _init_single_segtype(struct cmd_context *cmd,
+				struct segtype_library *seglib)
+{
+	struct segment_type *(*init_segtype_fn) (struct cmd_context *);
+	struct segment_type *segtype;
+
+	if (!(init_segtype_fn = dlsym(seglib->lib, "init_segtype"))) {
+		log_error("Shared library %s does not contain segment type "
+			  "functions", seglib->libname);
+		return 0;
+	}
+
+	if (!(segtype = init_segtype_fn(seglib->cmd)))
+		return_0;
+
+	return lvm_register_segtype(seglib, segtype);
+}
+
 static int _init_segtypes(struct cmd_context *cmd)
 {
 	struct segment_type *segtype;
+	struct segtype_library seglib = { .cmd = cmd };
 
 #ifdef HAVE_LIBDL
 	const struct config_node *cn;
@@ -747,41 +869,45 @@ static int _init_segtypes(struct cmd_context *cmd)
 	if (!(segtype = init_striped_segtype(cmd)))
 		return 0;
 	segtype->library = NULL;
-	list_add(&cmd->segtypes, &segtype->list);
+	dm_list_add(&cmd->segtypes, &segtype->list);
 
 	if (!(segtype = init_zero_segtype(cmd)))
 		return 0;
 	segtype->library = NULL;
-	list_add(&cmd->segtypes, &segtype->list);
+	dm_list_add(&cmd->segtypes, &segtype->list);
 
 	if (!(segtype = init_error_segtype(cmd)))
 		return 0;
 	segtype->library = NULL;
-	list_add(&cmd->segtypes, &segtype->list);
+	dm_list_add(&cmd->segtypes, &segtype->list);
+
+	if (!(segtype = init_free_segtype(cmd)))
+		return 0;
+	segtype->library = NULL;
+	dm_list_add(&cmd->segtypes, &segtype->list);
 
 #ifdef SNAPSHOT_INTERNAL
 	if (!(segtype = init_snapshot_segtype(cmd)))
 		return 0;
 	segtype->library = NULL;
-	list_add(&cmd->segtypes, &segtype->list);
+	dm_list_add(&cmd->segtypes, &segtype->list);
 #endif
 
 #ifdef MIRRORED_INTERNAL
 	if (!(segtype = init_mirrored_segtype(cmd)))
 		return 0;
 	segtype->library = NULL;
-	list_add(&cmd->segtypes, &segtype->list);
+	dm_list_add(&cmd->segtypes, &segtype->list);
 #endif
 
 #ifdef HAVE_LIBDL
 	/* Load any formats in shared libs unless static */
-	if (!cmd->is_static &&
+	if (!is_static() &&
 	    (cn = find_config_tree_node(cmd, "global/segment_libraries"))) {
 
 		struct config_value *cv;
-		struct segment_type *(*init_segtype_fn) (struct cmd_context *);
-		void *lib;
-		struct segment_type *segtype2;
+		int (*init_multiple_segtypes_fn) (struct cmd_context *,
+						  struct segtype_library *);
 
 		for (cv = cn->v; cv; cv = cv->next) {
 			if (cv->type != CFG_STRING) {
@@ -789,34 +915,37 @@ static int _init_segtypes(struct cmd_context *cmd)
 					  "global/segment_libraries");
 				return 0;
 			}
-			if (!(lib = load_shared_library(cmd, cv->v.str,
-							"segment type", 0))) {
-				stack;
-				return 0;
-			}
+			seglib.libname = cv->v.str;
+			if (!(seglib.lib = load_shared_library(cmd,
+							seglib.libname,
+							"segment type", 0)))
+				return_0;
 
-			if (!(init_segtype_fn = dlsym(lib, "init_segtype"))) {
-				log_error("Shared library %s does not contain "
-					  "segment type functions", cv->v.str);
-				dlclose(lib);
-				return 0;
-			}
-
-			if (!(segtype = init_segtype_fn(cmd)))
-				return 0;
-			segtype->library = lib;
-			list_add(&cmd->segtypes, &segtype->list);
-
-			list_iterate_items(segtype2, &cmd->segtypes) {
-				if ((segtype == segtype2) ||
-				     strcmp(segtype2->name, segtype->name))
-					continue;
-				log_error("Duplicate segment type %s: "
-					  "unloading shared library %s",
-					  segtype->name, cv->v.str);
-				list_del(&segtype->list);
-				segtype->ops->destroy(segtype);
-				dlclose(lib);
+			if ((init_multiple_segtypes_fn =
+			    dlsym(seglib.lib, "init_multiple_segtypes"))) {
+				if (dlsym(seglib.lib, "init_segtype"))
+					log_warn("WARNING: Shared lib %s has "
+						 "conflicting init fns.  Using"
+						 " init_multiple_segtypes().",
+						 seglib.libname);
+			} else
+				init_multiple_segtypes_fn =
+				    _init_single_segtype;
+ 
+			if (!init_multiple_segtypes_fn(cmd, &seglib)) {
+				struct dm_list *sgtl, *tmp;
+				log_error("init_multiple_segtypes() failed: "
+					  "Unloading shared library %s",
+					  seglib.libname);
+				dm_list_iterate_safe(sgtl, tmp, &cmd->segtypes) {
+					segtype = dm_list_item(sgtl, struct segment_type);
+					if (segtype->library == seglib.lib) {
+						dm_list_del(&segtype->list);
+						segtype->ops->destroy(segtype);
+					}
+				}
+				dlclose(seglib.lib);
+				return_0;
 			}
 		}
 	}
@@ -853,10 +982,10 @@ static int _init_backup(struct cmd_context *cmd)
 	char default_dir[PATH_MAX];
 	const char *dir;
 
-	if (!cmd->sys_dir) {
+	if (!cmd->system_dir) {
 		log_warn("WARNING: Metadata changes will NOT be backed up");
-		backup_init(cmd, "");
-		archive_init(cmd, "", 0, 0);
+		backup_init(cmd, "", 0);
+		archive_init(cmd, "", 0, 0, 0);
 		return 1;
 	}
 
@@ -872,18 +1001,19 @@ static int _init_backup(struct cmd_context *cmd)
 					 DEFAULT_ARCHIVE_NUMBER);
 
 	if (dm_snprintf
-	    (default_dir, sizeof(default_dir), "%s/%s", cmd->sys_dir,
+	    (default_dir, sizeof(default_dir), "%s/%s", cmd->system_dir,
 	     DEFAULT_ARCHIVE_SUBDIR) == -1) {
-		log_err("Couldn't create default archive path '%s/%s'.",
-			cmd->sys_dir, DEFAULT_ARCHIVE_SUBDIR);
+		log_error("Couldn't create default archive path '%s/%s'.",
+			  cmd->system_dir, DEFAULT_ARCHIVE_SUBDIR);
 		return 0;
 	}
 
 	dir = find_config_tree_str(cmd, "backup/archive_dir",
 			      default_dir);
 
-	if (!archive_init(cmd, dir, days, min)) {
-		log_debug("backup_init failed.");
+	if (!archive_init(cmd, dir, days, min,
+			  cmd->default_settings.archive)) {
+		log_debug("archive_init failed.");
 		return 0;
 	}
 
@@ -893,16 +1023,16 @@ static int _init_backup(struct cmd_context *cmd)
 			     DEFAULT_BACKUP_ENABLED);
 
 	if (dm_snprintf
-	    (default_dir, sizeof(default_dir), "%s/%s", cmd->sys_dir,
+	    (default_dir, sizeof(default_dir), "%s/%s", cmd->system_dir,
 	     DEFAULT_BACKUP_SUBDIR) == -1) {
-		log_err("Couldn't create default backup path '%s/%s'.",
-			cmd->sys_dir, DEFAULT_BACKUP_SUBDIR);
+		log_error("Couldn't create default backup path '%s/%s'.",
+			  cmd->system_dir, DEFAULT_BACKUP_SUBDIR);
 		return 0;
 	}
 
 	dir = find_config_tree_str(cmd, "backup/backup_dir", default_dir);
 
-	if (!backup_init(cmd, dir)) {
+	if (!backup_init(cmd, dir, cmd->default_settings.backup)) {
 		log_debug("backup_init failed.");
 		return 0;
 	}
@@ -910,9 +1040,24 @@ static int _init_backup(struct cmd_context *cmd)
 	return 1;
 }
 
+static void _init_rand(struct cmd_context *cmd)
+{
+	if (read_urandom(&cmd->rand_seed, sizeof(cmd->rand_seed)))
+		return;
+
+	cmd->rand_seed = (unsigned) time(NULL) + (unsigned) getpid();
+}
+
+static void _init_globals(struct cmd_context *cmd)
+{
+	init_full_scan_done(0);
+	init_mirror_in_sync(0);
+
+}
+
 /* Entry point */
-struct cmd_context *create_toolcontext(struct arg *the_args, unsigned is_static,
-				       unsigned is_long_lived)
+struct cmd_context *create_toolcontext(unsigned is_long_lived,
+				       const char *system_dir)
 {
 	struct cmd_context *cmd;
 
@@ -934,95 +1079,108 @@ struct cmd_context *create_toolcontext(struct arg *the_args, unsigned is_static,
 		return NULL;
 	}
 	memset(cmd, 0, sizeof(*cmd));
-	cmd->args = the_args;
-	cmd->is_static = is_static;
 	cmd->is_long_lived = is_long_lived;
+	cmd->handles_missing_pvs = 0;
 	cmd->hosttags = 0;
-	list_init(&cmd->formats);
-	list_init(&cmd->segtypes);
-	list_init(&cmd->tags);
-	list_init(&cmd->config_files);
+	dm_list_init(&cmd->formats);
+	dm_list_init(&cmd->segtypes);
+	dm_list_init(&cmd->tags);
+	dm_list_init(&cmd->config_files);
 
-	strcpy(cmd->sys_dir, DEFAULT_SYS_DIR);
+	/* FIXME Make this configurable? */
+	reset_lvm_errno(1);
+
+	/*
+	 * Environment variable LVM_SYSTEM_DIR overrides this below.
+	 */
+        if (system_dir)
+		strncpy(cmd->system_dir, system_dir, sizeof(cmd->system_dir) - 1);
+	else
+		strcpy(cmd->system_dir, DEFAULT_SYS_DIR);
 
 	if (!_get_env_vars(cmd))
-		goto error;
+		goto_out;
 
 	/* Create system directory if it doesn't already exist */
-	if (*cmd->sys_dir && !dm_create_dir(cmd->sys_dir)) {
+	if (*cmd->system_dir && !dm_create_dir(cmd->system_dir)) {
 		log_error("Failed to create LVM2 system dir for metadata backups, config "
 			  "files and internal cache.");
 		log_error("Set environment variable LVM_SYSTEM_DIR to alternative location "
 			  "or empty string.");
-		goto error;
+		goto out;
 	}
 
 	if (!(cmd->libmem = dm_pool_create("library", 4 * 1024))) {
 		log_error("Library memory pool creation failed");
-		goto error;
+		goto out;
 	}
 
 	if (!_init_lvm_conf(cmd))
-		goto error;
+		goto_out;
 
 	_init_logging(cmd);
 
 	if (!_init_hostname(cmd))
-		goto error;
+		goto_out;
 
 	if (!_init_tags(cmd, cmd->cft))
-		goto error;
+		goto_out;
 
 	if (!_init_tag_configs(cmd))
-		goto error;
+		goto_out;
 
 	if (!_merge_config_files(cmd))
-		goto error;
+		goto_out;
 
 	if (!_process_config(cmd))
-		goto error;
+		goto_out;
 
 	if (!_init_dev_cache(cmd))
-		goto error;
+		goto_out;
 
 	if (!_init_filters(cmd, 1))
-		goto error;
+		goto_out;
 
 	if (!(cmd->mem = dm_pool_create("command", 4 * 1024))) {
 		log_error("Command memory pool creation failed");
-		goto error;
+		goto out;
 	}
 
 	memlock_init(cmd);
 
 	if (!_init_formats(cmd))
-		goto error;
+		goto_out;
+
+	if (!init_lvmcache_orphans(cmd))
+		goto_out;
 
 	if (!_init_segtypes(cmd))
-		goto error;
+		goto_out;
 
 	if (!_init_backup(cmd))
-		goto error;
+		goto_out;
 
+	_init_rand(cmd);
+
+	_init_globals(cmd);
+
+	cmd->default_settings.cache_vgmetadata = 1;
 	cmd->current_settings = cmd->default_settings;
 
 	cmd->config_valid = 1;
+out:
 	return cmd;
-
-      error:
-	dm_free(cmd);
-	return NULL;
 }
 
-static void _destroy_formats(struct list *formats)
+static void _destroy_formats(struct dm_list *formats)
 {
-	struct list *fmtl, *tmp;
+	struct dm_list *fmtl, *tmp;
 	struct format_type *fmt;
 	void *lib;
 
-	list_iterate_safe(fmtl, tmp, formats) {
-		fmt = list_item(fmtl, struct format_type);
-		list_del(&fmt->list);
+	dm_list_iterate_safe(fmtl, tmp, formats) {
+		fmt = dm_list_item(fmtl, struct format_type);
+		dm_list_del(&fmt->list);
 		lib = fmt->library;
 		fmt->ops->destroy(fmt);
 #ifdef HAVE_LIBDL
@@ -1032,20 +1190,30 @@ static void _destroy_formats(struct list *formats)
 	}
 }
 
-static void _destroy_segtypes(struct list *segtypes)
+static void _destroy_segtypes(struct dm_list *segtypes)
 {
-	struct list *sgtl, *tmp;
+	struct dm_list *sgtl, *tmp;
 	struct segment_type *segtype;
 	void *lib;
 
-	list_iterate_safe(sgtl, tmp, segtypes) {
-		segtype = list_item(sgtl, struct segment_type);
-		list_del(&segtype->list);
+	dm_list_iterate_safe(sgtl, tmp, segtypes) {
+		segtype = dm_list_item(sgtl, struct segment_type);
+		dm_list_del(&segtype->list);
 		lib = segtype->library;
 		segtype->ops->destroy(segtype);
 #ifdef HAVE_LIBDL
-		if (lib)
+		/*
+		 * If no segtypes remain from this library, close it.
+		 */
+		if (lib) {
+			struct segment_type *segtype2;
+			dm_list_iterate_items(segtype2, segtypes)
+				if (segtype2->library == lib)
+					goto skip_dlclose;
 			dlclose(lib);
+skip_dlclose:
+			;
+		}
 #endif
 	}
 }
@@ -1060,7 +1228,7 @@ int refresh_toolcontext(struct cmd_context *cmd)
 	 */
 
 	activation_release();
-	lvmcache_destroy();
+	lvmcache_destroy(cmd, 0);
 	label_exit();
 	_destroy_segtypes(&cmd->segtypes);
 	_destroy_formats(&cmd->formats);
@@ -1102,6 +1270,9 @@ int refresh_toolcontext(struct cmd_context *cmd)
 	if (!_init_formats(cmd))
 		return 0;
 
+	if (!init_lvmcache_orphans(cmd))
+		return 0;
+
 	if (!_init_segtypes(cmd))
 		return 0;
 
@@ -1113,6 +1284,8 @@ int refresh_toolcontext(struct cmd_context *cmd)
 		persistent_filter_dump(cmd->filter);
 
 	cmd->config_valid = 1;
+
+	reset_lvm_errno(1);
 	return 1;
 }
 
@@ -1123,20 +1296,24 @@ void destroy_toolcontext(struct cmd_context *cmd)
 
 	archive_exit(cmd);
 	backup_exit(cmd);
-	lvmcache_destroy();
+	lvmcache_destroy(cmd, 0);
 	label_exit();
 	_destroy_segtypes(&cmd->segtypes);
 	_destroy_formats(&cmd->formats);
-	cmd->filter->destroy(cmd->filter);
-	dm_pool_destroy(cmd->mem);
+	if (cmd->filter)
+		cmd->filter->destroy(cmd->filter);
+	if (cmd->mem)
+		dm_pool_destroy(cmd->mem);
 	dev_cache_exit();
 	_destroy_tags(cmd);
 	_destroy_tag_configs(cmd);
-	dm_pool_destroy(cmd->libmem);
+	if (cmd->libmem)
+		dm_pool_destroy(cmd->libmem);
 	dm_free(cmd);
 
 	release_log_memory();
 	activation_exit();
 	fin_log();
 	fin_syslog();
+	reset_lvm_errno(0);
 }

lib/commands/toolcontext.h

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -34,6 +34,8 @@ struct config_info {
 	int archive;		/* should we archive ? */
 	int backup;		/* should we backup ? */
 	int read_ahead;		/* DM_READ_AHEAD_NONE or _AUTO */
+	int udev_sync;
+	int cache_vgmetadata;
 	const char *msg_prefix;
 	struct format_type *fmt;
 	uint64_t unit_factor;
@@ -56,22 +58,23 @@ struct cmd_context {
 	const struct format_type *fmt;	/* Current format to use by default */
 	struct format_type *fmt_backup;	/* Format to use for backups */
 
-	struct list formats;	/* Available formats */
-	struct list segtypes;	/* Available segment types */
+	struct dm_list formats;	/* Available formats */
+	struct dm_list segtypes;	/* Available segment types */
 	const char *hostname;
 	const char *kernel_vsn;
 
-	char *cmd_line;
+	unsigned rand_seed;
+	const char *cmd_line;
 	struct command *command;
-	struct arg *args;
 	char **argv;
-	unsigned is_static;	/* Static binary? */
-	unsigned is_long_lived;	/* Optimises persistent_filter handling */
+	unsigned is_long_lived:1;	/* Optimises persistent_filter handling */
+	unsigned handles_missing_pvs:1;
+	unsigned partial_activation:1;
 
 	struct dev_filter *filter;
 	int dump_filter;	/* Dump filter when exiting? */
 
-	struct list config_files;
+	struct dm_list config_files;
 	int config_valid;
 	struct config_tree *cft;
 	struct config_tree *cft_override;
@@ -80,19 +83,27 @@ struct cmd_context {
 
 	struct archive_params *archive_params;
 	struct backup_params *backup_params;
+	const char *stripe_filler;
 
 	/* List of defined tags */
-	struct list tags;
+	struct dm_list tags;
 	int hosttags;
 
-	char sys_dir[PATH_MAX];
+	char system_dir[PATH_MAX];
 	char dev_dir[PATH_MAX];
 	char proc_dir[PATH_MAX];
+	char sysfs_dir[PATH_MAX];
 };
 
-struct cmd_context *create_toolcontext(struct arg *the_args, unsigned is_static, unsigned is_long_lived);
+/*
+ * system_dir may be NULL to use the default value.
+ * The environment variable LVM_SYSTEM_DIR always takes precedence.
+ */
+struct cmd_context *create_toolcontext(unsigned is_long_lived,
+				       const char *system_dir);
 void destroy_toolcontext(struct cmd_context *cmd);
 int refresh_toolcontext(struct cmd_context *cmd);
 int config_files_changed(struct cmd_context *cmd);
+int init_lvmcache_orphans(struct cmd_context *cmd);
 
 #endif

lib/config/config.c

@@ -34,7 +34,8 @@
 enum {
 	TOK_INT,
 	TOK_FLOAT,
-	TOK_STRING,
+	TOK_STRING,		/* Single quotes */
+	TOK_STRING_ESCAPED,	/* Double quotes */
 	TOK_EQ,
 	TOK_SECTION_B,
 	TOK_SECTION_E,
@@ -185,6 +186,17 @@ struct config_tree *create_config_tree_from_string(struct cmd_context *cmd __att
 	return cft;
 }
 
+int override_config_tree_from_string(struct cmd_context *cmd,
+				     const char *config_settings)
+{
+	if (!(cmd->cft_override = create_config_tree_from_string(cmd,config_settings))) {
+		log_error("Failed to set overridden configuration entries.");
+		return 1;
+	}
+
+	return 0;
+}
+
 int read_config_fd(struct config_tree *cft, struct device *dev,
 		   off_t offset, size_t size, off_t offset2, size_t size2,
 		   checksum_fn_t checksum_fn, uint32_t checksum)
@@ -196,10 +208,8 @@ int read_config_fd(struct config_tree *cft, struct device *dev,
 	off_t mmap_offset = 0;
 	char *buf = NULL;
 
-	if (!(p = dm_pool_alloc(c->mem, sizeof(*p)))) {
-		stack;
-		return 0;
-	}
+	if (!(p = dm_pool_alloc(c->mem, sizeof(*p))))
+		return_0;
 	p->mem = c->mem;
 
 	/* Only use mmap with regular files */
@@ -217,10 +227,8 @@ int read_config_fd(struct config_tree *cft, struct device *dev,
 		}
 		p->fb = p->fb + mmap_offset;
 	} else {
-		if (!(buf = dm_malloc(size + size2))) {
-			stack;
-			return 0;
-		}
+		if (!(buf = dm_malloc(size + size2)))
+			return_0;
 		if (!dev_read_circular(dev, (uint64_t) offset, size,
 				       (uint64_t) offset2, size2, buf)) {
 			goto out;
@@ -237,10 +245,8 @@ int read_config_fd(struct config_tree *cft, struct device *dev,
 
 	p->fe = p->fb + size + size2;
 
-	if (!_parse_config_file(p, cft)) {
-		stack;
-		goto out;
-	}
+	if (!_parse_config_file(p, cft))
+		goto_out;
 
 	r = 1;
 
@@ -378,7 +384,7 @@ static int _line_append(struct output_line *outline, const char *fmt, ...)
 	va_end(ap);
 
 	if (!dm_pool_grow_object(outline->mem, &buf[0], strlen(buf))) {
-		log_error("dm_pool_grew_object failed for config line");
+		log_error("dm_pool_grow_object failed for config line");
 		return 0;
 	}
 
@@ -407,9 +413,16 @@ static int _line_end(struct output_line *outline)
 
 static int _write_value(struct output_line *outline, struct config_value *v)
 {
+	char *buf;
+
 	switch (v->type) {
 	case CFG_STRING:
-		line_append("\"%s\"", v->v.str);
+		if (!(buf = alloca(escaped_len(v->v.str)))) {
+			log_error("temporary stack allocation for a config "
+				  "string failed");
+			return 0;
+		}
+		line_append("\"%s\"", escape_double_quotes(buf, v->v.str));
 		break;
 
 	case CFG_FLOAT:
@@ -455,9 +468,9 @@ static int _write_config(struct config_node *n, int only_one,
 			line_append(" {");
 			if (!_line_end(outline))
 				return_0;
+			_write_config(n->child, 0, outline, level + 1);
 			if (!_line_start(outline))
 				return_0;
-			_write_config(n->child, 0, outline, level + 1);
 			line_append("%s}", space);
 		} else {
 			/* it's a value */
@@ -537,15 +550,14 @@ static struct config_node *_file(struct parser *p)
 {
 	struct config_node *root = NULL, *n, *l = NULL;
 	while (p->t != TOK_EOF) {
-		if (!(n = _section(p))) {
-			stack;
-			return 0;
-		}
+		if (!(n = _section(p)))
+			return_0;
 
 		if (!root)
 			root = n;
 		else
 			l->sib = n;
+		n->parent = root;
 		l = n;
 	}
 	return root;
@@ -555,39 +567,32 @@ static struct config_node *_section(struct parser *p)
 {
 	/* IDENTIFIER SECTION_B_CHAR VALUE* SECTION_E_CHAR */
 	struct config_node *root, *n, *l = NULL;
-	if (!(root = _create_node(p))) {
-		stack;
-		return 0;
-	}
+	if (!(root = _create_node(p)))
+		return_0;
 
-	if (!(root->key = _dup_tok(p))) {
-		stack;
-		return 0;
-	}
+	if (!(root->key = _dup_tok(p)))
+		return_0;
 
 	match(TOK_IDENTIFIER);
 
 	if (p->t == TOK_SECTION_B) {
 		match(TOK_SECTION_B);
 		while (p->t != TOK_SECTION_E) {
-			if (!(n = _section(p))) {
-				stack;
-				return 0;
-			}
+			if (!(n = _section(p)))
+				return_0;
 
 			if (!root->child)
 				root->child = n;
 			else
 				l->sib = n;
+			n->parent = root;
 			l = n;
 		}
 		match(TOK_SECTION_E);
 	} else {
 		match(TOK_EQ);
-		if (!(root->v = _value(p))) {
-			stack;
-			return 0;
-		}
+		if (!(root->v = _value(p)))
+			return_0;
 	}
 
 	return root;
@@ -600,10 +605,8 @@ static struct config_value *_value(struct parser *p)
 	if (p->t == TOK_ARRAY_B) {
 		match(TOK_ARRAY_B);
 		while (p->t != TOK_ARRAY_E) {
-			if (!(l = _type(p))) {
-				stack;
-				return 0;
-			}
+			if (!(l = _type(p)))
+				return_0;
 
 			if (!h)
 				h = l;
@@ -656,14 +659,23 @@ static struct config_value *_type(struct parser *p)
 		v->type = CFG_STRING;
 
 		p->tb++, p->te--;	/* strip "'s */
-		if (!(v->v.str = _dup_tok(p))) {
-			stack;
-			return 0;
-		}
+		if (!(v->v.str = _dup_tok(p)))
+			return_0;
 		p->te++;
 		match(TOK_STRING);
 		break;
 
+	case TOK_STRING_ESCAPED:
+		v->type = CFG_STRING;
+
+		p->tb++, p->te--;	/* strip "'s */
+		if (!(v->v.str = _dup_tok(p)))
+			return_0;
+		unescape_double_quotes(v->v.str);
+		p->te++;
+		match(TOK_STRING_ESCAPED);
+		break;
+
 	default:
 		log_error("Parse error at byte %" PRIptrdiff_t " (line %d): expected a value",
 			  p->tb - p->fb + 1, p->line);
@@ -734,7 +746,7 @@ static void _get_token(struct parser *p, int tok_prev)
 		break;
 
 	case '"':
-		p->t = TOK_STRING;
+		p->t = TOK_STRING_ESCAPED;
 		p->te++;
 		while ((p->te != p->fe) && (*p->te) && (*p->te != '"')) {
 			if ((*p->te == '\\') && (p->te + 1 != p->fe) &&
@@ -799,11 +811,9 @@ static void _get_token(struct parser *p, int tok_prev)
 static void _eat_space(struct parser *p)
 {
 	while ((p->tb != p->fe) && (*p->tb)) {
-		if (*p->te == '#') {
+		if (*p->te == '#')
 			while ((p->te != p->fe) && (*p->te) && (*p->te != '\n'))
 				p->te++;
-			p->line++;
-		}
 
 		else if (isspace(*p->te)) {
 			while ((p->te != p->fe) && (*p->te) && isspace(*p->te)) {
@@ -847,10 +857,8 @@ static char *_dup_tok(struct parser *p)
 {
 	size_t len = p->te - p->tb;
 	char *str = dm_pool_alloc(p->mem, len + 1);
-	if (!str) {
-		stack;
-		return 0;
-	}
+	if (!str)
+		return_0;
 	strncpy(str, p->tb, len);
 	str[len] = '\0';
 	return str;
@@ -1016,7 +1024,7 @@ float find_config_tree_float(struct cmd_context *cmd, const char *path,
 	return _find_config_float(cmd->cft_override ? cmd->cft_override->root : NULL, cmd->cft->root, path, fail);
 }
 
-static int _str_in_array(const char *str, const char *values[])
+static int _str_in_array(const char *str, const char * const values[])
 {
 	int i;
 
@@ -1029,9 +1037,8 @@ static int _str_in_array(const char *str, const char *values[])
 
 static int _str_to_bool(const char *str, int fail)
 {
-	static const char *_true_values[] = { "y", "yes", "on", "true", NULL };
-	static const char *_false_values[] =
-	    { "n", "no", "off", "false", NULL };
+	const char * const _true_values[]  = { "y", "yes", "on", "true", NULL };
+	const char * const _false_values[] = { "n", "no", "off", "false", NULL };
 
 	if (_str_in_array(str, _true_values))
 		return 1;
@@ -1171,7 +1178,7 @@ static void _merge_section(struct config_node *cn1, struct config_node *cn2)
 	}
 }
 
-static int _match_host_tags(struct list *tags, struct config_node *tn)
+static int _match_host_tags(struct dm_list *tags, struct config_node *tn)
 {
 	struct config_value *tv;
 	const char *str;
@@ -1254,9 +1261,13 @@ static unsigned _count_tokens(const char *str, unsigned len, int type)
 
 	c = _token_type_to_char(type);
 
-	return count_chars_len(str, len, c);
+	return count_chars(str, len, c);
 }
 
+const char *config_parent_name(const struct config_node *n)
+{
+	return (n->parent ? n->parent->key : "(root)");
+}
 /*
  * Heuristic function to make a quick guess as to whether a text
  * region probably contains a valid config "section".  (Useful for
@@ -1281,7 +1292,7 @@ unsigned maybe_config_section(const char *str, unsigned len)
 	begin_count = _count_tokens(str, len, TOK_SECTION_B);
 	end_count = _count_tokens(str, len, TOK_SECTION_E);
 
-	if (begin_count && end_count && (begin_count - end_count == 0))
+	if (begin_count && end_count && (begin_count == end_count))
 		return 1;
 	else
 		return 0;

lib/config/config.h

@@ -40,7 +40,7 @@ struct config_value {
 
 struct config_node {
 	char *key;
-	struct config_node *sib, *child;
+	struct config_node *parent, *sib, *child;
 	struct config_value *v;
 };
 
@@ -49,13 +49,15 @@ struct config_tree {
 };
 
 struct config_tree_list {
-	struct list list;
+	struct dm_list list;
 	struct config_tree *cft;
 };
 
 struct config_tree *create_config_tree(const char *filename, int keep_open);
 struct config_tree *create_config_tree_from_string(struct cmd_context *cmd,
 						   const char *config_settings);
+int override_config_tree_from_string(struct cmd_context *cmd,
+				     const char *config_settings);
 void destroy_config_tree(struct config_tree *cft);
 
 typedef uint32_t (*checksum_fn_t) (uint32_t initial, const void *buf, uint32_t size);
@@ -110,4 +112,6 @@ int get_config_str(const struct config_node *cn, const char *path,
 
 unsigned maybe_config_section(const char *str, unsigned len);
 
+const char *config_parent_name(const struct config_node *n);
+
 #endif

lib/config/defaults.h

@@ -1,6 +1,6 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
+ * Copyright (C) 2004-2009 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -32,12 +32,16 @@
 #define DEFAULT_PROC_DIR "/proc"
 #define DEFAULT_SYSFS_SCAN 1
 #define DEFAULT_MD_COMPONENT_DETECTION 1
+#define DEFAULT_MD_CHUNK_ALIGNMENT 1
 #define DEFAULT_IGNORE_SUSPENDED_DEVICES 1
+#define DEFAULT_DATA_ALIGNMENT_OFFSET_DETECTION 1
+#define DEFAULT_DATA_ALIGNMENT_DETECTION 1
 
 #define DEFAULT_LOCK_DIR "/var/lock/lvm"
 #define DEFAULT_LOCKING_LIB "liblvm2clusterlock.so"
 #define DEFAULT_FALLBACK_TO_LOCAL_LOCKING 1
 #define DEFAULT_FALLBACK_TO_CLUSTERED_LOCKING 1
+#define DEFAULT_WAIT_FOR_LOCKS 1
 
 #define DEFAULT_MIRRORLOG "disk"
 #define DEFAULT_MIRROR_LOG_FAULT_POLICY "allocate"
@@ -64,6 +68,12 @@
 #define DEFAULT_PVMETADATACOPIES 1
 #define DEFAULT_LABELSECTOR UINT64_C(1)
 #define DEFAULT_READ_AHEAD "auto"
+#define DEFAULT_UDEV_SYNC 0
+#define DEFAULT_EXTENT_SIZE 4096	/* In KB */
+#define DEFAULT_MAX_PV 0
+#define DEFAULT_MAX_LV 0
+#define DEFAULT_ALLOC_POLICY ALLOC_NORMAL
+#define DEFAULT_CLUSTERED 0
 
 #define DEFAULT_MSG_PREFIX "  "
 #define DEFAULT_CMD_NAME 0
@@ -90,7 +100,7 @@
 #  define DEFAULT_ACTIVATION 0
 #endif
 
-#define DEFAULT_STRIPE_FILLER "/dev/ioerror"
+#define DEFAULT_STRIPE_FILLER "error"
 #define DEFAULT_MIRROR_REGION_SIZE 512	/* KB */
 #define DEFAULT_INTERVAL 15
 
@@ -100,20 +110,23 @@
 
 #define DEFAULT_REP_ALIGNED 1
 #define DEFAULT_REP_BUFFERED 1
+#define DEFAULT_REP_COLUMNS_AS_ROWS 0
 #define DEFAULT_REP_HEADINGS 1
+#define DEFAULT_REP_PREFIXES 0
+#define DEFAULT_REP_QUOTED 1
 #define DEFAULT_REP_SEPARATOR " "
 
-#define DEFAULT_LVS_COLS "lv_name,vg_name,lv_attr,lv_size,origin,snap_percent,move_pv,mirror_log,copy_percent"
+#define DEFAULT_LVS_COLS "lv_name,vg_name,lv_attr,lv_size,origin,snap_percent,move_pv,mirror_log,copy_percent,convert_lv"
 #define DEFAULT_VGS_COLS "vg_name,pv_count,lv_count,snap_count,vg_attr,vg_size,vg_free"
 #define DEFAULT_PVS_COLS "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free"
 #define DEFAULT_SEGS_COLS "lv_name,vg_name,lv_attr,stripes,segtype,seg_size"
 #define DEFAULT_PVSEGS_COLS "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free,pvseg_start,pvseg_size"
 
-#define DEFAULT_LVS_COLS_VERB "lv_name,vg_name,seg_count,lv_attr,lv_size,lv_major,lv_minor,lv_kernel_major,lv_kernel_minor,origin,snap_percent,move_pv,copy_percent,mirror_log,lv_uuid"
+#define DEFAULT_LVS_COLS_VERB "lv_name,vg_name,seg_count,lv_attr,lv_size,lv_major,lv_minor,lv_kernel_major,lv_kernel_minor,origin,snap_percent,move_pv,copy_percent,mirror_log,convert_lv,lv_uuid"
 #define DEFAULT_VGS_COLS_VERB "vg_name,vg_attr,vg_extent_size,pv_count,lv_count,snap_count,vg_size,vg_free,vg_uuid"
 #define DEFAULT_PVS_COLS_VERB "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free,dev_size,pv_uuid"
 #define DEFAULT_SEGS_COLS_VERB "lv_name,vg_name,lv_attr,seg_start,seg_size,stripes,segtype,stripesize,chunksize"
-#define DEFAULT_PVSEGS_COLS_VERB "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free,pvseg_start,pvseg_size"
+#define DEFAULT_PVSEGS_COLS_VERB "pv_name,vg_name,pv_fmt,pv_attr,pv_size,pv_free,pvseg_start,pvseg_size,lv_name,seg_start_pe,segtype,seg_pe_ranges"
 
 #define DEFAULT_LVS_SORT "vg_name,lv_name"
 #define DEFAULT_VGS_SORT "vg_name"
@@ -121,4 +134,7 @@
 #define DEFAULT_SEGS_SORT "vg_name,lv_name,seg_start"
 #define DEFAULT_PVSEGS_SORT "pv_name,pvseg_start"
 
+#define DEFAULT_MIRROR_DEVICE_FAULT_POLICY "remove"
+#define DEFAULT_MIRROR_LOG_FAULT_POLICY "allocate"
+
 #endif				/* _LVM_DEFAULTS_H */

lib/datastruct/btree.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -87,10 +87,8 @@ int btree_insert(struct btree *t, uint32_t k, void *data)
 	struct node *p, **c = _lookup(&t->root, key, &p), *n;
 
 	if (!*c) {
-		if (!(n = dm_pool_alloc(t->mem, sizeof(*n)))) {
-			stack;
-			return 0;
-		}
+		if (!(n = dm_pool_alloc(t->mem, sizeof(*n))))
+			return_0;
 
 		n->key = key;
 		n->data = data;

lib/datastruct/list.c

@@ -19,7 +19,7 @@
  * Initialise a list before use.
  * The list head's next and previous pointers point back to itself.
  */
-void list_init(struct list *head)
+void dm_list_init(struct dm_list *head)
 {
 	head->n = head->p = head;
 }
@@ -28,7 +28,7 @@ void list_init(struct list *head)
  * Insert an element before 'head'.
  * If 'head' is the list head, this adds an element to the end of the list.
  */
-void list_add(struct list *head, struct list *elem)
+void dm_list_add(struct dm_list *head, struct dm_list *elem)
 {
 	assert(head->n);
 
@@ -43,7 +43,7 @@ void list_add(struct list *head, struct list *elem)
  * Insert an element after 'head'.
  * If 'head' is the list head, this adds an element to the front of the list.
  */
-void list_add_h(struct list *head, struct list *elem)
+void dm_list_add_h(struct dm_list *head, struct dm_list *elem)
 {
 	assert(head->n);
 
@@ -59,16 +59,25 @@ void list_add_h(struct list *head, struct list *elem)
  * Note that this doesn't change the element itself - it may still be safe
  * to follow its pointers.
  */
-void list_del(struct list *elem)
+void dm_list_del(struct dm_list *elem)
 {
 	elem->n->p = elem->p;
 	elem->p->n = elem->n;
 }
 
+/*
+ * Remove an element from existing list and insert before 'head'.
+ */
+void dm_list_move(struct dm_list *head, struct dm_list *elem)
+{
+        dm_list_del(elem);
+        dm_list_add(head, elem);
+}
+
 /*
  * Is the list empty?
  */
-int list_empty(const struct list *head)
+int dm_list_empty(const struct dm_list *head)
 {
 	return head->n == head;
 }
@@ -76,7 +85,7 @@ int list_empty(const struct list *head)
 /*
  * Is this the first element of the list?
  */
-int list_start(const struct list *head, const struct list *elem)
+int dm_list_start(const struct dm_list *head, const struct dm_list *elem)
 {
 	return elem->p == head;
 }
@@ -84,7 +93,7 @@ int list_start(const struct list *head, const struct list *elem)
 /*
  * Is this the last element of the list?
  */
-int list_end(const struct list *head, const struct list *elem)
+int dm_list_end(const struct dm_list *head, const struct dm_list *elem)
 {
 	return elem->n == head;
 }
@@ -92,44 +101,44 @@ int list_end(const struct list *head, const struct list *elem)
 /*
  * Return first element of the list or NULL if empty
  */
-struct list *list_first(const struct list *head)
+struct dm_list *dm_list_first(const struct dm_list *head)
 {
-	return (list_empty(head) ? NULL : head->n);
+	return (dm_list_empty(head) ? NULL : head->n);
 }
 
 /*
  * Return last element of the list or NULL if empty
  */
-struct list *list_last(const struct list *head)
+struct dm_list *dm_list_last(const struct dm_list *head)
 {
-	return (list_empty(head) ? NULL : head->p);
+	return (dm_list_empty(head) ? NULL : head->p);
 }
 
 /*
  * Return the previous element of the list, or NULL if we've reached the start.
  */
-struct list *list_prev(const struct list *head, const struct list *elem)
+struct dm_list *dm_list_prev(const struct dm_list *head, const struct dm_list *elem)
 {
-	return (list_start(head, elem) ? NULL : elem->p);
+	return (dm_list_start(head, elem) ? NULL : elem->p);
 }
 
 /*
  * Return the next element of the list, or NULL if we've reached the end.
  */
-struct list *list_next(const struct list *head, const struct list *elem)
+struct dm_list *dm_list_next(const struct dm_list *head, const struct dm_list *elem)
 {
-	return (list_end(head, elem) ? NULL : elem->n);
+	return (dm_list_end(head, elem) ? NULL : elem->n);
 }
 
 /*
  * Return the number of elements in a list by walking it.
  */
-unsigned int list_size(const struct list *head)
+unsigned int dm_list_size(const struct dm_list *head)
 {
 	unsigned int s = 0;
-	const struct list *v;
+	const struct dm_list *v;
 
-	list_iterate(v, head)
+	dm_list_iterate(v, head)
 	    s++;
 
 	return s;

lib/datastruct/list.h

@@ -24,101 +24,106 @@
  * The list head's pointers point to the first and the last element.
  */
 
-struct list {
-	struct list *n, *p;
+struct dm_list {
+	struct dm_list *n, *p;
 };
 
 /*
  * Initialise a list before use.
  * The list head's next and previous pointers point back to itself.
  */
-#define LIST_INIT(name)	struct list name = { &(name), &(name) }
-void list_init(struct list *head);
+#define DM_LIST_INIT(name)	struct dm_list name = { &(name), &(name) }
+void dm_list_init(struct dm_list *head);
 
 /*
  * Insert an element before 'head'.
  * If 'head' is the list head, this adds an element to the end of the list.
  */
-void list_add(struct list *head, struct list *elem);
+void dm_list_add(struct dm_list *head, struct dm_list *elem);
 
 /*
  * Insert an element after 'head'.
  * If 'head' is the list head, this adds an element to the front of the list.
  */
-void list_add_h(struct list *head, struct list *elem);
+void dm_list_add_h(struct dm_list *head, struct dm_list *elem);
 
 /*
  * Delete an element from its list.
  * Note that this doesn't change the element itself - it may still be safe
  * to follow its pointers.
  */
-void list_del(struct list *elem);
+void dm_list_del(struct dm_list *elem);
+
+/*
+ * Remove an element from existing list and insert before 'head'.
+ */
+void dm_list_move(struct dm_list *head, struct dm_list *elem);
 
 /*
  * Is the list empty?
  */
-int list_empty(const struct list *head);
+int dm_list_empty(const struct dm_list *head);
 
 /*
  * Is this the first element of the list?
  */
-int list_start(const struct list *head, const struct list *elem);
+int dm_list_start(const struct dm_list *head, const struct dm_list *elem);
 
 /*
  * Is this the last element of the list?
  */
-int list_end(const struct list *head, const struct list *elem);
+int dm_list_end(const struct dm_list *head, const struct dm_list *elem);
 
 /*
  * Return first element of the list or NULL if empty
  */
-struct list *list_first(const struct list *head);
+struct dm_list *dm_list_first(const struct dm_list *head);
 
 /*
  * Return last element of the list or NULL if empty
  */
-struct list *list_last(const struct list *head);
+struct dm_list *dm_list_last(const struct dm_list *head);
 
 /*
  * Return the previous element of the list, or NULL if we've reached the start.
  */
-struct list *list_prev(const struct list *head, const struct list *elem);
+struct dm_list *dm_list_prev(const struct dm_list *head, const struct dm_list *elem);
 
 /*
  * Return the next element of the list, or NULL if we've reached the end.
  */
-struct list *list_next(const struct list *head, const struct list *elem);
+struct dm_list *dm_list_next(const struct dm_list *head, const struct dm_list *elem);
 
 /*
- * Given the address v of an instance of 'struct list' called 'head' 
+ * Given the address v of an instance of 'struct dm_list' called 'head' 
  * contained in a structure of type t, return the containing structure.
  */
-#define list_struct_base(v, t, head) \
+#define dm_list_struct_base(v, t, head) \
     ((t *)((uintptr_t)(v) - (uintptr_t)&((t *) 0)->head))
 
 /*
- * Given the address v of an instance of 'struct list list' contained in
+ * Given the address v of an instance of 'struct dm_list list' contained in
  * a structure of type t, return the containing structure.
  */
-#define list_item(v, t) list_struct_base((v), t, list)
+#define dm_list_item(v, t) dm_list_struct_base((v), t, list)
 
 /*
  * Given the address v of one known element e in a known structure of type t,
  * return another element f.
  */
-#define struct_field(v, t, e, f) \
+#define dm_struct_field(v, t, e, f) \
     (((t *)((uintptr_t)(v) - (uintptr_t)&((t *) 0)->e))->f)
 
 /*
  * Given the address v of a known element e in a known structure of type t,
  * return the list head 'list'
  */
-#define list_head(v, t, e) struct_field(v, t, e, list)
+#define dm_list_head(v, t, e) dm_struct_field(v, t, e, list)
 
 /*
  * Set v to each element of a list in turn.
  */
-#define list_iterate(v, head) \
+#define dm_list_iterate(v, head) \
 	for (v = (head)->n; v != head; v = v->n)
 
 /*
@@ -128,7 +133,7 @@ struct list *list_next(const struct list *head, const struct list *elem);
  * already-processed elements.
  * If 'start' is 'head' it walks the list backwards.
  */
-#define list_uniterate(v, head, start) \
+#define dm_list_uniterate(v, head, start) \
 	for (v = (start)->p; v != head; v = v->p)
 
 /*
@@ -136,68 +141,68 @@ struct list *list_next(const struct list *head, const struct list *elem);
  * the way.
  * t must be defined as a temporary variable of the same type as v.
  */
-#define list_iterate_safe(v, t, head) \
+#define dm_list_iterate_safe(v, t, head) \
 	for (v = (head)->n, t = v->n; v != head; v = t, t = v->n)
 
 /*
  * Walk a list, setting 'v' in turn to the containing structure of each item.
  * The containing structure should be the same type as 'v'.
- * The 'struct list' variable within the containing structure is 'field'.
+ * The 'struct dm_list' variable within the containing structure is 'field'.
  */
-#define list_iterate_items_gen(v, head, field) \
-	for (v = list_struct_base((head)->n, typeof(*v), field); \
+#define dm_list_iterate_items_gen(v, head, field) \
+	for (v = dm_list_struct_base((head)->n, typeof(*v), field); \
 	     &v->field != (head); \
-	     v = list_struct_base(v->field.n, typeof(*v), field))
+	     v = dm_list_struct_base(v->field.n, typeof(*v), field))
 
 /*
  * Walk a list, setting 'v' in turn to the containing structure of each item.
  * The containing structure should be the same type as 'v'.
- * The list should be 'struct list list' within the containing structure.
+ * The list should be 'struct dm_list list' within the containing structure.
  */
-#define list_iterate_items(v, head) list_iterate_items_gen(v, (head), list)
+#define dm_list_iterate_items(v, head) dm_list_iterate_items_gen(v, (head), list)
 
 /*
  * Walk a list, setting 'v' in turn to the containing structure of each item.
  * The containing structure should be the same type as 'v'.
- * The 'struct list' variable within the containing structure is 'field'.
+ * The 'struct dm_list' variable within the containing structure is 'field'.
  * t must be defined as a temporary variable of the same type as v.
  */
-#define list_iterate_items_gen_safe(v, t, head, field) \
-	for (v = list_struct_base((head)->n, typeof(*v), field), \
-	     t = list_struct_base(v->field.n, typeof(*v), field); \
+#define dm_list_iterate_items_gen_safe(v, t, head, field) \
+	for (v = dm_list_struct_base((head)->n, typeof(*v), field), \
+	     t = dm_list_struct_base(v->field.n, typeof(*v), field); \
 	     &v->field != (head); \
-	     v = t, t = list_struct_base(v->field.n, typeof(*v), field))
+	     v = t, t = dm_list_struct_base(v->field.n, typeof(*v), field))
 /*
  * Walk a list, setting 'v' in turn to the containing structure of each item.
  * The containing structure should be the same type as 'v'.
- * The list should be 'struct list list' within the containing structure.
+ * The list should be 'struct dm_list list' within the containing structure.
  * t must be defined as a temporary variable of the same type as v.
  */
-#define list_iterate_items_safe(v, t, head) \
-	list_iterate_items_gen_safe(v, t, (head), list)
+#define dm_list_iterate_items_safe(v, t, head) \
+	dm_list_iterate_items_gen_safe(v, t, (head), list)
 
 /*
  * Walk a list backwards, setting 'v' in turn to the containing structure 
  * of each item.
  * The containing structure should be the same type as 'v'.
- * The 'struct list' variable within the containing structure is 'field'.
+ * The 'struct dm_list' variable within the containing structure is 'field'.
  */
-#define list_iterate_back_items_gen(v, head, field) \
-	for (v = list_struct_base((head)->p, typeof(*v), field); \
+#define dm_list_iterate_back_items_gen(v, head, field) \
+	for (v = dm_list_struct_base((head)->p, typeof(*v), field); \
 	     &v->field != (head); \
-	     v = list_struct_base(v->field.p, typeof(*v), field))
+	     v = dm_list_struct_base(v->field.p, typeof(*v), field))
 
 /*
  * Walk a list backwards, setting 'v' in turn to the containing structure 
  * of each item.
  * The containing structure should be the same type as 'v'.
- * The list should be 'struct list list' within the containing structure.
+ * The list should be 'struct dm_list list' within the containing structure.
  */
-#define list_iterate_back_items(v, head) list_iterate_back_items_gen(v, (head), list)
+#define dm_list_iterate_back_items(v, head) dm_list_iterate_back_items_gen(v, (head), list)
 
 /*
  * Return the number of elements in a list by walking it.
  */
-unsigned int list_size(const struct list *head);
+unsigned int dm_list_size(const struct dm_list *head);
 
 #endif

lib/datastruct/lvm-types.h

@@ -16,17 +16,16 @@
 #ifndef _LVM_TYPES_H
 #define _LVM_TYPES_H
 
-#include "list.h"
-
 #include <sys/types.h>
 #include <inttypes.h>
 
 /* Define some portable printing types */
 #define PRIsize_t "zu"
 #define PRIptrdiff_t "td"
+#define PRIpid_t PRId32
 
 struct str_list {
-	struct list list;
+	struct dm_list list;
 	const char *str;
 };
 

lib/datastruct/str_list.c

@@ -16,68 +16,62 @@
 #include "lib.h"
 #include "str_list.h"
 
-struct list *str_list_create(struct dm_pool *mem)
+struct dm_list *str_list_create(struct dm_pool *mem)
 {
-	struct list *sl;
+	struct dm_list *sl;
 
-	if (!(sl = dm_pool_alloc(mem, sizeof(struct list)))) {
-		stack;
+	if (!(sl = dm_pool_alloc(mem, sizeof(struct dm_list)))) {
+		log_errno(ENOMEM, "str_list allocation failed");
 		return NULL;
 	}
 
-	list_init(sl);
+	dm_list_init(sl);
 
 	return sl;
 }
 
-int str_list_add(struct dm_pool *mem, struct list *sll, const char *str)
+int str_list_add(struct dm_pool *mem, struct dm_list *sll, const char *str)
 {
 	struct str_list *sln;
 
-	if (!str) {
-		stack;
-		return 0;
-	}
+	if (!str)
+		return_0;
 
 	/* Already in list? */
 	if (str_list_match_item(sll, str))
 		return 1;
 
-	if (!(sln = dm_pool_alloc(mem, sizeof(*sln)))) {
-		stack;
-		return 0;
-	}
+	if (!(sln = dm_pool_alloc(mem, sizeof(*sln))))
+		return_0;
 
 	sln->str = str;
-	list_add(sll, &sln->list);
+	dm_list_add(sll, &sln->list);
 
 	return 1;
 }
 
-int str_list_del(struct list *sll, const char *str)
+int str_list_del(struct dm_list *sll, const char *str)
 {
-	struct list *slh, *slht;
+	struct dm_list *slh, *slht;
 
-	list_iterate_safe(slh, slht, sll) {
-		if (!strcmp(str, list_item(slh, struct str_list)->str))
-			 list_del(slh);
+	dm_list_iterate_safe(slh, slht, sll) {
+		if (!strcmp(str, dm_list_item(slh, struct str_list)->str))
+			 dm_list_del(slh);
 	}
 
 	return 1;
 }
 
-int str_list_dup(struct dm_pool *mem, struct list *sllnew,
-		 const struct list *sllold)
+int str_list_dup(struct dm_pool *mem, struct dm_list *sllnew,
+		 const struct dm_list *sllold)
 {
 	struct str_list *sl;
 
-	list_init(sllnew);
+	dm_list_init(sllnew);
 
-	list_iterate_items(sl, sllold) {
-		if (!str_list_add(mem, sllnew, dm_pool_strdup(mem, sl->str))) {
-			stack;
-			return 0;
-		}
+	dm_list_iterate_items(sl, sllold) {
+		if (!str_list_add(mem, sllnew, dm_pool_strdup(mem, sl->str)))
+			return_0;
 	}
 
 	return 1;
@@ -86,11 +80,11 @@ int str_list_dup(struct dm_pool *mem, struct list *sllnew,
 /*
  * Is item on list?
  */
-int str_list_match_item(const struct list *sll, const char *str)
+int str_list_match_item(const struct dm_list *sll, const char *str)
 {
 	struct str_list *sl;
 
-	list_iterate_items(sl, sll)
+	dm_list_iterate_items(sl, sll)
 	    if (!strcmp(str, sl->str))
 		return 1;
 
@@ -100,11 +94,11 @@ int str_list_match_item(const struct list *sll, const char *str)
 /*
  * Is at least one item on both lists?
  */
-int str_list_match_list(const struct list *sll, const struct list *sll2)
+int str_list_match_list(const struct dm_list *sll, const struct dm_list *sll2)
 {
 	struct str_list *sl;
 
-	list_iterate_items(sl, sll)
+	dm_list_iterate_items(sl, sll)
 	    if (str_list_match_item(sll2, sl->str))
 		return 1;
 
@@ -114,14 +108,14 @@ int str_list_match_list(const struct list *sll, const struct list *sll2)
 /*
  * Do both lists contain the same set of items?
  */
-int str_list_lists_equal(const struct list *sll, const struct list *sll2)
+int str_list_lists_equal(const struct dm_list *sll, const struct dm_list *sll2)
 {
 	struct str_list *sl;
 
-	if (list_size(sll) != list_size(sll2))
+	if (dm_list_size(sll) != dm_list_size(sll2))
 		return 0;
 
-	list_iterate_items(sl, sll)
+	dm_list_iterate_items(sl, sll)
 	    if (!str_list_match_item(sll2, sl->str))
 		return 0;
 

lib/datastruct/str_list.h

@@ -16,13 +16,13 @@
 #ifndef _LVM_STR_LIST_H
 #define _LVM_STR_LIST_H
 
-struct list *str_list_create(struct dm_pool *mem);
-int str_list_add(struct dm_pool *mem, struct list *sll, const char *str);
-int str_list_del(struct list *sll, const char *str);
-int str_list_match_item(const struct list *sll, const char *str);
-int str_list_match_list(const struct list *sll, const struct list *sll2);
-int str_list_lists_equal(const struct list *sll, const struct list *sll2);
-int str_list_dup(struct dm_pool *mem, struct list *sllnew,
-		 const struct list *sllold);
+struct dm_list *str_list_create(struct dm_pool *mem);
+int str_list_add(struct dm_pool *mem, struct dm_list *sll, const char *str);
+int str_list_del(struct dm_list *sll, const char *str);
+int str_list_match_item(const struct dm_list *sll, const char *str);
+int str_list_match_list(const struct dm_list *sll, const struct dm_list *sll2);
+int str_list_lists_equal(const struct dm_list *sll, const struct dm_list *sll2);
+int str_list_dup(struct dm_pool *mem, struct dm_list *sllnew,
+		 const struct dm_list *sllold);
 
 #endif

lib/device/dev-cache.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -31,7 +31,7 @@ struct dev_iter {
 };
 
 struct dir_list {
-	struct list list;
+	struct dm_list list;
 	char dir[0];
 };
 
@@ -42,8 +42,8 @@ static struct {
 	struct dm_regex *preferred_names_matcher;
 
 	int has_scanned;
-	struct list dirs;
-	struct list files;
+	struct dm_list dirs;
+	struct dm_list files;
 
 } _cache;
 
@@ -97,15 +97,16 @@ struct device *dev_create_file(const char *filename, struct device *dev,
 	}
 
 	dev->flags |= DEV_REGULAR;
-	list_init(&dev->aliases);
-	list_add(&dev->aliases, &alias->list);
+	dm_list_init(&dev->aliases);
+	dm_list_add(&dev->aliases, &alias->list);
 	dev->end = UINT64_C(0);
 	dev->dev = 0;
 	dev->fd = -1;
 	dev->open_count = 0;
 	dev->block_size = -1;
+	dev->read_ahead = -1;
 	memset(dev->pvid, 0, sizeof(dev->pvid));
-	list_init(&dev->open_list);
+	dm_list_init(&dev->open_list);
 
 	return dev;
 }
@@ -119,14 +120,15 @@ static struct device *_dev_create(dev_t d)
 		return NULL;
 	}
 	dev->flags = 0;
-	list_init(&dev->aliases);
+	dm_list_init(&dev->aliases);
 	dev->dev = d;
 	dev->fd = -1;
 	dev->open_count = 0;
 	dev->block_size = -1;
+	dev->read_ahead = -1;
 	dev->end = UINT64_C(0);
 	memset(dev->pvid, 0, sizeof(dev->pvid));
-	list_init(&dev->open_list);
+	dm_list_init(&dev->open_list);
 
 	return dev;
 }
@@ -140,8 +142,8 @@ void dev_set_preferred_name(struct str_list *sl, struct device *dev)
 		return;
 
 	log_debug("%s: New preferred name", sl->str);
-	list_del(&sl->list);
-	list_add_h(&dev->aliases, &sl->list);
+	dm_list_del(&sl->list);
+	dm_list_add_h(&dev->aliases, &sl->list);
 }
 
 /* Return 1 if we prefer path1 else return 0 */
@@ -236,26 +238,22 @@ static int _add_alias(struct device *dev, const char *path)
 	const char *oldpath;
 	int prefer_old = 1;
 
-	if (!sl) {
-		stack;
-		return 0;
-	}
+	if (!sl)
+		return_0;
 
 	/* Is name already there? */
-	list_iterate_items(strl, &dev->aliases) {
+	dm_list_iterate_items(strl, &dev->aliases) {
 		if (!strcmp(strl->str, path)) {
 			log_debug("%s: Already in device cache", path);
 			return 1;
 		}
 	}
 
-	if (!(sl->str = dm_pool_strdup(_cache.mem, path))) {
-		stack;
-		return 0;
-	}
+	if (!(sl->str = dm_pool_strdup(_cache.mem, path)))
+		return_0;
 
-	if (!list_empty(&dev->aliases)) {
-		oldpath = list_item(dev->aliases.n, struct str_list)->str;
+	if (!dm_list_empty(&dev->aliases)) {
+		oldpath = dm_list_item(dev->aliases.n, struct str_list)->str;
 		prefer_old = _compare_paths(path, oldpath);
 		log_debug("%s: Aliased to %s in device cache%s",
 			  path, oldpath, prefer_old ? "" : " (preferred name)");
@@ -264,9 +262,9 @@ static int _add_alias(struct device *dev, const char *path)
 		log_debug("%s: Added to device cache", path);
 
 	if (prefer_old)
-		list_add(&dev->aliases, &sl->list);
+		dm_list_add(&dev->aliases, &sl->list);
 	else
-		list_add_h(&dev->aliases, &sl->list);
+		dm_list_add_h(&dev->aliases, &sl->list);
 
 	return 1;
 }
@@ -294,29 +292,25 @@ static int _insert_dev(const char *path, dev_t d)
 						   (uint32_t) d))) {
 		/* create new device */
 		if (loopfile) {
-			if (!(dev = dev_create_file(path, NULL, NULL, 0))) {
-				stack;
-				return 0;
-			}
-		} else if (!(dev = _dev_create(d))) {
-			stack;
-			return 0;
-		}
+			if (!(dev = dev_create_file(path, NULL, NULL, 0)))
+				return_0;
+		} else if (!(dev = _dev_create(d)))
+			return_0;
 
 		if (!(btree_insert(_cache.devices, (uint32_t) d, dev))) {
-			log_err("Couldn't insert device into binary tree.");
+			log_error("Couldn't insert device into binary tree.");
 			_free(dev);
 			return 0;
 		}
 	}
 
 	if (!loopfile && !_add_alias(dev, path)) {
-		log_err("Couldn't add alias to dev cache.");
+		log_error("Couldn't add alias to dev cache.");
 		return 0;
 	}
 
 	if (!dm_hash_insert(_cache.names, path, dev)) {
-		log_err("Couldn't add name to hash in dev cache.");
+		log_error("Couldn't add name to hash in dev cache.");
 		return 0;
 	}
 
@@ -369,10 +363,8 @@ static int _insert_dir(const char *dir)
 				continue;
 			}
 
-			if (!(path = _join(dir, dirent[n]->d_name))) {
-				stack;
-				return 0;
-			}
+			if (!(path = _join(dir, dirent[n]->d_name)))
+				return_0;
 
 			_collapse_slashes(path);
 			r &= _insert(path, 1);
@@ -400,10 +392,8 @@ static int _insert_file(const char *path)
 		return 0;
 	}
 
-	if (!_insert_dev(path, 0)) {
-		stack;
-		return 0;
-	}
+	if (!_insert_dev(path, 0))
+		return_0;
 
 	return 1;
 }
@@ -439,10 +429,8 @@ static int _insert(const char *path, int rec)
 			return 0;
 		}
 
-		if (!_insert_dev(path, info.st_rdev)) {
-			stack;
-			return 0;
-		}
+		if (!_insert_dev(path, info.st_rdev))
+			return_0;
 
 		r = 1;
 	}
@@ -457,10 +445,10 @@ static void _full_scan(int dev_scan)
 	if (_cache.has_scanned && !dev_scan)
 		return;
 
-	list_iterate_items(dl, &_cache.dirs)
+	dm_list_iterate_items(dl, &_cache.dirs)
 		_insert_dir(dl->dir);
 
-	list_iterate_items(dl, &_cache.files)
+	dm_list_iterate_items(dl, &_cache.files)
 		_insert_file(dl->dir);
 
 	_cache.has_scanned = 1;
@@ -502,7 +490,7 @@ static int _init_preferred_names(struct cmd_context *cmd)
 		if (v->type != CFG_STRING) {
 			log_error("preferred_names patterns must be enclosed in quotes");
 			return 0;
-		}       
+		}
 
 		count++;
 	}
@@ -543,25 +531,22 @@ int dev_cache_init(struct cmd_context *cmd)
 	_cache.names = NULL;
 	_cache.has_scanned = 0;
 
-	if (!(_cache.mem = dm_pool_create("dev_cache", 10 * 1024))) {
-		stack;
-		return 0;
-	}
+	if (!(_cache.mem = dm_pool_create("dev_cache", 10 * 1024)))
+		return_0;
 
 	if (!(_cache.names = dm_hash_create(128))) {
-		stack;
 		dm_pool_destroy(_cache.mem);
 		_cache.mem = 0;
-		return 0;
+		return_0;
 	}
 
 	if (!(_cache.devices = btree_create(_cache.mem))) {
-		log_err("Couldn't create binary tree for dev-cache.");
+		log_error("Couldn't create binary tree for dev-cache.");
 		goto bad;
 	}
 
-	list_init(&_cache.dirs);
-	list_init(&_cache.files);
+	dm_list_init(&_cache.dirs);
+	dm_list_init(&_cache.files);
 
 	if (!_init_preferred_names(cmd))
 		goto_bad;
@@ -576,7 +561,7 @@ int dev_cache_init(struct cmd_context *cmd)
 static void _check_closed(struct device *dev)
 {
 	if (dev->fd >= 0)
-		log_err("Device '%s' has been left open.", dev_name(dev));
+		log_error("Device '%s' has been left open.", dev_name(dev));
 }
 
 static void _check_for_open_devices(void)
@@ -604,8 +589,8 @@ void dev_cache_exit(void)
 
 	_cache.devices = NULL;
 	_cache.has_scanned = 0;
-	list_init(&_cache.dirs);
-	list_init(&_cache.files);
+	dm_list_init(&_cache.dirs);
+	dm_list_init(&_cache.files);
 }
 
 int dev_cache_add_dir(const char *path)
@@ -630,7 +615,7 @@ int dev_cache_add_dir(const char *path)
 	}
 
 	strcpy(dl->dir, path);
-	list_add(&_cache.dirs, &dl->list);
+	dm_list_add(&_cache.dirs, &dl->list);
 	return 1;
 }
 
@@ -656,7 +641,7 @@ int dev_cache_add_loopfile(const char *path)
 	}
 
 	strcpy(dl->dir, path);
-	list_add(&_cache.files, &dl->list);
+	dm_list_add(&_cache.files, &dl->list);
 	return 1;
 }
 
@@ -673,7 +658,7 @@ const char *dev_name_confirmed(struct device *dev, int quiet)
 	if ((dev->flags & DEV_REGULAR))
 		return dev_name(dev);
 
-	while ((r = stat(name = list_item(dev->aliases.n,
+	while ((r = stat(name = dm_list_item(dev->aliases.n,
 					  struct str_list)->str, &buf)) ||
 	       (buf.st_rdev != dev->dev)) {
 		if (r < 0) {
@@ -697,14 +682,15 @@ const char *dev_name_confirmed(struct device *dev, int quiet)
 		/* Leave list alone if there isn't an alternative name */
 		/* so dev_name will always find something to return. */
 		/* Otherwise add the name to the correct device. */
-		if (list_size(&dev->aliases) > 1) {
-			list_del(dev->aliases.n);
+		if (dm_list_size(&dev->aliases) > 1) {
+			dm_list_del(dev->aliases.n);
 			if (!r)
 				_insert(name, 0);
 			continue;
 		}
 
-		log_error("Aborting - please provide new pathname for what "
+		/* Scanning issues this inappropriately sometimes. */
+		log_debug("Aborting - please provide new pathname for what "
 			  "used to be %s", name);
 		return NULL;
 	}
@@ -792,6 +778,6 @@ int dev_fd(struct device *dev)
 
 const char *dev_name(const struct device *dev)
 {
-	return (dev) ? list_item(dev->aliases.n, struct str_list)->str :
+	return (dev) ? dm_list_item(dev->aliases.n, struct str_list)->str :
 	    "unknown device";
 }

lib/device/dev-io.c

@@ -51,7 +51,7 @@
 #  endif
 #endif
 
-static LIST_INIT(_open_devices);
+static DM_LIST_INIT(_open_devices);
 
 /*-----------------------------------------------------------------
  * The standard io loop that keeps submitting an io until it's
@@ -170,10 +170,8 @@ static int _aligned_io(struct device_area *where, void *buffer,
 	struct device_area widened;
 
 	if (!(where->dev->flags & DEV_REGULAR) &&
-	    !_get_block_size(where->dev, &block_size)) {
-		stack;
-		return 0;
-	}
+	    !_get_block_size(where->dev, &block_size))
+		return_0;
 
 	if (!block_size)
 		block_size = lvm_getpagesize();
@@ -200,10 +198,8 @@ static int _aligned_io(struct device_area *where, void *buffer,
 
 	/* channel the io through the bounce buffer */
 	if (!_io(&widened, bounce, 0)) {
-		if (!should_write) {
-			stack;
-			return 0;
-		}
+		if (!should_write)
+			return_0;
 		/* FIXME pre-extend the file */
 		memset(bounce, '\n', widened.size);
 	}
@@ -266,18 +262,65 @@ static int _dev_get_size_dev(const struct device *dev, uint64_t *size)
 	return 1;
 }
 
+static int _dev_read_ahead_dev(struct device *dev, uint32_t *read_ahead)
+{
+	long read_ahead_long;
+
+	if (dev->read_ahead != -1) {
+		*read_ahead = (uint32_t) dev->read_ahead;
+		return 1;
+	}
+
+	if (!dev_open(dev))
+		return_0;
+
+	if (ioctl(dev->fd, BLKRAGET, &read_ahead_long) < 0) {
+		log_sys_error("ioctl BLKRAGET", dev_name(dev));
+		if (!dev_close(dev))
+			stack;
+		return 0;
+	}
+
+	if (!dev_close(dev))
+		stack;
+
+	*read_ahead = (uint32_t) read_ahead_long;
+	dev->read_ahead = read_ahead_long;
+
+	log_very_verbose("%s: read_ahead is %u sectors",
+			 dev_name(dev), *read_ahead);
+
+	return 1;
+}
+
 /*-----------------------------------------------------------------
  * Public functions
  *---------------------------------------------------------------*/
 
 int dev_get_size(const struct device *dev, uint64_t *size)
 {
+	if (!dev)
+		return 0;
+
 	if ((dev->flags & DEV_REGULAR))
 		return _dev_get_size_file(dev, size);
 	else
 		return _dev_get_size_dev(dev, size);
 }
 
+int dev_get_read_ahead(struct device *dev, uint32_t *read_ahead)
+{
+	if (!dev)
+		return 0;
+
+	if (dev->flags & DEV_REGULAR) {
+		*read_ahead = 0;
+		return 1;
+	}
+
+	return _dev_read_ahead_dev(dev, read_ahead);
+}
+
 /* FIXME Unused
 int dev_get_sectsize(struct device *dev, uint32_t *size)
 {
@@ -340,7 +383,7 @@ int dev_open_flags(struct device *dev, int flags, int direct, int quiet)
 
 		if (dev->open_count && !need_excl) {
 			/* FIXME Ensure we never get here */
-			log_debug("WARNING: %s already opened read-only", 
+			log_debug("WARNING: %s already opened read-only",
 				  dev_name(dev));
 			dev->open_count++;
 		}
@@ -354,10 +397,8 @@ int dev_open_flags(struct device *dev, int flags, int direct, int quiet)
 
 	if (dev->flags & DEV_REGULAR)
 		name = dev_name(dev);
-	else if (!(name = dev_name_confirmed(dev, quiet))) {
-		stack;
-		return 0;
-	}
+	else if (!(name = dev_name_confirmed(dev, quiet)))
+		return_0;
 
 	if (!(dev->flags & DEV_REGULAR)) {
 		if (stat(name, &buf) < 0) {
@@ -437,7 +478,7 @@ int dev_open_flags(struct device *dev, int flags, int direct, int quiet)
 	if ((flags & O_CREAT) && !(flags & O_TRUNC))
 		dev->end = lseek(dev->fd, (off_t) 0, SEEK_END);
 
-	list_add(&_open_devices, &dev->open_list);
+	dm_list_add(&_open_devices, &dev->open_list);
 
 	log_debug("Opened %s %s%s%s", dev_name(dev),
 		  dev->flags & DEV_OPENED_RW ? "RW" : "RO",
@@ -486,12 +527,12 @@ static void _close(struct device *dev)
 		log_sys_error("close", dev_name(dev));
 	dev->fd = -1;
 	dev->block_size = -1;
-	list_del(&dev->open_list);
+	dm_list_del(&dev->open_list);
 
 	log_debug("Closed %s", dev_name(dev));
 
 	if (dev->flags & DEV_ALLOCED) {
-		dm_free((void *) list_item(dev->aliases.n, struct str_list)->
+		dm_free((void *) dm_list_item(dev->aliases.n, struct str_list)->
 			 str);
 		dm_free(dev->aliases.n);
 		dm_free(dev);
@@ -522,8 +563,8 @@ static int _dev_close(struct device *dev, int immediate)
 
 	/* Close unless device is known to belong to a locked VG */
 	if (immediate ||
-	    (dev->open_count < 1 && 
-	     (!(info = info_from_pvid(dev->pvid)) ||
+	    (dev->open_count < 1 &&
+	     (!(info = info_from_pvid(dev->pvid, 0)) ||
 	      !info->vginfo ||
 	      !vgname_is_locked(info->vginfo->vgname))))
 		_close(dev);
@@ -543,11 +584,11 @@ int dev_close_immediate(struct device *dev)
 
 void dev_close_all(void)
 {
-	struct list *doh, *doht;
+	struct dm_list *doh, *doht;
 	struct device *dev;
 
-	list_iterate_safe(doh, doht, &_open_devices) {
-		dev = list_struct_base(doh, struct device, open_list);
+	dm_list_iterate_safe(doh, doht, &_open_devices) {
+		dev = dm_list_struct_base(doh, struct device, open_list);
 		if (dev->open_count < 1)
 			_close(dev);
 	}
@@ -557,10 +598,8 @@ int dev_read(struct device *dev, uint64_t offset, size_t len, void *buffer)
 {
 	struct device_area where;
 
-	if (!dev->open_count) {
-		stack;
-		return 0;
-	}
+	if (!dev->open_count)
+		return_0;
 
 	where.dev = dev;
 	where.start = offset;
@@ -607,10 +646,8 @@ int dev_append(struct device *dev, size_t len, void *buffer)
 {
 	int r;
 
-	if (!dev->open_count) {
-		stack;
-		return 0;
-	}
+	if (!dev->open_count)
+		return_0;
 
 	r = dev_write(dev, dev->end, len, buffer);
 	dev->end += (uint64_t) len;
@@ -625,10 +662,8 @@ int dev_write(struct device *dev, uint64_t offset, size_t len, void *buffer)
 {
 	struct device_area where;
 
-	if (!dev->open_count) {
-		stack;
-		return 0;
-	}
+	if (!dev->open_count)
+		return_0;
 
 	where.dev = dev;
 	where.start = offset;
@@ -644,10 +679,8 @@ int dev_set(struct device *dev, uint64_t offset, size_t len, int value)
 	size_t s;
 	char buffer[4096] __attribute((aligned(8)));
 
-	if (!dev_open(dev)) {
-		stack;
-		return 0;
-	}
+	if (!dev_open(dev))
+		return_0;
 
 	if ((offset % SECTOR_SIZE) || (len % SECTOR_SIZE))
 		log_debug("Wiping %s at %" PRIu64 " length %" PRIsize_t,

lib/device/dev-md.c

@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2004 Luca Berra
- * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
+ * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
  *
@@ -16,6 +16,7 @@
 #include "lib.h"
 #include "metadata.h"
 #include "xlate.h"
+#include "filter.h"
 
 #ifdef linux
 
@@ -48,18 +49,26 @@ static int _dev_has_md_magic(struct device *dev, uint64_t sb_offset)
  * 1: At start of device
  * 2: 4K from start of device.
  */
-static uint64_t _v1_sb_offset(uint64_t size, unsigned minor_version)
+typedef enum {
+	MD_MINOR_VERSION_MIN,
+	MD_MINOR_V0 = MD_MINOR_VERSION_MIN,
+	MD_MINOR_V1,
+	MD_MINOR_V2,
+	MD_MINOR_VERSION_MAX = MD_MINOR_V2
+} md_minor_version_t;
+
+static uint64_t _v1_sb_offset(uint64_t size, md_minor_version_t minor_version)
 {
-	uint64_t sb_offset;
+	uint64_t uninitialized_var(sb_offset);
 
 	switch(minor_version) {
-	case 0:
+	case MD_MINOR_V0:
 		sb_offset = (size - 8 * 2) & ~(4 * 2 - 1ULL);
 		break;
-	case 1:
+	case MD_MINOR_V1:
 		sb_offset = 0;
 		break;
-	case 2:
+	case MD_MINOR_V2:
 		sb_offset = 4 * 2;
 		break;
 	}
@@ -74,7 +83,7 @@ static uint64_t _v1_sb_offset(uint64_t size, unsigned minor_version)
 int dev_is_md(struct device *dev, uint64_t *sb)
 {
 	int ret = 1;
-	unsigned minor = 0;
+	md_minor_version_t minor;
 	uint64_t size, sb_offset;
 
 	if (!dev_get_size(dev, &size)) {
@@ -96,12 +105,13 @@ int dev_is_md(struct device *dev, uint64_t *sb)
 	if (_dev_has_md_magic(dev, sb_offset))
 		goto out;
 
+	minor = MD_MINOR_VERSION_MIN;
 	/* Version 1, try v1.0 -> v1.2 */
 	do {
 		sb_offset = _v1_sb_offset(size, minor);
 		if (_dev_has_md_magic(dev, sb_offset))
 			goto out;
-	} while (++minor <= 2);
+	} while (++minor <= MD_MINOR_VERSION_MAX);
 
 	ret = 0;
 
@@ -115,6 +125,197 @@ out:
 	return ret;
 }
 
+static int _md_sysfs_attribute_snprintf(char *path, size_t size,
+					const char *sysfs_dir,
+					struct device *blkdev,
+					const char *attribute)
+{
+	struct stat info;
+	dev_t dev = blkdev->dev;
+	int ret = -1;
+
+	if (!sysfs_dir || !*sysfs_dir)
+		return ret;
+
+check_md_major:
+	if (MAJOR(dev) != md_major()) {
+		if (get_primary_dev(sysfs_dir, blkdev, &dev))
+			goto check_md_major;
+		return ret;
+	}
+
+	ret = dm_snprintf(path, size, "%s/dev/block/%d:%d/md/%s", sysfs_dir,
+			  (int)MAJOR(dev), (int)MINOR(dev), attribute);
+	if (ret < 0) {
+		log_error("dm_snprintf md %s failed", attribute);
+		return ret;
+	}
+
+	if (stat(path, &info) == -1) {
+		if (errno != ENOENT) {
+			log_sys_error("stat", path);
+			return ret;
+		}
+		/* old sysfs structure */
+		ret = dm_snprintf(path, size, "%s/block/md%d/md/%s",
+				  sysfs_dir, (int)MINOR(dev), attribute);
+		if (ret < 0) {
+			log_error("dm_snprintf old md %s failed", attribute);
+			return ret;
+		}
+	}
+
+	return ret;
+}
+
+static int _md_sysfs_attribute_scanf(const char *sysfs_dir,
+				     struct device *dev,
+				     const char *attribute_name,
+				     const char *attribute_fmt,
+				     void *attribute_value)
+{
+	char path[PATH_MAX+1], buffer[64];
+	FILE *fp;
+	int ret = 0;
+
+	if (_md_sysfs_attribute_snprintf(path, PATH_MAX, sysfs_dir,
+					 dev, attribute_name) < 0)
+		return ret;
+
+	if (!(fp = fopen(path, "r"))) {
+		log_sys_error("fopen", path);
+		return ret;
+	}
+
+	if (!fgets(buffer, sizeof(buffer), fp)) {
+		log_sys_error("fgets", path);
+		goto out;
+	}
+
+	if ((ret = sscanf(buffer, attribute_fmt, attribute_value)) != 1) {
+		log_error("%s sysfs attr %s not in expected format: %s",
+			  dev_name(dev), attribute_name, buffer);
+		goto out;
+	}
+
+out:
+	if (fclose(fp))
+		log_sys_error("fclose", path);
+
+	return ret;
+}
+
+/*
+ * Retrieve chunk size from md device using sysfs.
+ */
+static unsigned long dev_md_chunk_size(const char *sysfs_dir,
+				       struct device *dev)
+{
+	const char *attribute = "chunk_size";
+	unsigned long chunk_size_bytes = 0UL;
+
+	if (_md_sysfs_attribute_scanf(sysfs_dir, dev, attribute,
+				      "%lu", &chunk_size_bytes) != 1)
+		return 0;
+
+	log_very_verbose("Device %s %s is %lu bytes.",
+			 dev_name(dev), attribute, chunk_size_bytes);
+
+	return chunk_size_bytes >> SECTOR_SHIFT;
+}
+
+/*
+ * Retrieve level from md device using sysfs.
+ */
+static int dev_md_level(const char *sysfs_dir, struct device *dev)
+{
+	const char *attribute = "level";
+	int level = -1;
+
+	if (_md_sysfs_attribute_scanf(sysfs_dir, dev, attribute,
+				      "raid%d", &level) != 1)
+		return -1;
+
+	log_very_verbose("Device %s %s is raid%d.",
+			 dev_name(dev), attribute, level);
+
+	return level;
+}
+
+/*
+ * Retrieve raid_disks from md device using sysfs.
+ */
+static int dev_md_raid_disks(const char *sysfs_dir, struct device *dev)
+{
+	const char *attribute = "raid_disks";
+	int raid_disks = 0;
+
+	if (_md_sysfs_attribute_scanf(sysfs_dir, dev, attribute,
+				      "%d", &raid_disks) != 1)
+		return 0;
+
+	log_very_verbose("Device %s %s is %d.",
+			 dev_name(dev), attribute, raid_disks);
+
+	return raid_disks;
+}
+
+/*
+ * Calculate stripe width of md device using its sysfs files.
+ */
+unsigned long dev_md_stripe_width(const char *sysfs_dir, struct device *dev)
+{
+	unsigned long chunk_size_sectors = 0UL;
+	unsigned long stripe_width_sectors = 0UL;
+	int level, raid_disks, data_disks;
+
+	chunk_size_sectors = dev_md_chunk_size(sysfs_dir, dev);
+	if (!chunk_size_sectors)
+		return 0;
+
+	level = dev_md_level(sysfs_dir, dev);
+	if (level < 0)
+		return 0;
+
+	raid_disks = dev_md_raid_disks(sysfs_dir, dev);
+	if (!raid_disks)
+		return 0;
+
+	/* The raid level governs the number of data disks. */
+	switch (level) {
+	case 0:
+		/* striped md does not have any parity disks */
+		data_disks = raid_disks;
+		break;
+	case 1:
+	case 10:
+		/* mirrored md effectively has 1 data disk */
+		data_disks = 1;
+		break;
+	case 4:
+	case 5:
+		/* both raid 4 and 5 have a single parity disk */
+		data_disks = raid_disks - 1;
+		break;
+	case 6:
+		/* raid 6 has 2 parity disks */
+		data_disks = raid_disks - 2;
+		break;
+	default:
+		log_error("Device %s has an unknown md raid level: %d",
+			  dev_name(dev), level);
+		return 0;
+	}
+
+	stripe_width_sectors = chunk_size_sectors * data_disks;
+
+	log_very_verbose("Device %s stripe-width is %lu bytes.",
+			 dev_name(dev),
+			 stripe_width_sectors << SECTOR_SHIFT);
+
+	return stripe_width_sectors;
+}
+
 #else
 
 int dev_is_md(struct device *dev __attribute((unused)),
@@ -123,4 +324,10 @@ int dev_is_md(struct device *dev __attribute((unused)),
 	return 0;
 }
 
+unsigned long dev_md_stripe_width(const char *sysfs_dir __attribute((unused)),
+				  struct device *dev  __attribute((unused)))
+{
+	return 0UL;
+}
+
 #endif

lib/device/dev-swap.c

@@ -0,0 +1,87 @@
+/*
+ * Copyright (C) 2009 Red Hat, Inc. All rights reserved.
+ *
+ * This file is part of LVM2.
+ *
+ * This copyrighted material is made available to anyone wishing to use,
+ * modify, copy, or redistribute it subject to the terms and conditions
+ * of the GNU Lesser General Public License v.2.1.
+ *
+ * You should have received a copy of the GNU Lesser General Public License
+ * along with this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include "lib.h"
+#include "metadata.h"
+#include "xlate.h"
+#include "filter.h"
+
+#ifdef linux
+
+#define MAX_PAGESIZE	(64 * 1024)
+#define SIGNATURE_SIZE  10
+
+static int
+_swap_detect_signature(const char *buf)
+{
+	if (memcmp(buf, "SWAP-SPACE", 10) == 0 ||
+            memcmp(buf, "SWAPSPACE2", 10) == 0)
+		return 1;
+
+	if (memcmp(buf, "S1SUSPEND", 9) == 0 ||
+	    memcmp(buf, "S2SUSPEND", 9) == 0 ||
+	    memcmp(buf, "ULSUSPEND", 9) == 0 ||
+	    memcmp(buf, "\xed\xc3\x02\xe9\x98\x56\xe5\x0c", 8) == 0)
+		return 1;
+
+	return 0;
+}
+
+int dev_is_swap(struct device *dev, uint64_t *signature)
+{
+	char buf[10];
+	uint64_t size;
+	int page;
+
+	if (!dev_get_size(dev, &size)) {
+		stack;
+		return -1;
+	}
+
+	if (!dev_open(dev)) {
+		stack;
+		return -1;
+	}
+
+	*signature = 0;
+
+	for (page = 0x1000; page <= MAX_PAGESIZE; page <<= 1) {
+		/*
+		 * skip 32k pagesize since this does not seem to be supported
+		 */
+		if (page == 0x8000)
+			continue;
+		if (size < page)
+			break;
+		if (!dev_read(dev, page - SIGNATURE_SIZE,
+			      SIGNATURE_SIZE, buf)) {
+			stack;
+			return -1;
+		}
+		if (_swap_detect_signature(buf)) {
+			*signature = page - SIGNATURE_SIZE;
+			break;
+		}
+	}
+
+	if (!dev_close(dev))
+		stack;
+
+	if (*signature)
+		return 1;
+
+	return 0;
+}
+
+#endif

lib/device/device.c

@@ -13,6 +13,7 @@
  * Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
 
+#include <libgen.h> /* dirname, basename */
 #include "lib.h"
 #include "lvm-types.h"
 #include "device.h"
@@ -27,22 +28,26 @@
 #define PART_OFFSET UINT64_C(0x1BE)
 
 struct partition {
-        uint8_t boot_ind;
-        uint8_t head;
-        uint8_t sector;
-        uint8_t cyl;
-        uint8_t sys_ind;	/* partition type */
-        uint8_t end_head;
-        uint8_t end_sector;
-        uint8_t end_cyl;
-        uint32_t start_sect;
-        uint32_t nr_sects;
+	uint8_t boot_ind;
+	uint8_t head;
+	uint8_t sector;
+	uint8_t cyl;
+	uint8_t sys_ind;	/* partition type */
+	uint8_t end_head;
+	uint8_t end_sector;
+	uint8_t end_cyl;
+	uint32_t start_sect;
+	uint32_t nr_sects;
 } __attribute__((packed));
 
 static int _is_partitionable(struct device *dev)
 {
 	int parts = max_partitions(MAJOR(dev->dev));
 
+	/* All MD devices are partitionable via blkext (as of 2.6.28) */
+	if (MAJOR(dev->dev) == md_major())
+		return 1;
+
 	if ((parts <= 1) || (MINOR(dev->dev) % parts))
 		return 0;
 
@@ -62,10 +67,8 @@ static int _has_partition_table(struct device *dev)
 		return -1;
 	}
 
-	if (!dev_read(dev, UINT64_C(0), sizeof(buf), &buf)) {
-		stack;
-		goto out;
-	}
+	if (!dev_read(dev, UINT64_C(0), sizeof(buf), &buf))
+		goto_out;
 
 	/* FIXME Check for other types of partition table too */
 
@@ -280,3 +283,208 @@ int _get_partition_type(struct dev_mgr *dm, struct device *d)
 	return 0;
 }
 #endif
+
+#ifdef linux
+
+int get_primary_dev(const char *sysfs_dir,
+		    struct device *dev, dev_t *result)
+{
+	char path[PATH_MAX+1];
+	char temp_path[PATH_MAX+1];
+	char buffer[64];
+	struct stat info;
+	FILE *fp;
+	uint32_t pri_maj, pri_min;
+	int ret = 0;
+
+	/* check if dev is a partition */
+	if (dm_snprintf(path, PATH_MAX, "%s/dev/block/%d:%d/partition",
+			sysfs_dir, (int)MAJOR(dev->dev), (int)MINOR(dev->dev)) < 0) {
+		log_error("dm_snprintf partition failed");
+		return ret;
+	}
+
+	if (stat(path, &info) == -1) {
+		if (errno != ENOENT)
+			log_sys_error("stat", path);
+		return ret;
+	}
+
+	/*
+	 * extract parent's path from the partition's symlink, e.g.:
+	 * - readlink /sys/dev/block/259:0 = ../../block/md0/md0p1
+	 * - dirname ../../block/md0/md0p1 = ../../block/md0
+	 * - basename ../../block/md0/md0  = md0
+	 * Parent's 'dev' sysfs attribute  = /sys/block/md0/dev
+	 */
+	if (readlink(dirname(path), temp_path, PATH_MAX) < 0) {
+		log_sys_error("readlink", path);
+		return ret;
+	}
+
+	if (dm_snprintf(path, PATH_MAX, "%s/block/%s/dev",
+			sysfs_dir, basename(dirname(temp_path))) < 0) {
+		log_error("dm_snprintf dev failed");
+		return ret;
+	}
+
+	/* finally, parse 'dev' attribute and create corresponding dev_t */
+	if (stat(path, &info) == -1) {
+		if (errno == ENOENT)
+			log_error("sysfs file %s does not exist", path);
+		else
+			log_sys_error("stat", path);
+		return ret;
+	}
+
+	fp = fopen(path, "r");
+	if (!fp) {
+		log_sys_error("fopen", path);
+		return ret;
+	}
+
+	if (!fgets(buffer, sizeof(buffer), fp)) {
+		log_sys_error("fgets", path);
+		goto out;
+	}
+
+	if (sscanf(buffer, "%d:%d", &pri_maj, &pri_min) != 2) {
+		log_error("sysfs file %s not in expected MAJ:MIN format: %s",
+			  path, buffer);
+		goto out;
+	}
+	*result = MKDEV(pri_maj, pri_min);
+	ret = 1;
+
+out:
+	if (fclose(fp))
+		log_sys_error("fclose", path);
+
+	return ret;
+}
+
+static unsigned long _dev_topology_attribute(const char *attribute,
+					     const char *sysfs_dir,
+					     struct device *dev)
+{
+	const char *sysfs_fmt_str = "%s/dev/block/%d:%d/%s";
+	char path[PATH_MAX+1], buffer[64];
+	FILE *fp;
+	struct stat info;
+	dev_t uninitialized_var(primary);
+	unsigned long result = 0UL;
+
+	if (!attribute || !*attribute)
+		return_0;
+
+	if (!sysfs_dir || !*sysfs_dir)
+		return_0;
+
+	if (dm_snprintf(path, PATH_MAX, sysfs_fmt_str, sysfs_dir,
+			(int)MAJOR(dev->dev), (int)MINOR(dev->dev),
+			attribute) < 0) {
+		log_error("dm_snprintf %s failed", attribute);
+		return 0;
+	}
+
+	/*
+	 * check if the desired sysfs attribute exists
+	 * - if not: either the kernel doesn't have topology support
+	 *   or the device could be a partition
+	 */
+	if (stat(path, &info) == -1) {
+		if (errno != ENOENT) {
+			log_sys_error("stat", path);
+			return 0;
+		}
+		if (!get_primary_dev(sysfs_dir, dev, &primary))
+			return 0;
+
+		/* get attribute from partition's primary device */
+		if (dm_snprintf(path, PATH_MAX, sysfs_fmt_str, sysfs_dir,
+				(int)MAJOR(primary), (int)MINOR(primary),
+				attribute) < 0) {
+			log_error("primary dm_snprintf %s failed", attribute);
+			return 0;
+		}
+		if (stat(path, &info) == -1) {
+			if (errno != ENOENT)
+				log_sys_error("stat", path);
+			return 0;
+		}
+	}
+
+	if (!(fp = fopen(path, "r"))) {
+		log_sys_error("fopen", path);
+		return 0;
+	}
+
+	if (!fgets(buffer, sizeof(buffer), fp)) {
+		log_sys_error("fgets", path);
+		goto out;
+	}
+
+	if (sscanf(buffer, "%lu", &result) != 1) {
+		log_error("sysfs file %s not in expected format: %s", path,
+			  buffer);
+		goto out;
+	}
+
+	log_very_verbose("Device %s %s is %lu bytes.",
+			 dev_name(dev), attribute, result);
+
+out:
+	if (fclose(fp))
+		log_sys_error("fclose", path);
+
+	return result >> SECTOR_SHIFT;
+}
+
+unsigned long dev_alignment_offset(const char *sysfs_dir,
+				   struct device *dev)
+{
+	return _dev_topology_attribute("alignment_offset",
+				       sysfs_dir, dev);
+}
+
+unsigned long dev_minimum_io_size(const char *sysfs_dir,
+				  struct device *dev)
+{
+	return _dev_topology_attribute("queue/minimum_io_size",
+				       sysfs_dir, dev);
+}
+
+unsigned long dev_optimal_io_size(const char *sysfs_dir,
+				  struct device *dev)
+{
+	return _dev_topology_attribute("queue/optimal_io_size",
+				       sysfs_dir, dev);
+}
+
+#else
+
+int get_primary_dev(const char *sysfs_dir,
+		    struct device *dev, dev_t *result)
+{
+	return 0;
+}
+
+unsigned long dev_alignment_offset(const char *sysfs_dir,
+				   struct device *dev)
+{
+	return 0UL;
+}
+
+unsigned long dev_minimum_io_size(const char *sysfs_dir,
+				  struct device *dev)
+{
+	return 0UL;
+}
+
+unsigned long dev_optimal_io_size(const char *sysfs_dir,
+				  struct device *dev)
+{
+	return 0UL;
+}
+
+#endif

lib/device/device.h

@@ -33,23 +33,24 @@
  * pointer comparisons are valid.
  */
 struct device {
-	struct list aliases;	/* struct str_list from lvm-types.h */
+	struct dm_list aliases;	/* struct str_list from lvm-types.h */
 	dev_t dev;
 
 	/* private */
 	int fd;
 	int open_count;
 	int block_size;
+	int read_ahead;
 	uint32_t flags;
 	uint64_t end;
-	struct list open_list;
+	struct dm_list open_list;
 
 	char pvid[ID_LEN + 1];
 	char _padding[7];
 };
 
 struct device_list {
-	struct list list;
+	struct dm_list list;
 	struct device *dev;
 };
 
@@ -64,6 +65,7 @@ struct device_area {
  */
 int dev_get_size(const struct device *dev, uint64_t *size);
 int dev_get_sectsize(struct device *dev, uint32_t *size);
+int dev_get_read_ahead(struct device *dev, uint32_t *read_ahead);
 
 /* Use quiet version if device number could change e.g. when opening LV */
 int dev_open(struct device *dev);
@@ -93,12 +95,21 @@ const char *dev_name_confirmed(struct device *dev, int quiet);
 
 /* Does device contain md superblock?  If so, where? */
 int dev_is_md(struct device *dev, uint64_t *sb);
-
-/* FIXME Check partition type if appropriate */
-
-#define is_lvm_partition(a) 1
-/* int is_lvm_partition(const char *name); */
+int dev_is_swap(struct device *dev, uint64_t *signature);
+unsigned long dev_md_stripe_width(const char *sysfs_dir, struct device *dev);
 
 int is_partitioned_dev(struct device *dev);
 
+int get_primary_dev(const char *sysfs_dir,
+		    struct device *dev, dev_t *result);
+
+unsigned long dev_alignment_offset(const char *sysfs_dir,
+				   struct device *dev);
+
+unsigned long dev_minimum_io_size(const char *sysfs_dir,
+				  struct device *dev);
+
+unsigned long dev_optimal_io_size(const char *sysfs_dir,
+				  struct device *dev);
+
 #endif

lib/display/display.c

@@ -24,9 +24,9 @@
 
 typedef enum { SIZE_LONG = 0, SIZE_SHORT = 1, SIZE_UNIT = 2 } size_len_t;
 
-static struct {
+static const struct {
 	alloc_policy_t alloc;
-	const char *str;
+	const char str[12]; /* must be changed when size extends 11 chars */
 } _policies[] = {
 	{
 	ALLOC_CONTIGUOUS, "contiguous"}, {
@@ -36,7 +36,7 @@ static struct {
 	ALLOC_INHERIT, "inherit"}
 };
 
-static int _num_policies = sizeof(_policies) / sizeof(*_policies);
+static const int _num_policies = sizeof(_policies) / sizeof(*_policies);
 
 uint64_t units_to_bytes(const char *units, char *unit_type)
 {
@@ -62,14 +62,14 @@ uint64_t units_to_bytes(const char *units, char *unit_type)
 		v = UINT64_C(1);
 		*unit_type = *units;
 		break;
-	case 's':
-		v *= SECTOR_SIZE;
-		break;
 	case 'b':
 	case 'B':
 		v *= UINT64_C(1);
 		break;
 #define KILO UINT64_C(1024)
+	case 's':
+		v *= (KILO/2);
+		break;
 	case 'k':
 		v *= KILO;
 		break;
@@ -90,6 +90,9 @@ uint64_t units_to_bytes(const char *units, char *unit_type)
 		break;
 #undef KILO
 #define KILO UINT64_C(1000)
+	case 'S':
+		v *= (KILO/2);
+		break;
 	case 'K':
 		v *= KILO;
 		break;
@@ -155,7 +158,7 @@ static const char *_display_size(const struct cmd_context *cmd,
 	uint64_t byte = UINT64_C(0);
 	uint64_t units = UINT64_C(1024);
 	char *size_buf = NULL;
-	const char *size_str[][3] = {
+	const char * const size_str[][3] = {
 		{" Exabyte", " EB", "E"},
 		{" Petabyte", " PB", "P"},
 		{" Terabyte", " TB", "T"},
@@ -266,11 +269,11 @@ void pvdisplay_segments(const struct physical_volume *pv)
 	if (pv->pe_size)
 		log_print("--- Physical Segments ---");
 
-	list_iterate_items(pvseg, &pv->segments) {
+	dm_list_iterate_items(pvseg, &pv->segments) {
 		log_print("Physical extent %u to %u:",
 			  pvseg->pe, pvseg->pe + pvseg->len - 1);
 
-		if (pvseg->lvseg) {
+		if (pvseg_is_allocated(pvseg)) {
 			log_print("  Logical volume\t%s%s/%s",
 				  pvseg->lvseg->lv->vg->cmd->dev_dir,
 				  pvseg->lvseg->lv->vg->name,
@@ -307,7 +310,8 @@ void pvdisplay_full(const struct cmd_context *cmd,
 
 	log_print("--- %sPhysical volume ---", pv->pe_size ? "" : "NEW ");
 	log_print("PV Name               %s", pv_dev_name(pv));
-	log_print("VG Name               %s%s", pv->vg_name,
+	log_print("VG Name               %s%s",
+		  is_orphan(pv) ? "" : pv->vg_name,
 		  pv->status & EXPORTED_VG ? " (exported)" : "");
 
 	data_size = (uint64_t) pv->pe_count * pv->pe_size;
@@ -360,10 +364,8 @@ int pvdisplay_short(const struct cmd_context *cmd __attribute((unused)),
 	if (!pv)
 		return 0;
 
-	if (!id_write_format(&pv->id, uuid, sizeof(uuid))) {
-		stack;
-		return 0;
-	}
+	if (!id_write_format(&pv->id, uuid, sizeof(uuid)))
+		return_0;
 
 	log_print("PV Name               %s     ", pv_dev_name(pv));
 	/* FIXME  pv->pv_number); */
@@ -407,10 +409,8 @@ int lvdisplay_full(struct cmd_context *cmd,
 	struct lv_segment *snap_seg = NULL;
 	float snap_percent;	/* fused, fsize; */
 
-	if (!id_write_format(&lv->lvid.id[1], uuid, sizeof(uuid))) {
-		stack;
-		return 0;
-	}
+	if (!id_write_format(&lv->lvid.id[1], uuid, sizeof(uuid)))
+		return_0;
 
 	inkernel = lv_info(cmd, lv, &info, 1, 1) && info.exists;
 
@@ -428,7 +428,7 @@ int lvdisplay_full(struct cmd_context *cmd,
 	if (lv_is_origin(lv)) {
 		log_print("LV snapshot status     source of");
 
-		list_iterate_items_gen(snap_seg, &lv->snapshot_segs,
+		dm_list_iterate_items_gen(snap_seg, &lv->snapshot_segs,
 				       origin_list) {
 			if (inkernel &&
 			    (snap_active = lv_snapshot_percent(snap_seg->cow,
@@ -486,7 +486,7 @@ int lvdisplay_full(struct cmd_context *cmd,
 			  display_size(cmd, (uint64_t) snap_seg->chunk_size));
 	}
 
-	log_print("Segments               %u", list_size(&lv->segments));
+	log_print("Segments               %u", dm_list_size(&lv->segments));
 
 /********* FIXME Stripes & stripesize for each segment
 	log_print("Stripe size (KByte)    %u", lv->stripesize / 2);
@@ -554,7 +554,7 @@ int lvdisplay_segments(const struct logical_volume *lv)
 
 	log_print("--- Segments ---");
 
-	list_iterate_items(seg, &lv->segments) {
+	dm_list_iterate_items(seg, &lv->segments) {
 		log_print("Logical extent %u to %u:",
 			  seg->le, seg->le + seg->len - 1);
 
@@ -575,14 +575,11 @@ void vgdisplay_extents(const struct volume_group *vg __attribute((unused)))
 
 void vgdisplay_full(const struct volume_group *vg)
 {
-	uint32_t access;
+	uint32_t access_str;
 	uint32_t active_pvs;
 	char uuid[64] __attribute((aligned(8)));
 
-	if (vg->status & PARTIAL_VG)
-		active_pvs = list_size(&vg->pvs);
-	else
-		active_pvs = vg->pv_count;
+	active_pvs = vg->pv_count - vg_missing_pv_count(vg);
 
 	log_print("--- Volume group ---");
 	log_print("VG Name               %s", vg->name);
@@ -590,28 +587,29 @@ void vgdisplay_full(const struct volume_group *vg)
 	log_print("Format                %s", vg->fid->fmt->name);
 	if (vg->fid->fmt->features & FMT_MDAS) {
 		log_print("Metadata Areas        %d",
-			  list_size(&vg->fid->metadata_areas));
+			  dm_list_size(&vg->fid->metadata_areas));
 		log_print("Metadata Sequence No  %d", vg->seqno);
 	}
-	access = vg->status & (LVM_READ | LVM_WRITE);
+	access_str = vg->status & (LVM_READ | LVM_WRITE);
 	log_print("VG Access             %s%s%s%s",
-		  access == (LVM_READ | LVM_WRITE) ? "read/write" : "",
-		  access == LVM_READ ? "read" : "",
-		  access == LVM_WRITE ? "write" : "",
-		  access == 0 ? "error" : "");
+		  access_str == (LVM_READ | LVM_WRITE) ? "read/write" : "",
+		  access_str == LVM_READ ? "read" : "",
+		  access_str == LVM_WRITE ? "write" : "",
+		  access_str == 0 ? "error" : "");
 	log_print("VG Status             %s%sresizable",
 		  vg->status & EXPORTED_VG ? "exported/" : "",
 		  vg->status & RESIZEABLE_VG ? "" : "NOT ");
 	/* vg number not part of LVM2 design
 	   log_print ("VG #                  %u\n", vg->vg_number);
 	 */
-	if (vg->status & CLUSTERED) {
+	if (vg_is_clustered(vg)) {
 		log_print("Clustered             yes");
 		log_print("Shared                %s",
 			  vg->status & SHARED ? "yes" : "no");
 	}
+
 	log_print("MAX LV                %u", vg->max_lv);
-	log_print("Cur LV                %u", vg->lv_count + vg->snapshot_count);
+	log_print("Cur LV                %u", vg_visible_lvs(vg));
 	log_print("Open LV               %u", lvs_in_vg_opened(vg));
 /****** FIXME Max LV Size
       log_print ( "MAX LV Size           %s",
@@ -638,8 +636,7 @@ void vgdisplay_full(const struct volume_group *vg)
 			       vg->extent_size));
 
 	log_print("Free  PE / Size       %u / %s", vg->free_count,
-		  display_size(vg->cmd,
-			       (uint64_t) vg->free_count * vg->extent_size));
+		  display_size(vg->cmd, vg_free(vg)));
 
 	if (!id_write_format(&vg->id, uuid, sizeof(uuid))) {
 		stack;
@@ -655,26 +652,23 @@ void vgdisplay_full(const struct volume_group *vg)
 void vgdisplay_colons(const struct volume_group *vg)
 {
 	uint32_t active_pvs;
-	const char *access;
+	const char *access_str;
 	char uuid[64] __attribute((aligned(8)));
 
-	if (vg->status & PARTIAL_VG)
-		active_pvs = list_size(&vg->pvs);
-	else
-		active_pvs = vg->pv_count;
+	active_pvs = vg->pv_count - vg_missing_pv_count(vg);
 
 	switch (vg->status & (LVM_READ | LVM_WRITE)) {
 		case LVM_READ | LVM_WRITE:
-			access = "r/w";
+			access_str = "r/w";
 			break;
 		case LVM_READ:
-			access = "r";
+			access_str = "r";
 			break;
 		case LVM_WRITE:
-			access = "w";
+			access_str = "w";
 			break;
 		default:
-			access = "";
+			access_str = "";
 	}
 
 	if (!id_write_format(&vg->id, uuid, sizeof(uuid))) {
@@ -685,11 +679,11 @@ void vgdisplay_colons(const struct volume_group *vg)
 	log_print("%s:%s:%d:-1:%u:%u:%u:-1:%u:%u:%u:%" PRIu64 ":%" PRIu32
 		  ":%u:%u:%u:%s",
 		vg->name,
-		access,
+		access_str,
 		vg->status,
 		/* internal volume group number; obsolete */
 		vg->max_lv,
-		vg->lv_count,
+		vg_visible_lvs(vg),
 		lvs_in_vg_opened(vg),
 		/* FIXME: maximum logical volume size */
 		vg->max_pv,
@@ -698,7 +692,7 @@ void vgdisplay_colons(const struct volume_group *vg)
 		(uint64_t) vg->extent_count * (vg->extent_size / 2),
 		vg->extent_size / 2,
 		vg->extent_count,
-		vg->extent_count - vg->free_count, 
+		vg->extent_count - vg->free_count,
 		vg->free_count,
 		uuid[0] ? uuid : "none");
 	return;
@@ -713,8 +707,7 @@ void vgdisplay_short(const struct volume_group *vg)
 		  display_size(vg->cmd,
 			       ((uint64_t) vg->extent_count -
 				vg->free_count) * vg->extent_size),
-		  display_size(vg->cmd,
-			       (uint64_t) vg->free_count * vg->extent_size));
+		  display_size(vg->cmd, vg_free(vg)));
 	return;
 }
 
@@ -722,7 +715,7 @@ void display_formats(const struct cmd_context *cmd)
 {
 	const struct format_type *fmt;
 
-	list_iterate_items(fmt, &cmd->formats) {
+	dm_list_iterate_items(fmt, &cmd->formats) {
 		log_print("%s", fmt->name);
 	}
 }
@@ -731,7 +724,7 @@ void display_segtypes(const struct cmd_context *cmd)
 {
 	const struct segment_type *segtype;
 
-	list_iterate_items(segtype, &cmd->segtypes) {
+	dm_list_iterate_items(segtype, &cmd->segtypes) {
 		log_print("%s", segtype->name);
 	}
 }

lib/error/errseg.c

@@ -51,14 +51,17 @@ static int _errseg_add_target_line(struct dev_manager *dm __attribute((unused)),
 	return dm_tree_node_add_error_target(node, len);
 }
 
-static int _errseg_target_present(const struct lv_segment *seg __attribute((unused)))
+static int _errseg_target_present(struct cmd_context *cmd,
+				  const struct lv_segment *seg __attribute((unused)),
+				  unsigned *attributes __attribute((unused)))
 {
 	static int _errseg_checked = 0;
 	static int _errseg_present = 0;
 
 	/* Reported truncated in older kernels */
 	if (!_errseg_checked &&
-	    (target_present("error", 0) || target_present("erro", 0)))
+	    (target_present(cmd, "error", 0) ||
+	     target_present(cmd, "erro", 0)))
 		_errseg_present = 1;
 
 	_errseg_checked = 1;
@@ -68,7 +71,7 @@ static int _errseg_target_present(const struct lv_segment *seg __attribute((unus
 
 static int _errseg_modules_needed(struct dm_pool *mem,
 				  const struct lv_segment *seg __attribute((unused)),
-				  struct list *modules)
+				  struct dm_list *modules)
 {
 	if (!str_list_add(mem, modules, "error")) {
 		log_error("error module string list allocation failed");
@@ -77,7 +80,7 @@ static int _errseg_modules_needed(struct dm_pool *mem,
 
 	return 1;
 }
- 
+
 static void _errseg_destroy(const struct segment_type *segtype)
 {
 	dm_free((void *)segtype);

lib/filters/filter-composite.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -50,10 +50,8 @@ struct dev_filter *composite_filter_create(int n, struct dev_filter **filters)
 {
 	struct dev_filter **filters_copy, *cft;
 
-	if (!filters) {
-		stack;
-		return NULL;
-	}
+	if (!filters)
+		return_NULL;
 
 	if (!(filters_copy = dm_malloc(sizeof(*filters) * (n + 1)))) {
 		log_error("composite filters allocation failed");

lib/filters/filter-md.c

@@ -19,7 +19,8 @@
 
 #ifdef linux
 
-static int _ignore_md(struct dev_filter *f, struct device *dev)
+static int _ignore_md(struct dev_filter *f __attribute((unused)),
+		      struct device *dev)
 {
 	int ret;
 	

lib/filters/filter-persistent.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -18,6 +18,7 @@
 #include "dev-cache.h"
 #include "filter-persistent.h"
 #include "lvm-file.h"
+#include "lvm-string.h"
 
 #include <sys/stat.h>
 #include <fcntl.h>
@@ -42,10 +43,8 @@ static int _init_hash(struct pfilter *pf)
 	if (pf->devices)
 		dm_hash_destroy(pf->devices);
 
-	if (!(pf->devices = dm_hash_create(128))) {
-		stack;
-		return 0;
-	}
+	if (!(pf->devices = dm_hash_create(128)))
+		return_0;
 
 	return 1;
 }
@@ -99,13 +98,13 @@ int persistent_filter_load(struct dev_filter *f, struct config_tree **cft_out)
 {
 	struct pfilter *pf = (struct pfilter *) f->private;
 	struct config_tree *cft;
-        struct stat info;
+	struct stat info;
 	int r = 0;
 
-        if (!stat(pf->file, &info))
+	if (!stat(pf->file, &info))
 		pf->ctime = info.st_ctime;
 	else {
-                log_very_verbose("%s: stat failed: %s", pf->file,
+		log_very_verbose("%s: stat failed: %s", pf->file,
 				 strerror(errno));
 		return_0;
 	}
@@ -144,6 +143,7 @@ static void _write_array(struct pfilter *pf, FILE *fp, const char *path,
 {
 	void *d;
 	int first = 1;
+	char *buf, *str;
 	struct dm_hash_node *n;
 
 	for (n = dm_hash_get_first(pf->devices); n;
@@ -160,7 +160,13 @@ static void _write_array(struct pfilter *pf, FILE *fp, const char *path,
 			first = 0;
 		}
 
-		fprintf(fp, "\t\t\"%s\"", dm_hash_get_key(pf->devices, n));
+		str = dm_hash_get_key(pf->devices, n);
+		if (!(buf = alloca(escaped_len(str)))) {
+			log_error("persistent filter device path stack "
+				  "allocation failed");
+			return;
+		}
+		fprintf(fp, "\t\t\"%s\"", escape_double_quotes(buf, str));
 	}
 
 	if (!first)
@@ -171,7 +177,7 @@ static void _write_array(struct pfilter *pf, FILE *fp, const char *path,
 
 int persistent_filter_dump(struct dev_filter *f)
 {
-	struct pfilter *pf = (struct pfilter *) f->private;
+	struct pfilter *pf;
 	char *tmp_file;
 	struct stat info, info2;
 	struct config_tree *cft = NULL;
@@ -179,6 +185,10 @@ int persistent_filter_dump(struct dev_filter *f)
 	int lockfd;
 	int r = 0;
 
+	if (!f)
+		return_0;
+	pf = (struct pfilter *) f->private;
+
 	if (!dm_hash_get_num_entries(pf->devices)) {
 		log_very_verbose("Internal persistent device cache empty "
 				 "- not writing to %s", pf->file);
@@ -267,7 +277,7 @@ static int _lookup_p(struct dev_filter *f, struct device *dev)
 		l = pf->real->passes_filter(pf->real, dev) ?
 		    PF_GOOD_DEVICE : PF_BAD_DEVICE;
 
-		list_iterate_items(sl, &dev->aliases)
+		dm_list_iterate_items(sl, &dev->aliases)
 			dm_hash_insert(pf->devices, sl->str, l);
 
 	} else if (l == PF_BAD_DEVICE)
@@ -293,10 +303,8 @@ struct dev_filter *persistent_filter_create(struct dev_filter *real,
 	struct pfilter *pf;
 	struct dev_filter *f = NULL;
 
-	if (!(pf = dm_malloc(sizeof(*pf)))) {
-		stack;
-		return NULL;
-	}
+	if (!(pf = dm_malloc(sizeof(*pf))))
+		return_NULL;
 	memset(pf, 0, sizeof(*pf));
 
 	if (!(pf->file = dm_malloc(strlen(file) + 1)))

lib/filters/filter-regex.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.  
+ * Copyright (C) 2001-2004 Sistina Software, Inc. All rights reserved.
  * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved.
  *
  * This file is part of LVM2.
@@ -70,10 +70,8 @@ static int _extract_pattern(struct dm_pool *mem, const char *pat,
 	/*
 	 * copy the regex
 	 */
-	if (!(r = dm_pool_strdup(mem, pat))) {
-		stack;
-		return 0;
-	}
+	if (!(r = dm_pool_strdup(mem, pat)))
+		return_0;
 
 	/*
 	 * trim the trailing character, having checked it's sep.
@@ -153,7 +151,7 @@ static int _accept_p(struct dev_filter *f, struct device *dev)
 	struct rfilter *rf = (struct rfilter *) f->private;
 	struct str_list *sl;
 
-	list_iterate_items(sl, &dev->aliases) {
+	dm_list_iterate_items(sl, &dev->aliases) {
 		m = dm_regex_match(rf->engine, sl->str);
 
 		if (m >= 0) {
@@ -192,10 +190,8 @@ struct dev_filter *regex_filter_create(struct config_value *patterns)
 	struct rfilter *rf;
 	struct dev_filter *f;
 
-	if (!mem) {
-		stack;
-		return NULL;
-	}
+	if (!mem)
+		return_NULL;
 
 	if (!(rf = dm_pool_alloc(mem, sizeof(*rf))))
 		goto_bad;

lib/filters/filter-sysfs.c

@@ -20,47 +20,11 @@
 
 #include <dirent.h>
 
-static int _locate_sysfs_blocks(const char *proc, char *path, size_t len,
+static int _locate_sysfs_blocks(const char *sysfs_dir, char *path, size_t len,
 				unsigned *sysfs_depth)
 {
-	char proc_mounts[PATH_MAX];
-	FILE *fp;
-	char *split[4], buffer[PATH_MAX + 16];
-	const char *sys_mnt = NULL;
 	struct stat info;
 
-	if (!*proc) {
-		log_verbose("No proc filesystem found: skipping sysfs filter");
-		return 0;
-	}
-		
-	if (dm_snprintf(proc_mounts, sizeof(proc_mounts),
-			 "%s/mounts", proc) < 0) {
-		log_error("Failed to create /proc/mounts string");
-		return 0;
-	}
-
-	if (!(fp = fopen(proc_mounts, "r"))) {
-		log_sys_error("fopen %s", proc_mounts);
-		return 0;
-	}
-
-	while (fgets(buffer, sizeof(buffer), fp)) {
-		if (dm_split_words(buffer, 4, 0, split) == 4 &&
-		    !strcmp(split[2], "sysfs")) {
-			sys_mnt = split[1];
-			break;
-		}
-	}
-
-	if (fclose(fp))
-		log_sys_error("fclose", proc_mounts);
-
-	if (!sys_mnt) {
-		log_error("Failed to find sysfs mount point");
-		return 0;
-	}
-
 	/*
 	 * unified classification directory for all kernel subsystems
 	 *
@@ -70,7 +34,7 @@ static int _locate_sysfs_blocks(const char *proc, char *path, size_t len,
 	 *  `-- sr0 -> ../../../devices/pci0000:00/0000:00:1f.2/host1/target1:0:0/1:0:0:0/block/sr0
 	 *
 	 */
-	if (dm_snprintf(path, len, "%s/%s", sys_mnt,
+	if (dm_snprintf(path, len, "%s/%s", sysfs_dir,
 			"subsystem/block/devices") >= 0) {
 		if (!stat(path, &info)) {
 			*sysfs_depth = 0;
@@ -87,7 +51,7 @@ static int _locate_sysfs_blocks(const char *proc, char *path, size_t len,
 	 *  `-- sr0 -> ../../devices/pci0000:00/0000:00:1f.2/host1/target1:0:0/1:0:0:0/block/sr0
 	 *
 	 */
-	if (dm_snprintf(path, len, "%s/%s", sys_mnt, "class/block") >= 0) {
+	if (dm_snprintf(path, len, "%s/%s", sysfs_dir, "class/block") >= 0) {
 		if (!stat(path, &info)) {
 			*sysfs_depth = 0;
 			return 1;
@@ -112,7 +76,7 @@ static int _locate_sysfs_blocks(const char *proc, char *path, size_t len,
 	 * ...
 	 *
 	 */
-	if (dm_snprintf(path, len, "%s/%s", sys_mnt, "block") >= 0) {
+	if (dm_snprintf(path, len, "%s/%s", sysfs_dir, "block") >= 0) {
 		if (!stat(path, &info)) {
 			*sysfs_depth = 1;
 			return 1;
@@ -236,21 +200,21 @@ static int _read_dev(const char *file, dev_t *result)
  */
 static int _read_devs(struct dev_set *ds, const char *dir, unsigned sysfs_depth)
 {
-        struct dirent *d;
-        DIR *dr;
+	struct dirent *d;
+	DIR *dr;
 	struct stat info;
 	char path[PATH_MAX];
 	char file[PATH_MAX];
 	dev_t dev = { 0 };
 	int r = 1;
 
-        if (!(dr = opendir(dir))) {
-                log_sys_error("opendir", dir);
-                return 0;
-        }
+	if (!(dr = opendir(dir))) {
+		log_sys_error("opendir", dir);
+		return 0;
+	}
 
-        while ((d = readdir(dr))) {
-                if (!strcmp(d->d_name, ".") || !strcmp(d->d_name, ".."))
+	while ((d = readdir(dr))) {
+		if (!strcmp(d->d_name, ".") || !strcmp(d->d_name, ".."))
 			continue;
 
 		if (dm_snprintf(path, sizeof(path), "%s/%s", dir,
@@ -278,8 +242,8 @@ static int _read_devs(struct dev_set *ds, const char *dir, unsigned sysfs_depth)
 		}
 	}
 
-        if (closedir(dr))
-                log_sys_error("closedir", dir);
+	if (closedir(dr))
+		log_sys_error("closedir", dir);
 
 	return r;
 }
@@ -321,7 +285,7 @@ static void _destroy(struct dev_filter *f)
 	dm_pool_destroy(ds->mem);
 }
 
-struct dev_filter *sysfs_filter_create(const char *proc)
+struct dev_filter *sysfs_filter_create(const char *sysfs_dir)
 {
 	char sys_block[PATH_MAX];
 	unsigned sysfs_depth;
@@ -329,7 +293,12 @@ struct dev_filter *sysfs_filter_create(const char *proc)
 	struct dev_set *ds;
 	struct dev_filter *f;
 
-	if (!_locate_sysfs_blocks(proc, sys_block, sizeof(sys_block), &sysfs_depth))
+	if (!*sysfs_dir) {
+		log_verbose("No proc filesystem found: skipping sysfs filter");
+		return NULL;
+	}
+
+	if (!_locate_sysfs_blocks(sysfs_dir, sys_block, sizeof(sys_block), &sysfs_depth))
 		return NULL;
 
 	if (!(mem = dm_pool_create("sysfs", 256))) {
@@ -357,7 +326,7 @@ struct dev_filter *sysfs_filter_create(const char *proc)
 
 #else
 
-struct dev_filter *sysfs_filter_create(const char *proc __attribute((unused)))
+struct dev_filter *sysfs_filter_create(const char *sysfs_dir __attribute((unused)))
 {
 	return NULL;
 }

lib/filters/filter-sysfs.h

@@ -18,6 +18,6 @@
 #include "config.h"
 #include "dev-cache.h"
 
-struct dev_filter *sysfs_filter_create(const char *proc);
+struct dev_filter *sysfs_filter_create(const char *sysfs_dir);
 
 #endif

lib/filters/filter.c

@@ -57,6 +57,7 @@ static const device_info_t device_info[] = {
 	{"ide", 64},		/* IDE disk */
 	{"sd", 16},		/* SCSI disk */
 	{"md", 1},		/* Multiple Disk driver (SoftRAID) */
+	{"mdp", 1},		/* Partitionable MD */
 	{"loop", 1},		/* Loop device */
 	{"dasd", 4},		/* DASD disk (IBM S/390, zSeries) */
 	{"dac960", 8},		/* DAC960 */
@@ -77,6 +78,9 @@ static const device_info_t device_info[] = {
 	{"xvd", 16},		/* Xen virtual block device */
 	{"vdisk", 8},		/* SUN's LDOM virtual block device */
 	{"ps3disk", 16},	/* PlayStation 3 internal disk */
+	{"virtblk", 8},		/* VirtIO disk */
+	{"mmc", 16},		/* MMC block device */
+	{"blkext", 1},		/* Extended device partitions */
 	{NULL, 0}
 };
 
@@ -277,9 +281,8 @@ struct dev_filter *lvm_type_filter_create(const char *proc,
 	f->private = NULL;
 
 	if (!_scan_proc_dev(proc, cn)) {
-		stack;
 		dm_free(f);
-		return NULL;
+		return_NULL;
 	}
 
 	return f;