1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-29 15:22:30 +03:00
Commit Graph

1342 Commits

Author SHA1 Message Date
Bryn M. Reeves
feb69966d4 libdm: use macro for boundary test in _stats_get_extents_for_file() 2016-07-08 22:21:14 +01:00
Bryn M. Reeves
2d1f03b616 libdm: use a constant for FIEMAP buffer size 2016-07-08 22:05:36 +01:00
Bryn M. Reeves
58bfea6a6e libdm: use SECTOR_SHIFT constant in _stats_add_extent() 2016-07-08 22:05:36 +01:00
Bryn M. Reeves
4bb57341bd libdm: enable creation of filemap regions with histograms 2016-07-08 17:27:52 +01:00
Bryn M. Reeves
db73d756e9 libdm: allow regions with histograms in dm_stats_create_group()
Allow regions with histograms to be grouped if all histograms have
the same number of bins and matching bounds.
2016-07-08 17:27:52 +01:00
Bryn M. Reeves
ae9cffba52 libdm: add aggregation support to dm_stats_get_histogram()
Support aggregate group and region histograms by allocating a new
histogram from the pool and populating it with a sum of the histogram
data for the areas contained in the region or group.

To avoid repeatedly summing the same histogram data, cache the pointer
in the group and regions structs for subsequent access. The aggregate
histograms are allocated from the same pool as the area histograms in
the corresponding handle and will be discarded at each list or populate
operation.
2016-07-08 17:27:52 +01:00
Bryn M. Reeves
e104825916 libdm: add dm_stats_create_regions_from_fd()
Add a call to create dmstats regions that correspond to the extents
present in a file descriptor open on a file in a local file system.
The file must reside on a file system type that correctly supports
physical extent location data in the FIEMAP ioctl.

Regions are optionally placed into a group with a user-defined alias.

File systems that do not support physical offsets in FIEMAP (btrfs
currently) are detected via fstatfs() - although attempting to map
a --filemap group on btrfs will fail anyway with the generic error
"Not on a device-mapper device" this is confusing; the file system
mount is on a device-mapper device, but btrfs' volume layer masks
this in the returned st_dev field since the returned logical file
extents may span multiple physical devices.
2016-07-08 14:34:41 +01:00
Bryn M. Reeves
ebc7fc67c8 libdm: fix group resource leak in dm_stats_delete_region()
The function _stats_remove_region_id_from_group() incorecctly set
the group_id to DM_STATS_GROUP_NOT_PRESENT _before_ the call to
_stats_group_destroy(). This will cause the destroy function to
return immediately without doing anything:

 339 static void _stats_group_destroy(struct dm_stats_group *group)
 340 {
 341         if (!_stats_group_present(group))
 342                 return;

Invalidating the ID in _stats_region_region_id_from_group() is
redundant anyway; it is rightly done as the last operation in
_stats_group_destroy (and it is not possible for anything to see
the old value between the two calls).

Remove the change to group_id to ensure that the alias and bitset
resources are correctly freed.
2016-07-08 12:30:09 +01:00
Bryn M. Reeves
cc4f036d36 libdm: improve comments in stats grouping functions
Add more detailed comments to dm_stats_create_group() and
_stats_group_check_overlap().
2016-07-08 11:16:12 +01:00
Bryn M. Reeves
059a383cf8 libdm: fix resource leak in dm_stats_set_alias()
When we fail to update aux_data the newly allocated group->alias must
be freed before reinstating old_alias.
2016-07-08 11:14:29 +01:00
Bryn M. Reeves
5e06b33c51 libdm: enclose dm_stats_walk_do/while() body in do..while
The call to dm_stats_walk_start() before the do statement makes
dm_stats_walk_do() behave inconsistently depending on context;
wrap them in an additional do { } while (0) so that the macro
always expands to a valid statement.
2016-07-08 11:14:22 +01:00
Bryn M. Reeves
bf1dfea393 libdm: check for empty aux_data in _parse_aux_data_group()
If after extracting stats arguments and group tags nothing remains
of aux_data but '-' set the region->aux_data field to the empty
string to match behaviour for non-grouped regions.
2016-07-06 16:31:30 +01:00
Bryn M. Reeves
a497b95db1 libdm: use log_err_once() for group histogram message 2016-07-06 11:16:12 +01:00
Bryn M. Reeves
28658541da libdm: do not permit grouping regions with histograms
Although not harmful do not allow a group containing regions with
histograms since it is not currently possible to present histogram
data aggregated for the group.
2016-07-06 11:10:23 +01:00
Bryn M. Reeves
95ef0cdb46 libdm: check non-zero io count in _average_{rd,wr}_wait_time (Coverity)
Although a non-zero value for the number of ticks spent doing IO
should imply a non-zero number of IOs in the interval test for
this explicitly to avoid a divide-by-zero in the event of bad
counter data.
2016-07-06 09:23:13 +01:00
Bryn M. Reeves
03e03e9c11 libdm: test for zero interval_ns in _utilization() (Coverity)
It's possible for interval_ns to be zero if the interval is not
set or the clock is misconfigured. Test for this before using the
value as the divisor in the utilisation calculation.
2016-07-06 09:14:43 +01:00
Bryn M. Reeves
5d3b136d38 libdm: restore missing braces in _stats_walk_end_areas
Jumping to the end of the region table must only happen if there
are no more present, non-skipped regions, and no group walk is
configured to begin.
2016-07-06 09:04:13 +01:00
Bryn M. Reeves
69c721dd68 libdm: fix mask leak in dm_bitset_parse_list
If an unexpected '-' is found jump to the error branch so that the
mask is properly freed before returning.
2016-07-06 08:59:09 +01:00
Bryn M. Reeves
4d4f48af9f libdm: cast walk flags to uint64_t when logging.
Walk flags are ULL constants; cast the result to a uint64_t before
logging with a FMTx64 format specifier to avoid a compiler warning:

  warning: format ‘%lx’ expects argument of type ‘long unsigned int’,
  but argument 5 has type ‘long long unsigned int’
2016-07-05 20:45:24 +01:00
Bryn M. Reeves
f1dd0258f1 libdm: ensure flags constants have ULL suffix
The walk flags used by libdm-stats use the upper portion of a 64b
value: use the ULL suffix to ensure the compiler knows the expected
size.
2016-07-05 20:21:49 +01:00
Bryn M. Reeves
0f64f2d5fc libdm: fix <backtrace> in dm_stats_populate 2016-07-05 19:53:17 +01:00
Bryn M. Reeves
53e92441a6 libdm: fix <backtrace> in dm_stats_get_nr_regions 2016-07-05 19:53:17 +01:00
Bryn M. Reeves
fef4832a85 libdm: clarify library's use of aux_data
Make it clear in libdevmapper.h, and in function argument names, that
libdm-stats uses the aux_data field internally and that any values set
for user_data are appended to the library values before being stored
with a region, and similarly, that internal data fields will be stripped
prior to returning any previously stored user_data.
2016-07-05 19:53:17 +01:00
Bryn M. Reeves
cda1622fef libdm: allow deleting regions with dm_stats_delete_group()
Add a flag to dm_stats_delete_group() to allow optional deletion
of all regions belonging to the group being removed.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
f1f2df7bc0 libdm: add stats group and region iterators and properties
Add support do dm_stats_walk*() to walk over the set of
available groups using the cursor embedded in the dm_stats
handle, and to obtain the type of the object at the current
stats cursor location. A set of flags is introduced to
control which objects are visited:

    DM_STATS_WALK_AREA
    DM_STATS_WALK_REGION
    DM_STATS_WALK_GROUP
    DM_STATS_WALK_ALL

A final flag suppresses visits to regions that contain only a
single area - since the aggregate of such a region is idential
to the area it contains this allows these duplicates to be
filtered out:

    DM_STATS_WALK_SKIP_SINGLE_AREA

If flags are not initialised before beginning a walk the default
set matches the behaviour of previous versions of the library.

Also accept group identifiers as immediate arguments to the
counter, metric, and property functions by adding control
flags to the region and area identifiers passed in.

Region and area properties are mapped to their equivalents for
the group (for example: group size is reported as the sum of
all regions contained in the group). Counter and metric values
are aggregated for the region or group.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
62050760aa libdm: use defined constants for buffer sizes
Introduce constants for the buffer sizes that libdm-stats uses:
one for messages sent to the kernel, one for rows of response data
returned, and a pair for the "start+len" range and histogram bounds
strings.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
2cb9794da2 libdm: add statistics groups
Add a grouping facility to the libdm-stats library that allows the
user to bind several regions together as a group. Groups may be
used to aggregate data from several regions for reporting, or to
select and sort among large sets of regions.

A textual descriptor ("group tag") is associated with each group
and is stored in the first group member's aux_data field. The
tag contains the group member list and an optional alias for the
group, allowing the user to assign meaningful names to groups of
regions.

These descriptors are parsed in @stats_list message responses and
populate the resulting region and area tables with the group
structure.

Groups with overlapping regions are permitted but since this will
result in some events being counted more than once a warning is
printed in this case.

Nested and overlapping groups are not currently supported and
attempting to create these configurations results in error.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
4cd3dcbbc2 libdm: rename 'region' to 'skip_region' in _stats_walk_next
In libdm-stats.c 'region' usually refers to a 'struct region*'.
Rename the argument to _stats_walk_start to avoid confusion.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
82e5766062 libdm: add enum based counter and metric calls
Add a new enum based interface for accessing counter and metric
values that uses a single function for each:

uint64_t dm_stats_get_counter(const struct dm_stats *dms,
                              dm_stats_counter_t counter
                              uint64_t region_id, uint64_t area_id);

int dm_stats_get_metric(const struct dm_stats *dms, int metric,
                        uint64_t region_id, uint64_t area_id,
                        double *value);

This simplifies the implementation of value aggregation for
groups of regions. The named function interface now calls the
enum interface internally so that all new functionality is
available regardless of the method used to retrieve values.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
69f808ac8d libdm: cache dm name in stats handle
Cache the device-mapper name of a bound device in the dm_stats
handle.

This will be used by stats groups to report a device name or
user defined alias for groups.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
4a66f02f88 libdm: rename dm_stats name, devno and uuid members
The device-mapper name, device numbers and uuid stored in the
dm_stats handle are used only to bind the handle to a specific
device in order to issue ioctls.

Rename them to "bind_*" to reflect this usage in preparation
for caching the device-mapper name of the bound device in the
dm_stats handle.

This will be used to allow optional aliases to be set for
dmstats groups.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
81fad9e853 libdm: add dm_bitset_parse_list()
Add a function to parse a list of integer values and ranges into
a dm_bitset representation. Individual values signify that that bit
is set in the resulting mask and ranges are given as a pair of
start and end values, M-N, such that M and N are the first and
last members of the range (inclusive).

The implementation is based on the kernel's __bitmap_parselist()
that is used for cpumasks and other set configuration passed in
string form from user space.
2016-07-05 19:53:16 +01:00
Bryn M. Reeves
d382e66035 libdm: fix histogram pool user-after-free (CWE-825) 2016-07-05 19:53:15 +01:00
Alasdair G Kergon
d8c2677ab9 raid0: Add raid0_meta segment type. 2016-07-01 22:20:54 +01:00
Peter Rajnoha
eac0706761 libdm: do not issue 'Failed to create directory' message for failure in dm_create_dir
There are detailed messages inside _create_dir_recursive that
dm_create_dir calls (except EROFS which where the message is not
generated, like anywhere else in the code).
2016-06-29 15:58:18 +02:00
Peter Rajnoha
3985b12a2d libdm: report: fix field width calculation when using dm_report_column_headings
This fixes commit 0ba5f4b8e9 which moved
field recalculation (field width and sort position) from
dm_report_object to dm_report_output but it didn't handle the case when
dm_report_column_headings was used separately to report headings (before
dm_report_outpout call) and hence we ended up with intial widths for
fields in the headings.

If we're using dm_report_column_headings, we need to recalculate
fields if we haven't done so yet, the same way as we do in
dm_report_output.
2016-06-28 02:28:40 +01:00
Peter Rajnoha
f0768f636e coverity: fix issues detected in recent code
Uninitialized variables  (UNINIT) /safe/guest2/covscan/LVM2.2.02.158/tools/toollib.c: 3520 in _process_pvs_in_vgs()
Uninitialized variables  (UNINIT) Using uninitialized value "do_report_ret_code".

Null pointer dereferences  (REVERSE_INULL) /safe/guest2/covscan/LVM2.2.02.158/libdm/libdm-report.c: 4745 in dm_report_output()
Null pointer dereferences  (REVERSE_INULL) Null-checking "rh" suggests that it may be null, but it has already been dereferenced on all paths leading to the check.

Incorrect expression  (MISSING_COMMA) /safe/guest2/covscan/LVM2.2.02.158/lib/log/log.c: 280 in _get_log_level_name()
Incorrect expression  (MISSING_COMMA) In the initialization of "log_level_names", a suspicious concatenated string ""noticeinfo"" is produced.

Null pointer dereferences  (FORWARD_NULL) /safe/guest2/covscan/LVM2.2.02.158/tools/reporter.c: 816 in_get_report_options()
Null pointer dereferences  (FORWARD_NULL) Comparing "mem" to null implies that "mem" might be null.
2016-06-28 02:26:54 +01:00
Peter Rajnoha
751163a743 libdm: log: remove log_print_bypass_report calls and register new print_log_libdm for libdm during lvm initialization instead
This fixes commit f50d4011cd which
introduced a problem when using older lvm2 code with newer libdm.
In this case, the old LVM didn't recognize new _LOG_BYPASS_REPORT flag
that libdm-report code used. This ended up with no output at all
from libdm where log_print_bypass_report was called because the
_LOG_BYPASS_REPORT was not masked properly in lvm2's print_log fn
which was called as callback function for logging.

With this patch, the lvm2 registers separate print_log_libdm logging
function for libdm instead. The print_log_libdm is exactly the same
as print_log (used throughout lvm2 code) but it checks whether we're
printing common line on output where "common" means not going to stderr,
not a warning and not an error and if we are, it adds the
_LOG_BYPASS_REPORT flag so the log_print goes directly to output, not
to any log report.

So this achieves the same goal as in f50d4011cd,
just doing it in a way that newer libdm is still compatible with older
lvm2 code (libdm-report is the only code using log_print).

Looking at the opposite mixture - older libdm with newer lvm2 code,
that won't be compilable because the new log report functionality
that is in lvm2 also requires new dm_report_group_* libdm functions
so we don't need to care here.
2016-06-23 14:45:52 +02:00
Peter Rajnoha
5b93db6566 libdm: select: recognize special selection string 'all' as an alias for blank selection string 2016-06-20 11:33:43 +02:00
Peter Rajnoha
2078b842fb libdm: report: add dm_report_set_selection
Since we can do repeated dm_report_output calls now, we also like
to be able to set selection for each of these outputs.
2016-06-20 11:33:43 +02:00
Peter Rajnoha
f2facdc1d0 libdm: report: add DM_REPORT_OUTPUT_MULTIPLE_TIMES report flag to keep report data even after output is done
The DM_REPORT_OUTPUT_MULTIPLE_TIMES instructs reporting code to
keep rows even after dm_report_output call - the rows are not
destroyed in this case which makes it possible to call dm_report_output
multiple times.
2016-06-20 11:33:43 +02:00
Peter Rajnoha
0ba5f4b8e9 refactor: move field width calculation and sort preparation from _do_report_object to dm_report_output
This also prepares code for repeated dm_report_output calls.
2016-06-20 11:33:42 +02:00
Peter Rajnoha
102cc4c1e2 libdm: report: remember special field to display selection status in struct row's field_sel_status variable
This allows for moving parts of the code from dm_report_object to
dm_report_output which is important for subsequent patches that allow
for repeated dm_report_output, not destroying rows on each
dm_report_output call.
2016-06-20 11:33:42 +02:00
Peter Rajnoha
f50d4011cd log: also pass log_print through report and add log_print_bypass_report for use in libdm-report for direct print without report
log_print is used during cmd line processing to log the result of the
operation (e.g. "Volume group vg successfully changed" and similar).

We don't want output from log_print to be interleaved with current
reports from group where log is reported as well. Also, the information
printed by log_print belongs to the log report too, so it should be
rerouted to log report if it's set.

Since the code in libdm-report which is responsible for doing the report
output uses log_print too, we need to use a different kind of log_print
which bypasses any log report currently used for logging (...simply,
we can't call log_print to output the log report itself which in turn
would again reroute to report - the report would never get on output
this way).
2016-06-20 11:33:42 +02:00
Peter Rajnoha
094fce3776 libdm: report: implement DM_REPORT_GROUP_JSON for JSON report output
This patch introduces DM_REPORT_GROUP_JSON report group type. When using
this group type and when pushing a report to such a group, these flags
are automatically unset:

   DM_REPORT_OUTPUT_ALIGNED
   DM_REPORT_OUTPUT_HEADINGS
   DM_REPORT_OUTPUT_COLUMNS_AS_ROWS

...and this flag is set:

   DM_REPORT_OUTPUT_BUFFERED

The whole group is encapsulated in { } for the outermost JSON object
and then each report is reported on output as array of objects where
each object is the row from report:

  {
     "report_name1": [
         {field1="value", field2="value",...},
         {field1="value", field2="value",...}
         ...
     ],
     "report_name2": [
         {field1="value", field2="value",...},
         {field1="value", field2="value",...}
         ...
     ]
     ...
  }
2016-06-20 10:42:26 +02:00
Peter Rajnoha
230b7ff0f6 libdm: report: implement DM_REPORT_GROUP_BASIC for extended report output
This patch introduces DM_REPORT_GROUP_BASIC report group type. This
type has exactly the classical output format as we know from before
introduction of report groups. However, in addition to that, it allows
to put several reports into a group - this is the very basic grouping
scheme that doesn't change the output format itself:

  Report: report1_name
  Header1  Header2 ...
  value    value   ...
  value    value   ...
  ...      ...     ...

  Report: report2_name
  Header1  Header2 ...
  value    value   ...
  value    value   ...
  ...      ...     ...
2016-06-20 10:42:26 +02:00
Peter Rajnoha
a9fe57db1c libdm: report: implement DM_REPORT_GROUP_SINGLE for a group containing single report only
There's no change in output for this report group type - with this type,
we only make sure there's always only one report in a group at a time,
not more.
2016-06-20 10:42:18 +02:00
Peter Rajnoha
9c8f912ea7 libdm: report: introduce dm_report_group
This patch introduces DM report group (represented by dm_report_group
structure) that is used to group several reports to make a whole. As a
whole, all the reports in the group follow the same settings and/or
formatting used on output and it controls that the output is properly
ordered (e.g. the output from different reports is not interleaved
which would break readability and/or syntax of target output format
used for the whole group).

To support this feature, there are 4 new functions:
  - dm_report_group_create
  - dm_report_group_push
  - dm_report_group_pop
  - dm_report_group_destroy

From the naming used (dm_report_group_push/pop), it's clear the reports
are pushed onto a stack. The rule then is that only the report on top
of the stack can be reported (that means calling dm_report_output).
This way we make sure that the output is not interleaved and provides
determinism and control over the output.

Different formats may allow or disallow some of the existing report
flags controlling output itself (DM_REPORT_OUTPUT_*) to be set or not so
once the report is pushed to a group, the grouping code makes sure that
all the reports have compatible flags set and then these flags are
restored once each report is popped from the report group stack.

We also allow to push/pop non-report item in which case such an item
creates a structure (e.g. to put several reports together with any
opening and/or closing lines needed on output which pose as extra
formatting structure besides formatting the reports).

The dm_report_group_push function accepts an argument to pass any
format-specific data needed (e.g. handle, name, structures passed
along while working with reports...).

We can call dm_report_output directly anytime we need (with the only
restriction that we can call dm_report_output only for the report that
is currently on top of the group's stack). Or we don't need to call
dm_report_output explicitly in which case all the reports in a stack are
reported on output automatically once we call dm_report_group_destroy.
2016-06-20 09:26:51 +02:00
Alasdair G Kergon
780424639a Revert "libdm: trace missing settings"
This reverts commit 8fd886f735.

This was a deliberate omission because logging token-by-token metadata
parsing greatly increases the amount of logging for hardly any benefit.

In general, only LVM config file settings need to be logged, and in
places where it's considered important to log particular elements of
metadata that should be done using specific log_* lines.

This area can be revisited.
2016-05-27 14:35:11 +01:00
Alasdair G Kergon
bf8d00985a raid0: Add raid0 segment type.
This remains experimental and quite restrictive so should only be used
for testing at this stage.  (E.g. lvreduce is not supported.)
2016-05-23 16:46:38 +01:00
Zdenek Kabelac
8fd886f735 libdm: trace missing settings
These settings have been missed in very verbose log.
2016-05-19 18:27:34 +02:00
Zdenek Kabelac
e8ba5c9bd4 libdm: cache status reports passthrough cache mode
Report passthrough mode instead of 'Unknown feature'.
2016-05-19 18:26:07 +02:00
Alasdair G Kergon
e6cafdad36 libdm: Show lib vsn even if driver vsn unavailable. 2016-05-12 01:14:25 +01:00
Alasdair G Kergon
16019b518e libdm: Add dm_udev_wait_immediate.
dm_udev_wait() waits inside the library.
dm_udev_wait_immediate allows the caller to do other things if the
cookie isn't yet ready to be decremented.
2016-04-28 00:54:27 +01:00
Bryn M. Reeves
17ad884836 libdm-stats: check for empty region and area lists
Check that @stats_list and @stats_print returned data in the
_stats_parse_list() and _stats_parse_region() functions before
attempting to operate on region and area values.

This avoids a coverity warning since fgets() could potentially
return no data from the memory buffer returned by the ioctl.

In both cases the ioctl would return an error, preventing these
functions from running, however it is cleaner to test for the
condition explicitly and fail in those cases.
2016-04-22 10:35:07 +01:00
Zdenek Kabelac
e2ceb90095 debug: update message in libdm
When dm_tree_find_node_by_uuid() fails to find passed uuid,
report in lof_debug the complete original uuid,
not the one stripped of LVM- prefix.

TODO: inspect manipulation with LVM- prefix here.
2016-04-18 12:32:56 +02:00
Zdenek Kabelac
1be74cfd7f cleanup: gcc warn about comparing int with uint 2016-04-12 11:47:51 +02:00
Peter Rajnoha
f2e59e05ed cleanup: use #define for field's quote and pair character and also for the error msg while extending output line 2016-04-08 10:55:13 +02:00
Alasdair G Kergon
7dcbf1dfd0 libdm: Correct typo. 2016-04-07 01:51:09 +01:00
Zdenek Kabelac
261a85ced9 libdm: improve debug message with ioctl
Make the debug message a less difficult to read:
Ioctl shows  [ noopencount flush ] instead of cryptic [NF].
2016-04-06 11:31:02 +02:00
Alasdair G Kergon
a5d53aec83 libdm: Raid status region units are sectors 2016-03-24 17:42:36 +00:00
Alasdair G Kergon
1056894f1f raid: Tidy dm_get_status_raid. [HM] 2016-03-22 21:39:52 +00:00
Alasdair G Kergon
cf39346697 libdm: Move _skip_fields within file. 2016-03-22 19:35:14 +00:00
Alasdair G Kergon
65d2d66d02 libdm: Change _advance_to_next_word to _skip_fields 2016-03-22 19:26:13 +00:00
Zdenek Kabelac
29d1533a49 libdm: parse more info from cache status
Parse Fail/Error/need_check/ro status info from cache.
2016-03-10 18:38:53 +01:00
Peter Rajnoha
9720898c8e libdm: config: fix dm_config_write_node and variants to properly return error on failures
The error was not propagated if _write_config (that is part of
dm_config_write_node and all its variants) was called recursively
for subsections.
2016-03-04 15:51:13 +01:00
Peter Rajnoha
67c5006e12 cleanup: previous patch with libdm config node buffer size 2016-03-04 15:19:09 +01:00
Peter Rajnoha
967a0889a0 libdm: config: remove 4096 char limit due to buffer size if writing dm_config_node 2016-03-04 14:59:22 +01:00
Zdenek Kabelac
36d0dcef38 coverity: eliminate DEADCODE
Coverity notice this cannot be NULL: cur = &dms->regions[*cur_r]
so avoid NULL checking and simplify form.
2016-03-01 14:02:43 +01:00
Zdenek Kabelac
79809d6cdc cleanup: use sizeof instead of macro.
Keep the buffer size defined at a single place and then use
its sizeof.
2016-02-23 21:40:17 +01:00
Zdenek Kabelac
6d6e063a0f libdm: fix string boundary
The test for string 'end' needs to account for ending \0,
which also needs to fit <SIZE.
2016-02-23 21:38:52 +01:00
Zdenek Kabelac
dbc71dc05e gcc: cleanup some sign warnings
When comparing unsigned with int, the comparision is made
as 'unsigned' type, so make it rather explicit which type
is being compared.
2016-02-23 12:25:25 +01:00
Zdenek Kabelac
3178cfc818 cleanup: use char arrays. 2016-02-22 14:27:44 +01:00
Zdenek Kabelac
fba54ae55e cleanup: libdm clang warnings
Add some extra clang pointer validation so we do not try deref NULL.
2016-02-22 14:18:28 +01:00
Zdenek Kabelac
275c9f7e77 libdm: grow with initialized struct content
Coverity noticed struct hist has been copied uninitalized into mempool.
2016-02-22 14:17:32 +01:00
Zdenek Kabelac
0fb3669d49 libdm: thin status update
Fix parsing of 'Fail' status (using capital letter) for thin-pool.
Add also parsing of 'Error' state for thin-pool.
Add needs_check test for thin-pool.

Detect Fail state for thin.
2016-02-18 16:45:42 +01:00
Zdenek Kabelac
0baf66a992 dm: alloc always 8byte aligned
Fixing regression caused by 197b5e6dc7.
So the 'TODO' part now finally know the answer - there is 'sparc64'
architecture which imposes limitation to read 64b words only through
64b aligned address.

Since we never could know how is the user going to use the returned
pointer and the userusually expects it's aligned on the highest CPU
required alignement, preserve it also for char*.

Fixes: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=809685
Reported-by: Anatoly Pugachev <matorola@gmail.com>
2016-02-11 18:35:05 +01:00
M.H. Tsai
f91622741f dm: fix thin-pool targer params order
Wrong thin-pool feature flag ordering in dm table: It will lead to
unnecessary table reload.

Fix it by placeing feature flags in order they are returned from the
kernel so current 'table line diff' code will not see a difference.
2016-02-11 18:32:24 +01:00
Zdenek Kabelac
fcbef05aae doc: change fsf address
Hmm rpmlint suggest fsf is using a different address these days,
so lets keep it up-to-date
2016-01-21 12:11:37 +01:00
Zdenek Kabelac
45781161f4 libdm: add some doc for mirror status
Comment content of struct for mirror status.
2015-12-04 22:10:30 +01:00
Zdenek Kabelac
fa87979004 libdm: introduce dm_get_status_mirror
Add missing function to parse mirror status.
2015-12-01 13:00:43 +01:00
Zdenek Kabelac
d582be43d4 libdm: const raid params and error for unsupported type
Accept const struct with raid params (No API change).
Also add extra error message when raid type is unsupported.
2015-11-26 09:27:04 +01:00
Zdenek Kabelac
6ca5447e0c libdm: enhance thin-pool preload
When preloading thin-pool device node for already
existing/running thin-pool do not resume such thin-pool.

This allows to properly schedule commit point for metadata,
when thin-pool data or metadata volume is resized.
2015-11-23 23:34:46 +01:00
Zdenek Kabelac
ddbf0075b1 libdm: drop extra space from cache target line
Extra space between 'cache' target and metadata device caused
string comparation being not equal and thus always causing
table reload even when uneeded.
2015-11-23 23:33:37 +01:00
Zdenek Kabelac
0614c63579 cleanup: cast resulting value explicitely 2015-11-19 11:59:19 +01:00
Zdenek Kabelac
0b2be60497 cleanup: add stack traces 2015-11-18 22:17:26 +01:00
Peter Rajnoha
0a2cadf6b8 libdm: report: consolidate use of string list selection structures 2015-11-18 10:54:09 +01:00
Zdenek Kabelac
83661c8f7f cleanup: use embeded list
Skip pointer and put list into selection_str_list.
2015-11-17 19:01:25 +01:00
Zdenek Kabelac
121341e52c cleanup: unify NULL custom check
Unify testing of NULL custom pointer.
Resolve 'factor' only in required if() branch.
2015-11-17 19:01:25 +01:00
Zdenek Kabelac
51dfba002b libdm: update error message
Correcting error message.
2015-11-17 19:01:25 +01:00
David Teigland
931fede81b hash: change name of new lookup function 2015-11-17 11:59:44 -06:00
David Teigland
485d2ca945 lvmetad: different style for hash functions
In lookup, return a count of entries with the
same key rather than the value from a second
entry with the same key.

Using some slightly different names.
2015-11-17 10:27:16 -06:00
David Teigland
920a281994 hash: add comment about multiple values 2015-11-16 11:02:25 -06:00
Zdenek Kabelac
b6a45963e3 libdm: fix check of pointer
Ahhh being blind here - wanted to check the pointer before dereference
not a dereferenced one.
2015-11-16 13:10:24 +01:00
Zdenek Kabelac
2a2487f02f libdm: better error reporting from dm_split_lvm_name
Report errors from all error paths correctly.
Validate passed args before dereferencing them.
2015-11-16 13:09:53 +01:00
Zdenek Kabelac
96d73dc6ea libdm: check for passed custom time value
Coverity reports custom should be checked before derefernce.
2015-11-16 01:16:11 +01:00
Zdenek Kabelac
d4288c9bdf libdm: check for null from pool strdup
Unlikely to happen, but Coverity shown we may have possible
derefer NULL pointer.
2015-11-16 01:16:09 +01:00
Zdenek Kabelac
422c7474ca libdm: check if passed return pointer is not NULL
Coverity: before storing return value, check passed space will
not dereference NULL pointer.
2015-11-16 01:15:04 +01:00
David Teigland
d9295410e9 lvmetad: change the new hash to take data len
If the data len is passed into the hash table
and saved there, then the hash table internals
do not need to assume that the data value is
a string at any point.
2015-11-13 16:54:22 -06:00
David Teigland
46193f4a59 lvmetad: handle duplicate VG names
New hash table functions are added that allow for
multiple entries with the same key.  Use of the
vgname_to_vgid hash table is converted to these
new functions since there are multiple entries
in vgname_to_vgid that have the same key (vgname).

When multiple VGs with the same name exist, commands
that reference only a VG name will fail saying the
VG could not be found (that error message could be
improved.)  Any command that works with the select
option can access one of the VGs with -S vg_uuid=X.
vgrename is a special case that allows the first VG
name arg to be replaced by a uuid, which also works.

(The existing hash table implementation is not well
suited for handling this case, but it works ok with
the new extensions.  Changing lvmetad to use its own
custom hash tables may be preferable at some point.)
2015-11-13 14:56:35 -06:00
Zdenek Kabelac
d74e1291cd libdm: put in secure check
Coverity complains about NULL deref - while this cannot currently
happen, put in secure INTERNAL_ERROR.
2015-11-13 11:17:05 +01:00
Zdenek Kabelac
6d0db97163 libdm: reorder error path
Coverity noticed recent fix of an error path missed to
release 'dmt' - reoder code to ensure 'dmt' is released.
2015-11-10 21:41:47 +01:00
Zdenek Kabelac
fb59847a0f libdm: replace assign with increment
Coverity didn't liked assign with && expression, so use trick.
It does not complain against this prefix incremenent operation.
2015-11-09 22:51:48 +01:00
Zdenek Kabelac
846adadbcc cleanup: use code in place
Pass const strings to printf(),
and use  struct names directly instead of creating unused vars on stack.
2015-11-09 10:22:52 +01:00
Zdenek Kabelac
3cadc1c87e libdm: add test for dm_task_get_message_response()
Coverity notices dm_task_get_message_response() result should be
checked for NULL which should not be passed to dm_pool_strdup().
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
84303dc17a libdm: exlicitly check for NULL
Coverity: another explict check for NULL, where coverity fails to
see it.
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
f6c140e200 libdm: ensure vars are initialized
Coverity found potential error path, where code could
have used some unset variables.
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
b1c4017743 libdm: add missing error path check
Coverity: do not continue with section cloning when root node
would a NULL.
2015-11-09 10:19:19 +01:00
Zdenek Kabelac
428ca9b120 libdm: enable no_flush for driver version > 11
It appears the driver version 11 has troubles with usage of no_flush
So require at least version 12.
2015-10-26 07:37:59 +01:00
Zdenek Kabelac
9ef820a2a5 libdm: dm_tree_node_size_changed recognizes reduction
Add more functionality to size_changed function.
While 'existing' API only detected  0 for
unchanged,  and  for changed,
new improved API will also detected if the
size has only went bigger - or there was
size reduction.

Function work for the whole dm-tree - so
no change is size is always 0.
only size extension  1.
and if some size reduction is there - returns -1.

This result can be used for better evaluation
whether we need to flush before suspend.
2015-10-25 21:05:15 +01:00
Zdenek Kabelac
39cffa4e9b cleanup: declare vars before code 2015-10-22 22:46:10 +02:00
Zdenek Kabelac
09a62cca0c libdm: add dm_hold_control_dev
Support hold of control device open.
Useful for daemons so the control device is not frequently reopenned.
2015-10-22 22:27:31 +02:00
Peter Rajnoha
c3bfe07f2a config: add report/compact_output_cols to control which columns to compact in report output
The new report/compact_output_cols setting has exactly the same effect
as report/compact_output setting. The difference is that with the new
setting it's possible to define which cols should be compacted exactly
in contrast to all cols in case of report/compact_output.

In case both compact_output and compact_output_cols is enabled/set,
the compact_output prevails.

For example:

$ lvmconfig --type full report/compact_output report/compact_output_cols
compact_output=0
compact_output_cols=""

$ lvs vg
  LV    VG   Attr       LSize Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert
  lvol0 vg   -wi-a----- 4.00m

---

$ lvmconfig --type full report/compact_output report/compact_output_cols
compact_output=0
compact_output_cols="data_percent,metadata_percent,pool_lv,move_pv,origin"

$ lvs vg
  LV    VG   Attr       LSize Log Cpy%Sync Convert
  lvol0 vg   -wi-a----- 4.00m

---

$ lvmconfig --type full report/compact_output report/compact_output_cols
compact_output=1
compact_output_cols="data_percent,metadata_percent,pool_lv,move_pv,origin"

$ lvs vg
  LV    VG   Attr       LSize
  lvol0 vg   -wi-a----- 4.00m
2015-10-16 17:05:54 +02:00
Peter Rajnoha
508f0f5a21 libdm: add dm_report_compact_given_fields
dm_report_compact_given_fields is the same as dm_report_compact_fields,
but it processes only given fields, not all the fields in the report
like dm_report_compact_field does.
2015-10-16 17:05:54 +02:00
Zdenek Kabelac
362558cd66 cleanup: typo in comment 2015-10-13 15:22:58 +02:00
Zdenek Kabelac
5695c6aca6 libdm: enforce writethrough mode for cleaner
With "cleaner" policy always use 'writethrough' mode.
2015-10-13 14:35:48 +02:00
Alasdair G Kergon
0173c260d8 libdm: Move status fns from deptree to targets.
libdm-deptree is only for functions working with dm tree nodes.
2015-09-28 20:28:31 +01:00
Heinz Mauelshagen
1945a0f504 libdm: fix bogus macro causing false parameter count 2015-09-24 14:22:52 +02:00
Heinz Mauelshagen
4e60e62444 raid: Fix raid target write_behind parameter.
Now uses correct "max_write_behind" instead of "writebehind".
(Includes some tidying up.)
2015-09-23 15:53:27 +01:00
Heinz Mauelshagen
96a6210198 libdm: Improve raid segment parameter handling. 2015-09-23 15:25:46 +01:00
Zdenek Kabelac
e0d915a873 libdm: parse Overflow string from snapshot status
This is likely to be a new 'info' provided by kernel
snapshot target.
For now just parse this string.
2015-09-18 17:45:45 +02:00
Peter Rajnoha
6c0b4a2769 libdm: file: add proper checks for directory components in dm_create_dir
Also make error messages more consistent:

Before this patch:

(/run/lock exists and is not a directory)
$ pvs
  /run/lock/lvm: mkdir failed: Not a directory
  File-based locking initialisation failed.

(/run/lock/lvm exists and is not a directory)
$ pvs
  Directory "/run/lock/lvm" not found
  File-based locking initialisation failed.

With this patch applied:

(/run/lock exists and is not a directory)
$ pvs
  Existing path /run/lock is not a directory.
  Failed to create directory /run/lock/lvm.
  File-based locking initialisation failed

(/run/lock/lvm exists and is not a directory)
$ pvs
  Existing path /run/lock/lvm is not a directory.
  Failed to create directory /run/lock/lvm.
  File-based locking initialisation failed.
2015-09-17 14:29:51 +02:00
Peter Rajnoha
afdae26c71 libdm: dev_node: use lstat instead of stat while removing and renaming nodes
When using udev, the /dev/mapper entries are symlinks - fix the code
to count with this.

This patch also fixes the dmsetup mknodes and vgmknodes to properly
repair /dev/mapper content if it sees dangling symlink in /dev/mapper.
2015-09-17 13:37:15 +02:00
Peter Rajnoha
b5022102bb libdm: report: make it possible to use blank value as selection for string list report field
$ lvs -o name,tags vg
  LV    LV Tags
  lvol0
  lvol1 mytag

Before this patch:
$ lvs -o name,tags vg -S 'tags=""'
  Failed to parse string list value for selection field lv_tags.
  Selection syntax error at 'tags=""'.
  Use 'help' for selection to get more help.

(and the same for -S 'tags={}' and -S 'tags=[]')

With this patch applied:
$ lvs -o name,tags vg -S 'tags=""'
  LV    LV Tags
  lvol0

(and the same for -S 'tags={}' and -S 'tags=[]')
2015-09-17 10:19:15 +02:00
Alasdair G Kergon
a729b1aa29 pre-release 2015-09-15 13:17:50 +01:00
Zdenek Kabelac
c356991fa8 libdm: no validate for pool without messages
Avoid validation of free space in pool, when no messages are passed.

Patch a3c7e326c3 add new check for
pool overload - but this check should not be made if there are
no messages and transaction_id is still within 'bounds' (bigger by 1).
2015-09-14 20:18:54 +02:00
Bryn M. Reeves
d7f45ebca5 libdm: clean up stats local variable use 2015-09-07 20:14:53 +01:00
Bryn M. Reeves
daa94eb792 libdm: fix display of nsec suffixes in histogram strings 2015-09-07 20:14:53 +01:00
Bryn M. Reeves
5f990473e4 libdm: clean up _build_histogram_arg()
Split up _build_histogram_arg() into separate functions to allocate
and fill the histogram arg string and remove nested local variable
declarations from the parent function.
2015-09-07 19:30:03 +01:00
Bryn M. Reeves
4bc7a86f3a libdm: only free the first histogram explicitly (Coverity)
Coverity flags a user-after-free in _stats_histograms_destroy():

>>>     Calling "dm_pool_free" frees pointer "mem->chunk" which has
>>>     already been freed.

This should not be possible since the histograms are destroyed in
reverse order of allocation:

 203         for (n = _nr_areas_region(region) - 1; n; n--)
 204                 if (region->counters[n].histogram)
 205                         dm_pool_free(mem, region->counters[n].histogram);

It appears that Coverity is unaware that pool->chunk is updated
during the call to dm_pool_free() and valgrind flags no errors in
this function when called with multiple allocated histograms.

Since there is no actual need to free the histograms individually
in this way simplify the code and just free the first allocated
object (which will also free all later allocated histograms in a
single call).
2015-09-07 17:53:56 +01:00
Bryn M. Reeves
ab1b54c3e3 libdm: fix dm_stats leak in dm_stats_create()
The histogram changes adds a new error path to dm_stats_create().
Make sure that the dm_stats handle is properly destroyed if we fail
to create the histogram pool and check for failures setting the
program_id.
2015-09-07 12:08:34 +01:00
Bryn M. Reeves
0f5933ecc1 libdm: handle pool errors in dm_histogram_to_string() 2015-09-07 12:01:20 +01:00
Bryn M. Reeves
e75b4bc2df libdm: check dm_pool_begin_object() return value. 2015-09-07 11:52:54 +01:00
Bryn M. Reeves
36b09fd147 libdm: add missing error handling in _stats_parse_histogram()
Since we are growing an object in the histogram pool the return
value of dm_pool_grow_object() must be checked and error paths need
to abandon the object before returning.
2015-09-07 11:44:53 +01:00
Alasdair G Kergon
fb12308416 style: Standardise some error paths. 2015-09-05 23:56:30 +01:00
Bryn M. Reeves
cdca2782d2 libdm: fix uninitialized variable warnings on older gcc
Older versions of gcc aren't able to track the assignments of
local variables as well as the latest versions leading to spurious
warnings like:

libdm-stats.c:2183: warning: "len" may be used uninitialized in this
function
libdm-stats.c:2177: warning: "minwidth" may be used uninitialized in
this function

Both of these variables are in fact assigned in all possible paths
through the function and later compilers do not produce these
warnings.

There's no reason to not initialize these variables though and
it makes the function slightly easier to follow.

Also fix one use of 'unsigned' for a nr_bins value.
2015-09-04 11:46:48 +01:00
Zdenek Kabelac
0a389691dc cleanup: avoid printing -0.00
Nice trick to not print -0.00 for some percent values.
2015-09-03 23:34:37 +02:00
Zdenek Kabelac
a3c7e326c3 libdm: relocate parsing of thin-pool status
Use single routine for parsing status.

Internally we do not need to allocate pool memory for
passed struct.
2015-09-03 23:34:36 +02:00
Bryn M. Reeves
f09e4f7b10 libdm: allow formatting histogram strings with no whitespace
Allow dm_histogram_to_string() to format histogram strings with
no whitespace by passing a width value less than zero.
2015-09-03 22:04:10 +01:00
Peter Rajnoha
fc35b6988d libdm: pkgconfig: fix devmapper.pc to not reference nonexistent rt.pc file
librt doesn't have a pkgconfig file so use Libs.private: -lrt instead
to declare the dependency directly.

The same applies for -lm which is also used and which hasn't been
defined in the devmapper.pc file yet.
2015-09-03 09:28:42 +02:00
Bryn M. Reeves
a0cf3d47f1 libdm: add latency histogram support
Add support for creating, parsing, and reporting dm-stats latency
histograms on kernels that support precise_timestamps.

Histograms are specified as a series of time values that give the
boundaries of the bins into which I/O counts accumulate (with
implicit lower and upper bounds on the first and last bins).

A new type, struct dm_histogram, is introduced to represent
histogram values and bin boundaries.

The boundary values may be given as either a string of values (with
optional unit suffixes) or as a zero terminated array of uint64_t
values expressing boundary times in nanoseconds.

A new bounds argument is added to dm_stats_create_region() which
accepts a pointer to a struct dm_histogram initialised with bounds
values.

Histogram data associated with a region is parsed during a call to
dm_stats_populate() and used to build a table of histogram values
that are pointed to from the containing area's counter set. The
histogram for a specified area may then be obtained and interogated
for values and properties.

This relies on kernel support to provide the boundary values in
a @stats_list response: this will be present in 4.3 and 4.2-stable. A
check for a minimum driver version of 4.33.0 is implemented to ensure
that this is present (4.32.0 has the necessary precise_timestamps and
histogram features but is unable to report these via @stats_list).

Access methods are provided to retrieve histogram values and bounds
as well as simple string representations of the counts and bin
boundaries.  Methods are also available to return the total count
for a histogram and the relative value (as a dm_percent_t) of a
specified bin.
2015-09-02 20:48:59 +01:00
Bryn M. Reeves
c4f3732c91 libdm: reset report field widths in _destroy_rows()
For repeating reports field widths should be re-calculated for
each report interval. Not doing so will cause a single row with
wide field data to cause all subsequent rows to share the width:

Name                                      RgID ArID R/s     W/s    Histogram                                     Bounds
vg_hex-lv_home                               0    0 4522.00 834.00 0s:   991, 2ms:   152, 4ms:   161, 6ms:  4052 0s, 2ms, 4ms, 6ms
vg_hex-lv_swap                               0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms
vg_hex-lv_root                               0    0 1754.00 683.00 0s:  369, 2ms:   65, 4ms:   90, 6ms: 1913     0s, 2ms, 4ms, 6ms
luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e    0    0 4522.00 868.00 0s:   985, 2ms:   152, 4ms:   161, 6ms:  4092 0s, 2ms, 4ms, 6ms
vg_hex-lv_images                             0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms

Name                                      RgID ArID R/s     W/s    Histogram                                     Bounds
vg_hex-lv_home                               0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms
vg_hex-lv_swap                               0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms
vg_hex-lv_root                               0    0    0.00   2.00 0s: 1, 2ms: 0, 4ms: 0, 6ms: 1                 0s, 2ms, 4ms, 6ms
luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e    0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms
vg_hex-lv_images                             0    0    0.00   0.00 0s: 0, 2ms: 0, 4ms: 0, 6ms: 0                 0s, 2ms, 4ms, 6ms
                                                                                                ^^^^^^^^^^^^^^^^^
This is especially significant for the current histogram fields:
depending on the time since the last clear operation the first
report iteration may contain very large values leading to a very
large minimum field width. Without resetting field widths this
large minimum field width value is used for all subsequent rows.
2015-09-02 20:48:59 +01:00
Alasdair G Kergon
746b1bcf2a libdm: Drop ignored duplicate export designation.
dm_stats_create_region is now assigned to DM_1_02_106 by default:
the DM_1_02_104 .exported_symbols file entry was moved into
libdm-stats.c as:
  DM_EXPORT_SYMBOL(dm_stats_create_region, 1_02_104)
so delete it from .exported_symbols.DM_1_02_104.
2015-08-26 17:30:36 +01:00
Bryn M. Reeves
567189cc76 libdm: add per region precise timestamps property methods 2015-08-24 20:03:21 +01:00
Bryn M. Reeves
f4262026b6 libdm: add precise timestamps support to libdm-stats
Add support for the kernel precise_timestamps feature. This allows
regions to be created using counters with nanosecond precision.

A new dm_stats method, dm_stats_set_precise_timestamps() causes all
future regions created with this handle to attempt to enable precise
counters.
2015-08-24 20:03:21 +01:00
Bryn M. Reeves
82a27a85b5 macros: fix default symbol export control
Fix the version export macros to make it possible to export two
different DM_* versions of a symbol: currently it is only possible for a
DM_* symbol to override a symbol in Base. Attempting to export two
symbols at different DM_* version levels (e.g. DM_1_02_104 and
DM_1_02_106) leads to a linker error due to a duplicate symbol
definition.

This is because the DM_EXPORTED_SYMBOL macro makes each exported symbol
the default (@@VERSION):

       __asm__(".symver " #func "_v" #ver ", " #func "@@DM_" #ver )

Fix the macro to use a single '@' for a symbols exported in multiple
versions and rename the macros to DM_EXPORT_*:

  DM_EXPORT_SYMBOL(func,ver)
  DM_EXPORT_SYMBOL_BASE(func,ver)

For functions that have multiple implementations these macros control
symbol export and versioning.

Function definitions that exist in only one version never need to use
these macros.

Backwards compatible implementations must include a version tag of
the form "_v1_02_104" as a suffix to the function name and use the
macro DM_EXPORT_SYMBOL to export the function and bind it to the
specified version string.

Since versioning is only available when compiling with GCC the entire
compatibility version should be enclosed in '#if defined(__GNUC__)',
for example:

  int dm_foo(int bar)
  {
    return bar;
  }

  #if defined(__GNUC__)
  // Backward compatible dm_foo() version 1.02.104
  int dm_foo_v1_02_104(void);
  int dm_foo_v1_02_104(void)
  {
    return 0;
  }
  DM_EXPORT_SYMBOL(dm_foo,1_02_104)
  #endif

A prototype for the compatibility version is required as these
functions must not be declared static.

The DM_EXPORT_SYMBOL_BASE macro is only used to export the base
versions of library symbols prior to the introduction of symbol
versioning: it must never be used for new symbols.
2015-08-24 20:03:21 +01:00
Bryn M. Reeves
386e91addb libdm: add dm_message_supports_precise_timestamps()
Add a function to test whether the kernel precise_timestamps
feature is available in the current device-mapper driver version.

Presence of precise_timestamps also implies the availability of
latency histograms.
2015-08-20 12:11:23 +01:00
Bryn M. Reeves
c1bd76d6fc configure: check for -lm and log10 function
We already use -lm functions in a couple of places (these are
satisfied by gcc built-ins for most builds): add a configure.in
check and explicitly link to -lm.
2015-08-18 15:25:54 +01:00
Zdenek Kabelac
6e1feb0f73 cleanup: preserve constness of some pointers 2015-08-18 16:05:04 +02:00
Zdenek Kabelac
55a9262bdb cleanup: unused header files (Coverity) 2015-08-18 15:00:08 +02:00
Zdenek Kabelac
ba94d0f144 libdm: simplify dmstats formula.
Since we check for stats for not being 0,
simplify the operation and use a single division.
2015-08-18 15:00:08 +02:00
Bryn M. Reeves
074b5de771 libdm: check for zero in _nr_areas() (Coverity) 2015-08-17 18:37:16 +01:00
Bryn M. Reeves
8967776713 libdm: do not read region before checking dms for NULL (Coverity)
dm_stats_get_area_start() attempts to assign a region pointer from
a stats handle before checking it is non-NULL: move the assignment
after the test.
2015-08-17 18:37:16 +01:00
Alasdair G Kergon
be1db6b6c1 pre-release 2015-08-17 17:20:14 +01:00
Bryn M. Reeves
4227b2ebb4 libdm-stats: only return uint64_t when required
Several interfaced in libdm-stats return a uint64_t when it is
only used to signal success/failure: change all these uses to
return a simple int instead.
2015-08-17 16:59:52 +01:00
Bryn M. Reeves
427d0a5e92 libdm: ensure dm_stats_get_area_offset() returns a value 2015-08-17 14:40:48 +01:00
Zdenek Kabelac
79ea81b8a8 thin: restore transaction_id handling
Revert back to already existing behavior which has been slightly
modified by a900d150e4.

At the end however it seem to be equal to change TID right with first
metadata write.

Existing code missed handling for 'unused' thin-pool which would
require to also check empty message list for TID==0.

So with the fix we now again preserve 'active' thin-pool volume
when first thin volume is created - this property was lost and caused
problems in cluster, where the lock was hold, but volume was no longer
active on the node.

Another missing part was the proper support for already increased,
but unfinished TID change.

So going back here with existing logic -

TID is increased with first MDA update.

Code allows start with either same TID or (TID-1).

If there are messages, TID must be lower by 1 for sending,
otherwise messages were already posted.
2015-08-17 11:25:03 +02:00
Bryn M. Reeves
00ed523659 libdm: add dm_stats_get_{current_}area_offset()
Add a method to retrieve the offset of an area within the
containing region (rather than the offset within the containing
device returned by dm_stats_get_area_start()).

Although users of the library can calculate this themselves it is
better to provide this through a method call to avoid users making
assumptions about the structure of regions and areas.
2015-08-14 22:03:37 +01:00
Bryn M. Reeves
77fae3d852 libdm: ensure dm_stats_get_area_start includes region offset
The dm_stats_get_area_start (and its '_current_' variant) methods
are expected to return the start sector of the area in the
containing device.

Make sure the call adds region->start to the returned value.
2015-08-14 22:03:37 +01:00
Natanael Copa
4534f0fbcf libdm: do not in include internal bits/time.h header
Do not include bits/time.h as it is an internal libc header file.

A comment at the top of the glibc specific bits/time.h says:
"Never include this file directly; use <time.h> instead."

This fixes the following build error with musl libc:
libdm-timestamp.c:37:23: fatal error: bits/time.h: No such file or directory
---
Compile tested with Alpine Linx (musl libc) and ubuntu 15.04

 libdm/libdm-timestamp.c | 1 -
 1 file changed, 1 deletion(-)
2015-08-14 11:33:12 +01:00
Alasdair G Kergon
043fb32c4b dmsetup: Restructure arg handling.
Introduce enums and global variables to record cleanly which command we
are processing and eliminate the historically inconsistent use of the
shifted argv[0] and fix assorted bugs discovered along the way.

Add dm_report_is_empty() to indicate there is no data awaiting output
and use this to suppress dmsetup report headings when no data is output
so we don't get a stray line saying 'Help' at the end of reporting help.

Define a report type (as the interface requires) so -o all selects
the right fields in splitname.  (A fix for stats list will follow.)

Exit immediately if no device is supplied to dmsetup wipe_table instead
of hitting errors later and failing.

Adjust the command name printed in usage/help output to match command
invoked (most of the time).
2015-08-13 22:30:39 +01:00
Bryn M. Reeves
99f55abc56 libdm: add dm_timestamp_copy() 2015-08-12 15:09:57 +01:00
Zdenek Kabelac
48ed8ac50c cleanup: indent 2015-08-12 14:33:16 +02:00
Zdenek Kabelac
79e9bde0ea libdm: rename to data_block_size
Use common name for pool device - as we use data_block_size
for thin pool metadata, use same name for cache_pool.

This change does not affect API.
2015-08-12 14:33:15 +02:00
Zdenek Kabelac
08f047eb51 libdm: cache target arg validation
Add some arg validation for dm_tree_node_add_cache_target().
2015-08-12 14:33:15 +02:00
Bryn M. Reeves
f072a76326 libdm-stats: backtrace if fclose fails (Coverity)
Since libdm-stats only uses fmemopen'd FILE objects the only way
that a close can fail is corruption of the memory containing the
FILE: check for this case and emit a backtrace if it occurs.

libdm/libdm-stats.c: 338 in _stats_parse_list()
libdm/libdm-stats.c: 341 in _stats_parse_list()
libdm/libdm-stats.c: 481 in _stats_parse_region()
libdm/libdm-stats.c: 487 in _stats_parse_region()
libdm/libdm-stats.c: 487 in _stats_parse_region()
 - Calling "fclose" without checking return value
2015-08-10 20:26:07 +01:00
Bryn M. Reeves
856f9cced8 libdm: simplify stats nr_areas calculation (Coverity)
Remove an unneccessary conditional operator and simplify the logic
in _nr_areas:

libdm/libdm-stats.c: 501 in _nr_areas() - Control flow issues  (DEADCODE)
2015-08-10 20:26:07 +01:00
Bryn M. Reeves
f9f5aac123 libdm: fix stats handle leak in dm_stats_create (Coverity)
Make sure the newly created handle is freed if we are unable to
also create the pool for it.

tools/dmsetup.c: 4255 in _stats_list() - Variable "dms" going out of scope leaks the storage it points to.
2015-08-10 20:20:30 +01:00
Bryn M. Reeves
1134de3c89 libdm: fix FILE leak in _program_id_from_proc() (Coverity)
Make sure comm is closed in the error path of _program_id_from_proc().

libdm/libdm-stats.c: 98 in dm_stats_create() - Variable "comm" going out of scope leaks the storage it points to.
2015-08-10 20:20:26 +01:00
Bryn M. Reeves
d62a8d2f15 dmstats: add libdm-stats library and 'dmsetup stats' command
Add the libdm-stats module to libdm: this implements a simple interface
for creating, managing and interrogating I/O statistics regions and
areas on device-mapper devices.

The library interface is documented in libdevmapper.h and provides a
'dm_stats' handle that is used to perform statistics operations and
obtain data.

Public methods are provided to create and destroy handles and to list,
create, and destroy statistics regions as well as to obtain and parse
counter data and calculate rate-based metrics.

This commit also adds a 'dmsetup stats' (aka 'dmstats') command with
'clear', 'create', 'delete', 'list', 'print', and 'report' sub-commands.

See the library documentation and the dmstats.8 manual page for detailed
API and command descriptions.
2015-08-09 14:37:58 +01:00
Bryn M. Reeves
3638c05ec9 libdm: do not attempt to output column headings with --rows
Columns-as-rows output does not use _report_headings(); don't
try to call it when rh->flags & DM_REPORT_OUTPUT_COLUMNS_AS_ROWS.
2015-08-08 17:48:30 +01:00
Bryn M. Reeves
54815fff06 libdm: remove report interval support
Don't do interval management and external timekeeping for stats in
dm_report: let applications handle this on their own.

Since this has not been included in a release remove it from the
library entirely and handle report timing directly inside dmsetup.
2015-08-08 11:48:12 +01:00
Bryn M. Reeves
666c77c0f2 libdm: add dm_report_column_headings
Add a function to print column headings regardless of whether they
have already been output. This will be used by dmstats to issue
periodic reminders of the column headings.

This patch removes a check for RH_HEADINGS_PRINTED from
_report_headings that prevents headings being displayed if the flag
is already set; this check is redundant since the only existing
caller (_output_as_columns()) already tests the flag before
calling the function.
2015-08-08 11:43:52 +01:00
Bryn M. Reeves
cafe145ba2 libdm: fix report rows and headings memory and state leaks
Not releasing objects back to the pool is fine for short-lived
pools since the memory will be freed when dm_pool_destroy() is
called.

Any pool that may be long-lived needs to be more careful to free
objects back to the pool to avoid leaking memory that will not be
reclaimed until the pool is destroyed at process exit time.

The report pool currently leaks each headings line and some row
data.

Although dm_report_output() tries to free the first allocated row
this may end up freeing a later row due to sorting of the row list
while reporting. Store a pointer to the first allocated row from
_do_report_obect() instead and free this at the end of
_output_as_columns(), _output_as_rows(), and dm_report_clear().

Also make sure to call dm_pool_free() for the headings line built
in _report_headings().

When dmstats is introduced it will maintain dm_report objects for
the whole lifetime of the process: without these changes a stats
report could leak around 600k in 10m (exact rate depends on field
selection and data values):

 top - 12:11:32 up 4 days,  3:16, 15 users,  load average: 0.01, 0.12, 0.14
  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 6473 root      20   0  130196   3124   2792 S   0.0  0.0   0:00.00 dmstats

 top - 12:22:04 up 4 days,  3:26, 15 users,  load average: 0.06, 0.11, 0.13
  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 6498 root      20   0  130836   3712   2752 S   0.0  0.0   0:00.60 dmstats

With this patch no increase in RSS is seen:

 top - 13:54:58 up 4 days,  4:59, 15 users,  load average: 0.12, 0.14, 0.14
  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
13962 root      20   0  130196   2996   2688 S   0.0  0.0   0:00.00 dmstats

 top - 14:04:31 up 4 days,  5:09, 15 users,  load average: 1.02, 0.67, 0.36
  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
13962 root      20   0  130196   2996   2688 S   0.3  0.0   0:00.32 dmstats

This also affects report output for repeating reports in the
DM_REPORT_OUTPUT_COLUMNS_AS_ROWS case; row state is not fully cleared for
the next iteration leading to progressive growth of the heading width:

vg_hex-lv_home:vg_hex-lv_swap:vg_hex-lv_root:luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e:vg_hex-lv_images
253:253:253:253:253
2:0:1:4:3
L--w:L--w:L--w:L--w:L--w
1:2:1:1:1
3:1:1:1:2
0:0:0:0:0
LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiv08BCGvF4WsJSqWUDUt7qtf2hEmjtVvo:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiKf7XIiwdAYOJfaGhQe9fu26cTEICGgFS:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiEZj7ZXbmrWDuGhd7vvi88VF0NdTMG8iA:CRYPT-LUKS1-797339213f684c929eb7d0aca4c6ba3e-luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOi2rKredlBPnw2X7v1BiCuEpFo6gaE7BRw

:::::vg_hex-lv_home:vg_hex-lv_swap:vg_hex-lv_root:luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e:vg_hex-lv_images
:::::253:253:253:253:253
:::::2:0:1:4:3
:::::L--w:L--w:L--w:L--w:L--w
:::::1:2:1:1:1
:::::3:1:1:1:2
:::::0:0:0:0:0
:::::LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiv08BCGvF4WsJSqWUDUt7qtf2hEmjtVvo:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiKf7XIiwdAYOJfaGhQe9fu26cTEICGgFS:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOiEZj7ZXbmrWDuGhd7vvi88VF0NdTMG8iA:CRYPT-LUKS1-797339213f684c929eb7d0aca4c6ba3e-luks-79733921-3f68-4c92-9eb7-d0aca4c6ba3e:LVM-9t8ITqLZa6AuuyVoz5Olp1KwF9ZDBfOi2rKredlBPnw2X7v1BiCuEpFo6gaE7BRw
2015-08-08 11:35:10 +01:00
Bryn M. Reeves
974e7b9220 libdm: ensure new dm_timestamp objects are initialized
Allocate a dm_timestamp with dm_zalloc() to ensure the memory is
initialized to a known state.
2015-08-08 10:42:14 +01:00
Alasdair G Kergon
559ca8bc65 dmsetup: Report timestamps of ioctls with -vvv.
If enabled, record timestamp immediately after the ioctl() returns.
2015-08-05 08:28:35 +01:00
Alasdair G Kergon
2f334afb98 libdm: Whitespace. 2015-08-05 05:21:58 +01:00
Alasdair G Kergon
88551add97 libdm: Whitespace. 2015-08-05 05:11:42 +01:00
Alasdair G Kergon
23e8e849e4 libdm: Sort new exported symbols. 2015-08-05 05:07:45 +01:00
Peter Rajnoha
6ac5689ce4 report: also recognize variants without underscores for <prefix>_all fields
For example: "pvs -o pv_all" and pvs -o pvall" are same.
2015-08-04 09:03:31 +02:00
Peter Rajnoha
71dbe47619 report: update comment for _is_same_field fn 2015-08-03 16:47:02 +02:00
Peter Rajnoha
a5b476a7d3 report: recognize report field name variants without any underscores too
Whenver reporting field name is registered with libdevmapper and if
the field name contains any number of underscores ('_'), libdm
can now automatically recognize any of its variant without any
underscores used.

For example:

..for underscores in prefixes:
  pvs -o pv_name
  pvs -o name
  pvs -o pvname (newly recognized besides pvname)

..for underscores in the name:
  lvs -o cache_mode
  lvs -o cachemode

..or even multiple underscores:
  pvs -o pv___na___me

It's all variant of the same field name.
2015-08-03 16:29:50 +02:00
Bryn M. Reeves
a161e29c59 dmsetup: Add --count and --interval to reports.
For example, to monitor active devices every second you can now run
dmsetup info -c --count 0.
2015-07-31 21:59:34 +01:00
Alasdair G Kergon
3cd644aeb5 libdm: Require librt for new dm_timestamp. 2015-07-29 19:30:22 +01:00
Alasdair G Kergon
a28fb37b9e libdm: Add dm_timestamp functions. 2015-07-29 19:21:07 +01:00
Alasdair G Kergon
a5491d3698 dmsetup: Accept vg/lv name format.
If there is exactly one / which is not the first character, check
for /dev/vg/lv (as dm_dir()/../$name i.e. /dev/mapper/../vg/lv.)
2015-07-29 12:24:36 +01:00
Alasdair G Kergon
af1c7bf0c7 libdm: Add dm_size_to_string to libdevmapper.
Moved out from lib/display and a little documentation added.
It's tuned to LVM's requirements historically and its behaviour
might not always be what you would expect.
2015-07-27 21:30:20 +01:00
Alasdair G Kergon
1612c570b6 libdm: Use wrappers for all malloc functions.
Move the DEBUG_MEM decision inside libdevmapper.so instead of exposing
it in libdevmapper.h which causes failures if the binary and library
were compiled with opposite debugging settings.
2015-07-22 23:11:48 +01:00
Zdenek Kabelac
c45e6e3c78 cleanup: avoid double assign
Variable n1 is assigned without using n1 before.
2015-07-15 13:10:22 +02:00
Peter Rajnoha
3ec4813ba2 coverity: fix missing initialization
... Using uninitialized value "lockd_state" when calling "lockd_vg"
(even though lockd_vg assigns 0 to the lockd_state, but it looks at
previous state of lockd_state just before that so we need to have
that properly initialized!)

libdm/libdm-report.c:2934: uninit_use_in_call: Using uninitialized value "tm". Field "tm.tm_gmtoff" is uninitialized when calling "_get_final_time".

daemons/lvmlockd/lvmlockctl.c:273: uninit_use_in_call: Using uninitialized element of array "r_name" when calling "format_info_r_action". (just added FIXME as this looks unfinished?)
2015-07-08 14:53:30 +02:00
Jonathan Brassow
4daea88516 clean-up: typos s/bellow/below/ 2015-07-06 10:15:11 -05:00
Alasdair G Kergon
810ab095e6 macros: Wrap PRI with FMT.
Create a set of wrappers with embedded % such as
  #define FMTu64 "%" PRIu64
2015-07-06 15:09:17 +01:00
Zdenek Kabelac
a900d150e4 thin: move pool messaging from resume to suspend
Existing messaging intarface for thin-pool has a few 'weak' points:

* Message were posted with each 'resume' operation, thus not allowing
activation of thin-pool with the existing state.

* Acceleration skipped suspend step has not worked in cluster,
since clvmd resumes only nodes which are suspended (have proper lock
state).

* Resume may fail and code is not really designed to 'fail' in this
phase (generic rule here is resume DOES NOT fail unless something serious
is wrong and lvm2 tool usually doesn't handle recovery path in this case.)

* Full thin-pool suspend happened, when taken a thin-volume snapshot.

With this patch the new method relocates message passing into suspend
state.

This has a few drawbacks with current API, but overal it performs
better and gives are more posibilities to deal with errors.

Patch introduces a new logic for 'origin-only' suspend of thin-pool and
this also relates to thin-volume when taking snapshot.

When suspend_origin_only operation is invoked on a pool with
queued messages then only those messages are posted to thin-pool and
actual suspend of thin pool and data and metadata volume is skipped.

This makes taking a snapshot of thin-volume lighter operation and
avoids blocking of other unrelated active thin volumes.

Also fail now happens in 'suspend' state where the 'Fail' is more expected
and it is better handled through error paths.

Activation of thin-pool is now not sending any message and leaves upto a tool
to decided later how to finish unfinished double-commit transaction.

Problem which needs some API improvements relates to the lvm2 tree
construction. For the suspend tree we do not add target table line
into the tree, but only a device is inserted into a tree.
Current mechanism to attach messages for thin-pool requires the libdm
to know about thin-pool target, so lvm2 currently takes assumption, node
is really a thin-pool and fills in the table line for this node (which
should be ensured by the PRELOAD phase, but it's a misuse of internal API)
we would possibly need to be able to attach message to 'any' node.

Other thing to notice - current messaging interface in thin-pool
target requires to suspend thin volume origin first and then send
a create message, but this could not have any 'nice' solution on lvm2
side and IMHO we should introduce something like 'create_after_resume'
message.

Patch also changes the moment, where lvm2 transaction id is increased.
Now it happens only after successful finish of kernel transaction id
change. This change was needed to handle properly activation of pool,
which is in the middle of unfinished transaction, and also this corrects
usage of thin-pool by external apps like Docker.
2015-07-03 16:13:14 +02:00
Zdenek Kabelac
5bef18f2eb libdm: support for posting messages in suspend
Add support for sending message in suspend tree for thin-pools.
When this operation is requested whole subtree suspend is then skipped.

This is experimantal support for new lvm2 code for sending message
in suspend phase where 'thin-pool origin-only suspend' will send
messages instead of really suspending thin-pool tree.

When suspening thin volume origin-only - only thin volume is suspended,
then messages are posted and thin-pool suspend is skipped.
2015-07-03 16:13:14 +02:00
Peter Rajnoha
9cee94372a report: select: add handler to recognize fuzzy time specification
Recognize date and time specification within selection criteria
that is formulated in a more free-form way besides to the original
basic YYYY-MM-DD HH:MM format that libdevmapper supports.

Currently, this free-form format is recognized for lv_time field.

Users are able to use expressions from this set:
  - weekday names ("Sunday" - "Saturday" or abbreviated as "Sun" - "Sat")
  - labels for points in time ("noon", "midnight")
  - labels for a day relative to current day ("today", "yesterday")
  - points back in time with relative offset from today (N is a number)
    ( "N" "seconds"/"minutes"/"hours"/"days"/"weeks"/"years" "ago")
    ( "N" "secs"/"mins"/"hrs" ... "ago")
    ( "N" "s"/"m"/"h" ... "ago")
  - time specification either in hh:mm:ss format or with AM/PM suffixes
  - month names ("January" - "December" or abbreviated as "Jan" - "Dec")

For example:

$ date
Fri Jul  3 10:11:13 CEST 2015

$ lvmconfig --type full report/time_format
time_format="%a %Y-%m-%d %T %z %Z [%s]"

$ lvs
  LV    VG     Time
  lvol0 vg     Fri 2014-08-22 21:25:41 +0200 CEST [1408735541]
  lvol2 vg     Sun 2015-04-26 14:52:20 +0200 CEST [1430052740]
  root  fedora Wed 2015-05-27 08:09:21 +0200 CEST [1432706961]
  swap  fedora Wed 2015-05-27 08:09:21 +0200 CEST [1432706961]
  lvol1 vg     Tue 2015-06-30 03:25:43 +0200 CEST [1435627543]
  lvol3 vg     Tue 2015-06-30 14:52:23 +0200 CEST [1435668743]
  lvol6 vg     Wed 2015-07-01 13:35:56 +0200 CEST [1435750556]
  lvol4 vg     Thu 2015-07-02 12:12:02 +0200 CEST [1435831922]
  lvol5 vg     Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

$ lvs -S 'time=yesterday'
  LV    VG   Time
  lvol4 vg   Thu 2015-07-02 12:12:02 +0200 CEST [1435831922]
  lvol5 vg   Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

$ lvs -S 'time since "June 30"'
  LV    VG   Time
  lvol1 vg   Tue 2015-06-30 03:25:43 +0200 CEST [1435627543]
  lvol3 vg   Tue 2015-06-30 14:52:23 +0200 CEST [1435668743]
  lvol6 vg   Wed 2015-07-01 13:35:56 +0200 CEST [1435750556]
  lvol4 vg   Thu 2015-07-02 12:12:02 +0200 CEST [1435831922]
  lvol5 vg   Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

$ lvs -S 'time since "noon June 30"'
  LV    VG   Time
  lvol3 vg   Tue 2015-06-30 14:52:23 +0200 CEST [1435668743]
  lvol6 vg   Wed 2015-07-01 13:35:56 +0200 CEST [1435750556]
  lvol4 vg   Thu 2015-07-02 12:12:02 +0200 CEST [1435831922]
  lvol5 vg   Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

$ lvs -S 'time since "2 July 9AM"'
  LV    VG   Time
  lvol4 vg   Thu 2015-07-02 12:12:02 +0200 CEST [1435831922]
  lvol5 vg   Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

$ lvs -S 'time since "2 July 1PM"'
  LV    VG   Time
  lvol5 vg   Thu 2015-07-02 14:30:32 +0200 CEST [1435840232]

...and so on.
2015-07-03 10:51:31 +02:00
Peter Rajnoha
3b1422c45c report: call appropriate handler to evaluate fuzzy reserved names and dynamic reserved values
Wire the dm_report_reserved_handler instance call in reporting/selection
infrastructure to handle reserved value actions (currently only
DM_REPORT_RESERVED_PARSE_FUZZY_NAME and DM_REPORT_RESERVED_GET_DYNAMIC_VALUE
actions).
2015-07-03 10:47:38 +02:00
Peter Rajnoha
335707b0e2 report: add infrastructure to recognize fuzzy reserved names and returning dynamic reserved values
With fuzzy names we mean the names for which it's hard or even impossible
to enumerate all possible variations of the name - the name needs to
be evaluated. An example of fuzzy name is a name which has a base
(substring) which matches and it can contain arbitrary variations
around this base. We can cover human language better with fuzzy
names as people may use several different names (or sentences) to
denote the same thing.

With dynamic values we mean the values which are not constants
and they need to be evaluated in runtime. An example of dynamic
value is a value which depends on current system state (e.g. time,
current configuration or any other state which may change and it
needs runtime evaluation).

There's a handler that can be registered with reporting/selection
using dm_report_reserved_handler instance. This is a central point
in which the computation/evaluation happens when processing reserved
values. Currently, there are two actions declared:

  DM_REPORT_RESERVED_PARSE_FUZZY_NAME
  (translates fuzzy name into canonical name)

  DM_REPORT_RESERVED_GET_DYNAMIC_VALUE
  (gets value for canonical name)

The handler is then registered as value in struct
dm_report_reserved_value (see explaining comments besided
the struct dm_report_reserved_value in libdevmapper.h).

Also, this patch provides support for simple caching of values
used during report/selection via dm_report_value_cache_{set,get}.
This is supposed to be used mainly in the dm_report_reserved_handler
instances to save values among calls so all the handler calls work
with the same base value used in computation/evaluation and/or
possibly to save resources if the evaluation is more time-consuming.
The cache is attached to the dm_report handle and so the cache is
dropped one dm_report is dropped.
2015-07-03 10:47:09 +02:00
Peter Rajnoha
a32d5a4afc report: adjust shared flags based on expected type for reserved values
Generic numbers and time values share some operators so make sure
we have the flags correctly adjusted based on expected type if
we're using reserved values.
2015-07-02 16:12:01 +02:00
Peter Rajnoha
454782f1a3 report: fix regression while selecting string fields using synonyms
$ lvs -o name,cache_policy vg/lvol0
  LV    Cache Policy
  lvol0

Before this patch:
$ lvs -o name,cache_policy -S 'cache_policy=undefined' vg/lvol0
  (no match)

With this patch applied:
$ lvs -o name,cache_policy -S 'cache_policy=undefined' vg/lvol0
  LV    Cache Policy
  lvol0
2015-07-02 11:31:54 +02:00
Zdenek Kabelac
21c0b1134f libdm: enhance tracing messages
Use new _node_name() and print name major:minor for thin-pool device.
2015-07-01 13:44:28 +02:00
Zdenek Kabelac
04ae5007e3 libdm: add helper function to print _node_name
_node_name() prepares into dm_tree internal buffer device
name and it (major:minor) for easy usage for debug messages.

To avoid any allocation a small buffer in struct dm_tree is preallocated
to store this message.
2015-07-01 13:41:40 +02:00
Alasdair G Kergon
f6ad48f0e5 libdm: Rename struct time_value variables.
warning: declaration of ‘time’ shadows a global declaration
2015-06-30 16:17:22 +01:00
Peter Rajnoha
ded279f826 report: add support for time (basic)
This patch adds support for time values used in reporting fields.
The raw values are always stored as number of seconds since epoch.

The support that comes with this patch is the basic one which allows
only for recognition of strictly formatted date and time in selection
criteria (the format follows a subset of formats defined by ISO 8601):

  date time timezone

  date:
    YYYY-MM-DD (or shortly YYYYMMDD)
    YYYY-MM (shortly YYYYMM), auto DD=1
    YYYY, auto MM=01 and DD=01

  time:
    hh:mm:ss (or shortly hhmmss)
    hh:mm (or shortly hhmm), auto ss=0
    hh (or shortly hh), auto mm=0, auto ss=0

  timezone (always with + or - sign):
    +hh:mm or -hh:mm (or shortly +hhmm or -hhmm)
    +hh or -hh

Or directly the time (number of seconds) since Epoch (1970-01-01 00:00:00 UTC)
when the number value is prefixed by "@":

   @number_of_seconds_since_epoch

This patch also adds aliases for comparison operators
used together with time values which are more intuitive
to use:
  since (as alias for >=)
  after (as alias for >)
  until (as alias for <=)
  before (as alias for <)

For example:

$ lvmconfig --type full report/time_format
time_format="%Y-%m-%d %T %z %Z [%s]"

$ lvs -o name,time vg
  LV    Time
  lvol0 2015-06-28 21:25:41 +0200 CEST [1435519541]
  lvol1 2015-06-30 03:25:43 +0200 CEST [1435627543]
  lvol2 2015-04-26 14:52:20 +0200 CEST [1430052740]
  lvol3 2015-06-30 14:52:23 +0200 CEST [1435668743]

$ lvs vg -o name,time -S 'time since "2015-04-26 15:00" && time until "2015-06-30"'
  LV    Time
  lvol0 2015-06-28 21:25:41 +0200 CEST [1435519541]
  lvol1 2015-06-30 03:25:43 +0200 CEST [1435627543]
  lvol3 2015-06-30 14:52:23 +0200 CEST [1435668743]

$ lvs vg -o name,time -S 'time since "2015-04-26 15:00" && time until "2015-06-30 6:00"'
  LV    Time
  lvol0 2015-06-28 21:25:41 +0200 CEST [1435519541]
  lvol1 2015-06-30 03:25:43 +0200 CEST [1435627543]

$ lvs vg -o name,time -S 'time since @1435519541'
  LV    Time
  lvol0 2015-06-28 21:25:41 +0200 CEST [1435519541]
  lvol1 2015-06-30 03:25:43 +0200 CEST [1435627543]
  lvol3 2015-06-30 14:52:23 +0200 CEST [1435668743]

This is basic time recognition support that is directly a part of
libdevmapper. Recognition of more free-form expressions will be a
part of subsequent patches.
2015-06-30 15:15:10 +02:00
Peter Rajnoha
d7b9349ce7 cleanup: report: use internal wrapper for various variables used for handling reserved values
Just a cleanup - wrap several variables we use to handle reserved
values into a structure for easier manipulation in the code.
2015-06-30 10:47:51 +02:00
Peter Rajnoha
d8996a17d1 select: add support for range reserved values and flagging named-only values
This patch allows for registration and recognition of reserved
values which are ranges, so they're composed of two values actually
to denote the lower and upper bound for the range (stored as an array
with exactly two items to define the boundaries).

Also, this patch allows for flagging reserved values as named-only
which means that such values are not strictly reserved. The strictly
reserved values are reserved values as used before this patch.

Distinction between strictly-reserved and named-only values
is clearly visible with comparisons. Normally, strictly reserved
value is not accounted for if we do "greater than" or "lower than"
comparisons, for example:

1  2  3 ....
   |
  abc

- we have "abc" as reserved value for field with value "2"
- the value reported for the field is "abc" (or "2", it doesn't matter here)
- the selection we're processing is -S 'field < abc'
- the result of the selection gives nothing as "abc" is strictly
reserved value (bound to "2") and there's no order defined for
it and it would only match if we directly compared the value
(so -S 'field = abc' would match)

With named-only values, the "abc" is named-only value for "2",
so selection -S 'field < abc" is the same as using -S 'field < 2'.
The "abc" is just an alias for some value so the value or its
assigned name can be used equally in selection criteria.
2015-06-30 10:47:50 +02:00
Peter Rajnoha
f6de196c21 config: also clone associated id when cloning node using dm_config_clone_node{_with_mem} 2015-06-25 10:21:07 +02:00
Peter Rajnoha
9465963faf config: add support for config value formatting flags
There are two basic groups of formatting flags (32 bits):
  - common ones applicable for all config value types (lower 16 bits)
  - type-related formatting flags (higher 16 bits)

With this patch, we initially support four new flags that
modify the the way the config value is displayed:

  Common flags:
  =============

  DM_CONFIG_VALUE_FMT_COMMON_ARRAY - causes array config values
    to be enclosed in "[ ]" even if there's only one item
    (previously, there was no way to recognize an array with one
     item and scalar value, hence array values with one member
     were always displayed without "[ ]" which libdm accepted
     when reading, but it may have been misleading for users)

  DM_CONFIG_VALUE_FMT_COMMON_EXTRA_SPACE - causes extra spaces to
    be inserted in "key = value" (or key = [ value, value, ... ] in
    case of arrays), compared to "key=value" seen on output before.
    This makes the output more readable for users.

  Type-related flags:
  ===================

  DM_CONFIG_VALUE_FMT_INT_OCTAL - prints integers in octal form with
    "0" as a prefix (libdm's config reading code can handle this via
    strtol just fine so it's properly recognized as number in octal
    form already if there's "0" used as prefix)

  DM_CONFIG_VALUE_FMT_STRING_NO_QUOTES - makes it possible to print
    strings without enclosing " "

This patch also adds dm_config_value_set_format_flags and
dm_config_value_get_format_flags functions to set and get
these formatting flags.
2015-06-24 11:13:37 +02:00
Zdenek Kabelac
69132f55ea libdm: add dm_tree_node_set_thin_pool_read_only
Support thin-pool tree node with activation in read-only mode.
(Native kernel API).
2015-06-18 15:15:39 +02:00
Zdenek Kabelac
9a06ae7b35 libdm: better debug message
Print reason for failing ioctl if thin pool message fails.
2015-06-15 14:48:04 +02:00
Alasdair G Kergon
eeb498627c libdm: Add dm_task_get_errno to return ioctl errno.
There are reports of unexplained ioctl failures when using dmeventd.
An explanation might be that the wrong value of errno is being used.

Change libdevmapper to store an errno set by from dm ioctl() directly
and provide it to the caller through a new dm_task_get_errno() function.

[Replaced f9510548667754d9209b232348ccd2d806c0f1d8]
2015-05-26 15:13:49 +01:00
Zdenek Kabelac
30c3bbcd9e makefiles: better clean
More exact clean of library exported symbols files.

Also use $(firstword) test to check for empty string
so 'make clean' has now cleaner condensed look.

Clean also created include links.
2015-05-18 12:45:42 +02:00
Zdenek Kabelac
fe00b163d6 configure: move DEFS to configure.h 2015-05-18 12:43:25 +02:00
Zdenek Kabelac
797c18d543 libdm: new dm_task_get_info with internal_suspend
Introduce new implmentation of dm_task_get_info() function
with support for reading internal_suspend.
.
This time it is done in a 'versioned' way.

We keep the old fashion dm_task_get_info(Base) to implement
the old behavior of 1.02.95 libdm code.

libdm version 1.02.96 introduced 'macro' wrapper
dm_task_get_info_with_deferred_remove with new implementation
of dm_task_get_info() - we cannot do anything else then to
provide compatible version of this symbol.

Now in version 1.02.97 we add new versioned implementation of
dm_task_get_info(DM_1_02_97) symbol.

This has the effect that i.e. rpm build will finaly resolve proper
dependency on a new symbol - so it will be no longer possible,
to build a new binary and use old library
(rpm -q --provides will show libdevmapper.so.1.02(DM_1_02_97)(64bit))

Also the history is now tracked. If a new function is added (or
reimplemented), it needs to be placed in proper file,
so it could be exported with right versioning symbol.
File .exported_symbols.Base should and any existing older DM
should be treated as read-only after a release.

Also - only libdm has been currently enhanced with versioned .Base
file, as soon as other libs (liblvm, libdevmapper-event) needs changes
they should also get their exported symbol files - meanwhile
make.tmpl handles both cases.
2015-05-15 16:48:22 +02:00
Zdenek Kabelac
a2c9ede6b3 makefiles: assign vars before include
Before we include, set INCLUDE and TARGETS.
Extend CFLAGS after include.
2015-05-14 00:19:33 +02:00
Ondrej Kozina
257f7febc7 libdm-common.c: remove trailing whitespace 2015-05-12 17:16:49 +02:00
Zdenek Kabelac
eadebc3b61 debug: show sys errors 2015-05-08 15:15:10 +02:00
Zdenek Kabelac
5232fd13f3 cleanup: cast minor to dev_t
Let the arithmetic run with a single dev_t type (Coverity).
2015-05-08 15:15:10 +02:00
Zdenek Kabelac
b8dfd7a53d cleanup: indent mismatch
Aling break (Coverity).
2015-05-08 15:15:10 +02:00
Peter Rajnoha
82f6dbfaf7 select: fix matching reserved values while <,<=,>,>= is used in selection criteria
Scenario:

$ vgs -o+vg_mda_copies
  VG     #PV #LV #SN Attr   VSize VFree #VMdaCps
  fedora   1   2   0 wz--n- 9.51g    0  unmanaged
  vg      16   9   0 wz--n- 1.94g 1.83g         2

$ lvs -o+read_ahead vg/lvol6 vg/lvol7
  LV    VG   Attr       LSize Pool Origin Data%  Rahead
  lvol6 vg   Vwi-a-tz-- 1.00g pool lvol5  0.00      auto
  lvol7 vg   Vwi---tz-k 1.00g pool lvol6         256.00k

Before this patch:

$vgs -o vg_name,vg_mda_copies -S 'vg_mda_copies < unmanaged'
  VG   #VMdaCps
  vg          2

Problem:
Reserved values can be only used with exact match = or !=, not <,<=,>,>=.
In the example above, the "unamanaged" is internally represented as
18446744073709551615, but this should be ignored while not comparing
field directly with "unmanaged" reserved name with = or !=. Users
should not be aware of this internal mapping of the reserved value
name to its internal value and hence it doesn't make sense for such
reserved value to take place in results of <,<=,> and >=.
There's no order defined for reserved values!!! It's a special
*reserved* value that is taken out of the usual value range
of that type.

This is very similar to what we have already fixed with
2f7f6932dc, but it's the other way round
now - we're using reserved value name in selection criteria now
(in the patch 2f7f693, we had concrete value and we compared it
with the reserved value). So this patch completes patch 2f7f693.

This patch also fixes this problem:

$ lvs -o+read_ahead vg/lvol6 vg/lvol7 -S 'read_ahead > 32k'
  LV    VG   Attr       LSize Pool Origin Data%  Rahead
  lvol6 vg   Vwi-a-tz-- 1.00g pool lvol5  0.00      auto
  lvol7 vg   Vwi---tz-k 1.00g pool lvol6         256.00k

Problem:
In the example above, the internal reserved value "auto" is in the
range of selection "> 32k" - it shouldn't match as well. Here the
"auto" is internally represented as MAX_DBL and of course, numerically,
MAX_DBL > 256k. But for users, the reserved value should be uncomparable
to any number so the mapping of the reserved value name to its interna
 value is transparent to users. Again, there's no order defined for
reserved values and hence it should never match if using <,<=,>,>=
operators.

This is actually exactly the same problem as already described in
2f7f6932dc, but that patch failed for
size field types because of incorrect internal representation used.

With this patch applied, both problematic scenarios mentioned
above are fixed now:

$ vgs -o vg_name,vg_mda_copies -S 'vg_mda_copies < unmanaged'
(blank)

$ lvs -o+read_ahead vg/lvol6 vg/lvol7 -S 'read_ahead > 32k'
  LV    VG   Attr       LSize Pool Origin Rahead
  lvol7 vg   Vwi---tz-k 1.00g pool lvol6  256.00k
2015-04-24 09:48:57 +02:00
Alasdair G Kergon
8853462528 libdm: Add uuid/devno to ioctl failure log message. 2015-04-23 19:26:52 +01:00
Alasdair G Kergon
1906619187 libdm: Add DM_INTERNAL_SUSPEND_FLAG.
Still needs to be reported by dmsetup.
2015-04-23 18:39:04 +01:00
Peter Rajnoha
f32973c78e select: mention { } use in lvm.8 man page and '-S help' properly 2015-04-08 11:14:16 +02:00
Zdenek Kabelac
4c184e9d6b cleanup: drop unused value assign
Dop unused value assignments.

Unknown is detected via other combination
(!linear && !striped).

Also change the log_error() message into a warning,
since the function is not really returning error,
but still keep the INTERNAL_ERROR.

Ret value is always set later.
2015-02-19 14:43:25 +01:00
Peter Rajnoha
123a3383a0 libdm: report: add dm_report_object_is_selected
The new dm_report_object_is_selected fn makes it possible to opt whether the
object reported should be displayed on output or not. Also, in addition to
that, it makes it possible to save the result of selection (either 0 or 1).

So dm_report_object_is_selected is simply  more general form of object
reporting fn - combinations now allow for:

  dm_report_object_is_selected(rh, object, 1, NULL):
  This is exactly the original dm_report_object fn and it's fully equal
  to it.

  dm_report_object_is_selected(rh, object, 0, selected):
  Do not display the result on output, but save info whether the object
  is selected or not in 'selected' variable.

  dm_report_object_is_selected(rh, object, 1, selected):
  Display the result on output (if it passes selection criteria) and save
  whether the object is selected or not in 'selected' variable.

  dm_report_object(rh, object, 0, NULL):
  This combination is not allowed - it will end up with internal error.
  We're either interested in selection status or we want to display the
  result on output or both, but never nothing of the two.
2015-02-10 16:04:38 +01:00
Petr Rockai
d571eab3b2 configure: Look for valgrind.h independently of VALGRIND_POOLS. 2015-02-05 13:50:34 +01:00
Alasdair G Kergon
25d906dbde dmeventd: Reduce waitevent EINTR message severity. 2015-01-21 12:54:00 +00:00
Alasdair G Kergon
7cfc9a4f64 libdevmapper: Improve incompatible version msg. 2015-01-21 12:23:56 +00:00
Alasdair G Kergon
651549594e libdm-report: Fix order of NULL dm_report check. 2015-01-17 10:50:54 +00:00
Zdenek Kabelac
2908ab3eed thin: errrorwhenfull support
Support error_if_no_space feature for thin pools.
Report more info about thinpool status:
(out_of_data (D), metadata_read_only (M), failed  (F) also as health
attribute.)
2015-01-14 14:52:05 +01:00
Zdenek Kabelac
d202f43fff cleanup: update API for segment reporting
API for seg reporting is breaking internal lvm coding - it cannot
use vgmem mem pool for allocation of reported value.
So use separate pool instead of 'vgmem' for non vg related allocations

Add consts for many function params - but still many other are left
for now as non-const - needs deeper level of change even on libdm side.
2015-01-14 14:50:08 +01:00
Peter Rajnoha
aaf25ec6bd libdm: report: also check whether field type is supported for field-specific reserved value
We only checked global per-report-type reserved values for compatibility
with selection code. This patch also adds a check for per-report-field
reserved values. This avoids problems where unsupported report type is
used as reserved value which could cause hard to debug problems
otherwise. So this additional check stops from registering unsupported
and unhandled per-field reserved values.

Registerting such unsupported reserved value is a programmatic error,
so report internal error in this case to stop us from making a mistake
here in the future or even today where STR_LIST fields can't have
reserved values yet.
2014-12-18 11:29:48 +01:00
Peter Rajnoha
f94f8463b0 libdm: report: fix incorrect memory use while using --select with --unbuffered for reporting
Under certain circumstances, the selection code can segfault:

$ vgs --select 'pv_name=~/dev/sda' --unbuffered vg0
  VG   #PV #LV #SN Attr   VSize   VFree
  vg0    6   3   0 wz--n- 744.00m 588.00m
Segmentation fault (core dumped)

The problem here is the use of --ubuffered together with regex used in
selection criteria. If the report output is not buffered, each row is
discarded as soon as it is reported. The bug is in the use of report
handle's memory - in the example above, what happens is:

  1) report handle is initialized together with its memory pool

  2) selection tree is initialized from selection criteria string
     (using the report handle's memory pool!)

    2a) this also means the regex is initialized from report handle's mem pool

  3) the object (row) is reported

    3a) any memory needed for output is intialized out of report handle's mem pool
    3b) selection criteria matching is executed - if the regex is checked the
        very first time (for the very first row reported), some more memory
        allocation happens as regex allocates internal structures "on-demand",
        it's allocating from report handle's mem pool (see also step 2a)

  4) the report output is executed

  5) the object (row) is discarded, meaning discarding all the mem pool
     memory used since step 3.

Now, with step 5) we have discarded the regex internal structures from step 3b.
When we execute reporting for another object (row), we're using the same
selection criteria (step 3b), but tihs is second time we're using the regex
and as such, it's already initialized completely. But the regex is missing the
internal structures now as they got discarded in step 5) from previous
object (row) reporting (because we're using "unbuffered" reporting).

To resolve this issue and to prevent any similar future issues where each
object/row memory is discarded after output (the unbuffered reporting) while
selection tree is global for all the object/rows, use separate memory pool
for report's selection.

This patch replaces "struct selection_node *selection_root" in struct
dm_report with new struct selection which contains both "selection_root"
and "mem" for separate mem pool used for selection.

We can change struct dm_report this way as it is not exposed via libdevmapper.

(This patch will have even more meaning for upcoming patches where selection
is used even for non-reporting commands where "internal" reporting and
selection criteria matching happens and where the internal reporting is
not buffered.)
2014-12-09 10:41:55 +01:00
Peter Rajnoha
42d71b9af3 libdm: report: return immediately from dm_report_compact_fields without error if there are no rows
Let's make dm_report_compact_fields consistent with dm_report_output fn
which also returns with success immediately if there are no rows.
2014-12-05 15:10:50 +01:00
Peter Rajnoha
f867dc6b29 libdm: report: compact output applicable only if report is buffered 2014-12-05 14:18:51 +01:00
Peter Rajnoha
5edf6a56c4 libdm: report: add dm_report_compact_fields
Add new dm_report_compact_fields function to cause report outout
(dm_report_output) to ignore fields which don't have any value set
in any of the rows reported. This provides support for compact report
output where only fields which have something to report are displayed.
2014-12-05 12:00:28 +01:00
Peter Rajnoha
44394cd246 libdm: remove unimplemented dm_report_set_output_selection fn
The dm_report_set_output_selection was not implemented in the end -
we have dm_report_init_with_selection instead. This is just a remnant
from development code that got into libdevmapper.h by mistake.
2014-12-05 11:54:25 +01:00
Zdenek Kabelac
8bc7b4f926 libdm: there is no element in item 0
Items[0] holds only counter in .len.
So don't zero already zeroed items[0].len assigned above.

(finishing fc935495c8)
2014-11-24 14:38:54 +01:00
Peter Rajnoha
fb220314ec report: add some comments about how string list is stored internally 2014-11-24 13:14:33 +01:00
Peter Rajnoha
b9601b8353 report: add some comments about how string list is stored internally 2014-11-24 10:48:01 +01:00
Zdenek Kabelac
ccdea661fa cleanup: warning: declaration of 'remove' shadows a global declaration
Don't shadow stdio.h declaration.
2014-11-23 00:49:04 +01:00
Zdenek Kabelac
678cc4e375 libdm: report fix memleak on error path
When _alloc_selection_node() fails, rh should be destroyed.

Use 'bad:' label since we have goto_bad with stack embeded.
2014-11-23 00:49:04 +01:00
Zdenek Kabelac
fc935495c8 libdm: fix reporting of empty string list
Don't write behind the allocated array when list is empty.
Use index 0 for the allocated element.

Error triggered by i.e.:  lvs -a -o all,lv_modules
2014-11-22 18:50:53 +01:00
Zdenek Kabelac
089bdc0be4 cleanup: drop unused vars 2014-11-20 17:49:32 +01:00
Petr Rockai
0050480c0e libdm-config: Warn about duplicate keys. 2014-11-20 16:51:06 +01:00
Petr Rockai
de2c5ab2ac libdm-config: Implement dm_config_remove_node. 2014-11-20 16:51:06 +01:00
Petr Rockai
8bc9966763 libdm-config: Interpret barewords after '=' as strings. 2014-11-20 16:51:06 +01:00
Petr Rockai
274a7a68b8 libdm-config: Implement dm_config_flatten. 2014-11-20 16:51:06 +01:00
Petr Rockai
956c192841 libdm-config: Re-link config trees to reflect file order of keys/sections. 2014-11-20 16:51:06 +01:00
Petr Rockai
687029cbbd libdm-config: Allow paths (section/key = value) in config files.
The order of the resulting tree is based on the first appearance of
sections. With no section repeats, the sections stay as listed in the
config file. Sections using the brace syntax 'section { key = value }' are
treated the same way: 'section { x = 1 } section { y = 2 }' is the same as
'section/x = 1 section/y = 2' is the same as 'section { x = 1 y = 2 }'
2014-11-20 16:47:30 +01:00
Alasdair G Kergon
9a5910bdf9 pre-release 2014-11-11 14:13:00 +00:00