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

168 Commits

Author SHA1 Message Date
David Teigland
76bd678ce3 devices: isolate fake device setup for config file
Freeing dev-cache from dumb fake device usage.
2024-07-02 16:18:09 -05:00
Zdenek Kabelac
85e0636623 cleanup: some more MAJOR MINOR unsigned
Finish remaing missing cases where MAJOR/MINOR was casted to (int)
while we can now use it directly as 'unsigned'.
2024-06-03 15:30:05 +02:00
Zdenek Kabelac
c451d7925d cleanup: correcting some log_print
Correcting to log_print_unless_silent(),
so -qq can do some work.
2023-05-06 20:38:49 +02:00
David Teigland
7e70041e32 devices: drop incorrect paths from aliases list
along with some basic checks for cases when a device
has no aliases.

lvm itself creates many situations where a struct device
has no valid paths, when it activates and opens an LV,
does something with it, e.g. zeroing, and then closes
and deactivates it.  (dev-cache is intended for PVs, and
the use of LVs should be moved out of dev-cache in a
future patch.)
2022-02-24 17:22:03 -06:00
Zdenek Kabelac
df70dca995 make: opened can be needed for either defines case
Reported-by: adamboardman of gemian
2021-09-27 18:56:14 +02:00
Zdenek Kabelac
bee9b5c1d8 cov: mask uninitialized value
Coverity doesn't track ioctl() too well, so let's just make it quiet.
2021-03-10 01:34:27 +01:00
Zdenek Kabelac
d422aa7924 dev-type: convert to use log_warn
Keep log_error designated only for 'erroring' condition of command
and replace these errors with log_warn() WARNING.

Also do some indent changes.
2021-02-08 23:43:38 +01:00
Zdenek Kabelac
dd8212365d debug: update messages 2020-10-02 21:04:16 +02:00
David Teigland
2f29765e7f devs: add some checks for a dev with no path name
It's possible for a dev-cache entry to remain after all
paths for it have been removed, and other parts of the
code expect that a dev always has a name.  A better fix
may be to remove a device from dev-cache after all paths
to it have been removed.
2020-05-13 16:26:26 -05:00
David Teigland
957904933b reduce device path error messsages
When /dev entries or sysfs entries are changing
due to concurrent lvm commands, it can cause
warning/error messages about missing paths.
2020-03-12 10:18:51 -05:00
David Teigland
eb6aa5fefe devices: put ifdef around BLKPBSZGET
BLKPBSZGET is not defined before kernel version 2.6.32
(e.g. rhel5)
2019-08-08 15:45:03 -05:00
David Teigland
09bc2d0fd1 devices: clean up block size functions
Replace calls to the old dev_get_block_size function
with calls to the new dev_get_direct_block_size function,
and remove the old function.
2019-08-07 11:48:10 -05:00
David Teigland
7f347698e3 Fix rounding writes up to sector size
Do this at two levels, although one would be enough to
fix the problem seen recently:

- Ignore any reported sector size other than 512 of 4096.
  If either sector size (physical or logical) is reported
  as 512, then use 512.  If neither are reported as 512,
  and one or the other is reported as 4096, then use 4096.
  If neither is reported as either 512 or 4096, then use 512.

- When rounding up a limited write in bcache to be a multiple
  of the sector size, check that the resulting write size is
  not larger than the bcache block itself.  (This shouldn't
  happen if the sector size is 512 or 4096.)
2019-07-26 14:21:08 -05:00
David Teigland
3ed9256985 remove unused io functions 2019-02-28 10:58:00 -06:00
David Teigland
1dc5603f73 devices: reuse bcache fd when getting block size
This avoids an unnecessary open() on the device.
2018-11-06 16:36:18 -06:00
Zdenek Kabelac
fdd76da33d cov: drop uneeded header files 2018-10-15 17:49:44 +02:00
Zdenek Kabelac
b1ff52ca14 cov: check dev_close_immediate
Function can report log_error() on fail path.
2018-10-15 17:49:44 +02:00
Zdenek Kabelac
c8b4f9414c dev_io: no discard in testmode
When lvm2 command is executed in test mode, discard ioctl is skipped.
This may cause even data-loose in case, issuing discard for released
areas was enabled and user 'tested'  lvreduce.
2018-07-09 00:19:30 +02:00
David Teigland
328303d4d4 Remove unused device error counting 2018-06-15 14:04:39 -05:00
Joe Thornber
d5da55ed85 device_mapper: remove dbg_malloc.
I wrote dbg_malloc before we had valgrind.  These days there's just
no need.
2018-06-08 13:40:53 +01:00
Joe Thornber
89fdc0b588 Merge branch 'master' into 2018-05-11-fork-libdm 2018-05-16 13:43:02 +01:00
Joe Thornber
7f97c7ea9a build: Don't generate symlinks in include/ dir
As we start refactoring the code to break dependencies (see doc/refactoring.txt),
I want us to use full paths in the includes (eg, #include "base/data-struct/list.h").
This makes it more obvious when we're breaking abstraction boundaries, eg, including a file in
metadata/ from base/
2018-05-14 10:30:20 +01:00
David Teigland
73578e36fa dev_cache: remove the lvmcache check when closing fd
This is no longer used since devices are not held
open in dev_cache.
2018-05-11 14:30:10 -05:00
David Teigland
3e3cb22f2a dev_cache: fix close in utility functions
All these functions are now used as utilities,
e.g. for ioctl (not for io), and need to
open/close the device each time they are called.
(Many of the opens can probably be eliminated by
just using the bcache fd for the ioctl.)
2018-05-11 14:25:08 -05:00
David Teigland
ccab54677c dev_cache: fix close in dev_get_block_size 2018-05-11 13:53:19 -05:00
David Teigland
bbb8040456 dev_cache: drop open_list
devices are now held open only in bcache,
so drop the dev_cache list of open devices
which is unused.
2018-05-11 12:47:56 -05:00
David Teigland
57bb46c5e7 filter: use bcache for filter reads
Filters are still applied before any device reading or
the label scan, but any filter checks that want to read
the device are skipped and the device is flagged.

After bcache is populated, but before lvm looks for
devices (i.e. before label scan), the filters are
reapplied to the devices that were flagged above.
The filters will then find the data they need in
bcache.
2018-05-10 16:03:19 -05:00
David Teigland
6c67c7557c scan: use separate fd for bcache
Create a new dev->bcache_fd that the scanning code owns
and is in charge of opening/closing.  This prevents other
parts of lvm code (which do various open/close) from
interfering with the bcache fd.  A number of dev_open
and dev_close are removed from the reading path since
the read path now uses the bcache.

With that in place, open(O_EXCL) for pvcreate/pvremove
can then be fixed.  That wouldn't work previously because
of other open fds.
2018-04-20 11:22:46 -05:00
Joe Thornber
00f1b208a1 [io paths] Unpick agk's aio stuff 2018-04-20 11:03:58 -05:00
Zdenek Kabelac
e878c3fc32 cleanup: correct casting 2018-04-20 12:17:01 +02:00
Zdenek Kabelac
f2d0eefa77 coverity: make use of defined variable
Since we declare 'r', let's use the value for something.
2018-03-17 23:33:58 +01:00
Zdenek Kabelac
285413b502 cleanup: missing dots and indent 2018-03-15 11:01:04 +01:00
Zdenek Kabelac
70ad633638 devcache: add reason and always log_error
With these read errors it's useful to know the reason.
Also avoid to log error just once so we know exactly
how many times we did failing read.

On the other hand reduce repeated log_error() on code 'backtrace'
path and change severity of message to just log_debug() so the
actual read error is printed once for one read.
2018-03-15 10:50:28 +01:00
Zdenek Kabelac
6b48868cf0 io: keep 64b arithmetic
Widen to 64b arithmetic from start.
2018-02-28 21:05:18 +01:00
Alasdair G Kergon
d6cabbbc53 device: Fix basic async I/O error handling 2018-02-08 20:19:21 +00:00
Alasdair G Kergon
3e29c80122 device: Queue any aio beyond defined limits. 2018-02-08 20:15:37 +00:00
Alasdair G Kergon
db41fe6c5d lvmcache: Use asynchronous I/O when scanning devices. 2018-02-08 20:15:29 +00:00
Alasdair G Kergon
8c7bbcfb0f device: Basic config and setup to support async I/O. 2018-02-08 20:15:14 +00:00
Alasdair G Kergon
e869a52cc4 callbacks: Miscellaneous fixes for recent changes 2018-02-06 01:09:39 +00:00
Zdenek Kabelac
a1cfef9f26 dev_io: fix writes for unaligned buffers
Actually the removed code is necessary - since not all writes are
getting alligned buffer - older compilers seems to be not able
to create 4K aligned buffers on stack - this the aligning code still
need to be present for write path.
2018-01-23 13:36:12 +01:00
Zdenek Kabelac
6e9148e7ab debug: drop DEBUG_MEM path
Memory is not allocated so no DEBUG_MEM part is needed.
2018-01-23 11:45:18 +01:00
Alasdair G Kergon
9194610f42 device: Add ioflags parameter to transfer additional state.
Flags are set on the initial I/O and passed to any callbacks that
may in turn issue further I/O using the inherited flags.
2018-01-21 21:10:23 +00:00
Alasdair G Kergon
c26458339e device: Move buffer allocation nearer to the I/O.
Don't allocate memory until it's needed - later we'll add
some of the I/O to an internal queue instead of issuing it
immediately.
2018-01-16 01:12:08 +00:00
Alasdair G Kergon
081902b4c1 device: Merge _dev_read and dev_read_callback. 2018-01-16 00:41:42 +00:00
Alasdair G Kergon
b825987b2f device: Rearrange _aligned_io(). 2018-01-15 20:10:54 +00:00
Alasdair G Kergon
c90582344d device: Add reason to devbuf. 2018-01-15 19:38:18 +00:00
Alasdair G Kergon
1f01eaa612 device: Store offset to data instead of pointer.
We want to save the relative offset before we've allocated the
buffer's memory.
2018-01-15 19:32:59 +00:00
Alasdair G Kergon
6210c1ec28 device: Mark read-only device buffers const. 2018-01-10 19:57:10 +00:00
Alasdair G Kergon
c350f96c09 device: Eliminate unnecessary buffer from dev_read. 2018-01-10 18:48:01 +00:00
Alasdair G Kergon
366493a1d1 device: Suppress repeated reads of the same data.
If the data being requested is present in last_[extra_]devbuf,
return that directly instead of reading it from disk again.

Typical LVM2 access patterns request data within two adjacent 4k blocks
so we eliminate some read() system calls by always reading at least 8k.
2018-01-10 15:52:03 +00:00