Before this patch, various errors and messages were reported using
the pr_* functions: pr_err, pr_warn, pr_info, etc., but that does
not tell you which gfs2 mount had the problem, which is often vital
to debugging. This patch changes the calls from pr_* to fs_* in
most of the messages so that the file system id is printed along
with the message.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
The GFS2_RDF_UPTODATE flag in the rgrp is used to determine when
a rgrp buffer is valid. It's cleared when the glock is invalidated,
signifying that the buffer data is now invalid. But before this
patch, function update_rgrp_lvb was setting the flag when it
determined it had a valid lvb. But that's an invalid assumption:
just because you have a valid lvb doesn't mean you have valid
buffers. After all, another node may have made the lvb valid,
and this node just fetched it from the glock via dlm.
Consider this scenario:
1. The file system is mounted with RGRPLVB option.
2. In gfs2_inplace_reserve it locks the rgrp glock EX, but thanks
to GL_SKIP, it skips the gfs2_rgrp_bh_get.
3. Since loops == 0 and the allocation target (ap->target) is
bigger than the largest known chunk of blocks in the rgrp
(rs->rs_rbm.rgd->rd_extfail_pt) it skips that rgrp and bypasses
the call to gfs2_rgrp_bh_get there as well.
4. update_rgrp_lvb sees the lvb MAGIC number is valid, so bypasses
gfs2_rgrp_bh_get, but it still sets sets GFS2_RDF_UPTODATE due
to this invalid assumption.
5. The next time update_rgrp_lvb is called, it sees the bit is set
and just returns 0, assuming both the lvb and rgrp are both
uptodate. But since this is a smaller allocation, or space has
been freed by another node, thus adjusting the lvb values,
it decides to use the rgrp for allocations, with invalid rd_free
due to the fact it was never updated.
This patch changes update_rgrp_lvb so it doesn't set the UPTODATE
flag anymore. That way, it has no choice but to fetch the latest
values.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Before this patch, gfs2_rgrp_bh_get would check for lvb mismatches,
but it wouldn't tell you what was actually wrong. This patch adds
more information to help us debug it. It also makes rgrp consistency
checks dump any bad rgrps, and the rgrp dump code dump any lvbs
as well as the rgrp itself.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
Function update_rgrp_lvb_unlinked used to do the same thing as
be32_add_cpu. This patch removes it in favor of using be32_add_cpu
directly.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andrew Price <anprice@redhat.com>
Function gfs2_testbit is called in three places. Two of those places,
gfs2_alloc_extent and gfs2_unaligned_extlen, should be using the clone
bitmaps, not the "real" bitmaps. Function gfs2_unaligned_extlen is used
by the block reservations scheme to determine the length of an extent of
free blocks. Before this patch, it wasn't using the clone bitmap, which
means recently-freed blocks were treated as free blocks for the purposes
of an allocation.
This patch adds a new parameter to gfs2_testbit to indicate whether or
not the clone bitmaps should be used (if available).
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch gfs2_rgrp_ondisk2lvb was called after every call
to gfs2_rgrp_out. This patch just calls it directly from within
gfs2_rgrp_out, and moves the function to be before it so we don't
need a function prototype.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch, several functions in rgrp.c checked the value of
rgd->rd_free_clone. That does not take into account blocks that were
reserved by a multi-block reservation. This causes a problem when
space gets tight in the file system. For example, when function
gfs2_inplace_reserve checks to see if a rgrp has enough blocks to
satisfy the request, it can accept a rgrp that it should reject
because, although there are enough blocks to satisfy the request
_now_, those blocks may be reserved for another running process.
A second problem with this occurs when we've reserved the remaining
blocks in an rgrp: function rg_mblk_search() can reject an rgrp
improperly because it calculates:
u32 free_blocks = rgd->rd_free_clone - rgd->rd_reserved;
But rd_reserved includes blocks that the current process just
reserved in its own call to inplace_reserve. For example, it can
reserve the last 128 blocks of an rgrp, then reject that same rgrp
because the above calculates out to free_blocks = 0;
Consequences include, but are not limited to, (1) leaving holes,
and thus increasing file system fragmentation, and (2) reporting
file system is full long before it actually is.
This patch introduces a new function, rgd_free, which returns the
number of clone-free blocks (blocks that are truly free as opposed
to blocks that are still being used because an unlinked file is
still open) minus the number of blocks reserved by processes, but
not counting the blocks we ourselves reserved (because obviously
we need to allocate them).
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Pull in the gfs2 iomap-write changes: Tweak the existing code to
properly support iomap write and eliminate an unnecessary special case
in gfs2_block_map. Implement iomap write support for buffered and
direct I/O. Simplify some of the existing code and eliminate code that
is no longer used:
gfs2: Remove gfs2_write_{begin,end}
gfs2: iomap direct I/O support
gfs2: gfs2_extent_length cleanup
gfs2: iomap buffered write support
gfs2: Further iomap cleanups
This is based on the following changes on the xfs 'iomap-4.19-merge'
branch:
iomap: add private pointer to struct iomap
iomap: add a page_done callback
iomap: generic inline data handling
iomap: complete partial direct I/O writes synchronously
iomap: mark newly allocated buffer heads as new
fs: factor out a __generic_write_end helper
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Before this patch, you could get into situations like this:
1. Process 1 searches for X free blocks, finds them, makes a reservation
2. Process 2 searches for free blocks in the same rgrp, but now the
bitmap is full because process 1's reservation is skipped over.
So it marks the bitmap as GBF_FULL.
3. Process 1 tries to allocate blocks from its own reservation, but
since the GBF_FULL bit is set, it skips over the rgrp and searches
elsewhere, thus not using its own reservation.
This patch adds an additional check to allow processes to use their
own reservations.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
GFS2 remembers the last rgrp used for allocations in ip->i_rgd.
However, block allocations are made by way of a reservations structure,
ip->i_res, which keeps the last rgrp in ip->i_res.rs_rgd, and ip->i_res
is kept in sync with ip->i_res.rs_rgd, so it's redundant. Get rid of
ip->i_rgd and just use ip->i_res.rs_rgd in its place.
Based on patches by Robert Peterson.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
In the resource group list code, keep the last resource group added in
the last position in the array. Check against that instead of messing
with ip->i_rgd.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Before this patch, block reservations kept track of the inode
number. At one point, that was a valid thing to do. However, since
we made the reservation a part of the inode (rather than a pointer
to a separate allocated object) the reservation can determine the
inode number by using container_of. This saves us a little memory
in our inode.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Function gfs2_free_extlen calculates the length of an extent of
free blocks that may be reserved. The end pointer was calculated as
end = start + bh->b_size but b_size is incorrect because the
bitmap usually stops prior to the end of the buffer data on
the last bitmap.
What this means is that when you do a write, you can reserve a
chunk of blocks that runs off the end of the last bitmap. For
example, I've got a file system where there is only one bitmap
for each rgrp, so ri_length==1. I saw cases in which iozone
tried to do a big write, grabbed a large block reservation,
chose rgrp 5464152, which has ri_data0 5464153 and ri_data 8188.
So 5464153 + 8188 = 5472341 which is the end of the rgrp.
When it grabbed a reservation it got back: 5470936, length 7229.
But 5470936 + 7229 = 5478165. So the reservation starts inside
the rgrp but runs 5824 blocks past the end of the bitmap.
This patch fixes the calculation so it won't exceed the last
bitmap. It also adds a BUG_ON to guard against overflows in the
future.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Some of the info, warning, and error messages are missing their trailing
newline.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch just adds the capability for GFS2 to track which function
called gfs2_log_flush. This should make it easier to diagnose
problems based on the sequence of events found in the journals.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
This patch adds a new structure called gfs2_log_header_v2 which is used
to store expanded fields into previously unused areas of the log headers
(i.e., this change is backwards compatible). Some of these are used for
debug purposes so we can backtrack when problems occur. Others are
reserved for future expansion.
This patch is based on a prototype from Steve Whitehouse.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Document when to use gfs2_blk2rgrpd for "inexact" resource group
matching. Based on that, fix an incorrect use of gfs2_blk2rgrpd in
sweep_bh_for_rgrps.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Add the rg_crc field to store a crc32 of the gfs2_rgrp structure. This
allows us to check resource group headers' integrity and removes the
requirement to check them against the rindex entries in fsck. If this
field is found to be zero, it should be ignored (or updated with an
accurate value).
Signed-off-by: Andrew Price <anprice@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Add rg_data0, rg_data and rg_bitbytes to struct gfs2_rgrp. The fields
are identical to their counterparts in struct gfs2_rindex and are
intended to reduce the use of the rindex. For now the fields are only
written back as the in-memory equivalents in struct gfs2_rgrpd are set
using values from the rindex. However, they are needed at this point so
that userspace can make use of them, allowing a migration away from the
rindex over time.
The new fields take up previously reserved space which was explicitly
zeroed on write so, in clusters with mixed kernels, these fields could
get zeroed after being set and this should not be treated as an error.
Signed-off-by: Andrew Price <anprice@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Add a new rg_skip field to struct gfs2_rgrp, replacing __pad. The
rg_skip field has the following meaning:
- If rg_skip is zero, it is considered unset and not useful.
- If rg_skip is non-zero, its value will be the number of blocks between
this rgrp's address and the next rgrp's address. This can be used as a
hint by fsck.gfs2 when rebuilding a bad rindex, for example.
This will provide less dependency on the rindex in future, and allow
tools such as fsck.gfs2 to iterate the resource groups without keeping
the rindex around.
The field is updated in gfs2_rgrp_out() so that existing file systems
will have it set. This means that any resource groups that aren't ever
written will not be updated. The final rgrp is a special case as there
is no next rgrp, so it will always have a rg_skip of 0 (unless the fs is
extended).
Before this patch, gfs2_rgrp_out() zeroes the __pad field explicitly, so
the rg_skip field can get set back to 0 in cases where nodes with and
without this patch are mixed in a cluster. In some cases, the field may
bounce between being set by one node and then zeroed by another which
may harm performance slightly, e.g. when two nodes create many small
files. In testing this situation is rare but it becomes more likely as
the filesystem fills up and there are fewer resource groups to choose
from. The problem goes away when all nodes are running with this patch.
Dipping into the space currently occupied by the rg_reserved field would
have resulted in the same problem as it is also explicitly zeroed, so
unfortunately there is no other way around it.
Signed-off-by: Andrew Price <anprice@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Before this patch, function gfs2_free_di was 4 lines of code, and
one of those lines was to call gfs2_free_uninit_di. Although
unlikely, if function gfs2_free_uninit_di encountered an error
finding the block to be freed, the error was silently ignored by the
caller, which went ahead and improperly did a quota-change operation
and meta_wipe despite the error. This patch combines the two
functions into one to make the code more readable and fixes the bug
by returning from the combined function before it takes those next
incorrect steps.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
The following cleanup is needed to avoid spilling the syslog with
false warnings.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch removes a call to gfs2_glock_add_to_lru from function
gfs2_clear_rgrpd. The call is just a waste of time because as soon
as it adds it to the lru_list, the call to gfs2_glock_put takes it
back off again.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Put all remaining accesses to gl->gl_object under the
gl->gl_lockref.lock spinlock to prevent races.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Implement truncate/delete as a non-recursive algorithm. The older
algorithm was implemented with recursion to strip off each layer
at a time (going by height, starting with the maximum height.
This version tries to do the same thing but without recursion,
and without needing to allocate new structures or lists in memory.
For example, say you want to truncate a very large file to 1 byte,
and its end-of-file metapath is: 0.505.463.428. The starting
metapath would be 0.0.0.0. Since it's a truncate to non-zero, it
needs to preserve that byte, and all metadata pointing to it.
So it would start at 0.0.0.0, look up all its metadata buffers,
then free all data blocks pointed to at the highest level.
After that buffer is "swept", it moves on to 0.0.0.1, then
0.0.0.2, etc., reading in buffers and sweeping them clean.
When it gets to the end of the 0.0.0 metadata buffer (for 4K
blocks the last valid one is 0.0.0.508), it backs up to the
previous height and starts working on 0.0.1.0, then 0.0.1.1,
and so forth. After it reaches the end and sweeps 0.0.1.508,
it continues with 0.0.2.0, and so on. When that height is
exhausted, and it reaches 0.0.508.508 it backs up another level,
to 0.1.0.0, then 0.1.0.1, through 0.1.0.508. So it has to keep
marching backwards and forwards through the metadata until it's
all swept clean. Once it has all the data blocks freed, it
lowers the strip height, and begins the process all over again,
but with one less height. This time it sweeps 0.0.0 through
0.505.463. When that's clean, it lowers the strip height again
and works to free 0.505. Eventually it strips the lowest height, 0.
For a delete or truncate to 0, all metadata for all heights of
0.0.0.0 would be freed. For a truncate to 1 byte, 0.0.0.0 would
be preserved.
This isn't much different from normal integer incrementing,
where an integer gets incremented from 0000 (0.0.0.0) to 3021
(3.0.2.1). So 0000 gets increments to 0001, 0002, up to 0009,
then on to 0010, 0011 up to 0099, then 0100 and so forth. It's
just that each "digit" goes from 0 to 508 (for a total of 509
pointers) rather than from 0 to 9.
Note that the dinode will only have 483 pointers due to the
dinode structure itself.
Also note: this is just an example. These numbers (509 and 483)
are based on a standard 4K block size. Smaller block sizes will
yield smaller numbers of indirect pointers accordingly.
The truncation process is accomplished with the help of two
major functions and a few helper functions.
Functions do_strip and recursive_scan are obsolete, so removed.
New function sweep_bh_for_rgrps cleans a buffer_head pointed to
by the given metapath and height. By cleaning, I mean it frees
all blocks starting at the offset passed in metapath. It starts
at the first block in the buffer pointed to by the metapath and
identifies its resource group (rgrp). From there it frees all
subsequent block pointers that lie within that rgrp. If it's
already inside a transaction, it stays within it as long as it
can. In other words, it doesn't close a transaction until it knows
it's freed what it can from the resource group. In this way,
multiple buffers may be cleaned in a single transaction, as long
as those blocks in the buffer all lie within the same rgrp.
If it's not in a transaction, it starts one. If the buffer_head
has references to blocks within multiple rgrps, it frees all the
blocks inside the first rgrp it finds, then closes the
transaction. Then it repeats the cycle: identifies the next
unfreed block, uses it to find its rgrp, then starts a new
transaction for that set. It repeats this process repeatedly
until the buffer_head contains no more references to any blocks
past the given metapath.
Function trunc_dealloc has been reworked into a finite state
automaton. It has basically 3 active states:
DEALLOC_MP_FULL, DEALLOC_MP_LOWER, and DEALLOC_FILL_MP:
The DEALLOC_MP_FULL state implies the metapath has a full set
of buffers out to the "shrink height", and therefore, it can
call function sweep_bh_for_rgrps to free the blocks within the
highest height of the metapath. If it's just swept the lowest
level (or an error has occurred) the state machine is ended.
Otherwise it proceeds to the DEALLOC_MP_LOWER state.
The DEALLOC_MP_LOWER state implies we are finished with a given
buffer_head, which may now be released, and therefore we are
then missing some buffer information from the metapath. So we
need to find more buffers to read in. In most cases, this is
just a matter of releasing the buffer_head and moving to the
next pointer from the previous height, so it may be read in and
swept as well. If it can't find another non-null pointer to
process, it checks whether it's reached the end of a height
and needs to lower the strip height, or whether it still needs
move forward through the previous height's metadata. In this
state, all zero-pointers are skipped. From this state, it can
only loop around (once more backing up another height) or,
once a valid metapath is found (one that has non-zero
pointers), proceed to state DEALLOC_FILL_MP.
The DEALLOC_FILL_MP state implies that we have a metapath
but not all its buffers are read in. So we must proceed to read
in buffer_heads until the metapath has a valid buffer for every
height. If the previous state backed us up 3 heights, we may
need to read in a buffer, increment the height, then repeat the
process until buffers have been read in for all required heights.
If it's successful reading a buffer, and it's at the highest
height we need, it proceeds back to the DEALLOC_MP_FULL state.
If it's unable to fill in a buffer, (encounters a hole, etc.)
it tries to find another non-zero block pointer. If they're all
zero, it lowers the height and returns to the DEALLOC_MP_LOWER
state. If it finds a good non-null pointer, it loops around and
reads it in, while keeping the metapath in lock-step with the
pointers it examines.
The state machine runs until the truncation request is
satisfied. Then any transactions are ended, the quota and
statfs data are updated, and the function is complete.
Helper function metaptr1 was introduced to be an easy way to
determine the start of a buffer_head's indirect pointers.
Helper function lookup_mp_height was introduced to find a
metapath index and read in the buffer that corresponds to it.
In this way, function lookup_metapath becomes a simple loop to
call it for every height.
Helper function fillup_metapath is similar to lookup_metapath
except it can do partial lookups. If the state machine
backed up multiple levels (like 2999 wrapping to 3000) it
needs to find out the next starting point and start issuing
metadata reads at that point.
Helper function hptrs is a shortcut to determine how many
pointers should be expected in a buffer. Height 0 is the dinode
which has fewer pointers than the others.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
For the last process to close a file opened for write, function
gfs2_rsqa_delete was deleting the file's inode's block reservation
out of the rgrp reservations tree. Then it was checking to make sure
rs_free was 0, but it was performing the check outside the protection
of rd_rsspin spin_lock. The rd_rsspin spin_lock protection is needed
to prevent a race between the process freeing the reservation and
another who is allocating a new set of blocks inside the same rgrp
for the same inode, thus changing its value.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Make the code more readable by cleaning up the different ways of
initializing lock holders and checking for initialized lock holders:
mark lock holders as uninitialized by setting the holder's glock to NULL
(gfs2_holder_mark_uninitialized) instead of zeroing out the entire
object or using a separate flag. Recognize initialized holders by their
non-NULL glock (gfs2_holder_initialized). Don't zero out holder objects
which are immeditiately initialized via gfs2_holder_init or
gfs2_glock_nq_init.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Before this patch, function read_rindex_entry would set a rgrp
glock's gl_object pointer to itself before inserting the rgrp into
the rgrp rbtree. The problem is: if another process was also reading
the rgrp in, and had already inserted its newly created rgrp, then
the second call to read_rindex_entry would overwrite that value,
then return a bad return code to the caller. Later, other functions
would reference the now-freed rgrp memory by way of gl_object.
In some cases, that could result in gfs2_rgrp_brelse being called
twice for the same rgrp: once for the failed attempt and once for
the "real" rgrp release. Eventually the kernel would panic.
There are also a number of other things that could go wrong when
a kernel module is accessing freed storage. For example, this could
result in rgrp corruption because the fake rgrp would point to a
fake bitmap in memory too, causing gfs2_inplace_reserve to search
some random memory for free blocks, and find some, since we were
never setting rgd->rd_bits to NULL before freeing it.
This patch fixes the problem by not setting gl_object until we
have successfully inserted the rgrp into the rbtree. Also, it sets
rd_bits to NULL as it frees them, which will ensure any accidental
access to the wrong rgrp will result in a kernel panic rather than
file system corruption, which is preferred.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
- Abhi Das has two patches that fix a GFS2 splice issue (and an adjustment).
- Ben Marzinski has a patch which allows the proper unmount of a GFS2
file system after hitting a withdraw error.
- I have a patch to fix a problem where GFS2 would dereference an error
value, plus three cosmetic / refactoring patches.
- Daniel DeFreez has a patch to fix two glock reference count problems,
where GFS2 was not properly "uninitializing" its glock holder on error
paths.
- Denys Vlasenko has a patch to change a function to not be inlined,
thus reducing the memory footprint of the GFS2 module.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJXP2KLAAoJENeLYdPf93o7nScH/jlDCv+JzwBwAiFEVelgB657
lFY+Q3bPvmcTaHiOqXzco7U1EeyvAuZ+x9duaFBBEGC2/NydIbkAdCP1XjDOTAGZ
aVVAP8qov4adFiXTfBdzMYjJ7DArOy/JRnlC8WK5OiD/3yxsXJZNlAHTgjjL/ZMq
PkhjxX2Lc9ZHGF9YIFTyeoBlqf1qO6WbhkxvSNefgq3c702QPM+T9XND5T1duGDF
ThEcXS2DRRoE+6wr3h+ruvS5kcfxXRLLausAgqNpj2FYQOFCQ6XXgGK8pyXA5V/t
3Vh8mHW0VmSjAfrRWslCRjd+APXoOf+7QHqOHUAA1dvQ6CGfC9qTETH9Wp32JCI=
=8g+a
-----END PGP SIGNATURE-----
Merge tag 'gfs2-4.7.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2
Pull GFS2 updates from Bob Peterson:
"We've got nine patches this time:
- Abhi Das has two patches that fix a GFS2 splice issue (and an
adjustment).
- Ben Marzinski has a patch which allows the proper unmount of a GFS2
file system after hitting a withdraw error.
- I have a patch to fix a problem where GFS2 would dereference an
error value, plus three cosmetic / refactoring patches.
- Daniel DeFreez has a patch to fix two glock reference count
problems, where GFS2 was not properly "uninitializing" its glock
holder on error paths.
- Denys Vlasenko has a patch to change a function to not be inlined,
thus reducing the memory footprint of the GFS2 module"
* tag 'gfs2-4.7.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
GFS2: Refactor gfs2_remove_from_journal
GFS2: Remove allocation parms from gfs2_rbm_find
gfs2: use inode_lock/unlock instead of accessing i_mutex directly
GFS2: Add calls to gfs2_holder_uninit in two error handlers
GFS2: Don't dereference inode in gfs2_inode_lookup until it's valid
GFS2: fs/gfs2/glock.c: Deinline do_error, save 1856 bytes
gfs2: Use gfs2 wrapper to sync inode before calling generic_file_splice_read()
GFS2: Get rid of dead code in inode_go_demote_ok
GFS2: ignore unlock failures after withdraw
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Before this patch, when function try_rgrp_unlink queued a glock for
delete_work to reclaim the space, it used the inode glock to do so.
That's different from the iopen callback which uses the iopen glock
for the same purpose. We should be consistent and always use the
iopen glock. This may also save us reference counting problems with
the inode glock, since clear_glock does an extra glock_put() for the
inode glock.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Before this patch, multi-block reservation structures were allocated
from a special slab. This patch folds the structure into the gfs2_inode
structure. The disadvantage is that the gfs2_inode needs more memory,
even when a file is opened read-only. The advantages are: (a) we don't
need the special slab and the extra time it takes to allocate and
deallocate from it. (b) we no longer need to worry that the structure
exists for things like quota management. (c) This also allows us to
remove the calls to get_write_access and put_write_access since we
know the structure will exist.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch basically reverts the majority of patch 5407e24.
That patch eliminated the gfs2_qadata structure in favor of just
using the reservations structure. The problem with doing that is that
it increases the size of the reservations structure. That is not an
issue until it comes time to fold the reservations structure into the
inode in memory so we know it's always there. By separating out the
quota structure again, we aren't punishing the non-quota users by
making all the inodes bigger, requiring more slab space. This patch
creates a new slab area to allocate the quota stuff so it's managed
a little more sanely.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
When gfs2 allocates an inode and its extended attribute block next to
each other at inode create time, the inode's directory entry indicates
that in de_rahead. In that case, we can readahead the extended
attribute block when we read in the inode.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch fixes a bug introduced by commit 7005c3e. That patch
tries to map a vm range for resource groups, but the calculation
breaks down when the block size is less than the page size.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Commit e66cf161 replaced the gl_spin spinlock in struct gfs2_glock with a
gl_lockref lockref and defined gl_spin as gl_lockref.lock (the spinlock in
gl_lockref). Remove that define to make the references to gl_lockref.lock more
obvious.
Signed-off-by: Andreas Gruenbacher <andreas.gruenbacher@gmail.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
None of these statistics can meaningfully be negative, and the
numerator for do_div() must have the type u64. The generic
implementation of do_div() used on some 32-bit architectures asserts
that, resulting in a compiler error in gfs2_rgrp_congested().
Fixes: 0166b197c2 ("GFS2: Average in only non-zero round-trip times ...")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Andreas Gruenbacher <agruenba@redhat.com>
What uniquely identifies a glock in the glock hash table is not
gl_name, but gl_name and its superblock pointer. This patch makes
the gl_name field correspond to a unique glock identifier. That will
allow us to simplify hashing with a future patch, since the hash
algorithm can then take the gl_name and hash its components in one
operation.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
This patch allows the block allocation code to retain the buffers
for the resource groups so they don't need to be re-read from buffer
cache with every request. This is a performance improvement that's
especially noticeable when resource groups are very large. For
example, with 2GB resource groups and 4K blocks, there can be 33
blocks for every resource group. This patch allows those 33 buffers
to be kept around and not read in and thrown away with every
operation. The buffers are released when the resource group is
either synced or invalidated.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Steven Whitehouse <swhiteho@redhat.com>
Reviewed-by: Benjamin Marzinski <bmarzins@redhat.com>
bi was already declared and initialized globally in gfs2_rbm_find()
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
The function set_rgrp_preferences() does not handle the (rarely
returned) NULL value from gfs2_rgrpd_get_next() and this patch
fixes that.
The fs image in question is only 150MB in size which allows for
only 1 rgrp to be created. The in-memory rb tree has only 1 node
and when gfs2_rgrpd_get_next() is called on this sole rgrp, it
returns NULL. (Default behavior is to wrap around the rb tree and
return the first node to give the illusion of a circular linked
list. In the case of only 1 rgrp, we can't have
gfs2_rgrpd_get_next() return the same rgrp (first, last, next all
point to the same rgrp)... that would cause unintended consequences
and infinite loops.)
Signed-off-by: Abhi Das <adas@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch changes function gfs2_rgrp_congested so that it only factors
in non-zero values into its average round trip time. If the round-trip
time is zero for a particular cpu, that cpu has obviously never dealt
with bouncing the resource group in question, so factoring in a zero
value will only skew the numbers. It also fixes a compile error on
some arches related to division.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
This patch changes function gfs2_rgrp_congested so that it uses an
average srttb (smoothed round trip time for blocking rgrp glocks)
rather than the CPU-specific value. If we use the CPU-specific value
it can incorrectly report no contention when there really is contention
due to the glock processing occurring on a different CPU.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
struct gfs2_alloc_parms is passed to gfs2_quota_check() and
gfs2_inplace_reserve() with ap->target containing the number of
blocks being requested for allocation in the current operation.
We add a new field to struct gfs2_alloc_parms called 'allowed'.
gfs2_quota_check() and gfs2_inplace_reserve() return the max
blocks allowed by quota and the max blocks allowed by the chosen
rgrp respectively in 'allowed'.
A new field 'min_target', when non-zero, tells gfs2_quota_check()
and gfs2_inplace_reserve() to not return -EDQUOT/-ENOSPC when
there are atleast 'min_target' blocks allowable/available. The
assumption is that the caller is ok with just 'min_target' blocks
and will likely proceed with allocating them.
Signed-off-by: Abhi Das <adas@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Acked-by: Steven Whitehouse <swhiteho@redhat.com>
If we run out of blocks for a given multi-block allocation, we obviously
did not reserve enough. We should reserve more blocks for the next
reservation to reduce fragmentation. This patch increases the size hint
for reservations when they run out.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch tries to use the journal numbers to evenly distribute
which node prefers which resource group for block allocations. This
is to help performance.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch speeds up GFS2 unlink operations by using function
gfs2_rbm_incr rather than continuously calculating the rbm.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch checks if i_goal is either zero or if doesn't exist
within any rgrp (i.e gfs2_blk2rgrpd() returns NULL). If so, it
assigns the ip->i_no_addr block as the i_goal.
There are two scenarios where a bad i_goal can result in a
-EBADSLT error.
1. Attempting to allocate to an existing inode:
Control reaches gfs2_inplace_reserve() and ip->i_goal is bad.
We need to fix i_goal here.
2. A new inode is created in a directory whose i_goal is hosed:
In this case, the parent dir's i_goal is copied onto the new
inode. Since the new inode is not yet created, the ip->i_no_addr
field is invalid and so, the fix in gfs2_inplace_reserve() as per
1) won't work in this scenario. We need to catch and fix it sooner
in the parent dir itself (gfs2_create_inode()), before it is
copied to the new inode.
Signed-off-by: Abhi Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
GFS2 has a transaction glock, which must be grabbed for every
transaction, whose purpose is to deal with freezing the filesystem.
Aside from this involving a large amount of locking, it is very easy to
make the current fsfreeze code hang on unfreezing.
This patch rewrites how gfs2 handles freezing the filesystem. The
transaction glock is removed. In it's place is a freeze glock, which is
cached (but not held) in a shared state by every node in the cluster
when the filesystem is mounted. This lock only needs to be grabbed on
freezing, and actions which need to be safe from freezing, like
recovery.
When a node wants to freeze the filesystem, it grabs this glock
exclusively. When the freeze glock state changes on the nodes (either
from shared to unlocked, or shared to exclusive), the filesystem does a
special log flush. gfs2_log_flush() does all the work for flushing out
the and shutting down the incore log, and then it tries to grab the
freeze glock in a shared state again. Since the filesystem is stuck in
gfs2_log_flush, no new transaction can start, and nothing can be written
to disk. Unfreezing the filesytem simply involes dropping the freeze
glock, allowing gfs2_log_flush() to grab and then release the shared
lock, so it is cached for next time.
However, in order for the unfreezing ioctl to occur, gfs2 needs to get a
shared lock on the filesystem root directory inode to check permissions.
If that glock has already been grabbed exclusively, fsfreeze will be
unable to get the shared lock and unfreeze the filesystem.
In order to allow the unfreeze, this patch makes gfs2 grab a shared lock
on the filesystem root directory during the freeze, and hold it until it
unfreezes the filesystem. The functions which need to grab a shared
lock in order to allow the unfreeze ioctl to be issued now use the lock
grabbed by the freeze code instead.
The freeze and unfreeze code take care to make sure that this shared
lock will not be dropped while another process is using it.
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Add pr_fmt, remove embedded "GFS2: " prefixes.
This now consistently emits lower case "gfs2: " for each message.
Other miscellanea around these changes:
o Add missing newlines
o Coalesce formats
o Realign arguments
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
-All printk(KERN_foo converted to pr_foo().
-Messages updated to fit in 80 columns.
-fs_macros converted as well.
-fs_printk removed.
Signed-off-by: Fabian Frederick <fabf@skynet.be>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Mark functions as static in gfs2/rgrp.c because they are not used
outside this file.
This eliminates the following warning in gfs2/rgrp.c:
fs/gfs2/rgrp.c:1092:5: warning: no previous prototype for ‘gfs2_rgrp_bh_get’ [-Wmissing-prototypes]
fs/gfs2/rgrp.c:1157:5: warning: no previous prototype for ‘update_rgrp_lvb’ [-Wmissing-prototypes]
Signed-off-by: Rashika Kheria <rashika.kheria@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is another step towards improving the allocation of xattr
blocks at inode allocation time. Here we take advantage of
Christoph's recent work on ACLs to allocate a block for the
xattrs early if we know that we will be adding ACLs to the
inode later on. The advantage of that is that it is much
more likely that we'll get a contiguous run of two blocks
where the first is the inode and the second is the xattr block.
We still have to fall back to the original system in case we
don't get the requested two contiguous blocks, or in case the
ACLs are too large to fit into the block.
Future patches will move more of the ACL setting code further
up the gfs2_inode_create() function. Also, I'd like to be
able to do the same thing with the xattrs from LSMs in
due course, too. That way we should be able to slowly reduce
the number of independent transactions, at least in the
most common cases.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is a small cleanup to function gfs2_rgrp_go_lock so that it
uses rgd instead of its more complicated twin.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Al Viro has tactfully pointed out that we are using the incorrect
error code in some cases. This patch fixes that, and also removes
the (unused) return value for glock dumping.
> * gfs2_iget() - ENOBUFS instead of ENOMEM. ENOBUFS is
> "No buffer space available (POSIX.1 (XSI STREAMS option))" and since
> we don't support STREAMS it's probably fair game, but... what the hell?
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Each rgrp header is represented as a single extent on disk, so we
can calculate the position within the address space, since we are
using address spaces mapped 1:1 to the disk. This means that it
is possible to use the range based versions of filemap_fdatawrite/wait
and for invalidating the page cache.
Our eventual intent is to then be able to merge the address spaces
used for rgrps into a single address space, rather than to have
one for each glock, saving memory and reducing complexity.
Since during umount, the rgrp structures are disposed of before
the glocks, we need to store the extent information in the glock
so that is is available for a final invalidation. This patch uses
a field which is otherwise unused in rgrp glocks to do that, so
that we do not have to expand the size of a glock.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Since gfs2_inplace_reserve() is always called with a valid
alloc parms structure, there is no need to test for this
within the function itself - and in any case, after we've
all ready dereferenced it anyway.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
With the preceding patch, we started accepting block reservations
smaller than the ideal size, which requires a lot more parsing of the
bitmaps. To reduce the amount of bitmap searching, this patch
implements a scheme whereby each rgrp keeps track of the point
at this multi-block reservations will fail.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is just basically a resend of a patch I posted earlier.
It didn't change from its original, except in diff offsets, etc:
This patch fixes a bug in the GFS2 block allocation code. The problem
starts if a process already has a multi-block reservation, but for
some reason, another process disqualifies it from further allocations.
For example, the other process might set on the GFS2_RDF_ERROR bit.
The process holding the reservation jumps to label skip_rgrp, but
that label comes after the code that removes the reservation from the
tree. Therefore, the no longer usable reservation is not removed from
the rgrp's reservations tree; it's lost. Eventually, the lost reservation
causes the count of reserved blocks to get off, and eventually that
causes a BUG_ON(rs->rs_rbm.rgd->rd_reserved < rs->rs_free) to trigger.
This patch moves the call to after label skip_rgrp so that the
disqualified reservation is properly removed from the tree, thus keeping
the rgrp rd_reserved count sane.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Here is a second try at a patch I posted earlier, which also implements
suggestions Steve made:
Before this patch, GFS2 would keep searching through all the rgrps
until it found one that had a chunk of free blocks big enough to
satisfy the size hint, which is based on the file write size,
regardless of whether the chunk was big enough to perform the write.
However, when doing big writes there may not be a large enough
chunk of free blocks in any rgrp, due to file system fragmentation.
The largest chunk may be big enough to satisfy the write request,
but it may not meet the ideal reservation size from the "size hint".
The writes would slow to a crawl because every write would search
every rgrp, then finally give up and default to a single-block write.
In my case, performance would drop from 425MB/s to 18KB/s, or 24000
times slower.
This patch basically makes it so that if we can't find a contiguous
chunk of blocks big enough to satisfy the sizehint, we'll use the
largest chunk of blocks we found that will still contain the write.
It does so by keeping track of the largest run of blocks within the
rgrp.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When setting the starting point for block allocation, there were calls
to both gfs2_rbm_to_block() and gfs2_rbm_from_block() in the common case
of there being an active reservation. The gfs2_rbm_from_block() function
can be quite slow, and since the two conversions were effectively a
no-op, it makes sense to avoid them entirely in this case.
There is no functional change here, but the code should be a bit more
efficient after this patch.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch adds a structure to contain allocation parameters with
the intention of future expansion of this structure. The idea is
that we should be able to add more information about the allocation
in the future in order to allow the allocator to make a better job
of placing the requests on-disk.
There is no functional difference from applying this patch.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The reservation for an inode should be cleared when it is truncated so
that we can start again at a different offset for future allocations.
We could try and do better than that, by resetting the search based on
where the truncation started from, but this is only a first step.
In addition, there are three callers of gfs2_rs_delete() but only one
of those should really be testing the value of i_writecount. While
we get away with that in the other cases currently, I think it would
be better if we made that test specific to the one case which
requires it.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Since the previous patch eliminated bi in favor of bii, this follow-on
patch needed to be adjusted accordingly. Here is the revised version.
This patch adds a new function, gfs2_rbm_incr, which increments
an rbm structure. This is more efficient than calling gfs2_rbm_to_block,
incrementing, then calling gfs2_rbm_from_block.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This is a respin of the original patch. As Steve pointed out, the
introduction of field bii makes it easy to eliminate bi itself.
This revised patch does just that, replacing bi with bii.
This patch adds a new field to the rbm structure, called bii,
which is an index into the array of bitmaps for an rgrp.
This replaces *bi which was a pointer to the bitmap.
This is being done for further optimizations.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When we used try locks for rgrps on block allocations, it was important
to clear the flags field so that we used a blocking hold on the glock.
Now that we're not doing try locks, clearing flags is unnecessary, and
a waste of time. In fact, it's probably doing the wrong thing because
it clears the GL_SKIP bit that was set for the lvb tracking purposes.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch introduces a new field in the bitmap structure called
bi_blocks. Its purpose is to save us from constantly multiplying
bi_len by the constant GFS2_NBBY. It also paves the way for more
optimization in a future patch.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
In function gfs2_rbm_from_block, it starts by checking if the block
falls within the first bitmap. It does so by checking if the rbm's
offset is less than (rbm->bi->bi_start + rbm->bi->bi_len) * GFS2_NBBY.
However, the first bitmap will always have bi_start==0. Therefore
this is an unnecessary calculation in a function that gets called
billions of times. This patch removes the reference to bi_start.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch correctly distinguishes two boundary conditions:
1. When the given range is entire within the unaccounted space between
two rgrps, and
2. The range begins beyond the end of the filesystem
Also fix the unit of the returned value r.len (total trimming) to be in bytes
instead of the (incorrect) 512 byte blocks
With this patch, GFS2 passes multiple iterations of all the relevant xfstests
(251, 260, 288) with different fs block sizes.
Signed-off-by: Abhi Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch calls get_write_access in a few functions. This
merely increases inode->i_writecount for the duration of the function.
That will ensure that any file closes won't delete the inode's
multi-block reservation while the function is running.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch changes the multi-block allocation code, such that
directory inodes only get a single block reserved in the bitmap.
That way, the bitmaps are more tightly packed together, and there
are fewer spans of free blocks for in-use block reservations.
This means it takes less time to find a free span of blocks in the
bitmap, which speeds things up. This increases the performance of
some workloads by almost 2X. In Nate's mockup.py script (which does
(1) create dir, (2) create dir in dir, (3) create file in that dir)
the test executes in 23 steps rather than 43 steps, a 47%
performance improvement.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The functions that delete block reservations from the rgrp block
reservations rbtree no longer use the ip parameter. This patch
eliminates the parameter.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch cleans up the inode creation code path in GFS2. After the
Orlov allocator was merged, a number of potential improvements are
now possible, and this is a first set of these.
The quota handling is now updated so that it matches the point in
the code where the allocation takes place. This means that the one
exception in gfs2_alloc_blocks relating to quota is now no longer
required, and we can use the generic code everywhere.
In addition the call to figure out whether we need to allocate any
extra blocks in order to add a directory entry is moved higher up
gfs2_create_inode. This means that if it returns an error, we
can deal with that at a stage where it is easier to handle that case.
The returned status cannot change during the function since we hold
an exclusive lock on the directory.
Two calls to gfs2_rindex_update have been changed to one, again at
the top of gfs2_create_inode to simplify error handling.
The time stamps are also now initialised earlier in the creation
process, this is gradually moving towards being able to remove the
call to gfs2_refresh_inode in gfs2_inode_create once we have all the
fields covered.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch changes GFS2's discard issuing code so that it calls
function sb_issue_discard rather than blkdev_issue_discard. The
code was calling blkdev_issue_discard and specifying the correct
sector offset and sector size, but blkdev_issue_discard expects
these values to be in terms of 512 byte sectors, even if the native
sector size for the device is different. Calling sb_issue_discard
with the BLOCK size instead ensures the correct block-to-512b-sector
translation. I verified that "minlen" is specified in blocks, so
comparing it to a number of blocks is correct.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The error code in gfs2_rs_alloc() is set to ENOMEM when error
but never be used, instead, gfs2_rs_alloc() always return 0.
Fix to return 'error'.
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Pull vfs pile (part one) from Al Viro:
"Assorted stuff - cleaning namei.c up a bit, fixing ->d_name/->d_parent
locking violations, etc.
The most visible changes here are death of FS_REVAL_DOT (replaced with
"has ->d_weak_revalidate()") and a new helper getting from struct file
to inode. Some bits of preparation to xattr method interface changes.
Misc patches by various people sent this cycle *and* ocfs2 fixes from
several cycles ago that should've been upstream right then.
PS: the next vfs pile will be xattr stuff."
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (46 commits)
saner proc_get_inode() calling conventions
proc: avoid extra pde_put() in proc_fill_super()
fs: change return values from -EACCES to -EPERM
fs/exec.c: make bprm_mm_init() static
ocfs2/dlm: use GFP_ATOMIC inside a spin_lock
ocfs2: fix possible use-after-free with AIO
ocfs2: Fix oops in ocfs2_fast_symlink_readpage() code path
get_empty_filp()/alloc_file() leave both ->f_pos and ->f_version zero
target: writev() on single-element vector is pointless
export kernel_write(), convert open-coded instances
fs: encode_fh: return FILEID_INVALID if invalid fid_type
kill f_vfsmnt
vfs: kill FS_REVAL_DOT by adding a d_weak_revalidate dentry op
nfsd: handle vfs_getattr errors in acl protocol
switch vfs_getattr() to struct path
default SET_PERSONALITY() in linux/elf.h
ceph: prepopulate inodes only when request is aborted
d_hash_and_lookup(): export, switch open-coded instances
9p: switch v9fs_set_create_acl() to inode+fid, do it before d_instantiate()
9p: split dropping the acls from v9fs_set_create_acl()
...
There is little common content in gfs2_trans_add_bh() between the data
and meta classes by the time that the functions which it calls are
taken into account. The intent here is to split this into two
separate functions. Stage one is to introduce gfs2_trans_add_data()
and gfs2_trans_add_meta() and update the callers accordingly.
Later patches will then pull in the content of gfs2_trans_add_bh()
and its dependent functions in order to clean up the code in this
area.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
In function rg_mblk_search, it's searching for multiple blocks in
a given state (e.g. "free"). If there's an active block reservation
its goal is the next free block of that. If the resource group
contains the dinode's goal block, that's used for the search. But
if neither is the case, it uses the rgrp's last allocated block.
That way, consecutive allocations appear after one another on media.
The problem comes in when you hit the end of the rgrp; it would never
start over and search from the beginning. This became a problem,
since if you deleted all the files and data from the rgrp, it would
never start over and find free blocks. So it had to keep searching
further out on the media to allocate blocks. This patch resets the
rd_last_alloc after it does an unsuccessful search at the end of
the rgrp.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch adds a return code check after calling function
gfs2_rbm_from_block while determining the free extent size.
That way, when the end of an rgrp is reached, it won't try
to process unaligned blocks after the end.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
QE aio tests uncovered a race condition in gfs2_rs_alloc where it's possible
to come out of the function with a valid ip->i_res allocation but it gets
freed before use resulting in a NULL ptr dereference.
This patch envelopes the initial short-circuit check for non-NULL ip->i_res
into the mutex lock. With this patch, I was able to successfully run the
reproducer test multiple times.
Resolves: rhbz#878476
Signed-off-by: Abhi Das <adas@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
The lksb struct already contains a pointer to the lvb,
so another directly from the glock struct is not needed.
Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
For filesystems with only a single resource group, we need to be careful
that the allocation loop will not land up with a NULL resource group. This
fixes a bug in a previous patch where the gfs2_rgrpd_get_next() function
was being used instead of gfs2_rgrpd_get_first()
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Just like ext3, this works on the root directory and any directory
with the +T flag set. Also, just like ext3, any subdirectory created
in one of the just mentioned cases will be allocated to a random
resource group (GFS2 equivalent of a block group).
If you are creating a set of directories, each of which will contain a
job running on a different node, then by setting +T on the parent
directory before creating the subdirectories, each will land up in a
different resource group, and thus resource group contention between
nodes will be kept to a minimum.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch uses information gathered by the recent glock statistics
patch in order to derrive a boolean verdict on the congestion
status of a resource group. This is then used when making decisions
on which resource group to choose during block allocation.
The aim is to avoid resource groups which are heavily contended
by other nodes, while still ensuring locality of access wherever
possible.
Once a reservation has been made in a particular resource group
we continue to use that resource group until a new reservation is
required. This should help to ensure that we do not change resource
groups too often.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch is a rewrite of function gfs2_rbm_from_block. Rather than
looping to find the right bitmap, the code now does a few simple
math calculations.
I compared the performance of both algorithms side by side and the new
algorithm is noticeably faster. Sample instrumentation output from a
"fast" machine:
5 million calls: millisec spent: Orig: 166 New: 113
5 million calls: millisec spent: Orig: 189 New: 114
In addition, I ran postmark (on a somewhat slowr CPU) before the after
the new algorithm was put in place and postmark showed a decent
improvement:
Before the new algorithm:
-------------------------
Time:
645 seconds total
584 seconds of transactions (171 per second)
Files:
150087 created (232 per second)
Creation alone: 100000 files (2083 per second)
Mixed with transactions: 50087 files (85 per second)
49995 read (85 per second)
49991 appended (85 per second)
150087 deleted (232 per second)
Deletion alone: 100174 files (7705 per second)
Mixed with transactions: 49913 files (85 per second)
Data:
273.42 megabytes read (434.08 kilobytes per second)
852.13 megabytes written (1.32 megabytes per second)
With the new algorithm:
-----------------------
Time:
599 seconds total
530 seconds of transactions (188 per second)
Files:
150087 created (250 per second)
Creation alone: 100000 files (1886 per second)
Mixed with transactions: 50087 files (94 per second)
49995 read (94 per second)
49991 appended (94 per second)
150087 deleted (250 per second)
Deletion alone: 100174 files (6260 per second)
Mixed with transactions: 49913 files (94 per second)
Data:
273.42 megabytes read (467.42 kilobytes per second)
852.13 megabytes written (1.42 megabytes per second)
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Currently implementation in gfs2 uses FITRIM arguments as it were in
file system blocks units which is wrong. The FITRIM arguments
(fstrim_range.start, fstrim_range.len and fstrim_range.minlen) are
actually in bytes.
Moreover, check for start argument beyond the end of file system, len
argument being smaller than file system block and minlen argument being
bigger than biggest resource group were missing.
This commit converts the code to convert FITRIM argument to file system
blocks and also adds appropriate checks mentioned above.
All the problems were recognised by xfstests 251 and 260.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When the fstrim_range argument is not provided by user in FITRIM ioctl
we should just return EFAULT and not promoting bad behaviour by filling
the structure in kernel. Let the user deal with it.
Signed-off-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Despite the return value from kmem_cache_zalloc() being checked, the
error wasn't being returned until after a possible null pointer
dereference. This patch returns the error immediately, allowing the
removal of the error variable.
Signed-off-by: Andrew Price <anprice@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch fixes an infinite loop in gfs2_rbm_find that was introduced
by the previous patch. The problem occurred when the length was less
than 3 but the rbm block was byte-aligned, causing it to improperly
return a extent length of zero, which caused it to spin.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Tested-by: Bob Peterson <rpeterso@redhat.com>
Tested-by: Barry Marson <bmarson@redhat.com>
With the recently added block reservation code, an additional function
was added to search for free blocks. This had a restriction of only being
able to search for aligned extents of free blocks. As a result the
allocation patterns when reserving blocks were suboptimal when the
existing allocation of blocks for an inode was not aligned to the same
boundary.
This patch resolves that problem by adding the ability for gfs2_rbm_find
to search for extents of a particular minimum size. We can then use
gfs2_rbm_find for both looking for reservations, and also looking for
free blocks on an individual basis when we actually come to do the
allocation later on. As a result we only need a single set of code
to deal with both situations.
The function gfs2_rbm_from_block() is moved up rgrp.c so that it
occurs before all of its callers.
Many thanks are due to Bob for helping track down the final issue in
this patch. That fix to the rb_tree traversal and to not share
block reservations from a dirctory to its children is included here.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
This patch stops multiple block allocations if a nonzero
return code is received from gfs2_rbm_from_block. Without
this patch, if enough pressure is put on the file system,
you get a kernel warning quickly followed by:
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<ffffffffa04f47e8>] gfs2_alloc_blocks+0x2c8/0x880 [gfs2]
With this patch, things run normally.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When rgd->rd_free_clone is less than rgd->rd_reserved, the
unclaimed_blocks() calculation would wrap and produce
incorrect results. This patch checks for this condition
when this function is called from gfs2_mblk_search()
In addition, the use of this particular function in other
places in the code has been dropped by means of a general
clean up of gfs2_inplace_reserve(). This function is now
much easier to follow.
Also the setting of the rgd->rd_last_alloc field is corrected.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch improves the tracing of block reservations by
removing some corner cases and also providing more useful
detail in the traces.
A new field is added to the reservation structure to contain
the inode number. This is used since in certain contexts it is
not possible to access the inode itself to obtain this information.
As a result we can then display the inode number for all tracepoints
and also in case we dump the resource group.
The "del" tracepoint operation has been removed. This could be called
with the reservation rgrp set to NULL. That resulted in not printing
the device number, and thus making the information largely useless
anyway. Also, the conditional on the rgrp being NULL can then be
removed from the tracepoint. After this change, all the block
reservation tracepoint calls will be called with the rgrp information.
The existing ins,clm and tdel calls to the block reservation tracepoint
are sufficient to track the entire life of the block reservation.
In gfs2_block_alloc() the error detection is updated to print out
the inode number of the problematic inode. This can then be compared
against the information in the glock dump,tracepoints, etc.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
When we get to the stage of allocating blocks, we know that the
resource group in question must contain enough free blocks, otherwise
gfs2_inplace_reserve() would have failed. So if we are left with only
free blocks which are reserved, then we must use those. This can happen
if another node has sneeked in and use some blocks reserved on this
node, for example. Generally this will happen very rarely and only
when the resouce group is nearly full.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
