linux/fs/btrfs/locking.c
Tejun Heo 5a0e3ad6af include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files.  percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed.  Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability.  As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

  http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
  only the necessary includes are there.  ie. if only gfp is used,
  gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
  blocks and try to put the new include such that its order conforms
  to its surrounding.  It's put in the include block which contains
  core kernel includes, in the same order that the rest are ordered -
  alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
  doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
  because the file doesn't have fitting include block), it prints out
  an error message indicating which .h file needs to be added to the
  file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
   over 4000 files, deleting around 700 includes and adding ~480 gfp.h
   and ~3000 slab.h inclusions.  The script emitted errors for ~400
   files.

2. Each error was manually checked.  Some didn't need the inclusion,
   some needed manual addition while adding it to implementation .h or
   embedding .c file was more appropriate for others.  This step added
   inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
   from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
   e.g. lib/decompress_*.c used malloc/free() wrappers around slab
   APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
   editing them as sprinkling gfp.h and slab.h inclusions around .h
   files could easily lead to inclusion dependency hell.  Most gfp.h
   inclusion directives were ignored as stuff from gfp.h was usually
   wildly available and often used in preprocessor macros.  Each
   slab.h inclusion directive was examined and added manually as
   necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
   were fixed.  CONFIG_GCOV_KERNEL was turned off for all tests (as my
   distributed build env didn't work with gcov compiles) and a few
   more options had to be turned off depending on archs to make things
   build (like ipr on powerpc/64 which failed due to missing writeq).

   * x86 and x86_64 UP and SMP allmodconfig and a custom test config.
   * powerpc and powerpc64 SMP allmodconfig
   * sparc and sparc64 SMP allmodconfig
   * ia64 SMP allmodconfig
   * s390 SMP allmodconfig
   * alpha SMP allmodconfig
   * um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
   a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.

Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-30 22:02:32 +09:00

233 lines
6.1 KiB
C

/*
* Copyright (C) 2008 Oracle. All rights reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public
* License v2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public
* License along with this program; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*/
#include <linux/sched.h>
#include <linux/pagemap.h>
#include <linux/spinlock.h>
#include <linux/page-flags.h>
#include <asm/bug.h>
#include "ctree.h"
#include "extent_io.h"
#include "locking.h"
static inline void spin_nested(struct extent_buffer *eb)
{
spin_lock(&eb->lock);
}
/*
* Setting a lock to blocking will drop the spinlock and set the
* flag that forces other procs who want the lock to wait. After
* this you can safely schedule with the lock held.
*/
void btrfs_set_lock_blocking(struct extent_buffer *eb)
{
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags)) {
set_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags);
spin_unlock(&eb->lock);
}
/* exit with the spin lock released and the bit set */
}
/*
* clearing the blocking flag will take the spinlock again.
* After this you can't safely schedule
*/
void btrfs_clear_lock_blocking(struct extent_buffer *eb)
{
if (test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags)) {
spin_nested(eb);
clear_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags);
smp_mb__after_clear_bit();
}
/* exit with the spin lock held */
}
/*
* unfortunately, many of the places that currently set a lock to blocking
* don't end up blocking for very long, and often they don't block
* at all. For a dbench 50 run, if we don't spin on the blocking bit
* at all, the context switch rate can jump up to 400,000/sec or more.
*
* So, we're still stuck with this crummy spin on the blocking bit,
* at least until the most common causes of the short blocks
* can be dealt with.
*/
static int btrfs_spin_on_block(struct extent_buffer *eb)
{
int i;
for (i = 0; i < 512; i++) {
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
return 1;
if (need_resched())
break;
cpu_relax();
}
return 0;
}
/*
* This is somewhat different from trylock. It will take the
* spinlock but if it finds the lock is set to blocking, it will
* return without the lock held.
*
* returns 1 if it was able to take the lock and zero otherwise
*
* After this call, scheduling is not safe without first calling
* btrfs_set_lock_blocking()
*/
int btrfs_try_spin_lock(struct extent_buffer *eb)
{
int i;
if (btrfs_spin_on_block(eb)) {
spin_nested(eb);
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
return 1;
spin_unlock(&eb->lock);
}
/* spin for a bit on the BLOCKING flag */
for (i = 0; i < 2; i++) {
cpu_relax();
if (!btrfs_spin_on_block(eb))
break;
spin_nested(eb);
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
return 1;
spin_unlock(&eb->lock);
}
return 0;
}
/*
* the autoremove wake function will return 0 if it tried to wake up
* a process that was already awake, which means that process won't
* count as an exclusive wakeup. The waitq code will continue waking
* procs until it finds one that was actually sleeping.
*
* For btrfs, this isn't quite what we want. We want a single proc
* to be notified that the lock is ready for taking. If that proc
* already happen to be awake, great, it will loop around and try for
* the lock.
*
* So, btrfs_wake_function always returns 1, even when the proc that we
* tried to wake up was already awake.
*/
static int btrfs_wake_function(wait_queue_t *wait, unsigned mode,
int sync, void *key)
{
autoremove_wake_function(wait, mode, sync, key);
return 1;
}
/*
* returns with the extent buffer spinlocked.
*
* This will spin and/or wait as required to take the lock, and then
* return with the spinlock held.
*
* After this call, scheduling is not safe without first calling
* btrfs_set_lock_blocking()
*/
int btrfs_tree_lock(struct extent_buffer *eb)
{
DEFINE_WAIT(wait);
wait.func = btrfs_wake_function;
if (!btrfs_spin_on_block(eb))
goto sleep;
while(1) {
spin_nested(eb);
/* nobody is blocking, exit with the spinlock held */
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
return 0;
/*
* we have the spinlock, but the real owner is blocking.
* wait for them
*/
spin_unlock(&eb->lock);
/*
* spin for a bit, and if the blocking flag goes away,
* loop around
*/
cpu_relax();
if (btrfs_spin_on_block(eb))
continue;
sleep:
prepare_to_wait_exclusive(&eb->lock_wq, &wait,
TASK_UNINTERRUPTIBLE);
if (test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
schedule();
finish_wait(&eb->lock_wq, &wait);
}
return 0;
}
/*
* Very quick trylock, this does not spin or schedule. It returns
* 1 with the spinlock held if it was able to take the lock, or it
* returns zero if it was unable to take the lock.
*
* After this call, scheduling is not safe without first calling
* btrfs_set_lock_blocking()
*/
int btrfs_try_tree_lock(struct extent_buffer *eb)
{
if (spin_trylock(&eb->lock)) {
if (test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags)) {
/*
* we've got the spinlock, but the real owner is
* blocking. Drop the spinlock and return failure
*/
spin_unlock(&eb->lock);
return 0;
}
return 1;
}
/* someone else has the spinlock giveup */
return 0;
}
int btrfs_tree_unlock(struct extent_buffer *eb)
{
/*
* if we were a blocking owner, we don't have the spinlock held
* just clear the bit and look for waiters
*/
if (test_and_clear_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
smp_mb__after_clear_bit();
else
spin_unlock(&eb->lock);
if (waitqueue_active(&eb->lock_wq))
wake_up(&eb->lock_wq);
return 0;
}
void btrfs_assert_tree_locked(struct extent_buffer *eb)
{
if (!test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags))
assert_spin_locked(&eb->lock);
}