linux.git/drivers/md/raid10.c, branch v2.6.14

[PATCH] gfp flags annotations - part 1

2005-10-08T22:00:57+00:00

 - added typedef unsigned int __nocast gfp_t;

 - replaced __nocast uses for gfp flags with gfp_t - it gives exactly
   the same warnings as far as sparse is concerned, doesn't change
   generated code (from gcc point of view we replaced unsigned int with
   typedef) and documents what's going on far better.

Signed-off-by: Al Viro 
Signed-off-by: Linus Torvalds

[PATCH] md: fix BUG when raid10 rebuilds without enough drives

2005-09-09T23:39:15+00:00

This shouldn't be a BUG.  We should cope.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: fix raid10 assembly when too many devices are missing

2005-09-09T23:39:14+00:00

If you try to assemble an array with too many missing devices, raid10 will now
reject the attempt, instead of allowing it.

Also check when hot-adding a drive and refuse the hot-add if the array is
beyond hope.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: fail IO request to md that require a barrier.

2005-09-09T23:39:09+00:00

md does not yet support BIO_RW_BARRIER, so be honest about it and fail
(-EOPNOTSUPP) any such requests.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: fix minor error in raid10 read-balancing calculation.

2005-09-09T23:39:09+00:00

'this_sector' is a virtual (array) address while 'head_position' is a physical
(device) address, so substraction doesn't make any sense.  devs[slot].addr
should be used instead of this_sector.

However, this patch doesn't make much practical different to the read
balancing due to the effects of later code.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: remove unneeded NULL checks before kfree

2005-06-22T02:07:48+00:00

This patch removes some unneeded checks of pointers being NULL before
calling kfree() on them.  kfree() handles NULL pointers just fine, checking
first is pointless.

Signed-off-by: Jesper Juhl 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: fix deadlock due to md thread processing delayed requests.

2005-06-22T02:07:46+00:00

Before completing a 'write' the md superblock might need to be updated.
This is best done by the md_thread.

The current code schedules this up and queues the write request for later
handling by the md_thread.

However some personalities (Raid5/raid6) will deadlock if the md_thread
tries to submit requests to its own array.

So this patch changes things so the processes submitting the request waits
for the superblock to be written and then submits the request itself.

This fixes a recently-created deadlock in raid5/raid6

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: improve the interface to sync_request

2005-06-22T02:07:43+00:00

1/ change the return value (which is number-of-sectors synced)
 from 'int' to 'sector_t'.
 The number of sectors is usually easily small enough to fit
 in an int, but if resync needs to abort, it may want to return
 the total number of remaining sectors, which could be large.
 Also errors cannot be returned as negative numbers now, so use
 0 instead
2/ Add a 'skipped' return parameter to allow the array to report
 that it skipped the sectors.  This allows md to take this into account
 in the speed calculations.
 Currently there is no important skipping, but the bitmap-based-resync
 that is coming will use this.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: improve locking on 'safemode' and move superblock writes

2005-06-22T02:07:43+00:00

When md marks the superblock dirty before a write, it calls
generic_make_request (to write the superblock) from within
generic_make_request (to write the first dirty block), which could cause
problems later.

With this patch, the superblock write is always done by the helper thread, and
write request are delayed until that write completes.

Also, the locking around marking the array dirty and writing the superblock is
improved to avoid possible races.

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[PATCH] md: merge md_enter_safemode into md_check_recovery

2005-06-22T02:07:42+00:00

md_enter_safemode checks if it is time to mark the md superblock as 'clean'.
i.e.  if all writes have completed and a suitable delay has passed.

This is currently called from md_handle_safemode which in-turn is called
(almost) every time md_check_recovery is called, and from the end of
md_do_sync which causes the mddev->thread to run, which will always call
md_check_recovery as well.

So it doesn't need to be a separate function and fits quite well into
md_check_recovery.

The "almost" is because multipathd calls md_check_recovery but not
md_handle_safemode.  This is OK because the code from md_enter_safemode is a
no-op if mddev->safemode == 0, which it always is for a multipathd (providing
we don't allow it to be set to 2 on a signal...)

Signed-off-by: Neil Brown 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds