[PATCH] [GFS2] Simplify multiple glock aquisition

From: swhiteho
Date: Mon Jul 09 2007 - 12:38:58 EST


From: Steven Whitehouse <swhiteho@xxxxxxxxxx>

There is a bug in the code which acquires multiple glocks where if the
initial out-of-order attempt fails part way though we can land up trying
to acquire the wrong number of glocks. This is part of the fix for red
hat bz #239737. The other part of the bz doesn't apply to upstream
kernels since it was fixed by:

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=d3717bdf8f08a0e1039158c8bab2c24d20f492b6

Since the out-of-order code doesn't appear to add anything to the
performance of GFS2, this patch just removed it rather than trying to
fix it. It should be much easier to see whats going on here now. In
addition, we don't allocate any memory unless we are using a lot of
glocks (which is a relatively uncommon case).

Signed-off-by: Steven Whitehouse <swhiteho@xxxxxxxxxx>

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 384cae6..3f0974e 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -1327,10 +1327,6 @@ static int nq_m_sync(unsigned int num_gh, struct gfs2_holder *ghs,
* @num_gh: the number of structures
* @ghs: an array of struct gfs2_holder structures
*
- * Figure out how big an impact this function has. Either:
- * 1) Replace this code with code that calls gfs2_glock_prefetch()
- * 2) Forget async stuff and just call nq_m_sync()
- * 3) Leave it like it is
*
* Returns: 0 on success (all glocks acquired),
* errno on failure (no glocks acquired)
@@ -1338,62 +1334,28 @@ static int nq_m_sync(unsigned int num_gh, struct gfs2_holder *ghs,

int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs)
{
- int *e;
- unsigned int x;
- int borked = 0, serious = 0;
+ struct gfs2_holder *tmp[4];
+ struct gfs2_holder **pph = tmp;
int error = 0;

- if (!num_gh)
+ switch(num_gh) {
+ case 0:
return 0;
-
- if (num_gh == 1) {
+ case 1:
ghs->gh_flags &= ~(LM_FLAG_TRY | GL_ASYNC);
return gfs2_glock_nq(ghs);
- }
-
- e = kcalloc(num_gh, sizeof(struct gfs2_holder *), GFP_KERNEL);
- if (!e)
- return -ENOMEM;
-
- for (x = 0; x < num_gh; x++) {
- ghs[x].gh_flags |= LM_FLAG_TRY | GL_ASYNC;
- error = gfs2_glock_nq(&ghs[x]);
- if (error) {
- borked = 1;
- serious = error;
- num_gh = x;
+ default:
+ if (num_gh <= 4)
break;
- }
+ pph = kmalloc(num_gh * sizeof(struct gfs2_holder *), GFP_NOFS);
+ if (!pph)
+ return -ENOMEM;
}

- for (x = 0; x < num_gh; x++) {
- error = e[x] = glock_wait_internal(&ghs[x]);
- if (error) {
- borked = 1;
- if (error != GLR_TRYFAILED && error != GLR_CANCELED)
- serious = error;
- }
- }
-
- if (!borked) {
- kfree(e);
- return 0;
- }
-
- for (x = 0; x < num_gh; x++)
- if (!e[x])
- gfs2_glock_dq(&ghs[x]);
-
- if (serious)
- error = serious;
- else {
- for (x = 0; x < num_gh; x++)
- gfs2_holder_reinit(ghs[x].gh_state, ghs[x].gh_flags,
- &ghs[x]);
- error = nq_m_sync(num_gh, ghs, (struct gfs2_holder **)e);
- }
+ error = nq_m_sync(num_gh, ghs, pph);

- kfree(e);
+ if (pph != tmp)
+ kfree(pph);

return error;
}
--
1.5.1.2

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/