The patch titled
mm: clean up and kernelify shrinker registration
has been removed from the -mm tree. Its filename was
mm-clean-up-and-kernelify-shrinker-registration-reiser4.patch
This patch was dropped because it was folded into reiser4.patch
------------------------------------------------------
Subject: mm: clean up and kernelify shrinker registration
From: Rusty Russell <rusty@xxxxxxxxxxxxxxx>
I can never remember what the function to register to receive VM pressure
is called. I have to trace down from __alloc_pages() to find it.
It's called "set_shrinker()", and it needs Your Help.
1) Don't hide struct shrinker. It contains no magic.
2) Don't allocate "struct shrinker". It's not helpful.
3) Call them "register_shrinker" and "unregister_shrinker".
4) Call the function "shrink" not "shrinker".
5) Reduce the 17 lines of waffly comments to 13, but document it properly.
The comment in reiser4 makes me a little queasy.
Signed-off-by: Rusty Russell <rusty@xxxxxxxxxxxxxxx>
Cc: Vladimir Saveliev <vs@xxxxxxxxxxx>
Cc: Edward Shishkin <edward@xxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/reiser4/fsdata.c | 32 ++++++++++++++------------------
1 file changed, 14 insertions(+), 18 deletions(-)
diff -puN fs/reiser4/fsdata.c~mm-clean-up-and-kernelify-shrinker-registration-reiser4 fs/reiser4/fsdata.c
--- a/fs/reiser4/fsdata.c~mm-clean-up-and-kernelify-shrinker-registration-reiser4
+++ a/fs/reiser4/fsdata.c
@@ -7,7 +7,6 @@
/* cache or dir_cursors */
static struct kmem_cache *d_cursor_cache;
-static struct shrinker *d_cursor_shrinker;
/* list of unused cursors */
static LIST_HEAD(cursor_cache);
@@ -53,6 +52,18 @@ static int d_cursor_shrink(int nr, gfp_t
return d_cursor_unused;
}
+/*
+ * actually, d_cursors are "priceless", because there is no way to
+ * recover information stored in them. On the other hand, we don't
+ * want to consume all kernel memory by them. As a compromise, just
+ * assign higher "seeks" value to d_cursor cache, so that it will be
+ * shrunk only if system is really tight on memory.
+ */
+static struct shrinker d_cursor_shrinker = {
+ .shrink = d_cursor_shrink,
+ .seeks = DEFAULT_SEEKS << 3,
+};
+
/**
* reiser4_init_d_cursor - create d_cursor cache
*
@@ -66,20 +77,7 @@ int reiser4_init_d_cursor(void)
if (d_cursor_cache == NULL)
return RETERR(-ENOMEM);
- /*
- * actually, d_cursors are "priceless", because there is no way to
- * recover information stored in them. On the other hand, we don't
- * want to consume all kernel memory by them. As a compromise, just
- * assign higher "seeks" value to d_cursor cache, so that it will be
- * shrunk only if system is really tight on memory.
- */
- d_cursor_shrinker = set_shrinker(DEFAULT_SEEKS << 3,
- d_cursor_shrink);
- if (d_cursor_shrinker == NULL) {
- destroy_reiser4_cache(&d_cursor_cache);
- d_cursor_cache = NULL;
- return RETERR(-ENOMEM);
- }
+ register_shrinker(&d_cursor_shrinker);
return 0;
}
@@ -90,9 +88,7 @@ int reiser4_init_d_cursor(void)
*/
void reiser4_done_d_cursor(void)
{
- BUG_ON(d_cursor_shrinker == NULL);
- remove_shrinker(d_cursor_shrinker);
- d_cursor_shrinker = NULL;
+ unregister_shrinker(&d_cursor_shrinker);
destroy_reiser4_cache(&d_cursor_cache);
}
_
Patches currently in -mm which might be from rusty@xxxxxxxxxxxxxxx are
origin.patch
update-boot-spec-to-207.patch
add-weak-for-creating-weak-asm-labels.patch
i386-paravirt-boot-sequence.patch
git-kvm.patch
reiser4.patch
mm-clean-up-and-kernelify-shrinker-registration-reiser4.patch
-
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
