The patch titled
Manage jbd allocations from its own slabs
has been added to the -mm tree. Its filename is
manage-jbd-allocations-from-its-own-slabs.patch
See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to find
out what to do about this
------------------------------------------------------
Subject: Manage jbd allocations from its own slabs
From: Badari Pulavarty <pbadari@xxxxxxxxxx>
JBD currently allocates commit and frozen buffers from slabs. With
CONFIG_SLAB_DEBUG, its possible for an allocation to cross the page
boundary causing IO problems.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=200127
So, instead of allocating these from regular slabs - manage allocation from
its own slabs and disable slab debug for these slabs.
Signed-off-by: Badari Pulavarty <pbadari@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxx>
---
fs/jbd/commit.c | 6 +-
fs/jbd/journal.c | 87 +++++++++++++++++++++++++++++++++++++++--
fs/jbd/transaction.c | 9 ++--
include/linux/jbd.h | 3 +
4 files changed, 94 insertions(+), 11 deletions(-)
diff -puN fs/jbd/commit.c~manage-jbd-allocations-from-its-own-slabs fs/jbd/commit.c
--- a/fs/jbd/commit.c~manage-jbd-allocations-from-its-own-slabs
+++ a/fs/jbd/commit.c
@@ -261,7 +261,7 @@ void journal_commit_transaction(journal_
struct buffer_head *bh = jh2bh(jh);
jbd_lock_bh_state(bh);
- kfree(jh->b_committed_data);
+ jbd_slab_free(jh->b_committed_data, bh->b_size);
jh->b_committed_data = NULL;
jbd_unlock_bh_state(bh);
}
@@ -745,14 +745,14 @@ restart_loop:
* Otherwise, we can just throw away the frozen data now.
*/
if (jh->b_committed_data) {
- kfree(jh->b_committed_data);
+ jbd_slab_free(jh->b_committed_data, bh->b_size);
jh->b_committed_data = NULL;
if (jh->b_frozen_data) {
jh->b_committed_data = jh->b_frozen_data;
jh->b_frozen_data = NULL;
}
} else if (jh->b_frozen_data) {
- kfree(jh->b_frozen_data);
+ jbd_slab_free(jh->b_frozen_data, bh->b_size);
jh->b_frozen_data = NULL;
}
diff -puN fs/jbd/journal.c~manage-jbd-allocations-from-its-own-slabs fs/jbd/journal.c
--- a/fs/jbd/journal.c~manage-jbd-allocations-from-its-own-slabs
+++ a/fs/jbd/journal.c
@@ -84,6 +84,7 @@ EXPORT_SYMBOL(journal_force_commit);
static int journal_convert_superblock_v1(journal_t *, journal_superblock_t *);
static void __journal_abort_soft (journal_t *journal, int errno);
+static int journal_create_jbd_slab(size_t slab_size);
/*
* Helper function used to manage commit timeouts
@@ -328,10 +329,10 @@ repeat:
char *tmp;
jbd_unlock_bh_state(bh_in);
- tmp = jbd_rep_kmalloc(bh_in->b_size, GFP_NOFS);
+ tmp = jbd_slab_alloc(bh_in->b_size, GFP_NOFS);
jbd_lock_bh_state(bh_in);
if (jh_in->b_frozen_data) {
- kfree(tmp);
+ jbd_slab_free(tmp, bh_in->b_size);
goto repeat;
}
@@ -1090,6 +1091,13 @@ int journal_load(journal_t *journal)
}
}
+ /*
+ * Make sure to create a slab for this blocksize
+ */
+ err = journal_create_jbd_slab(cpu_to_be32(journal->j_superblock->s_blocksize));
+ if (err)
+ return err;
+
/* Let the recovery code check whether it needs to recover any
* data from the journal. */
if (journal_recover(journal))
@@ -1612,6 +1620,76 @@ void * __jbd_kmalloc (const char *where,
}
/*
+ * jbd slab management: create 1k, 2k, 4k, 8k slabs as needed
+ * and allocate frozen and commit buffers from these slabs.
+ *
+ * Reason for doing this is to avoid, SLAB_DEBUG - since it could
+ * cause bh to cross page boundary.
+ */
+
+#define JBD_MAX_SLABS 5
+#define JBD_SLAB_INDEX(size) (size >> 11)
+
+static kmem_cache_t *jbd_slab[JBD_MAX_SLABS];
+static const char *jbd_slab_names[JBD_MAX_SLABS] = {
+ "jbd_1k", "jbd_2k", "jbd_4k", NULL, "jbd_8k"
+};
+
+static void journal_destroy_jbd_slabs(void)
+{
+ int i;
+
+ for (i=0; i<JBD_MAX_SLABS; i++) {
+ if (jbd_slab[i])
+ kmem_cache_destroy(jbd_slab[i]);
+ jbd_slab[i] = NULL;
+ }
+}
+
+static int journal_create_jbd_slab(size_t slab_size)
+{
+ int i = JBD_SLAB_INDEX(slab_size);
+
+ BUG_ON(i >= JBD_MAX_SLABS);
+ /*
+ * Check if we already have a slab created for this size
+ */
+ if (jbd_slab[i])
+ return 0;
+
+ /*
+ * Create a slab and force alignment to be same as slabsize -
+ * this will make sure that allocations won't cross the page
+ * boundary.
+ */
+ jbd_slab[i] = kmem_cache_create(jbd_slab_names[i],
+ slab_size, slab_size, 0, NULL, NULL);
+ if (!jbd_slab[i]) {
+ printk(KERN_EMERG "JBD: no memory for jbd_slab cache\n");
+ return -ENOMEM;
+ }
+ return 0;
+}
+
+void * jbd_slab_alloc(size_t size, gfp_t flags)
+{
+ int idx;
+
+ idx = JBD_SLAB_INDEX(size);
+ BUG_ON(jbd_slab[idx] == NULL);
+ return kmem_cache_alloc(jbd_slab[idx], flags | __GFP_NOFAIL);
+}
+
+void jbd_slab_free(void *ptr, size_t size)
+{
+ int idx;
+
+ idx = JBD_SLAB_INDEX(size);
+ BUG_ON(jbd_slab[idx] == NULL);
+ kmem_cache_free(jbd_slab[idx], ptr);
+}
+
+/*
* Journal_head storage management
*/
static kmem_cache_t *journal_head_cache;
@@ -1799,13 +1877,13 @@ static void __journal_remove_journal_hea
printk(KERN_WARNING "%s: freeing "
"b_frozen_data\n",
__FUNCTION__);
- kfree(jh->b_frozen_data);
+ jbd_slab_free(jh->b_frozen_data, bh->b_size);
}
if (jh->b_committed_data) {
printk(KERN_WARNING "%s: freeing "
"b_committed_data\n",
__FUNCTION__);
- kfree(jh->b_committed_data);
+ jbd_slab_free(jh->b_committed_data, bh->b_size);
}
bh->b_private = NULL;
jh->b_bh = NULL; /* debug, really */
@@ -1961,6 +2039,7 @@ static void journal_destroy_caches(void)
journal_destroy_revoke_caches();
journal_destroy_journal_head_cache();
journal_destroy_handle_cache();
+ journal_destroy_jbd_slabs();
}
static int __init journal_init(void)
diff -puN fs/jbd/transaction.c~manage-jbd-allocations-from-its-own-slabs fs/jbd/transaction.c
--- a/fs/jbd/transaction.c~manage-jbd-allocations-from-its-own-slabs
+++ a/fs/jbd/transaction.c
@@ -666,8 +666,9 @@ repeat:
if (!frozen_buffer) {
JBUFFER_TRACE(jh, "allocate memory for buffer");
jbd_unlock_bh_state(bh);
- frozen_buffer = jbd_kmalloc(jh2bh(jh)->b_size,
- GFP_NOFS);
+ frozen_buffer =
+ jbd_slab_alloc(jh2bh(jh)->b_size,
+ GFP_NOFS);
if (!frozen_buffer) {
printk(KERN_EMERG
"%s: OOM for frozen_buffer\n",
@@ -879,7 +880,7 @@ int journal_get_undo_access(handle_t *ha
repeat:
if (!jh->b_committed_data) {
- committed_data = jbd_kmalloc(jh2bh(jh)->b_size, GFP_NOFS);
+ committed_data = jbd_slab_alloc(jh2bh(jh)->b_size, GFP_NOFS);
if (!committed_data) {
printk(KERN_EMERG "%s: No memory for committed data\n",
__FUNCTION__);
@@ -906,7 +907,7 @@ repeat:
out:
journal_put_journal_head(jh);
if (unlikely(committed_data))
- kfree(committed_data);
+ jbd_slab_free(committed_data, bh->b_size);
return err;
}
diff -puN include/linux/jbd.h~manage-jbd-allocations-from-its-own-slabs include/linux/jbd.h
--- a/include/linux/jbd.h~manage-jbd-allocations-from-its-own-slabs
+++ a/include/linux/jbd.h
@@ -72,6 +72,9 @@ extern int journal_enable_debug;
#endif
extern void * __jbd_kmalloc (const char *where, size_t size, gfp_t flags, int retry);
+extern void * jbd_slab_alloc(size_t size, gfp_t flags);
+extern void jbd_slab_free(void *ptr, size_t size);
+
#define jbd_kmalloc(size, flags) \
__jbd_kmalloc(__FUNCTION__, (size), (flags), journal_oom_retry)
#define jbd_rep_kmalloc(size, flags) \
_
Patches currently in -mm which might be from pbadari@xxxxxxxxxx are
manage-jbd-allocations-from-its-own-slabs.patch
pass-io-size-to-batch_write-address-space-operation.patch
vectorize-aio_read-aio_write-fileop-methods.patch
vectorize-aio_read-aio_write-fileop-methods-xfs-fix.patch
vectorize-aio_read-aio_write-fileop-methods-hypfs-fix.patch
remove-readv-writev-methods-and-use-aio_read-aio_write.patch
streamline-generic_file_-interfaces-and-filemap.patch
add-vector-aio-support.patch
add-vector-aio-support-fix.patch
streamline-generic_file_-interfaces-and-filemap-ecryptfs.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
