LU-1115 kernel: software raid6 related BUG

Software raid6 hit BUGON in fs/bio.c:222 when raid chunk > 64k.
We pull upstream patch: 5b99c2ffa980528a197f26c7d876cceeccce8dd5
to deal with this issue.

Signed-off-by: yang sheng <ys@whamcloud.com>
Change-Id: I5330bc161e7cf5364a614547949323fc9a3ee7e3
Reviewed-on: http://review.whamcloud.com/2625
Reviewed-by: Bobi Jam <bobijam@whamcloud.com>
Reviewed-by: Johann Lombardi <johann@whamcloud.com>
Tested-by: Hudson
Tested-by: Maloo <whamcloud.maloo@gmail.com>

diff --git a/lustre/kernel_patches/patches/make-bi_phys_segments-uint.patch b/lustre/kernel_patches/patches/make-bi_phys_segments-uint.patch
new file mode 100644 (file)
index 0000000..44a88f8
--- /dev/null
+++ b/lustre/kernel_patches/patches/make-bi_phys_segments-uint.patch
@@ -0,0 +1,73 @@
+From 5b99c2ffa980528a197f26c7d876cceeccce8dd5 Mon Sep 17 00:00:00 2001
+From: Jens Axboe <jens.axboe@oracle.com>
+Date: Fri, 15 Aug 2008 10:56:11 +0200
+Subject: [PATCH] block: make bi_phys_segments an unsigned int instead of
+ short
+
+raid5 can overflow with more than 255 stripes, and we can increase it
+to an int for free on both 32 and 64-bit archs due to the padding.
+
+Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
+---
+ drivers/md/raid5.c  |   12 ++++++------
+ include/linux/bio.h |    2 +-
+ 2 files changed, 7 insertions(+), 7 deletions(-)
+
+diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
+index 05b2292..37e5465 100644
+--- a/drivers/md/raid5.c
++++ b/drivers/md/raid5.c
+@@ -102,17 +102,17 @@ const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
+ #endif
+ 
+ /*
+- * We maintain a biased count of active stripes in the bottom 8 bits of
+- * bi_phys_segments, and a count of processed stripes in the upper 8 bits
++ * We maintain a biased count of active stripes in the bottom 16 bits of
++ * bi_phys_segments, and a count of processed stripes in the upper 16 bits
+  */
+ static inline int raid5_bi_phys_segments(struct bio *bio)
+ {
+-      return bio->bi_phys_segments & 0xff;
++      return bio->bi_phys_segments & 0xffff;
+ }
+ 
+ static inline int raid5_bi_hw_segments(struct bio *bio)
+ {
+-      return (bio->bi_phys_segments >> 8) & 0xff;
++      return (bio->bi_phys_segments >> 16) & 0xffff;
+ }
+ 
+ static inline int raid5_dec_bi_phys_segments(struct bio *bio)
+@@ -126,13 +126,13 @@ static inline int raid5_dec_bi_hw_segments(struct bio *bio)
+       unsigned short val = raid5_bi_hw_segments(bio);
+ 
+       --val;
+-      bio->bi_phys_segments = (val << 8) | raid5_bi_phys_segments(bio);
++      bio->bi_phys_segments = (val << 16) | raid5_bi_phys_segments(bio);
+       return val;
+ }
+ 
+ static inline void raid5_set_bi_hw_segments(struct bio *bio, unsigned int cnt)
+ {
+-      bio->bi_phys_segments = raid5_bi_phys_segments(bio) || (cnt << 8);
++      bio->bi_phys_segments = raid5_bi_phys_segments(bio) || (cnt << 16);
+ }
+ 
+ static inline int raid6_next_disk(int disk, int raid_disks)
+diff --git a/include/linux/bio.h b/include/linux/bio.h
+index dfc3556..2c0c090 100644
+--- a/include/linux/bio.h
++++ b/include/linux/bio.h
+@@ -75,7 +75,7 @@ struct bio {
+       /* Number of segments in this BIO after
+        * physical address coalescing is performed.
+        */
+-      unsigned short          bi_phys_segments;
++      unsigned int            bi_phys_segments;
+ 
+       unsigned int            bi_size;        /* residual I/O count */
+ 
+-- 
+1.7.6.5
+

diff --git a/lustre/kernel_patches/series/2.6-rhel5.series b/lustre/kernel_patches/series/2.6-rhel5.series
index f81f552..6e22e25 100644 (file)
--- a/lustre/kernel_patches/series/2.6-rhel5.series
+++ b/lustre/kernel_patches/series/2.6-rhel5.series
@@ -24,3 +24,4 @@ md-mmp-unplug-dev.patch
 mpt-fusion-max-sge.patch
 jbd2-jcberr-2.6-rhel5.patch
 jbd2-commit-timer-no-jiffies-rounding.diff
+make-bi_phys_segments-uint.patch