Hi,
I got these errors when switch the kernel from 3.1-rc8 to linux-next. Commit
4fd30c0330ba951 looks most likely the culprit in the linux-next change set.
[ 301.322279] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.322401] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370008 with max blocks 2 with error -5
[ 301.322403] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.322404]
[ 301.322580] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.322702] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370010 with max blocks 5 with error -5
[ 301.322704] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.322705]
[ 301.322885] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.323006] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370015 with max blocks 3 with error -5
[ 301.323008] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.323010]
[ 301.323414] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #13: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.323489] EXT4-fs (sda7): delayed block allocation failed for inode 13 at logical offset 3417483 with max blocks 1461 with error -5
[ 301.323490] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.323491]
[ 301.366611] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.366694] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370018 with max blocks 344 with error -5
[ 301.366696] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.366696]
[ 301.367028] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.367117] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370362 with max blocks 13 with error -5
[ 301.367118] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.367119]
[ 301.367257] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.367341] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370375 with max blocks 3 with error -5
[ 301.367342] EXT4-fs (sda7): This should not happen!! Data will be lost
[ 301.367343]
[ 301.367487] EXT4-fs error (device sda7): ext4_ext_insert_index:783: inode #12: comm flush-8:0: ix > EXT_LAST_INDEX!
[ 301.367606] EXT4-fs (sda7): delayed block allocation failed for inode 12 at logical offset 3370378 with max blocks 2 with error -5
[ 301.367609] EXT4-fs (sda7): This should not happen!! Data will be lost
It all starts when comparing the below kernels' performance. I noticed
two outstanding regressions in the dirty_thresh=1M,8M cases of ext4.
The two cases with regressions are both running 2 sequential write dd's.
3.1.0-rc8-ioless6a+ 3.1.0-rc9-ioless-full-next-20111014+
------------------------ ------------------------
58.23 -3.0% 56.47 thresh=100M/btrfs-10dd-4k-8p-4096M-100M:10-X
58.43 -3.7% 56.28 thresh=100M/btrfs-1dd-4k-8p-4096M-100M:10-X
58.53 -4.1% 56.11 thresh=100M/btrfs-2dd-4k-8p-4096M-100M:10-X
37.34 +1.4% 37.86 thresh=100M/ext3-10dd-4k-8p-4096M-100M:10-X
44.44 +3.3% 45.91 thresh=100M/ext3-1dd-4k-8p-4096M-100M:10-X
41.70 +0.4% 41.87 thresh=100M/ext3-2dd-4k-8p-4096M-100M:10-X
46.45 -1.7% 45.68 thresh=100M/ext4-10dd-4k-8p-4096M-100M:10-X
56.60 -1.5% 55.74 thresh=100M/ext4-1dd-4k-8p-4096M-100M:10-X
54.14 -1.5% 53.32 thresh=100M/ext4-2dd-4k-8p-4096M-100M:10-X
44.53 +3.8% 46.20 thresh=100M/xfs-10dd-4k-8p-4096M-100M:10-X
55.89 -0.3% 55.72 thresh=100M/xfs-1dd-4k-8p-4096M-100M:10-X
51.11 +5.7% 54.01 thresh=100M/xfs-2dd-4k-8p-4096M-100M:10-X
56.55 -2.6% 55.08 thresh=1G/btrfs-100dd-4k-8p-4096M-1024M:10-X
56.11 -1.1% 55.49 thresh=1G/btrfs-10dd-4k-8p-4096M-1024M:10-X
56.21 -1.5% 55.38 thresh=1G/btrfs-1dd-4k-8p-4096M-1024M:10-X
30.66 +19.7% 36.70 thresh=1G/ext3-100dd-4k-8p-4096M-1024M:10-X
35.24 +15.3% 40.64 thresh=1G/ext3-10dd-4k-8p-4096M-1024M:10-X
43.58 +11.6% 48.65 thresh=1G/ext3-1dd-4k-8p-4096M-1024M:10-X
50.42 -1.2% 49.84 thresh=1G/ext4-100dd-4k-8p-4096M-1024M:10-X
56.23 -0.4% 56.03 thresh=1G/ext4-10dd-4k-8p-4096M-1024M:10-X
58.12 -1.2% 57.42 thresh=1G/ext4-1dd-4k-8p-4096M-1024M:10-X
41.76 +9.5% 45.74 thresh=1G/xfs-100dd-4k-8p-4096M-1024M:10-X
48.34 +12.1% 54.19 thresh=1G/xfs-10dd-4k-8p-4096M-1024M:10-X
52.36 +6.8% 55.93 thresh=1G/xfs-1dd-4k-8p-4096M-1024M:10-X
32.09 -0.8% 31.82 thresh=1M/ext4-10dd-4k-8p-4096M-1M:10-X
51.36 +1.9% 52.33 thresh=1M/ext4-1dd-4k-8p-4096M-1M:10-X
==> 46.93 -59.4% 19.05 thresh=1M/ext4-2dd-4k-8p-4096M-1M:10-X
28.68 -0.9% 28.43 thresh=1M/xfs-10dd-4k-8p-4096M-1M:10-X
51.95 +1.9% 52.93 thresh=1M/xfs-1dd-4k-8p-4096M-1M:10-X
47.07 -0.4% 46.87 thresh=1M/xfs-2dd-4k-8p-4096M-1M:10-X
54.37 +0.3% 54.54 thresh=8M/btrfs-10dd-4k-8p-4096M-8M:10-X
56.12 +0.9% 56.60 thresh=8M/btrfs-1dd-4k-8p-4096M-8M:10-X
56.22 -0.0% 56.21 thresh=8M/btrfs-2dd-4k-8p-4096M-8M:10-X
32.21 +1.0% 32.54 thresh=8M/ext3-10dd-4k-8p-4096M-8M:10-X
45.37 +1.4% 46.01 thresh=8M/ext3-1dd-4k-8p-4096M-8M:10-X
43.71 +1.0% 44.13 thresh=8M/ext3-2dd-4k-8p-4096M-8M:10-X
35.58 +0.5% 35.78 thresh=8M/ext4-10dd-4k-8p-4096M-8M:10-X
56.39 -1.9% 55.29 thresh=8M/ext4-1dd-4k-8p-4096M-8M:10-X
==> 51.26 -63.2% 18.85 thresh=8M/ext4-2dd-4k-8p-4096M-8M:10-X
31.07 +0.5% 31.21 thresh=8M/xfs-10dd-4k-8p-4096M-8M:10-X
55.44 -2.4% 54.10 thresh=8M/xfs-1dd-4k-8p-4096M-8M:10-X
47.59 -1.3% 46.97 thresh=8M/xfs-2dd-4k-8p-4096M-8M:10-X
2016.38 -1.8% 1979.91 TOTAL write_bw
It's not likely media error because the other cases work fine. And
grep shows that the error messages only show up in these two cases:
wfg@bee /export/writeback% grep -l ext4_ext_insert_index thresh*/*-ioless6*+/dmesg
wfg@bee /export/writeback% grep -l ext4_ext_insert_index thresh*/*-ioless-full-next-20111014+/dmesg
thresh=1M/ext4-2dd-4k-8p-2941M-1M:10-3.1.0-rc9-ioless-full-next-20111014+/dmesg
thresh=8M/ext4-2dd-4k-8p-2941M-8M:10-3.1.0-rc9-ioless-full-next-20111014+/dmesg
Thanks,
Fengguang
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
