Re: 2.6.17-rc5-mm2 md cause BUGs, and readahead speedup

[Helge Hafting <helge.hafting@aitel.hist.no>]

The good stuff first, bootup went from 55s to 43s according to 
bootchart. :-)
Probably the readahead stuff.

I got some BUG messages in dmesg though, they
seem related to md initialization:

Freeing unused kernel memory: 236k freed
md: Autodetecting RAID arrays.
BUG: warning at fs/block_dev.c:944/do_open()
<c015748d> do_open+0x2e8/0x2ed  <c025d905> cfb_imageblit+0x85/0x550
<c02537e9> soft_cursor+0x13d/0x1a4  <c0166446> iget5_locked+0xe6/0x14e
<c0157668> bdev_test+0x0/0xc  <c0156aec> bdget+0xd6/0xde
<c015763d> open_partition_by_devnum+0x69/0x7c  <c034dcfa> 
md_import_device+0x78/0x24b
<c023aedc> kobject_register+0x30/0x35  <c034cdf3> md_probe+0x12e/0x16b
<c03506ef> md_ioctl+0x27a/0x154b  <c034f857> md_open+0x48/0x4f
<c01573d2> do_open+0x22d/0x2ed  <c015c659> do_lookup+0x47/0x126
<c02a84f4> scrup+0xca/0xd4  <c016422f> dput+0xba/0x177
<c015cee9> __link_path_walk+0x7b1/0xc37  <c0252d37> bit_cursor+0x34d/0x5c3
<c01674a0> mntput_no_expire+0x13/0x52  <c0350475> md_ioctl+0x0/0x154b
<c023413c> blkdev_driver_ioctl+0x42/0x44  <c02343a9> 
blkdev_ioctl+0x242/0x759
<c0166446> iget5_locked+0xe6/0x14e  <c0157668> bdev_test+0x0/0xc
<c0156aec> bdget+0xd6/0xde  <c0157674> bdev_set+0x0/0x8
<c0157492> blkdev_open+0x0/0x4c  <c01574ae> blkdev_open+0x1c/0x4c
<c014fb0c> __dentry_open+0xe4/0x1a3  <c014fc58> nameidata_to_filp+0x31/0x3a
<c014fc9a> do_filp_open+0x39/0x40  <c0156907> block_ioctl+0x18/0x1d
<c01568ef> block_ioctl+0x0/0x1d  <c015f5c9> do_ioctl+0x19/0x55
<c015f657> vfs_ioctl+0x52/0x247  <c015f880> sys_ioctl+0x34/0x50
<c0468027> syscall_call+0x7/0xb  <c046007b> sctp_getsockopt+0x7a0/0x147f
BUG: warning at fs/block_dev.c:944/do_open()
<c015748d> do_open+0x2e8/0x2ed  <c02ec589> ide_do_request+0x6c2/0x894
<c0237dd7> cfq_insert_request+0x6b/0x4a5  <c0230fa3> 
blk_remove_plug+0x25/0x5c
<c023100b> __generic_unplug_device+0x1f/0x25  <c0233e63> 
__make_request+0x108/0x395
<c0166446> iget5_locked+0xe6/0x14e  <c0157668> bdev_test+0x0/0xc
<c0156aec> bdget+0xd6/0xde  <c015763d> open_partition_by_devnum+0x69/0x7c
<c034dcfa> md_import_device+0x78/0x24b  <c023aedc> 
kobject_register+0x30/0x35
<c034cdf3> md_probe+0x12e/0x16b  <c03506ef> md_ioctl+0x27a/0x154b
<c034f857> md_open+0x48/0x4f  <c01573d2> do_open+0x22d/0x2ed
<c015c659> do_lookup+0x47/0x126  <c02a84f4> scrup+0xca/0xd4
<c016422f> dput+0xba/0x177  <c015cee9> __link_path_walk+0x7b1/0xc37
<c0252d37> bit_cursor+0x34d/0x5c3  <c01674a0> mntput_no_expire+0x13/0x52
<c0350475> md_ioctl+0x0/0x154b  <c023413c> blkdev_driver_ioctl+0x42/0x44
<c02343a9> blkdev_ioctl+0x242/0x759  <c0166446> iget5_locked+0xe6/0x14e
<c0157668> bdev_test+0x0/0xc  <c0156aec> bdget+0xd6/0xde
<c0157674> bdev_set+0x0/0x8  <c0157492> blkdev_open+0x0/0x4c
<c01574ae> blkdev_open+0x1c/0x4c  <c014fb0c> __dentry_open+0xe4/0x1a3
<c014fc58> nameidata_to_filp+0x31/0x3a  <c014fc9a> do_filp_open+0x39/0x40
<c0156907> block_ioctl+0x18/0x1d  <c01568ef> block_ioctl+0x0/0x1d
<c015f5c9> do_ioctl+0x19/0x55  <c015f657> vfs_ioctl+0x52/0x247
<c015f880> sys_ioctl+0x34/0x50  <c0468027> syscall_call+0x7/0xb
<c046007b> sctp_getsockopt+0x7a0/0x147f
md: autorun ...
md: considering hdb1 ...
.
.
.
hda: cache flushes not supported
hdb: cache flushes not supported
md: md_d0 stopped.
BUG: warning at fs/block_dev.c:944/do_open()
<c015748d> do_open+0x2e8/0x2ed  <c0238ea7> cfq_set_request+0x1d3/0x33f
<c0237dd7> cfq_insert_request+0x6b/0x4a5  <c022ffe4> elv_insert+0xdd/0x142
<c0154561> bio_phys_segments+0x14/0x16  <c0233e49> 
__make_request+0xee/0x395
<c016598b> find_inode+0x37/0x61  <c01661ce> ifind+0x29/0x61
<c01663bc> iget5_locked+0x5c/0x14e  <c0157668> bdev_test+0x0/0xc
<c0156a4c> bdget+0x36/0xde  <c015763d> open_partition_by_devnum+0x69/0x7c
<c034dcfa> md_import_device+0x78/0x24b  <c04664f8> schedule+0x2a2/0x5ef
<c01663bc> iget5_locked+0x5c/0x14e  <c035162f> md_ioctl+0x11ba/0x154b
<c0231000> __generic_unplug_device+0x14/0x25  <c0466853> 
io_schedule+0xe/0x16
<c0132a5f> find_get_pages_tag+0x27/0x66  <c0350475> md_ioctl+0x0/0x154b
<c023413c> blkdev_driver_ioctl+0x42/0x44  <c02343a9> 
blkdev_ioctl+0x242/0x759
<c013a42a> release_pages+0x2a/0x175  <c0132990> find_get_pages+0x19/0x49
<c0467089> mutex_lock+0xb/0x1a  <c02f46dc> ide_disk_put+0x1c/0x27
<c02f52bd> idedisk_release+0x45/0xc6  <c0165383> iput+0x35/0x62
<c0156f14> __blkdev_put+0x77/0x187  <c0165383> iput+0x35/0x62
<c0156f84> __blkdev_put+0xe7/0x187  <c0156907> block_ioctl+0x18/0x1d
<c01568ef> block_ioctl+0x0/0x1d  <c015f5c9> do_ioctl+0x19/0x55
<c015f657> vfs_ioctl+0x52/0x247  <c015f880> sys_ioctl+0x34/0x50
<c0467fbd> sysenter_past_esp+0x56/0x79
md: bind<hdb2>
raid1: raid set md_d0 active with 1 out of 2 mirrors

The machine seems to work fine.  (That degraded array
is degraded on purpose.)

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