2.6.15-mm2: Kernel BUG at fs/sysfs/symlink.c:87
Hi all,
I got the following when I tried to create raid1 over nine sata disks. One on onboard NVidia and eight on a Promise SX8.
Both the motherboard (Tyan K8SE) and the Promise SX8 are flashed with the latest bios, FWIW.
NCQ is disabled in the Promise controller (although the current firmware should be able to handle it).
OS is up to date Debian Sid in an initramfs which I use to install the system.
Please let me know if this is of any use, if so, whom I should contact, and if more info is needed.
Btw, this system is fresh out of the box, so hardware can be a problem too.
I'll try plain 2.6.15 now.
Kind regards, Sander
PS, I had to create the /dev/sx8 directory and the nodes (0, 1, 0p1, 1p1, etc) for the disks and partitions by hand (mknod). Is there a more easy way? It seems MAKEDEV in Debian Sid doesn't know how to. Do I miss something? (likely, but I did my google homework ;-)
[ 1059.591796] Kernel BUG at fs/sysfs/symlink.c:87
[ 1059.591820] invalid opcode: 0000 [1] SMP
[ 1059.591845] CPU 1
[ 1059.591865] Modules linked in:
[ 1059.591888] Pid: 982, comm: mdadm Not tainted 2.6.15-mm2 #1
[ 1059.591915] RIP: 0010:[<ffffffff80191903>] <ffffffff80191903>{sysfs_create_link+37}
[ 1059.591948] RSP: 0018:ffff810138d8fbf8 EFLAGS: 00010202
[ 1059.591990] RAX: ffff810138845300 RBX: ffff81013fc74618 RCX: 0000000000000034
[ 1059.592020] RDX: ffffffff80401c01 RSI: ffff810138845350 RDI: ffff81013fc74618
[ 1059.592050] RBP: ffffffff80401ca7 R08: 000000000000000b R09: 0000000807aa308c
[ 1059.592080] R10: ffff81013fc74618 R11: fffffffffffffff3 R12: 0000000000000000
[ 1059.592111] R13: ffff810138d8fc58 R14: ffff81013af70018 R15: ffff810138845350
[ 1059.592142] FS: 00002aaaaae01ae0(0000) GS:ffffffff805c5080(0000) knlGS:0000000000000000
[ 1059.592187] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 1059.592214] CR2: 00007fffffcf7dd8 CR3: 000000013828c000 CR4: 00000000000006e0
[ 1059.592244] Process mdadm (pid: 982, threadinfo ffff810138d8e000, task ffff810138cb30d0)
[ 1059.592288] Stack: 0000000000000034 ffff81013fc74618 ffff81013fc745c0 ffff81013af70000
[ 1059.592320] ffff810138d8fc58 ffff81013af70018 ffff81013fc42a40 ffffffff8033d5cd
[ 1059.592365] ffff810138d8fd18 0000000000000000
[ 1059.592390] Call Trace:<ffffffff8033d5cd>{bind_rdev_to_array+493} <ffffffff8033ea27>{md_ioctl+3652}
[ 1059.592440] <ffffffff8027d86a>{blkdev_driver_ioctl+91} <ffffffff8027df81>{blkdev_ioctl+1699}
[ 1059.592492] <ffffffff801475e9>{invalidate_mapping_pages+183} <ffffffff801198f5>{flat_send_IPI_allbutself+20}
[ 1059.592547] <ffffffff80116ded>{__smp_call_function+101} <ffffffff80164133>{invalidate_bh_lru+0}
[ 1059.592596] <ffffffff80168412>{block_ioctl+25} <ffffffff80171f59>{do_ioctl+33}
[ 1059.592645] <ffffffff801721f8>{vfs_ioctl+597} <ffffffff8017224d>{sys_ioctl+60}
[ 1059.592691] <ffffffff8010d7da>{system_call+126}
[ 1059.592733]
[ 1059.592734] Code: 0f 0b 68 73 6c 3f 80 c2 57 00 49 8b 7c 24 10 41 bd f4 ff ff
[ 1059.592811] RIP <ffffffff80191903>{sysfs_create_link+37} RSP <ffff810138d8fbf8>
[ 1059.592996]Signature · 2 lines
Humilis IT Services and Solutions http://www.humilis.net
Kernel BUG at fs/sysfs/symlink.c:87 happens with 2.6.15 too
Sander wrote (ao):
# I'll try plain 2.6.15 now.
The same thing happens. I try to create a raid1 over nine disks (eight on Promise SX8):
mdadm -C -l1 -n9 /dev/md0 /dev/sda1 /dev/sx8/0p1 /dev/sx8/1p1 /dev/sx8/2p1 /dev/sx8/3p1 /dev/sx8/4p1 /dev/sx8/5p1 /dev/sx8/6p1 /dev/sx8/7p1
It Segfaults and gives this:
Difference with 2.6.15-mm2 is that 2.6.15 keeps on running.
I'll enable some debug options now. Any advice on what I can try besides this? Am I doing something wrong?
[ 2281.547827] md: bind<sda1>
[ 2281.547877] md: bind<sx8/0p1>
[ 2281.547902] ----------- [cut here ] --------- [please bite here ] ---------
[ 2281.547931] Kernel BUG at fs/sysfs/symlink.c:87
[ 2281.547955] invalid operand: 0000 [1] SMP
[ 2281.547980] CPU 1
[ 2281.548000] Modules linked in:
[ 2281.548023] Pid: 874, comm: mdadm Not tainted 2.6.15 #1
[ 2281.548049] RIP: 0010:[<ffffffff8018fb07>] <ffffffff8018fb07>{sysfs_create_link+37}
[ 2281.548082] RSP: 0018:ffff81013fcc7bf8 EFLAGS: 00010202
[ 2281.548123] RAX: ffff81013fc9e700 RBX: ffff81013a55add8 RCX: 0000000000000034
[ 2281.548153] RDX: ffffffff80370001 RSI: ffff81013fc9e750 RDI: ffff81013a55add8
[ 2281.548183] RBP: ffffffff80370043 R08: 000000000000000b R09: 0000000807aa308c
[ 2281.548214] R10: ffff81013a55add8 R11: ffff81013a55add8 R12: 0000000000000000
[ 2281.548244] R13: ffff81013fcc7c58 R14: ffff81013a481c18 R15: ffff81013fc9e750
[ 2281.548275] FS: 00002aaaaae01ae0(0000) GS:ffffffff804f7880(0000) knlGS:0000000000000000
[ 2281.548319] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 2281.548346] CR2: 00007fffffdd29c8 CR3: 000000013a4c8000 CR4: 00000000000006e0
[ 2281.548376] Process mdadm (pid: 874, threadinfo ffff81013fcc6000, task ffff81013b2948e0)
[ 2281.548420] Stack: 0000000000000034 ffff81013a55add8 ffff81013a55ad80 ffff81013a481c00
[ 2281.548452] ffff81013fcc7c58 ffff81013a481c18 ffff81013fc42440 ffffffff802ba1a3
[ 2281.548499] 0000000000000000 0000808000000000
[ 2281.548524] Call Trace:<ffffffff802ba1a3>{bind_rdev_to_array+441} <ffffffff802bb2ef>{md_ioctl+3217}
[ 2281.548577] <ffffffff801ec3ee>{blkdev_driver_ioctl+91} <ffffffff801ecae1>{blkdev_ioctl+1663}
[ 2281.548629] <ffffffff8014914e>{pagevec_lookup+23} <ffffffff80149ceb>{invalidate_mapping_pages+188}
[ 2281.548681] <ffffffff80119664>{flat_send_IPI_allbutself+20} <ffffffff80116c50>{__smp_call_function+101}
[ 2281.548733] <ffffffff80161e34>{invalidate_bh_lru+0} <ffffffff80166287>{block_ioctl+25}
[ 2281.548784] <ffffffff80170179>{do_ioctl+33} <ffffffff80170431>{vfs_ioctl+622}
[ 2281.548833] <ffffffff80170486>{sys_ioctl+60} <ffffffff8010d6ce>{system_call+126}
[ 2281.548882]
[ 2281.548916]
[ 2281.548917] Code: 0f 0b 68 ea ae 36 80 c2 57 00 49 8b 44 24 10 48 8d b8 c0 00
[ 2281.548994] RIP <ffffffff8018fb07>{sysfs_create_link+37} RSP <ffff81013fcc7bf8>
[ 2281.549178]2.6.15-mm2:
# [ 1059.591796] Kernel BUG at fs/sysfs/symlink.c:87
# [ 1059.591820] invalid opcode: 0000 [1] SMP
# [ 1059.591845] CPU 1
# [ 1059.591865] Modules linked in:
# [ 1059.591888] Pid: 982, comm: mdadm Not tainted 2.6.15-mm2 #1
# [ 1059.591915] RIP: 0010:[<ffffffff80191903>] <ffffffff80191903>{sysfs_create_link+37} # [ 1059.591948] RSP: 0018:ffff810138d8fbf8 EFLAGS: 00010202
# [ 1059.591990] RAX: ffff810138845300 RBX: ffff81013fc74618 RCX: 0000000000000034
# [ 1059.592020] RDX: ffffffff80401c01 RSI: ffff810138845350 RDI: ffff81013fc74618
# [ 1059.592050] RBP: ffffffff80401ca7 R08: 000000000000000b R09: 0000000807aa308c
# [ 1059.592080] R10: ffff81013fc74618 R11: fffffffffffffff3 R12: 0000000000000000
# [ 1059.592111] R13: ffff810138d8fc58 R14: ffff81013af70018 R15: ffff810138845350
# [ 1059.592142] FS: 00002aaaaae01ae0(0000) GS:ffffffff805c5080(0000) knlGS:0000000000000000 # [ 1059.592187] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
# [ 1059.592214] CR2: 00007fffffcf7dd8 CR3: 000000013828c000 CR4: 00000000000006e0
# [ 1059.592244] Process mdadm (pid: 982, threadinfo ffff810138d8e000, task ffff810138cb30d0) # [ 1059.592288] Stack: 0000000000000034 ffff81013fc74618 ffff81013fc745c0 ffff81013af70000
# [ 1059.592320] ffff810138d8fc58 ffff81013af70018 ffff81013fc42a40 ffffffff8033d5cd
# [ 1059.592365] ffff810138d8fd18 0000000000000000
# [ 1059.592390] Call Trace:<ffffffff8033d5cd>{bind_rdev_to_array+493} <ffffffff8033ea27>{md_ioctl+3652}
# [ 1059.592440] <ffffffff8027d86a>{blkdev_driver_ioctl+91} <ffffffff8027df81>{blkdev_ioctl+1699}
# [ 1059.592492] <ffffffff801475e9>{invalidate_mapping_pages+183} <ffffffff801198f5>{flat_send_IPI_allbutself+20}
# [ 1059.592547] <ffffffff80116ded>{__smp_call_function+101} <ffffffff80164133>{invalidate_bh_lru+0}
# [ 1059.592596] <ffffffff80168412>{block_ioctl+25} <ffffffff80171f59>{do_ioctl+33}
# [ 1059.592645] <ffffffff801721f8>{vfs_ioctl+597} <ffffffff8017224d>{sys_ioctl+60}
# [ 1059.592691] <ffffffff8010d7da>{system_call+126}# [ 1059.592733]
# [ 1059.592734] Code: 0f 0b 68 73 6c 3f 80 c2 57 00 49 8b 7c 24 10 41 bd f4 ff ff # [ 1059.592811] RIP <ffffffff80191903>{sysfs_create_link+37} RSP <ffff810138d8fbf8> # [ 1059.592996]
Re: Kernel BUG at fs/sysfs/symlink.c:87 happens with 2.6.15 too
Quoted 6 lines · On Wednesday January 11, [email protected] wrote:
Sander wrote (ao):
# I'll try plain 2.6.15 now.The same thing happens. I try to create a raid1 over nine disks (eight on Promise SX8):
Quoted 8 lines
mdadm -C -l1 -n9
/dev/md0/dev/sda1/dev/sx8/0p1/dev/sx8/1p1/dev/sx8/2p1/dev/sx8/3p1/dev/sx8/4p1/dev/sx8/5p1/dev/sx8/6p1/dev/sx8/7p1It Segfaults and gives this:
Difference with 2.6.15-mm2 is that 2.6.15 keeps on running.
I'll enable some debug options now. Any advice on what I can try besides this? Am I doing something wrong?
No, you aren't doing anything wrong.
It appears that the 'kobject_add' call in bind_rdev_to_array is not setting the ->dentry field for the kobject.
i.e. it is failing to create
/sys/block/md0/md/dev-XXXX
where 'XXXX' is the internal name for the component device, as returned by 'bdevname'. The most likely reason I can think of for this is that the name is already in use. i.e. different sx devices get the same name.... let's see....
Close, but not quite.
sx8 creates device names with a '/' in them as we can see from:
Quoted 3 lines
[ 2281.547827] md: bind<sda1> [ 2281.547877] md: bind<sx8/0p1>
^
That's unfortunate. We will have to remove them. You will find that the names in /sys/block look like
/sys/block/sx8!0
The following patch should achieve the same substitution for md.
NeilBrown
--------------------------
Remove slashed from disk names when creation dev names in sysfs
e.g. The sx8 driver uses names like sx8/0.
This would make a md component dev name like
/sys/block/md0/md/dev-sx8/0
which is no allowed. So we change the '/' to '!' just like fs/partitions/check.c(register_disk) does.
### Diffstat output
./drivers/md/md.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/md/md.c renamed from drivers/md/md.c~current~ +3 -0
@@ -1238,6 +1238,7 @@ static int bind_rdev_to_array(mdk_rdev_tmdk_rdev_t *same_pdev;char b[BDEVNAME_SIZE ], b2[BDEVNAME_SIZE ];struct kobject *ko; +char *s;if (rdev->mddev) {MD_BUG ();
@@ -1277,6 +1278,8 @@ static int bind_rdev_to_array(mdk_rdev_tbdevname (rdev->bdev,b);if (kobject_set_name (&rdev->kobj,"dev-%s" , b) <0 )return -ENOMEM ; +while ( (s=strchr (rdev->kobj.k_name,'/' )) !=NULL ) + *s ='!' ;list_add (&rdev->same_set, &mddev->disks); rdev->mddev = mddev;
SOLVED: Re: Kernel BUG at fs/sysfs/symlink.c:87 happens with 2.6.15 too
Neil Brown wrote (ao):
Quoted 4 lines
sx8 creates device names with a '/' in them as we can see from:
Quoted 3 lines
[ 2281.547827] md: bind<sda1> [ 2281.547877] md: bind<sx8/0p1>
Quoted 8 lines
^That's unfortunate. We will have to remove them. You will find that the names in
/sys/blocklook like/sys/block/sx8!0The following patch should achieve the same substitution for md.
Thanks Neil!
2.6.15 with your patch works (lets me create raid on the Promise SX8 SATA controller).
I don't quite understand why I'm the first one to hit this (the Promise SX8 is good value for money and with decent Linux support?), but it seems your patch is needed in both vanilla and -mm.
Thanks again, Sander
Quoted 35 lines
--------------------------
Remove slashed from disk names when creation dev names in sysfse.g. The sx8 driver uses names like sx8/0.
This would make a md component dev name like
/sys/block/md0/md/dev-sx8/0
which is no allowed. So we change the '/' to '!' just like fs/partitions/check.c(register_disk) does.Signed-off-by Neil Brown <[email protected]>### Diffstat output
./drivers/md/md.c | 3 +++ 1 file changed, 3 insertions(+)diff
./drivers/md/md.c~current~./drivers/md/md.cdrivers/md/md.c renamed from drivers/md/md.c~current~ +3
-0@@ -1238,6 +1238,7 @@ static int bind_rdev_to_array(mdk_rdev_tmdk_rdev_t *same_pdev;char b[BDEVNAME_SIZE ], b2[BDEVNAME_SIZE ];struct kobject *ko; +char *s;if (rdev->mddev) {MD_BUG ();@@ -1277,6 +1278,8 @@ static int bind_rdev_to_array(mdk_rdev_tbdevname (rdev->bdev,b);if (kobject_set_name (&rdev->kobj,"dev-%s" , b) <0 )return -ENOMEM ; +while ( (s=strchr (rdev->kobj.k_name,'/' )) !=NULL ) + *s ='!' ;list_add (&rdev->same_set, &mddev->disks); rdev->mddev = mddev;
Signature · 2 lines
Humilis IT Services and Solutions http://www.humilis.net