Re: Issues with 20080128/Ubuntu Hardy (8.04)
Julian Andres Klode wrote: > Am Donnerstag, den 15.05.2008, 16:30 +0200 schrieb "Jørgen P. Tjernø": >> How is a user supposed to figure out that this is the issue >> with prebuilt packages? :( > The config file should be available via /sys/fs/aufs/config. Ah, neat. >> (This is a binary package from ubuntu) > It's just not good for a distribution to enable options that damage the > performance and are only useful to a small group of people. Does it really damage the performance to _enable_ the option if udba=none or udba=reval is the mount option? > BTW, if you want to use a newer version of aufs, simply install the > aufs-source package from Debian unstable or Ubuntu Intrepid and run m-a > a-i aufs (you may also need to install module-assistant). But you also > have to remove the prebuilt module in /lib/modules/$(uname > -r)/ubuntu/fs/aufs/aufs.ko. Thanks. How can I enable inotify in this package? I tried downloading it and unpacking it, modifying the makefile to have CONFIG_AUFS_HINOTIFY = y, then doing m-a build -fO aufs, and it error'd out with this: /usr/src/modules/aufs/hinotify.c:32: error: redefinition of ‘au_hin_alloc’ /usr/src/modules/aufs/hinode.h:109: error: previous definition of ‘au_hin_alloc’ was here Plus a lot more. Another way I should be doing this? Kindest regards, Jørgen P. Tjernø. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Issues with 20080128/Ubuntu Hardy (8.04)
[EMAIL PROTECTED] wrote:\ > Maybe you are right. But it might be a nature of mount(2) systemcall. > When the systemcall meets an illegal option, it just returns an error > with EINVAL error code. It should be mount(8) command that prints the > error message. But there is no way for the systemcall to tell userspace > which option is wrong. > I guess every filesystem returns EINVAL simply in this situation and > mount(8) prints "mount: wrong fs type, bad option, ..." Yes, but often dmesg has information about the error details - could perhaps printk (or what it's called?) be used in aufs to show information about what error occured? :) Kindest regards, Jørgen P. Tjernø. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Issues with 20080128/Ubuntu Hardy (8.04)
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> [EMAIL PROTECTED]:~$ sudo mount -t aufs -o = >> >> br:/vault/disk2=3Drw:/vault/disk5=3Drw:/vault/disk4=3Drw:/vault/disk3=3Drw:= >> /vault/disk6=3Drw:/vault/disk1=3Drw,udba=3Dinotify = >> >> none /storage >> mount: wrong fs type, bad option, bad superblock on none, >> missing codepage or helper program, or other error >> (for several filesystems (e.g. nfs, cifs) you might >> need a /sbin/mount. helper program) >> In some cases useful info is found in syslog - try >> dmesg | tail or so > > Is CONFIG_AUFS_HINOTIFY enabled? I guess not, if I removed udba=inotify it worked. The error messages were not dechiperable at _ALL_, I think this is something that should be changed? How is a user supposed to figure out that this is the issue with prebuilt packages? :( (This is a binary package from ubuntu) -- Kindest regards, Jørgen P. Tjernø - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Issues with 20080128/Ubuntu Hardy (8.04)
I upgraded from edgy (where I used 2.6.20.3 and built aufs from CVS) to hardy (or rather edgy->feisty->gutsy->hardy), and I'm now using the stock ubuntu 2.6.24-16-generic and their build of aufs. My previous mount-command does no longer work, it says "failure bad /proc/mounts 1". Any tips? Here is my command, the output, dmesg|tail -n 1 and /proc/mounts: [EMAIL PROTECTED]:~$ sudo mount -t aufs -o br:/vault/disk2=rw:/vault/disk5=rw:/vault/disk4=rw:/vault/disk3=rw:/vault/disk6=rw:/vault/disk1=rw,udba=inotify none /storage mount: wrong fs type, bad option, bad superblock on none, missing codepage or helper program, or other error (for several filesystems (e.g. nfs, cifs) you might need a /sbin/mount. helper program) In some cases useful info is found in syslog - try dmesg | tail or so failure bad /proc/mounts 1 [EMAIL PROTECTED]:~$ dmesg|tail -n 1 [ 71.800060] Bluetooth: RFCOMM ver 1.8 [EMAIL PROTECTED]:~$ cat /proc/mounts rootfs / rootfs rw 0 0 none /sys sysfs rw,nosuid,nodev,noexec 0 0 none /proc proc rw,nosuid,nodev,noexec 0 0 udev /dev tmpfs rw,relatime 0 0 fusectl /sys/fs/fuse/connections fusectl rw,relatime 0 0 /dev/disk/by-uuid/5c6fcc95-c3c3-4526-a868-33cb76d79471 / ext3 rw,relatime,errors=remount-ro,data=ordered 0 0 /dev/disk/by-uuid/5c6fcc95-c3c3-4526-a868-33cb76d79471 /dev/.static/dev ext3 rw,relatime,errors=remount-ro,data=ordered 0 0 tmpfs /var/run tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /var/lock tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /lib/modules/2.6.24-16-generic/volatile tmpfs rw,relatime 0 0 tmpfs /dev/shm tmpfs rw,relatime 0 0 devpts /dev/pts devpts rw,relatime 0 0 tmpfs /var/run tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /var/lock tmpfs rw,nosuid,nodev,noexec 0 0 /dev/sda1 /vault/disk1 xfs rw,relatime,ikeep,noquota 0 0 /dev/sdc1 /vault/disk2 xfs rw,relatime,ikeep,noquota 0 0 /dev/sdb1 /vault/disk3 xfs rw,relatime,ikeep,noquota 0 0 /dev/sdd1 /vault/disk4 xfs rw,relatime,ikeep,noquota 0 0 /dev/sde1 /vault/disk5 xfs rw,relatime,ikeep,noquota 0 0 /dev/sdf1 /vault/disk6 xfs rw,relatime,ikeep,noquota 0 0 /dev/sdg3 /vault/misc ext3 rw,relatime,data=ordered 0 0 rpc_pipefs /var/lib/nfs/rpc_pipefs rpc_pipefs rw,relatime 0 0 nfsd /proc/fs/nfsd nfsd rw,relatime 0 0 [EMAIL PROTECTED]:~$ -- Kindest regards, Jørgen P. Tjernø. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
aufs & linux 2.6.17 (Ubuntu Edgy)
Does aufs support Ubuntu Edgy's 2.6.17? If not, what version of Ubuntu kernels can I run to get proper aufs support? :-) Kindest regards, Jørgen P. Tjernø. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Slab corruption
(Sorry, mailed it directly to sfjro first) [EMAIL PROTECTED] wrote: > Do you mean you have the problem still? > Does your system crash after a while? My system was running totally stable for a long time, then this happened. I have not had it crash after this slab corruption thing - it's stable now too. -- Kindest regards, Jørgen P. Tjernø - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
Re: Slab corruption
For what it's worth; I've upgraded to 20080421 now, and I've never gotten this problem earlier. :-) Kindest regards, Jørgen P. Tjernø. - This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
Slab corruption
I did a large number of mkdir -p's directly on a branch, so that mkdir created up to three levels of directories on the branch (I.e. mkdir -p /vault/disk1/foo/bar/baz when /vault/disk1/foo did not exist). dmesg gave me all this output, then the box froze (I rebooted). Kernel version: 2.6.20.3 GNU/Linux AUFS version: 20080121 Default local.mk, except for CONFIG_AUFS_HINOTIFY = y CONFIG_AUFS_EXPORT = y /proc/mounts says: none /storage aufs rw,xino=/vault/disk2/.aufs.xino,udba=inotify,br:/vault/disk1=rw:/vault/disk2=rw:/vault/disk5=rw:/vault/disk4=rw:/vault/disk3=rw:/vault/disk6=rw 0 0 But it's incorrect, as it has been remounted using: mount -lo remount,del:/vault/disk1,prepend:/vault/disk1=rw /storage (It's done automatically, since it has more free space) dmesg output attached in log.txt. :) I haven't had any other problems though. :-) Kindest regards, Jørgen P. Tjernø. [878571.742915] Slab corruption: start=c69b34b4, len=32 [878571.745786] Redzone: 0x5a2cf071/0x5a2cf071. [878571.748791] Last user: [](au_kzrealloc+0xe3/0x150 [aufs]) [878571.751886] 000: 6b 6b 6b 6b 6b 6b 6b 6b 00 04 00 00 00 e0 5d d9 [878571.755137] Prev obj: start=c69b3488, len=32 [878571.758159] Redzone: 0x170fc2a5/0x170fc2a5. [878571.761050] Last user: [](cache_alloc_refill+0x564/0x660) [878571.764115] 000: 30 34 9b c6 30 df 8d c1 00 00 00 00 00 a0 e4 c2 [878571.767371] 010: 01 00 00 00 ff ff ff ff 00 00 5a 5a fe ff ff ff [878571.770611] Next obj: start=c69b34e0, len=32 [878571.773635] Redzone: 0x170fc2a5/0x170fc2a5. [878571.776645] Last user: [](au_kzrealloc+0x99/0x150 [aufs]) [878571.779599] 000: 00 04 00 00 00 40 1d cd 00 00 00 00 00 00 00 00 [878571.782847] 010: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a a5 [878571.786302] Slab corruption: start=c69b34b4, len=32 [878571.789307] Redzone: 0x5a2cf071/0x5a2cf071. [878571.792313] Last user: [](au_kzrealloc+0xe3/0x150 [aufs]) [878571.795251] 010: 00 04 00 00 00 c0 b5 f2 6b 6b 6b 6b 6b 6b 6b a5 [878571.798500] Prev obj: start=c69b3488, len=32 [878571.801519] Redzone: 0x170fc2a5/0x170fc2a5. [878571.804535] Last user: [](cache_alloc_refill+0x564/0x660) [878571.807609] 000: 30 34 9b c6 30 df 8d c1 00 00 00 00 00 a0 e4 c2 [878571.810773] 010: 01 00 00 00 ff ff ff ff 00 00 5a 5a fe ff ff ff [878571.813993] Next obj: start=c69b34e0, len=32 [878571.817007] Redzone: 0x170fc2a5/0x170fc2a5. [878571.820018] Last user: [](au_kzrealloc+0x99/0x150 [aufs]) [878571.823104] 000: 00 04 00 00 00 40 1d cd 00 00 00 00 00 00 00 00 [878571.826404] 010: 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a 5a a5 [878571.829685] slab error in verify_redzone_free(): cache `size-32': double free detected [878571.835666] [] show_trace_log_lvl+0x1a/0x30 [878571.835671] [] show_trace+0x12/0x20 [878571.835674] [] dump_stack+0x16/0x20 [878571.835677] [] __slab_error+0x26/0x30 [878571.835680] [] cache_free_debugcheck+0x19c/0x1d0 [878571.835684] [] kfree+0x55/0xb0 [878571.835686] [] au_kzrealloc+0xe3/0x150 [aufs] [878571.835704] [] au_dpages_append+0x71/0x100 [aufs] [878571.835718] [] au_dcsub_pages+0xdf/0x1c0 [aufs] [878571.835732] [] hin_digen_tree+0x48/0x230 [aufs] [878571.835748] [] hin_job+0x550/0x690 [aufs] [878571.835764] [] postproc+0x200/0x660 [aufs] [878571.835780] [] wkq_func+0x19/0xf0 [aufs] [878571.835794] [] run_workqueue+0x94/0x140 [878571.835797] [] worker_thread+0x138/0x160 [878571.835800] [] kthread+0xb7/0xe0 [878571.835804] [] kernel_thread_helper+0x7/0x14 [878571.835807] === [878571.835809] c69b34b0: redzone 1:0x5a2cf071, redzone 2:0x5a2cf071. [878571.839023] slab error in verify_redzone_free(): cache `size-32': memory outside object was overwritten [878571.844881] [] show_trace_log_lvl+0x1a/0x30 [878571.844884] [] show_trace+0x12/0x20 [878571.844887] [] dump_stack+0x16/0x20 [878571.844890] [] __slab_error+0x26/0x30 [878571.844893] [] cache_free_debugcheck+0x123/0x1d0 [878571.844896] [] kfree+0x55/0xb0 [878571.844899] [] au_kzrealloc+0xe3/0x150 [aufs] [878571.844916] [] au_dpages_append+0x71/0x100 [aufs] [878571.844930] [] au_dcsub_pages+0xdf/0x1c0 [aufs] [878571.844944] [] hin_digen_tree+0x48/0x230 [aufs] [878571.844960] [] hin_job+0x550/0x690 [aufs] [878571.844976] [] postproc+0x200/0x660 [aufs] [878571.844992] [] wkq_func+0x19/0xf0 [aufs] [878571.845006] [] run_workqueue+0x94/0x140 [878571.845008] [] worker_thread+0x138/0x160 [878571.845011] [] kthread+0xb7/0xe0 [878571.845014] [] kernel_thread_helper+0x7/0x14 [878571.845017] === [878571.845019] c69b34b0: redzone 1:0x5a2cf071, redzone 2:0x400. [878571.848204] slab error in verify_redzone_free(): cache `size-32': double free detected [878571.854189] [] show_trace_log_lvl+0x1a/0x30 [878571.854192] [] show_trace+0x12/0x20 [878571.854195] [] dump_stack+0x16/0x20 [878571.854198] [] __slab_error+0x26/0x30 [878571.854200] [] cache_free_debugcheck+0x19c/0x1d0 [878571.854204] [] kfree+0x55/0xb0 [878571.854206] [] au_kzreall
Re: Kernel BUG with du -s & direct branch modification (with udba=inotify)
[EMAIL PROTECTED] wrote: > This line shows that aufs found a problem about aufs inode. I guess it > is caused by direct (bypassing aufs) modifications on branches. Although > I don't make the cause clear, I want you to try latest aufs and see what > will happen, since udba=notify feature is massivly refined after > 20070917. Huuhm. I'm terribly sorry, I _thought_ I was running the latest version (I had done cvs up and such to get the latest) - but it seems I did cvs update -D once to get a version other than the latest (you did some changes that I wanted to wait for others to test). I didn't know using -D set a "sticky date" on all files, and you had to cvs update -A to get the newest. :-) I'll upgrade at once! Thanks. :) Kindest regards, Jørgen. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Kernel BUG / umount -lf + mount --bind
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> I have a aufs mount at /storage, and a few mount --binds pointing to >> subdirectories of /storage. I did a umount -lf /storage, which happened >> without errors, but then did umount -lf >> /some/mount/--bind/pointing/to/storage, and the following kernel BUG >> appeared. :-) > > Let me make sure first since I cound not reproduce the problem. > > What you did was, > - on a plain linux-2.6.20.3 > - aufs 20070917 > - mount -t aufs ... /storage > - mount --bind /storage /some/mount/--bind/pointing/to/storage > - umount /storage > - umount /some/mount/--bind/pointing/to/storage > and the bug appeared. > > Am I right? Sounds about right, except that I used umount -lf because I think some files were still open. :-) Kindest regards, Jørgen. - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Kernel BUG with du -s & direct branch modification (with udba=inotify)
Mount options: none on /storage type aufs (rw,xino=/vault/disk6/.aufs.xino,udba=inotify,br:/vault/disk6=rw:/vault/disk5=rw:/vault/disk4=rw:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw) AUFS version: aufs 20070917 (Ps. the above goes for the other kernel BUG I posted) I did the following: cd /storage mv foo\ bar foo.bar ^C mv foo.bar/* foo\ bar/ && rmdir foo.bar cd /vault/disk5 mv foo\ bar/ foo.bar cd ../disk6 mv foo\ bar/ foo.bar cd /storage du -s * And then got the following kernel BUG: [ 4593.717866] [ cut here ] [ 4593.720868] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/iinfo.c:71! [ 4593.723914] invalid opcode: [#1] [ 4593.726757] SMP [ 4593.729738] Modules linked in: bnep rfcomm l2cap aufs sata_via sata_promise skge [ 4593.735822] CPU:0 [ 4593.735823] EIP:0060:[]Not tainted VLI [ 4593.735824] EFLAGS: 00010206 (2.6.20.3 #9) [ 4593.744712] EIP is at au_h_iptr_i+0x46/0x80 [aufs] [ 4593.747711] eax: 0001 ebx: c4e74af8 ecx: 0001 edx: 0004 [ 4593.750774] esi: 0004 edi: c4e74af8 ebp: c645ddd8 esp: c645ddd0 [ 4593.753835] ds: 007b es: 007b ss: 0068 [ 4593.756843] Process du (pid: 4874, ti=c645c000 task=f78b1560 task.ti=c645c000) [ 4593.759804] Stack: 0004 c645ddec f912ccd5 f5cd7b78 f49a8ea0 f4bedc30 c645de38 [ 4593.765917]f911e27d c645de1c c01615f5 f5cd7b78 c645de0c f910c07e f5cd7b78 c645de1c [ 4593.772033]f910c0e2 db40fedc 0004 00018800 cb8c5a68 f49a7838 f4bedc40 0004 [ 4593.778026] Call Trace: [ 4593.783892] [] show_trace_log_lvl+0x1a/0x30 [ 4593.786937] [] show_stack_log_lvl+0xb1/0xe0 [ 4593.789977] [] show_registers+0x1c1/0x2b0 [ 4593.792863] [] die+0x11c/0x270 [ 4593.795881] [] do_trap+0x82/0xb0 [ 4593.798901] [] do_invalid_op+0x97/0xb0 [ 4593.801934] [] error_code+0x7c/0x84 [ 4593.804963] [] test_ro+0x35/0x60 [aufs] [ 4593.807966] [] hidden_open+0x8d/0x1e0 [aufs] [ 4593.810925] [] do_open_dir+0x101/0x180 [aufs] [ 4593.813980] [] au_do_open+0x9e/0x510 [aufs] [ 4593.817033] [] aufs_open_dir+0xd/0x10 [aufs] [ 4593.820087] [] __dentry_open+0xad/0x1a0 [ 4593.823122] [] nameidata_to_filp+0x35/0x40 [ 4593.826006] [] do_filp_open+0x49/0x50 [ 4593.829035] [] do_sys_open+0x47/0xe0 [ 4593.832063] [] sys_open+0x1c/0x20 [ 4593.835085] [] sysenter_past_esp+0x5d/0x81 [ 4593.838122] === [ 4593.841103] Code: c7 85 c0 75 2e 89 d8 e8 79 fb ff ff 83 c0 08 e8 51 a8 00 c7 85 c0 75 17 89 f0 84 c0 78 0d 89 d8 e8 f0 fe ff ff 89 f2 38 c2 7e 0c <0f> 0b eb fe 0f 0b eb fe 0f 0b eb fe 89 d8 e8 47 fb ff ff 89 f1 [ 4593.850923] EIP: [] au_h_iptr_i+0x46/0x80 [aufs] SS:ESP 0068:c645ddd0 [ 4593.861744] - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Kernel BUG / umount -lf + mount --bind
I have a aufs mount at /storage, and a few mount --binds pointing to subdirectories of /storage. I did a umount -lf /storage, which happened without errors, but then did umount -lf /some/mount/--bind/pointing/to/storage, and the following kernel BUG appeared. :-) Kindest regards, Jørgen. [82345.550450] [ cut here ] [82345.553451] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/super.c:268! [82345.556498] invalid opcode: [#1] [82345.559495] SMP [82345.562321] Modules linked in: bnep rfcomm l2cap aufs skge sata_promise sata_via [82345.568408] CPU:0 [82345.568409] EIP:0060:[]Not tainted VLI [82345.568410] EFLAGS: 00010246 (2.6.20.3 #9) [82345.577352] EIP is at aufs_umount_begin+0x12a/0x1f0 [aufs] [82345.580312] eax: ebx: f746b9dc ecx: c18e9768 edx: f79d8454 [82345.583375] esi: f79d8c84 edi: f74c4550 ebp: cdd1df38 esp: cdd1df08 [82345.586436] ds: 007b es: 007b ss: 0068 [82345.589444] Process umount (pid: 28676, ti=cdd1c000 task=c3955a90 task.ti=cdd1c000) [82345.592558] Stack: 0001 cdd1df2c c016fc86 cdd1df4c ff9c c18de330 f79d8454 [82345.598514]f746b9dc 0002 cdd1df4c cdd1dfb0 c017c51e f74c4550 f74c4550 [82345.604625]f79d8c84 f7bbb34c f74c4550 c0169071 cdd1dfb0 c0169094 0001 0001 [82345.610671] Call Trace: [82345.616462] [] show_trace_log_lvl+0x1a/0x30 [82345.619505] [] show_stack_log_lvl+0xb1/0xe0 [82345.622548] [] show_registers+0x1c1/0x2b0 [82345.625583] [] die+0x11c/0x270 [82345.628446] [] do_trap+0x82/0xb0 [82345.631467] [] do_invalid_op+0x97/0xb0 [82345.634498] [] error_code+0x7c/0x84 [82345.637526] [] sys_umount+0xfe/0x250 [82345.640556] [] sysenter_past_esp+0x5d/0x81 [82345.643575] === [82345.646447] Code: 58 38 0f 84 be 00 00 00 8d 51 c8 8b 4a 38 0f 18 01 90 8d 42 38 39 45 e8 75 db 8b 55 ec 8b 82 80 00 00 00 85 c0 0f 85 6a ff ff ff <0f> 0b eb fe 89 f6 65 a1 08 00 00 00 c7 44 24 14 5c c0 13 f9 8b [82345.656364] EIP: [] aufs_umount_begin+0x12a/0x1f0 [aufs] SS:ESP 0068:cdd1df08 - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: mfsrr freespace question
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> Is "low" specified in bytes? Does "low" support units? (3T, 10G, 5M, >> 10k, etc) > > In bytes, and it doesn't support unit. > Try an arithmetic expansion of shell which is defined by POSIX. > For example, $((10 * 1024 * 1024)) for 10M. > (I will add this description in the manual) Thanks for clarifying, great. :-) >> Also, is it possible to create something mfsrr for mkdir(2)? The thing >> is that I'd like to keep all files in a dir on the same drive, which is >> best achieved by making sure the directory is on the right drive when >> you make it. It could also combine mfsrr & tdp, so that it only checks >> the branches that _have_ the dir you're mkdir(2)-ing inside - but finds >> the one with the most free space. :-) > > Unfortunately, my poor English doesn't allow me to understand the > difference between what you want and create=tdp. Well, the concept is a bit hard to explain ;-) > Do you want this policy? > - find the all writable branches where the parent dir exists. > - and then find the most free space among those branches. > But this policy will select a different branch after the first-selected > branch has less free space. Oh, right. Well, then I'd really like it if you could make it so mkdir(2) use the described policy for choosing what path to create a dir on. For my purposes, that'd suffice, and it would keep all writes in a dir on a specific branch. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFHRDd1UMzc1WGo4zgRAof6AJ0TGgdHeZFgYa0Sdm6LAm9GVwSozwCgg5G+ WHqErcMzJCBClIMUKbgLAao= =l4qq -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
mfsrr freespace question
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 create=mfsrr:low[:second] | most-free-space-round-robin:low[:second] Is "low" specified in bytes? Does "low" support units? (3T, 10G, 5M, 10k, etc) Also, is it possible to create something mfsrr for mkdir(2)? The thing is that I'd like to keep all files in a dir on the same drive, which is best achieved by making sure the directory is on the right drive when you make it. It could also combine mfsrr & tdp, so that it only checks the branches that _have_ the dir you're mkdir(2)-ing inside - but finds the one with the most free space. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFHQ/yWUMzc1WGo4zgRAvjvAJ9SUc2zsoV3ttN3ohsJxy4NHae9gwCfe+Rd HKodvamX2BRcUrbrarCTRwg= =4uS+ -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: BUG: kernel NULL pointer dereference
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >>> Well, it is interesting to me. >>> I will consider again later while it may not be within this year. >> Ok, thanks. :-) > > How do you think this approach? > - current xino files share one filename. the name is specified by xino > option or given by default. aufs creates the file per a branch and > unlink it. additonally aufs create one more xino file for an inode > number bitmap. > - introduce a new branch attribute, named brxino. its syntax is > ++brxino= > the specified xino file will not be unlinked. > - introduce a new option, xib=. this spcifies an inode bitmap, and > the file will not be unlinked. > > Ex) > # mount -t aufs -o xib=/xino/for/aufs_mntpnt, \ > br:/branchA=rw+brxino=/xino/for/branchA \ > :/branchB=ro+brxino=/xino/for/branchB \ > none /aufs_mntpnt > > - when the xib option is specified, brxino must be given for all > branches. > - the xino option is un-usable when xib or brxino is specified. > > I gave up assigning the xino filename automatically based upon uuid or > something. > So the filename is entirely user's choise and responsiblity. :-) > If he/she type a wrong filename, aufs inode numbers will be perfectly > crazy. So these options are dangerous. This sounds like a great idea, would be perfectly viable for my use. :-) > I see. > But now I have a temporary part time job for my daily food. > Please wait for a while. Ah, glad you've got a job. There's no rush, there're no really "breaking" bugs in aufs - good job! I really like it! :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+3KRUMzc1WGo4zgRAoeVAJ47peYZ3eHFd9Raaa0FDuncRug3EQCgn2Dh nvxr65YOn8CW4wMJxtpad6M= =0GKj -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: BUG: kernel NULL pointer dereference
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> When I mount again, it will find files in /tmp/.aufs.xino.dir/ that >> matche the uuid of the members of the branch set specified, and use >> those as a base (but creating new xino-files for uuids that are not in >> /tmp/.aufs.xino.dir/). >> >> Does that make sense? Would it be doable? > > Well, it is interesting to me. > I will consider again later while it may not be within this year. Ok, thanks. :-) > kernel BUG may happen on nfs client, or it cannot shutdown itself since > even the root dir on nfs client may be ESTALE too. > It means you may have to reboot(power off/on) your nfs client. > Finally, it is meaningless to return ESTALE, isn't it? Well, I need some kind of quick fix, even if it's a local hack; The NFS server is way more important for everyday service than the NFS clients (no problem if they have to be rebooted, but it's a problem if the NFS server has to be rebooted). Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+hJqUMzc1WGo4zgRAsvAAJ4/ZyUgOLLTXSQRY29QI+FX0nna/ACdEH2E UFzoJkbxjwzx/rqBtrJC8aE= =Fn6E -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: BUG: kernel NULL pointer dereference
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Jorgen, > Thank you for sending logs. > Let's change the thread and continue the discussion a little more. > > I think that make the xino files to be persistent is most hopeful, but > it may be difficult to ensure the branch set is perfectly the same. > > With the persistent xino files will allow you to: > # mount -t aufs -o br:,keep_xino none /aufs > # umount /aufs and keep the xino files > # mount -t aufs -o br:,use_old_xino none /aufs > > But if you give different branch set to the second aufs mounting, you > may be totally lose since it is hard to judge the correctness of > matching xino files and the branch set. > The uuid on a filesystem may be helpful, but the recent vmware/xen users > may not pay attentions to uuid. If the uuid is illegally set to the > branch fs, it must be useless. Well, I'm not too good at this, but my suggestion is to have an option (compiletime or runtime) to separate to one xino-file per underlying FS (indexed by UUID), and to be able to specify what dir to write (or read - - using your suggested keep_xino-option) them from. Example: # mount -t aufs -o br:,keep_xino,xino=/tmp/.aufs.xino.dir/ /aufs # umount /aufs keeping the xino files # mount -t aufs -o br:,keep_xino,xino=/tmp/.aufs.xino.dir/ /aufs When I mount again, it will find files in /tmp/.aufs.xino.dir/ that matche the uuid of the members of the branch set specified, and use those as a base (but creating new xino-files for uuids that are not in /tmp/.aufs.xino.dir/). Does that make sense? Would it be doable? > About the approach you wrote which is to make nfs client to re-lookup > the inodes after rebooting, I am afraid it means nfs server will return > ESTALE for almost all accesses, and the nfs-root client will not work > anymore. Ah, darn. This might though be a better "default behaviour" than kernel BUG on nfs server? Or is there another "temporary fix" that atleast won't make the aufs BUG when this happen? Kindest regards, Jørgen P. Tjernø -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+g2uUMzc1WGo4zgRAgHSAJ43YCMHDJBRHAUyw8CDtyxngk6XlwCfeTeq NDrU7uDnh18JkPrmxNKaVCE= =yljS -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Kernel BUG during / after file write (?)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > I see. > Such situation is beyond my expectations or imaginations. :-) > I guess the root cause is related to the inode number in aufs. > Basically, aufs stores its inode numbers to the xino files, and the xino > files are volatile which means the xino files are totally gone after the > aufs is unmounted. Mounting aufs again with the same branches, the new > aufs will have new, different set of inode numbers. > When the nfs client accesses to nfs server by oboleted inode number, > aufs detects their inconsistency and call BUG_ON macro. > I guess this is the story of your case. Okay, I think I understand. :-) > If it is true, the the quiestion is: does aufs needs persistent xino > files? Well, is there any other workaround for this? Can you somehow invalidate the inode number to the NFS client, so that it looks up the inode number again? I don't know if it's of any interest, but I posted the dmesg output with your patch to the other thread. Continue discussion there? Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+gKpUMzc1WGo4zgRAmrbAJ9pX54YDPe79YIXdQPSndwaNAP8DgCfcjYl EyTtbvWiW7ykZ4Mq0757fJ0= =aROT -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: BUG: kernel NULL pointer dereference
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Hi Jorgen, > > "Jrgen_P._Tjern": >> I'm not sure what triggered this, but I guess I should post it just in case. > > What is your kernel version and aufs's? > If you can, please apply this debug patch and let me know the result. > The patch is for finding out where the BUG happened in a function. > After boot of nfs client: [EMAIL PROTECTED]:~# dmesg|tail -n 5 [52425.250149] skge eth2: Link is up at 100 Mbps, full duplex, flow control both [52425.349766] br0: port 3(eth2) entering learning state [52425.353753] br0: topology change detected, propagating [52425.353755] br0: port 3(eth2) entering forwarding state [52455.193058] aufs aufs_encode_fh:564:nfsd[2966]: set default find_exported_dentry for xfs [EMAIL PROTECTED]:~# reboot The system is going down for reboot NOW! [EMAIL PROTECTED]:~# logout After reboot of NFS server: [ 58.234729] skge eth2: Link is up at 100 Mbps, full duplex, flow control both [ 58.333991] br0: port 3(eth2) entering learning state [ 58.337976] br0: topology change detected, propagating [ 58.337978] br0: port 3(eth2) entering forwarding state [ 81.300262] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory [ 81.391964] NFSD: starting 90-second grace period [ 84.050123] BUG: unable to handle kernel NULL pointer dereference at virtual address [ 84.056266] printing eip: [ 84.059253] [ 84.059255] *pde = [ 84.062397] Oops: [#1] [ 84.065384] SMP [ 84.068387] Modules linked in: aufs skge sata_via sata_promise [ 84.071529] CPU:0 [ 84.071530] EIP:0060:[<>]Not tainted VLI [ 84.071531] EFLAGS: 00010297 (2.6.20.3 #8) [ 84.080428] EIP is at 0x0 [ 84.083414] eax: f7932760 ebx: c05e4a00 ecx: edx: f6ddbd94 [ 84.086483] esi: f7932760 edi: f7932760 ebp: f6ddbdac esp: f6ddbd78 [ 84.089552] ds: 007b es: 007b ss: 0068 [ 84.092566] Process nfsd (pid: 2964, ti=f6dda000 task=f7b72a90 task.ti=f6dda000) [ 84.095495] Stack: c02eb98d f90da8b0 f6ddbd9c c017871a f7952248 f6ddbd9c 000e [ 84.101677] a178 c02eb920 0001 f6ddbe88 f90db78a 0001 [ 84.107857]f90da8b0 8000 c042eb69 0001 c1806d40 [ 84.113923] Call Trace: [ 84.119841] [] show_trace_log_lvl+0x1a/0x30 [ 84.122900] [] show_stack_log_lvl+0xb1/0xe0 [ 84.125955] [] show_registers+0x1c1/0x2b0 [ 84.128855] [] die+0x11c/0x270 [ 84.131888] [] do_page_fault+0x2af/0x580 [ 84.134937] [] error_code+0x7c/0x84 [ 84.137978] [] aufs_decode_fh+0x83a/0xf10 [aufs] [ 84.141059] [] fh_verify+0x213/0x590 [ 84.144000] [] nfsd_access+0x1c/0x100 [ 84.147010] [] nfsd3_proc_access+0x8e/0xf0 [ 84.150062] [] nfsd_dispatch+0x88/0x1d0 [ 84.153108] [] svc_process+0x3b6/0x710 [ 84.156156] [] nfsd+0x164/0x280 [ 84.159189] [] kernel_thread_helper+0x7/0x14 [ 84.162095] === [ 84.165097] Code: Bad EIP value. [ 84.168130] EIP: [<>] 0x0 SS:ESP 0068:f6ddbd78 [ 84.171204] <1>BUG: unable to handle kernel NULL pointer dereference at virtual address [ .. snip .. ] It seems this kernel BUG repeats infinitely. This is with your supplied patch to export.c. Kindest regards, Jørgen P. Tjernø. :-) -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+gH0UMzc1WGo4zgRAt95AJwMGeVI69AUKbKfIcdU4LpHrk5zCwCeLOJ1 15izMk0+/AY7tMqSX79DzBw= =QolD -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Kernel BUG during / after file write (?)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> Just so you know: Even though my kernel is SMP, the server is running a >> single CPU (not with hyperthreading either, it's a simple AMD Athlon 64). >> I'll try your patch, and see if it happens again. :-) > > It means I was totally wrong... > By the way, your log in previous mail showed that the cpu registers are > 32bit. Are you sure your cpu is 64bit? (just to make sure) Yes, but the kernel / userspace is 32bit: [EMAIL PROTECTED]:~$ cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 12 model name : AMD Athlon(tm) 64 Processor 3200+ stepping: 0 cpu MHz : 2202.909 cache size : 512 KB fdiv_bug: no hlt_bug : no f00f_bug: no coma_bug: no fpu : yes fpu_exception : yes cpuid level : 1 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow up ts fid vid ttp bogomips: 4410.74 clflush size: 64 >> The issue is that the server that uses aufs BUGs if you boot the server >> while a client which uses an NFS-export from the server as root is >> *turned on*. (Very strange) > > Unfortunately, I cannot get what you wrote. > You have two different problems, nfsd and glftpd, don't you? That's right. :-) The nfsd issue (not reported in this thread), is triggered when the NFS server (which exports aufs) is booting while the nfs client (which uses the nfs server for root) is already on. Example: * NFS server is turned on * NFS-root client is turned on - Everything works fine * NFS server is rebooted, NFS-root client is not rebooted * NFS server gets a *lot* of BUGs in dmesg I'll get you more info with the patch you gave me earlier, ASAP. Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+f9gUMzc1WGo4zgRAowxAJ0XwabyZxi62oUqwbJY7ONvsIyRVgCgoOcJ +LpkTcTeSFgRQVPXuB1OppM= =wL2S -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: Kernel BUG during / after file write (?)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> I had a kernel BUG today, using a slightly old AUFS (20070917). It seems >> to have been caused by glftpd (an ftp daemon), while I was uploading >> files. I believe it happened during (or after) a file transfer, not an >> mkdir. Here's the dmesg output: (I'll upgrade my aufs soon) > > The aufs version you are using is not old. > Because your kernel is 2.6.20.3, I don't think the newer aufs is not > important for you. > > The bug you met is interesting to me. > The line dentry.c:737 shows the problem is related to the branch > management of aufs. It might be an SMP specific problem. But I am not > sure. Did you remount,del or add while you were uploading? > Won't you try this patch against 20070917? Just so you know: Even though my kernel is SMP, the server is running a single CPU (not with hyperthreading either, it's a simple AMD Athlon 64). I'll try your patch, and see if it happens again. :-) I have another bug to report too, but I'll have to spend a bit more time on it. I think you gave me a patch with BUG_ON's to locate the problem. The issue is that the server that uses aufs BUGs if you boot the server while a client which uses an NFS-export from the server as root is *turned on*. (Very strange) Kindest regards, Jørgen. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+foKUMzc1WGo4zgRAhk6AJ9kPKzYUd6Cf3WYGtFCNynf38w7WwCgjfDq /WrnhzdSONuC9D6UB1FZhyg= =jBRb -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Kernel BUG during / after file write (?)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I had a kernel BUG today, using a slightly old AUFS (20070917). It seems to have been caused by glftpd (an ftp daemon), while I was uploading files. I believe it happened during (or after) a file transfer, not an mkdir. Here's the dmesg output: (I'll upgrade my aufs soon) [ 63.667977] aufs 20070917 [ .. snip .. ] [592019.792625] [ cut here ] [592019.795626] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/dentry.c:737! [592019.798676] invalid opcode: [#1] [592019.801678] SMP [592019.804509] Modules linked in: bnep rfcomm l2cap aufs skge sata_via sata_promise [592019.810601] CPU:0 [592019.810602] EIP:0060:[]Not tainted VLI [592019.810603] EFLAGS: 00210217 (2.6.20.3 #8) [592019.819629] EIP is at au_reval_dpath+0x444/0x4c0 [aufs] [592019.822510] eax: f0507394 ebx: f2ec9694 ecx: 0001 edx: 00ff [592019.825574] esi: c372b3e0 edi: 0002 ebp: f2f85cf8 esp: f2f85ca8 [592019.828637] ds: 007b es: 007b ss: 0068 [592019.831646] Process glftpd (pid: 27489, ti=f2f84000 task=dfb8f560 task.ti=f2f84000) [592019.834762] Stack: f2f85cbc c02e0cf2 c05e48e0 e2a9cd34 f2f85cc8 c02ee692 c05e4be0 [592019.840739]c387c3d8 c01e3a4a f782a298 c372b3e0 f2f85ce8 f90c7c55 [592019.846852]f2f85cf8 c372b3e0 0002 f2f85de0 f90c6b5c 0001 f2f85dc0 [592019.852964] Call Trace: [592019.858686] [] show_trace_log_lvl+0x1a/0x30 [592019.861731] [] show_stack_log_lvl+0xb1/0xe0 [592019.864774] [] show_registers+0x1c1/0x2b0 [592019.867811] [] die+0x11c/0x270 [592019.870715] [] do_trap+0x82/0xb0 [592019.873717] [] do_invalid_op+0x97/0xb0 [592019.876751] [] error_code+0x7c/0x84 [592019.879780] [] aufs_d_revalidate+0x13c/0x9f0 [aufs] [592019.882847] [] __link_path_walk+0x5ff/0xdd0 [592019.885891] [] link_path_walk+0x46/0xd0 [592019.888775] [] do_path_lookup+0x86/0x1c0 [592019.891810] [] __user_walk_fd+0x38/0x60 [592019.894846] [] sys_statfs64+0x48/0x90 [592019.897879] [] syscall_call+0x7/0xb [592019.900906] === [592019.903819] Code: 12 0e f9 e8 7f a7 05 c7 e9 73 fc ff ff 0f 0b eb fe 8d b6 00 00 00 00 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe <0f> 0b eb fe 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe a1 [592019.913720] EIP: [] au_reval_dpath+0x444/0x4c0 [aufs] SS:ESP 0068:f2f85ca8 [592019.924569] -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG+Q+WUMzc1WGo4zgRAs69AJ9Dtg4xGdoGKQqHQ+4qBb+riOtIQACffPSP vzkI82EasgAkn1y0nef5XcI= =+7LL -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: remount with delete-prepend, slow?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > I had such plan to support multiple writable branches, but it is still a > plan. :-) What do you mean by 'multiple writable branches'? > In you case, a new option such like, > remount,reorder:: > may be better. If you like it, I will try design/implement it. That sounds like a good idea, I'd use it. Is there any way I can get the ordering of the disks at the current time? If I del and add using remount, I either have to store the new order myself, or parse /proc/mounts (which is a bit of a hassle) :-) > Aufs truncates the External Inode Number Bitmap by default at deleting a > branch. I hope this description in aufs manual will help you. Ah, thanks. Can I use notrunc_xib like this: mount -o remount,notrunc_xib,del:/vault/disk6,prepend:/vault/disk6=rw,trunc_xib /storage Or can I not turn off and then on external ib-file in the same remount? Thanks for your help! Kindest regards, Jørgen P. Tjernø. Ps. I'll post the 'make sure aufs has enough free space'-script that I've written when it's done, in case anyone's interested. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG83neUMzc1WGo4zgRAjgnAKCWVySXNS+zf1tptiVhAFKoG8ELrACfeAcB W7zhsJM4VN3ITg/HHEEtiHQ= =uXQI -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
remount with delete-prepend, slow?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm trying to write a script that adds the drive with the most free space as the 'first' drive, by using command like this: $ mount -o remount,del:/vault/disk6,prepend:/vault/disk6=rw /storage (I'm making sure the directory structure matches, so that it'll simply pick the first drive to write the file to. For the future: Would it be possible to make it create the dir-structure on a particular "write-first"-drive, and create new files on it, instead of finding a "matching dir structure"?) This command takes a couple of minutes to finish, and I get this in dmesg: [61632.277643] aufs do_xib_restore:643:mount[23548]: testing a large xino file 3758097024 [61642.668549] aufs do_xib_restore:643:mount[23548]: testing a large xino file 8466517632 [61665.501141] aufs do_xib_restore:643:mount[23548]: testing a large xino file 8519454512 [61688.857910] aufs do_xib_restore:643:mount[23548]: testing a large xino file 8116821632 [61710.718058] aufs do_xib_restore:643:mount[23548]: testing a large xino file 4037693024 Why does it take so long, and what does this mean? Relevant /proc/mounts: /dev/disk/by-uuid/e2bd2748-1050-4fd2-8c98-e90a6dc863e2 /vault/disk1 xfs rw 0 0 /dev/disk/by-uuid/2e35bd30-8879-4127-8ec0-1a2dcbdceca8 /vault/disk2 xfs rw 0 0 /dev/disk/by-uuid/16b24ed5-5947-4610-9598-278e677ad1cc /vault/disk3 xfs rw 0 0 /dev/disk/by-uuid/f38c7e6d-9dac-4763-8fbf-95311a3acfcf /vault/disk4 xfs rw 0 0 /dev/disk/by-uuid/71f1489f-f197-4912-ad29-be062903a6c1 /vault/disk5 xfs rw 0 0 /dev/disk/by-uuid/0b2955d8-ccdb-40c5-965c-e75119d2f234 /vault/disk6 xfs rw 0 0 /dev/disk/by-uuid/88457bda-800b-4b64-89b7-c0d0096eea5d /vault/misc ext3 rw,data=ordered 0 0 none /storage aufs rw,xino=/vault/disk5/.aufs.xino,udba=inotify,br:/vault/disk6=rw:/vault/disk5=rw:/vault/disk4=rw:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw 0 0 Thanks in advance! Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG8PNeUMzc1WGo4zgRAlDyAJ0cRmChSGCrQ2tkcFF/eFjLd1fnDwCghc42 hW1eF1hzfK4IuMb+sMd1tzA= =mGne -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
BUG: kernel NULL pointer dereference
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm not sure what triggered this, but I guess I should post it just in case. Relevant /proc/mounts-part: /dev/disk/by-uuid/e2bd2748-1050-4fd2-8c98-e90a6dc863e2 /vault/disk1 xfs rw 0 0 /dev/disk/by-uuid/2e35bd30-8879-4127-8ec0-1a2dcbdceca8 /vault/disk2 xfs rw 0 0 /dev/disk/by-uuid/16b24ed5-5947-4610-9598-278e677ad1cc /vault/disk3 xfs rw 0 0 /dev/disk/by-uuid/f38c7e6d-9dac-4763-8fbf-95311a3acfcf /vault/disk4 xfs rw 0 0 none /storage aufs rw,xino=/vault/disk4/.aufs.xino,udba=inotify,br:/vault/disk4=rw:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw 0 0 Relevant /etc/exports-part: /storage/art-root 192.168.0.24(rw,async,no_root_squash,fsid=6996) dmesg info: [ 225.019306] BUG: unable to handle kernel NULL pointer dereference at virtual address [ 225.025376] printing eip: [ 225.028385] [ 225.028387] *pde = [ 225.031234] Oops: [#1] [ 225.034221] SMP [ 225.037223] Modules linked in: bnep rfcomm l2cap aufs sata_promise skge sata_via [ 225.043372] CPU:0 [ 225.043372] EIP:0060:[<>]Not tainted VLI [ 225.043374] EFLAGS: 00010297 (2.6.20.3 #8) [ 225.052271] EIP is at 0x0 [ 225.055256] eax: f792d0bc ebx: c05e4a00 ecx: edx: f72cfdc4 [ 225.058325] esi: f792d0bc edi: f792d0bc ebp: f72cfddc esp: f72cfda8 [ 225.061394] ds: 007b es: 007b ss: 0068 [ 225.064258] Process nfsd (pid: 2902, ti=f72ce000 task=f7d0ea90 task.ti=f72ce000) [ 225.067341] Stack: c02eb98d f9078620 f72cfdcc c017871a f7938d84 f72cfdcc 000e [ 225.073523]97c949d1 064c0039 c02eb920 f73b52d4 f72cfeb4 f9079514 0001 [ 225.079610]f9078620 c18df500 f6a68094 f72cfe2c c0161799 c06aec74 [ 225.085741] Call Trace: [ 225.091653] [] show_trace_log_lvl+0x1a/0x30 [ 225.094711] [] show_stack_log_lvl+0xb1/0xe0 [ 225.097616] [] show_registers+0x1c1/0x2b0 [ 225.100665] [] die+0x11c/0x270 [ 225.103698] [] do_page_fault+0x2af/0x580 [ 225.106746] [] error_code+0x7c/0x84 [ 225.109788] [] aufs_decode_fh+0x864/0xed0 [aufs] [ 225.112793] [] fh_verify+0x213/0x590 [ 225.115781] [] nfsd3_proc_getattr+0x7a/0xf0 [ 225.118836] [] nfsd_dispatch+0x88/0x1d0 [ 225.121883] [] svc_process+0x3b6/0x710 [ 225.124930] [] nfsd+0x164/0x280 [ 225.127963] [] kernel_thread_helper+0x7/0x14 [ 225.130870] === [ 225.133871] Code: Bad EIP value. [ 225.136904] EIP: [<>] 0x0 SS:ESP 0068:f72cfda8 Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG7b2wUMzc1WGo4zgRAs76AJ9Z0XFfZBh3zdsDY/X7zqdy8wFvvACeMfaK 30qV4/fJQfniKbT5pWHDDtA= =RWgp -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
Re: How to NFS-export AUFS?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Hello Jorgen, > > "Jrgen_P._Tjern": >> I'm trying to export an aufs over nfs, but it always yields >> NFSERR_ACCESS when I try to mount (mount call failed: 13). > ::: > > Essentially you need, > - specify fsid in your /etc/export. > - enable CONFIG_AUFS_EXPORT. > - The branch filesystem must support NFS-exporting. > - NFSv2 is not supported. > > Please refer to the aufs manual in detail. Ah, thanks. Sorry, should've read the manual. ;-) I'll test this when I get home, but I guess this is all I needed. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG57C0UMzc1WGo4zgRAkt2AJ0Y2wh6+sOpaX87z3S7cu59fRrMGACfW6fB BaGciirGAcmJl/jTmIbmMe0= =Te7U -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
How to NFS-export AUFS?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm trying to export an aufs over nfs, but it always yields NFSERR_ACCESS when I try to mount (mount call failed: 13). The way I've tried is: /etc/exports: /storage/some-path 192.168.0.24(rw,no_root_squash,async) /etc/fstab: none/storageaufs br:/vault/lvm=rw:/vault/disk2=rw:/vault/disk1=rw,udba=inotify 0 0 If I mount 192.168.0.3:/storage/some-path, it fails. If I use the following: /etc/exports: /vault/disk3/some-path 192.168.0.24(rw,no_root_squash,async) And mount 192.168.0.3:/vault/disk3/some-path, it works. Any suggestions? I'd rather not export one of the underlying fs-es directly. :-/ Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG55VvUMzc1WGo4zgRAoYoAKCiR7CKHPEqA9GceU7murpVAzUXhQCfYxai MeabazWAz1BpwRPJngWSb10= =3rBV -END PGP SIGNATURE- - This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
renaming dir who has child(ren) on multiple branches, is not supported
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I found this in my dmesg: (aufs 20070709) aufs may_rename_srcdir:401:mv[28278]: renaming dir who has child(ren) on multiple branches, is not supported I guess this is a non-implemented feature? It should give more information about what dir this occured on. Also, what would happen to the 'mv'-process that tried to do this move? Will it manually copy the files, then delete at original branch? Thanks in advance! Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGrzgWUMzc1WGo4zgRAicXAJwIUfL06KM4ZJcOuDjjExGsOmnfoACfapw+ gXy2ZA8XPTRxRkuvXYMB3gk= =Dw2E -END PGP SIGNATURE- - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/
Re: Inclusion in the kernel?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Russell Harmon wrote: > I was just wondering, has anyone tried to get aufs included in the > kernel? If not, why not? I've got nothing against AUFS (I use it myself, and I'm quite content with it), but I think it'll have to mature a bit before it would be accepted into the 2.6-tree. I've had almost no serious issues - a couple of kernel panics, but mr. Okajima is very quick to put out fixes. :-) I hope mr. Okajima will keep up the good work, and we'll probably see aufs in 2.6 soon :-) OTOH, I guess you could try getting it added with an EXPERIMENTAL tag or something. But aufs still has very rapid development cycles, so the code in the kernel release would get "outdated" very soon. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGnjj7wqQbW3my7pcRAoXgAJkBvQ0KUpsF1P+n0awIkqxhloaLiwCeO3xm tu8A3K3/Gc98Nzv8NSL3DOU= =+VGE -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: >> "Jrgen P. Tjern": >>> [2297447.631266] aufs 20070702 >>> [2297509.899071] aufs au_new_inode:347:smbd[27416]: Un-notified UDBA or >>> directly renamed dir, b0, xfs, Simon - Garfunkel - The Best Of Simon - >>> Garfunkel - Song For The Asking.mp3, hi1610612893, i16. > > Please show me your /sys/fs/aufs/*/xino file, if it exists. [EMAIL PROTECTED]:~$ cat /sys/fs/aufs/*/xino 0x4096 8192 0: 1, 0x4096 3821854336 1: 1, 0x4096 4037693024 2: 1, 0x4096 57868296 3: 1, 0x4096 494731296 > And please try this patch. I'm updating to the latest monday release now (20070709). I'll let you know if the problem is still there. :-) Thanks! Kindest regards, Jørgen P. Tjernø Ps. Sorry for the slow reply, kinda busy. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGlgvywqQbW3my7pcRAjLvAKCErgexWWnQj6SVjlFlu9SWW/E+fACfcTQz 3kp7wS2c+sDK+Brs7c5V4D0= =5uxn -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Hello Jorgen, > > "Jrgen_P._Tjern": >> [EMAIL PROTECTED] wrote: >>> Thank you very much for your tests. >>> Currently I am considering about the bug in aufs inotify handler. >>> Do you know whether this file 'Simon and Garfunkel ... .mp3' has ever >>> been renamed or not? >> (bah, sorry for misposting this reply directly to you, Junjiro) >> >> It shouldn't have been renamed ever, no. :-) > > I think the root cause of this problem is mis-managing inode number > after unlink/rename. > I have re-implemented inode management in aufs, please try the latetest > code. And if the problem (the warning) arise again, please let me know. > > Thank you > Junjiro Okajima I just upgraded to the newest version, and this is my dmesg, shortly after mounting my stuff: [2297447.631266] aufs 20070702 [2297509.899071] aufs au_new_inode:347:smbd[27416]: Un-notified UDBA or directly renamed dir, b0, xfs, Simon - Garfunkel - The Best Of Simon - Garfunkel - Song For The Asking.mp3, hi1610612893, i16. It's not an access directly to the underlying fs, so how come this happens? Is this anything like hte broken_ino? Thanks for your quick responses. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGiOWAwqQbW3my7pcRAqaKAJ4zqZKoIyku8aRXxLu5cRq72h/uvgCfbRGm 6NzoaeTneN+5zqAjbAIS84w= =/CGV -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Thank you very much for your tests. > Currently I am considering about the bug in aufs inotify handler. > Do you know whether this file 'Simon and Garfunkel ... .mp3' has ever > been renamed or not? (bah, sorry for misposting this reply directly to you, Junjiro) It shouldn't have been renamed ever, no. :-) Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGe9qdwqQbW3my7pcRAglgAJkBDpxPoi1kp2egOrzZayQQrB1WZgCg4chb yu4TBPnLyNuoZNIvH2dmDdM= =/qvr -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > Here is a debug patch to print more information when the message 'broken > ino' was produced. Please reproduce the message and send me the debug > log. Here is the dmesg, it has only occured once after I applied the patch. I also updated to the newest aufs. :-) The machine isn't being used as regularly during summer, so use-pattern is a bit different now. :-) I hope this helps! Kindest regards, Jørgen P. Tjernø [1160867.631996] aufs 20070618 [1204700.027800] aufs au_new_inode:322:smbd[18549]: broken ino, b0, xfs, Simon - Garfunkel - The Best Of Simon - Garfunkel - Song For The Asking.mp3, hi1610612893, i17. Try udba=inotify. [1204700.027835] aufs au_new_inode:328:smbd[18549]: dentry [1204700.027853] d-1: Simon and Garfunkel - The Best Of/Simon - Garfunkel - The Best Of Simon - Garfunkel - Song For The Asking.mp3, aufs, cnt 1, flags 0x10 [1204700.027871] i-1: err 0 [1204700.027881] d-1: bstart 0, bend 0, bwh -1, bdiropq -1, gen 0 [1204700.027896] d0: Simon and Garfunkel - The Best Of/Simon - Garfunkel - - The Best Of Simon - Garfunkel - Song For The Asking.mp3, xfs, cnt 1, flags 0x20 [1204700.027924] i0: i1610612893, xfs, cnt 1, nl 1, 0100664, sz 4481042, blk 8768, ct 59950, np 0, st 0x0, g 2 [1204700.027940] aufs au_new_inode:329:smbd[18549]: inode [1204700.027959] i-1: i17, aufs, cnt 2, nl 24, 040777, sz 4096, blk 8, ct 9767, np 0, st 0x0, g 0 [1204700.027973] i-1: bstart 0, bend 3, gen 0 [1204700.027991] i0: i536871069, xfs, cnt 3, nl 12, 040777, sz 4096, blk 8, ct 9767, np 0, st 0x0, g 0 [1204700.028004] i1: err 0 [1204700.028011] i2: err 0 [1204700.028030] i3: i118685697, ext3, cnt 3, nl 14, 040777, sz 4096, blk 8, ct 18432, np 0, st 0x0, g a4259808 [1204700.028046] aufs au_new_inode:331:smbd[18549]: d [1204700.028059] d-1: //applications, aufs, cnt 2, flags 0x8 [1204700.028076] i-1: i17, aufs, cnt 2, nl 24, 040777, sz 4096, blk 8, ct 9767, np 0, st 0x0, g 0 [1204700.028092] d-1: bstart 0, bend 3, bwh -1, bdiropq -1, gen 0 [1204700.028105] d0: //applications, xfs, cnt 1, flags 0x20 [1204700.028124] i0: i536871069, xfs, cnt 3, nl 12, 040777, sz 4096, blk 8, ct 9767, np 0, st 0x0, g 0 [1204700.028137] d1: err 0 [1204700.028145] d2: err 0 [1204700.028156] d3: //applications, ext3, cnt 1, flags 0x20 [1204700.028176] i3: i118685697, ext3, cnt 3, nl 14, 040777, sz 4096, blk 8, ct 18432, np 0, st 0x0, g a4259808 -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGeQsSwqQbW3my7pcRAuevAJ9oUOz6sUgnmaLIu1HIYAhcyy0grwCfd6Go jQJNlKs1FxH+PzNx0HC27OM= =VaPi -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
[EMAIL PROTECTED] wrote:
> "Jrgen_P._Tjern":
>> I removed 3.Lbs as it's not on the fs any more. I ran your script with
>> the changes, produced a lot of output. See the attached file. :-)
>
> Sorry, I should write more correctly.
> The path you set is /storage and it is aufs mount point, isn't it?
> It should be branch path. In your case, probably /vault/disk3 or
> something.
>
>
> Junjiro Okajima
I picked the most recent entry in my dmesg, and added that hi and fname
to the commands. (added 536871063 to the egrep)
Entry was:
[1081101.429555] aufs au_new_inode:325:find[31722]: broken ino, b0,
linux-meta/linux-image-generic_2.6.17.11_i386.deb, hi536871063, i50306.
Try udba=inotify.
Check attached file for output. :-)
Kindest regards, Jørgen P. Tjernø
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGdovuwqQbW3my7pcRAnIpAKDSdXmGqmVtOcuJ4llwkC5WNrQXkgCfZ3/F
zTSxhiT7l6e50zA1pDg8yuI=
=OFAm
-END PGP SIGNATURE-
[EMAIL PROTECTED]:~$ bash inode.sh
+ df -i /vault/disk3
FilesystemInodes IUsed IFree IUse% Mounted on
/dev/sdd13125686407382 3125612581% /vault/disk3
+ for i in 1 2
+ sudo mount -o remount /vault/disk3
+ sudo find /vault/disk3 '(' '(' -name linux-image-generic_2.6.17.11_i386.deb
-o -name libxtst6_1.0.1-3build1_i386.deb -o -name
samba-common_3.0.22-1ubuntu4.2_i386.deb ')' -fls /tmp/2674.1a ')' -o -printf
'%i\n'
+ sort -n
+ for i in 1 2
+ sudo mount -o remount /vault/disk3
+ sort -n
+ sudo find /vault/disk3 '(' '(' -name linux-image-generic_2.6.17.11_i386.deb
-o -name libxtst6_1.0.1-3build1_i386.deb -o -name
samba-common_3.0.22-1ubuntu4.2_i386.deb ')' -fls /tmp/2674.2a ')' -o -printf
'%i\n'
+ diff -u /tmp/2674.1a /tmp/2674.2a
+ diff -u /tmp/2674.1 /tmp/2674.2
+ cat /tmp/2674.1a
134217865 12 -rw-rw-rw- 1 daxxar storage 9774 Jul 7 2006
/vault/disk3/apt-proxy/ubuntu/pool/main/libx/libxtst/libxtst6_1.0.1-3build1_i386.deb
134217862 2148 -rw--- 1 aptproxy nogroup 2196940 May 16 03:06
/vault/disk3/apt-proxy/ubuntu-security/pool/main/s/samba/samba-common_3.0.22-1ubuntu4.2_i386.deb
536871063 24 -rw-rw-rw- 1 daxxar storage 23806 Feb 8 11:03
/vault/disk3/apt-proxy/ubuntu-security/pool/main/l/linux-meta/linux-image-generic_2.6.17.11_i386.deb
+ tail /tmp/2674.1
2029205398
2029205399
2029205400
2029205401
2029205402
2029205403
2029205404
2029205405
2029205406
2029205407
+ egrep -w '(536871063|134217865|134217862)' /tmp/2674.1 /tmp/2674.2
+ rm -f /tmp/2674.1 /tmp/2674.1a /tmp/2674.2 /tmp/2674.2a
rm: cannot remove `/tmp/2674.1a': Operation not permitted
rm: cannot remove `/tmp/2674.2a': Operation not permitted
-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
[EMAIL PROTECTED] wrote:
>> "Jrgen_P._Tjern":
>>> Branch 0 would be the first branch on the /proc/mounts-line? If so,
>>> that's /vault/disk3, which is xfs. How do I find largest inode number?
>
> Sorry, please add one line.
>
> [ .. snip .. ]
I removed 3.Lbs as it's not on the fs any more. I ran your script with
the changes, produced a lot of output. See the attached file. :-)
Kindest regards, Jørgen P. Tjernø.
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGdlUewqQbW3my7pcRAkpcAKDpHYCIPqCdoALOmod0W0TaVcxtgwCeOQz0
PeEXoKbETNgHcpMIqymYHUk=
=X3gE
-END PGP SIGNATURE-
[EMAIL PROTECTED]:~$ bash inode.sh
+ df -i /storage
FilesystemInodes IUsed IFree IUse% Mounted on
none 3125686407382 3125612581% /storage
+ for i in 1 2
+ sudo mount -o remount /storage
+ sudo find /storage '(' '(' -name libxtst6_1.0.1-3build1_i386.deb -o -name
samba-common_3.0.22-1ubuntu4.2_i386.deb ')' -fls /tmp/31648.1a ')' -o -printf
'%i\n'
+ sort -n
+ for i in 1 2
+ sudo mount -o remount /storage
+ sort -n
+ sudo find /storage '(' '(' -name libxtst6_1.0.1-3build1_i386.deb -o -name
samba-common_3.0.22-1ubuntu4.2_i386.deb ')' -fls /tmp/31648.2a ')' -o -printf
'%i\n'
+ diff -u /tmp/31648.1a /tmp/31648.2a
--- /tmp/31648.1a 2007-06-18 11:35:21.0 +0200
+++ /tmp/31648.2a 2007-06-18 11:37:56.0 +0200
@@ -1,2 +1,2 @@
- 50255 12 -rw-rw-rw- 1 daxxar storage 9774 Jul 7 2006
/storage/apt-proxy/ubuntu/pool/main/libx/libxtst/libxtst6_1.0.1-3build1_i386.deb
- 50336 2148 -rw--- 1 aptproxy nogroup 2196940 May 16 03:06
/storage/apt-proxy/ubuntu-security/pool/main/s/samba/samba-common_3.0.22-1ubuntu4.2_i386.deb
+ 50409 12 -rw-rw-rw- 1 daxxar storage 9774 Jul 7 2006
/storage/apt-proxy/ubuntu/pool/main/libx/libxtst/libxtst6_1.0.1-3build1_i386.deb
+ 50490 2148 -rw--- 1 aptproxy nogroup 2196940 May 16 03:06
/storage/apt-proxy/ubuntu-security/pool/main/s/samba/samba-common_3.0.22-1ubuntu4.2_i386.deb
+ diff -u /tmp/31648.1 /tmp/31648.2
--- /tmp/31648.12007-06-18 11:35:21.0 +0200
+++ /tmp/31648.22007-06-18 11:37:56.0 +0200
@@ -45576,154 +45576,6 @@
49946
49973
50029
-50188
-50189
-50190
-50191
-50192
-50193
-50194
-50195
-50196
-50197
-50198
-50199
-50200
-50201
-50202
-50203
-50204
-50205
-50206
-50207
-50208
-50209
-50210
-50211
-50212
-50213
-50214
-50215
-50216
-50217
-50218
-50220
-50221
-50222
-50223
-50224
-50225
-50226
-50227
-50228
-50229
-50230
-50231
-50232
-50233
-50234
-50235
-50236
-50237
-50238
-50239
-50240
-50241
-50242
-50243
-50244
-50245
-50246
-50247
-50248
-50249
-50250
-50251
-50252
-50253
-50254
-50256
-50257
-50258
-50259
-50260
-50261
-50262
-50263
-50264
-50265
-50266
-50267
-50268
-50269
-50270
-50271
-50272
-50273
-50274
-50275
-50276
-50277
-50279
-50280
-50281
-50282
-50283
-50284
-50285
-50286
-50287
-50288
-50289
-50290
-50291
-50292
-50293
-50294
-50295
-50296
-50297
-50299
-50300
-50301
-50302
-50303
-50304
-50305
-50306
-50307
-50308
-50309
-50310
-50311
-50312
-50313
-50314
-50315
-50316
-50317
-50318
-50319
-50320
-50321
-50322
-50323
-50324
-50325
-50326
-50327
-50328
-50329
-50330
-50331
-50332
-50333
-50334
-50335
-50337
-50338
-50340
-50341
50342
50343
50344
@@ -45755,7 +45607,6 @@
50370
50371
50372
-50373
50374
50375
50376
@@ -45791,7 +45642,6 @@
50406
50407
50408
-50409
50410
50411
50412
@@ -45814,7 +45664,6 @@
50429
50430
50431
-50432
50433
50434
50435
@@ -45834,7 +45683,6 @@
50449
50450
50451
-50452
50453
50454
50455
@@ -45872,9 +45720,161 @@
50487
50488
50489
-50490
50491
50492
-50493
50494
50495
+50496
+50497
+50498
+50499
+50500
+50501
+50502
+50503
+50504
+50505
+50506
+50507
+50508
+50509
+50510
+50511
+50512
+50513
+50514
+50515
+50516
+50517
+50518
+50519
+50520
+50521
+50522
+50523
+50524
+50525
+50526
+50527
+50528
+50529
+50530
+50531
+50532
+50533
+50534
+50535
+50536
+50537
+50538
+50539
+50540
+50541
+50542
+50543
+50544
+50545
+50546
+50547
+50548
+50549
+50550
+50551
+50552
+50553
+50554
+50555
+50556
+50557
+50558
+50559
+50560
+50561
+50562
+50563
+50564
+50565
+50566
+50567
+50568
+50569
+50570
+50571
+50572
+50573
+50574
+50575
+50576
+50577
+50578
+50579
+50580
+50581
+50582
+50583
+50584
+50585
+50586
+50587
+50588
+50589
+50590
+50591
+50592
+50593
+50594
+50595
+50596
+50597
+50598
+50599
+50600
+50601
+50602
+50603
+50604
+50605
+50606
+50607
+50608
+50609
+50610
+50611
+50612
+50613
+50614
+50615
+50616
+50617
+50618
+50619
+50620
+50621
+50622
+50623
+50624
+50625
+50626
+50627
+50628
+50629
+50630
+50631
+50632
+50633
+50634
+50635
+50636
+50637
+50638
+50639
+50640
+50641
+50642
+50643
+50644
+50645
+50646
+
Re: "au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 [EMAIL PROTECTED] wrote: > [ .. snip .. ] > Actually you already specified udba=inotify, so it must be a bug. > Additionally the inode numbers #805306524, #134217865 and #134217862 > seem to be too large. Are they correct numbers? Generally speaking, they > are not illegal numbers, but I want to check whether they are correct or > not. > Please tell me about your branch 0. What filesystem is it? What is the > largest inode number on it? (Sorry about sending this email directly to you first, Junjiro, please ignore that) Thank you for your quick response. :) Branch 0 would be the first branch on the /proc/mounts-line? If so, that's /vault/disk3, which is xfs. How do I find largest inode number? The other branches are disk2: xfs, disk1: ext3, lvm: ext3. Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGc+3MwqQbW3my7pcRAquiAJ9zCL5mWxdrhAdLGFsPnzfWnifvKgCcDN3a XEkLpMKcxl8BJEAgCCB0U3M= =5TmO -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
"au_new_inode: broken ino"
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Using 20070528 release, I'm at random intervals getting "aufs au_new_inode:...: broken ino", e.g.: [ 349.301408] aufs au_new_inode:325:glftpd[3180]: broken ino, b0, files/3.Lbs, hi805306524, i1078. Try udba=inotify. Also: [631021.668293] aufs au_new_inode:325:twistd[27583]: broken ino, b0, libxtst/libxtst6_1.0.1-3build1_i386.deb, hi134217865, i48493. Try udba=inotify. [717395.849214] aufs au_new_inode:325:twistd[27583]: broken ino, b0, samba/samba-common_3.0.22-1ubuntu4.2_i386.deb, hi134217862, i48869. Try udba=inotify. Not the same path each time. Also, it says 'try udba=inotify': I'm already using it: [EMAIL PROTECTED]:~$ grep aufs /proc/mounts none /storage aufs rw,xino=/vault/disk3/.aufs.xino,udba=inotify,br:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw:/vault/lvm=rw 0 0 What's the cause? Is this a problem? Kindest regards, Jørgen P. Tjernø. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGcoVMwqQbW3my7pcRAtVBAJ437yDeq03fN/UEoBQJb0zRLIO3sgCgjyr5 0WRIett+qcVvUbwdyIaC5wI= =PKck -END PGP SIGNATURE- - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Possible bug (?) - hard to reproduce
Hiya. :-) I'm having some issues with a piece of software I've written. It works fine in general, but at my local file server I've got some problems. The thing that's acting up is a dir handle. I opendir() it, and I can read fine from it. A bit later in the programs execution, and readdir() returns NULL (right after I rewinddir()), and errno is not set. So, basically it's telling me the dir is empty - which it isn't. I've looked at a gazillion other causes, even run the whole program through valgrind - no errors there. The dir-pointer is correct when it's being read from, everything seems to be normal. I'm trying to figure out what's wrong, and I'm starting to consider two viable options: memory error or aufs. Could this be the doing of aufs? Any way to debug it on a bit lower level? Any suggestions? This is x86, linux 2.6.20.3, slightly old aufs: 20070402. In my dmesg, the only aufs-related messages are these: [275196.616129] aufs rmdir_whtmp:773:aufsd[2191]: failed removing .wh..wh.Mary j blige.0060(-21), ignored [275196.616178] aufs do_rmdir_whtmp:810:aufsd[2191]: I/O Error, err -21 But they occured a *long* time ago. Thanks in advance, sorry for not having any simple way of reproducing it. :-( Kindest regards, Jørgen Tjernø. - This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
Re: Stale NFS file handle (wtf?)
Jørgen P. Tjernø wrote: > I'm toying around, and I want to ls a dir on the aufs. > [EMAIL PROTECTED]:/storage/isos# ls > ls: .: Stale NFS file handle > [EMAIL PROTECTED]:/storage/isos# grep /storage /proc/mounts > none /storage aufs > rw,xino=/vault/disk3/.aufs.xino,udba=inotify,br:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw:/vault/lvm=rw > > 0 0 > [ .. snip .. ] What the ..? [EMAIL PROTECTED]:/storage/isos# ls . ls: .: Stale NFS file handle [EMAIL PROTECTED]:/storage/isos# ls /storage/isos Strange. :-) Kindest regards, Jørgen P. Tjernø. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Stale NFS file handle (wtf?)
I'm toying around, and I want to ls a dir on the aufs. [EMAIL PROTECTED]:/storage/isos# ls ls: .: Stale NFS file handle [EMAIL PROTECTED]:/storage/isos# grep /storage /proc/mounts none /storage aufs rw,xino=/vault/disk3/.aufs.xino,udba=inotify,br:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw:/vault/lvm=rw 0 0 [EMAIL PROTECTED]:/storage/isos# /etc/init.d/nfs-kernel-server stop * Stopping NFS kernel daemon [ ok ] * Unexporting directories for NFS kernel daemon... [ ok ] [EMAIL PROTECTED]:/storage/isos# ls ls: .: Stale NFS file handle It's obviously not the nfs causing the problems. since it's not running. Nothing interesting in dmesg. Compiled using gcc 3.4, local.mk and kernel v2.6.20.3. Not using latest daily, will recompile to that and reload module later on. I'm guessing that will solve the problem temporarily, but I've got no clue what's causing this. Doing ls on a bind --mount of /storage/isos shows the correct directory listing. Any clues as to what could cause this? Should it be fixed in the tuesday update? Kindest regards Jørgen P. Tjernø. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: aufs Tuesday release (Re: skip aufs Monday release)
Tomas M wrote: > Hello, > I think the Tuesday's release can't be compiled. > At least it doesn't work for me. > > Using make -f local.mk, no other changes. Kernel 2.6.20: > [ .. snip .. ] > > Is my environment broken or is it aufs-compile problem? > > > Thank you very much > > Tomas M > slax.org Confirmed, setup: compiler: gcc (GCC) 4.1.2 20060928 (prerelease) (Ubuntu 4.1.1-13ubuntu5) kernel : vanilla 2.6.20.3 options : default local.mk, except for CONFIG_AUFS_HINOTIFY = y Same error: /root/aufs.wcvs/aufs/fs/aufs/misc.h: In function ‘au_is_kthread’: /root/aufs.wcvs/aufs/fs/aufs/misc.h:150: error: dereferencing pointer to incomplete type - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> Well, the bug still occurs: > ::: >> [78022.858961] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/iinfo.c:62! > ::: >> Any ideas? > > Unfortunately, I have no idea. > Since there is no reason to stop at iinfo.c:62, I must check your > environment again. But your kernel is plain and no symbol conflict with > unionfs. > > Additionally, > >> [78022.915130] Call Trace: >> [78022.921074] [] test_ro+0x32/0x60 [aufs] >> [78022.924274] [] hidden_open+0x7b/0x110 [aufs] >> [78022.927381] [] do_open_dir+0xa9/0x120 [aufs] >> [78022.930578] [] au_do_open+0x9c/0x110 [aufs] >> [78022.933781] [] do_open_dir+0x0/0x120 [aufs] >> [78022.936988] [] aufs_open_dir+0x0/0x10 [aufs] >> [78022.940198] [] __dentry_open+0xba/0x1c0 >> [78022.943339] [] nameidata_to_filp+0x35/0x40 >> [78022.946403] [] do_filp_open+0x4b/0x60 >> [78022.949555] [] do_sys_open+0x4a/0xe0 >> [78022.952698] [] sys_open+0x1c/0x20 > > this trace is strange. But you might disable CONFIG_FRAME_POINTER. If so > that may be reasonable. > > If your GCC is version 4, try version 3. > And I suggest you to enable CONFIG_DEBUG_KERNEL, CONFIG_DEBUG_SLAB, > CONFIG_FRAME_POINTER, and CONFIG_DEBUG_STACKOVERFLOW. $ gcc --version gcc (GCC) 4.1.2 20060928 (prerelease) (Ubuntu 4.1.1-13ubuntu5) $ for i in DEBUG_KERNEL DEBUG_SLAB FRAME_POINTER DEBUG_STACKOVERFLOW; do > grep CONFIG_$i /usr/src/linux/.config > done CONFIG_DEBUG_KERNEL=y # CONFIG_DEBUG_SLAB is not set # CONFIG_FRAME_POINTER is not set # CONFIG_DEBUG_STACKOVERFLOW is not set I will try setting those three other options and compiling with gcc-3.4 :-) I will let you know if the problem appears again. I wish I knew how to reproduce them, but they just happen during normal use. :-) Again, thank you for your swift and helpful response. Kindest regards, Jørgen P. Tjernø. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
BUG @ hinotify.c:123!
As I was uploading some files via FTP to /storage, I got a kernel BUG in my dmesg. No UnionFS support enabled or loaded. Also, notice it *hasn't* been mounted with udba=inotify! :o $ grep /storage /proc/mounts none /storage aufs rw,xino=/vault/disk3/.aufs.xino,br:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw:/vault/lvm=rw 0 0 $ uname -r 2.6.20.3 $ lsmod Module Size Used by aufs 122624 1 $ [14664.831134] [ cut here ] [14664.834300] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/hinotify.c:123! [14664.837355] invalid opcode: [#1] [14664.840433] SMP [14664.843462] Modules linked in: aufs [14664.846574] CPU:0 [14664.846579] EIP:0060:[]Not tainted VLI [14664.846586] EFLAGS: 00010202 (2.6.20.3 #5) [14664.855785] EIP is at do_hdir_lock+0x48/0x50 [aufs] [14664.858908] eax: c20c1e70 ebx: cdf6d520 ecx: edx: [14664.862121] esi: c4eedb7c edi: caf598b4 ebp: c4524dd4 esp: c8279e74 [14664.865334] ds: 007b es: 007b ss: 0068 [14664.868385] Process zipscript-c (pid: 7318, ti=c8278000 task=c8776ab0 task.ti=c8278000) [14664.871613] Stack: c45240d4 d0c2fb10 0006 c8279ee8 [14664.877965]c20c1e90 c4eedb7c c20c1e90 caf598b4 c8279f14 c20c1e90 c3fd3000 [14664.884320]d0c2fd0a c45240d4 030081b6 caf598d0 0002 0003 [14664.890517] Call Trace: [14664.896465] [] lock_hdir_lkup_wh+0x80/0x100 [aufs] [14664.899707] [] add_simple+0x4a/0x280 [aufs] [14664.902759] [] di_read_unlock+0x35/0x70 [aufs] [14664.905972] [] aufs_read_unlock+0x15/0x30 [aufs] [14664.909192] [] aufs_symlink+0x1d/0x30 [aufs] [14664.912405] [] vfs_symlink+0xa5/0xf0 [14664.915559] [] sys_symlinkat+0xb2/0xe0 [14664.918646] [] remove_vma+0x39/0x50 [14664.921735] [] sys_symlink+0x1f/0x30 [14664.924878] [] sysenter_past_esp+0x5d/0x81 [14664.928048] === [14664.931125] Code: 0f be d3 8d 14 52 8d 1c 95 00 00 00 00 03 58 18 39 33 75 17 8d 46 70 e8 27 55 88 ef 89 d8 31 d2 5b 5e e9 9c fe ff ff 0f 0b eb fe <0f> 0b eb fe 8d 74 26 00 55 57 89 d7 56 89 c6 53 8b 02 89 cb 8b [14664.941608] EIP: [] do_hdir_lock+0x48/0x50 [aufs] SS:ESP 0068:c8279e74 - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
Jørgen P. Tjernø wrote: > [EMAIL PROTECTED] wrote: >> "Jrgen_P._Tjern": >>> It has also been remounted using mount --bind to 6 other places. I've = >>> >>> tried without the unionfs module loaded now, and there seems to be no = >>> >>> problem currently, but dmesg has the following: >> Currently, I don't think that the bound mount is a problem. >> But you still have strange aufs behaviour. Won't you try vanilla kernel, >> if you can? > Now I'm running a vanilla 2.6.20.3-kernel from kernel.org, no patches, > and newest aufs, no patches. (make -f local.mk) > Took a bit of time to fix it, had some problems with my ADSL2+-modem > dieing and such, then some problems getting the kernel to boot properly. > > Since the only way I've been able to produce the bug is by "normal > usage", I'll let it run now, and send a mail here if I can reproduce it. > If not, I guess we can dismiss the bug as being caused by the ubuntu > kernel patchset. :-) Well, the bug still occurs: [61860.305642] aufs dtime_revert:151:smbd[4450]: restoring timestamps failed(-1). ignored [61860.305768] aufs dtime_revert:151:smbd[4450]: restoring timestamps failed(-1). ignored [78022.855799] [ cut here ] [78022.858961] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/iinfo.c:62! [78022.861984] invalid opcode: [#1] [78022.865061] SMP [78022.868093] Modules linked in: aufs [78022.871206] CPU:0 [78022.871211] EIP:0060:[]Not tainted VLI [78022.871217] EFLAGS: 00010206 (2.6.20.3 #5) [78022.880434] EIP is at h_iptr_i+0x3f/0x80 [aufs] [78022.883545] eax: ebx: c0744068 ecx: 0001 edx: 0003 [78022.886758] esi: 0003 edi: c0744068 ebp: cf79a720 esp: cf34de70 [78022.889971] ds: 007b es: 007b ss: 0068 [78022.893065] Process find (pid: 11246, ti=cf34c000 task=c1a6e050 task.ti=cf34c000) [78022.896230] Stack: 0003 d0c33402 cf6fc800 c20f3904 cfbf7b94 d0c2adeb 00018800 [78022.902583]cfbf7e14 0003 cf79a738 0003 0003 c924e4a0 cfbf7e14 d0c2cb59 [78022.908935] 00018800 03002d2e cf36061c cfbf7e14 cf364a24 [78022.915130] Call Trace: [78022.921074] [] test_ro+0x32/0x60 [aufs] [78022.924274] [] hidden_open+0x7b/0x110 [aufs] [78022.927381] [] do_open_dir+0xa9/0x120 [aufs] [78022.930578] [] au_do_open+0x9c/0x110 [aufs] [78022.933781] [] do_open_dir+0x0/0x120 [aufs] [78022.936988] [] aufs_open_dir+0x0/0x10 [aufs] [78022.940198] [] __dentry_open+0xba/0x1c0 [78022.943339] [] nameidata_to_filp+0x35/0x40 [78022.946403] [] do_filp_open+0x4b/0x60 [78022.949555] [] do_sys_open+0x4a/0xe0 [78022.952698] [] sys_open+0x1c/0x20 [78022.955829] [] sysenter_past_esp+0x5d/0x81 [78022.959002] === [78022.961924] Code: 50 ef 85 c0 75 2d 89 d8 e8 bf fc ff ff e8 6a 3e 50 ef 85 c0 75 19 89 f0 84 c0 89 f6 78 0d 89 d8 e8 07 ff ff ff 89 f2 38 c2 7e 0c <0f> 0b eb fe 0f 0b eb fe 0f 0b eb fe 89 d8 e8 8e fc ff ff 89 f1 [78022.972494] EIP: [] h_iptr_i+0x3f/0x80 [aufs] SS:ESP 0068:cf34de70 Any ideas? Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: kernel panic / BUG
[EMAIL PROTECTED] wrote: > That was my fault. Sorry. > I've fixed just a few minutes before, please update your CVS tree. Ah, thanks. Applied. :-) For future reference; is there any way to get the kernel panic log? :) Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
kernel panic / BUG (unkown cause, smb?)
Hi! I've gotten atleast a couple of kernel panics on my server, but I can't get the full backtrace and such - when I connect a keyboard, it just blinks the num/caps/scroll LEDs, and refuses to accept any kind of page up / page down, so all I can see is the bottom of the call stack and a few messages. I've faithfully typed down the messages from the monitor, in the hope that it'll help you debug it. :) The below lines are repeated upwards till the top of the screen: (the callstack, i.e. it repeats those 4 calls a *lot* of times) [] vfs_statfs+0x68/0x80 [] vfsub_statfs+0xf/0x40 [aufs] [] di_read_lock+0x2c/0x80 [aufs] [] aufs_statfs+x02a/0x70 [aufs] [] vfs_statfs+0x68/0x80 [] vfsub_statfs+0xf/0x40 [aufs] BUG: unable to handle kernel paging request at virtual address 693c printing eip: c0103f92 *pde = Recursive die() failure, output suppressed <0>Kernel panic - not syncing: Fatal exception in interrupt. Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Possible to 'invalidate cache'?
[EMAIL PROTECTED] wrote: >> BUG at /root/aufs.wcvs/aufs/fs/aufs/super.c:357! >> must be a bug in aufs. I will fix it as soon as possible. > > Please apply this patch, which is against the latest source in CVS. > If you success, it will be included in next release. > > This patch is only for BUG at super.c:357 only. I applied the patch, and it seems to have worked great. But I also tried reproducing the bug with the old source, and it doesn't seem to work. That is, I can't get a new BUG at super.c:357. I mount /storage, do a mount -o remount,udba=inotify /storage, and it seems to work. (NOTE: this is this weeks monday release, *not* the version I was running when I first stumbled upon BUG at super.c:357) So either there were other conditions to cause my BUG @ super.c:357, or the newest CVS fixed the problem by itself. In any case, thanks a lot. :-) Kindest regards, Jørgen Tjernø. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> I've changed my kernel, and the bug still happens, see attached >> aufs.log. I'm running the newest cvs from 18th march (i.e. not the one > > Did you link unionfs statically? > If so, there is a poosibility of symbol confilict. No, not loaded nor compiled unionfs for this kernel. :-) Kindest regards, Jørgen. Ps. I'll upgrade my aufs now. :) Pss. If the mail directly to you came through; disregard it. (I tried stopping it before it sent it, but I'm not certain I made it) - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
Jørgen P. Tjernø wrote: I'm not 100% certain what caused this, but as I mounted my samba-shares (which is an exported aufs) and tried accessing them, I got the following: ls: reading directory /home/daxxar/samba/videos.datamania: Input/output error So I tried checking on datamania if I could find the problem, and I found the following in my dmesg: I've changed my kernel, and the bug still happens, see attached aufs.log. I'm running the newest cvs from 18th march (i.e. not the one you just released), and vanilla 2.6.20.3-kernel. unmodified aufs. compiled using make -f local.mk, INOTIFY support enabled - but not mounted with it. Relevant /proc/mounts: /dev/mapper/main-storage /vault/lvm ext3 rw,data=ordered 0 0 /dev/disk/by-uuid/d61a06e4-2830-44e6-95ba-5e54e64a9736 /vault/disk1 ext3 rw,data=ordered 0 0 /dev/disk/by-uuid/e2bd2748-1050-4fd2-8c98-e90a6dc863e2 /vault/disk2 xfs rw 0 0 /dev/disk/by-uuid/2e35bd30-8879-4127-8ec0-1a2dcbdceca8 /vault/disk3 xfs rw 0 0 none /storage aufs rw,xino=/vault/disk3/.aufs.xino,br:/vault/disk3=rw:/vault/disk2=rw:/vault/disk1=rw:/vault/lvm=rw 0 0 To my knowledge there wasn't much activity on the system at the time, the closest sign of activity (the kernel BUG happened at 6.25AM according to syslog) was the following in a sambalog: write_data: write failure in writing to client 192.168.0.21. Error Broken pipe I have no other ideas as to what could've triggered it. :-( Kindest regards, Jørgen. [26065.464684] [ cut here ] [26065.467726] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/iinfo.c:62! [26065.470901] invalid opcode: [#1] [26065.473979] SMP [26065.477012] Modules linked in: aufs [26065.480125] CPU:0 [26065.480130] EIP:0060:[]Not tainted VLI [26065.480137] EFLAGS: 00010206 (2.6.20.3 #5) [26065.489349] EIP is at h_iptr_i+0x3f/0x80 [aufs] [26065.492461] eax: ebx: c0fa7638 ecx: 0001 edx: 0003 [26065.495672] esi: 0003 edi: c0fa7638 ebp: cef90760 esp: c20d5e74 [26065.498885] ds: 007b es: 007b ss: 0068 [26065.501833] Process find (pid: 6100, ti=c20d4000 task=c239f580 task.ti=c20d4000) [26065.505103] Stack: 0003 d0c32c52 cedc0800 ce738c5c cd8ca254 d0c2a80b 00018800 [26065.511459]cd8cae54 0003 cef90778 0003 0003 c2fa0640 cd8cae54 d0c2c569 [26065.517713] 00018800 0300257e cdfc661c cd8cae54 ce726a24 [26065.524023] Call Trace: [26065.529968] [] test_ro+0x32/0x60 [aufs] [26065.533167] [] hidden_open+0x7b/0x110 [aufs] [26065.536229] [] do_open_dir+0xa9/0x120 [aufs] [26065.539439] [] do_open+0x96/0x110 [aufs] [26065.542631] [] do_open_dir+0x0/0x120 [aufs] [26065.545836] [] aufs_open_dir+0x0/0x10 [aufs] [26065.549042] [] __dentry_open+0xba/0x1c0 [26065.552054] [] nameidata_to_filp+0x35/0x40 [26065.555221] [] do_filp_open+0x4b/0x60 [26065.558372] [] do_sys_open+0x4a/0xe0 [26065.561516] [] sys_open+0x1c/0x20 [26065.564647] [] sysenter_past_esp+0x5d/0x81 [26065.567726] === [26065.570761] Code: 50 ef 85 c0 75 2d 89 d8 e8 bf fc ff ff e8 1a 46 50 ef 85 c0 75 19 89 f0 84 c0 89 f6 78 0d 89 d8 e8 07 ff ff ff 89 f2 38 c2 7e 0c <0f> 0b eb fe 0f 0b eb fe 0f 0b eb fe 89 d8 e8 8e fc ff ff 89 f1 [26065.581338] EIP: [] h_iptr_i+0x3f/0x80 [aufs] SS:ESP 0068:c20d5e74 [26065.593937] - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
BUG: remount / udba=inotify
I tried a simple umount -o remount,udba=inotify /storage, and the following popped up in my dmesg (see attached file). Not certain if mounting (i.e. not -o remount) with udba=inotify works, can't test it right now. SYSINFO ~~~ kernel : 2.6.20.3 (Vanilla, inotify-support) aufs : latest cvs as of 18. march 2007 aufs compiled: make -f local.mk, w/ CONFIG_AUFS_HINOTIFY = y Kindest regards, Jørgen. [ 121.467484] [ cut here ] [ 121.470667] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/super.c:357! [ 121.473883] invalid opcode: [#1] [ 121.476984] SMP [ 121.479900] Modules linked in: aufs [ 121.483050] CPU:0 [ 121.483054] EIP:0060:[]Not tainted VLI [ 121.483061] EFLAGS: 00010246 (2.6.20.3 #5) [ 121.492459] EIP is at aufs_remount_fs+0x2aa/0x360 [aufs] [ 121.495548] eax: cf3dc2ac ebx: cf9daf14 ecx: 0002 edx: 0005 [ 121.498767] esi: cf9daf4c edi: cfa37754 ebp: cf9daf4c esp: c8851da4 [ 121.502002] ds: 007b es: 007b ss: 0068 [ 121.505130] Process mount (pid: 3227, ti=c885 task=c885a580 task.ti=c885) [ 121.508398] Stack: 0001 cf757e84 2348 cf757e84 2348 cf757e00 ce555ec0 [ 121.514737]cfcf0860 0003 0005 cf60b994 0005 0001 0005 [ 121.521177]0020 0001 c8819000 002a d0c3d6a0 cf757e00 [ 121.527617] Call Trace: [ 121.533467] [] do_remount_sb+0xbd/0x120 [ 121.536670] [] do_mount+0x246/0x6f0 [ 121.539849] [] get_page_from_freelist+0x27c/0x340 [ 121.543087] [] __handle_mm_fault+0x52d/0x910 [ 121.546160] [] __alloc_pages+0x57/0x310 [ 121.549346] [] error_code+0x7c/0x84 [ 121.552513] [] copy_mount_options+0x40/0x150 [ 121.555719] [] sys_mount+0x77/0xc0 [ 121.558885] [] sysenter_past_esp+0x5d/0x81 [ 121.562032] [] svcauth_gss_accept+0x453/0xb50 [ 121.565174] === [ 121.568266] Code: 8b 47 0c 85 c0 74 b8 0f b7 40 6a 25 00 f0 00 00 3d 00 40 00 00 75 a8 3b 7f 18 74 0f 89 f8 e8 ce 7e 00 00 8b 54 24 28 3b 10 75 7d <0f> 0b eb fe 89 f6 89 d8 e8 b9 7e 00 00 8b 54 24 28 3b 10 74 a7 [ 121.579314] EIP: [] aufs_remount_fs+0x2aa/0x360 [aufs] SS:ESP 0068:c8851da4 [ 121.585614] BUG: at kernel/exit.c:860 do_exit() [ 121.588856] [] do_exit+0x757/0x7e0 [ 121.592060] [] oops_may_print+0xb/0x10 [ 121.595228] [] do_trap+0x0/0xf0 [ 121.598304] [] do_invalid_op+0x0/0xb0 [ 121.601504] [] do_invalid_op+0xa2/0xb0 [ 121.604707] [] aufs_remount_fs+0x2aa/0x360 [aufs] [ 121.607984] [] alloc_hinotify+0x52/0x90 [aufs] [ 121.611245] [] set_h_iptr+0xf3/0x1e0 [aufs] [ 121.614345] [] iput+0x35/0x70 [ 121.617513] [] reset_hinotify+0x7d/0xa0 [aufs] [ 121.620783] [] error_code+0x7c/0x84 [ 121.623979] [] __kill_fasync+0x3b/0x50 [ 121.627176] [] aufs_remount_fs+0x2aa/0x360 [aufs] [ 121.630294] [] do_remount_sb+0xbd/0x120 [ 121.633505] [] do_mount+0x246/0x6f0 [ 121.636700] [] get_page_from_freelist+0x27c/0x340 [ 121.639950] [] __handle_mm_fault+0x52d/0x910 [ 121.643166] [] __alloc_pages+0x57/0x310 [ 121.646255] [] error_code+0x7c/0x84 [ 121.649443] [] copy_mount_options+0x40/0x150 [ 121.652673] [] sys_mount+0x77/0xc0 [ 121.655860] [] sysenter_past_esp+0x5d/0x81 [ 121.659062] [] svcauth_gss_accept+0x453/0xb50 [ 121.662262] === - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Possible to 'invalidate cache'?
I've got a problem I stumble upon with certain intervals; I've copied some files using samba to a directory that exists on a branch that's running out of space. So, I want to put files on another branch, and create the directory structure neccessary in another branch. example: /storage is a merge of /vault/disk1 and /vault/disk2, both rw. I want to put files under /storage/misc on disk2, but disk2 has the directory structure already (/vault/disk2/misc). So I create /vault/disk1/misc. Now, the problem is, /storage does not notice this! When I now copy files to /storage/misc, it still puts them on disk2! This is my example /proc/mounts (as you can see, disk1 has the highest priority, and should be used). none /storage aufs rw,xino=/vault/disk1/.aufs.xino,br:/vault/disk1=rw:/vault/disk2=rw 0 0 Any way I can manually invalidate whatever cache is used to perform the aufs -> branch lookup? Or should I use udba=inotify? Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> It has also been remounted using mount --bind to 6 other places. I've = >> >> tried without the unionfs module loaded now, and there seems to be no = >> >> problem currently, but dmesg has the following: > > Currently, I don't think that the bound mount is a problem. > But you still have strange aufs behaviour. Won't you try vanilla kernel, > if you can? Now I'm running a vanilla 2.6.20.3-kernel from kernel.org, no patches, and newest aufs, no patches. (make -f local.mk) Took a bit of time to fix it, had some problems with my ADSL2+-modem dieing and such, then some problems getting the kernel to boot properly. Since the only way I've been able to produce the bug is by "normal usage", I'll let it run now, and send a mail here if I can reproduce it. If not, I guess we can dismiss the bug as being caused by the ubuntu kernel patchset. :-) Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> [EMAIL PROTECTED]:~$ grep /storage /proc/mounts >> none /storage aufs = >> >> rw,xino=3D/vault/sdd/.aufs.xino,br:/vault/sdd=3Drw:/vault/sda=3Drw:/vault/h= >> dc=3Drw:/vault/lvm=3Drw = >> >> 0 0 > > Thanks. > I recognized that a bug around the maximum number of branches is not > related to your problem. That's good news atleast, with only .. 4 branches ;-) >> It has also been remounted using mount --bind to 6 other places. I've = >> >> tried without the unionfs module loaded now, and there seems to be no = >> >> problem currently, but dmesg has the following: > > Currently, I don't think that the bound mount is a problem. > But you still have strange aufs behaviour. Won't you try vanilla kernel, > if you can? Okay, just thought I'd mention it. I'll try compiling a vanilla kernel as soon as I get home, to my knowledge I do not use any of the ubuntu-specific options in the kernel. :-) Then I'd get away from the strange mount-bug too! Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: Samba / AUFS / kernel BUG iinfo.c:62
[EMAIL PROTECTED] wrote: > And how many branches did you add into your aufs? > > Junjiro Okajima [EMAIL PROTECTED]:~$ grep /storage /proc/mounts none /storage aufs rw,xino=/vault/sdd/.aufs.xino,br:/vault/sdd=rw:/vault/sda=rw:/vault/hdc=rw:/vault/lvm=rw 0 0 It has also been remounted using mount --bind to 6 other places. I've tried without the unionfs module loaded now, and there seems to be no problem currently, but dmesg has the following: [49417.002933] [ cut here ] [49417.002979] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/vdir.c:605! [49417.003008] invalid opcode: [#1] [49417.003029] SMP [49417.003054] Modules linked in: nfsd exportfs lockd sunrpc bridge aufs xfs dm_mod md_mod fuse lp i2c_piix4 psmouse evdev sg serio_raw floppy parport_pc parport i2c_core pcspkr r1000 r8169 3c59x mii shpchp pci_hotplug intel_agp agpgart ext3 jbd uhci_hcd usbcore ide_generic sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcursor vesafb capability commoncap [49417.003328] CPU:0 [49417.00] EIP:0060:[]Not tainted VLI [49417.003339] EFLAGS: 00010286 (2.6.17-11-generic #2) [49417.003473] EIP is at init_vdir+0x698/0x730 [aufs] [49417.003503] eax: ebx: ecx: ca48d6a0 edx: c693ade0 [49417.003536] esi: 0001 edi: c693ae00 ebp: cfe44c00 esp: c794bef4 [49417.003565] ds: 007b es: 007b ss: 0068 [49417.003593] Process find (pid: 10860, threadinfo=c794a000 task=cd985560) [49417.003611] Stack: 0002 c2f4dc80 d0b96794 c693ae00 0005 0005 [49417.003680]d0b97360 c693ade0 cd985560 cd985560 cfe44c00 c02da7b5 c693ade8 [49417.003750]00100100 00200200 0002 c693ade0 c693ade0 c693ae00 c2f4dc80 [49417.003819] Call Trace: [49417.003859] .text.lock.dinfo+0x9e/0xda [aufs] reval_and_lock_finfo+0xa0/0x6fd [aufs] [49417.004007] rwsem_down_write_failed+0x165/0x180 aufs_readdir+0x8d/0x140 [aufs] [49417.004143] filldir64+0x0/0xe0 filldir64+0x0/0xe0 [49417.004217] vfs_readdir+0x94/0xb0 sys_getdents64+0x6f/0xc0 [49417.004305] sysenter_past_esp+0x54/0x79 [49417.004372] Code: d6 f9 ff ff 0f 0b 5b 02 38 2b ba d0 e9 dd f9 ff ff 0f 0b 5d 02 38 2b ba d0 e9 f9 f9 ff ff 0f 0b 5d 02 38 2b ba d0 e9 05 fa ff ff <0f> 0b 5d 02 38 2b ba d0 e9 0c fa ff ff 0f 0b 5e 02 38 2b ba d0 [49417.004636] EIP: [] init_vdir+0x698/0x730 [aufs] SS:ESP 0068:c794bef4 [49417.005106] I'll revert my copy to current cvs and only apply the mount-patch :) Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Samba / AUFS / kernel BUG iinfo.c:62
I'm not 100% certain what caused this, but as I mounted my samba-shares (which is an exported aufs) and tried accessing them, I got the following: ls: reading directory /home/daxxar/samba/videos.datamania: Input/output error So I tried checking on datamania if I could find the problem, and I found the following in my dmesg: [958904.932570] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:506/alloc_sbinfo() [958904.932615] aufs_fill_super+0x86/0x630 [aufs] get_filesystem+0xf/0x40 [958904.932751] get_sb_nodev+0x40/0x80 vfs_kern_mount+0xb0/0x150 [958904.932801] aufs_fill_super+0x0/0x630 [aufs] do_kern_mount+0x39/0x60 [958904.932883] do_mount+0x450/0x760 get_page_from_freelist+0x3a2/0x410 [958904.932953] __handle_mm_fault+0x4ea/0x900 __alloc_pages+0x56/0x330 [958904.933085] copy_mount_options+0x40/0x150 sys_mount+0x77/0xc0 [958904.933143] sysenter_past_esp+0x54/0x79 [958911.271223] sb c13c9000, sbinfo c1e61ac0 [958911.271250] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:243/aufs_umount_begin() [958911.271290] aufs_umount_begin+0x52/0x9f [aufs] sys_umount+0x110/0x2c0 [958911.271433] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958911.271504] sysenter_past_esp+0x54/0x79 [958911.275059] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:274/aufs_put_super() [958911.275090] aufs_put_super+0x2f/0x50 [aufs] generic_shutdown_super+0x87/0x140 [958911.275191] deactivate_super+0x6e/0x90 sys_umount+0x4a/0x2c0 [958911.275248] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958911.275308] sysenter_past_esp+0x54/0x79 [958911.275355] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:259/free_sbinfo() [958911.275371] free_sbinfo+0x3b/0x70 [aufs] generic_shutdown_super+0x87/0x140 [958911.275450] deactivate_super+0x6e/0x90 sys_umount+0x4a/0x2c0 [958911.275506] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958911.275563] sysenter_past_esp+0x54/0x79 [958916.443795] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:506/alloc_sbinfo() [958916.443848] aufs_fill_super+0x86/0x630 [aufs] get_filesystem+0xf/0x40 [958916.443991] get_sb_nodev+0x40/0x80 vfs_kern_mount+0xb0/0x150 [958916.444042] aufs_fill_super+0x0/0x630 [aufs] do_kern_mount+0x39/0x60 [958916.444125] do_mount+0x450/0x760 get_page_from_freelist+0x3a2/0x410 [958916.444194] __handle_mm_fault+0x4ea/0x900 __alloc_pages+0x56/0x330 [958916.444324] copy_mount_options+0x40/0x150 sys_mount+0x77/0xc0 [958916.444382] sysenter_past_esp+0x54/0x79 [958983.382473] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:506/alloc_sbinfo() [958983.382525] aufs_fill_super+0x86/0x630 [aufs] get_filesystem+0xf/0x40 [958983.382667] get_sb_nodev+0x40/0x80 vfs_kern_mount+0xb0/0x150 [958983.382716] aufs_fill_super+0x0/0x630 [aufs] do_kern_mount+0x39/0x60 [958983.382797] do_mount+0x450/0x760 get_page_from_freelist+0x3a2/0x410 [958983.382864] __handle_mm_fault+0x4ea/0x900 __alloc_pages+0x56/0x330 [958983.382988] copy_mount_options+0x40/0x150 sys_mount+0x77/0xc0 [958983.383045] sysenter_past_esp+0x54/0x79 [958985.516819] sb cf5a5c00, sbinfo cf310520 [958985.516847] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:243/aufs_umount_begin() [958985.516886] aufs_umount_begin+0x52/0x9f [aufs] sys_umount+0x110/0x2c0 [958985.517028] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958985.517098] sysenter_past_esp+0x54/0x79 [958985.520623] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:274/aufs_put_super() [958985.520655] aufs_put_super+0x2f/0x50 [aufs] generic_shutdown_super+0x87/0x140 [958985.520754] deactivate_super+0x6e/0x90 sys_umount+0x4a/0x2c0 [958985.520812] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958985.520872] sysenter_past_esp+0x54/0x79 [958985.520919] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:259/free_sbinfo() [958985.520935] free_sbinfo+0x3b/0x70 [aufs] generic_shutdown_super+0x87/0x140 [958985.521015] deactivate_super+0x6e/0x90 sys_umount+0x4a/0x2c0 [958985.521070] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [958985.521127] sysenter_past_esp+0x54/0x79 [958985.861527] sb c13c9000, sbinfo cce83e20 [958985.861553] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:243/aufs_umount_begin() [958985.861591] aufs_umount_begin+0x52/0x9f [aufs] sys_umount+0x110/0x2c0 [958985.861733] do_page_fault+0x3db/0x6f0 sys_oldumount+0x17/0x20 [958985.861803] sysenter_past_esp+0x54/0x79 [958985.863663] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:274/aufs_put_super() [958985.863691] aufs_put_super+0x2f/0x50 [aufs] generic_shutdown_super+0x87/0x140 [958985.863785] deactivate_super+0x6e/0x90 sys_umount+0x4a/0x2c0 [958985.863842] do_page_fault+0x3db/0x6f0 sys_oldumount+0x17/0x20 [958985.863900] sysenter_past_esp+0x54/0x79 [958985.863947] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:259/free_sbinfo() [958985.863962] free_sbinfo+0x3b
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> I had indeed stopped using aufs for the time being, as I needed my = >> >> system up and running, but now I see why you're developing aufs; for me, = >> >> unionfs had serious issues with samba-shares and such. :-) > > Honestly speaking, I have never tried a samba branch. :-) > But I think it should work. Sorry I was a bit unclear, not a samba branch, but sharing the whole aufs over samba, so e.g. windows-clients can access the aufs. :-) Jørgen P. Tjernø. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > Hello Jorgen, > >> Since you didn't seem to understand what I wrote. >> I suggest you one more thing. >> - build your kernel by yourself. >> If you succeed it, then you have a correct environment. > > While I don't know that you are still trying aufs, I have found one > thing about ubuntu edgy kernel. > > The version of this kernel seems to be 2.6.17 basically, but one > interface (at least) is patched by a part of 2.6.18. It is the arguments > of super_block operation umount_begin(). > > Since aufs supports both of 2.6.17 and 18, it uses, > #if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18) > /* code for linux-2.6.17 and earlier */ > #endif > In ubuntu edgy kernel, this condition didn't work. > Since the LINUX_VERSION_CODE in ubuntu edgy is defineed as 2.6.17, but a > part of code/interface is 2.6.18 actually. Perhaps I should just compile a vanilla kernel. I hardly have any spectacular hardware on this machine. ;-) > Here is a patch for it. > But I am afraid you may meet another problem which is related to the > kernel interface which is mixed with 2.6.17 and 2.6.18 in ubuntu edgy. Thanks a lot! :-) I had indeed stopped using aufs for the time being, as I needed my system up and running, but now I see why you're developing aufs; for me, unionfs had serious issues with samba-shares and such. :-) I'm now running latest cvs + that patch, and umounting works great! I'll let you know if I stumble into any other problems. ;-) Thanks for your work, appreciate it :-) Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> Sure! See the attached aufs.log. :-) > > Thank you. > It seems your environment is broken. > >> [135136.953541] BUG: warning at >> /root/aufs.wcvs/aufs/fs/aufs/super.c:503/alloc_sbinfo() > ::: >> [135147.517991] sb cea18e40, sbinfo >> [135147.518018] BUG: warning at >> /root/aufs.wcvs/aufs/fs/aufs/super.c:240/aufs_umount_begin() > ::: > > This shows, > - a memory area (sbinfo) is allocated correctly and aufs uses it. > - aufs doesn't free it. > - but it is freed by someone, somewhere. and aufs cannot work. > > It may be possible to check whether the pointer is NULL or not, and stop > crashing at umounting aufs. But I am afraid soon you will meet another > problem. > I suggest you to compile other module by yourself and test it, or to > check the compile environment. Unloading uncessary modules may be > meaningful. Are you saying the default precompiled modules from Ubuntu are broken? :O (Again, sorry for sending it directly to you, darn email-software) Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: "Jrgen_P._Tjern": I thought CONFIG_AUFS_LHASH_PATCH was for using NFS-filesystems as a = branch of your aufs? I haven't touched AUFS_LHASH_PATCH, I just fetched = source and did make -f local.mk. :-) Ok. You didn't apply kernel patches from aufs and used the default aufs configuration. That's correct. :) Here are the files for the kernel: Thank you. I will check them. And will you apply this debug patch and send the log? Sure! See the attached aufs.log. :-) Kindest regards, Jørgen P. Tjernø. [135136.952027] aufs 20070226 [135136.953541] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:503/alloc_sbinfo() [135136.953587] aufs_fill_super+0x86/0x630 [aufs] get_filesystem+0xf/0x40 [135136.953703] get_sb_nodev+0x40/0x80 vfs_kern_mount+0xb0/0x150 [135136.953751] aufs_fill_super+0x0/0x630 [aufs] do_kern_mount+0x39/0x60 [135136.953829] do_mount+0x450/0x760 get_page_from_freelist+0x3a2/0x410 [135136.953890] __handle_mm_fault+0x4ea/0x900 __alloc_pages+0x56/0x330 [135136.953999] copy_mount_options+0x40/0x150 sys_mount+0x77/0xc0 [135136.954051] sysenter_past_esp+0x54/0x79 [135147.517991] sb cea18e40, sbinfo [135147.518018] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:240/aufs_umount_begin() [135147.518057] aufs_umount_begin+0x51/0xb0 [aufs] sys_umount+0x110/0x2c0 [135147.518191] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [135147.518259] sysenter_past_esp+0x54/0x79 [135147.521758] BUG: unable to handle kernel NULL pointer dereference at virtual address [135147.521816] printing eip: [135147.521837] d0bf21f7 [135147.521844] *pde = [135147.521868] Oops: 0002 [#1] [135147.521888] SMP [135147.521912] Modules linked in: aufs nfsd exportfs lockd sunrpc bridge unionfs xfs dm_mod md_mod fuse lp pcspkr i2c_piix4 psmouse i2c_core serio_raw parport_pc parport evdev shpchp pci_hotplug sg r1000 r8169 3c59x mii floppy intel_agp agpgart ext3 jbd uhci_hcd usbcore ide_generic sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcursor vesafb capability commoncap [135147.522192] CPU:0 [135147.522197] EIP:0060:[]Not tainted VLI [135147.522204] EFLAGS: 00010246 (2.6.17-11-generic #2) [135147.522295] EIP is at aufs_umount_begin+0x67/0xb0 [aufs] [135147.522325] eax: ebx: ecx: edx: 0001 [135147.522358] esi: cea18e40 edi: ebp: cec0df50 esp: cec0df28 [135147.522388] ds: 007b es: 007b ss: 0068 [135147.522415] Process umount (pid: 22876, threadinfo=cec0c000 task=cd1c1ab0) [135147.522434] Stack: d0c07855 d0c07d64 00f0 d0c07500 cec0c000 c0188970 [135147.522504]cea18e40 cfa21e00 c32adb64 cea18e40 0286 cf9fb5f4 0001 [135147.522574]0001 b7d21000 cf4c1900 c015bce6 b7d22000 cf4c1904 [135147.522644] Call Trace: [135147.522686] sys_umount+0x110/0x2c0 do_munmap+0x186/0x1e0 [135147.522799] sys_oldumount+0x17/0x20 sysenter_past_esp+0x54/0x79 [135147.522895] Code: 64 7d c0 d0 c7 04 24 55 78 c0 d0 e8 a4 0c 53 ef e8 4f 2c 51 ef 89 f0 e8 f8 10 00 00 89 c3 e8 71 76 6e ef ba 01 00 ff ff 89 d8 90 <0f> c1 10 85 d2 0f 85 55 10 00 00 89 f0 e8 d7 10 00 00 f6 40 20 [135147.523158] EIP: [] aufs_umount_begin+0x67/0xb0 [aufs] SS:ESP 0068:cec0df28 [135147.523624] sb cea18e40, sbinfo [135147.597692] BUG: warning at /root/aufs.wcvs/aufs/fs/aufs/super.c:240/aufs_umount_begin() [135147.597790] aufs_umount_begin+0x51/0xb0 [aufs] sys_umount+0x110/0x2c0 [135147.598143] do_munmap+0x186/0x1e0 sys_oldumount+0x17/0x20 [135147.598432] sysenter_past_esp+0x54/0x79 [135147.600596] BUG: unable to handle kernel NULL pointer dereference at virtual address [135147.600707] printing eip: [135147.600759] d0bf21f7 [135147.600809] *pde = [135147.600862] Oops: 0002 [#2] [135147.600913] SMP [135147.601030] Modules linked in: aufs nfsd exportfs lockd sunrpc bridge unionfs xfs dm_mod md_mod fuse lp pcspkr i2c_piix4 psmouse i2c_core serio_raw parport_pc parport evdev shpchp pci_hotplug sg r1000 r8169 3c59x mii floppy intel_agp agpgart ext3 jbd uhci_hcd usbcore ide_generic sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcursor vesafb capability commoncap [135147.602922] CPU:0 [135147.602927] EIP:0060:[]Not tainted VLI [135147.602934] EFLAGS: 00010246 (2.6.17-11-generic #2) [135147.603115] EIP is at aufs_umount_begin+0x67/0xb0 [aufs] [135147.603174] eax: ebx: ecx: edx: 0001 [135147.603237] esi: cea18e40 edi: ebp: c3b4df50 esp: c3b4df28 [135147.603298] ds: 007b es: 007b ss: 0068 [135147.603358] Process umount (pid: 22893, threadinfo=c3b4c000 task=c9b54560) [135147.603408] Stack: d0c07855 d0c07d64 00f0 d0c07500 c3b4c000 c0188970 [135147.603786]cea18e40 cfa21e00 c32adb64 cea1
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> I've now stumbled upon more problems. I switched from 2.6.17-10-server >> to 2.6.17-11-generic, recompiled aufs-module with no issues, and mounted >> /storage. (I guess this problem existed before I switched kernel) >> I went into /storage/apt-proxy (which resides on /vault/sdd/apt-proxy), >> and did du -sh: (It still has the umount bug) >> >> [ 421.415679] aufs 20070226 >> [ 482.034285] [ cut here ] >> [ 482.034364] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/vdir.c:608! > > A few weeks ago another user reported the same thing which is an > internal checking in vdir.c. His cause seemed his mis-configuration > about CONFIG_AUFS_LHASH_PATCH. > I suggest you to check your configuration. I thought CONFIG_AUFS_LHASH_PATCH was for using NFS-filesystems as a branch of your aufs? I haven't touched AUFS_LHASH_PATCH, I just fetched source and did make -f local.mk. :-) > And please select your kernel and tell me where I can get its source > code. Currectly I guess something around umount was changed by ubuntu > people and it caused your umount problem. > I could not reproduce your problem on plain linux-2.6.17. Here are the files for the kernel: http://archive.ubuntu.com/ubuntu/pool/main/l/linux-source-2.6.17/linux-source-2.6.17_2.6.17.1.orig.tar.gz http://archive.ubuntu.com/ubuntu/pool/main/l/linux-source-2.6.17/linux-source-2.6.17_2.6.17.1-11.35.diff.gz Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> [ 170.818591] sb cc657dc0, sbinfo > > Is your kernel plain 2.6.10, or something modified? > > Junjiro Okajima I'm running 2.6.17-11-generic from ubuntu edgy, see below for apt-cache show linux-image-2.6.7-11-generic. I've now stumbled upon more problems. I switched from 2.6.17-10-server to 2.6.17-11-generic, recompiled aufs-module with no issues, and mounted /storage. (I guess this problem existed before I switched kernel) I went into /storage/apt-proxy (which resides on /vault/sdd/apt-proxy), and did du -sh: (It still has the umount bug) [ 421.415679] aufs 20070226 [ 482.034285] [ cut here ] [ 482.034364] kernel BUG at /root/aufs.wcvs/aufs/fs/aufs/vdir.c:608! [ 482.034423] invalid opcode: [#1] [ 482.034474] SMP [ 482.034590] Modules linked in: aufs nfsd exportfs lockd sunrpc bridge xfs dm_mod md_mod fuse lp sg parport_pc evdev psmouse serio_raw parport intel_agp agpgart floppy r1000 r8169 shpchp pci_hotplug 3c59x mii pcspkr i2c_piix4 i2c_core ext3 jbd ide_generic uhci_hcd usbcore sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcursor vesafb capability commoncap [ 482.036455] CPU:0 [ 482.036459] EIP:0060:[]Not tainted VLI [ 482.036466] EFLAGS: 00010286 (2.6.17-11-generic #2) [ 482.036689] EIP is at init_vdir+0x698/0x730 [aufs] [ 482.036748] eax: ebx: ecx: c37cbe80 edx: c982d6b4 [ 482.036813] esi: 0001 edi: c982d6d4 ebp: cf5e6800 esp: c27e3ef4 [ 482.036872] ds: 007b es: 007b ss: 0068 [ 482.036928] Process du (pid: 4902, threadinfo=c27e2000 task=c360da90) [ 482.036977] Stack: c64750a0 c2bc4ae0 d0af46f4 d0af4549 c982d6d4 0005 d0af52c0 [ 482.037350]d0afa60b c982d6b4 c360da90 c360da90 cf5e6800 c02da7b5 c982d6bc [ 482.037722]00100100 00200200 0002 c982d6b4 c982d6b4 c982d6d4 c2bc4ae0 [ 482.038099] Call Trace: [ 482.038200] .text.lock.dinfo+0x9e/0xda [aufs] di_read_unlock+0x49/0xc0 [aufs] [ 482.038548] reval_and_lock_finfo+0xa0/0x6fd [aufs] aufs_permission+0x17b/0x2c0 [aufs] [ 482.038896] rwsem_down_write_failed+0x165/0x180 aufs_readdir+0x8d/0x140 [aufs] [ 482.039244] filldir64+0x0/0xe0 filldir64+0x0/0xe0 [ 482.039531] vfs_readdir+0x94/0xb0 sys_getdents64+0x6f/0xc0 [ 482.039834] sysenter_past_esp+0x54/0x79 [ 482.040054] Code: d6 f9 ff ff 0f 0b 5e 02 08 09 b0 d0 e9 dd f9 ff ff 0f 0b 60 02 08 09 b0 d0 e9 f9 f9 ff ff 0f 0b 60 02 08 09 b0 d0 e9 05 fa ff ff <0f> 0b 60 02 08 09 b0 d0 e9 0c fa ff ff 0f 0b 61 02 08 09 b0 d0 [ 482.042343] EIP: [] init_vdir+0x698/0x730 [aufs] SS:ESP 0068:c27e3ef4 [ 482.042516] Package: linux-image-2.6.17-11-generic Priority: optional Section: base Installed-Size: 65972 Maintainer: Ben Collins <[EMAIL PROTECTED]> Architecture: i386 Source: linux-source-2.6.17 Version: 2.6.17.1-11.35 Provides: linux-image, linux-image-2.6, fuse-module, ndiswrapper-modules-1.22 Depends: initramfs-tools (>= 0.36ubuntu6), coreutils | fileutils (>= 4.0), module-init-tools (>= 3.2.1-0ubuntu1) Recommends: lilo (>= 19.1) | grub Suggests: fdutils, linux-doc-2.6.17 | linux-source-2.6.17 Conflicts: hotplug (<< 0.0.20040105-1) Filename: pool/main/l/linux-source-2.6.17/linux-image-2.6.17-11-generic_2.6.17.1-11.35_i386.deb Size: 22984370 MD5sum: f25214fa56fc95e102a378b4ecc0400f SHA1: 10e949f28d14450de688ef9a781b552291d4cbe3 SHA256: 98b735c88d6ed48eba4853c8a7e99d2cbecde8780d243d6145940e10cf16ae13 Description: Linux kernel image for version 2.6.17 on x86/x86_64 This package contains the Linux kernel image for version 2.6.17 on x86/x86_64. . Also includes the corresponding System.map file, the modules built by the packager, and scripts that try to ensure that the system is not left in an unbootable state after an update. . Supports Generic processors. . Geared toward desktop systems. . You likely do not want to install this package directly. Instead, install the linux-generic meta-package, which will ensure that upgrades work correctly, and that supporting packages are also installed. Bugs: mailto:[EMAIL PROTECTED] Origin: Ubuntu - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
[EMAIL PROTECTED] wrote: > "Jrgen_P._Tjern": >> Mounting with only xfs gives me the same error, /proc/mounts: >> /dev/sdd1 /vault/sdd xfs rw 0 0 >> none /storage aufs rw,xino=3D/vault/sdd/.aufs.xino,br:/vault/sdd=3Drw 0 0 >> >> When you say 'a while', how long do you mean? > > A typical use case of yours. > Or nothing, just mount and umount. > > >> Here're the oops: > > Will you apply this patch and test umounting? This yields: [ 170.818591] sb cc657dc0, sbinfo - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
Re: umount crash / BUG: unable to handle kernel NULL pointer
(Sorry about sending this directly to you, my thunderbird doesn't have a proper "reply to list"-feature) [EMAIL PROTECTED] wrote: > Currently, I have no idea what is wrong. > I could not reproduce your problem on my linux-2.6.17. > In order to solve this problem, will you try these things? > - update aufs, I have updated CVS a few minuts ago. > - mount aufs without lvm. plain ext3 or xfs. > - check your syslog, including mount-time. > - use aufs for a while and unmount it. Okay, updated to 20070226 now. Syslog says nothing interesting at mount-time: Feb 26 06:07:44 localhost kernel: [ 137.822506] aufs 20070226 Mounting with only xfs gives me the same error, /proc/mounts: /dev/sdd1 /vault/sdd xfs rw 0 0 none /storage aufs rw,xino=/vault/sdd/.aufs.xino,br:/vault/sdd=rw 0 0 When you say 'a while', how long do you mean? Here're the oops: [ 267.997133] BUG: unable to handle kernel NULL pointer dereference at virtual address [ 267.997258] printing eip: [ 267.997310] d0ae105e [ 267.997317] *pde = 089aa001 [ 267.997371] Oops: 0002 [#1] [ 267.997420] SMP [ 267.997535] Modules linked in: aufs nfsd exportfs lockd sunrpc bridge xfs dm_mod md_mod fuse lp pcspkr psmouse serio_raw i 2c_piix4 i2c_core floppy sg evdev shpchp pci_hotplug intel_agp agpgart parport_pc r1000 3c59x parport r8169 mii ext3 jbd ide_ generic uhci_hcd usbcore sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcurs or vesafb capability commoncap [ 267.999414] CPU:0 [ 267.999418] EIP:0060:[]Not tainted VLI [ 267.999425] EFLAGS: 00010286 (2.6.17-10-server #2) [ 267.999641] EIP is at aufs_umount_begin+0xe/0x50 [aufs] [ 267.999701] eax: ebx: ce296640 ecx: d0ae1050 edx: 0001 [ 267.999764] esi: c833e000 edi: ebp: c833ff50 esp: c833ff3c [ 267.999825] ds: 007b es: 007b ss: 0068 [ 267.999882] Process umount (pid: 4386, threadinfo=c833e000 task=c8d09a90) [ 267.32] Stack: c0187ea0 ce296640 cfd36e00 c844d86c ce296640 [ 268.000310]0286 c8994234 0001 0001 b7d06000 ceeefc80 c015b236 [ 268.000685] b7d07000 ceeefc84 c8994120 b7d07000 c899412c c833ff94 c833ff94 [ 268.001061] Call Trace: [ 268.001157] sys_umount+0x110/0x2a0 do_munmap+0x186/0x1e0 [ 268.001501] sys_oldumount+0x17/0x20 sysenter_past_esp+0x54/0x75 [ 268.001816] Code: 80 04 00 b8 ea ff ff ff 0f 45 d0 89 d0 c3 eb 0d 90 90 90 90 90 90 90 90 90 90 90 90 90 53 89 c3 e8 58 11 00 00 ba 01 00 ff ff 90 <0f> c1 10 85 d2 0f 85 bf 10 00 00 89 d8 e8 40 11 00 00 f6 40 20 [ 268.004100] EIP: [] aufs_umount_begin+0xe/0x50 [aufs] SS:ESP 0068:c833ff3c [ 268.004269] <1>BUG: unable to handle kernel NULL pointer dereference at virtual address [ 268.038235] printing eip: [ 268.038289] d0ae105e [ 268.038338] *pde = 08490001 [ 268.038391] Oops: 0002 [#2] [ 268.038441] SMP [ 268.038556] Modules linked in: aufs nfsd exportfs lockd sunrpc bridge xfs dm_mod md_mod fuse lp pcspkr psmouse serio_raw i 2c_piix4 i2c_core floppy sg evdev shpchp pci_hotplug intel_agp agpgart parport_pc r1000 3c59x parport r8169 mii ext3 jbd ide_ generic uhci_hcd usbcore sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcurs or vesafb capability commoncap [ 268.040435] CPU:0 [ 268.040439] EIP:0060:[]Not tainted VLI [ 268.040445] EFLAGS: 00010286 (2.6.17-10-server #2) [ 268.040660] EIP is at aufs_umount_begin+0xe/0x50 [aufs] [ 268.040722] eax: ebx: ce296640 ecx: d0ae1050 edx: 0001 [ 268.040784] esi: c852c000 edi: ebp: c852df50 esp: c852df3c [ 268.040844] ds: 007b es: 007b ss: 0068 [ 268.040901] Process umount (pid: 4391, threadinfo=c852c000 task=cfe3fa90) [ 268.040950] Stack: c0187ea0 ce296640 cfd36e00 c844d86c ce296640 [ 268.041328]0286 c8994234 0001 0001 b7cb2000 ceeefac0 c015b236 [ 268.041703] b7cb3000 ceeefac4 c89940c4 b7cb3000 c89940d0 c852df94 c852df94 [ 268.042080] Call Trace: [ 268.042178] sys_umount+0x110/0x2a0 do_munmap+0x186/0x1e0 [ 268.042524] sys_oldumount+0x17/0x20 sysenter_past_esp+0x54/0x75 [ 268.042843] Code: 80 04 00 b8 ea ff ff ff 0f 45 d0 89 d0 c3 eb 0d 90 90 90 90 90 90 90 90 90 90 90 90 90 53 89 c3 e8 58 11 00 00 ba 01 00 ff ff 90 <0f> c1 10 85 d2 0f 85 bf 10 00 00 89 d8 e8 40 11 00 00 f6 40 20 [ 268.045117] EIP: [] aufs_umount_begin+0xe/0x50 [aufs] SS:ESP 0068:c852df3c [ 268.045286] Kindest regards, Jørgen. - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
NFS-exporting my AUFS.
Hi!
I was up until recently using a pure LVM for my storage, and today
switched it to a LVM + 3x loose partitions as a AUFS - and now my
previously fine NFS-setup stopped working.
I read that using NFS-mounts as a branch of the AUFS could be a problem,
but nothing about exporting an AUFS using NFS.
The setup is like this:
* datamania, storage server with AUFS and the NFS export:
[EMAIL PROTECTED]:~$ ifconfig|grep 'inet addr'|head -1
inet addr:192.168.0.3 Bcast:192.168.0.255 Mask:255.255.255.0
[EMAIL PROTECTED]:~$ grep /storage /proc/mounts
none /storage aufs
rw,xino=/vault/sdd/.aufs.xino,br:/vault/sdd=rw:/vault/sda=rw:/vault/hdc=rw:/vault/lvm=rw
0 0
[EMAIL PROTECTED]:~$ grep ^/storage/music /var/lib/nfs/etab
/storage/music
gullars(rw,async,wdelay,hide,nocrossmnt,secure,root_squash,no_all_squash,subtree_check,secure_locks,acl,mapping=identity,anonuid=65534,anongid=65534)
[EMAIL PROTECTED]:~$ getent hosts gullars
192.168.0.2 gullars
[EMAIL PROTECTED]:~$ ls -ld /storage/{,music}
drwxrwxrwx 31 daxxar storage 104 2007-02-26 00:03 /storage/
drwxrwxrwx 440 daxxar storage 6 2007-02-25 20:26 /storage/music
* gullars, the server trying to mount the nfs:
[EMAIL PROTECTED]:~$ ifconfig|grep 'inet addr'|head -1
inet addr:192.168.0.2 Bcast:192.168.0.255 Mask:255.255.255.0
[EMAIL PROTECTED]:~$ getent hosts datamania
192.168.0.3 datamania
[EMAIL PROTECTED]:~$ grep ^datamania /etc/fstab
datamania:/storage/music /mnt/music nfs
auto,soft,intr,rsize=8192,wsize=8192,async,nfsvers=3,tcp 0 2
* now, trying to mount:
[EMAIL PROTECTED]:~$ sudo mount /mnt/music
mount: datamania:/storage/music failed, reason given by server:
Permission denied
* checking logs:
[EMAIL PROTECTED]:~$ tail -n 1 /var/log/daemon.log
Feb 26 04:39:23 localhost mountd[5094]: authenticated mount request from
gullars:926 for /storage/music (/storage/music)
No errors, no nothing. Anyone know what could cause this? I could be
wrong in assuming this is something related to aufs, but it worked fine
up until I switched. :-)
Thanks in advance!
Kindest regards, Jørgen P. Tjernø.
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
umount crash / BUG: unable to handle kernel NULL pointer
Basically I have three loose disks and one lvm combined to form a aufs-filesystem. I've then mounted this into /storage. When I try to unmount the main aufs mount, I get a kernel oops and no change in the mtab. (nor on the disk) I've also had a problem with a directory suddenly appearing as neither a directory nor a file. (ls -l shows the directory as a file (no 'd' flag), rm on the directory says 'rm: cannot remove `dir': Is a directory' - while rmdir on it says 'rmdir: dir: Not a directory') The latter is not reproducible (I started using aufs today), but the former is reproducible. Information: [EMAIL PROTECTED]:~$ cat /proc/mounts rootfs / rootfs rw 0 0 none /sys sysfs rw,nosuid,nodev,noexec 0 0 none /proc proc rw,nosuid,nodev,noexec 0 0 udev /dev tmpfs rw 0 0 /dev/disk/by-uuid/5c6fcc95-c3c3-4526-a868-33cb76d79471 / ext3 rw,data=ordered 0 0 /dev/disk/by-uuid/5c6fcc95-c3c3-4526-a868-33cb76d79471 /dev/.static/dev ext3 rw,data=ordered 0 0 tmpfs /var/run tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /var/lock tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /dev/shm tmpfs rw 0 0 devpts /dev/pts devpts rw 0 0 usbfs /dev/bus/usb/.usbfs usbfs rw 0 0 udev /proc/bus/usb tmpfs rw 0 0 usbfs /proc/bus/usb/.usbfs usbfs rw 0 0 tmpfs /var/run tmpfs rw,nosuid,nodev,noexec 0 0 tmpfs /var/lock tmpfs rw,nosuid,nodev,noexec 0 0 /dev/mapper/main-storage /vault/lvm ext3 rw,data=ordered 0 0 /dev/disk/by-uuid/d61a06e4-2830-44e6-95ba-5e54e64a9736 /vault/hdc ext3 rw,data=ordered 0 0 /dev/sdd1 /vault/sdd xfs rw 0 0 /dev/sda1 /vault/sda xfs rw 0 0 none /storage aufs rw,xino=/vault/sdd/.aufs.xino,br:/vault/sdd=rw:/vault/sda=rw:/vault/hdc=rw:/vault/lvm=rw 0 0 [EMAIL PROTECTED]:~$ sudo umount /storage Segmentation fault [EMAIL PROTECTED]:~$ grep AUFS_VERSION /root/aufs.wcvs/aufs/include/linux/aufs_type.h #define AUFS_VERSION"20070219" [EMAIL PROTECTED]:~$ uname -a Linux datamania 2.6.17-10-server #2 SMP Tue Dec 5 22:29:32 UTC 2006 i686 GNU/Linux [EMAIL PROTECTED]:~$ dmesg|tail -n 24 [ 78.453998] br0: port 1(eth0) entering forwarding state [ 209.826244] BUG: unable to handle kernel NULL pointer dereference at virtual address [ 209.826369] printing eip: [ 209.826421] d0b8a06e [ 209.826428] *pde = 08792001 [ 209.826484] Oops: 0002 [#1] [ 209.826535] SMP [ 209.826650] Modules linked in: bridge aufs xfs dm_mod md_mod lp sg psmouse parport_pc serio_raw evdev i2c_piix4 parport floppy i2c_core r1000 r8169 shpchp pci_hotplug intel_agp agpgart 3c59x mii pcspkr ext3 jbd uhci_hcd usbcore ide_generic sd_mod sata_sil libata scsi_mod ide_disk piix generic processor fbcon tileblit font bitblit softcursor vesafb capability commoncap [ 209.828348] CPU:0 [ 209.828353] EIP:0060:[]Not tainted VLI [ 209.828360] EFLAGS: 00010286 (2.6.17-10-server #2) [ 209.828569] EIP is at aufs_umount_begin+0xe/0x50 [aufs] [ 209.828630] eax: ebx: ceb3ce40 ecx: d0b8a060 edx: 0001 [ 209.828693] esi: c87a6000 edi: ebp: c87a7f50 esp: c87a7f3c [ 209.828752] ds: 007b es: 007b ss: 0068 [ 209.828810] Process umount (pid: 4296, threadinfo=c87a6000 task=cca4ba90) [ 209.828859] Stack: c0187ea0 ceb3ce40 cf6bd600 c1326df8 ceb3ce40 [ 209.829237]0286 c86011d8 0001 0001 b7d41000 cfaf1580 c015b236 [ 209.829617] b7d42000 cfaf1584 c86013a4 b7d42000 c86013b0 c87a7f94 c87a7f94 [ 209.829995] Call Trace: [ 209.830092] sys_umount+0x110/0x2a0 do_munmap+0x186/0x1e0 [ 209.830429] sys_oldumount+0x17/0x20 sysenter_past_esp+0x54/0x75 [ 209.830738] Code: 90 90 90 90 90 90 90 90 90 90 90 90 90 b9 30 f4 b9 d0 e9 56 11 60 ef 8d b6 00 00 00 00 53 89 c3 e8 d8 10 00 00 ba 01 00 ff ff 90 <0f> c1 10 85 d2 0f 85 50 10 00 00 89 d8 e8 c0 10 00 00 f6 40 20 [ 209.833015] EIP: [] aufs_umount_begin+0xe/0x50 [aufs] SS:ESP 0068:c87a7f3c - Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
