Re: [Nbd] NBD, DISCARD/TRIM
- To: Paul Clements <paul.clements@...856...>
- Cc: nbd-general@lists.sourceforge.net
- Subject: Re: [Nbd] NBD, DISCARD/TRIM
- From: folkert <folkert@...421...>
- Date: Tue, 19 Mar 2013 19:06:42 +0100
- Message-id: <20130319180641.GD4508@...855...>
- In-reply-to: <20130319160149.GB4508@...855...>
- References: <20130319145642.GA4508@...855...> <CAECXXi5Co05GJrHe-8gskN-=ZBwHW-jchUb9CNwLWkOH67v=EQ@...18...> <20130319160149.GB4508@...855...>
> > > Am I right that discard/trim is in kernel 3.9? Or is it already in 3.8?
> > Patch was merged into Linus's tree on 5Oct12, so it should have made 3.7.
>
> There are problems with it, it seems.
>
> nbd-client occasionally crashes.
> Last time it did so, it did that with:
>
> root@...1280...:~# nbd-client -n 172.16.86.1 12341 /dev/nbd0
> Negotiation: ..size = 10240MB
> bs=1024, sz=10737418240 bytes
> nbd,2854: Kernel call returned: 22Closing: que, sock, done
This is 100% reproducable with:
- nbd-client & server version 1:3.2-2 (debian package)
- server linux kernel: 2.6.38-2-amd64
- client linux kernel: 3.8-trunk-amd64
This is with the original(!) stock(!) nbd client/server.
dmesg output:
[ 4881.378784] block nbd0: shutting down socket
[ 4881.379605] end_request: I/O error, dev nbd0, sector 271352
[ 4881.379827] block nbd0: queue cleared
[ 6762.039173] block nbd0: Attempted send on closed socket
[ 6762.039419] end_request: I/O error, dev nbd0, sector 73992
[ 6762.040301] ------------[ cut here ]------------
[ 6762.040428] kernel BUG at /build/buildd-linux_3.8.2-1~experimental.1-amd64-9iFbgL/linux-3.8.2/fs/buffer.c:2948!
[ 6762.040704] invalid opcode: 0000 [#1] SMP
[ 6762.040825] Modules linked in: ext4 crc16 jbd2 nfsd auth_rpcgss nfs_acl nfs lockd dns_resolver fscache sunrpc vsock(O) ext2 loop coretemp parport_pc parport snd_pcm snd_page_alloc snd_timer snd soundcore vmw_balloon microcode pcspkr pd
[ 6762.042305] CPU 1
[ 6762.042377] Pid: 2868, comm: bash Tainted: G O 3.8-trunk-amd64 #1 Debian 3.8.2-1~experimental.1 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform
[ 6762.042797] RIP: 0010:[<ffffffff81130558>] [<ffffffff81130558>] submit_bh+0x1f/0x1af
[ 6762.043823] RSP: 0018:ffff88001bdd9a08 EFLAGS: 00010246
[ 6762.043983] RAX: 0000000000000000 RBX: ffff88001ab3c130 RCX: 0000000000000000
[ 6762.044223] RDX: 0000000000000000 RSI: ffff88001ab3c130 RDI: ffff88001ab3c130
[ 6762.044459] RBP: 0000000000000411 R08: ffffffff8168e0a0 R09: 0000000000000000
[ 6762.044696] R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000411
[ 6762.044932] R13: 0000000000000001 R14: 00000000000ffff4 R15: 0000000000461439
[ 6762.045203] FS: 00007faf824a9700(0000) GS:ffff88001d820000(0000) knlGS:0000000000000000
[ 6762.045508] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 6762.045705] CR2: 0000000000466460 CR3: 000000001ed65000 CR4: 00000000000007e0
[ 6762.045984] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 6762.046275] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 6762.046515] Process bash (pid: 2868, threadinfo ffff88001bdd8000, task ffff88001ca34140)
[ 6762.046816] Stack:
[ 6762.046912] ffff88001ab3c130 0000000000000411 ffff88001fc43400 ffffffff8113084f
[ 6762.047232] ffff88001fbbe770 ffffffff811d892f 0000000000000000 ffff88001efa4800
[ 6762.047534] ffff88001ab3c130 ffffffffa039a1e0 0000000000000000 ffff880007578800
[ 6762.047831] Call Trace:
[ 6762.047944] [<ffffffff8113084f>] ? __sync_dirty_buffer+0x51/0x83
[ 6762.048194] [<ffffffff811d892f>] ? __percpu_counter_sum+0x40/0x55
[ 6762.048460] [<ffffffffa039a1e0>] ? ext4_commit_super+0x191/0x1dc [ext4]
[ 6762.048702] [<ffffffffa039b0a5>] ? ext4_error_inode+0x4f/0x113 [ext4]
[ 6762.048932] [<ffffffffa0387e10>] ? ext4_find_entry+0x2d5/0x3c3 [ext4]
[ 6762.049198] [<ffffffff810fd239>] ? kmem_cache_alloc+0x8e/0xfd
[ 6762.049415] [<ffffffffa0387f34>] ? ext4_lookup+0x36/0x10b [ext4]
[ 6762.049637] [<ffffffff811126d5>] ? lookup_real+0x24/0x3f
[ 6762.049827] [<ffffffff81112b1c>] ? __lookup_hash+0x2a/0x31
[ 6762.050022] [<ffffffff8111371d>] ? walk_component+0x72/0x1aa
[ 6762.050221] [<ffffffff811144d1>] ? path_lookupat+0x8b/0x2b1
[ 6762.050418] [<ffffffff81114718>] ? filename_lookup+0x21/0x79
[ 6762.050617] [<ffffffff81113b46>] ? getname_flags.part.23+0xb8/0x10a
[ 6762.050833] [<ffffffff811166ca>] ? user_path_at_empty+0x6f/0xa8
[ 6762.051067] [<ffffffff810e2610>] ? handle_mm_fault+0x1eb/0x239
[ 6762.051321] [<ffffffff81384dde>] ? __do_page_fault+0x2d7/0x375
[ 6762.051528] [<ffffffff8110eb76>] ? vfs_fstatat+0x3e/0x92
[ 6762.051718] [<ffffffff8110ee52>] ? sys_newstat+0x12/0x2d
[ 6762.051915] [<ffffffff81382218>] ? page_fault+0x28/0x30
[ 6762.052104] [<ffffffff81386fe9>] ? system_call_fastpath+0x16/0x1b
[ 6762.052312] Code: ff 41 59 89 d8 5b 5d 41 5c 41 5d c3 41 54 55 53 48 8b 06 48 89 f3 a8 04 75 02 0f 0b 41 89 fc 48 89 f7 e8 9d e7 ff ff 85 c0 75 02 <0f> 0b 48 83 7b 38 00 75 02 0f 0b 48 8b 03 f6 c4 02 74 02 0f 0b
[ 6762.053176] RIP [<ffffffff81130558>] submit_bh+0x1f/0x1af
[ 6762.053371] RSP <ffff88001bdd9a08>
[ 6762.054343] ---[ end trace 9932bdfb7f9817a6 ]---
And the nbd-client output was again this '22' error which means EINVAL
which is a bit strange.
Folkert van Heusden
--
Multitail est un outil permettant la visualisation de fichiers de
journalisation et/ou le suivi de l'exécution de commandes. Filtrage,
mise en couleur de mot-clé, fusions, visualisation de différences
(diff-view), etc. http://www.vanheusden.com/multitail/
----------------------------------------------------------------------
Phone: +31-6-41278122, PGP-key: 1F28D8AE, www.vanheusden.com
Reply to: