Re: reiserfs bug?

[Frederic Weisbecker <fweisbec@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 2771 bytes --]

On Sat, Aug 14, 2010 at 01:34:07PM +0100, Andrew Benton wrote:
> Hello World!
> I've just recompiled my kernel (current Linus tree), rebooted into it, 
> everything seems to work Ok. I wanted to change 1 option so I 
> recompiled again. Right at the end of the build Xorg server died and on 
> the screen was printed:
> 
> ------------[ cut here ]------------
> kernel BUG at /home/andy/save/src/linux-2.6/fs/inode.c:298!
> invalid opcode: 0000 [#1] SMP 
> last sysfs file: /sys/devices/virtual/vc/vcsa2/uevent
> CPU 0 
> Pid: 17243, comm: mv Not tainted 2.6.35+ #1 MS-7521/MS-7521
> RIP: 0010:[<ffffffff810b325b>]  [<ffffffff810b325b>] end_writeback+0x5b/0x70
> RSP: 0018:ffff880114d7dcf8  EFLAGS: 00010202
> RAX: 0000000000000060 RBX: ffff88013fab00f8 RCX: 0000000000000000
> RDX: 0000000000000000 RSI: 00000000ffffffff RDI: ffff88013fab00f8
> RBP: 00000000ffffffff R08: 0000000000000f36 R09: 0000000000000f36
> R10: 0000000000000bac R11: 0000000000000000 R12: ffff88013f9ac540
> R13: 0000000000000000 R14: ffff88010fd31500 R15: ffff88010fd31440
> FS:  00007f306e5be700(0000) GS:ffff880001a00000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> CR2: 00007f306debb080 CR3: 000000013d730000 CR4: 00000000000406f0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> Process mv (pid: 17243, threadinfo ffff880114d7c000, task ffff88013cf73430)
> Stack:
>  ffff88013fab00f8 ffffffff810f57a7 ffff88013e608400 0000000000000000
> <0> 0000000000000000 0000000000000000 0000000000000000 0000000000000000
> <0> 0000000000000000 0000000000000000 ffff88010fd31500 ffff88013fab00f8
> Call Trace:
>  [<ffffffff810f57a7>] ? reiserfs_evict_inode+0x47/0x130
>  [<ffffffff810b328a>] ? evict+0x1a/0xa0
>  [<ffffffff810b3f6c>] ? iput+0x19c/0x280
>  [<ffffffff810afdb7>] ? d_kill+0x47/0x80
>  [<ffffffff810b0dbf>] ? dput+0x9f/0x180
>  [<ffffffff810abb1b>] ? sys_renameat+0x21b/0x270
>  [<ffffffff81088449>] ? handle_mm_fault+0x1b9/0x940
>  [<ffffffff8102118f>] ? do_page_fault+0x1cf/0x320
>  [<ffffffff810021ab>] ? system_call_fastpath+0x16/0x1b
> Code: 1d 48 8d bf d0 01 00 00 b9 02 00 00 00 48 c7 c2 70 2b 0b 81 be 07 00 00 00 e8 42 b0 49 00 48 c7 83 d0 01 00 00 60 00 00 00 5b c3 <0f> 0b 0f 0b 0f 0b 0f 0b 66 66 66 66 2e 0f 1f 84 00 00 00 00 00 
> RIP  [<ffffffff810b325b>] end_writeback+0x5b/0x70
>  RSP <ffff880114d7dcf8>
> ---[ end trace fdd5f8a966d540a4 ]---
> 
> The Opps mentions reiserfs. The only reiserfs partition I have is /boot 
> (grub won't boot from btrfs) and the script I use to recompile the 
> kernel copies it into /boot when the new kernel is ready.
> 
> Andy


Oh, looks like a return has been forgotten.

Does that patch fix your issue?

Thanks.

[-- Attachment #2: reiserfs.diff --]
[-- Type: text/x-diff, Size: 346 bytes --]

diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index ae35413..5b18f93 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -84,6 +84,8 @@ void reiserfs_evict_inode(struct inode *inode)
 	inode->i_blocks = 0;
 	reiserfs_write_unlock_once(inode->i_sb, depth);
 
+	return;
+
 no_delete:
 	end_writeback(inode);
 	dquot_drop(inode);