linux-kernel - Re: ftrace 'failed to modify' bug when loading reiserfs.ko

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 6 Sep 2013 09:32:52 -0400
From:	Steven Rostedt <rostedt@...dmis.org>
To:	Dave Jones <davej@...hat.com>
Cc:	Linux Kernel <linux-kernel@...r.kernel.org>
Subject: Re: ftrace 'failed to modify' bug when loading reiserfs.ko

On Thu, 5 Sep 2013 23:58:27 -0400
Dave Jones <davej@...hat.com> wrote:

> On Thu, Sep 05, 2013 at 09:51:54PM -0400, Steven Rostedt wrote:
>  > On Thu, 5 Sep 2013 21:48:59 -0400
>  > Dave Jones <davej@...hat.com> wrote:
>  > 
>  > > On Thu, Sep 05, 2013 at 09:44:55PM -0400, Steven Rostedt wrote:
>  > >  > On Thu, 5 Sep 2013 21:34:55 -0400
>  > >  > Dave Jones <davej@...hat.com> wrote:
>  > >  > 
>  > >  > > On Thu, Sep 05, 2013 at 09:28:34PM -0400, Steven Rostedt wrote:
>  > 
>  > >  > Did you change a config option, or update your gcc?
>  > > 
>  > > Yeah, changed CONFIG_DEBUG_KOBJECT, which rebuilt the world.
>  > 
>  > Still doesn't explain why it gave you that splat there.
>  > 
>  > Do you still have that binary module, and can you show me what's at
>  > reiserfs_init_bitmap_cache+0x0 with objdump?
> 
> I didn't, but it turns out I can recreate this. A little convoluted but..
> 
> disable DEBUG_KOBJECT_RELEASE
> build, install and boot into kernel
> 
> enable DEBUG_KOBJECT_RELEASE
> build kernel
> install -> boom

Did it report failing on the same function as before?

> 
> 
> 00000000000028b0 <reiserfs_init_bitmap_cache>:
> 
>         return bh;
> }
> 
> int reiserfs_init_bitmap_cache(struct super_block *sb)
> {
>     28b0:       e8 00 00 00 00          callq  28b5 <reiserfs_init_bitmap_cache+0x5>

That looks to be the call to fentry, but without being resolved. What
we saw in the previous report wasn't 0xe8 but 0x14, and it was
unresolved after loading!

I'm surprised that the module built with a different
DEBUG_KOBJECT_RELEASE config would load. Does it not affect module
versions?

-- Steve

>     28b5:       55                      push   %rbp
> 
> /* Don't trust REISERFS_SB(sb)->s_bmap_nr, it's a u16
>  * which overflows on large file systems. */
> static inline __u32 reiserfs_bmap_count(struct super_block *sb)
> {
>         return (SB_BLOCK_COUNT(sb) - 1) / (sb->s_blocksize * 8) + 1;
>     28b6:       31 d2                   xor    %edx,%edx
>     28b8:       48 89 e5                mov    %rsp,%rbp
>     28bb:       41 54                   push   %r12
>     28bd:       53                      push   %rbx
>     28be:       48 89 fb                mov    %rdi,%rbx
>     28c1:       48 8b 87 50 07 00 00    mov    0x750(%rdi),%rax
>     28c8:       48 8b 77 18             mov    0x18(%rdi),%rsi
>     28cc:       48 8b 40 08             mov    0x8(%rax),%rax
>     28d0:       48 8d 0c f5 00 00 00    lea    0x0(,%rsi,8),%rcx
>     28d7:       00 
>     28d8:       8b 00                   mov    (%rax),%eax
>     28da:       83 e8 01                sub    $0x1,%eax
>     28dd:       48 f7 f1                div    %rcx

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