[alsa-devel] KMSAN: uninit-value in linear_transfer (2)
Alexander Potapenko
glider at google.com
Fri Nov 9 10:31:03 CET 2018
On Thu, Nov 8, 2018 at 9:38 PM syzbot
<syzbot+1cb36954e127c98dd037 at syzkaller.appspotmail.com> wrote:
>
> Hello,
>
> syzbot found the following crash on:
>
> HEAD commit: 7438a3b20295 kmsan: print user address when reporting info..
> git tree: https://github.com/google/kmsan.git/master
> console output: https://syzkaller.appspot.com/x/log.txt?x=15b42133400000
> kernel config: https://syzkaller.appspot.com/x/.config?x=8df5fc509a1b351b
> dashboard link: https://syzkaller.appspot.com/bug?extid=1cb36954e127c98dd037
> compiler: clang version 8.0.0 (trunk 343298)
> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=12be9825400000
> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=13e2c425400000
>
> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> Reported-by: syzbot+1cb36954e127c98dd037 at syzkaller.appspotmail.com
>
> ==================================================================
> BUG: KMSAN: uninit-value in __arch_swab32
> arch/x86/include/uapi/asm/swab.h:10 [inline]
> BUG: KMSAN: uninit-value in __fswab32 include/uapi/linux/swab.h:59 [inline]
> BUG: KMSAN: uninit-value in do_convert sound/core/oss/linear.c:50 [inline]
> BUG: KMSAN: uninit-value in convert sound/core/oss/linear.c:81 [inline]
> BUG: KMSAN: uninit-value in linear_transfer+0x92d/0xca0
> sound/core/oss/linear.c:110
> CPU: 1 PID: 6835 Comm: syz-executor866 Not tainted 4.19.0+ #78
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
> Google 01/01/2011
> Call Trace:
> __dump_stack lib/dump_stack.c:77 [inline]
> dump_stack+0x32d/0x480 lib/dump_stack.c:113
> kmsan_report+0x19f/0x300 mm/kmsan/kmsan.c:911
> __msan_warning+0x76/0xd0 mm/kmsan/kmsan_instr.c:415
> __arch_swab32 arch/x86/include/uapi/asm/swab.h:10 [inline]
> __fswab32 include/uapi/linux/swab.h:59 [inline]
> do_convert sound/core/oss/linear.c:50 [inline]
> convert sound/core/oss/linear.c:81 [inline]
> linear_transfer+0x92d/0xca0 sound/core/oss/linear.c:110
> snd_pcm_plug_read_transfer+0x3bf/0x590 sound/core/oss/pcm_plugin.c:659
> snd_pcm_oss_read2 sound/core/oss/pcm_oss.c:1480 [inline]
> snd_pcm_oss_read1 sound/core/oss/pcm_oss.c:1518 [inline]
> snd_pcm_oss_read+0xe6d/0x1cb0 sound/core/oss/pcm_oss.c:2758
> __vfs_read+0x1e2/0xb10 fs/read_write.c:416
> vfs_read+0x380/0x6b0 fs/read_write.c:452
> ksys_read fs/read_write.c:578 [inline]
> __do_sys_read fs/read_write.c:588 [inline]
> __se_sys_read+0x17a/0x370 fs/read_write.c:586
> __x64_sys_read+0x4a/0x70 fs/read_write.c:586
> do_syscall_64+0xcf/0x110 arch/x86/entry/common.c:291
> entry_SYSCALL_64_after_hwframe+0x63/0xe7
> RIP: 0033:0x446379
> Code: e8 0c e8 ff ff 48 83 c4 18 c3 0f 1f 80 00 00 00 00 48 89 f8 48 89 f7
> 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff
> ff 0f 83 7b 08 fc ff c3 66 2e 0f 1f 84 00 00 00 00
> RSP: 002b:00007fcd5b97cda8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
> RAX: ffffffffffffffda RBX: 00000000006dac38 RCX: 0000000000446379
> RDX: 0000000000000184 RSI: 0000000020002180 RDI: 0000000000000003
> RBP: 00000000006dac30 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000246 R12: 00000000006dac3c
> R13: 7073642f7665642f R14: 00800000c0045002 R15: 0000000000000001
>
> Uninit was stored to memory at:
> kmsan_save_stack_with_flags mm/kmsan/kmsan.c:252 [inline]
> kmsan_save_stack mm/kmsan/kmsan.c:267 [inline]
> kmsan_internal_chain_origin+0x136/0x240 mm/kmsan/kmsan.c:569
> kmsan_memcpy_origins+0x13d/0x1b0 mm/kmsan/kmsan.c:393
> __msan_memcpy+0x6f/0x80 mm/kmsan/kmsan_instr.c:242
> do_convert sound/core/oss/linear.c:48 [inline]
> convert sound/core/oss/linear.c:81 [inline]
> linear_transfer+0x74c/0xca0 sound/core/oss/linear.c:110
> snd_pcm_plug_read_transfer+0x3bf/0x590 sound/core/oss/pcm_plugin.c:659
> snd_pcm_oss_read2 sound/core/oss/pcm_oss.c:1480 [inline]
> snd_pcm_oss_read1 sound/core/oss/pcm_oss.c:1518 [inline]
> snd_pcm_oss_read+0xe6d/0x1cb0 sound/core/oss/pcm_oss.c:2758
> __vfs_read+0x1e2/0xb10 fs/read_write.c:416
> vfs_read+0x380/0x6b0 fs/read_write.c:452
> ksys_read fs/read_write.c:578 [inline]
> __do_sys_read fs/read_write.c:588 [inline]
> __se_sys_read+0x17a/0x370 fs/read_write.c:586
> __x64_sys_read+0x4a/0x70 fs/read_write.c:586
> do_syscall_64+0xcf/0x110 arch/x86/entry/common.c:291
> entry_SYSCALL_64_after_hwframe+0x63/0xe7
>
> Uninit was created at:
> kmsan_save_stack_with_flags mm/kmsan/kmsan.c:252 [inline]
> kmsan_internal_alloc_meta_for_pages+0x155/0x740 mm/kmsan/kmsan.c:689
> kmsan_alloc_page+0x77/0xe0 mm/kmsan/kmsan_hooks.c:320
> __alloc_pages_nodemask+0x12cc/0x6640 mm/page_alloc.c:4416
> alloc_pages_current+0x584/0x7e0 mm/mempolicy.c:2093
> alloc_pages include/linux/gfp.h:511 [inline]
> __vmalloc_area_node mm/vmalloc.c:1689 [inline]
> __vmalloc_node_range+0x879/0x12a0 mm/vmalloc.c:1752
> __vmalloc_node mm/vmalloc.c:1797 [inline]
> __vmalloc_node_flags mm/vmalloc.c:1811 [inline]
> vmalloc+0xd8/0xf0 mm/vmalloc.c:1833
> snd_pcm_plugin_alloc+0x255/0xc80 sound/core/oss/pcm_plugin.c:71
> snd_pcm_plug_alloc+0x281/0x600 sound/core/oss/pcm_plugin.c:137
> snd_pcm_oss_change_params_locked+0x5e40/0x6e30
> sound/core/oss/pcm_oss.c:1039
> snd_pcm_oss_change_params sound/core/oss/pcm_oss.c:1107 [inline]
> snd_pcm_oss_get_active_substream+0x4f7/0x5a0 sound/core/oss/pcm_oss.c:1124
> snd_pcm_oss_get_rate sound/core/oss/pcm_oss.c:1774 [inline]
> snd_pcm_oss_set_rate sound/core/oss/pcm_oss.c:1766 [inline]
> snd_pcm_oss_ioctl+0x4adb/0x8860 sound/core/oss/pcm_oss.c:2614
> do_vfs_ioctl+0xf77/0x2d30 fs/ioctl.c:46
> ksys_ioctl fs/ioctl.c:702 [inline]
> __do_sys_ioctl fs/ioctl.c:709 [inline]
> __se_sys_ioctl+0x1da/0x270 fs/ioctl.c:707
> __x64_sys_ioctl+0x4a/0x70 fs/ioctl.c:707
> do_syscall_64+0xcf/0x110 arch/x86/entry/common.c:291
> entry_SYSCALL_64_after_hwframe+0x63/0xe7
> ==================================================================
I am actually unsure whether this is a true positive or not.
snd_pcm_plugin_alloc() allocates plugin->buf using vmalloc(size) (see
https://elixir.bootlin.com/linux/v4.20-rc1/source/sound/core/oss/pcm_plugin.c#L70)
KMSAN doesn't know where this buffer is initialized (any help finding
this out is welcome!), so we mark |size| bytes starting from
plugin->buf as initialized:
https://github.com/google/kmsan/blob/master/sound/core/oss/pcm_plugin.c#L78
But when |size| is not page-aligned vmalloc() still allocates whole
page, and in the report above linear_transfer() appears to run past
|size| bytes and read the tail bytes.
KASAN doesn't detect this bug, because it doesn't handle vmalloc'ed
(and these bytes are addressable anyway).
The reproducer that triggers this bug is attached.
>
> ---
> This bug is generated by a bot. It may contain errors.
> See https://goo.gl/tpsmEJ for more information about syzbot.
> syzbot engineers can be reached at syzkaller at googlegroups.com.
>
> syzbot will keep track of this bug report. See:
> https://goo.gl/tpsmEJ#bug-status-tracking for how to communicate with
> syzbot.
> syzbot can test patches for this bug, for details see:
> https://goo.gl/tpsmEJ#testing-patches
>
> --
> You received this message because you are subscribed to the Google Groups "syzkaller-bugs" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to syzkaller-bugs+unsubscribe at googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/syzkaller-bugs/00000000000021d9a9057a2d364e%40google.com.
> For more options, visit https://groups.google.com/d/optout.
--
Alexander Potapenko
Software Engineer
Google Germany GmbH
Erika-Mann-Straße, 33
80636 München
Geschäftsführer: Paul Manicle, Halimah DeLaine Prado
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dev_dsp.c
Type: text/x-csrc
Size: 671 bytes
Desc: not available
URL: <http://mailman.alsa-project.org/pipermail/alsa-devel/attachments/20181109/2dc56901/attachment-0001.bin>
More information about the Alsa-devel
mailing list