general protection fault in ucma_connect

[syzbot]

Hello,

syzbot hit the following crash on upstream commit
e2c15aff5f353ba80bd3bb49840837f65fa5cc43 (Thu Mar 15 18:07:35 2018 +0000)
Merge tag 'sound-4.16-rc6' of
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound

So far this crash happened 2 times on upstream.
C reproducer is attached.
syzkaller reproducer is attached.
Raw console output is attached.
compiler: gcc (GCC) 7.1.1 20170620
.config is attached.

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+36712f...@syzkaller.appspotmail.com
It will help syzbot understand when the bug is fixed. See footer for
details.
If you forward the report, please keep this part and the footer.

kasan: CONFIG_KASAN_INLINE enabled
kasan: GPF could be caused by NULL-ptr deref or user memory access
general protection fault: 0000 [#1] SMP KASAN
Dumping ftrace buffer:
(ftrace buffer empty)
Modules linked in:
CPU: 0 PID: 4215 Comm: syzkaller255546 Not tainted 4.16.0-rc5+ #355
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
Google 01/01/2011
RIP: 0010:ucma_copy_conn_param drivers/infiniband/core/ucma.c:1021 [inline]
RIP: 0010:ucma_connect+0x49d/0x870 drivers/infiniband/core/ucma.c:1042
RSP: 0018:ffff8801b7d678f8 EFLAGS: 00010202
RAX: 0000000000000000 RBX: ffff8801b7d67aa8 RCX: 0000000000000004
RDX: dffffc0000000000 RSI: 1ffff10036facf27 RDI: 0000000000000020
RBP: ffff8801b7d67ad0 R08: 0000000000000000 R09: 0000000000000000
R10: ffffffff886133c0 R11: 0000000000000000 R12: 1ffff10036facf21
R13: ffff8801b7d67968 R14: ffff8801bc250e40 R15: ffff8801b7d67928
FS: 00007f8b2af61700(0000) GS:ffff8801db200000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f8b2af3fe78 CR3: 00000001bb359004 CR4: 00000000001606f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
ucma_write+0x2d6/0x3d0 drivers/infiniband/core/ucma.c:1633
__vfs_write+0xef/0x970 fs/read_write.c:480
vfs_write+0x189/0x510 fs/read_write.c:544
SYSC_write fs/read_write.c:589 [inline]
SyS_write+0xef/0x220 fs/read_write.c:581
do_syscall_64+0x281/0x940 arch/x86/entry/common.c:287
entry_SYSCALL_64_after_hwframe+0x42/0xb7
RIP: 0033:0x445ad9
RSP: 002b:00007f8b2af60da8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 00000000006dac3c RCX: 0000000000445ad9
RDX: 0000000000000120 RSI: 00000000200001c0 RDI: 0000000000000003
RBP: 00000000006dac38 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 006d635f616d6472
R13: 2f646e6162696e69 R14: 666e692f7665642f R15: 0000000000000004
Code: b6 14 16 84 d2 74 09 80 fa 03 0f 8e a5 01 00 00 48 8d 78 20 89 8b 90
fe ff ff 48 ba 00 00 00 00 00 fc ff df 48 89 f9 48 c1 e9 03 <0f> b6 0c 11
48 89 fa 83 e2 07 83 c2 01 38 ca 7c 08 84 c9 0f 85
RIP: ucma_copy_conn_param drivers/infiniband/core/ucma.c:1021 [inline] RSP:
ffff8801b7d678f8
RIP: ucma_connect+0x49d/0x870 drivers/infiniband/core/ucma.c:1042 RSP:
ffff8801b7d678f8
---[ end trace 038c1f5808ef1630 ]---
Kernel panic - not syncing: Fatal exception
Dumping ftrace buffer:
(ftrace buffer empty)
Kernel Offset: disabled
Rebooting in 86400 seconds..


---
This bug is generated by a dumb bot. It may contain errors.
See https://goo.gl/tpsmEJ for details.
Direct all questions to syzk...@googlegroups.com.

syzbot will keep track of this bug report.
If you forgot to add the Reported-by tag, once the fix for this bug is
merged
into any tree, please reply to this email with:
#syz fix: exact-commit-title
If you want to test a patch for this bug, please reply with:
#syz test: git://repo/address.git branch
and provide the patch inline or as an attachment.
To mark this as a duplicate of another syzbot report, please reply with:
#syz dup: exact-subject-of-another-report
If it's a one-off invalid bug report, please reply with:
#syz invalid
Note: if the crash happens again, it will cause creation of a new bug
report.
Note: all commands must start from beginning of the line in the email body.

[Al Viro]

On Fri, Mar 16, 2018 at 04:59:02PM -0700, syzbot wrote:
> Hello,
>
> syzbot hit the following crash on upstream commit
> e2c15aff5f353ba80bd3bb49840837f65fa5cc43 (Thu Mar 15 18:07:35 2018 +0000)
> Merge tag 'sound-4.16-rc6' of
> git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
>
> So far this crash happened 2 times on upstream.
> C reproducer is attached.
> syzkaller reproducer is attached.
> Raw console output is attached.
> compiler: gcc (GCC) 7.1.1 20170620
> .config is attached.

May I politely inquire why am I getting Cc on the slew of ucma reports?
My last involvement with that thing that wouldn't have been absolutely
trivial had been back in 2012, and even that had been a conversion of
fget() to fdget() - nowhere near the areas implicated by those. Anything
more recent would have no impact on the object code - replacement of
POLL{IN,RDNORM} with EPOLL{IN,RDNORM} (equal values on x86) and two
replacements of unsigned int with __poll_t, which is typedefed to
unsigned.

It's not that I've objections against helping to debug that thing (other
than general aversion to drivers/infinibarf), but I'm really curious -
just what got me volunteered from the syzbot POV?

Al, digging through tons of unpleasant code in fs/{config,debug}fs
at the moment...