kernel BUG at /usr/src/linux-2.5.37/include/asm/spinlock.h:123!
This happened on 2.5.37 with KDE artsd and OSS sound drivers.
MikaL
Sep 21 22:12:10 devil kernel: eip: c02471a0 Sep 21 22:12:10 devil kernel: kernel BUG at /usr/src/linux-2.5.37/include/asm/spinlock.h:123! Sep 21 22:12:10 devil kernel: invalid operand: 0000 Sep 21 22:12:10 devil kernel: CPU: 1 Sep 21 22:12:10 devil kernel: EIP: 0060:[dma_ioctl+920/2634] Not tainted Sep 21 22:12:10 devil kernel: EFLAGS: 00210082 Sep 21 22:12:10 devil kernel: eax: 0000000e ebx: 40045010 ecx: d2a87e70 edx: c0368800 Sep 21 22:12:10 devil kernel: esi: 40045038 edi: d98220fc ebp: 40045038 esp: d2a87f00 Sep 21 22:12:10 devil kernel: ds: 0068 es: 0068 ss: 0068 Sep 21 22:12:10 devil kernel: Process artsd (pid: 545, threadinfo=d2a86000 task=d2dca020) Sep 21 22:12:10 devil kernel: Stack: c032bacd c02471a0 40045010 bffff6ec 00000000 40045010 d9822000 00000000 Sep 21 22:12:10 devil kernel: 00200202 00000002 d9822574 00000000 d5135a20 c0125226 00000000 00001000 Sep 21 22:12:10 devil kernel: 00000000 00000000 c0246a4c 00000000 40045010 bffff6ec 00000004 00000003 Sep 21 22:12:10 devil kernel: Call Trace: [dma_ioctl+896/2634] [update_wall_time+18/60] [audio_ioctl+1452/1512] [sound_ioctl+341/380] [sys_ioctl+639/748] Sep 21 22:12:10 devil kernel: [syscall_call+7/11] Sep 21 22:12:10 devil kernel: Sep 21 22:12:10 devil kernel: Code: 0f 0b 7b 00 a0 ba 32 c0 83 c4 08 f0 fe 0e 0f 88 aa 06 00 00
Quoted 21 lines · Mika Liljeberg
This happened on 2.5.37 with KDE artsd and OSS sound drivers.
MikaL
Sep 21 22:12:10 devil kernel: eip: c02471a0 Sep 21 22:12:10 devil kernel: kernel BUG at /usr/src/linux-2.5.37/include/asm/spinlock.h:123! Sep 21 22:12:10 devil kernel: invalid operand: 0000 Sep 21 22:12:10 devil kernel: CPU: 1 Sep 21 22:12:10 devil kernel: EIP: 0060:[dma_ioctl+920/2634] Not tainted Sep 21 22:12:10 devil kernel: EFLAGS: 00210082 Sep 21 22:12:10 devil kernel: eax: 0000000e ebx: 40045010 ecx: d2a87e70 edx: c0368800 Sep 21 22:12:10 devil kernel: esi: 40045038 edi: d98220fc ebp: 40045038 esp: d2a87f00 Sep 21 22:12:10 devil kernel: ds: 0068 es: 0068 ss: 0068 Sep 21 22:12:10 devil kernel: Process artsd (pid: 545, threadinfo=d2a86000 task=d2dca020) Sep 21 22:12:10 devil kernel: Stack: c032bacd c02471a0 40045010 bffff6ec 00000000 40045010 d9822000 00000000 Sep 21 22:12:10 devil kernel: 00200202 00000002 d9822574 00000000 d5135a20 c0125226 00000000 00001000 Sep 21 22:12:10 devil kernel: 00000000 00000000 c0246a4c 00000000 40045010 bffff6ec 00000004 00000003 Sep 21 22:12:10 devil kernel: Call Trace: [dma_ioctl+896/2634] [update_wall_time+18/60] [audio_ioctl+1452/1512] [sound_ioctl+341/380] [sys_ioctl+639/748] Sep 21 22:12:10 devil kernel: [syscall_call+7/11] Sep 21 22:12:10 devil kernel: Sep 21 22:12:10 devil kernel: Code: 0f 0b 7b 00 a0 ba 32 c0 83 c4 08 f0 fe 0e 0f 88 aa 06 00 00
Quoted 5 lines
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Mika,
I haven't looked at the code yet... but most of the time when you see this error it is because the kernel is compiled with CONFIG_DEBUG_SPINLOCK and the code is using a lock that hasn't be initilized correctly (i.e.:SPIN_LOCK_UNLOCKED().
Signature · 2 lines
Bob Miller Email: [email protected] Open Source Development Lab Phone: 503.626.2455 Ext. 17
Quoted 1 line · Bob Miller
Mika,
Hi Bob,
Quoted 4 lines
I haven't looked at the code yet... but most of the time when you see this error it is because the kernel is compiled with
CONFIG_DEBUG_SPINLOCKand the code is using a lock that hasn't be initilized correctly (i.e.:SPIN_LOCK_UNLOCKED().
Yup. I usually compile new kernels with spinlock debugging.
Looks like a number of the ioctls in oss/audio.c have locking problems. At least the following ones look buggy:
SNDCTL_DSP_SETTRIGGERSNDCTL_DSP_GETIPTRSNDCTL_DSP_GETOPTRSNDCTL_DSP_GETODELAY
Each of these tries to manipulate a spinlock through the uninitialized dmap pointer. SNDCTL_DSP_SETTRIGGER actually operates on both dmap_in and dmap_out. It should presumably acquire locks for both.
Looks like a copy/paste error, probably a result of a hasty conversion cli()/sti()'s to spinlocks.
Regards,
MikaL
Quoted 4 lines
Signature · 3 lines
-- Bob Miller Email: [email protected] Open Source Development Lab Phone: 503.626.2455 Ext. 17