Hi, > No luck. I had hoped we could at least see it it hangs in the DVB driver > or somewhere else in the USB stack. If you can produce a trace with > DVB or USB functions in it then please post. > > > Thanks, > Johannes Is that better? I thought "Lets try 2.6.9-mm1", maybe the problem magically disappears, which it does not :-( But, at least, I see some dvb messages in the "Call trace" section, so maybe this is more helpful? Dvb-kernel from today's CVS. eip: e09d7434 kernel BUG at include/asm/spinlock.h:149! invalid operand: 0000 [#1] CPU: 1 EIP: 0060:[<c02b843f>] Not tainted VLI Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00010082 (2.6.9-mm1) eax: 0000000e ebx: 00000286 ecx: c02f648c edx: 00000082 esi: d8562024 edi: 00000000 ebp: da8c3e28 esp: da8c3e18 ds: 007b es: 007b ss: 0068 Stack: c02c637f e09d7434 d8562000 d8562024 da8c3e3c e09d7434 e0a5a000 d8562000 d8562648 da8c3e58 e09d77d7 df072520 e0a5a000 00000000 00000000 e0a5a000 da8c3e70 e09f5174 d8562464 00000000 00000010 e0a76000 da8c3ebc e09f2bb4 Call Trace: [<c0104d9a>] show_stack+0x7a/0x90 [<c0104f22>] show_registers+0x152/0x1d0 [<c0105130>] die+0xf0/0x180 [<c01055e4>] do_invalid_op+0xd4/0xf0 [<c0104a1d>] error_code+0x2d/0x38 [<e09d7434>] dibusb_get_free_pid+0x14/0x60 [dvb_dibusb] [<e09d77d7>] dibusb_start_feed+0x27/0x90 [dvb_dibusb] [<e09f5174>] dmx_section_feed_start_filtering+0xa4/0x160 [dvb_core] [<e09f2bb4>] dvb_dmxdev_filter_start+0x2c4/0x3c0 [dvb_core] [<e09f3425>] dvb_demux_do_ioctl+0x1a5/0x280 [dvb_core] [<e09f183a>] dvb_usercopy+0xaa/0x119 [dvb_core] [<e09f351a>] dvb_demux_ioctl+0x1a/0x20 [dvb_core] [<c0163c45>] sys_ioctl+0x1d5/0x230 [<c0103f61>] sysenter_past_esp+0x52/0x71 Code: 00 74 01 fb f3 90 80 3e 00 7e f9 fa eb e8 89 d8 8b 75 fc 8b 5d f8 c9 c3 8b 45 04 c7 04 24 7f 63 2c c0 89 44 24 04 e8 71 16 e6 ff <0f> 0b 95 00 64 5b 2c c0 eb c1 8d b4 26 00 00 00 00 55 89 e5 53 >>EIP; c02b843f <_read_lock_irq+1f/30> <===== >>ecx; c02f648c <log_wait+0/10> >>esi; d8562024 <pg0+18169024/3fc05400> >>ebp; da8c3e28 <pg0+1a4cae28/3fc05400> >>esp; da8c3e18 <pg0+1a4cae18/3fc05400> Trace; c0104d9a <show_stack+7a/90> Trace; c0104f22 <show_registers+152/1d0> Trace; c0105130 <die+f0/180> Trace; c01055e4 <do_invalid_op+d4/f0> Trace; c0104a1d <error_code+2d/38> Trace; e09d7434 <pg0+205de434/3fc05400> Trace; e09d77d7 <pg0+205de7d7/3fc05400> Trace; e09f5174 <pg0+205fc174/3fc05400> Trace; e09f2bb4 <pg0+205f9bb4/3fc05400> Trace; e09f3425 <pg0+205fa425/3fc05400> Trace; e09f183a <pg0+205f883a/3fc05400> Trace; e09f351a <pg0+205fa51a/3fc05400> Trace; c0163c45 <sys_ioctl+1d5/230> Trace; c0103f61 <sysenter_past_esp+52/71> This architecture has variable length instructions, decoding before eip is unreliable, take these instructions with a pinch of salt. Code; c02b8414 <_read_lock_irqsave+24/30> 00000000 <_EIP>: Code; c02b8414 <_read_lock_irqsave+24/30> 0: 00 74 01 fb add %dh,0xfffffffb(%ecx,%eax,1) Code; c02b8418 <_read_lock_irqsave+28/30> 4: f3 90 pause Code; c02b841a <_read_lock_irqsave+2a/30> 6: 80 3e 00 cmpb $0x0,(%esi) Code; c02b841d <_read_lock_irqsave+2d/30> 9: 7e f9 jle 4 <_EIP+0x4> Code; c02b841f <_read_lock_irqsave+2f/30> b: fa cli Code; c02b8420 <_read_lock_irq+0/30> c: eb e8 jmp fffffff6 <_EIP+0xfffffff6> Code; c02b8422 <_read_lock_irq+2/30> e: 89 d8 mov %ebx,%eax Code; c02b8424 <_read_lock_irq+4/30> 10: 8b 75 fc mov 0xfffffffc(%ebp),%esi Code; c02b8427 <_read_lock_irq+7/30> 13: 8b 5d f8 mov 0xfffffff8(%ebp),%ebx Code; c02b842a <_read_lock_irq+a/30> 16: c9 leave Code; c02b842b <_read_lock_irq+b/30> 17: c3 ret Code; c02b842c <_read_lock_irq+c/30> 18: 8b 45 04 mov 0x4(%ebp),%eax Code; c02b842f <_read_lock_irq+f/30> 1b: c7 04 24 7f 63 2c c0 movl $0xc02c637f,(%esp) Code; c02b8436 <_read_lock_irq+16/30> 22: 89 44 24 04 mov %eax,0x4(%esp) Code; c02b843a <_read_lock_irq+1a/30> 26: e8 71 16 e6 ff call ffe6169c <_EIP+0xffe6169c> This decode from eip onwards should be reliable Code; c02b843f <_read_lock_irq+1f/30> 00000000 <_EIP>: Code; c02b843f <_read_lock_irq+1f/30> <===== 0: 0f 0b ud2a <===== Code; c02b8441 <_read_lock_irq+21/30> 2: 95 xchg %eax,%ebp Code; c02b8442 <_read_lock_irq+22/30> 3: 00 64 5b 2c add %ah,0x2c(%ebx,%ebx,2) Code; c02b8446 <_read_lock_irq+26/30> 7: c0 eb c1 shr $0xc1,%bl Code; c02b8449 <_read_lock_irq+29/30> a: 8d b4 26 00 00 00 00 lea 0x0(%esi),%esi Code; c02b8450 <_read_lock_bh+0/40> 11: 55 push %ebp Code; c02b8451 <_read_lock_bh+1/40> 12: 89 e5 mov %esp,%ebp Code; c02b8453 <_read_lock_bh+3/40> 14: 53 push %ebx With kind regards, Phil
Attachment:
signature.asc
Description: This is a digitally signed message part