[ofw] CM ref counting issues...
Fab Tillier
ftillier at microsoft.com
Tue Dec 8 19:45:10 PST 2009
Hi Sean
Hefty, Sean wrote on Tue, 8 Dec 2009 at 17:13:05
> Here are stack dumps that touch the IB stack when the application
> hangs. The trace for the hanging process is the last one listed. The
> application is exiting, causing winverbs to cleanup in the kernel. I
> have cm_destroy_id call DbgBreakPoint after it times out. Also, note
> that both the client and server are running on the same system.
>
> Looking at the list, nothing obvious stands out that should cause a
> deadlock. Is there anything special about how MADs are processed in
> the loopback case, or the handling of the CM timewait state?
>
> 4.00005c fffffa80036c2720 fffe752f Blocked nt!KiSwapContext+0x7f
Why is this thread blocked trying to disconnect the QP? Could this be holding the ref on the CM ID?
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> winverbs!WvProviderDisableRemove+0x4a
> winverbs!WvQpAcquire+0x2a
> winverbs!WvEpDisconnectQp+0x31
> winverbs!WvEpDisconnectHandler+0xcd
> winverbs!WorkQueueHandler+0x4b
> nt!IopProcessWorkItem+0x27
> nt!ExpWorkerThread+0x11a
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000cc fffffa8003b7f890
> ffffff9b Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> ibbus!cl_event_wait_on+0x33
> ibbus!__cl_thread_pool_routine+0x2b
> ibbus!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000e0 fffffa8003b87040
> ffffff4b Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForMultipleObjects+0x2d6
> mthca!eq_polling_thread+0x77
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000e4 fffffa8004028040
> ffff3baf Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> ibbus!cl_event_wait_on+0x33
> ibbus!__cl_thread_pool_routine+0x2b
> ibbus!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000e8 fffffa8004028bb0
> fffe7791 Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> ibbus!cl_event_wait_on+0x33
> ibbus!__cl_thread_pool_routine+0x2b
> ibbus!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000ec fffffa8004028720
> fffe7442 Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> ibbus!cl_event_wait_on+0x33
> ibbus!__cl_thread_pool_routine+0x2b
> ibbus!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000f0 fffffa8004031040
> ffffff4b Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> mthca!__cl_thread_pool_routine+0x31
> mthca!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 4.0000f8 fffffa80043eabb0
> ffff3baf Blocked
> nt!KiSwapContext+0x7f
> nt!KiSwapThread+0x2fa
> nt!KeWaitForSingleObject+0x2da
> ipoib!__cl_thread_pool_routine+0x31
> ipoib!__thread_callback+0x28
> nt!PspSystemThreadStartup+0x57
> nt!KxStartSystemThread+0x16
> 6dc.000458 fffffa80076f4bb0
> fffe7329 RUNNING
> nt!DbgBreakPoint
> ibbus!cm_destroy_id+0x6c
> winverbs!WvEpFree+0x74
> winverbs!WvProviderCleanup+0x76
> Wdf01000!FxPkgGeneral::OnCleanup+0x82
>
> Wdf01000!FxPkgGeneral::Dispatch+0x1ce
>
> Wdf01000!FxDevice::Dispatch+0xa9
> nt!IopCloseFile+0x184
> nt!ObpDecrementHandleCount+0xc0
> nt!ObpCloseHandleTableEntry+0xb7
> nt!ObpCloseHandleProcedure+0x30
> nt!ExSweepHandleTable+0x73
> nt!ObKillProcess+0x61
> nt!PspExitThread+0x4b8
> nt!PspTerminateThreadByPointer+0x4d
> nt!NtTerminateProcess+0xfa
> nt!KiSystemServiceCopyEnd+0x13
> ntdll!ZwTerminateProcess+0xa
> ntdll!RtlExitUserProcess+0x8b
> msvcrt!cinit+0x13b +0xff474be8
More information about the ofw
mailing list