topic Re: DECthreads bugcheck in Operating System - OpenVMS

DECthreads bugcheck

Feige — Sun, 13 Jan 2008 04:46:08 GMT

When I start the EXMAN process in OpenVMS 8.2 on HP Alpha DS25 Platform,the pthread.dump files as follows:
%DECthreads bugcheck (version V3.22-077), terminating execution.
%Reason: lckMcsLock: deadlock detected, cell = 0x0000000000AD1300
%Running on OpenVMS V8.2 on AlphaServer DS25, 4096Mb; 1 CPUs
% The bugcheck occurred at 12-JAN-2008 17:20:35.28, running image
% $1$DGA1:[BBS.EXECUTABLE]CSROUTER.EXE;1 in process 202018CA (named
% "csrouter"), under username "SYSTEM". AST delivery is enabled for all modes;
% no ASTs active. Upcalls are disabled. Multiple kernel threads are disabled.
% The current thread sequence number is 2, at 0x00AD1300
% Current thread traceback:
% 0: PC 0x80A3BAA4, FP 0x00ACF070, DESC 0x7BB96B80
% 1: PC 0x80A421AC, FP 0x00ACF130, DESC 0x7BB97F20
% 2: PC 0x80A3DE18, FP 0x00ACF1E0, DESC 0x7BB97558
% 3: PC 0x80A3FCE4, FP 0x00ACF200, DESC 0x7BB977C8
% 4: PC 0x80A447E4, FP 0x00ACF240, DESC 0x7BB99048
% 5: PC 0x8099FCD4, FP 0x00ACF2A0, DESC 0x7B978658
% 6: PC 0x0004F248, FP 0x00ACF2D0, DESC 0x00018530
% 7: PC 0x8007D774, FP 0x00ACF340, DESC 0x81945D18
% 8: PC 0x801468E0, FP 0x00ACF390, DESC 0x818E3FD0
% 9: PC 0x8015BA50, FP 0x00ACF400, DESC 0x818E8E78
% 10: PC 0x80A371FC, FP 0x00ACF5F0, DESC 0x7BB96300
% 11: PC 0x80A36FCC, FP 0x00ACF640, DESC 0x7BB96590
% 12: PC 0x80A3A4EC, FP 0x00ACF6D0, DESC 0x7BB96690
% 13: PC 0x80A6A90C, FP 0x00ACF700, DESC 0x7BB9E018
% 14: PC 0x8015BA50, FP 0x00ACF740, DESC 0x818E8E78
% 15: PC 0x8015F5C4, FP 0x00ACF940, DESC 0x818E6D10
% 16: PC 0x00104664, FP 0x00ACF9C0, DESC 0x0012FA30
% 17: PC 0x80ABB444, FP 0x00ACFA60, DESC 0x7BE97948
% 18: PC 0x00033124, FP 0x00ACFB50, DESC 0x00010C30
% 19: PC 0x00035204, FP 0x00ACFD70, DESC 0x00011500
% 20: PC 0x0003B068, FP 0x00ACFDA0, DESC 0x00012B48
% 21: PC 0x80A5117C, FP 0x00ACFDE0, DESC 0x7BB9A670
% 22: PC 0x80A3D940, FP 0x00ACFFE0, DESC 0x7BB973B0
% 23: PC 0x00000000, FP 0x7AE69A50, DESC 0x7BB94380
% 24: PC 0x80339F94, FP 0x7AE6DB30, DESC 0x81946050
% 25: PC 0x7AF97A64, FP 0x7AE6DBB0, DESC 0x7AF1CFC0

The attachment is OpenVMS OS patch.
Please help me, It's very urgent.

Re: DECthreads bugcheck

Karl Rohwedder — Sun, 13 Jan 2008 13:11:32 GMT

AS a general advice I would get current with VMS patches and then try again.

regards Kalle

Re: DECthreads bugcheck

Feige — Sun, 13 Jan 2008 13:47:20 GMT

Hi Kalle,

Thank you very much!

which patches? your mearn "VMS82A_PTHREAD-V0200.ZIPEXE"? any other patches?

Re: DECthreads bugcheck

Hoff — Sun, 13 Jan 2008 14:39:53 GMT

There's a list of available and of mandatory patches at the ITRC FTP server, posted in a text file the patch directory for the particular release. The text filename contains Master Index or some such.

If the failure is not repaired by a DECthreads ECO (and it may not be), you'll probably have to do debugging around what threads are contending for the specified lock(s) involved. (And debugging that requires a whole lot more application context and application knowledge than is provided by a DECthreads stackdump.)

Re: DECthreads bugcheck

Ian Miller. — Mon, 14 Jan 2008 09:59:09 GMT

The list of patches for OpenVMS Alpha V8.2 is at
ftp://ftp.itrc.hp.com/openvms_patches/alpha/V8.2/ALPHA_V82A_MASTER_ECO_LIST.txt

I would start with
VMS83A_UPDATE-V1000

which does include the aforementioned PTHREADS patch.

Re: DECthreads bugcheck

John Travell — Mon, 14 Jan 2008 20:41:20 GMT

Ian, you're slipping! VMS83A_UPDATE-V1000 on V8.2 ??

VMS82A_PTHREAD-V0200 appears to be the latest relevant kit for V8.2, and is included with a whole load of other good stuff in VMS82A_UPDATE-V1000, available as specified earlier at:
ftp://ftp.itrc.hp.com/openvms_patches/alpha/V8.2/
JT:

Re: DECthreads bugcheck

Feige — Tue, 15 Jan 2008 00:59:10 GMT

Thanks Everyone

the VMS82A_PTHREAD-V0200 patch fix the problem.

Thanks

Re: DECthreads bugcheck

Feige — Tue, 15 Jan 2008 01:00:00 GMT

Thanks everyone.

VMS82A_PTHREAD-V0200 fix the problem.

Re: DECthreads bugcheck

Ian Miller. — Tue, 15 Jan 2008 10:56:26 GMT

I guess I meant VMS82A_UPDATE-V1000 not
VMS83A_UPDATE-V1000.

Anyway it appears a fix has been found.