postmortem-debugging

Crashdump analyzing: CxxUnhandledExceptionFilter

由 若如初见. 提交于 2019-12-07 13:45:02
问题 I've got a crash minidump to analyze. My program is multithread Qt5 application. I'm not a debugging guru but usualy I can easily find place where program failed, but this time I can't. I opened dump file in Visual Studio 2010, clicked "Debug with native only" and it shows me where problem is: it is thread with location "__CxxUnhandledExceptionFilter". Call stack is like this: msvcr100.dll()!_abort() msvcr100.dll()!terminate() program.exe!__CxxUnhandledExceptionFilter(_EXCEPTION_POINTERS *

Tool for analyzing .Net app memory dumps

由 ☆樱花仙子☆ 提交于 2019-12-05 10:19:22
问题 Can somebody suggest a good free tool for analyzing .Net memory dumps other than Adplus/windbg/sos ? 回答1: You can load sos and your memory dump into Visual Studio to at least insulate you from the 'interesting' ui that WinDbg presents. 回答2: You can try out DebugDiag 1.1 回答3: Take a look at SOS Assist, it provides a GUI around SOS. 回答4: I fully recommend .Net Memory Profiler. Beside being a great live memory profiler for .Net applications, it can also load memory dumps, and let you traverse

How do you debug a LONG RUNNING php script?

由 ぃ、小莉子 提交于 2019-12-02 09:15:20
The problem is that the script hangs up after some long time. strace returns something like this and nothing else: Process 7286 attached - interrupt to quit restart_syscall(<... resuming interrupted call ...>) = 0 poll([{fd=13, events=POLLIN|POLLPRI|POLLRDNORM|POLLRDBAND}], 1, 0) = 0 (Timeout) clock_gettime(CLOCK_MONOTONIC, {1817569, 74651533}) = 0 clock_gettime(CLOCK_MONOTONIC, {1817569, 74734744}) = 0 clock_gettime(CLOCK_MONOTONIC, {1817569, 74812047}) = 0 poll([{fd=13, events=POLLIN|POLLPRI|POLLRDNORM|POLLRDBAND}], 1, 1000) = 0 (Timeout) poll([{fd=13, events=POLLIN|POLLPRI|POLLRDNORM

Get the right .net native symbols for Windbg

由 早过忘川 提交于 2019-12-01 06:58:30
I'm doing some crash dump debugging, where I am looking a dump taken from a production server. The machine I'm running WinDbg on must have a slightly different version of the .NET runtime installed -- I'm getting errors loading the native images of .NET system assemblies (so can't load for example System.Data.Linq ). What is the best way to ensure that my debug machine has access to all the right symbols? Edit Added output of lmv for Thomas Weller 000007fb`68660000 000007fb`68993000 System_Data_Linq_ni C (pdb symbols) C:\Program Files\Debugging Tools for Windows (x64)\sym\System.Data.Linq.pdb

Get the right .net native symbols for Windbg

由 本秂侑毒 提交于 2019-12-01 04:30:22
问题 I'm doing some crash dump debugging, where I am looking a dump taken from a production server. The machine I'm running WinDbg on must have a slightly different version of the .NET runtime installed -- I'm getting errors loading the native images of .NET system assemblies (so can't load for example System.Data.Linq ). What is the best way to ensure that my debug machine has access to all the right symbols? Edit Added output of lmv for Thomas Weller 000007fb`68660000 000007fb`68993000 System

Cross-platform crash handler

由 女生的网名这么多〃 提交于 2019-11-30 11:27:23
I'm looking for a cross-platform crash handler. Google Breakpad looks promising, but it is sorely lacking any documentation, and requires a reasonable amount of fiddling to actually get going. What is a better alternative? All I need is the ability to reliably record crash dumps, stack traces, and CPU information at the time of a crash. Alternatively, what is the experience using Google Breakpad? Has it been great or horrible? Well, it turns out that google-breakpad is pretty nice after all. It's not totally easy to set up, but it's OK for what I need. I wrote a article about dumping the stack
订阅 postmortem-debugging