suyu-github/breakpad
0
0 Fork 0
mirror of https://github.com/suyu-emu/breakpad.git synced 2026-09-21 13:08:49 -07:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • 515f92cd61 Remove warnings about uninitialized fields. jimblandy@gmail.com 2009-07-27 21:28:57 +00:00
  • 22734848ea Port fixes from internal Google Breakpad to SVN. nealsid 2009-07-21 00:10:57 +00:00
  • 23c364a2b4 Fix for http://breakpad.appspot.com/18009 - run dump_syms on both PPC & i386 machines correctly, and process STABS/DWARF information in the same binary nealsid 2009-07-17 20:32:08 +00:00
  • f4f249e544 Integration test for Windows exception handler/minidump generation. nealsid 2009-07-17 01:15:48 +00:00
  • 6e525cbfbb Add stack-dumping logic to crash_report with -t switch nealsid 2009-07-02 00:30:44 +00:00
  • df55cf4313 Updated google mock external revision so we can use their own tuple implementation, so we can compile test cases under vs 2005 nealsid 2009-06-26 21:50:48 +00:00
  • 90a6505b03 The symbol files necessary for the processor to walk stacks on Linux that have __kernel_vsyscall at innermost frame. Special symbol files ar eneeded because EBP does not point to the frame when a thread is inside a system call. nealsid 2009-06-25 21:36:39 +00:00
  • ae4d8370ee Fix for crashs generation test app to not deadlock nealsid 2009-06-22 21:45:31 +00:00
  • 1d8cba25c7 Fix build break in crash_report caused by my checkin that moved minidump processing error codes to a different namespace nealsid 2009-06-16 22:31:51 +00:00
  • 61e88c7ad7 issue 323 - OS X exception handler / minidump generator should set exception address correctly for EXC_BAD_ACCESS . r=nealsid at http://breakpad.appspot.com/15002 ted.mielczarek 2009-06-12 20:36:53 +00:00
  • c52715f32f Fix for generating dumps on-demand to set the requesting thread to the current thread nealsid 2009-06-12 18:16:09 +00:00
  • dae47f0d16 Support overriding makefile variables, changed defaults compilation flags. This patch from Jim Blandy was Merged with a previous change from nealsid to force 32-bit compilation even on 64-bit machines. nealsid 2009-06-05 23:06:54 +00:00
  • 0eb52ff8cc Use ctsdio streams for dump_syms for significant speedup. Also contains a makefile fix to build in 32-bit mode, even on 64-bit systems. nealsid 2009-06-05 22:40:28 +00:00
  • 2eb356a68d Support custom URL parameters. Added unit tests for Breakpad. Added a way to specify server parameters in app plist file, as well. nealsid 2009-06-03 21:51:33 +00:00
  • 1a997295b6 Removed logging init macro from minidump_processor_unittest, since with the Google Test integration, logging is handled through it's facilities nealsid 2009-05-29 20:49:04 +00:00
  • e328c2dc58 Updated autoconf build files, and fixed externals dependency at revision 158(current as of 5/28) nealsid 2009-05-29 01:07:25 +00:00
  • b56cfa067a Add more error information to minidump processing return code. Also added dependency on google test, and modified minidump processing unit tests to use google test nealsid 2009-05-29 00:53:02 +00:00
  • aaecb48b3b Fix minidump_stackwalk compilation on gcc 4.3. Patch by Jim Blandy ted.mielczarek 2009-05-28 17:50:33 +00:00
  • 2a78191cab Show the reporter UI for Breakpad.framework regardless of whether continuing the crashed process succeeded (issue 319) stuartmorgan 2009-05-28 16:26:30 +00:00
  • 2a4698319a Fix up warnings when building http_upload.cc. Changed warning level to /W4, and also turned off deprecated 64-bit compatability warning flag(since some people are regularly building Breakpad in 64-bit mode, if we turn off this flag we get rid of a warning in 32 bit mode under VS2008) nealsid 2009-05-27 16:54:23 +00:00
  • bec07f6972 Patch from Jeremy to have better error reporting, and workaround a Cocoa bug in different locales nealsid 2009-05-15 17:54:20 +00:00
  • fc26f4a9b5 Fix memory leak in test case when calling into basic source line resolver. nealsid 2009-05-14 21:31:03 +00:00
  • dad1acc2d0 Deleting obsolete file(when I added DWARF support I had to make this a .MM but I forgot to delete the old file) nealsid 2009-05-12 17:19:59 +00:00
  • f9c5540ea6 Add a missing include (for gettimeofday) to Inspector.mm stuartmorgan 2009-04-30 01:11:36 +00:00
  • 33e8fad61d Mac reporter improvements: - Made localization for UI entirely string-based, with flexible layout based on the size of the strings inserted. - Made the request for an email address optional. - Fixed a bug that would prevent comments or email from being collected if the text field were still focused. - Refactored askUserPermissionToSend. stuartmorgan 2009-04-29 13:50:53 +00:00
  • 73afbc7302 Mac framework fixes: Propagate server type from plist, and fix precedence of defaults vs. plist stuart.morgan 2009-04-29 13:27:16 +00:00
  • ed1f6e754a Code review issue 9002: Add paranoid logging to Inspector & Reporter nealsid 2009-04-23 07:56:16 +00:00
  • e438d9cc0b Add a parameter to control the Mac reporter timeout. stuartmorgan 2009-04-22 13:22:08 +00:00
  • 1cb4ad73a6 Fix to support extensible parameter handling, and add process crashtime/uptime support nealsid 2009-04-22 06:25:21 +00:00
  • 04a7df7acf Give the Mac reporter app a localized name of 'Crash Reporter' stuartmorgan 2009-04-22 03:10:03 +00:00
  • ad6543e4e7 Fix memory leak when using the basic source line resolver, plus the optimization to load using in-memory buffers. Moved from manually allocating/deallocating memory to using a scoped_array nealsid 2009-04-22 00:55:31 +00:00
  • a0a0de39a8 Patch from Jeremy Moskovich to build with 10.5 SDK correctly. Also removed externals directory. nealsid 2009-04-21 08:29:46 +00:00
  • ea78675fbd AppVerifier leak test fix for Windows 7. doshimun 2009-04-08 18:58:12 +00:00
  • 841f5f7390 Sender tweaks; strip whitespace from server-returned ID, and change the title of the dialog a bit. stuart.morgan 2009-04-07 16:31:27 +00:00
  • 9290b7062f Fix framework bundle ID and remove useless autogenerated InfoPlist.strings stuart.morgan 2009-04-06 20:17:09 +00:00
  • 77d851c356 Fix my build break due to misnamed constant nealsid 2009-04-06 00:57:06 +00:00
  • a3d4c97336 Incorporate feedback from first Breakpad integration. nealsid 2009-04-06 00:25:29 +00:00
  • 3ebdb1bd7a Open sourcing the Breakpad framework from Google. nealsid 2009-04-01 03:18:49 +00:00
  • 0abe34ce5d issue 305 - breakpad Linux handler doesn't build with compiler built from latest GCC sources. Trivial patch by Jim Blandy ted.mielczarek 2009-03-24 19:57:46 +00:00
  • a18f4f9a06 Removing SVN external of GOogle Toolbox for mac, since we only need a few files that don't change very often nealsid 2009-03-23 20:22:45 +00:00
  • 3366749ee7 Fix for issue 304: symupload needs to support timeout specifications(wininet can timeout when sending large symbol files). R=doshimun W=nealsid nealsid 2009-03-20 19:02:12 +00:00
  • 2ad976ef0b Fix for issues 296, 297. Various symbol supplier classes need to be updated with new overload('make check' was failing, as well as crash_report), and remove logging that was flooding output nealsid 2009-02-27 23:59:40 +00:00
  • 02c244f1b2 Fix for issue 242, plus a redo of how old exception handlers are tracked, and called. nealsid 2009-02-26 21:31:53 +00:00
  • 4af5fe0b59 Fix build breaks on Solaris using Sun Studio. Written by Ginn Chen & Eagle.Lu@ R=nealsid (although I don't have a Solaris machine to build on, & these changes look localized to Sun-only builds) nealsid 2009-02-23 09:28:29 +00:00
  • e1a7efca7d See https://bugzilla.mozilla.org/show_bug.cgi?id=397200 The method of calculating a binary ID using the LC_ID command isn't compatible with non-default build processes, most Mac consumers use LC_UUID anyway but for those that don't, MD5 is a better choice R=nealsid W=Ted.Mielczarek nealsid 2009-02-22 08:16:05 +00:00
  • cadc8ddde9 Issue 294: mmap error checking is not correct. R=Liu nealsid 2009-02-22 01:31:43 +00:00
  • 0fd2f1ae21 Modify symbol supplier interface to support an overload that takes a symbol data buffer, to get around an extraneous read/write of symbol data nealsid 2009-02-19 21:26:20 +00:00
  • a285652cba Issue 275: patch to include list of loaded modules in crash_report output Written by: jum@mac.com R=nealsid nealsid 2009-02-10 21:29:54 +00:00
  • c9944d96e8 Reinstnate the call to TerminateThread inside ExceptionHandler destructor and put the graceful code inside a #ifdef. doshimun 2009-01-20 20:45:28 +00:00
  • 104e4e0114 Fix an AppVerifier STOP in OOP server code. In the destructor of the OOP server, we need to wait for any pending I/O to be done. doshimun 2009-01-16 22:37:48 +00:00
  • 308947d1ce Fix some typos in comments. doshimun 2009-01-14 21:54:08 +00:00
  • a6f58a1ac8 Minor fixes to windows exception handler. doshimun 2009-01-14 20:51:51 +00:00
  • 8b6271ce2c New test data to reflect Ted's changes that add function parameters to symbols when using DWARF nealsid 2008-12-19 22:28:45 +00:00
  • ab819f0680 Updating externals to include all of Google Toolbox for Mac nealsid 2008-12-11 03:25:39 +00:00
  • 3751b05354 Issue 283 - DWARF dumper doesn't handle DW_AT_specification. r=nealsid ted.mielczarek 2008-12-08 13:12:45 +00:00
  • d3441c2c96 Issue 284 - DWARF dumper doesn't output function names including arguments. r=nealsid ted.mielczarek 2008-12-08 13:00:29 +00:00
  • 7837cb7236 issue 286 - clean up some demangling code in dump_syms.mm. r=chris rogers ted.mielczarek 2008-12-03 18:23:09 +00:00
  • e5c401467b ExceptionHandler class is not handling initialization errors, such as the handler thread not being created. Protect the exception handler against failure to create semaphores and a thread handle. mmentovai 2008-12-03 17:21:34 +00:00
  • af553e22cb follow up for issue 281 - dump_syms fails to find dylib symbol file inside of a bundle. handle files without an extension properly ted.mielczarek 2008-12-01 18:42:16 +00:00
  • 7b8eb04745 issue 281 - dump_syms fails to find dylib symbol file inside of a bundle. r=nealsid ted.mielczarek 2008-12-01 17:05:43 +00:00
  • 5f27d9125c Followup to address some review comments from Issue 259 ted.mielczarek 2008-11-25 13:41:41 +00:00
  • 324d84c2f6 Issue 259 - crash when generating large .sym files with dump_syms on Linux. Patch by login_ing@yahoo.com, r=Liu Li ted.mielczarek 2008-11-24 20:47:37 +00:00
  • e055207058 Added some sanity checks when iterating over threads in a crashed process. In same cases the ESP could be 0. In other cases it would be borked completely(i.e. pointing to an invalid range). nealsid 2008-11-19 19:07:34 +00:00
  • f32d028110 Committing an externals directory to pull in google toolbox for mac nealsid 2008-11-14 20:16:13 +00:00
  • c85fb043ed Fix for dump_syms to ignore line number information for addresses that don't have an enclosing function nealsid 2008-11-03 23:23:50 +00:00
  • 3a283d8f8b Undo suspend/resume feature since it was meant as a workaround to a bug in how we handled child process exceptions. Currently we don't return to the kernel when we take an exception from a child process, causing a hang. Now we return KERN_FAILURE, indicating to the kernel to move on to the host-level exception handler(usually Crash Reporter) nealsid 2008-10-27 04:52:16 +00:00
  • 32441cc060 Issue 181: Add version info for Mac OS X modules. Found by iterating over load commands until I found LC_ID_DYLIB. Also modified crash_report to generate version number. Also added suspend/resume capability to exception handler, necessary because exception handling can behave strangely across fork() calls. Also added fix for filtering out functions with no line number information, and for filtering out some multiple inheritance glue the compiler generates. nealsid 2008-10-22 05:08:50 +00:00
  • c5f46b2f4b Issue 276 - generate GUIDs ahead of time in Linux handler. r=Liu Li ted.mielczarek 2008-10-14 11:21:34 +00:00
  • 1de70760db Fix for DWARF processing when inline functions were encountered. Inline functions were part of the debugging information as sequential in the PC range of the function but the source file was where the function was written(as you might expect). However the old dump_syms code, once it switched files as it was iterating over a range of addresses, never switched back, causing inline functions to cause all file numbers for addresses after the PC of the inline function to be wrong nealsid 2008-10-14 01:24:02 +00:00
  • bb067bf185 Fix warning when converting int to bool nealsid 2008-10-10 17:51:32 +00:00
  • 596e476fc2 fix a typo in the makefile, HANDLER_SRC->HANDLER_OBJ ted.mielczarek 2008-10-10 15:48:06 +00:00
  • cb4aa6b804 Added DWARF support to Breakpad client. Thanks to Google for open sourcing their DWARF code! Modified dump_syms to detect dSYM bundles or a binary with DWARF data appropriately, and convert data from DWARF reader to dump_syms native structures nealsid 2008-10-08 23:56:02 +00:00
  • a509f4ec4b Generate Windows full memory dumps as a separate file. Patch by Huan Ren <huanr@google.com>. r=me mmentovai 2008-09-16 18:50:02 +00:00
  • 9fcf4db315 Include what you use: include <algorithm> and <string.h> as needed. Patch by Robert Henry. r=me mmentovai 2008-09-15 18:16:49 +00:00
  • 11dc028694 Log a failure message in case of LOCATE_NOT_FOUND error. Without this change, no error is logged in this case. doshimun 2008-07-30 19:39:48 +00:00
  • 85534c2eaf Now we provide our own implementation of the MIG function exc_server, as recommended by Apple. See http://groups.google.com/group/google-breakpad-dev/browse_thread/thread/8b363b1f8a404714 For more information nealsid 2008-07-11 17:39:52 +00:00
  • 2d7664e8d4 In the destructor of ClientInfo, currently the wait on the clent_crashed_event is unregistered after the handle is closed. MSDN clearly says that the behavior is undefined in this case. See http://msdn.microsoft.com/en-us/library/ms685061(VS.85).aspx doshimun 2008-06-09 23:38:24 +00:00
  • 3a516e4177 Client process uptime is a very useful metric for crashes. Instead of each client process having to keep track of uptime and report it whenever a crash happens, integrate this feature into breakpad OOP code so that all clients get it for free. doshimun 2008-06-06 21:41:32 +00:00
  • 38bc56cfa8 Modified symupload to correctly handle spaces in module names when processing symbol file Reviewer=mmentovai nealsid 2008-06-05 21:20:53 +00:00
  • cc6cd1ff37 Fix the following bug: The sender updates the checkpoint file even when the report is not successfully sent. doshimun 2008-06-05 20:59:35 +00:00
  • d55398a492 Filter out duplicates from the image list when creating a minidump. See issue 266 for more details. nealsid 2008-06-04 22:57:09 +00:00
  • f6b3e16ef8 Set the dump thread id and the requesting thread id properties of MDRawBreakpadInfo only in case of in-process dump generation; otherwise the dump processing code of breakpad doesn't identify the crashing thread properly. doshimun 2008-05-31 19:23:20 +00:00
  • b91bb881d5 Fix to ignore source files that have a text address of 0 in their stabs entry nealsid 2008-05-19 17:27:40 +00:00
  • 72bafd378d Microsoft broke PROCESS_ALL_ACCESS when we build with WINVER=0x600. See http://nsylvain.blogspot.com/2008/01/winverwin32winnt-mayhem.html for details (second half of the post). doshimun 2008-05-15 04:12:34 +00:00
  • 808ff9bc70 64-bit processor can't handle 32-bit ppc minidumps due to struct padding differences on 64- and 32-bit CPUs. Force 32-bit alignment rules for 32-bit ppc context struct. r=nealsid mmentovai 2008-05-14 18:24:23 +00:00
  • d318f9efee Fix to stop unit tests from being compiled & run for non-native architectures nealsid 2008-05-08 23:15:00 +00:00
  • 7ab126ecdd Replace the user of wcscpy_s with safe_wcscpy in common/windows/string_utils-inl.h to make the code compatible for teams that still use VS 2003 to build. doshimun 2008-05-08 18:15:07 +00:00
  • 8602aa71ec Some style guide compliance changes and changed the prototype of the method to get custom client info in ClientInfo class. doshimun 2008-05-07 21:54:12 +00:00
  • c2bdaa613c Bug fix for issue 263: mach_vm_region_recurse calls have invalid parameter specified, which succeeds on Leopard but fails on Tiger Reviewer: Chris Rogers(ladderbreaker) nealsid 2008-05-07 21:15:12 +00:00
  • 3d4dc1f201 Add a convenience method to CustomClientInfo struct to set both name and value in one call. doshimun 2008-05-05 21:03:51 +00:00
  • 0ded3d718f Add a way for the client apps to specify custom information in case of out-of-process scenarios that the OOP server can use in whatever way it wants to. doshimun 2008-05-05 20:03:56 +00:00
  • 47df365bf8 Issue 261/262: 64-bit bug fix when iterating through load commands of a 64-bit binary, and added new test case for ReadImageInfoForTask nealsid 2008-05-05 18:20:04 +00:00
  • d6f0f92de3 Style changes nealsid 2008-04-30 00:56:17 +00:00
  • d906d0f7c9 Updated code to be compliant with style guidelines nealsid 2008-04-29 23:49:39 +00:00
  • 4c39c138fe Issue 258: Added test cases for ReadTaskMemory, reorganized project file, renamed filenames inside comments nealsid 2008-04-25 00:37:19 +00:00
  • 9033edcd7b Add one more parameter to the ClientDumpRequestCallback in crash generation server to pass in the path of the dump file if the dump was generated successfully. doshimun 2008-04-17 21:21:48 +00:00
  • 9609db8173 Fix a bug in start address calculation nealsid 2008-04-17 04:15:30 +00:00
  • 5da03791a7 As part of resolving issue 256 I'd like to check in some unit tests(ok, just one) for the private copy of nlist that I checked in last week, plus lay some of the ground work for collecting code coverage numbers. Both of these are accomplished by leveraging the built-in facilities of Xcode & Developer Tools(namely, CPlusTest and gcov integration; however, eventually I will also add a way to get lcov results from the gcov results). I also: nealsid 2008-04-15 23:09:30 +00:00
  • b0d807666f Issue 254: The problem is that nlist() is compiled out of libc in 64-bit builds of Leopard. I ported the code over myself and will check it into the Breakpad tree until we decide there's a better longer-term solution. If you want to diff the changes, the Apple source is in libc/gen/nlist.c(I used 498 from 10.5.2). nealsid 2008-04-12 01:16:18 +00:00
  • 35c34d6bb9 Add a parameter to CrashGenerationServer to let the callers set the security attributes on the pipe. This is useful when the process hosting the crash server wants to let processes from other users to connect to it and generate dumps for them. doshimun 2008-04-12 00:00:34 +00:00