Jeremy Rifkin
b364d37f78
Implement better dwarf rangelist base address logic ( #157 )
...
Related to https://github.com/davea42/libdwarf-code/issues/255 , allows
re-enabling the clang+sanitizer+rel+dsym tests.
2024-08-18 16:51:40 -05:00
Jeremy Rifkin
64e0210449
Better handle resolution of safe object frames with empty object paths
2024-08-18 13:38:06 -05:00
Jeremy Rifkin
0e2c3a130d
Fix windows build
2024-08-18 12:11:20 -05:00
Jeremy Rifkin
973309cb22
Some more platform refactoring/cleanup, also add a note to readme
2024-08-18 12:07:34 -05:00
Jeremy Rifkin
164cc75681
A little refactoring and reorganization. Remove relative includes.
2024-08-18 11:58:49 -05:00
Jeremy Rifkin
7621f2b277
A couple notes/comments/documentation tweaks
2024-08-18 11:20:59 -05:00
Jeremy Rifkin
26e009c688
Add cpptrace::from_current ( #155 )
2024-08-18 11:11:47 -05:00
Jeremy
cc2b3fce40
Add another try/catch to prevent any potential TCO
2024-07-13 19:07:58 -05:00
Jeremy
dcf89743f6
Block TCO on wrapper functions
2024-07-13 19:06:33 -05:00
Jeremy
c35ad6887a
Fix issue with incorrect reported object addresses on macos when using debug maps
2024-07-13 19:06:13 -05:00
Jeremy
e1fb3527a4
Fix an issue with split dwarf on clang
2024-07-11 19:10:16 -05:00
Jeremy
611660de5b
Fix use after free during cleanup of split dwarf information, fixes #141
2024-07-02 21:56:35 -06:00
Jeremy
1f19c31e02
Fix bug with dladdr1 not being used, silly typo
2024-06-19 15:15:51 -05:00
Jeremy Rifkin
2fb0a6fd2e
Fix trace generation for cpptrace exception objects on armv7hf ( #138 )
...
noexcept on get_raw_trace_and_absorb seems to somehow interfere with
unwinding on armv7hf, possibly due to unwind tables not being generated
for the function. Fixes #134 .
2024-06-13 11:17:44 -05:00
Jeremy
bb83fa922c
Merge branch 'dev'
2024-06-11 22:30:27 -05:00
Jeremy
8d7a85c447
Update flatten_inlines to properly move inline frames instead of copying them and also clarify a comment
2024-06-04 23:21:14 -05:00
Infko
72b0e467d7
fix compile error in gcc 4.8.5. ( #133 )
2024-06-04 23:16:34 -05:00
Jeremy
b8a6e8446a
Quick microfmt fix for an old msvc bug
2024-06-02 12:47:53 -05:00
Jeremy
44ddc1b72f
Merge branch 'main' into dev
2024-06-02 11:35:01 -05:00
Jeremy
59ef49f1a3
Update microfmt
2024-06-02 11:31:43 -05:00
Jeremy
70634cb6f4
Undo temporary change printing object addresses in stack traces
2024-05-28 21:44:13 -05:00
Jeremy
9d1c592f0c
Remove CPPTRACE_CONDA_LIBDWARF_WEIRDNESS bodge and better support libdwarf via pkgconfig, resolves #131
2024-05-28 21:32:54 -05:00
Jeremy Rifkin
b2bff57d81
Debug Fission ( #132 )
...
Implement support for debug fission, closes #129
2024-05-28 20:39:34 -05:00
Jeremy
5556aedddb
Refactor cu lookup out of pc resolution
2024-05-25 16:20:41 -05:00
Jeremy
8bff5dc9fe
Quick fix
2024-05-25 14:32:34 -05:00
Jeremy
810b74ecaa
Refactor symbols_with_libdwarf indo separate files
2024-05-25 14:29:55 -05:00
Jeremy
5171e87a51
Refactor logic in the libdwarf resolver's resolve_frames, pull out trace flattening
2024-05-25 13:38:06 -05:00
Jeremy
a9f72c6f67
Refactor logic in the libdwarf resolver's resolve_frames, pull out resolver management
2024-05-25 13:02:25 -05:00
Jeremy
dcd9a31ce4
Refactor collate_frames
2024-05-25 12:38:34 -05:00
Jeremy
dad3bd1843
Some comments regarding dlfcn logistics and comment cleanup
2024-05-25 00:40:08 -05:00
Jeremy
7916f10278
Attempt to fallback to dladdr if dladdr1 is not available
2024-05-24 23:04:20 -05:00
Jeremy
f8d28a6469
Fix incorrect handling of lazy_trace_holder's union, I can't believe I wrote such egregious code
2024-05-23 22:33:56 -05:00
Jeremy
6307700710
Fix ASSERT macro under -Wpedantic
2024-05-23 22:33:33 -05:00
Jeremy Rifkin
b3836e9318
Support -Wpedantic if someone sets the flag globally while using FetchContent ( #127 )
...
Closes #107
2024-05-23 21:18:11 -05:00
Jeremy
7c9b24ca72
Fix dwarf_ranges handling of lowpc == pc, it's not correct to jump the gun and not check the high-pc. I think I erroneously added this in c94b03ae6e. Discovered as part of #123 .
2024-05-19 17:00:44 -05:00
Jeremy
4884c4e99b
Fix frame pointer calculation for symbol frames under libunwind, #123
2024-05-19 16:36:07 -05:00
Jeremy
b2fe396d70
Fix address formatting, somehow it was left-aligned
2024-05-19 16:28:41 -05:00
Jeremy
323ab5814e
Merge branch 'main' into dev
2024-05-12 22:00:45 -05:00
Ben Dunkin
a70cf7935a
Fix missing cxxabi.h include on Windows when using libc++ ( #122 )
...
Make operator precedence explicit so the
`CPPTRACE_HAS_CXX_EXCEPTION_TYPE` define is a dominant switch to use
cxxabi.h, rather than one that only applies when `__GLIBCXX__` is
defined.
Fixes https://github.com/jeremy-rifkin/cpptrace/issues/121
2024-05-12 21:56:38 -05:00
Jeremy
74ed6afc0a
Add cpptrace::system_error
2024-05-06 23:01:11 -05:00
Jeremy
da0aa4d5c0
Split a bunch of internal headers into .cpp/.hpp
2024-05-05 17:44:27 -05:00
Jeremy
d7c19a5544
Updates for conda
2024-05-03 20:07:09 -05:00
Jeremy
abbc77dcea
Clean up logic a bit
2024-05-02 11:17:41 -05:00
Jeremy
d3c7e7351c
Remove guard for resolve_l_name
2024-05-02 11:09:37 -05:00
Jeremy
ce19421bb5
Update for macos
2024-05-02 11:05:22 -05:00
Jeremy
b125248b32
Update dladdr path to not use dli_fname as it's unreliable in the case of a non-standard argv[0]
2024-05-02 10:47:31 -05:00
Jeremy
0527cea39f
For the libbacktrace back-end don't treat missing debug info in ELFs as a critical error; fixes #114
2024-04-28 19:51:52 -05:00
Jeremy
c871c2a43c
Grab new microfmt changes and fix #113
2024-04-27 01:29:39 -05:00
Jeremy
599d6abd6c
Workaround bug for old msvc
2024-03-31 23:51:41 -05:00
Jeremy
0dd71cebb7
Mac fixes
2024-03-31 22:01:14 -05:00
Jeremy
06372b772f
Fix another warning
2024-03-31 16:42:51 -05:00
Jeremy
40696e8d02
Handle MSVC warning treated as error
2024-03-31 16:37:29 -05:00
Jeremy
ffb1b8ed68
A little more error handling cleanup
2024-03-31 16:02:39 -05:00
Jeremy
dc9cee98fb
Move error when returning
2024-03-31 15:52:45 -05:00
Jeremy
124607e7bf
Resolve some TODOs with error handling for mach-o stuff
2024-03-31 15:24:01 -05:00
Jeremy
b04fc09682
Update internal error messages
2024-03-31 14:18:21 -05:00
Jeremy
8007413ff6
Print Result(E) if not absorbing trace exceptions
2024-03-31 14:15:19 -05:00
Jeremy
4db73b1856
Fix assertion failure format string
2024-03-30 23:19:13 -05:00
Jeremy
50483783cb
Fix bug with non-recursive inline walking
2024-03-30 23:18:55 -05:00
Jeremy
ab63c34827
Replace most stream use with new formatting system
2024-03-30 20:25:06 -05:00
Jeremy Rifkin
acaa4f42e6
Add a Result type and replace some exceptions with it ( #109 )
2024-03-30 13:35:52 -05:00
Jeremy
47e7ee79eb
Formatting update
2024-03-30 11:11:07 -05:00
Jeremy
fa6155ff47
Remove awful stringf system
2024-03-29 23:12:26 -05:00
Jeremy
411fdff8fa
Merge branch 'dev'
2024-03-20 22:08:33 -05:00
Jeremy
d1199dc325
Fix potential null dereference issue, thanks @eyalgolan1337. Resolves #106 .
2024-03-20 22:08:21 -05:00
Billy O'Neal
37e6cef4f9
Defend against min and max macros from windows.h ( #105 )
...
Alternative to
https://github.com/microsoft/vcpkg/pull/37512/files#diff-9f533b43a5faabaa6b5a0e046f0ae425cd85736808604dd61dc9a955db3d060aR9
I left the examples in mach-o.hpp as they are guarded by #if IS_APPLE
2024-03-18 23:03:26 -05:00
Jeremy
a841a1b74a
Fix msvc warning treated as error
2024-03-16 19:23:03 -05:00
Jeremy
a528aa8e0b
Fix object address resolution with _dl_find_object, #104
2024-03-16 16:17:27 -05:00
Jeremy
3e0689a5e6
Resolve object address for dbghelp frames by default, handles #100
2024-03-09 22:15:41 -06:00
Jeremy
8d6326da07
Update locking in dbghelp_syminit_manager
2024-03-08 17:39:08 -06:00
Jeremy
c896d70f79
Small reorganization
2024-03-05 17:47:55 -06:00
Jeremy
da739d30c5
Add can_signal_safe_unwind and update some documentation surrounding signal-safe stack tracing
2024-03-03 12:15:27 -06:00
Jeremy
ec264aa0eb
Cleanup and improve snippet code a bit
2024-03-03 11:49:15 -06:00
Jeremy
862b546532
Update margin handling
2024-03-03 11:34:09 -06:00
Jeremy
9da2b3f556
Refactor object.hpp, separate out module base logic which greatly cleans things up
2024-03-03 11:31:06 -06:00
Jeremy
f11e119d32
Refactor stacktrace conversion code
2024-03-03 10:54:23 -06:00
Jeremy
389f788b57
Add get_object_info method to stacktrace frames, adding onto the previous work for #97
2024-03-03 10:54:03 -06:00
Jeremy Rifkin
d99f1745d2
Basic source code snippet system ( #98 )
2024-02-28 23:29:24 -06:00
Jeremy
736643358b
Implement a non-recursive get_inlines_info
2024-02-28 22:30:47 -06:00
Jeremy
0fda4a88ed
Refactor stack trace frame printing
2024-02-28 20:53:28 -06:00
Jeremy
a31d35c04e
Fixes
2024-02-27 19:12:55 -06:00
Jeremy
a24c140baf
Add object address resolution for libdl backend
2024-02-27 19:00:08 -06:00
Jeremy
d17fe33abf
Try to fix build errors
2024-02-27 00:39:58 -06:00
Jeremy
3c7a677689
Add an object address to stacktrace_frame entries, addresses #97 (no pun intended)
2024-02-27 00:34:45 -06:00
Jeremy
d5b2646283
Merge branch 'main' into dev
2024-02-27 00:23:10 -06:00
Bruce Mitchener
a144002bf0
Tweaks from clang-tidy ( #92 )
...
This is (perhaps) mainly for discussion as I saw you removed
`clang-tidy` checks last year. These fix a variety of minor things.
With one of the options that I was using, these still remain:
```
/Users/bruce/Development/custodian/cpptrace/src/symbols/../utils/utils.hpp:235:22: warning: noexcept specifier on the move constructor evaluates to 'false' [performance-noexcept-move-constructor]
235 | noexcept(std::is_nothrow_move_constructible<T>::value)
| ^
/Users/bruce/Development/custodian/cpptrace/src/symbols/../utils/utils.hpp:250:64: warning: noexcept specifier on the move assignment operator evaluates to 'false' [performance-noexcept-move-constructor]
250 | noexcept(std::is_nothrow_move_assignable<T>::value && std::is_nothrow_move_constructible<T>::value)
| ^
```
2024-02-18 11:21:48 -06:00
Bruce Mitchener
1488460172
Use defined(CPPTRACE_HAS_CXX_EXCEPTION_TYPE). ( #94 )
...
This is set by the build system when it is available and not set when
not available (and not set to 1/0 respectively).
The current code can generate a warning when this preprocessor
definition is not defined.
2024-02-18 11:16:04 -06:00
Bruce Mitchener
aa446b0540
Fix typos. ( #90 )
2024-02-18 00:10:21 -06:00
Jeremy
d52b80301a
Make sure to at least show object frame info even if resolution fails for libdwarf, resolves #87
...
Co-authored-by: eyalgolan1337
2024-02-15 19:40:11 -06:00
Jeremy
22b326ba7e
Small optimization, doesn't really matter but doesn't hurt
2024-02-11 22:22:29 -06:00
Jeremy
e6627b760d
Implement better logic for handling scrfile indices, related to #86
2024-02-11 21:02:48 -06:00
Jeremy
cb92c9fdfa
Small optimization to reduce unnecessary frame copying
2024-02-11 15:59:56 -06:00
Jeremy
36a16df075
Handle an edge case with dwarf file indices, identified in #86
2024-02-11 15:25:18 -06:00
Jeremy
9a7c45bda7
remove return that wasn't needed
2024-02-11 02:05:00 -06:00
Jeremy
cdaab64be2
Revert "Printbug"
...
This reverts commit 2e9c586ba7
2024-02-11 01:58:29 -06:00
Jeremy
612d5e6d65
Fix architecture handling in load_mach
2024-02-11 01:53:57 -06:00
Jeremy
2e9c586ba7
Printbug
2024-02-11 01:49:46 -06:00
Jeremy
37630f479d
Try to make fat mach architecture handling more robust
2024-02-08 00:38:56 -06:00
Jeremy
59272bccd6
printbug...
2024-02-07 22:37:55 -06:00
Jeremy
68f919f292
Add ctrace documentation
2024-02-07 00:11:48 -06:00
