[crashtracking] Little improvements to the FFI api #842
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
95c725e to
a21b4c8
Compare
BenchmarksComparisonBenchmark execution time: 2025-02-06 08:53:32 Comparing candidate commit 1987c10 in PR branch Found 0 performance improvements and 0 performance regressions! Performance is the same for 52 metrics, 2 unstable metrics. CandidateCandidate benchmark detailsGroup 1
Group 2
Group 3
Group 4
Group 5
Group 6
Group 7
Group 8
Group 9
Group 10
Group 11
Group 12
Group 13
BaselineOmitted due to size. |
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #842 +/- ##
==========================================
- Coverage 71.48% 71.45% -0.03%
==========================================
Files 320 319 -1
Lines 47115 47124 +9
==========================================
- Hits 33680 33674 -6
- Misses 13435 13450 +15
|
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
a21b4c8 to
6d17ec9
Compare
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
11 times, most recently
from
8e1814b to
202ffdc
Compare
gleocadie
changed the title
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
06e271c to
92216f0
Compare
gleocadie
commented
Jan 31, 2025
| ) -> VoidResult { | ||
| wrap_with_void_ffi_result!({ | ||
| if thread.crashed { | ||
| let stack = (thread.stack.to_inner_mut())?.clone(); | ||
| builder.to_inner_mut()?.with_stack(stack)?; |
There was a problem hiding this comment.
I wonder if ddog_crasht_CrashInfoBuilder_with_stack shouldn't be removed ? 🤔
There was a problem hiding this comment.
If there is only one executing thread, does it make sense to have it twice?
There was a problem hiding this comment.
But you're the main user of this API, so do what makes sense to you!
danielsn
reviewed
Jan 31, 2025
| ) -> VoidResult { | ||
| wrap_with_void_ffi_result!({ | ||
| if thread.crashed { | ||
| let stack = (thread.stack.to_inner_mut())?.clone(); | ||
| builder.to_inner_mut()?.with_stack(stack)?; |
There was a problem hiding this comment.
If there is only one executing thread, does it make sense to have it twice?
| ) -> VoidResult { | ||
| wrap_with_void_ffi_result!({ | ||
| if thread.crashed { | ||
| let stack = (thread.stack.to_inner_mut())?.clone(); | ||
| builder.to_inner_mut()?.with_stack(stack)?; |
There was a problem hiding this comment.
But you're the main user of this API, so do what makes sense to you!
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
6 times, most recently
from
e604600 to
82cd9e1
Compare
danielsn
reviewed
Feb 1, 2025
|
|
||
| //////////////////////////////////////////////////////////////////////////////////////////////////// | ||
| // FFI API // | ||
| //////////////////////////////////////////////////////////////////////////////////////////////////// | ||
|
|
||
| #[allow(dead_code)] | ||
| #[repr(C)] | ||
| pub enum StackTraceNewResult { |
There was a problem hiding this comment.
you can use ddcommon_ffi::Result e.g. ddcommon_ffi::Result<Handle<StackTrace>>
There was a problem hiding this comment.
If I use ddcommon_ffi::Result<Handle<StackFrame>> instead of StackFrameNewResult
The enum item name is
DDOG_CRASHT_STACK_FRAME_NEW_RESULT_OK_HANDLE_STACK_FRAME (with renaming)
or
DDOG_CRASHT_RESULT_HANDLE_FRAME_OK_HANDLE_STACK_FRAME
If I use the new enum StackFrameNewResult, I got this
DDOG_CRASHT_STACK_FRAME_NEW_RESULT_OK which is shorter
There was a problem hiding this comment.
If you know a way to prevent a new enum to have shorter name, I'll be happy :)
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
82cd9e1 to
5d0dbee
Compare
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
3 times, most recently
from
a2a6b6f to
e14a873
Compare
danielsn
reviewed
Feb 5, 2025
ddcommon-ffi/src/string.rs
Outdated
|
|
||
| impl ToHexStr for usize { | ||
| fn to_hex_str(&self) -> String { | ||
| format!("0x{:X}", self) |
There was a problem hiding this comment.
Do we want this to be fixed length? i.e. should 2 be 0x2 or 0x0000000000000002?
There was a problem hiding this comment.
🤷
the specification says that it's a hexadecimal string prefixed with 0x but nothing about the length.
There was a problem hiding this comment.
Agreed that either is standards compliant. Its just aesthetically, which do we prefer?
danielsn
approved these changes
Feb 5, 2025
paullegranddc
approved these changes
Feb 5, 2025
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
6678fa3 to
cbbb00a
Compare
gleocadie
force-pushed
the
gleocadie/crashtracker-new-format-improvements
branch
from
cbbb00a to
1987c10
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
This PR tries to improve a bit the crashtracker ffi APIs.
Motivation
The goal is to simplify the crashtracker ffi API and adding code in the
crashinfo.cppexample.Implementation
usize(expecting it to be 64bit on x86_64, and 32bit on x86 architecture) and build a0x-prefixed string from it.ddog_crasht_CrashInfoBuilder_with_stackadds the callstack of the crashing threads but only the callstack.If we want to add the crashing thread (
ThreadData), we have to duplicate the callstack and make one calls toddog_crasht_CrashInfoBuilder_with_stackandddog_crasht_CrashInfoBuilder_with_thread.Instead, we check the
crashedfield, and call intowith_stackto add a duplicate of the callstack thread (if it's true) and call intowith_threadwith the thread dataMaybe we could remove the
ddog_crasht_CrashInfoBuilder_with_stack?XXXNewResulttypes to have shorter symbols for the generated code.Additional Notes
Anything else we should know when reviewing?
How to test the change?
Describe here in detail how the change can be validated.