Fix virtual to binary addr conversion for processes that have an uncommon virtual memory mapping #1637
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Outdated suggestions cannot be applied.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.
Summary: Fix virtual to binary addr conversion for processes that have an uncommon virtual memory mapping
Our previous virtual to binary address conversion logic assumed that the first offset within
/proc/$PID/maps
was the correct offset to apply for PIE binaries. There are certain cases, such as when an unlimited stack ulimit is applied, where this assumption doesn't hold true (see the linked issue before for more details). This change adjusts our conversion logic to take into account the correct/proc/$PID/maps
entry so address conversion works in all known cases.Relevant Issues: #1630
Type of change: /kind bug
Test Plan: Verified the following:
perf_profiler_bpf_test
passes when the perf profiler uses the ELF symbolizer