[SERVER-64439] Investigate building libunwind with --enable-per-thread-cache Created: 11/Mar/22 Updated: 02/Feb/24 |
|
| Status: | Backlog |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | None |
| Fix Version/s: | 6.0 Desired |
| Type: | Improvement | Priority: | Major - P3 |
| Reporter: | Daniel Moody | Assignee: | Unassigned |
| Resolution: | Unresolved | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||
| Assigned Teams: |
Build
|
||||
| Participants: | |||||
| Description |
|
By default, libunwind uses a global cache, which potentially has some performance issues, due to needing to lock the cache frequently. Possibly there is some performance gains with --enable-per-thread-cache. I could not find much discussion about potential perf improvements besides on this PR: https://github.com/libunwind/libunwind/pull/8 I would also note, I saw several past issues around per thread caching in my searches, which I did not investigate deep enough to determine if all issues were real and resolved, but its worth investigating in this ticket. |