clock_gettime(CLOCK_MONOTONIC, &ts) returns unexpected timestamp
Nobody has claimed this yet.
- Dominant language
- C++
- Stars
- 10.7k
- Forks
- 662
- Avg merge
- 2d 3h
- Merged PRs (30d)
- 2
Description
During recording, I observed that the system monotonic clock may return strange values. In particular, the timestamp ts returned by clock_gettime(CLOCK_MONOTONIC, &ts) may be 0. This is rather unexpected and software relying on monotonic behavior might behave weird.
When replaying the trace, I see 2 threads calling clock_gettime simultaneously. Thread T1 gets interrupted during its call to clock_gettime. T2 is scheduled and returns a sensible value. T1 resumes and returns a timestamp of 0.
I'm not familiar with the rr code base, maybe the above information is sufficient to identify the problem. I build rr from source on commit 97997c263ad82c76f5705fb6b2f0e45de417b085.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the provided recording and replay scenario from commit 97997c263ad82c76f5705fb6b2f0e45de417b085, focusing on simultaneous clock_gettime(CLOCK_MONOTONIC, &ts) calls and the interrupted T1 thread. Trace the relevant rr handling until the zero timestamp is explained; done means the replay no longer returns 0 and the behavior is covered by a regression test.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp, linux
- Domain
- devtools, operating-systems
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100