-
Notifications
You must be signed in to change notification settings - Fork 546
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Provide control over which replayed core is which output drmemtrace stream #6726
Labels
Comments
derekbruening
added a commit
that referenced
this issue
Mar 26, 2024
Currently, in as-traced mode the output streams are assigned to the cores in the as-traced schedule file in file order. But that order is essentially random, which scrambles key arrangements like which core is on which socket. Here we sort by the recorded cpuid to recreate the same cpuid order as before. Adds a unit test. Also tested on larger cases with > 100 cores. Issue: #6726
derekbruening
added a commit
that referenced
this issue
Mar 27, 2024
Currently, in as-traced mode the output streams are assigned to the cores in the as-traced schedule file in file order. But that order is essentially random, which scrambles key arrangements like which core is on which socket. Here we sort by the recorded cpuid to recreate the same cpuid order as before. Adds a unit test. Also tested on larger cases with > 100 cores. Issue: #6726
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
In the drmemtrace scheduler, some sort of full control over which replayed core should be which output stream would be good, as the comment in read_traced_schedule() says:
As a first step, we should at least sort the cpuids, as raw2trace inserts them in whatever hashtable order they were in, which is essentially random. E.g., on 2 sockets, the order for one trace has the 2 sockets completely intermixed: 4185,70,4207,4126,4129,69,... (the top bit 4096 marks the 2nd socket). This means the replay will not match the original order.
The text was updated successfully, but these errors were encountered: