Page MenuHomeFreeBSD

uma: fixup some ktr messages
ClosedPublic

Authored by rlibby on Jan 12 2020, 8:04 PM.
Tags
None
Referenced Files
F171444061: D23148.id66714.diff
Fri, Sep 11, 4:06 AM
Unknown Object (File)
Sun, Sep 6, 6:10 AM
Unknown Object (File)
Sat, Sep 5, 11:46 AM
Unknown Object (File)
Sat, Sep 5, 9:28 AM
Unknown Object (File)
Sat, Sep 5, 5:41 AM
Unknown Object (File)
Sat, Sep 5, 2:44 AM
Unknown Object (File)
Thu, Sep 3, 6:10 PM
Unknown Object (File)
Thu, Sep 3, 8:46 AM
Subscribers

Details

Summary

Fixup some ktr format issues noticed in a recent debugging session.

Test Plan

make buildkernel

Diff Detail

Lint
Lint Passed
Unit
No Test Coverage
Build Status
Buildable 28625
Build 26657: arc lint + arc unit

Event Timeline

markj added inline comments.
sys/vm/uma_core.c
3526

I think curthread is somewhat redundant since KTR always logs thread IDs.

... ah, I see it is in other KTR messages, so we should at least be consistent.

This revision is now accepted and ready to land.Jan 12 2020, 8:20 PM
sys/vm/uma_core.c
3526

Yes. Although, it looks like there are only 4 others. We could fix those not to print thread instead? How thread is printed in uma_core.c is kind of weird anyway ("%x").

Separately it looks like ktr_tracepoint/ktr_verbose doesn't print the thread. I don't know if we care to change that.

sys/vm/uma_core.c
3526

I'd be fine with dropping the thread from the KTR messages. I think they used to be plain printf()s and were converted mechanically, which would explain why they're there.

Hmm, I don't think I've ever used ktr_verbose before. I agree that it should print the thread pointer as well.

markj feedback, don't print threads

This revision now requires review to proceed.Jan 13 2020, 12:34 AM

missed one, and messed one up

This revision is now accepted and ready to land.Jan 13 2020, 5:11 AM
This revision was automatically updated to reflect the committed changes.