sched_4bsd: remove dumping from maybe_preempt()
ClosedPublic
Actions

Authored by mchoo on Fri, Sep 4, 5:36 PM.

Details

Reviewers

jhb
olce

Group Reviewers

scheduler

Commits

rG1dfefffcbdc6: sched_4bsd: remove dumping from maybe_preempt()

Summary

'dumping' is true only when kernel is dumping after crash (see
minidumpsys()) so KERNEL_PANICKED() will catch this.

MFC after: 2 weeks
MFC to: stable/14, stable/15
Sponsored by: FreeBSD Foundation

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

No Test Coverage

Build Status

Buildable 76609
Build 73492: arc lint + arc unit

Event Timeline

mchoo created this revision.Fri, Sep 4, 5:36 PM

Herald added subscribers: olce, imp. · View Herald TranscriptFri, Sep 4, 5:36 PM

mchoo requested review of this revision.Fri, Sep 4, 5:36 PM

Harbormaster completed remote builds in B76503: Diff 185831.Fri, Sep 4, 5:36 PM

mchoo added a child revision: D59396: sched_4bsd: add static assertion for nice weight.Fri, Sep 4, 5:36 PM

Mmm, that's not true since we have been having live (mini)dump. On the contrary, it is perhaps time to put back dumping here and in ULE.

This revision now requires changes to proceed.Sun, Sep 6, 3:53 PM

mchoo added a parent revision: D59394: sched_4bsd: remove kern.sched.4bsd.followon.Sun, Sep 6, 6:45 PM

In D59395#1363961, @olce wrote:

Mmm, that's not true since we have been having live (mini)dump. On the contrary, it is perhaps time to put back dumping here and in ULE.

Makes sense, I'll open a new revision for that.

Actually, dumping is not set to true during live dump, so I think the patch is still correct.

int
minidumpsys(struct dumperinfo *di, bool livedump)
{
	struct minidumpstate state;
	struct msgbuf mb_copy;
	char *msg_ptr;
	int error;

	if (livedump) {
		KASSERT(!dumping, ("live dump invoked from incorrect context"));

This revision now requires changes to proceed.Sun, Sep 6, 8:39 PM

Update commit message

Harbormaster completed remote builds in B76609: Diff 186035.Sun, Sep 6, 8:48 PM

Remove duplicated commit message template

Harbormaster completed remote builds in B76610: Diff 186036.Sun, Sep 6, 8:50 PM

mchoo edited the summary of this revision. (Show Details)Sun, Sep 6, 8:50 PM

On second thought, there is a balance between getting a consistent dump image and temporary disruption of the live system for live dumps. It's already the case that they do not guarantee a fully consistent image. And they don't try to stop CPUs, the scheduler, etc. As you point out, minidump is not set, which in this line of thinking is not surprising.

So let's proceed with that.

This revision is now accepted and ready to land.Sun, Sep 6, 9:18 PM

Closed by commit rG1dfefffcbdc6: sched_4bsd: remove dumping from maybe_preempt() (authored by mchoo). · Explain WhyFri, Sep 18, 2:15 PM

This revision was automatically updated to reflect the committed changes.

mchoo added a commit: rG1dfefffcbdc6: sched_4bsd: remove dumping from maybe_preempt().