So IIUC the only race that could really happen here is being called with KD_TEXT n times in quick succession so we end up calling vd_postswitch and resuming the flush timer n times?

This looks okay, nice catch

Agreed. But it got me looking at the other places we touch vw_flags and I confirmed some races with KCSAN. I’ll put that up in a separate review.

I confirmed some races with KCSAN. I’ll put that up in a separate review.

Great, thanks!

Closed by commit rG82dcd3ab12f0: vt: Move VWF_GRAPHICS clearing under VT_LOCK. · Explain WhySep 20 2026, 12:20 PM

This revision was automatically updated to reflect the committed changes.

emaste added a commit: rG82dcd3ab12f0: vt: Move VWF_GRAPHICS clearing under VT_LOCK.

Herald added a subscriber: imp. · View Herald TranscriptSep 20 2026, 12:20 PM

Revision Contents
Changeset List

Path

Size

sys/

dev/

vt/

vt_core.c

2 lines

Diff 187273

View Options

vt: Clear VWF_GRAPHICS under VT_LOCKClosedPublicActions

Details

Diff Detail

Event Timeline

Revision ContentsChangeset List

Diff 187273

sys/dev/vt/vt_core.c

vt: Clear VWF_GRAPHICS under VT_LOCK
ClosedPublic
Actions

Revision Contents
Changeset List