storage/spanlatch: add test for overlapping spans #40260
Conversation
|
This change is |
There was a problem hiding this comment.
Reviewed 2 of 2 files at r1.
Reviewable status:complete! 0 of 0 LGTMs obtained (waiting on @irfansharif)
pkg/storage/spanlatch/manager_test.go, line 169 at r1 (raw file):
latchCs = append(latchCs, m.MustAcquireCh(spans("a", "b", write), zeroTS)) latchCs = append(latchCs, m.MustAcquireCh(spans("b", "c", read), zeroTS)) latchCs = append(latchCs, m.MustAcquireCh(spans("b", "c", write), zeroTS))
Right now this case and the last case are getting blocked on their previous cases, not the original latches, right? Can we avoid that by putting the reads at lower timestamps than the writes?
When overlapping spans are acquired, the spanlatch Manager should grant access based on the more restrictive latch over the specific overlapped sub-span. Release note: None
irfansharif
force-pushed
the
spanlatch-test
branch
from
9bacbe8
to
b98fef3
Compare
There was a problem hiding this comment.
Reviewable status:
pkg/storage/spanlatch/manager_test.go, line 169 at r1 (raw file):
Previously, nvanbenschoten (Nathan VanBenschoten) wrote…
Right now this case and the last case are getting blocked on their previous cases, not the original latches, right? Can we avoid that by putting the reads at lower timestamps than the writes?
oh woops, and yep that does it.
There was a problem hiding this comment.
Reviewable status:
|
bors r+ |
40260: storage/spanlatch: add test for overlapping spans r=irfansharif a=irfansharif When overlapping spans are acquired, the spanlatch Manager should grant access based on the more restrictive latch over the specific overlapped sub-span. Release note: None Co-authored-by: irfan sharif <irfanmahmoudsharif@gmail.com>
Build succeeded |
When overlapping spans are acquired, the spanlatch Manager should
grant access based on the more restrictive latch over the specific
overlapped sub-span.
Release note: None