diff options
| author | Tom Lane <tgl@sss.pgh.pa.us> | 2019-09-10 12:45:32 -0400 |
|---|---|---|
| committer | Tom Lane <tgl@sss.pgh.pa.us> | 2019-09-10 12:45:32 -0400 |
| commit | b5784042ae7a8c15dc44a7d0112b575dfcc101dc (patch) | |
| tree | 39760d5c306bb26506cbf0244c60c7178ff27f66 /src/backend/access/gist/gistvalidate.c | |
| parent | 8459f908c6166e0249c310a03a3d1265ff7b7972 (diff) | |
| download | postgresql-b5784042ae7a8c15dc44a7d0112b575dfcc101dc.tar.gz postgresql-b5784042ae7a8c15dc44a7d0112b575dfcc101dc.zip | |
Sync isolationtester's handling of notice/warning messages with HEAD.
Back-patch relevant parts of these commits:
30717637c Fix isolationtester race condition for notices sent before blocking
ebd499282 Don't drop NOTICE messages in isolation tests
a28e10e82 Indicate session name in isolationtester notices
This ensures that older versions of the isolationtester will handle
NOTICE/WARNING messages the same way as HEAD and v12 do. While this
isn't fixing any critical problem right now, it seems like a prudent
change to prevent surprises (like we had yesterday...) with
back-patches of future isolation test changes.
Back-patch as far as 9.6. Due to the significant changes we made in
isolationtester in 9.6, back-patching isolation tests further than
that is going to be risky anyway; besides, this patch doesn't apply
cleanly before that.
Discussion: https://postgr.es/m/E1i7IqC-0000Uc-5H@gemulon.postgresql.org
Diffstat (limited to 'src/backend/access/gist/gistvalidate.c')
0 files changed, 0 insertions, 0 deletions