[PATCH 7 of 7 STABLE] bookmarks: actual fix for race condition deleting bookmark

Yuya Nishihara yuya at tcha.org
Sun Jun 23 19:18:19 EDT 2019


On Sun, 23 Jun 2019 18:42:16 -0400, Matt Harbison wrote:
> On Thu, 20 Jun 2019 22:04:14 -0400, Pierre-Yves David  
> <pierre-yves.david at ens-lyon.org> wrote:
> 
> > # HG changeset patch
> > # User Pierre-Yves David <pierre-yves.david at octobus.net>
> > # Date 1561081840 -7200
> > #      Fri Jun 21 03:50:40 2019 +0200
> > # Branch stable
> > # Node ID 8611e1183db406463655af0b196cba6fcd2cf8a9
> > # Parent  ddd62edfa406e3b84483cafcaa93927ca917d0cd
> > # EXP-Topic book-del-stable
> > # Available At https://bitbucket.org/octobus/mercurial-devel/
> > #              hg pull https://bitbucket.org/octobus/mercurial-devel/ -r  
> > 8611e1183db4
> > bookmarks: actual fix for race condition deleting bookmark
> 
> I have no idea why, but this is the commit that broke the Windows tests.
> 
> https://buildbot.mercurial-scm.org/builders/Win7%20x86_64%20hg%20tests/builds/1480

Maybe backout this patch? I suspect that we still rely on in-memory changelog
data to be flushed out of the transaction.

Windows has no filecache, which behavior can be tested with this change:

diff --git a/mercurial/scmutil.py b/mercurial/scmutil.py
--- a/mercurial/scmutil.py
+++ b/mercurial/scmutil.py
@@ -1356,6 +1356,7 @@ class filecachesubentry(object):
             self.cachestat = filecachesubentry.stat(self.path)
 
     def cacheable(self):
+        return False
         if self._cacheable is not None:
             return self._cacheable
 


More information about the Mercurial-devel mailing list