On Mon, Jun 19, 2017 at 7:39 AM, Yuya Nishihara <y...@tcha.org> wrote:
> On Sun, 18 Jun 2017 00:02:26 -0700, Martin von Zweigbergk via Mercurial-devel
> wrote:
>> # HG changeset patch
>> # User Martin von Zweigbergk <martinv...@google.com>
>> # Date 1497593354 25200
>> # Thu Jun 15 23:09:14 2017 -0700
>> # Node ID 1a23893eb7f7ea55df8835bd6208f8427fa36d13
>> # Parent b6b0928d0128d2b424b40bf4c4d2a59acc20f9ac
>> repair: create transaction for bundle1 unbundling earlier
>>
>> See earlier patch for motivation.
>>
>> diff --git a/mercurial/repair.py b/mercurial/repair.py
>> --- a/mercurial/repair.py
>> +++ b/mercurial/repair.py
>> @@ -202,13 +202,15 @@
>> if not repo.ui.verbose:
>> # silence internal shuffling chatter
>> repo.ui.pushbuffer()
>> + tmpbundleurl = 'bundle:' + vfs.join(tmpbundlefile)
>> if isinstance(gen, bundle2.unbundle20):
>> with repo.transaction('strip') as tr:
>> bundle2.applybundle(repo, gen, tr, source='strip',
>> - url='bundle:' +
>> vfs.join(tmpbundlefile))
>> + url=tmpbundleurl)
>> else:
>> - gen.apply(repo, 'strip', 'bundle:' +
>> vfs.join(tmpbundlefile),
>> - True)
>> + txnname = "strip\n%s" % util.hidepassword(tmpbundleurl)
>> + with repo.lock(), repo.transaction(txnname):
>
> I'm not sure if this repo.lock() is intentional or a copy-paste error.
I don't think that was intentional. Good catch. Will you fix in flight
it want me to send a v2?
>
> FWIW, repair.strip() says "does not take any lock on the repo", but we already
> have one for bookmark processing.
Hmm, good point about that comment. I wonder if that's even correct.
If it is, why don't we see the 'transaction requires locking'
exception from localrepo.transaction()? I'll look into it. I suspect
it will simply result in a patch that removes the comment.
>
> Other than that, the series looks good to me.
_______________________________________________
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel
