[PATCH] verify: clarify misleading fncache message

Gregory Szorc gregory.szorc at gmail.com
Sat Jun 20 22:22:21 CDT 2015


On Fri, Jun 19, 2015 at 2:55 PM, Matt Mackall <mpm at selenic.com> wrote:

> On Fri, 2015-06-19 at 19:58 +0200, Adrian Buehlmann wrote:
> > On 2015-06-19 19:00, Matt Mackall wrote:
> > > # HG changeset patch
> > > # User Matt Mackall <mpm at selenic.com>
> > > # Date 1434733206 18000
> > > #      Fri Jun 19 12:00:06 2015 -0500
> > > # Node ID c1ad1589a44b519411b6f209c1e81322f4ccfd4e
> > > # Parent  85294076adceb3263056e366023cba2e88ace6ca
> > > verify: clarify misleading fncache message
>
> > I agree that "missing revlog" for the case where the fncache file lacks
> > entries, is less than ideal.
>
> Ok, I've queued this, thanks.
>
> > Demoting such a case to a warning is probably justified.
> >
> > Remark: I think that if the fncache file in a repo A lacks entries,
> > using A as the source in a 'hg clone --uncompressed' (for example) will
> > then produce a repo B which indeed has missing revlogs. 'hg verify' on B
> > will then fail.
>
> Yep. We should probably have a debug command which can rebuild this file
> from changelog metadata and add a hint to verify to use it.
>

I just patchbombed a series to do this.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://selenic.com/pipermail/mercurial-devel/attachments/20150620/dba4b0e6/attachment.html>


More information about the Mercurial-devel mailing list