Date: Fri, 11 Jul 2014 19:54:42 +0200 From: Baptiste Daroussin <bapt@FreeBSD.org> To: John Baldwin <jhb@freebsd.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: Phabric IDs / URLs in commits Message-ID: <20140711175442.GJ93051@ivaldir.etoilebsd.net> In-Reply-To: <201407111238.23391.jhb@freebsd.org> References: <201407111616.s6BGGQFW060195@svn.freebsd.org> <201407111238.23391.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--ELVYuRnMxQ5nnKRy Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Fri, Jul 11, 2014 at 12:38:23PM -0400, John Baldwin wrote: > On Friday, July 11, 2014 12:16:26 pm John Baldwin wrote: > > Author: jhb > > Date: Fri Jul 11 16:16:26 2014 > > New Revision: 268531 > > URL: http://svnweb.freebsd.org/changeset/base/268531 > >=20 > > Log: > > Fix some edge cases with rewinddir(): > > - In the unionfs case, opendir() and fdopendir() read the directory's= full > > contents and cache it. This cache is not refreshed when rewinddir(= ) is > > called, so rewinddir() will not notice updates to a directory. Fix= this > > by splitting the code to fetch a directory's contents out of > > __opendir_common() into a new _filldir() function and call this from > > rewinddir() when operating on a unionfs directory. > > - If rewinddir() is called on a directory opened with fdopendir() bef= ore > > any directory entries are fetched, rewinddir() will not adjust the = seek > > location of the backing file descriptor. If the file descriptor pa= ssed > > to fdopendir() had a non-zero offset, the rewinddir() will not rewi= nd to > > the beginning. Fix this by always seeking back to 0 in rewinddir(). > > This means the dd_rewind hack can also be removed. > > =20 > > While here, add missing locking to rewinddir(). > > =20 > > CR: https://phabric.freebsd.org/D312 > > Reviewed by: jilles > > MFC after: 1 week >=20 > Just picking my own commit here as a sample case. >=20 > I think we should be annotating commits with phabricator code reviews in = some=20 > way when a change has gone through that review. It is very useful to get= back > to the review details from the commit log message in svnweb, etc. >=20 > I can see a number of different ways to do this, but I do think it would = be > nice to pick a consistent way to do it. >=20 > Things to consider: >=20 > 1) The tag ("CR:" is what I used above). I don't care, just pick one. I > chose CR since Warner used it previously. Whatever we decide, we shou= ld > add it to the template. >=20 > 2) ID vs full URL. For PRs we just list the bug ID and not the full URL > (same for Coverity). I would be fine with that so long as someone hac= ks > up svnweb to convert the IDs into links (the way it handles PR bug > numbers). OTOH, if you use the full URL you get that for free in svnw= eb, > and you also get it in mail clients, etc. It helps that the URL isn't= but > so long. for bugs we could use http://bugs.FreeBSD.org/<number>; that also works and = it is short :) >=20 > This is more of a pie-in-the-sky, but it would be _really_ nice if arcani= st=20 > were hacked up to support our local commit template and would auto popula= te=20 > the 'Reviewed by' and 'CR' (or whatever it ends up being called) fields s= o one=20 > could use 'arc commit'. I'm planning to work on this but I first need to finish tracking 2 bugs it = has with svn. regards, Bapt --ELVYuRnMxQ5nnKRy Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iEYEARECAAYFAlPAJOIACgkQ8kTtMUmk6EzmRQCfTYu5FgBQq4grYkus3Aq0wxBD v5MAn2GhQC7eOfSzOVhEs16bgK+ryBA9 =GqKd -----END PGP SIGNATURE----- --ELVYuRnMxQ5nnKRy--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20140711175442.GJ93051>