Date: Tue, 11 Jan 2011 05:58:45 +0000 From: Alexey Dokuchaev <danfe@FreeBSD.org> To: wen heping <wenheping@gmail.com> Cc: Sahil Tandon <sahil@freebsd.org>, Wen Heping <wen@freebsd.org>, cvs-all@freebsd.org, ports-committers@freebsd.org, Boris Samorodov <bsam@ipt.ru>, cvs-ports@freebsd.org Subject: Re: cvs commit: ports/sysutils/tmux Makefile ports/sysutils/tmux/files extra-patch-configure Message-ID: <20110111055845.GA19115@FreeBSD.org> In-Reply-To: <AANLkTi=VDj5PTYy6bJCk8Yb%2BN6hYYjFj%2BCZHM5L_DGXC@mail.gmail.com> References: <201101100135.p0A1ZLEL028834@repoman.freebsd.org> <20110110014602.GB92966@FreeBSD.org> <20110110041847.GB5992@magic.hamla.org> <20110110065210.GA24349@FreeBSD.org> <95244161@h30.sp.ipt.ru> <AANLkTi=VDj5PTYy6bJCk8Yb%2BN6hYYjFj%2BCZHM5L_DGXC@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jan 11, 2011 at 08:03:40AM +0800, wen heping wrote: > 2011/1/11 Boris Samorodov <bsam@ipt.ru>: > > On Mon, 10 Jan 2011 06:52:10 +0000 Alexey Dokuchaev wrote: > >> On Sun, Jan 09, 2011 at 11:18:48PM -0500, Sahil Tandon wrote: > >> > > >> > http://lists.freebsd.org/pipermail/freebsd-ports/2010-March/060003.html > > > >> Kudos to gj@ for his investigation; I hope clueful guys will eventually > >> solve the problem properly. But commit message could be improved > >> considerably if it included that text from ports@ at least partially. > >> This would save the list from my stupid questions, and interested folks > >> might decide to help Gary and Andriy fixing this issue. > > > > Well, after that commit the port is broken: freezes in minutes: > > [...] > > I agree commit message should include more message of the run error to > help others resolve the problem. But: some days ago, garga@ send me > a mail with a scipts written by him to confirm me there would be run error > if tmux build with kqueue on some special platform. I think it is not easy > to include this message in serveral lines in commit log. I haven't seen the message sent to you by garga@, but for example for the above analysis given by gj@ I think approprite shorter message could be: Disable kqueue support for now; apparently, devel/libevent thinks our kqueue(2) implementation is broken: it assumes that kevent() will always return 1 on success. According to kqueue(2) kevent will return up to nevents on success. Analyzed by: gj@ When in doubt, there is community to ask. :-) ./danfe
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110111055845.GA19115>