Date: Mon, 2 Jan 2012 14:36:02 +0100 From: =?ISO-8859-1?Q?Ren=E9_Ladan?= <rene@freebsd.org> To: Luigi Rizzo <luigi@onelab2.iet.unipi.it> Cc: chromium@freebsd.org Subject: Re: ports/161737: chromium 14.0.835.x stalls (probably javascript-related Message-ID: <CADL2u4hmm9jYzfNCHFVV=%2BNU7DjUQ6no1xeLrv4z5ZGymuP7fA@mail.gmail.com> In-Reply-To: <20111214230716.GA60789@onelab2.iet.unipi.it> References: <20111017114140.B38037300A@onelab2.iet.unipi.it> <4EE91F46.40801@freebsd.org> <20111214230716.GA60789@onelab2.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
2011/12/15 Luigi Rizzo <luigi@onelab2.iet.unipi.it>: > On Wed, Dec 14, 2011 at 11:12:22PM +0100, Rene Ladan wrote: >> On 17-10-2011 13:41, Luigi Rizzo wrote: >> >> Number: =A0 =A0 =A0 =A0 161737 >> >> Category: =A0 =A0 =A0 ports >> >> Synopsis: =A0 =A0 =A0 chromium 14.0.835.x stalls (probably javascript= -related >> > >> > On both of my RELENG_8 machines, since i updated >> > from chromium 13.0.782.112 to 14.0.835.* I am experiencing >> > frequent stalls on most pages -- from gmail to google docs to >> > online forms for accessing various services, after a >> > short time the tab containing the page become insensitive to >> > keyboard and mouse input events for large amount of times >> > (from 30s to minutes) before processing them. >> > >> > This does not affect just input: on some other sites >> > i am seeing mouse clicks being only partially >> > processed, i.e. the click causes part of the content to >> > be updated, but another part remain stale. >> > I am seeing this in particular on JS-based pages (99.99% these days). >> > >> > Unfortunately this makes chromium unusable for services where >> > i cannot tolerate crashing the tab in the middle of a transaction >> > (bank, shops, reservations etc.) >> > >> > Haven't done a deep investigation, but it worked fine with >> > 13.0.782.112 and broke when i updated to 14.0.835.X (both versions). >> > The commit logs show some changes >> > (tcmalloc, v8) that might be related to what i am seeing. >> > >> > >> >> How-To-Repeat: >> > =A0 =A0 open a spreadsheet in google docs, play with it for a while, >> > =A0 =A0 until you'll see that keypresses and clicks are not >> > =A0 =A0 processed anymore. >> >> Can you try with chromium 16.0.912.63 ? > > i think this PR can be closed. > Things seem to have improved with > chromium-15.0.874.121 and i forgot about the PR. > Now building 16.0.X > Can you confirm that it works with chromium 16 and that this PR can be clos= ed? Regards, Ren=E9
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CADL2u4hmm9jYzfNCHFVV=%2BNU7DjUQ6no1xeLrv4z5ZGymuP7fA>