Comments on: jemalloc builds http://blog.pavlov.net/2008/01/12/jemalloc-builds/ Ramblings from the mind of Stuart Parmenter Sat, 19 Jul 2008 11:14:54 +0000 http://wordpress.org/?v=MU By: Mark Finkle’s Weblog » XULRunner 1.9 RC3 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-3523 Mark Finkle’s Weblog » XULRunner 1.9 RC3 Fri, 13 Jun 2008 14:40:00 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-3523 [...] build available. In addition to using the same source as Firefox 3 RC3, XULRunner RC3 also enables jemalloc support on Windows, just like [...] [...] build available. In addition to using the same source as Firefox 3 RC3, XULRunner RC3 also enables jemalloc support on Windows, just like [...]

]]> By: cgk http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-3094 cgk Thu, 13 Mar 2008 18:36:03 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-3094 So, simply re#defining malloc/free/realloc and then mapping them to an own no-/low-fragmentation heap implementation (e.g., like http://msdn2.microsoft.com/en-us/library/aa366750(VS.85).aspx ) was not an option? Or just replacing the malloc/free/realloc calls with project-unique names with the same signature. So, simply re#defining malloc/free/realloc and then mapping them to an own no-/low-fragmentation heap implementation (e.g., like http://msdn2.microsoft.com/en-us/library/aa366750(VS.85).aspx ) was not an option?

Or just replacing the malloc/free/realloc calls with project-unique names with the same signature.

]]> By: Firefox 3 Memory Usage « pavlov.net http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2886 Firefox 3 Memory Usage « pavlov.net Wed, 12 Mar 2008 06:19:31 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2886 [...] carefully studied the fragmentation effects of various allocators and concluded that jemalloc gave us the smallest amount of fragmentation after running for a long period of time. I’ve [...] [...] carefully studied the fragmentation effects of various allocators and concluded that jemalloc gave us the smallest amount of fragmentation after running for a long period of time. I’ve [...]

]]> By: Just Browsing » Browsers and Commoditization http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2862 Just Browsing » Browsers and Commoditization Fri, 29 Feb 2008 12:40:33 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2862 [...] a renewed commitment to addressing these criticisms, as does Stuart Parmenter’s work on reducing memory fragmentation. Stories like this should help to win over geek hearts and minds, although end users will withhold [...] [...] a renewed commitment to addressing these criticisms, as does Stuart Parmenter’s work on reducing memory fragmentation. Stories like this should help to win over geek hearts and minds, although end users will withhold [...]

]]> By: jemalloc on trunk — linux edition « pavlov.net http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2854 jemalloc on trunk — linux edition « pavlov.net Thu, 28 Feb 2008 00:46:35 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2854 [...] on trunk — linux edition Back on the 12th of February we turned jemalloc on in our Linux builds.  Sorry for not posting sooner!  We saw a good performance increase and a [...] [...] on trunk — linux edition Back on the 12th of February we turned jemalloc on in our Linux builds.  Sorry for not posting sooner!  We saw a good performance increase and a [...]

]]> By: Pola http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2848 Pola Fri, 15 Feb 2008 16:37:11 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2848 I've tried to compile te mozdev19.dll. After fixing the missing file unhandld.obj; I am still unable to compile it. I got fatal error LNK1181: cannot open input file 'build\intel\dll_obj\asctime.obj' and NMAKE : fatal error U1073: don't know how to make 'build\intel\dll_obj\_setargv.obj' Do you have any sugestion. Thanks in advance. I’ve tried to compile te mozdev19.dll.
After fixing the missing file unhandld.obj; I am still unable to compile it.

I got
fatal error LNK1181: cannot open input file ‘build\intel\dll_obj\asctime.obj’
and
NMAKE : fatal error U1073: don’t know how to make ‘build\intel\dll_obj\_setargv.obj’

Do you have any sugestion.

Thanks in advance.

]]> By: pavlov http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2808 pavlov Tue, 05 Feb 2008 09:54:53 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2808 jimis: I haven't yet. We're just now looking at Linux so I'll take a look here shortly. jimis: I haven’t yet. We’re just now looking at Linux so I’ll take a look here shortly.

]]> By: jemalloc now on the trunk « pavlov.net http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2807 jemalloc now on the trunk « pavlov.net Tue, 05 Feb 2008 09:37:32 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2807 [...] now on the trunk Our Windows nightlies (beta4pre, this is not in beta 3) now include jemalloc.  These builds are leaps and bounds better than the last build I [...] [...] now on the trunk Our Windows nightlies (beta4pre, this is not in beta 3) now include jemalloc.  These builds are leaps and bounds better than the last build I [...]

]]> By: Roger http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2761 Roger Wed, 23 Jan 2008 21:43:38 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2761 It keeps crashing with some MSVC++ error dialog after some time (several hours), even when I'm not using it at the moment (crashed a while ago while I was watching a movie). I'm using my FF3beta profile, so take with a grain of salt. It keeps crashing with some MSVC++ error dialog after some time (several hours), even when I’m not using it at the moment (crashed a while ago while I was watching a movie).
I’m using my FF3beta profile, so take with a grain of salt.

]]> By: jimis http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2643 jimis Sun, 13 Jan 2008 20:21:28 +0000 http://blog.pavlov.net/2008/01/12/jemalloc-builds/#comment-2643 Did you try setting various glibc mallopt() options, as I had proposed in previous comments? Did you try setting various glibc mallopt() options, as I had proposed in previous comments?

]]>