You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@trafficserver.apache.org by "Adam W. Dace" <co...@gmail.com> on 2013/12/13 07:02:48 UTC
Quick Thoughts On TCMalloc
I know James was interested in this, so I finally got tired of awaiting the
next release of ATS and thought I'd give it a try.
On CentOS Linux VM: Up and running, definitely seems faster...but on a
parent cache that's really a guess.
On Mac OS X: Unfortunately, the farthest I can get with it ends in a
segfault. It looks like this:
CXX CompileParseRules.o
CXXLD CompileParseRules
./CompileParseRules
make[3]: *** [ParseRulesCType] Segmentation fault: 11
make[2]: *** [all] Error 2
make[1]: *** [all-recursive] Error 1
make: *** [all-recursive] Error 1
I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
supported enough?
Thanks,
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by "Adam W. Dace" <co...@gmail.com>.
Sorry, that was confusing. Dang you, GMail.
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
Unfortunately, I have no idea how to "turn on" coredumps in Mac OS X.
I'll dig around and see if I can figure it out.
On Sun, Dec 15, 2013 at 5:26 AM, Adam W. Dace <co...@gmail.com>wrote:
> Unfortunately, I have no idea how to "turn on" coredumps in Mac OS X.
>
> I'll dig around and see if I can figure it out.
>
>
>
> On Fri, Dec 13, 2013 at 8:38 AM, Igor Galić <i....@brainsware.org>wrote:
>
>>
>> ------------------------------
>>
>> I know James was interested in this, so I finally got tired of awaiting
>> the next release of ATS and thought I'd give it a try.
>>
>> On CentOS Linux VM: Up and running, definitely seems faster...but on a
>> parent cache that's really a guess.
>>
>> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
>> segfault. It looks like this:
>>
>> CXX CompileParseRules.o
>> CXXLD CompileParseRules
>> ./CompileParseRules
>> make[3]: *** [ParseRulesCType] Segmentation fault: 11
>> make[2]: *** [all] Error 2
>> make[1]: *** [all-recursive] Error 1
>> make: *** [all-recursive] Error 1
>>
>> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
>> supported enough?
>>
>>
>> That's usually related to flex/bison in the wccp build..
>> But I'd check with Dr Allen, he knows best.
>> If you get a coredump, then a traceback would be greatly appreciated
>>
>>
>>
>> Thanks,
>>
>> --
>> ____________________________________________________________
>> Adam W. Dace <co...@gmail.com>
>>
>> Phone: (815) 355-5848
>> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
>> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>>
>> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>>
>>
>>
>>
>> --
>> Igor Galić
>>
>> Tel: +43 (0) 664 886 22 883
>> Mail: i.galic@brainsware.org
>> URL: http://brainsware.org/
>> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>>
>>
>
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by "Adam W. Dace" <co...@gmail.com>.
Unfortunately, I have no idea how to "turn on" coredumps in Mac OS X.
I'll dig around and see if I can figure it out.
On Fri, Dec 13, 2013 at 8:38 AM, Igor Galić <i....@brainsware.org> wrote:
>
> ------------------------------
>
> I know James was interested in this, so I finally got tired of awaiting
> the next release of ATS and thought I'd give it a try.
>
> On CentOS Linux VM: Up and running, definitely seems faster...but on a
> parent cache that's really a guess.
>
> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
> segfault. It looks like this:
>
> CXX CompileParseRules.o
> CXXLD CompileParseRules
> ./CompileParseRules
> make[3]: *** [ParseRulesCType] Segmentation fault: 11
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all-recursive] Error 1
>
> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
> supported enough?
>
>
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
>
>
>
> Thanks,
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
>
>
>
> --
> Igor Galić
>
> Tel: +43 (0) 664 886 22 883
> Mail: i.galic@brainsware.org
> URL: http://brainsware.org/
> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
>
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by "Adam W. Dace" <co...@gmail.com>.
I'd happy to maintain that Wiki page, maybe simplify it a bit(i.e. just a
list of applicable OS's and versions)...
but I just clicked through to the README file. It doesn't mention Solaris
either. ):
On Sun, Dec 15, 2013 at 2:41 AM, Igor Galić <i....@brainsware.org> wrote:
> ------------------------------
>
> Currently, Apache Traffic Server only supports FreeBSD, Linux, and Mac OS
> X.
>
> See:
> https://cwiki.apache.org/confluence/display/TS/Supported+Operating+Systems
>
> lies!
> We also support Solaris.
> That wiki needs to be either updated, or deleted, in favour of the overly
> verbose README
>
> https://git-wip-us.apache.org/repos/asf?p=trafficserver.git;a=blob;f=README;hb=HEAD
>
>
>
> On Fri, Dec 13, 2013 at 10:53 AM, James Johnson <ja...@gmail.com> wrote:
>
>> Is there a Windows implementation? I have someone interested but the
>> organization is primarily Windows based.
>>
>> Jim
>>
>> ------------------------------
>>
>> I know James was interested in this, so I finally got tired of awaiting
>> the next release of ATS and thought I'd give it a try.
>>
>> On CentOS Linux VM: Up and running, definitely seems faster...but on a
>> parent cache that's really a guess.
>>
>> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
>> segfault. It looks like this:
>>
>> CXX CompileParseRules.o
>> CXXLD CompileParseRules
>> ./CompileParseRules
>> make[3]: *** [ParseRulesCType] Segmentation fault: 11
>> make[2]: *** [all] Error 2
>> make[1]: *** [all-recursive] Error 1
>> make: *** [all-recursive] Error 1
>>
>> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
>> supported enough?
>>
>>
>> That's usually related to flex/bison in the wccp build..
>> But I'd check with Dr Allen, he knows best.
>> If you get a coredump, then a traceback would be greatly appreciated
>>
>>
>>
>> Thanks,
>>
>> --
>> ____________________________________________________________
>> Adam W. Dace <co...@gmail.com>
>>
>> Phone: (815) 355-5848
>> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
>> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>>
>> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>>
>>
>>
>>
>> --
>> Igor Galić
>>
>> Tel: +43 (0) 664 886 22 883
>> Mail: i.galic@brainsware.org
>> URL: http://brainsware.org/
>> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>>
>>
>
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
>
>
>
> --
> Igor Galić
>
> Tel: +43 (0) 664 886 22 883
> Mail: i.galic@brainsware.org
> URL: http://brainsware.org/
> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
>
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by Igor Galić <i....@brainsware.org>.
----- Original Message -----
> Currently, Apache Traffic Server only supports FreeBSD, Linux, and Mac OS X.
> See:
> https://cwiki.apache.org/confluence/display/TS/Supported+Operating+Systems
lies!
We also support Solaris.
That wiki needs to be either updated, or deleted, in favour of the overly verbose README
https://git-wip-us.apache.org/repos/asf?p=trafficserver.git;a=blob;f=README;hb=HEAD
> On Fri, Dec 13, 2013 at 10:53 AM, James Johnson < jajohga@gmail.com > wrote:
> > Is there a Windows implementation? I have someone interested but the
> > organization is primarily Windows based.
>
> > Jim
>
> > > I know James was interested in this, so I finally got tired of awaiting
> > > the
> > > next release of ATS and thought I'd give it a try.
> >
>
> > > On CentOS Linux VM: Up and running, definitely seems faster...but on a
> > > parent
> > > cache that's really a guess.
> >
>
> > > On Mac OS X: Unfortunately, the farthest I can get with it ends in a
> > > segfault. It looks like this:
> >
>
> > > CXX CompileParseRules.o
> >
>
> > > CXXLD CompileParseRules
> >
>
> > > ./CompileParseRules
> >
>
> > > make[3]: *** [ParseRulesCType] Segmentation fault: 11
> >
>
> > > make[2]: *** [all] Error 2
> >
>
> > > make[1]: *** [all-recursive] Error 1
> >
>
> > > make: *** [all-recursive] Error 1
> >
>
> > > I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
> > > supported enough?
> >
>
> > That's usually related to flex/bison in the wccp build..
>
> > But I'd check with Dr Allen, he knows best.
>
> > If you get a coredump, then a traceback would be greatly appreciated
>
> > > Thanks,
> >
>
> > > --
> >
>
> > > ____________________________________________________________
> >
>
> > > Adam W. Dace < colonelforbin74@gmail.com >
> >
>
> > > Phone: (815) 355-5848
> >
>
> > > Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> >
>
> > > Microsoft Messenger - colonelforbin74@live.com
> >
>
> > > Google Profile: https://plus.google.com/u/0/109309036874332290399/about
> >
>
> > --
>
> > Igor Galić
>
> > Tel: +43 (0) 664 886 22 883
>
> > Mail: i.galic@brainsware.org
>
> > URL: http://brainsware.org/
>
> > GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
> --
> ____________________________________________________________
> Adam W. Dace < colonelforbin74@gmail.com >
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
--
Igor Galić
Tel: +43 (0) 664 886 22 883
Mail: i.galic@brainsware.org
URL: http://brainsware.org/
GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
Re: Quick Thoughts On TCMalloc
Posted by "Adam W. Dace" <co...@gmail.com>.
Currently, Apache Traffic Server only supports FreeBSD, Linux, and Mac OS X.
See:
https://cwiki.apache.org/confluence/display/TS/Supported+Operating+Systems
On Fri, Dec 13, 2013 at 10:53 AM, James Johnson <ja...@gmail.com> wrote:
> Is there a Windows implementation? I have someone interested but the
> organization is primarily Windows based.
>
> Jim
>
> ------------------------------
>
> I know James was interested in this, so I finally got tired of awaiting
> the next release of ATS and thought I'd give it a try.
>
> On CentOS Linux VM: Up and running, definitely seems faster...but on a
> parent cache that's really a guess.
>
> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
> segfault. It looks like this:
>
> CXX CompileParseRules.o
> CXXLD CompileParseRules
> ./CompileParseRules
> make[3]: *** [ParseRulesCType] Segmentation fault: 11
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all-recursive] Error 1
>
> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
> supported enough?
>
>
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
>
>
>
> Thanks,
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
>
>
>
> --
> Igor Galić
>
> Tel: +43 (0) 664 886 22 883
> Mail: i.galic@brainsware.org
> URL: http://brainsware.org/
> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
>
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by James Johnson <ja...@gmail.com>.
Is there a Windows implementation? I have someone interested but the
organization is primarily Windows based.
Jim
------------------------------
I know James was interested in this, so I finally got tired of awaiting the
next release of ATS and thought I'd give it a try.
On CentOS Linux VM: Up and running, definitely seems faster...but on a
parent cache that's really a guess.
On Mac OS X: Unfortunately, the farthest I can get with it ends in a
segfault. It looks like this:
CXX CompileParseRules.o
CXXLD CompileParseRules
./CompileParseRules
make[3]: *** [ParseRulesCType] Segmentation fault: 11
make[2]: *** [all] Error 2
make[1]: *** [all-recursive] Error 1
make: *** [all-recursive] Error 1
I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
supported enough?
That's usually related to flex/bison in the wccp build..
But I'd check with Dr Allen, he knows best.
If you get a coredump, then a traceback would be greatly appreciated
Thanks,
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
--
Igor Galić
Tel: +43 (0) 664 886 22 883
Mail: i.galic@brainsware.org
URL: http://brainsware.org/
GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
Re: Quick Thoughts On TCMalloc
Posted by Yongming Zhao <mi...@gmail.com>.
we, the alibaba team, meet the memory issue since the later Y2011. all those memory problem can be one of:
1, memory leak
2, httpsm hang
3, ram cache used memory out of control
and most of the memory leak problem is introduced by our self, so we can argue that most of the memory issues here is other httpsm hang, or ram cache, and that finally get into two directions:
1, ram: we use the standard lru instead of the default CLFUS, which will cause more memory as it will hold more memory during swap in small files and swap out the big one.
2, TS-1006, replaim freelist: to make memory in the free list reclaimable by others or return to the system.
after many bug fix for the httpsm hang problems. we finally get the memory usage under control, and performance is even better due to the reclaim free list is mostly a in-thread allocator. I think we have now have a strong memory management system in the core, as all of our effort already in the tree. if it does not, there is bug somewhere.
I’d advise you try the reclaim free list for the core memory issues. but in plugins, there is no way to benefit from the freelist in the core, if you make many alloc/free in the plugin, tcmalloc may help.
a little out of topic, just FYI
thanks
Yongming Zhao
赵永明
aka 永豪 yonghao@taobao.com
在 2013年12月16日，下午9:35，Adam W. Dace <co...@gmail.com> 写道：
> > That's usually related to flex/bison in the wccp build..
> > But I'd check with Dr Allen, he knows best.
> > If you get a coredump, then a traceback would be greatly appreciated
>
> Stupid question...how do I limit the output of "bt" from within gdb?
>
> <adace@mound:/cores> gdb -c core.18423
> ...
> #0 0x000000010b700eec in tcmalloc::ThreadCache::MinSizeForSlowPath () at /Users/adace/Src/gperftools-2.1/src/thread_cache.h:1094
> 1094 LIKELY(size < ThreadCache::MinSizeForSlowPath())) {
> (gdb)
>
> Apologies, I've never really used gdb much. ):
>
>
>
> On Fri, Dec 13, 2013 at 8:38 AM, Igor Galić <i....@brainsware.org> wrote:
>
> I know James was interested in this, so I finally got tired of awaiting the next release of ATS and thought I'd give it a try.
>
> On CentOS Linux VM: Up and running, definitely seems faster...but on a parent cache that's really a guess.
>
> On Mac OS X: Unfortunately, the farthest I can get with it ends in a segfault. It looks like this:
>
> CXX CompileParseRules.o
> CXXLD CompileParseRules
> ./CompileParseRules
> make[3]: *** [ParseRulesCType] Segmentation fault: 11
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all-recursive] Error 1
>
> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc supported enough?
>
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
>
>
>
> Thanks,
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
>
>
> --
> Igor Galić
>
> Tel: +43 (0) 664 886 22 883
> Mail: i.galic@brainsware.org
> URL: http://brainsware.org/
> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
>
>
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by "Adam W. Dace" <co...@gmail.com>.
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
Stupid question...how do I limit the output of "bt" from within gdb?
<adace@mound:/cores> gdb -c core.18423
...
#0 0x000000010b700eec in tcmalloc::ThreadCache::MinSizeForSlowPath () at
/Users/adace/Src/gperftools-2.1/src/thread_cache.h:1094
1094 LIKELY(size < ThreadCache::MinSizeForSlowPath())) {
(gdb)
Apologies, I've never really used gdb much. ):
On Fri, Dec 13, 2013 at 8:38 AM, Igor Galić <i....@brainsware.org> wrote:
>
> ------------------------------
>
> I know James was interested in this, so I finally got tired of awaiting
> the next release of ATS and thought I'd give it a try.
>
> On CentOS Linux VM: Up and running, definitely seems faster...but on a
> parent cache that's really a guess.
>
> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
> segfault. It looks like this:
>
> CXX CompileParseRules.o
> CXXLD CompileParseRules
> ./CompileParseRules
> make[3]: *** [ParseRulesCType] Segmentation fault: 11
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all-recursive] Error 1
>
> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
> supported enough?
>
>
> That's usually related to flex/bison in the wccp build..
> But I'd check with Dr Allen, he knows best.
> If you get a coredump, then a traceback would be greatly appreciated
>
>
>
> Thanks,
>
> --
> ____________________________________________________________
> Adam W. Dace <co...@gmail.com>
>
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
>
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
>
>
>
>
> --
> Igor Galić
>
> Tel: +43 (0) 664 886 22 883
> Mail: i.galic@brainsware.org
> URL: http://brainsware.org/
> GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641
>
>
--
____________________________________________________________
Adam W. Dace <co...@gmail.com>
Phone: (815) 355-5848
Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
Microsoft Messenger - colonelforbin74@live.com <ad...@turing.com>
Google Profile: https://plus.google.com/u/0/109309036874332290399/about
Re: Quick Thoughts On TCMalloc
Posted by Igor Galić <i....@brainsware.org>.
----- Original Message -----
> I know James was interested in this, so I finally got tired of awaiting the
> next release of ATS and thought I'd give it a try.
> On CentOS Linux VM: Up and running, definitely seems faster...but on a parent
> cache that's really a guess.
> On Mac OS X: Unfortunately, the farthest I can get with it ends in a
> segfault. It looks like this:
> CXX CompileParseRules.o
> CXXLD CompileParseRules
> ./CompileParseRules
> make[3]: *** [ParseRulesCType] Segmentation fault: 11
> make[2]: *** [all] Error 2
> make[1]: *** [all-recursive] Error 1
> make: *** [all-recursive] Error 1
> I'm not sure if I should go ahead and file a Jira on this...is tcmalloc
> supported enough?
That's usually related to flex/bison in the wccp build..
But I'd check with Dr Allen, he knows best.
If you get a coredump, then a traceback would be greatly appreciated
> Thanks,
> --
> ____________________________________________________________
> Adam W. Dace < colonelforbin74@gmail.com >
> Phone: (815) 355-5848
> Instant Messenger: AIM & Yahoo! IM - colonelforbin74 | ICQ - #39374451
> Microsoft Messenger - colonelforbin74@live.com
> Google Profile: https://plus.google.com/u/0/109309036874332290399/about
--
Igor Galić
Tel: +43 (0) 664 886 22 883
Mail: i.galic@brainsware.org
URL: http://brainsware.org/
GPG: 8716 7A9F 989B ABD5 100F 4008 F266 55D6 2998 1641