This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PING] [v9] malloc: Consistently apply trim_threshold to all heaps [BZ #17195]
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: Siddhesh Poyarekar <siddhesh at redhat dot com>
- Cc: Mel Gorman <mgorman at suse dot de>, Ond??ej B?lka <neleai at seznam dot cz>, Mike Frysinger <vapier at gentoo dot org>, Julian Taylor <jtaylor dot debian at googlemail dot com>, Chris Metcalf <cmetcalf at ezchip dot com>, libc-alpha at sourceware dot org
- Date: Thu, 07 May 2015 01:08:29 -0400
- Subject: Re: [PING] [v9] malloc: Consistently apply trim_threshold to all heaps [BZ #17195]
- Authentication-results: sourceware.org; auth=none
- References: <55071DB5 dot 5010702 at redhat dot com> <20150324154645 dot GA20062 at suse dot de> <20150401170238 dot GC32115 at domone> <20150401225419 dot GB20397 at suse dot de> <2106100772 dot 11112686 dot 1427957233362 dot JavaMail dot zimbra at redhat dot com> <20150402071721 dot GF20397 at suse dot de> <5542E696 dot 2060402 at redhat dot com> <20150501100546 dot GE2449 at suse dot de> <20150504150958 dot GU15457 at spoyarek dot pnq dot redhat dot com> <5547B879 dot 8030102 at redhat dot com> <20150507050102 dot GC15457 at spoyarek dot pnq dot redhat dot com>
On 05/07/2015 01:01 AM, Siddhesh Poyarekar wrote:
> On Mon, May 04, 2015 at 02:20:41PM -0400, Carlos O'Donell wrote:
>> On 05/04/2015 11:09 AM, Siddhesh Poyarekar wrote:
>>> On Fri, May 01, 2015 at 11:05:46AM +0100, Mel Gorman wrote:
>>>> Also FWIW, my trim patch is still making its way into opensuse. It got held
>>>> up by a bug in another package that was using uninitialsed memory. In the
>>>> future if there is a problem that points the finger at this patch, it's
>>>> worth testing with MALLOC_CHECK_ to see if that "fixes" it by zeroing memory.
>>>
>>> Is this other package java? We had openjdk crashing due to this patch
>>> in rawhide.
>>
>> What openjdk crash? :-)
>
> https://bugzilla.redhat.com/show_bug.cgi?id=1209451
>
> This became very frequent due to this patch. I'm going to leave out
> this patch till they figure out what the problem is.
Or put it back in to force them to figure it out :}
c.