<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Feb 5, 2015 at 8:00 AM, James Cameron <span dir="ltr"><<a href="mailto:quozl@laptop.org" target="_blank">quozl@laptop.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Thanks.<br>
<br>
Can I make it happen more often?<br>
<br>
Is there a later version of the driver?<br>
<br>
We have a different version that I may look into, on arm-3.5-android<br>
branch.<br>
<span class=""><br></span></blockquote><div><br></div><div>run memtester against the majority of your machines memory and then run gtkperf in an X session. That is usually enough to trigger it.</div><div><br></div><div>Considering that bug exists in all the 4.xx vivante galcore drivers I have seen I doubt it is fixed in the other version. Android is much simpler on memory because it runs everything through a single GL context against a framebuffer.</div><div><br></div><div>I have some tentative patches to fix parts of it in my trees but I doubt a lot of them would apply to 3.5 without backporting a lot of upstream work.</div><div><br></div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">
On Wed, Feb 04, 2015 at 12:14:02PM +0100, Jon Nettleton wrote:<br>
> It is a problem with the v4 version of the galcore driver. We have replicated<br>
> it on a couple of platforms.<br>
><br>
</span>> On Wed, Feb 4, 2015 at 11:26 AM, Peter Robinson <[1]<a href="mailto:pbrobinson@gmail.com">pbrobinson@gmail.com</a>><br>
> wrote:<br>
<span class="">><br>
> On Wed, Feb 4, 2015 at 8:10 AM, James Cameron <[2]<a href="mailto:quozl@laptop.org">quozl@laptop.org</a>> wrote:<br>
> > Following up a thread from last September.<br>
> ><br>
> > This problem has just become more interesting, because it hit during<br>
> > an activity startup.<br>
> ><br>
> > I'm quite used to seeing it with yum. But seeing it without yum now<br>
> > points us at kernel, glibc or python.<br>
><br>
> We've not seen this in the wider F-20 Fedora ARM distro so my bet<br>
> would be on the kernel.<br>
><br>
> Peter<br>
><br>
</span>> > [3]<a href="http://dev.laptop.org/ticket/12837#comment:4" target="_blank">http://dev.laptop.org/ticket/12837#comment:4</a> has the details of the<br>
<span class="">> > most recent event.<br>
> ><br>
> > On Wed, Sep 10, 2014 at 01:56:27PM +1000, James Cameron wrote:<br>
> >> G'day Peter,<br>
> >><br>
> >> Thanks for any ideas you may have.<br>
> >><br>
> >> The problem also reproduces on OLPC Fedora 20 image for XO-4:<br>
> >><br>
</span>> >> [4]<a href="http://build.laptop.org/14.1.0/os1/xo-4/41001o4.zd" target="_blank">http://build.laptop.org/14.1.0/os1/xo-4/41001o4.zd</a> (552 MB)<br>
<span class="">> >><br>
> >> *** Error in `/usr/bin/python': free(): invalid pointer: 0x047c79ae ***<br>
> >> ======= Backtrace: =========<br>
> >> /lib/libc.so.6(+0x6c8b4)[0xb6c828b4]<br>
> >> /lib/libc.so.6(+0x754e8)[0xb6c8b4e8]<br>
> >> ======= Memory map: ========<br>
> >> [...]<br>
> >><br>
> >> The error varies in detail, but always suggests corruption of heap or<br>
> >> pointers to heap.<br>
> >><br>
> >> The triggering conditions are interactive use of yum, yum update, or<br>
> >> yum used by olpc-os-builder. The latter is a simple reproducer for me.<br>
> >><br>
> >> I'm reproducing it on an XO-4, with 2GB of RAM, no swap, 8 GB eMMC, 8<br>
> >> GB USB flash drive.<br>
> >><br>
> >> While memory demand by yum is large by comparison to other programs,<br>
> >> the available memory at the time of failure is ample. There are no<br>
> >> kernel out of memory (OOM) events. It seems more likely to occur when<br>
> >> the filesystem cache is under heavy demand.<br>
> >><br>
> >> The method to recreate the problem was:<br>
> >><br>
> >> 1. install the system image 41001o4.zd using fs-update and then boot,<br>
> >><br>
> >> 2. configure wireless network,<br>
> >><br>
> >> 3. "yum install -y git olpc-os-builder"<br>
> >><br>
> >> 4. clone the master branch of<br>
</span>> >> git://[5]<a href="http://dev.laptop.org/projects/olpc-os-builder" target="_blank">dev.laptop.org/projects/olpc-os-builder</a><br>
<span class="">> >> (last verified with b87e6ee)<br>
> >><br>
> >> 5. run "./osbuilder.py examples/olpc-os-14.1.0-xo4.ini" repeatedly<br>
> >> until the error occurs (usually within about five attempts),<br>
> >><br>
> >><br>
> >> I've also tried running under valgrind, but that causes illegal<br>
> >> instruction. It is quite likely I'm not using valgrind correctly.<br>
</span>> >> [6]<a href="http://dev.laptop.org/~quozl/z/1XRYtO.txt" target="_blank">http://dev.laptop.org/~quozl/z/1XRYtO.txt</a><br>
<span class="">> >><br>
> >> The workaround at the moment is to build our Fedora 20 images on<br>
> >> Fedora 18. Fedora 18 shows no sign of the problem. I'm worried that<br>
> >> a low probability heap corruptor may cause instability of applications<br>
> >> in the field.<br>
> >><br>
> >> The exact same kernel is being used for Fedora 18 and Fedora 20.<br>
> >><br>
> >> On Tue, Sep 09, 2014 at 03:55:24PM +0100, Peter Robinson wrote:<br>
> >> > What version of OOB are you using, and what config files? I can try<br>
> >> > and recreate the problem here on other devices.<br>
> >><br>
> >> --<br>
> >> James Cameron<br>
</span>> >> [7]<a href="http://quozl.linux.org.au/" target="_blank">http://quozl.linux.org.au/</a><br>
> ><br>
> > --<br>
> > James Cameron<br>
> > [8]<a href="http://quozl.linux.org.au/" target="_blank">http://quozl.linux.org.au/</a><br>
> _______________________________________________<br>
> Devel mailing list<br>
> [9]<a href="mailto:Devel@lists.laptop.org">Devel@lists.laptop.org</a><br>
> [10]<a href="http://lists.laptop.org/listinfo/devel" target="_blank">http://lists.laptop.org/listinfo/devel</a><br>
><br>
> References:<br>
><br>
> [1] mailto:<a href="mailto:pbrobinson@gmail.com">pbrobinson@gmail.com</a><br>
> [2] mailto:<a href="mailto:quozl@laptop.org">quozl@laptop.org</a><br>
> [3] <a href="http://dev.laptop.org/ticket/12837#comment:4" target="_blank">http://dev.laptop.org/ticket/12837#comment:4</a><br>
> [4] <a href="http://build.laptop.org/14.1.0/os1/xo-4/41001o4.zd" target="_blank">http://build.laptop.org/14.1.0/os1/xo-4/41001o4.zd</a><br>
> [5] <a href="http://dev.laptop.org/projects/olpc-os-builder" target="_blank">http://dev.laptop.org/projects/olpc-os-builder</a><br>
> [6] <a href="http://dev.laptop.org/~quozl/z/1XRYtO.txt" target="_blank">http://dev.laptop.org/~quozl/z/1XRYtO.txt</a><br>
> [7] <a href="http://quozl.linux.org.au/" target="_blank">http://quozl.linux.org.au/</a><br>
> [8] <a href="http://quozl.linux.org.au/" target="_blank">http://quozl.linux.org.au/</a><br>
> [9] mailto:<a href="mailto:Devel@lists.laptop.org">Devel@lists.laptop.org</a><br>
> [10] <a href="http://lists.laptop.org/listinfo/devel" target="_blank">http://lists.laptop.org/listinfo/devel</a><br>
<div class="HOEnZb"><div class="h5"><br>
--<br>
James Cameron<br>
<a href="http://quozl.linux.org.au/" target="_blank">http://quozl.linux.org.au/</a><br>
</div></div></blockquote></div><br></div></div>