1

I use gnus (by way of gwene.org) to read RSS feeds. However, for the first time I've ever seen, Emacs crashed when trying to open a feed, giving the following dump at the terminal:

Fatal error 11: Segmentation fault Backtrace: emacs[0x5032b3] emacs[0x4e99be] emacs[0x50206e] emacs[0x502293] /usr/lib/libpthread.so.0(+0x10d60)[0x7f55c97b7d60] /usr/lib/libOpenCL.so(+0x1a7c6a)[0x7f55b5be8c6a] /usr/lib/libOpenCL.so(clGetDeviceInfo+0x8f0)[0x7f55b5bc4150] /usr/lib/libMagickCore-6.Q16HDRI.so.2(+0x14c4a9)[0x7f55cb9fa4a9] /usr/lib/libMagickCore-6.Q16HDRI.so.2(InitOpenCLEnv+0x7c)[0x7f55cb9fb3ec] /usr/lib/libMagickCore-6.Q16HDRI.so.2(+0x445c0)[0x7f55cb8f25c0] /usr/lib/libMagickCore-6.Q16HDRI.so.2(AccelerateCompositeImage+0x79)[0x7f55cb901839] /usr/lib/libMagickCore-6.Q16HDRI.so.2(CompositeImageChannel+0x2aa)[0x7f55cb94060a] /usr/lib/libMagickCore-6.Q16HDRI.so.2(MergeImageLayers+0x1e8)[0x7f55cb9ddbc8] /usr/lib/libMagickWand-6.Q16HDRI.so.2(MagickMergeImageLayers+0x4a)[0x7f55cbe286ca] emacs[0x5cd29c] emacs[0x5d0b11] emacs[0x5d0f4e] emacs[0x55d18c] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] emacs[0x55cfab] emacs[0x5927b3] emacs[0x55cc0f] ... Segmentation fault (core dumped)

What could possibly be the case here?

Koz Ross
  • 425
  • 3
  • 13
  • This backtrace is close to useless. If you can reproduce the issue reliably, recompile Emacs to come with debug symbols and no optimizations enabled, then show us the backtrace again. – wasamasa Oct 11 '15 at 08:22
  • [This answer](http://emacs.stackexchange.com/a/14376/11) describes how to get more information for a crash like this. – legoscia Oct 11 '15 at 09:14

1 Answers1

1

I was getting similar backtraces when trying to open any buffer with inline images.

Following wasamasa's comment I ran emacs under GDB. While I am not able to fix the bug right now, I found it is related to libpng12 and my openCL runtime (intel). Removing libpng12 solved the issue for me.

gabrielhidasy
  • 11
  • 1