[SOLVED] Firefox 89.0.1 segmentation fault on slarm64?
Slackware - ARMThis forum is for the discussion of Slackware ARM.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
After upgrading to mozilla-firefox-89.0.1-aarch64-1, it no longer launches and instead spits out a "Segmentation fault". Is anyone else getting this? It is happening on all of my slarm64 machines that have firefox installed. I thought I would just check in to see if anyone else is experiencing this before I try to reset the user profile, etc.
After upgrading to mozilla-firefox-89.0.1-aarch64-1, it no longer launches and instead spits out a "Segmentation fault". Is anyone else getting this? It is happening on all of my slarm64 machines that have firefox installed. I thought I would just check in to see if anyone else is experiencing this before I try to reset the user profile, etc.
thanks much.
did not observe such problems and communicating with another user after the update also works, without resetting the profile.
Ok, that is interesting to know, thank you. I have tried creating a new user, using a different DE, refreshing the profile, creating a new profile, etc. This is consistently happening on a Raspberry Pi 4 (legacy kernel) and a Pinebook (legacy kernel) no matter what the environment is like.
This is the error message:
Code:
bash-5.1$ firefox
Crash Annotation GraphicsCriticalError: |[0][GFX1-]: No GPUs detected via PCI (t=7.29735) [GFX1-]: No GPUs detected via PCI
###!!! [Parent][MessageChannel] Error: (msgtype=0x39015D,name=PContent::Msg_CommitWindowContextTransaction) Channel error: cannot send/recv
###!!! [Parent][MessageChannel] Error: (msgtype=0x39015D,name=PContent::Msg_CommitWindowContextTransaction) Channel error: cannot send/recv
Exiting due to channel error.
Segmentation fault
Exiting due to channel error.
Exiting due to channel error.
bash-5.1$
I have tried to eliminate any obvious variable such as GTK theme, desktop environment, add-ons, browser themes, etc. I will try a fresh image on an SD card to see if it happens there. So far it is pretty consistent. Like I said, even a fresh profile with a brand new user gives the same output.
The strace suggestion almost worked. It spits out the same errors above plus a lot of other stuff that is not an error, just sys calls. But it gave me the idea to try:
Code:
firefox --safe-mode
which actually works fine without crashing and gives me access to my bookmarks and saved passwords. No other option I tried prevented the browser from crashing (even --headless). Only safe-mode does the trick, so far. I will report back if I notice anything interesting on a fresh SD card image.
Thanks for the feedback in any case, it helps me troubleshoot.
Here is my corresponding output. Note that mesa recently updated, and the problem was happening before that update, and is still happening.
Code:
bash-5.1# glxinfo | grep -i mesa
GLX_INTEL_swap_event, GLX_MESA_copy_sub_buffer, GLX_OML_swap_method,
client glx vendor string: Mesa Project and SGI
GLX_MESA_copy_sub_buffer, GLX_MESA_multithread_makecurrent,
GLX_MESA_query_renderer, GLX_MESA_swap_control, GLX_OML_swap_method,
GLX_MESA_copy_sub_buffer, GLX_MESA_query_renderer, GLX_MESA_swap_control,
Extended renderer info (GLX_MESA_query_renderer):
OpenGL version string: 2.1 Mesa 21.1.3
GL_KHR_texture_compression_astc_sliced_3d, GL_MESA_pack_invert,
GL_MESA_tile_raster_order, GL_MESA_window_pos, GL_NV_blend_square,
OpenGL ES profile version string: OpenGL ES 2.0 Mesa 21.1.3
GL_KHR_texture_compression_astc_sliced_3d, GL_MESA_tile_raster_order,
I was able to narrow it down to the last Firefox update. I have an SD image created on 2021-06-17 (with an enlightenment DE, just for reference) that has Firefox 89.0 installed, as well as mesa 21.1.2. Before updating anything on that image, Firefox works fine. This is on a Pinebook (although I can reproduce it on a Raspberry Pi 4). I then run a slackpkg update (after grabbing the GPG-KEY from a repository using wget, this was not pre-installed in the image for some reason) and allow only Firefox to update to 89.0.1 (skipping everything else, including mesa). Now Firefox segfaults like it does on my other systems. Then I went ahead and updated everything else, including mesa to 21.1.3, but that does not help the Firefox issue, sadly.
Historically, from what I can find around the web, this seems to crop up from time to time over the years on different platforms and seems to be processor-specific, even in the x86 world, from what I gather from the forums. The symptoms are the random crash shortly after launching with a segmentation fault, while working in safe-mode avoids the problem somehow, despite not having any add-ons installed in the other mode. Eventually an upgrade to the next version makes the problem go away. I am guessing/hoping that this is a similar instance. It might just go away with the next update.
Since I only use Firefox for a few sites (banking and such, sites that complain if you are not using Firefox or Chrome specifically), and I use SeaMonkey and/or text browsers for everything else. the safe-mode workaround actually does not bother me that much for the time being.
The strange part is that when I launch Firefox 89.0 from the command line on the fresh unaltered enlightenment image, I get the same errors I listed above, minus the "segmentation fault". Everything else that 89.0.1 complains about is also present in the output of 89.0, with the one exception of "segmentation fault". So all the other error messages are probably completely unrelated to this issue.
UPDATE: I was able to grab a full strace output this time (capture was not working for me the other day, not sure why, might have been tired...). The raw text is somewhat large. I uploaded a gzipped version of it as well for quicker downloading.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.