FEDORA-2020-f04e9f1683 has been submitted as an update to Fedora 33. https://bodhi.fedoraproject.org/updates/FEDORA-2020-f04e9f1683

After updating to version 20.2.6.1 gnome-session-check-acelerated-gl-helper is still segfaulting, and the stack backtrace is:
#0  0x0000000000000000 in ?? ()
#1  0x00007f983d5e60bb in cso_destroy_context () from /usr/lib64/dri/r300_dri.so
#2  0x00007f983d167f34 in st_destroy_context_priv () from /usr/lib64/dri/r300_dri.so
#3  0x00007f983d1691d4 in st_destroy_context () from /usr/lib64/dri/r300_dri.so
#4  0x00007f983d14a982 in dri_destroy_context () from /usr/lib64/dri/r300_dri.so
#5  0x00007f983d5e4787 in driDestroyContext () from /usr/lib64/dri/r300_dri.so
#6  0x00007f983e932c33 in dri2_destroy_context () from /lib64/libGLX_mesa.so.0
#7  0x00007f983e921559 in glXDestroyContext () from /lib64/libGLX_mesa.so.0
#8  0x0000557b43c3793e in main ()

The journal entry is:
Dec 17 08:38:14 cain.armitage.org.uk kernel: gnome-session-c[1118]: segfault at 0 ip 0000000000000000 sp 00007ffd1ebd4ca8 error 14 in gnome-session-check-accelerated-gl-helper[557b43c35000+2000]
Dec 17 08:38:14 cain.armitage.org.uk kernel: Code: Unable to access opcode bytes at RIP 0xffffffffffffffd6.


This is followed by gnome-shell segfaulting, and the stack backtrace is:
(gdb) bt
#0  0x0000000000000000 in  ()
#1  0x00007ff956fe00bb in cso_destroy_context () at /usr/lib64/dri/r300_dri.so
#2  0x00007ff956b61f34 in st_destroy_context_priv () at /usr/lib64/dri/r300_dri.so
#3  0x00007ff956b631d4 in st_destroy_context () at /usr/lib64/dri/r300_dri.so
#4  0x00007ff956b44982 in dri_destroy_context () at /usr/lib64/dri/r300_dri.so
#5  0x00007ff956fde787 in driDestroyContext () at /usr/lib64/dri/r300_dri.so
#6  0x00007ff9651fbc33 in dri2_destroy_context () at /lib64/libGLX_mesa.so.0
#7  0x00007ff9651ea559 in glXDestroyContext () at /lib64/libGLX_mesa.so.0
#8  0x00007ff97c531aa2 in _cogl_winsys_display_destroy.lto_priv.0 () at /usr/lib64/mutter-7/libmutter-cogl-7.so.0
#9  0x00007ff97c500d8d in _cogl_object_display_indirect_free () at /usr/lib64/mutter-7/libmutter-cogl-7.so.0
#10 0x00007ff97c503875 in cogl_renderer_check_onscreen_template () at /usr/lib64/mutter-7/libmutter-cogl-7.so.0
#11 0x00007ff97cceaf24 in clutter_backend_x11_get_display () at /usr/lib64/mutter-7/libmutter-clutter-7.so.0
#12 0x00007ff97cc8e31e in clutter_backend_real_create_context () at /usr/lib64/mutter-7/libmutter-clutter-7.so.0
#13 0x00007ff97ccb6327 in clutter_init_real () at /usr/lib64/mutter-7/libmutter-clutter-7.so.0
#14 0x00007ff97ccb66c8 in post_parse_hook () at /usr/lib64/mutter-7/libmutter-clutter-7.so.0
#15 0x00007ff97d6a41c7 in g_option_context_parse () at /lib64/libglib-2.0.so.0
#16 0x00007ff97ccb69bd in clutter_init () at /usr/lib64/mutter-7/libmutter-clutter-7.so.0
#17 0x00007ff97ca8dec3 in meta_backend_initable_init () at /lib64/libmutter-7.so.0
#18 0x00007ff97cae504c in meta_init () at /lib64/libmutter-7.so.0
#19 0x000055987a34f9dc in main ()
(gdb) 

and then another segfault in gnome-shell with the same stack backtrace (except for the addresses)

Prior to the gnome-shell segfaults are the following entries in the journal (twice for each segfault):
Dec 17 08:38:20 cain.armitage.org.uk /usr/libexec/gdm-x-session[1068]: (II) RADEON(0): EDID vendor "SEC", prod id 0
Dec 17 08:38:20 cain.armitage.org.uk /usr/libexec/gdm-x-session[1068]: (II) RADEON(0): Printing DDC gathered Modelines:
Dec 17 08:38:20 cain.armitage.org.uk /usr/libexec/gdm-x-session[1068]: (II) RADEON(0): Modeline "1280x800"x0.0   71.11  1280 1328 1360 1440  800 802 808 823 -hsync -vsync (49.4 kHz eP)

I suspect it could be a fallout from https://gitlab.freedesktop.org/mesa/mesa/-/commit/bd32ac29bbbe26e035c76a0d85c664be4c3ec0e4 that added an assert to cso_destroy_context(). Any chance you could file this upstream and say that it's a regression in 20.2?

I have built the mesa packages v20.2.6 with commit bd32ac29 reverted, and the segfault still occurs.

I have reported this upstream at https://gitlab.freedesktop.org/mesa/mesa/-/issues/3990

With v20.2.6 if I revert bd32ac29 as before and also revert 5f9912a4 and d6fd7acf it then works successfully.

After upgrading to v20.3.3 the segfault starts occurring again. Applying the following patch stops the segfault occurring, although I don't expect that the patch is the right way to resolve the problem. I have posted these details at https://gitlab.freedesktop.org/mesa/mesa/-/issues/3990.

diff --git a/src/gallium/auxiliary/cso_cache/cso_context.c b/src/gallium/auxiliary/cso_cache/cso_context.c
index 1eef6aac70c..4dca8b41d38 100644
--- a/src/gallium/auxiliary/cso_cache/cso_context.c
+++ b/src/gallium/auxiliary/cso_cache/cso_context.c
@@ -412,7 +412,7 @@ void cso_destroy_context( struct cso_context *ctx )
             if (maxview > 0) {
                ctx->pipe->set_sampler_views(ctx->pipe, sh, 0, maxview, views);
             }
-            if (maxssbo > 0) {
+            if (maxssbo > 0 && ctx->pipe-set_shader_buffers) {
                ctx->pipe->set_shader_buffers(ctx->pipe, sh, 0, maxssbo, ssbos, 0);
             }
             for (int i = 0; i < maxcb; i++) {

Upstream commit https://gitlab.freedesktop.org/mesa/mesa/-/commit/58e43594fc457eaaf1b1e01e48948959a82080bc resolves this issue.

Would it be possible to backport the patch to Fedora 32 and 33?

This message is a reminder that Fedora 33 is nearing its end of life.
Fedora will stop maintaining and issuing updates for Fedora 33 on 2021-11-30.
It is Fedora's policy to close all bug reports from releases that are no longer
maintained. At that time this bug will be closed as EOL if it remains open with a
Fedora 'version' of '33'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version.

Thank you for reporting this issue and we are sorry that we were not 
able to fix it before Fedora 33 is end of life. If you would still like 
to see this bug fixed and are able to reproduce it against a later version 
of Fedora, you are encouraged  change the 'version' to a later Fedora 
version prior this bug is closed as described in the policy above.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events. Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

Fedora 33 changed to end-of-life (EOL) status on 2021-11-30. Fedora 33 is
no longer maintained, which means that it will not receive any further
security or bug fix updates. As a result we are closing this bug.

If you can reproduce this bug against a currently maintained version of
Fedora please feel free to reopen this bug against that version. If you
are unable to reopen this bug, please file a new report against the
current release. If you experience problems, please add a comment to this
bug.

Thank you for reporting this bug and we are sorry it could not be fixed.