[Mesa-dev] [Bug 99591] Segmentation fault when running vulkaninfo with RADV Radeon Vulkan driver

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=99591

--- Comment #8 from Vasilis LIaskovitis <vliaskov at gmail.com> ---
I am getting a very similar backtrace in radv (with git master
Vulkan-LoaderAndValidationLayers):

(gdb) bt
#0  0x000000000001e006 in ?? ()
#1  0x00007fffedcc7e42 in radv_lookup_entrypoint (name=<optimized out>) at
radv_entrypoints.c:933
#2  0x00007ffff796e2bb in loader_scanned_icd_add () from
/home/vliaskovitis/Vulkan-LoaderAndValidationLayers/loader/libvulkan.so.1
#3  0x00007ffff7972904 in loader_icd_scan () from
/home/vliaskovitis/Vulkan-LoaderAndValidationLayers/loader/libvulkan.so.1
#4  0x00007ffff797a017 in vkCreateInstance () from
/home/vliaskovitis/Vulkan-LoaderAndValidationLayers/loader/libvulkan.so.1
#5  0x0000000000401837 in ?? ()
#6  0x00007ffff75d8541 in __libc_start_main () from /lib64/libc.so.6
#7  0x0000000000403eaa in _start ()
(gdb) up
#1  0x00007fffedcc7e42 in radv_lookup_entrypoint (name=<optimized out>) at
radv_entrypoints.c:933
933     radv_entrypoints.c: No such file or directory.
(gdb) print e
$1 = (const struct radv_entrypoint *) 0x7fffedd967c0 <entrypoints>
(gdb) print *e
$2 = {name = 0, hash = 950370726}
(gdb) print i
$3 = 0

I think it happens when loader_scanned_icd_add attempts to find the radv entry
point:
fp_create_inst = (PFN_vkCreateInstance)fp_get_proc_addr(NULL,
"vkCreateInstance");

Mesa is 17.0.4-160.1

HW setup is completely different. I am testing on a laptop with intel GPU only,
but radv is still checked during vulkaninfo (if installed) and segfaults.

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20170503/51539be3/attachment.html>