11:22 <MoeIcenowy> simon-perretta-img: BTW the GCC incompatibility issue (pvr_device_info_33_15_11_3 repeated initialization of .ident) looks to still exist in newest dev/bxs

11:29 <simon-perretta-img> MoeIcenowy: Ah I was actually going to ask about that - do you mind if I include the fix you linked in the next MR that'll contain the device info change?

14:47 <MoeIcenowy> simon-perretta-img: quite okay

14:49 <MoeIcenowy> https://gitlab.freedesktop.org/imagination/mesa/-/commit/d23425877f979674f270919807f2b2bbf05799a8#e76bd5bf52504b3eec7dbe76b7e23b2966334176_128_130 oops this really messed up

14:50 <MoeIcenowy> all changes in this commit have n and v reversed

14:50 <MoeIcenowy> and the value for 36.53.104.796 looks bogus (a wild 40 appears)

14:59 <simon-perretta-img> Oops yeah not sure how I missed that the first time round haha, I'll adjust it, cheers!

15:06 <MoeIcenowy> (and on GPUs I added, when I unset PVR_I_WANT_A_BROKEN_VULKAN_DRIVER, no warning is shown as expected

15:07 <MoeIcenowy> it silently fails

15:07 <MoeIcenowy> (and on all other mesa drivers being non-conformant is only a warning, no failing will happen, a helper function for warning even exists(vk_warn_non_conformant_implementation)

15:27 <simon-perretta-img> Ah, I'm guessing that's happening with a release build? Will feed it back

15:36 <MoeIcenowy> oh I may forgot to set it to debug then

15:47 <simon-perretta-img> I think you're right in that we probably still want something to be printed in release builds though