Hello, i suspect itis precisely R300 vs R600 features precisely of the driver s 3d part.
if it consists of 2d with its exa emphasis and glamor part emphasis on 3dfunctions in 2d and one the api .. ithinkitwas... .. dri ? #keypoint thenthe relevant part is the R300 driver part 3d part vs its features likely in R600 #keypoint isuggest shifting rv530 into the acceptance/features sampling/ applying of r600 including the agp to pcie translator chip allthebest Christian KISS Original Message From: Diederik de Haas Sent: Freitag, 23. Februar 2024 16:48 To: Christian Kiss; 1035...@bugs.debian.org Subject: Re: Bug#1035880: radeon: radeon driver only partial support of pipelines Control: reassign -1 src:linux 5.10.178-1 On Wednesday, 10 May 2023 15:41:44 CET Christian Kiss wrote: > Package: firmware-amd-graphics > Version: 20210315-3 > > kernel.log reports that radeon: 1 quad pipes, 2 z pipes initialized. > the gpu an rv530 has atleast double to 4 of these each. > > the rudimentary partial support of the hardware features includes oddity > clock at 85.5Mhz instead 500 / 1100ram > > while the bit width 128seems to becorrect > > additionally to this the realtime kernel scheduling seems to break the gpu/ > non optimised coordination cpu gpu scheduling of cpucycles This seems to be a kernel issue, reassigning accordingly.
