RPCS3/rpcs3
1
0
Fork 0
mirror of https://github.com/RPCS3/rpcs3.git synced 2026-01-07 17:20:28 +01:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

Move "glMultiDrawArrays is broken on the old AMD OpenGL Windows driver" to solved

Ani 2025-11-18 06:10:47 +01:00
parent 5371d00fa5
commit 9e21071b37
1 changed files with 9 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
Graphics-Driver-Issues.md

@ -15,19 +15,15 @@ This page quickly summarizes graphics driver quirks and issues present in GPU dr
» **Workaround** implemented: No <br>
» **Reported**: No <br>
» **Note**: Intel proprietary OpenGL drivers are unsupported by RPCS3 due to their poor state.
- _There are likely other issues as well..._
- _Unsupported. There are likely other issues as well..._
**Vulkan**
- _No known problems at this time_
- _Unsupported. There are currently known issues..._
## AMD (Proprietary) <a name="amd-proprietary"/>
**OpenGL**
- [glMultiDrawArrays is broken](https://community.amd.com/message/2858799) <br>
» **Workaround** implemented: **Yes** <br>
» **Reported**: **Yes**, on 2018-04-25. No replies. <br>
» **Note**: This seems to be undefined behavior in the spec. The gl_VertexID propagation behavior across MultiDrawArrays is not defined, i.e whether it should continue incrementing of reset to 0 on a new instance. However, due to the comparison with a loop of [first, count] and implied functional equivalence, this may be interpreted as a bug. Either way, RPCS3 provides a workaround using index lookups.
- _No known problems at this time_
**Vulkan**
- Primitive restart is 'broken', this seems to be a GCN hardware bug as it affects both proprietary and open-source drivers <br>
@ -109,6 +105,12 @@ List of reported graphics driver bugs that are now fixed. We'd like to thank the
» **Workaround** used: None <br>
» **Fixed** on: NVIDIA Windows 527.37
- **High:** [glMultiDrawArrays is broken on the old AMD OpenGL Windows driver](https://github.com/RPCS3/rpcs3/issues/17711) <br>
» **Affected** drivers: Radeon Software Adrenalin 22.6.1 and older <br>
» **Workaround** implemented: **Yes** - this seems to be undefined behavior in the spec. The gl_VertexID propagation behavior across MultiDrawArrays is not defined, i.e whether it should continue incrementing of reset to 0 on a new instance. However, due to the comparison with a loop of [first, count] and implied functional equivalence, this may be interpreted as a bug. Either way, RPCS3 provides a workaround using index lookups. <br>
» **Reported**: **Yes**, on 2018-04-25. [No replies.](https://community.amd.com/message/2858799) <br>
» **Fixed** on: Radeon Software Adrenalin 22.7.1
- **High:** Performance regression on AMD Windows with very high submit and flip latency per frame <br>
» **Affected** drivers: Radeon Software Adrenalin 23.3.2, 23.4.1, 23.4.2 <br>
» **Workaround** used: None <br>