2eQTu

Results 22 comments of 2eQTu

@cgmb Thanks for the ISA incompatibility heads up for Navi 31/32/33. Good to know. I actually had just started going through the [RDNA 3 ISA doc](https://www.amd.com/content/dam/amd/en/documents/radeon-tech-docs/instruction-set-architectures/rdna3-shader-instruction-set-architecture-feb-2023_0.pdf), but did not notice...

FYI to AMD folks, the POC issue bot assigned the wrong label for hardware. Per output above, submitter is talking about an "AMD Radeon RX 7800 XT" , but no...

Nice! Good to hear. Yes, I did pick up a 9070 XT. Happy to try any test builds or whatever if you are delayed in getting one yourself.

Will give it a try. Is there any subset of libs/apps and some sanity check tests which should be tried first?

also. FYI, that branch build dies on a CMake ROCM_ROOT_DIR not found error when it hits rocprofiler. Have not looked in to it yet.

FYI that original error was from a fresh Ubuntu 24.04 container. Will try again on latest changes. btw, is there any suggested way to specify the gfx arch to build...

FYI branch now breaks on: ``` Building wheel torch-2.7.0-rc2 -- Building version 2.7.0-rc2 -- Checkout nccl release tag: v2.25.1-1 cmake -GNinja -DBUILD_CPU_COUNT_DEFAULT=24 -DBUILD_CPU_COUNT_MAX=32 -DBUILD_CPU_COUNT_MIN=1 -DBUILD_CPU_COUNT_MODERATE=20 -DBUILD_CPU_COUNT_SAFE=12 -DBUILD_PYTHON=True -DBUILD_ROOT_DIR=/rocm_sdk_builder_633/builddir -DBUILD_RULE_ROOT_DIR=/rocm_sdk_builder_633/binfo -DBUILD_SCRIPT_ROOT_DIR=/rocm_sdk_builder_633/build...

fwiw, looking back at the log I do see a few: ``` CMake Error at /opt/rocm_sdk_633/lib/cmake/hiprtc/hiprtc-config.cmake:63 (include): include could not find requested file: /opt/rocm_sdk_633/lib/cmake/hiprtc/hiprtc-targets.cmake Call Stack (most recent call first):...

Just tried again. looks like a break related to, or in the area of, zstd. This is on 6b78f42 with the zstd update. ``` In file included from /rocm_sdk_builder_633/src_projects/SuiteSparse/GraphBLAS/Source/GB_zstd.c:63: /rocm_sdk_builder_633/src_projects/SuiteSparse/GraphBLAS/zstd/zstd_subset/compress/zstd_compress.c:943:24:...

Seems to have gotten past that point after pulling latest, but still building. I will try again anyway in a fresh env and report back either way. FYI from looking...