Bug#1118618: Updating rocr-runtime libraries to 6.4 causes llama.cpp/ggml to fail
Hi Cory,
On 2025-10-23 01:16, Cordell Bloor wrote:
> I was testing with packages built against newer versions of the HIP Runtime, so I didn't catch this regression. I'd only tested the basics with rocminfo, which worked fine for this case because it doesn't use the broken function.
Regression tests like these should be testable in our CI, by testing the
reverse dependencies from testing while pinning the new package from
unstable. This is what the official debci does to prevent migration.
For reverse build dependencies, bin:ratt helps.
On 2025-10-23 05:10, Cordell Bloor wrote:
> Control: reassign -1 libamdhip64-5 5.7.1-6
>
> On 2025-10-22 17:16, Cordell Bloor wrote:
>> The problem seems to be that upstream changed the ABI of
>> hsa_agent_iterate_isas by changing the types for the callback function
>
> I was mistaken. I tried to fix the ABI breakage only to find nothing
> there. AFAICT, the expected data layout for the callback function is
> defined by the caller, not by libhsa-runtime64-1. I created a minimal
> reproducer to compare the traced output when linked libamdhip64-5 and
> libamdhip64-6.
> [...]
> This bad assumption was fixed in libamdhip64-6 6.4.3-1~exp2.Unless I'm missing something, this will still need an additional fix
because there is nothing yet stopping rocr-runtime from being rebuilt
from source with the "wrong" libamdhip64-dev.
I believe either rocr-runtime's B-D of libamdhip64-dev needs to be
bumped to 6.4.3-1~exp2, or libamdhip64-5 needs to be fixed in stable.
Best,
Christian
Reply to: