Violet
5185138596
Create bindings for hipblasLt ( #510 )
...
ZLUDA / Build (Linux) (push) Waiting to run
ZLUDA / Build (Windows) (push) Waiting to run
ZLUDA / Build AMD GPU unit tests (push) Waiting to run
ZLUDA / Run AMD GPU unit tests (push) Blocked by required conditions
Generate bindings for hipblasLt and make some changes to the bindings for cublasLt. Notably, the `hip_type` `Option` is changed to a `Vec`, so that multiple `From` implementations (for `rocblas_error` and `hipblasLtError`) can be created for `cublasError_t`.
2025-09-16 16:23:15 -07:00
Andrzej Janik
6c811a55d2
Random fixes ( #504 )
...
This is a collection of random changes coming from the workload I'm working on. The most important change is better support for `.params`: PTX uses .param namespace both for some local variables and kernel args. This is a problem for us because those are different address spaces on AMDGPU. So far we've made an effort to convert to local and const namespaces whenever possible, but this commit tries to handle more patterns, which are impossible to track precisely, by converting to generic space.
2025-09-12 13:52:33 -07:00
Andrzej Janik
3632f2bf03
Some fixes to BLASLt ( #482 )
ZLUDA / Build (Linux) (push) Has been cancelled
ZLUDA / Build (Windows) (push) Has been cancelled
ZLUDA / Build AMD GPU unit tests (push) Has been cancelled
ZLUDA / Run AMD GPU unit tests (push) Has been cancelled
2025-08-26 14:28:36 -07:00
Andrzej Janik
ec1358af1c
Add more NVML and cuBLAS coverage ( #481 )
ZLUDA / Build (Linux) (push) Waiting to run
ZLUDA / Build (Windows) (push) Waiting to run
ZLUDA / Build AMD GPU unit tests (push) Waiting to run
ZLUDA / Run AMD GPU unit tests (push) Blocked by required conditions
2025-08-26 12:55:12 -07:00
Andrzej Janik
547536de4a
Add more instructions, tighten generated assembly ( #475 )
ZLUDA / Build (Linux) (push) Waiting to run
ZLUDA / Build (Windows) (push) Waiting to run
ZLUDA / Build AMD GPU unit tests (push) Waiting to run
ZLUDA / Run AMD GPU unit tests (push) Blocked by required conditions
2025-08-18 11:12:57 -07:00
Violet
21ef5f60a3
Check Rust formatting on pull requests ( #451 )
...
* Check Rust formatting on pull requests
This should help us maintain consistent style, without having unrelated style changes in pull requests from running `rustfmt`.
* cargo fmt non-generated files
* Ignore generated files
2025-07-30 14:55:09 -07:00
Violet
481c3550fa
Convert CUDA performance lib statuses to Rust result types ( #444 )
...
ZLUDA / Build (Linux) (push) Waiting to run
ZLUDA / Build (Windows) (push) Waiting to run
ZLUDA / Build AMD GPU unit tests (push) Waiting to run
ZLUDA / Run AMD GPU unit tests (push) Blocked by required conditions
These changes replicate how the main library is handled. cuDNN still needs to have zluda_bindgen run and zluda_dump_common updated
2025-07-29 14:28:14 -07:00
Violet
4ffa669cce
Fix Windows linkage ( #445 )
2025-07-29 10:15:32 -07:00
Violet
d81404eb70
Add support for rocblas to zluda_bindgen ( #440 )
...
ZLUDA / Build (Linux) (push) Waiting to run
ZLUDA / Build (Windows) (push) Waiting to run
ZLUDA / Build AMD GPU unit tests (push) Waiting to run
ZLUDA / Run AMD GPU unit tests (push) Blocked by required conditions
One step of several for adding cublas support
2025-07-28 15:07:22 -07:00
Andrzej Janik
5935cfec78
Work around broken AMD Adrenalin 25.5.1 driver ( #366 )
...
For reasons unknown AMD Adrenalin 25.5.1 ships with comgr that presents itself as version 2, but expects ABI for veersion 3. Add a workaround
2025-05-13 02:20:23 +02:00
Andrzej Janik
3d3e38aadc
Fix ROCm 6.4 failures ( #364 )
...
Lazy load comgr and dispatch to different code paths based on the name of the comgr .dll/.so
2025-05-02 00:38:22 +02:00
Andrzej Janik
cc83b9f1f6
Create infrastructure for performance libraries ( #363 )
2025-05-01 22:37:18 +02:00
Andrzej Janik
adc4673a20
Explicitly fail compilation on ROCm 6.4 ( #361 )
...
AMD broke comgr ABI in 6.4. This is a temporary solution.
2025-04-20 17:02:05 +02:00
Andrzej Janik
7a6df9dcbf
Fix host code and update to CUDA 12.4 ( #299 )
2024-12-02 00:29:57 +01:00
Andrzej Janik
46def3e7e0
Connect new parser to LLVM bitcode backend ( #269 )
...
Rust / Build and publish (Linux) (push) Has been cancelled
Rust / Build and publish (Windows) (push) Has been cancelled
This is very incomplete. Just enough code to emit LLVM bitcode and continue further development
2024-09-13 01:07:31 +02:00
Andrzej Janik
659b2c6ec4
Merge commit ' 4b96dbc8f4'
2021-01-03 17:54:01 +01:00
Andrzej Janik
77523940b3
Merge commit ' dabc40cb19' as 'ext/detours'
2021-01-03 17:52:14 +01:00
Andrzej Janik
f2f3eeb303
Validate PTX build tests against SPIR-V sources
2020-07-12 14:58:54 +02:00
