happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Adjust workaround for ROCWMMA_FATTN/GFX9 to only newer ROCm veresions (#19591) 

Avoids issues with ROCm 6.4.4.

Closes: https://github.com/ggml-org/llama.cpp/issues/19580
Fixes: 6845f7f87 ("Add a workaround for compilation with ROCWMMA_FATTN and gfx9 (#19461)")

Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
This commit is contained in:
Mario Limonciello 2026-02-16 07:46:08 -06:00 committed by GitHub
parent cc45f2ada6
commit 2ba9adc093
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-cuda/fattn-wmma-f16.cu
View File

@ -63,7 +63,7 @@ static __global__ void flash_attn_ext_f16(
constexpr int frag_m = ncols == 8 ? 32 : 16;
constexpr int frag_n = ncols == 8 ? 8 : 16;
static_assert(D % frag_m == 0, "If ncols == 8 then D % frag_m must be 0.");
#if defined(GGML_USE_HIP)
#if defined(GGML_USE_HIP) && HIP_VERSION >= 60500000
typedef wmma::fragment<wmma::matrix_a, frag_m, frag_n, 16, _Float16, wmma::row_major> frag_a_K;
typedef wmma::fragment<wmma::matrix_a, frag_m, frag_n, 16, _Float16, wmma::col_major> frag_a_V;
typedef wmma::fragment<wmma::matrix_b, frag_m, frag_n, 16, _Float16, wmma::col_major> frag_b;
@ -135,7 +135,7 @@ static __global__ void flash_attn_ext_f16(
__shared__ half VKQ[ncols*D_padded]; // Accumulator for final VKQ slice.
half2 * VKQ2 = (half2 *) VKQ;
#if defined(GGML_USE_HIP)
#if defined(GGML_USE_HIP) && HIP_VERSION >= 60500000
const _Float16 * K_h_f16 = reinterpret_cast<const _Float16 *>(K_h);
const _Float16 * V_h_f16 = reinterpret_cast<const _Float16 *>(V_h);
_Float16 * KQ_f16 = reinterpret_cast<_Float16 *>(KQ);