derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
1245 lines
29 KiB
YAML
1245 lines
29 KiB
YAML
---
|
|
include: hipblas_common.yaml
|
|
#include: known_bugs.yaml
|
|
|
|
# This is a minimal smoke test of rocblas functionality
|
|
# It will run at least a single small sized function test for all of the L1, L2, and L3 functions.
|
|
# Algorithm variations controlled by enums will be tested but those that are selected by size will not be
|
|
# exhaustive as very small sizes are only used
|
|
|
|
Definitions:
|
|
|
|
############
|
|
# Level 1+ #
|
|
############
|
|
|
|
- &incx_range
|
|
- [1]
|
|
|
|
- &N_range
|
|
- [513]
|
|
|
|
- &batch_count_range
|
|
- [2]
|
|
|
|
- &incx_incy_range
|
|
- { incx: 1, incy: 1 }
|
|
|
|
- &incx_incy_range_x_y_output
|
|
- { incx: 1, incy: 1 }
|
|
|
|
- &incx_incy_range_y_output
|
|
- { incx: 1, incy: 1 }
|
|
|
|
- &alpha_range
|
|
- [0.5]
|
|
|
|
- &alpha_beta_range
|
|
- { alpha: 0.5, beta: -1.0 }
|
|
|
|
- &alphai_betai_range
|
|
- { alphai: 0.5 }
|
|
|
|
- &complex_alpha_range
|
|
- { alpha: 1.5, alphai: -1.0 }
|
|
|
|
- &complex_alpha_beta_range
|
|
- { alpha: -0.5, beta: -0.5, alphai: 0.5, betai: 0.5 }
|
|
|
|
###########
|
|
# Level 2 #
|
|
###########
|
|
|
|
- &gbmv_dims
|
|
- { M: 20, N: 30, lda: 50, KL: 14, KU: 14 }
|
|
|
|
- &gbmv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
transA: [ N, T ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *alpha_beta_range
|
|
matrix_size: *gbmv_dims
|
|
|
|
- &gemv_dims
|
|
- { M: 65, N: 65, lda: 65, stride_a: 1 }
|
|
|
|
- &gemv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
transA: [ N, T ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *alpha_beta_range
|
|
matrix_size: *gemv_dims
|
|
|
|
- &hbmv_dims
|
|
- { N: 65, K: 64, lda: 65 }
|
|
|
|
- &hbmv_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *hbmv_dims
|
|
|
|
- &hemv_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &hemv_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *hemv_dims
|
|
|
|
- &her_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &her_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
incx: *incx_range
|
|
matrix_size: *her_dims
|
|
|
|
- &her2_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &her2_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
incx_incy: *incx_incy_range
|
|
matrix_size: *her2_dims
|
|
|
|
- &hpmv_dims
|
|
- { N: 65 }
|
|
|
|
- &hpmv_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *hpmv_dims
|
|
|
|
- &hpr_dims
|
|
- { N: 65 }
|
|
|
|
- &hpr_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_range
|
|
incx: *incx_range
|
|
matrix_size: *hpr_dims
|
|
|
|
- &hpr2_dims
|
|
- { N: 65 }
|
|
|
|
- &hpr2_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_range
|
|
incx_incy: *incx_incy_range
|
|
matrix_size: *hpr2_dims
|
|
|
|
|
|
- &ger_dims
|
|
- { M: 65, N: 65, lda: 65, stride_a: 4225}
|
|
|
|
- &ger_common_args
|
|
precision: *single_double_precisions
|
|
incx_incy: *incx_incy_range
|
|
alpha: *alpha_range
|
|
matrix_size: *ger_dims
|
|
|
|
- &geruc_dims
|
|
- { M: 65, N: 65, lda: 65, stride_a: 4225}
|
|
|
|
- &geruc_common_args
|
|
precision: *single_double_precisions_complex
|
|
incx_incy: *incx_incy_range
|
|
alpha: *alpha_range
|
|
matrix_size: *geruc_dims
|
|
|
|
- &spr_dims
|
|
- { N: 65 }
|
|
|
|
- &spr_common_args
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_range
|
|
incx: *incx_range
|
|
matrix_size: *spr_dims
|
|
|
|
- &spr2_dims
|
|
- { N: 65 }
|
|
|
|
- &spr2_common_args
|
|
precision: *single_double_precisions
|
|
uplo: [ U, L ]
|
|
alpha: *alpha_range
|
|
incx_incy: *incx_incy_range
|
|
matrix_size: *spr2_dims
|
|
|
|
- &syr_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &syr_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_range
|
|
incx: *incx_range
|
|
matrix_size: *syr_dims
|
|
|
|
- &syr2_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &syr2_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_range
|
|
incx_incy: *incx_incy_range
|
|
matrix_size: *syr2_dims
|
|
|
|
- &symv_dims
|
|
- { N: 65, lda: 65 }
|
|
|
|
- &symv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *symv_dims
|
|
|
|
- &sbmv_dims
|
|
- { N: 65, K: 11, lda: 65 }
|
|
|
|
- &sbmv_common_args
|
|
precision: *single_double_precisions
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *alpha_beta_range
|
|
matrix_size: *sbmv_dims
|
|
|
|
- &spmv_dims
|
|
- { N: 65, K: 11, lda: 65 }
|
|
|
|
- &spmv_common_args
|
|
precision: *single_double_precisions
|
|
uplo: [ U, L ]
|
|
incx_incy: *incx_incy_range
|
|
alpha_beta: *alpha_beta_range
|
|
matrix_size: *spmv_dims
|
|
|
|
- &tbmv_dims
|
|
- { M: 257, K: 11, lda: 257 }
|
|
|
|
- &tbmv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *tbmv_dims
|
|
|
|
- &tpmv_dims
|
|
- { M: 290, stride_a: 1 }
|
|
|
|
- &tpmv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *tpmv_dims
|
|
|
|
- &trmv_dims
|
|
- { M: 257, lda: 257, stride_a: 1 }
|
|
|
|
- &trmv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T, C]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *trmv_dims
|
|
|
|
- &tbsv_dims
|
|
- { N: 65, K: 5, lda: 65 }
|
|
|
|
- &tbsv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *tbsv_dims
|
|
|
|
- &tpsv_dims
|
|
- { N: 65 }
|
|
|
|
- &tpsv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *tpsv_dims
|
|
|
|
- &trsv_dims
|
|
- { M: 33, lda: 33 }
|
|
|
|
- &trsv_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [U, L]
|
|
transA: [N, T]
|
|
diag: [N, U]
|
|
incx: *incx_range
|
|
matrix_size: *trsv_dims
|
|
|
|
###########
|
|
# Level 3 #
|
|
###########
|
|
|
|
- &transA_transB_range
|
|
- { transA: [N,T,C], transB: [N,T,C] }
|
|
|
|
- &dgmm_dims
|
|
- { M: 33, N: 33, lda: 35, incx: 2, ldc: 35 }
|
|
|
|
- &dgmm_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
side: [L, R]
|
|
matrix_size: *dgmm_dims
|
|
|
|
- &geam_dims
|
|
- { M: 3, N: 33, lda: 35, ldb: 35, ldc: 35 }
|
|
|
|
- &geam_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
transA_transB: *transA_transB_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *geam_dims
|
|
|
|
- &hemm_dims
|
|
- { M: 12, N: 33, lda: 33, ldb: 33, ldc: 33 }
|
|
|
|
- &hemm_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
side: [L, R]
|
|
alpha_beta: *alpha_beta_range
|
|
matrix_size: *hemm_dims
|
|
|
|
- &herk_dims
|
|
- { N: 65, lda: 65, K: 33, ldc: 65 }
|
|
|
|
- &herk_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
transA: [ N, C ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *herk_dims
|
|
|
|
- &herk_ex_dims
|
|
- { N: 65, lda: 65, K: 33, ldc: 65 }
|
|
|
|
- &herk_ex_common_args
|
|
precision: *herk_ex_precisions
|
|
uplo: [ U, L ]
|
|
transA: [ N, C ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *herk_ex_dims
|
|
|
|
- &her2k_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &her2k_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
transA: [ N, C ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *her2k_dims
|
|
|
|
- &herkx_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &herkx_common_args
|
|
precision: *single_double_precisions_complex
|
|
uplo: [ U, L ]
|
|
transA: [ N, C ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *herkx_dims
|
|
|
|
- &symm_dims
|
|
- { M: 12, N: 33, lda: 33, ldb: 33, ldc: 33 }
|
|
|
|
- &symm_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
side: [ L, R ]
|
|
uplo: [ U, L ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *symm_dims
|
|
|
|
- &syrk_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &syrk_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [ U, L ]
|
|
transA: [ N, T ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *syrk_dims
|
|
|
|
- &syrk_ex_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &syrk_ex_common_args
|
|
precision: *syrk_ex_precisions
|
|
uplo: [ U, L ]
|
|
transA: [ N, T ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *syrk_ex_dims
|
|
|
|
- &syr2k_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &syr2k_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [ U, L ]
|
|
transA: [ N, T ]
|
|
matrix_size: *syr2k_dims
|
|
|
|
- &syrkx_dims
|
|
- { N: 12, K: 9, lda: 12, ldb: 12, ldc: 12 }
|
|
|
|
- &syrkx_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
uplo: [ U, L ]
|
|
transA: [ N, T ]
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *syrkx_dims
|
|
|
|
- &gemm_dims
|
|
- { M: 65, N: 33, K: 33, lda: 128, ldb: 128, ldc: 128 }
|
|
|
|
- &gemm_common_args
|
|
precision: *single_double_precisions_complex_real
|
|
transA_transB: *transA_transB_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *gemm_dims
|
|
|
|
- &gemm_ex_dims
|
|
- { M: 65, N: 33, K: 33, lda: 128, ldb: 128, ldc: 128, ld: 128 }
|
|
|
|
- &gemm_ex_common_args
|
|
precision: *hpa_half_precision
|
|
transA_transB: *transA_transB_range
|
|
alpha_beta: *complex_alpha_beta_range
|
|
matrix_size: *gemm_ex_dims
|
|
|
|
Tests:
|
|
|
|
###########
|
|
# Level 1 #
|
|
###########
|
|
|
|
# All functions with no parameters to scan over
|
|
|
|
- name: blas1
|
|
category: quick
|
|
function:
|
|
- rotg: *rotg_precisions
|
|
- rotmg: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
|
|
- name: blas1_batched
|
|
category: quick
|
|
batch_count: *batch_count_range
|
|
function:
|
|
- rotg_batched: *rotg_precisions
|
|
- rotmg_batched: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
- name: blas1_strided_batched
|
|
category: quick
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1.0 ]
|
|
function:
|
|
- rotg_strided_batched: *rotg_precisions
|
|
- rotmg_strided_batched: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
# All functions with alpha and incx and incy
|
|
|
|
- name: blas1_with_alpha
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_y_output
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
function:
|
|
#- axpy: *half_single_precisions_complex_real
|
|
- axpy: *single_double_precisions_complex_real
|
|
- axpy_ex: *axpy_ex_precision_all
|
|
|
|
- name: blas1_batched_with_alpha
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_y_output
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
batch_count: *batch_count_range
|
|
function:
|
|
#- axpy_batched: *half_single_precisions_complex_real
|
|
- axpy_batched: *single_double_precisions_complex_real
|
|
- axpy_batched_ex: *half_single_double_complex_real_precisions
|
|
backend_flags: AMD
|
|
|
|
- name: blas1_strided_batched_with_alpha
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_y_output
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
function:
|
|
#- axpy_strided_batched: *half_single_precisions_complex_real
|
|
- axpy_strided_batched: *single_double_precisions_complex_real
|
|
- axpy_strided_batched_ex: *half_single_double_complex_real_precisions
|
|
backend_flags: AMD
|
|
|
|
|
|
# All functions with alpha and incx but no incy
|
|
|
|
- name: blas1_with_alpha
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
function:
|
|
- scal: *single_double_precisions_complex_real
|
|
#- scal: *scal_ex_single_double_complex_real_alpha_complex_out
|
|
- scal_ex: *scal_ex_precision_all
|
|
#- scal_ex: *scal_ex_single_double_complex_real_alpha_complex_out
|
|
|
|
- name: blas1_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
batch_count: *batch_count_range
|
|
function:
|
|
- scal_batched: *single_double_precisions_complex_real
|
|
#- scal_batched: *scal_ex_single_double_complex_real_alpha_complex_out
|
|
- scal_batched_ex: *scal_ex_precision_all
|
|
- scal_batched_ex: *scal_ex_precision_amd
|
|
backend_flags: AMD
|
|
|
|
- name: blas1_strided_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
alpha_beta: *alpha_beta_range
|
|
alphai_betai: *alphai_betai_range
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
function:
|
|
- scal_strided_batched: *single_double_precisions_complex_real
|
|
#- scal_strided_batched: *scal_ex_single_double_complex_real_alpha_complex_out
|
|
- scal_strided_batched_ex: *scal_ex_precision_all
|
|
- scal_strided_batched_ex: *scal_ex_precision_amd
|
|
backend_flags: AMD
|
|
|
|
# All functions with incx, no incy, no alpha
|
|
# nrm2,asum,iamax,iamin all variants
|
|
|
|
- name: blas1
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
function:
|
|
- nrm2: *single_double_precisions_complex_real
|
|
- asum: *single_double_precisions_complex_real
|
|
- iamax: *single_double_precisions_complex_real
|
|
- iamin: *single_double_precisions_complex_real
|
|
|
|
- name: blas1_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
batch_count: *batch_count_range
|
|
function:
|
|
- asum_batched: *single_double_precisions_complex_real
|
|
- nrm2_batched: *single_double_precisions_complex_real
|
|
- iamax_batched: *single_double_precisions_complex_real
|
|
- iamin_batched: *single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
- name: blas1_strided_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx: *incx_range
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1.0 ]
|
|
function:
|
|
- asum_strided_batched: *single_double_precisions_complex_real
|
|
- nrm2_strided_batched: *single_double_precisions_complex_real
|
|
- iamax_strided_batched: *single_double_precisions_complex_real
|
|
- iamin_strided_batched: *single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
# All functions with incx, incy, no alpha
|
|
|
|
- name: blas1
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range
|
|
function:
|
|
- copy: *single_double_precisions_complex_real
|
|
#- dot: *half_bfloat_single_double_complex_real_precisions
|
|
- dot: *single_double_precisions_complex_real
|
|
- dotc: *single_double_precisions_complex
|
|
#*single_double_precisions_complex_real
|
|
|
|
|
|
- name: blas1_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range
|
|
batch_count: *batch_count_range
|
|
function:
|
|
- copy_batched: *single_double_precisions_complex_real
|
|
#- dot_batched: *half_bfloat_single_double_complex_real_precisions
|
|
- dot_batched: *single_double_precisions_complex_real
|
|
- dotc_batched: *single_double_precisions_complex
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
- name: blas1_strided_batched
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range
|
|
batch_count: *batch_count_range
|
|
stride_scale: [1]
|
|
function:
|
|
- copy_strided_batched: *single_double_precisions_complex_real
|
|
#- dot_strided_batched: *half_bfloat_single_double_complex_real_precisions
|
|
- dot_strided_batched: *single_double_precisions_complex_real
|
|
- dotc_strided_batched: *single_double_precisions_complex
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
# Swap, rot, and rotm get their own for quick tests as they write to x and y
|
|
|
|
- name: blas1
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_x_y_output
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
function:
|
|
- swap: *single_double_precisions_complex_real
|
|
- rot: *rot_precisions
|
|
- rotm: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
|
|
- name: blas1
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_x_y_output
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
function:
|
|
- swap_batched: *single_double_precisions_complex_real
|
|
- rot_batched: *rot_precisions
|
|
- rotm_batched: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
- name: blas1
|
|
category: quick
|
|
N: *N_range
|
|
incx_incy: *incx_incy_range_x_y_output
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
function:
|
|
- swap_strided_batched: *single_double_precisions_complex_real
|
|
- rot_strided_batched: *rot_precisions
|
|
- rotm_strided_batched: *single_double_precisions
|
|
#*single_double_precisions_complex_real
|
|
backend_flags: AMD
|
|
|
|
###########
|
|
# Level 2 #
|
|
###########
|
|
|
|
# gbmv
|
|
- name: gbmv
|
|
category: quick
|
|
function: gbmv
|
|
arguments: *gbmv_common_args
|
|
|
|
- name: gbmv_batched
|
|
category: quick
|
|
function:
|
|
- gbmv_batched
|
|
- gbmv_strided_batched
|
|
arguments: *gbmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: gemv
|
|
category: quick
|
|
function: gemv
|
|
arguments: *gemv_common_args
|
|
|
|
- name: gemv_batched
|
|
category: quick
|
|
function:
|
|
- gemv_batched
|
|
- gemv_strided_batched
|
|
arguments: *gemv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: ger
|
|
category: quick
|
|
function: ger
|
|
arguments: *ger_common_args
|
|
|
|
- name: ger_batched
|
|
category: quick
|
|
function:
|
|
- ger_batched
|
|
- ger_strided_batched
|
|
arguments: *ger_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: geruc
|
|
category: quick
|
|
function:
|
|
- gerc
|
|
- geru
|
|
arguments: *geruc_common_args
|
|
|
|
- name: geruc_batched
|
|
category: quick
|
|
function:
|
|
- gerc_batched
|
|
- gerc_strided_batched
|
|
- geru_batched
|
|
- geru_strided_batched
|
|
arguments: *geruc_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hbmv
|
|
category: quick
|
|
function: hbmv
|
|
arguments: *hbmv_common_args
|
|
|
|
- name: hbmv_batched
|
|
category: quick
|
|
function:
|
|
- hbmv_batched
|
|
- hbmv_strided_batched
|
|
arguments: *hbmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hemv
|
|
category: quick
|
|
function: hemv
|
|
arguments: *hemv_common_args
|
|
|
|
- name: hemv_batched
|
|
category: quick
|
|
function:
|
|
- hemv_batched
|
|
- hemv_strided_batched
|
|
arguments: *hemv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: her
|
|
category: quick
|
|
function: her
|
|
arguments: *her_common_args
|
|
|
|
- name: her_batched
|
|
category: quick
|
|
function:
|
|
- her_batched
|
|
- her_strided_batched
|
|
arguments: *her_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: her2
|
|
category: quick
|
|
function: her2
|
|
arguments: *her2_common_args
|
|
|
|
- name: her2_batched
|
|
category: quick
|
|
function:
|
|
- her2_batched
|
|
- her2_strided_batched
|
|
arguments: *her2_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hpmv
|
|
category: quick
|
|
function: hpmv
|
|
arguments: *hpmv_common_args
|
|
|
|
- name: hpmv_batched
|
|
category: quick
|
|
function:
|
|
- hpmv_batched
|
|
- hpmv_strided_batched
|
|
arguments: *hpmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hpr
|
|
category: quick
|
|
function: hpr
|
|
arguments: *hpr_common_args
|
|
|
|
- name: hpr_batched
|
|
category: quick
|
|
function:
|
|
- hpr_batched
|
|
- hpr_strided_batched
|
|
arguments: *hpr_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hpr2
|
|
category: quick
|
|
function: hpr2
|
|
arguments: *hpr2_common_args
|
|
|
|
- name: hpr2_batched
|
|
category: quick
|
|
function:
|
|
- hpr2_batched
|
|
- hpr2_strided_batched
|
|
arguments: *hpr2_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: sbmv
|
|
category: quick
|
|
function: sbmv
|
|
arguments: *sbmv_common_args
|
|
|
|
- name: sbmv_batched
|
|
category: quick
|
|
function:
|
|
- sbmv_batched
|
|
- sbmv_strided_batched
|
|
arguments: *sbmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: spmv
|
|
category: quick
|
|
function: spmv
|
|
arguments: *spmv_common_args
|
|
|
|
- name: spmv_batched
|
|
category: quick
|
|
function:
|
|
- spmv_batched
|
|
- spmv_strided_batched
|
|
arguments: *spmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: spr_NVIDIA
|
|
category: quick
|
|
precision: *single_double_precisions
|
|
function: spr
|
|
arguments: *spr_common_args
|
|
backend_flags: NVIDIA
|
|
|
|
- name: spr_AMD
|
|
category: quick
|
|
precision: *single_double_precisions_complex_real
|
|
function:
|
|
- spr
|
|
- spr_batched
|
|
- spr_strided_batched
|
|
arguments: *spr_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: spr2
|
|
category: quick
|
|
function: spr2
|
|
arguments: *spr2_common_args
|
|
|
|
- name: spr2_batched
|
|
category: quick
|
|
function:
|
|
- spr2_batched
|
|
- spr2_strided_batched
|
|
arguments: *spr2_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: symv
|
|
category: quick
|
|
function: symv
|
|
arguments: *symv_common_args
|
|
|
|
- name: symv_batched
|
|
category: quick
|
|
function:
|
|
- symv_batched
|
|
- symv_strided_batched
|
|
arguments: *symv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: syr
|
|
category: quick
|
|
function: syr
|
|
arguments: *syr_common_args
|
|
|
|
- name: syr_batched
|
|
category: quick
|
|
function:
|
|
- syr_batched
|
|
- syr_strided_batched
|
|
arguments: *syr_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: syr2
|
|
category: quick
|
|
function: syr2
|
|
arguments: *syr2_common_args
|
|
|
|
- name: syr2_batched
|
|
category: quick
|
|
function:
|
|
- syr2_batched
|
|
- syr2_strided_batched
|
|
arguments: *syr2_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: tbmv
|
|
category: quick
|
|
function: tbmv
|
|
arguments: *tbmv_common_args
|
|
|
|
- name: tbmv_batched
|
|
category: quick
|
|
function:
|
|
- tbmv_batched
|
|
- tbmv_strided_batched
|
|
arguments: *tbmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: tpmv
|
|
category: quick
|
|
function: tpmv
|
|
arguments: *tpmv_common_args
|
|
|
|
- name: tpmv_batched
|
|
category: quick
|
|
function:
|
|
- tpmv_batched
|
|
- tpmv_strided_batched
|
|
arguments: *tpmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: trmv
|
|
category: quick
|
|
function: trmv
|
|
arguments: *trmv_common_args
|
|
|
|
- name: trmv_batched
|
|
category: quick
|
|
function:
|
|
- trmv_batched
|
|
- trmv_strided_batched
|
|
arguments: *trmv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: tbsv
|
|
category: quick
|
|
function: tbsv
|
|
arguments: *tbsv_common_args
|
|
|
|
- name: tbsv_batched
|
|
category: quick
|
|
function:
|
|
- tbsv_batched
|
|
- tbsv_strided_batched
|
|
arguments: *tbsv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: tpsv
|
|
category: quick
|
|
function: tpsv
|
|
arguments: *tpsv_common_args
|
|
|
|
- name: tpsv_batched
|
|
category: quick
|
|
function:
|
|
- tpsv_batched
|
|
- tpsv_strided_batched
|
|
arguments: *tpsv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: trsv
|
|
category: quick
|
|
function: trsv
|
|
arguments: *trsv_common_args
|
|
|
|
- name: trsv_batched
|
|
category: quick
|
|
function:
|
|
- trsv_batched
|
|
- trsv_strided_batched
|
|
arguments: *trsv_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
###########
|
|
# Level 3 #
|
|
###########
|
|
|
|
- name: dgmm
|
|
category: quick
|
|
function: dgmm
|
|
arguments: *dgmm_common_args
|
|
|
|
- name: dgmm_batched
|
|
category: quick
|
|
function:
|
|
- dgmm_batched
|
|
- dgmm_strided_batched
|
|
arguments: *dgmm_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: geam
|
|
category: quick
|
|
function: geam
|
|
arguments: *geam_common_args
|
|
|
|
- name: geam_batched
|
|
category: quick
|
|
function:
|
|
- geam_batched
|
|
- geam_strided_batched
|
|
arguments: *geam_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: hemm
|
|
category: quick
|
|
function: hemm
|
|
arguments: *hemm_common_args
|
|
|
|
- name: hemm_batched
|
|
category: quick
|
|
function:
|
|
- hemm_batched
|
|
- hemm_strided_batched
|
|
arguments: *hemm_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: herk
|
|
category: quick
|
|
function: herk
|
|
arguments: *herk_common_args
|
|
|
|
- name: herk_batched
|
|
category: quick
|
|
function:
|
|
- herk_batched
|
|
- herk_strided_batched
|
|
arguments: *herk_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: herk_ex
|
|
category: quick
|
|
function: herk_ex
|
|
arguments: *herk_ex_common_args
|
|
|
|
- name: her2k
|
|
category: quick
|
|
function: her2k
|
|
arguments: *her2k_common_args
|
|
|
|
- name: her2k_batched
|
|
category: quick
|
|
function:
|
|
- her2k_batched
|
|
- her2k_strided_batched
|
|
arguments: *her2k_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: herkx
|
|
category: quick
|
|
function: herkx
|
|
arguments: *herkx_common_args
|
|
|
|
- name: herkx_batched
|
|
category: quick
|
|
function:
|
|
- herkx_batched
|
|
- herkx_strided_batched
|
|
arguments: *herkx_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: symm
|
|
category: quick
|
|
function: symm
|
|
arguments: *symm_common_args
|
|
|
|
- name: symm_batched
|
|
category: quick
|
|
function:
|
|
- symm_batched
|
|
- symm_strided_batched
|
|
arguments: *symm_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: syrk
|
|
category: quick
|
|
function: syrk
|
|
arguments: *syrk_common_args
|
|
|
|
- name: syrk_batched
|
|
category: quick
|
|
function:
|
|
- syrk_batched
|
|
- syrk_strided_batched
|
|
arguments: *syrk_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: syrk_ex
|
|
category: quick
|
|
function: syrk_ex
|
|
arguments: *syrk_ex_common_args
|
|
|
|
- name: syr2k
|
|
category: quick
|
|
function: syr2k
|
|
arguments: *syr2k_common_args
|
|
|
|
- name: syr2k_batched
|
|
category: quick
|
|
function:
|
|
- syr2k_batched
|
|
- syr2k_strided_batched
|
|
arguments: *syr2k_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: syrkx
|
|
category: quick
|
|
function: syrkx
|
|
arguments: *syrkx_common_args
|
|
|
|
- name: syrkx_batched
|
|
category: quick
|
|
function:
|
|
- syrkx_batched
|
|
- syrkx_strided_batched
|
|
arguments: *syrkx_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
# gemms
|
|
|
|
- name: gemm
|
|
category: quick
|
|
function: gemm
|
|
arguments: *gemm_common_args
|
|
|
|
- name: gemm_batched
|
|
category: quick
|
|
function:
|
|
- gemm_batched
|
|
- gemm_strided_batched
|
|
arguments: *gemm_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
- name: gemm_ex
|
|
category: quick
|
|
function: gemm_ex
|
|
arguments: *gemm_ex_common_args
|
|
|
|
- name: gemm_batched_ex
|
|
category: quick
|
|
function:
|
|
- gemm_batched_ex
|
|
- gemm_strided_batched_ex
|
|
arguments: *gemm_ex_common_args
|
|
batch_count: *batch_count_range
|
|
stride_scale: [ 1 ]
|
|
backend_flags: AMD
|
|
|
|
...
|