LOADDUP variant of c2 microkernel
- instead of 1 LD1 and 4 DUP, use 1 LD4R
PiperOrigin-RevId: 410282504
diff --git a/test/qs8-igemm-minmax-fp32.yaml b/test/qs8-igemm-minmax-fp32.yaml
index c486df6..174e5a0 100644
--- a/test/qs8-igemm-minmax-fp32.yaml
+++ b/test/qs8-igemm-minmax-fp32.yaml
@@ -3,6 +3,18 @@
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.
+- name: xnn_qs8_igemm_minmax_fp32_ukernel_1x8c2__neon_mlal_ld4r
+ init: xnn_init_qs8_conv_minmax_fp32_neon_params
+ k-block: 16
+- name: xnn_qs8_igemm_minmax_fp32_ukernel_2x8c2__neon_mlal_ld4r
+ init: xnn_init_qs8_conv_minmax_fp32_neon_params
+ k-block: 16
+- name: xnn_qs8_igemm_minmax_fp32_ukernel_1x8c2__neonv8_mlal_ld4r
+ init: xnn_init_qs8_conv_minmax_fp32_neonv8_params
+ k-block: 16
+- name: xnn_qs8_igemm_minmax_fp32_ukernel_2x8c2__neonv8_mlal_ld4r
+ init: xnn_init_qs8_conv_minmax_fp32_neonv8_params
+ k-block: 16
- name: xnn_qs8_igemm_minmax_fp32_ukernel_1x8c4s2__neon_mlal
init: xnn_init_qs8_conv_minmax_fp32_neon_params
k-block: 16