Google Git
Sign in
opensecura/sw/kelvin/06d10c6e58c873e3df6a480f933e7b91f16fb79f^!/.
commit
06d10c6e58c873e3df6a480f933e7b91f16fb79f
[log]
author
Cindy Liu <hcindyl@google.com>
Fri Jul 21 19:32:50 2023 -0700
committer
Cindy Liu <hcindyl@google.com>
Mon Jul 24 09:19:05 2023 -0700
tree
986320341d5ed7772d8e47f23e2c8a51e50649f6
parent
73536fb821603705ecc4c2ca935aa4b596ce4338 [diff]
Add vsliden and vslidep intrinstics to the sw header

Change-Id: I4aaf1dbb76f7aab28021120f1fd987f7e3839350

diff --git a/crt/kelvin_intrinsics.h b/crt/kelvin_intrinsics.h
index db22707..11371ad 100644
--- a/crt/kelvin_intrinsics.h
+++ b/crt/kelvin_intrinsics.h

@@ -1025,6 +1025,30 @@
 #define vmadd_w_vx_m(Vd, Vs, t)     __asm__ __volatile__(ARGS_F_A_A_A("vmadd.w.vx.m", Vd, Vs, %0) : : "r"(t))
 
 // 110 Shuffle
+#define vsliden_b_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.1.vv", Vd, Vs, Vt))
+#define vsliden_b_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_b_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.2.vv", Vd, Vs, Vt))
+#define vsliden_b_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_b_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.3.vv", Vd, Vs, Vt))
+#define vsliden_b_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_b_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.4.vv", Vd, Vs, Vt))
+#define vsliden_b_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.b.4.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_h_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.1.vv", Vd, Vs, Vt))
+#define vsliden_h_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_h_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.2.vv", Vd, Vs, Vt))
+#define vsliden_h_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_h_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.3.vv", Vd, Vs, Vt))
+#define vsliden_h_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_h_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.4.vv", Vd, Vs, Vt))
+#define vsliden_h_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.h.4.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_w_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.1.vv", Vd, Vs, Vt))
+#define vsliden_w_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_w_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.2.vv", Vd, Vs, Vt))
+#define vsliden_w_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_w_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.3.vv", Vd, Vs, Vt))
+#define vsliden_w_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vsliden_w_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.4.vv", Vd, Vs, Vt))
+#define vsliden_w_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vsliden.w.4.vx", Vd, Vs, %0) : : "r"(t))
 #define vslidevn_b_1_vv_m(Vd, Vs, Vt)__asm__ __volatile__(ARGS_F_A_A_A("vslidevn.b.1.vv.m", Vd, Vs, Vt))
 #define vslidevn_b_1_vx_m(Vd, Vs, t)__asm__ __volatile__(ARGS_F_A_A_A("vslidevn.b.1.vx.m", Vd, Vs, %0) : : "r"(t))
 #define vslidevn_b_2_vv_m(Vd, Vs, Vt)__asm__ __volatile__(ARGS_F_A_A_A("vslidevn.b.2.vv.m", Vd, Vs, Vt))
@@ -1073,6 +1097,30 @@
 #define vslidehn_w_3_vx_m(Vd, Vs, t)__asm__ __volatile__(ARGS_F_A_A_A("vslidehn.w.3.vx.m", Vd, Vs, %0) : : "r"(t))
 #define vslidehn_w_4_vv_m(Vd, Vs, Vt)__asm__ __volatile__(ARGS_F_A_A_A("vslidehn.w.4.vv.m", Vd, Vs, Vt))
 #define vslidehn_w_4_vx_m(Vd, Vs, t)__asm__ __volatile__(ARGS_F_A_A_A("vslidehn.w.4.vx.m", Vd, Vs, %0) : : "r"(t))
+#define vslidep_b_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.1.vv", Vd, Vs, Vt))
+#define vslidep_b_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_b_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.2.vv", Vd, Vs, Vt))
+#define vslidep_b_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_b_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.3.vv", Vd, Vs, Vt))
+#define vslidep_b_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_b_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.4.vv", Vd, Vs, Vt))
+#define vslidep_b_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.b.4.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_h_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.1.vv", Vd, Vs, Vt))
+#define vslidep_h_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_h_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.2.vv", Vd, Vs, Vt))
+#define vslidep_h_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_h_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.3.vv", Vd, Vs, Vt))
+#define vslidep_h_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_h_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.4.vv", Vd, Vs, Vt))
+#define vslidep_h_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.h.4.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_w_1_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.1.vv", Vd, Vs, Vt))
+#define vslidep_w_1_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.1.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_w_2_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.2.vv", Vd, Vs, Vt))
+#define vslidep_w_2_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.2.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_w_3_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.3.vv", Vd, Vs, Vt))
+#define vslidep_w_3_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.3.vx", Vd, Vs, %0) : : "r"(t))
+#define vslidep_w_4_vv(Vd, Vs, Vt)  __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.4.vv", Vd, Vs, Vt))
+#define vslidep_w_4_vx(Vd, Vs, t)   __asm__ __volatile__(ARGS_F_A_A_A("vslidep.w.4.vx", Vd, Vs, %0) : : "r"(t))
 #define vslidevp_b_1_vv_m(Vd, Vs, Vt)__asm__ __volatile__(ARGS_F_A_A_A("vslidevp.b.1.vv.m", Vd, Vs, Vt))
 #define vslidevp_b_1_vx_m(Vd, Vs, t)__asm__ __volatile__(ARGS_F_A_A_A("vslidevp.b.1.vx.m", Vd, Vs, %0) : : "r"(t))
 #define vslidevp_b_2_vv_m(Vd, Vs, Vt)__asm__ __volatile__(ARGS_F_A_A_A("vslidevp.b.2.vv.m", Vd, Vs, Vt))