lavu/riscv: require B or zba explicitly

libavutil/riscv/fixed_dsp_rvv.S

@@ -20,7 +20,7 @@
 
 #include "asm.S"
 
-func ff_vector_fmul_window_scaled_rvv, zve64x
+func ff_vector_fmul_window_scaled_rvv, zve64x, zba
         csrwi   vxrm, 0
         vsetvli t0, zero, e16, m1, ta, ma
         sh2add  a2, a4, a2
@@ -68,7 +68,7 @@ func ff_vector_fmul_window_scaled_rvv, zve64x
         ret
 endfunc
 
-func ff_vector_fmul_window_fixed_rvv, zve64x
+func ff_vector_fmul_window_fixed_rvv, zve64x, zba
         csrwi   vxrm, 0
         vsetvli t0, zero, e16, m1, ta, ma
         sh2add  a2, a4, a2
@@ -112,7 +112,7 @@ func ff_vector_fmul_window_fixed_rvv, zve64x
         ret
 endfunc
 
-func ff_vector_fmul_fixed_rvv, zve32x
+func ff_vector_fmul_fixed_rvv, zve32x, zba
         csrwi   vxrm, 0
 1:
         vsetvli t0, a3, e32, m4, ta, ma
@@ -129,7 +129,7 @@ func ff_vector_fmul_fixed_rvv, zve32x
         ret
 endfunc
 
-func ff_vector_fmul_reverse_fixed_rvv, zve32x
+func ff_vector_fmul_reverse_fixed_rvv, zve32x, zba
         csrwi   vxrm, 0
         // e16/m4 and e32/m8 are possible but slow the gathers down.
         vsetvli t0, zero, e16, m1, ta, ma
@@ -155,7 +155,7 @@ func ff_vector_fmul_reverse_fixed_rvv, zve32x
         ret
 endfunc
 
-func ff_vector_fmul_add_fixed_rvv, zve32x
+func ff_vector_fmul_add_fixed_rvv, zve32x, zba
         csrwi   vxrm, 0
 1:
         vsetvli t0, a4, e32, m8, ta, ma
@@ -175,7 +175,7 @@ func ff_vector_fmul_add_fixed_rvv, zve32x
         ret
 endfunc
 
-func ff_scalarproduct_fixed_rvv, zve64x
+func ff_scalarproduct_fixed_rvv, zve64x, zba
         li      t1, 1 << 30
         vsetvli t0, zero, e64, m8, ta, ma
         vmv.v.x v8, zero
@@ -198,7 +198,7 @@ func ff_scalarproduct_fixed_rvv, zve64x
 endfunc
 
 // (a0) = (a0) + (a1), (a1) = (a0) - (a1) [0..a2-1]
-func ff_butterflies_fixed_rvv, zve32x
+func ff_butterflies_fixed_rvv, zve32x, zba
 1:
         vsetvli t0, a2, e32, m4, ta, ma
         vle32.v v16, (a0)

libavutil/riscv/float_dsp_rvv.S

@@ -21,7 +21,7 @@
 #include "asm.S"
 
 // (a0) = (a1) * (a2) [0..a3-1]
-func ff_vector_fmul_rvv, zve32f
+func ff_vector_fmul_rvv, zve32f, zba
 1:
         vsetvli  t0, a3, e32, m8, ta, ma
         vle32.v  v16, (a1)
@@ -38,7 +38,7 @@ func ff_vector_fmul_rvv, zve32f
 endfunc
 
 // (a0) += (a1) * fa0 [0..a2-1]
-func ff_vector_fmac_scalar_rvv, zve32f
+func ff_vector_fmac_scalar_rvv, zve32f, zba
 NOHWF   fmv.w.x   fa0, a2
 NOHWF   mv        a2, a3
 1:
@@ -57,7 +57,7 @@ NOHWF   mv        a2, a3
 endfunc
 
 // (a0) = (a1) * fa0 [0..a2-1]
-func ff_vector_fmul_scalar_rvv, zve32f
+func ff_vector_fmul_scalar_rvv, zve32f, zba
 NOHWF   fmv.w.x  fa0, a2
 NOHWF   mv       a2, a3
 1:
@@ -73,7 +73,7 @@ NOHWF   mv       a2, a3
         ret
 endfunc
 
-func ff_vector_fmul_window_rvv, zve32f
+func ff_vector_fmul_window_rvv, zve32f, zba
         // a0: dst, a1: src0, a2: src1, a3: window, a4: length
         // e16/m2 and e32/m4 are possible but slower due to gather.
         vsetvli    t0, zero, e16, m1, ta, ma
@@ -113,7 +113,7 @@ func ff_vector_fmul_window_rvv, zve32f
 endfunc
 
 // (a0) = (a1) * (a2) + (a3) [0..a4-1]
-func ff_vector_fmul_add_rvv, zve32f
+func ff_vector_fmul_add_rvv, zve32f, zba
 1:
         vsetvli   t0, a4, e32, m8, ta, ma
         vle32.v   v8, (a1)
@@ -133,7 +133,7 @@ endfunc
 
 // TODO factor vrsub, separate last iteration?
 // (a0) = (a1) * reverse(a2) [0..a3-1]
-func ff_vector_fmul_reverse_rvv, zve32f
+func ff_vector_fmul_reverse_rvv, zve32f, zba
         // e16/m4 and e32/m8 are possible but slower due to gather.
         vsetvli  t0, zero, e16, m1, ta, ma
         sh2add   a2, a3, a2
@@ -159,7 +159,7 @@ func ff_vector_fmul_reverse_rvv, zve32f
 endfunc
 
 // (a0) = (a0) + (a1), (a1) = (a0) - (a1) [0..a2-1]
-func ff_butterflies_float_rvv, zve32f
+func ff_butterflies_float_rvv, zve32f, zba
 1:
         vsetvli  t0, a2, e32, m8, ta, ma
         vle32.v  v16, (a0)
@@ -177,7 +177,7 @@ func ff_butterflies_float_rvv, zve32f
 endfunc
 
 // a0 = (a0).(a1) [0..a2-1]
-func ff_scalarproduct_float_rvv, zve32f
+func ff_scalarproduct_float_rvv, zve32f, zba
         vsetvli      t0, zero, e32, m8, ta, ma
         vmv.v.x      v8, zero
         vmv.s.x      v0, zero
@@ -199,7 +199,7 @@ NOHWF   fmv.x.w  a0, fa0
 endfunc
 
 // (a0) = (a1) * (a2) [0..a3-1]
-func ff_vector_dmul_rvv, zve64d
+func ff_vector_dmul_rvv, zve64d, zba
 1:
         vsetvli  t0, a3, e64, m8, ta, ma
         vle64.v  v16, (a1)
@@ -216,7 +216,7 @@ func ff_vector_dmul_rvv, zve64d
 endfunc
 
 // (a0) += (a1) * fa0 [0..a2-1]
-func ff_vector_dmac_scalar_rvv, zve64d
+func ff_vector_dmac_scalar_rvv, zve64d, zba
 NOHWD   fmv.d.x   fa0, a2
 NOHWD   mv        a2, a3
 1:
@@ -234,7 +234,7 @@ NOHWD   mv        a2, a3
 endfunc
 
 // (a0) = (a1) * fa0 [0..a2-1]
-func ff_vector_dmul_scalar_rvv, zve64d
+func ff_vector_dmul_scalar_rvv, zve64d, zba
 NOHWD   fmv.d.x  fa0, a2
 NOHWD   mv       a2, a3
 1:
@@ -250,7 +250,7 @@ NOHWD   mv       a2, a3
         ret
 endfunc
 
-func ff_scalarproduct_double_rvv, zve64f
+func ff_scalarproduct_double_rvv, zve64f, zba
         vsetvli      t0, zero, e64, m8, ta, ma
         vmv.v.x      v8, zero
         vmv.s.x      v0, zero