[PATCH] Middle-end: Adjust decrement IV style partial vectorization COST model

[Juzhe-Zhong <juzhe.zhong@rivai.ai>]

Hi, before this patch, a simple conversion case for RVV codegen:

foo:
        ble     a2,zero,.L8
        addiw   a5,a2,-1
        li      a4,6
        bleu    a5,a4,.L6
        srliw   a3,a2,3
        slli    a3,a3,3
        add     a3,a3,a0
        mv      a5,a0
        mv      a4,a1
        vsetivli        zero,8,e16,m1,ta,ma
.L4:
        vle8.v  v2,0(a5)
        addi    a5,a5,8
        vzext.vf2       v1,v2
        vse16.v v1,0(a4)
        addi    a4,a4,16
        bne     a3,a5,.L4
        andi    a5,a2,-8
        beq     a2,a5,.L10
.L3:
        slli    a4,a5,32
        srli    a4,a4,32
        subw    a2,a2,a5
        slli    a2,a2,32
        slli    a5,a4,1
        srli    a2,a2,32
        add     a0,a0,a4
        add     a1,a1,a5
        vsetvli zero,a2,e16,m1,ta,ma
        vle8.v  v2,0(a0)
        vzext.vf2       v1,v2
        vse16.v v1,0(a1)
.L8:
        ret
.L10:
        ret
.L6:
        li      a5,0
        j       .L3

This vectorization go through first loop:

        vsetivli        zero,8,e16,m1,ta,ma
.L4:
        vle8.v  v2,0(a5)
        addi    a5,a5,8
        vzext.vf2       v1,v2
        vse16.v v1,0(a4)
        addi    a4,a4,16
        bne     a3,a5,.L4

Each iteration processes 8 elements.

For a scalable vectorization with VLEN > 128 bits CPU, it's ok when VLEN = 128.
But, as long as VLEN > 128 bits, it will waste the CPU resources. That is, e.g. VLEN = 256bits.
only half of the vector units are working and another half is idle.

After investigation, I realize that I forgot to adjust COST for SELECT_VL.
So, adjust COST for SELECT_VL styple length vectorization. We adjust COST from 3 to 2. since
after this patch:

foo:
	ble	a2,zero,.L5
.L3:
	vsetvli	a5,a2,e16,m1,ta,ma     -----> SELECT_VL cost.
	vle8.v	v2,0(a0)
	slli	a4,a5,1                -----> additional shift of outcome SELECT_VL for memory address calculation.
	vzext.vf2	v1,v2
	sub	a2,a2,a5
	vse16.v	v1,0(a1)
	add	a0,a0,a5
	add	a1,a1,a4
	bne	a2,zero,.L3
.L5:
	ret

This patch is a simple fix that I previous forgot.

Ok for trunk ?

If not, I am going to adjust cost in backend cost model.

	PR target/111317

gcc/ChangeLog:

	* tree-vect-loop.cc (vect_estimate_min_profitable_iters): Adjust for COST for decrement IV.

gcc/testsuite/ChangeLog:

	* gcc.dg/vect/costmodel/riscv/rvv/pr111317.c: New test.

---
 .../gcc.dg/vect/costmodel/riscv/rvv/pr111317.c  | 12 ++++++++++++
 gcc/tree-vect-loop.cc                           | 17 ++++++++++++++---
 2 files changed, 26 insertions(+), 3 deletions(-)
 create mode 100644 gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/pr111317.c

diff --git a/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/pr111317.c b/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/pr111317.c
new file mode 100644
index 00000000000..d4bea242a9a
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/vect/costmodel/riscv/rvv/pr111317.c
@@ -0,0 +1,12 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3 -ftree-vectorize --param=riscv-autovec-lmul=m1" } */
+
+void
+foo (char *__restrict a, short *__restrict b, int n)
+{
+  for (int i = 0; i < n; i++)
+    b[i] = (short) a[i];
+}
+
+/* { dg-final { scan-assembler-times {vsetvli\s+[a-x0-9]+,\s*[a-x0-9]+,\s*e16,\s*m1,\s*t[au],\s*m[au]} 1 } } */
+/* { dg-final { scan-assembler-times {vsetvli} 1 } } */
diff --git a/gcc/tree-vect-loop.cc b/gcc/tree-vect-loop.cc
index 6261cd1be1d..19e38b8637b 100644
--- a/gcc/tree-vect-loop.cc
+++ b/gcc/tree-vect-loop.cc
@@ -4870,10 +4870,21 @@ vect_estimate_min_profitable_iters (loop_vec_info loop_vinfo,
 	    if (partial_load_store_bias != 0)
 	      body_stmts += 1;
 
-	    /* Each may need two MINs and one MINUS to update lengths in body
-	       for next iteration.  */
+	    unsigned int length_update_cost = 0;
+	    if (LOOP_VINFO_USING_DECREMENTING_IV_P (loop_vinfo))
+	      /* For decrement IV style, we use a single SELECT_VL since
+		 beginning to calculate the number of elements need to be
+		 processed in current iteration, and a SHIFT operation to
+		 compute the next memory address instead of adding vectorization
+		 factor.  */
+	      length_update_cost = 2;
+	    else
+	      /* For increment IV stype, Each may need two MINs and one MINUS to
+		 update lengths in body for next iteration.  */
+	      length_update_cost = 3;
+
 	    if (need_iterate_p)
-	      body_stmts += 3 * num_vectors;
+	      body_stmts += length_update_cost * num_vectors;
 	  }
 
       (void) add_stmt_cost (target_cost_data, prologue_stmts,
-- 
2.36.3