riscv: memcpy_noalignment: Fold SZREG/BLOCK_SIZE alignment to single andi

Simplify the alignment steps for SZREG and BLOCK_SIZE multiples. The previous
three-instruction sequences

    addi    a7, a2, -SZREG
    andi    a7, a7, -SZREG
    addi    a7, a7, SZREG

and

    addi    a7, a2, -BLOCK_SIZE
    andi    a7, a7, -BLOCK_SIZE
    addi    a7, a7, BLOCK_SIZE

are equivalent to a single

    andi    a7, a2, -SZREG
    andi    a7, a2, -BLOCK_SIZE

because SZREG and BLOCK_SIZE are powers of two in this context, making the
surrounding addi steps cancel out. Folding to one instruction reduces code
size with identical semantics.

No functional change.

sysdeps/riscv/multiarch/memcpy_noalignment.S: Remove redundant addi around
alignment; keep a single andi for SZREG/BLOCK_SIZE rounding.

Signed-off-by: Yao Zihong <zihong.plct@isrc.iscas.ac.cn>
Reviewed-by: Peter Bergner <bergner@tenstorrent.com>

diff --git a/sysdeps/riscv/multiarch/memcpy_noalignment.S b/sysdeps/riscv/multiarch/memcpy_noalignment.S
index db8654fb594759affceacc2f366fb295e89077ac..6917fc435b009ce52b1245dc316f99453f5487e4 100644 (file)
--- a/sysdeps/riscv/multiarch/memcpy_noalignment.S
+++ b/sysdeps/riscv/multiarch/memcpy_noalignment.S
@@ -57,9 +57,7 @@ ENTRY (__memcpy_noalignment)
        add     a5, a0, a4
        add     a1, a1, a4
        bleu    a2, a3, L(word_copy_adjust)
-       addi    a7, a2, -BLOCK_SIZE
-       andi    a7, a7, -BLOCK_SIZE
-       addi    a7, a7, BLOCK_SIZE
+       andi    a7, a2, -BLOCK_SIZE
        add     a3, a5, a7
        mv      a4, a1
 L(block_copy):
@@ -106,9 +104,7 @@ L(word_copy):
        li      a5, SZREG-1
        /* if LEN < SZREG jump to tail handling.  */
        bleu    a2, a5, L(tail_adjust)
-       addi    a7, a2, -SZREG
-       andi    a7, a7, -SZREG
-       addi    a7, a7, SZREG
+       andi    a7, a2, -SZREG
        add     a6, a3, a7
        mv      a5, a1
 L(word_copy_loop):