LoongArch: Improve cpymemsi expansion [PR109465]

We'd been generating really bad block move sequences which is recently
complained by kernel developers who tried __builtin_memcpy.  To improve
it:

1. Take the advantage of -mno-strict-align.  When it is set, set mode
   size to UNITS_PER_WORD regardless of the alignment.
2. Half the mode size when (block size) % (mode size) != 0, instead of
   falling back to ld.bu/st.b at once.
3. Limit the length of block move sequence considering the number of
   instructions, not the size of block.  When -mstrict-align is set and
   the block is not aligned, the old size limit for straight-line
   implementation (64 bytes) was definitely too large (we don't have 64
   registers anyway).

Change since v1: add a comment about the calculation of num_reg.

gcc/ChangeLog:

	PR target/109465
	* config/loongarch/loongarch-protos.h
	(loongarch_expand_block_move): Add a parameter as alignment RTX.
	* config/loongarch/loongarch.h:
	(LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER): Remove.
	(LARCH_MAX_MOVE_BYTES_STRAIGHT): Remove.
	(LARCH_MAX_MOVE_OPS_PER_LOOP_ITER): Define.
	(LARCH_MAX_MOVE_OPS_STRAIGHT): Define.
	(MOVE_RATIO): Use LARCH_MAX_MOVE_OPS_PER_LOOP_ITER instead of
	LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER.
	* config/loongarch/loongarch.cc (loongarch_expand_block_move):
	Take the alignment from the parameter, but set it to
	UNITS_PER_WORD if !TARGET_STRICT_ALIGN.  Limit the length of
	straight-line implementation with LARCH_MAX_MOVE_OPS_STRAIGHT
	instead of LARCH_MAX_MOVE_BYTES_STRAIGHT.
	(loongarch_block_move_straight): When there are left-over bytes,
	half the mode size instead of falling back to byte mode at once.
	(loongarch_block_move_loop): Limit the length of loop body with
	LARCH_MAX_MOVE_OPS_PER_LOOP_ITER instead of
	LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER.
	* config/loongarch/loongarch.md (cpymemsi): Pass the alignment
	to loongarch_expand_block_move.

gcc/testsuite/ChangeLog:

	PR target/109465
	* gcc.target/loongarch/pr109465-1.c: New test.
	* gcc.target/loongarch/pr109465-2.c: New test.
	* gcc.target/loongarch/pr109465-3.c: New test.

1 files changed, 4 insertions, 6 deletions

diff --git a/gcc/config/loongarch/loongarch.h b/gcc/config/loongarch/loongarch.h
index 277facb..a9eff6a 100644
--- a/gcc/config/loongarch/loongarch.h
+++ b/gcc/config/loongarch/loongarch.h
@@ -1062,13 +1062,13 @@ typedef struct {
 
 /* The maximum number of bytes that can be copied by one iteration of
    a cpymemsi loop; see loongarch_block_move_loop.  */
-#define LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER (UNITS_PER_WORD * 4)
+#define LARCH_MAX_MOVE_OPS_PER_LOOP_ITER 4
 
 /* The maximum number of bytes that can be copied by a straight-line
    implementation of cpymemsi; see loongarch_block_move_straight.  We want
    to make sure that any loop-based implementation will iterate at
    least twice.  */
-#define LARCH_MAX_MOVE_BYTES_STRAIGHT (LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER * 2)
+#define LARCH_MAX_MOVE_OPS_STRAIGHT (LARCH_MAX_MOVE_OPS_PER_LOOP_ITER * 2)
 
 /* The base cost of a memcpy call, for MOVE_RATIO and friends.  These
    values were determined experimentally by benchmarking with CSiBE.
@@ -1076,7 +1076,7 @@ typedef struct {
 #define LARCH_CALL_RATIO 8
 
 /* Any loop-based implementation of cpymemsi will have at least
-   LARCH_MAX_MOVE_BYTES_STRAIGHT / UNITS_PER_WORD memory-to-memory
+   LARCH_MAX_MOVE_OPS_PER_LOOP_ITER memory-to-memory
    moves, so allow individual copies of fewer elements.
 
    When cpymemsi is not available, use a value approximating
@@ -1087,9 +1087,7 @@ typedef struct {
    value of LARCH_CALL_RATIO to take that into account.  */
 
 #define MOVE_RATIO(speed) \
-  (HAVE_cpymemsi \
-   ? LARCH_MAX_MOVE_BYTES_PER_LOOP_ITER / UNITS_PER_WORD \
-   : CLEAR_RATIO (speed) / 2)
+  (HAVE_cpymemsi ? LARCH_MAX_MOVE_OPS_PER_LOOP_ITER : CLEAR_RATIO (speed) / 2)
 
 /* For CLEAR_RATIO, when optimizing for size, give a better estimate
    of the length of a memset call, but use the default otherwise.  */