public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r13-2580] xtensa: constantsynth: Add new 3-insns synthesis pattern
@ 2022-09-10 18:25 Max Filippov
0 siblings, 0 replies; only message in thread
From: Max Filippov @ 2022-09-10 18:25 UTC (permalink / raw)
To: gcc-cvs
https://gcc.gnu.org/g:75e5cc9c3aba943819c284902b3792f7150749cf
commit r13-2580-g75e5cc9c3aba943819c284902b3792f7150749cf
Author: Takayuki 'January June' Suwa <jjsuwa_sys3175@yahoo.co.jp>
Date: Sat Sep 10 18:29:45 2022 +0900
xtensa: constantsynth: Add new 3-insns synthesis pattern
This patch adds a new 3-instructions constant synthesis pattern:
- A value that can fit into a signed 12-bit after a number of either bitwise
left or right rotations:
=> "MOVI(.N) Ax, simm12" + "SSAI (1 ... 11) or (21 ... 31)"
+ "SRC Ax, Ax, Ax"
gcc/ChangeLog:
* config/xtensa/xtensa.cc (xtensa_constantsynth):
Add new pattern for the abovementioned case.
gcc/testsuite/ChangeLog:
* gcc.target/xtensa/constsynth_3insns.c (test_4):
Add new test function.
Diff:
---
gcc/config/xtensa/xtensa.cc | 31 ++++++++++++++++++++++
.../gcc.target/xtensa/constsynth_3insns.c | 11 ++++++++
2 files changed, 42 insertions(+)
diff --git a/gcc/config/xtensa/xtensa.cc b/gcc/config/xtensa/xtensa.cc
index 0f586b09dfb..ac52c015a94 100644
--- a/gcc/config/xtensa/xtensa.cc
+++ b/gcc/config/xtensa/xtensa.cc
@@ -1142,6 +1142,37 @@ xtensa_constantsynth (rtx dst, HOST_WIDE_INT srcval)
xtensa_constantsynth_rtx_ADDSUBX,
divisor))
return 1;
+
+ /* loading simm12 followed by left/right bitwise rotation:
+ MOVI + SSAI + SRC. */
+ if ((srcval & 0x001FF800) == 0
+ || (srcval & 0x001FF800) == 0x001FF800)
+ {
+ int32_t v;
+
+ for (shift = 1; shift < 12; ++shift)
+ {
+ v = (int32_t)(((uint32_t)srcval >> shift)
+ | ((uint32_t)srcval << (32 - shift)));
+ if (xtensa_simm12b(v))
+ {
+ emit_move_insn (dst, GEN_INT (v));
+ emit_insn (gen_rotlsi3 (dst, dst, GEN_INT (shift)));
+ return 1;
+ }
+ }
+ for (shift = 1; shift < 12; ++shift)
+ {
+ v = (int32_t)(((uint32_t)srcval << shift)
+ | ((uint32_t)srcval >> (32 - shift)));
+ if (xtensa_simm12b(v))
+ {
+ emit_move_insn (dst, GEN_INT (v));
+ emit_insn (gen_rotrsi3 (dst, dst, GEN_INT (shift)));
+ return 1;
+ }
+ }
+ }
}
return 0;
diff --git a/gcc/testsuite/gcc.target/xtensa/constsynth_3insns.c b/gcc/testsuite/gcc.target/xtensa/constsynth_3insns.c
index f3c4a1c7c15..831288c7ddd 100644
--- a/gcc/testsuite/gcc.target/xtensa/constsynth_3insns.c
+++ b/gcc/testsuite/gcc.target/xtensa/constsynth_3insns.c
@@ -21,4 +21,15 @@ void test_3(int *p)
*p = 192437;
}
+struct foo
+{
+ unsigned int b : 10;
+ unsigned int g : 11;
+ unsigned int r : 11;
+};
+void test_4(struct foo *p, unsigned int v)
+{
+ p->g = v;
+}
+
/* { dg-final { scan-assembler-not "l32r" } } */
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2022-09-10 18:25 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-09-10 18:25 [gcc r13-2580] xtensa: constantsynth: Add new 3-insns synthesis pattern Max Filippov
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).