aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorAndrew Pinski <apinski@marvell.com>2023-04-28 05:22:34 +0000
committerAndrew Pinski <apinski@marvell.com>2023-05-02 14:10:19 -0700
commit7cc33d12dca84befde69259a17fab8224e2e4025 (patch)
treec220f6e80cc25d3c66e6891f38516a57b680c7cc
parentf9861511a1fa0f9e386f3f7bcee84b6e3ca3c579 (diff)
downloadgcc-7cc33d12dca84befde69259a17fab8224e2e4025.zip
gcc-7cc33d12dca84befde69259a17fab8224e2e4025.tar.gz
gcc-7cc33d12dca84befde69259a17fab8224e2e4025.tar.bz2
target: [PR109657] (a ? -1 : 0) | b could be optimized better for aarch64
There is no canonical form for this case defined. So the aarch64 backend needs a pattern to match both of these forms. The forms are: (set (reg/i:SI 0 x0) (if_then_else:SI (eq (reg:CC 66 cc) (const_int 0 [0])) (reg:SI 97) (const_int -1 [0xffffffffffffffff]))) and (set (reg/i:SI 0 x0) (ior:SI (neg:SI (ne:SI (reg:CC 66 cc) (const_int 0 [0]))) (reg:SI 102))) Currently the aarch64 backend matches the first form so this patch adds a insn_and_split to match the second form and convert it to the first form. OK? Bootstrapped and tested on aarch64-linux-gnu with no regressions PR target/109657 gcc/ChangeLog: * config/aarch64/aarch64.md (*cmov<mode>_insn_m1): New insn_and_split pattern. gcc/testsuite/ChangeLog: * gcc.target/aarch64/csinv-2.c: New test.
-rw-r--r--gcc/config/aarch64/aarch64.md21
-rw-r--r--gcc/testsuite/gcc.target/aarch64/csinv-2.c26
2 files changed, 47 insertions, 0 deletions
diff --git a/gcc/config/aarch64/aarch64.md b/gcc/config/aarch64/aarch64.md
index e1a2b26..286f044 100644
--- a/gcc/config/aarch64/aarch64.md
+++ b/gcc/config/aarch64/aarch64.md
@@ -4194,6 +4194,27 @@
[(set_attr "type" "csel, csel, csel, csel, csel, mov_imm, mov_imm")]
)
+;; There are two canonical forms for `cmp ? -1 : a`.
+;; This is the second form and is here to help combine.
+;; Support `-(cmp) | a` into `cmp ? -1 : a` to be canonical in the backend.
+(define_insn_and_split "*cmov<mode>_insn_m1"
+ [(set (match_operand:GPI 0 "register_operand" "=r")
+ (ior:GPI
+ (neg:GPI
+ (match_operator:GPI 1 "aarch64_comparison_operator"
+ [(match_operand 2 "cc_register" "") (const_int 0)]))
+ (match_operand 3 "register_operand" "r")))]
+ ""
+ "#"
+ "&& true"
+ [(set (match_dup 0)
+ (if_then_else:GPI (match_dup 1)
+ (const_int -1)
+ (match_dup 3)))]
+ {}
+ [(set_attr "type" "csel")]
+)
+
(define_insn "*cmovdi_insn_uxtw"
[(set (match_operand:DI 0 "register_operand" "=r")
(if_then_else:DI
diff --git a/gcc/testsuite/gcc.target/aarch64/csinv-2.c b/gcc/testsuite/gcc.target/aarch64/csinv-2.c
new file mode 100644
index 0000000..89132ac
--- /dev/null
+++ b/gcc/testsuite/gcc.target/aarch64/csinv-2.c
@@ -0,0 +1,26 @@
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+/* PR target/109657: (a ? -1 : 0) | b could be better */
+
+/* Both functions should have the same assembly of:
+ cmp w1, 0
+ csinv w0, w0, wzr, eq
+
+ We should not get:
+ cmp w1, 0
+ csetm w1, ne
+ orr w0, w1, w0
+ */
+/* { dg-final { scan-assembler-times "csinv\tw\[0-9\]" 2 } } */
+/* { dg-final { scan-assembler-not "csetm\tw\[0-9\]" } } */
+unsigned b(unsigned a, unsigned b)
+{
+ if(b)
+ return -1;
+ return a;
+}
+unsigned b1(unsigned a, unsigned b)
+{
+ unsigned t = b ? -1 : 0;
+ return a | t;
+}