Take: int f(int a) { return a < 4 ? a - 4 : 0; } int f1(int a) { int x = a - 4; if (a < 4) return x; return 0; } int f2(int a) { int t = a < 4 ? a : 4; return t - 4; } In GCC 11, we produce: f(int): mov w1, 4 cmp w0, w1 csel w0, w0, w1, le sub w0, w0, #4 ret f1(int): subs w0, w0, 4 csel w0, w0, wzr, lt ret f2(int): mov w1, 4 cmp w0, w1 csel w0, w0, w1, le sub w0, w0, #4 ret On the trunk all three give the same code gen (due to PHI-OPT being improved) but of what f and f2 used to give. All three should produce what f1 had produed instead. This is gcc.target/aarch64/subs_compare_2.c
Note clang even does produce the same code gen for all three: https://godbolt.org/z/hq357jdMd Also I put how we could fix this in https://gcc.gnu.org/pipermail/gcc-patches/2021-June/571699.html
*** Bug 100873 has been marked as a duplicate of this bug. ***
Note here is a 4th variant of the function: int f1a(int a) { int x = a - 4; return (a < 4) ? x : 0; }
The trunk branch has been updated by Richard Sandiford <rsandifo@gcc.gnu.org>: https://gcc.gnu.org/g:8e84b2b37a541b27feea69769fc314d534464ebd commit r12-7249-g8e84b2b37a541b27feea69769fc314d534464ebd Author: Richard Sandiford <richard.sandiford@arm.com> Date: Tue Feb 15 18:09:35 2022 +0000 aarch64: Fix subs_compare_2.c regression [PR100874] subs_compare_2.c tests that we can use a SUBS+CSEL sequence for: unsigned int foo (unsigned int a, unsigned int b) { unsigned int x = a - 4; if (a < 4) return x; else return 0; } As Andrew notes in the PR, this is effectively MIN (x, 4) - 4, and it is now recognised as such by phiopt. Previously it was if-converted in RTL instead. I tried to look for ways to generalise this to other situations and to other ?:-style operations, not just max and min. However, for general ?: we tend to push an outer â- CSTâ into the arms of the ?: -- at least if one of them simplifies -- so I didn't find any useful abstraction. This patch therefore adds a pattern specifically for max/min(a,cst)-cst. I'm not thrilled at having to do this, but it seems like the least worst fix in the circumstances. Also, max(a,cst)-cst for unsigned a is a useful saturating subtraction idiom and so is arguably worth its own code for that reason. gcc/ PR target/100874 * config/aarch64/aarch64-protos.h (aarch64_maxmin_plus_const): Declare. * config/aarch64/aarch64.cc (aarch64_maxmin_plus_const): New function. * config/aarch64/aarch64.md (*aarch64_minmax_plus): New pattern. gcc/testsuite/ * gcc.target/aarch64/max_plus_1.c: New test. * gcc.target/aarch64/max_plus_2.c: Likewise. * gcc.target/aarch64/max_plus_3.c: Likewise. * gcc.target/aarch64/max_plus_4.c: Likewise. * gcc.target/aarch64/max_plus_5.c: Likewise. * gcc.target/aarch64/max_plus_6.c: Likewise. * gcc.target/aarch64/max_plus_7.c: Likewise. * gcc.target/aarch64/min_plus_1.c: Likewise. * gcc.target/aarch64/min_plus_2.c: Likewise. * gcc.target/aarch64/min_plus_3.c: Likewise. * gcc.target/aarch64/min_plus_4.c: Likewise. * gcc.target/aarch64/min_plus_5.c: Likewise. * gcc.target/aarch64/min_plus_6.c: Likewise. * gcc.target/aarch64/min_plus_7.c: Likewise.
Fixed.