https://bugs.llvm.org/show_bug.cgi?id=47053
Bug ID: 47053
Summary: Missed opportunity to optimize pointer range checks
Product: libraries
Version: trunk
Hardware: PC
OS: Linux
Status: NEW
Severity: enhancement
Priority: P
Component: Common Code Generator Code
Assignee: unassignedb...@nondot.org
Reporter: david.bolvan...@gmail.com
CC: llvm-bugs@lists.llvm.org
bool f1 (char *a, char *b)
{
return (a + 16 <= b) || (b + 16 <= a);
}
Clang:
f1(char*, char*): # @f1(char*, char*)
lea rax, [rdi + 16]
cmp rax, rsi
setbe cl
add rsi, 16
cmp rsi, rdi
setbe al
or al, cl
ret
GCC:
f1(char*, char*):
add rdi, 15
sub rdi, rsi
cmp rdi, 30
seta al
ret
GCC transforms
A + size <= B || B + size <= A
to
(size_t) (A + (size - 1) - B) > (size - 1) * 2
Could be useful to optimize better RT checks from vectorizer.
Godbolt: https://godbolt.org/z/9jsKa4
--
You are receiving this mail because:
You are on the CC list for the bug._______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
