search for: shlx

I suspect that this is because the mask in your example is the result of a variable shift, which (a) has it’s own performance and flags hazards pre-SHLX and (b) requires additional µops to do with TEST. I expect that ICC is putting a dummy TEST or XOR ahead of the BT to break the false flags dependency, as well. If the mask were constant, I expect ICC would generate TEST instead (but I don’t have it handy to check). – Steve > On Jan 23, 2015...

On Thu Jan 22 2015 at 3:32:53 PM Chris Sears <chris.sears at gmail.com> wrote: > The status quo is: > > a) 40b REX+BT instruction for the 64b case > b) 48b TEST for the 32b case > c) unless it's small TEST > > > You are currently paying a 16b penalty for TEST vs BT in the 32b case. > That may be worth testing the -Os flag. > You'll want -Oz here, Os

...f 83 00 04 00 00 e9 17 ff ff ff 41 54 55 53 48 89 fb <8b> 47 28 85 c0 0f 84 cf 00 00 00 48 8b bb c0 01 00 00 31 f6 4c 8b All code ========    0:    89 d9                    mov    %ebx,%ecx    2:    48 c7 c6 50 04 e5 c0     mov    $0xffffffffc0e50450,%rsi    9:    c4 42 79 f7 c0           shlx   %eax,%r8d,%r8d    e:    bd f0 ff ff ff           mov    $0xfffffff0,%ebp   13:    e8 42 d5 7a c6           callq  0xffffffffc67ad55a   18:    ff 83 00 04 00 00        incl   0x400(%rbx)   1e:    e9 17 ff ff ff           jmpq   0xffffffffffffff3a   23:    41 54                    push   %r12...