Cmpldi vs li+cmpld

0-report.txt

Compiling cmpldi.s and running with "perf stat" the output was "195.508661159 seconds time elapsed".
Doing the same with nocmpldi.s the output was "195.614111789 seconds".

The files were compiled using gcc version 4.9.2 and the following line was used to compile:
gcc cmpldi.s -o cmpldi.o
gcc nocmpldi.s -o nocmpldi.o

Based on these results, no performance gain was detected.

cmpldi.s

        .file   "inline_branch_cmpldi.c"
        .abiversion 2
        .section        ".toc","aw"
        .section        ".text"
        .align 2
        .globl main
        .type   main, @function
main:
        std 31,-8(1)
        stdu 1,-64(1)
        mr 31,1
        lis 9,32767
        ori 9,9,32767
        sldi 9,9,7
L1:
        addi 9,9,-1
        cmpldi 9,0
        bne L1
        li 9,0
        mr 3,9
        addi 1,31,64
        ld 31,-8(1)
        blr
        .long 0
        .byte 0,0,0,0,128,1,0,1
        .size   main,.-main
        .ident  "GCC: (Ubuntu 4.9.2-10ubuntu13) 4.9.2"
        .section        .note.GNU-stack,"",@progbits

nocmpldi.s

        .file   "inline_branch_nocmpldi.c"
        .abiversion 2
        .section        ".toc","aw"
        .section        ".text"
        .align 2
        .globl main
        .type   main, @function
main:
        std 31,-8(1)
        stdu 1,-64(1)
        mr 31,1
        lis 9,32767
        ori 9,9,32767
        sldi 9,9,7
L1:
        addi 9,9,-1
        li 8, 0
        cmpld 9,8
        bne L1
        li 9,0
        mr 3,9
        addi 1,31,64
        ld 31,-8(1)
        blr
        .long 0
        .byte 0,0,0,0,128,1,0,1
        .size   main,.-main
        .ident  "GCC: (Ubuntu 4.9.2-10ubuntu13) 4.9.2"
        .section        .note.GNU-stack,"",@progbits