[llvm-dev] Loop invariant code motion optimization does not happen for the input code

[Sudakshina Dutta via llvm-dev]

Thanks a lot for your help.

On Sun, 2 May 2021, 12:38 Craig Topper, <craig.topper at gmail.com> wrote:

> Does your function in prog1.ll have the optnone attribute on it? If so try
> adding "-Xclang -disable-O0-optnone" to your clang command line
>
> ~Craig
>
>
> On Sat, May 1, 2021 at 11:35 PM Sudakshina Dutta via llvm-dev <
> llvm-dev at lists.llvm.org> wrote:
>
>> Dear all,
>>
>> Greetings. I have been trying to have a loop invariant code motion on the
>> LLVM IR (.ll) file which has been generated from the following .c file.
>>
>> int main()
>> {
>>     int a = 5, b = 20, c = 50, d = 1, x;
>>     while(a < b)
>>     {
>>         x = d + 1;
>>         x = d + 2;
>>         c = c + 5;
>>         a = a + 1;
>>     }
>> }
>> Both the first two statements should qualify for loop invariant code
>> motion. However, loop invariant code does not move out of the loop in the
>> generated .ll file. Note that I have applied the following commands.
>>
>> clang -S -emit-llvm prog1.c  -o prog1.ll
>> opt -licm -S prog1.ll -o prog1-opt3.ll -opt-bisect-limit=300
>>
>> Can you please help in this regard ?
>>
>> Thanks and regards,
>> Sudakshina
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20210502/a238eb90/attachment.html>