https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106458
--- Comment #15 from dave.anglin at bell dot net ---
On 2022-08-10 9:30 a.m., rguenth at gcc dot gnu.org wrote:
> hen I try with a cc1 cross I see
>
>> ./cc1 -quiet t.i -fpreprocessed -O2 -g -std=gnu11 -fgnu89-inline
>> -fmerge-all-constants -frounding-math -fno-stack-protector -fno-common
>> -fmath-errno -fno-pie -fopt-info-vec
> and nothing actually vectorized. Besides the PRE there seems to be a missed
> DSE at the end (when vectorization is enabled as after the change):
>
> # DEBUG BEGIN_STMT
> memset (&buf.__space.__c, 64, 1024);
> # DEBUG BEGIN_STMT
> + sizes[0] = 16;
> sizes[1] = 1024;
> sizes[2] = 1040;
>
> otherwise I see nothing suspicious in differences in .optimized when
> comparing -fdump-tree-optimized with/without -fno-tree-vectorize.
>
I'm not sure what's going on with the sizes[0] statement but the main
difference in .optimize with and without pre
seems to be in the handling of the old_length variable in do_test:
Removing basic block 86
int do_test ()
{
- unsigned int ivtmp.23;
- size_t old_length;
+ unsigned int ivtmp.28;
void * r;
void * c;
size_t l;
