public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug ipa/100801] New: Aggressive loop optimizations cause incorrect warning
@ 2021-05-27 14:03 jl_gccbugs at conductive dot de
  2021-05-28 10:54 ` [Bug tree-optimization/100801] " rguenth at gcc dot gnu.org
                   ` (8 more replies)
  0 siblings, 9 replies; 10+ messages in thread
From: jl_gccbugs at conductive dot de @ 2021-05-27 14:03 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100801

            Bug ID: 100801
           Summary: Aggressive loop optimizations cause incorrect warning
           Product: gcc
           Version: 11.1.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: ipa
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jl_gccbugs at conductive dot de
                CC: marxin at gcc dot gnu.org
  Target Milestone: ---

The following warning is triggered

> $ gcc-11 -c constproploopopt.c -O2 -Wall -mavx -g
> constproploopopt.c: In function ‘test’:
> constproploopopt.c:22:18: warning: iteration 4611686018427387903 invokes undefined behavior [-Waggressive-loop-optimizations]
>    22 |     dest[i] = src[i];
>       |                  ^
> constproploopopt.c:21:12: note: within this loop
>    21 |   for (; i < count; ++i) {  // handle residual elements
>       |          ~~^~~~~~~

by this (minimal) code:

> #include <stdint.h>
> #include <stdio.h>
> #if defined(_MSC_VER)
> #include <intrin.h>
> #else
> #include <x86intrin.h>
> #endif
> 
> void copy_32_unaligned(uint32_t* dest, const uint32_t* src, size_t count) {
>   // invariant/nop
>   __m128i shufmask =  _mm_set_epi8(15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
> 
>   size_t i;
>   for (i = 0; i + 4 <= count; i += 4) {
>     __m128i input = _mm_loadu_si128((const __m128i*)(&src[i]));
>     __m128i output = input;
>     // no warning without the shuffle:
>     output = _mm_shuffle_epi8(input, shufmask);
>     _mm_storeu_si128((__m128i*)(&dest[i]), output);
>   }
>   for (; i < count; ++i) {  // handle residual elements
>     dest[i] = src[i];
>   }
> }
> 
> void test(uint32_t* buf1, uint32_t* buf2) {
>     copy_32_unaligned(buf2, buf1,
>                       // multiples of 4 and greater or equal then 12 trigger it:
>                       12);
> }

>From objdump output I believe the generated code is correct though. The warning
seems to be incorrect in this context, especially since the "residual" loop
should be skipped for count=n*4 anyways.

^ permalink raw reply	[flat|nested] 10+ messages in thread
end of thread, other threads:[~2024-03-11  3:42 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-27 14:03 [Bug ipa/100801] New: Aggressive loop optimizations cause incorrect warning jl_gccbugs at conductive dot de
2021-05-28 10:54 ` [Bug tree-optimization/100801] " rguenth at gcc dot gnu.org
2021-05-28 12:45 ` jl_gccbugs at conductive dot de
2021-07-31  0:41 ` matthijsvanduin at gmail dot com
2021-12-20  0:55 ` [Bug tree-optimization/100801] [9/10/11/12 Regression] " pinskia at gcc dot gnu.org
2022-01-05 14:29 ` rguenth at gcc dot gnu.org
2022-05-27  9:45 ` [Bug tree-optimization/100801] [10/11/12/13 " rguenth at gcc dot gnu.org
2022-06-28 10:45 ` jakub at gcc dot gnu.org
2023-07-07 10:40 ` [Bug tree-optimization/100801] [11/12/13/14 " rguenth at gcc dot gnu.org
2024-03-11  3:42 ` [Bug tree-optimization/100801] [11/12 " law at gcc dot gnu.org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).