public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/98880] New: [10/11 Regression] assignment unexpectedly optimized away with -O
@ 2021-01-29  9:14 gcc at jochen dot sprickerhof.de
  2021-01-29  9:18 ` [Bug c/98880] " jakub at gcc dot gnu.org
                   ` (2 more replies)
  0 siblings, 3 replies; 4+ messages in thread
From: gcc at jochen dot sprickerhof.de @ 2021-01-29  9:14 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98880

            Bug ID: 98880
           Summary: [10/11 Regression] assignment unexpectedly optimized
                    away with -O
           Product: gcc
           Version: 10.2.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gcc at jochen dot sprickerhof.de
  Target Milestone: ---

The following code:

#include <assert.h>
#include <unistd.h>

struct sh_lq {
  ssize_t slh_first;
};

struct sh_le {
  ssize_t sle_next;
};

static void sh_l_discard(struct sh_lq *list) {
  struct sh_le *ele;

  ele = (struct sh_le *)((char *)list + list->slh_first);
  assert(ele->sle_next == -1);
}

int main() {
  struct sh_lq list;
  struct sh_le ele;

  ele.sle_next = -1;
  list.slh_first = (ssize_t)((char *)&ele - (char *)&list);

  sh_l_discard(&list);

  return 0;
}

compiled with gcc-10 -O fails on the assert(). It works fine with gcc-9 -O and
with gcc-10 without -O.

Tested with gcc -v:
Using built-in specs.
COLLECT_GCC=/usr/bin/gcc
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/10/lto-wrapper
OFFLOAD_TARGET_NAMES=nvptx-none:amdgcn-amdhsa:hsa
OFFLOAD_TARGET_DEFAULT=1
Target: x86_64-linux-gnu
Configured with: ../src/configure -v --with-pkgversion='Debian 10.2.1-6'
--with-bugurl=file:///usr/share/doc/gcc-10/README.Bugs
--enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --prefix=/usr
--with-gcc-major-version-only --program-suffix=-10
--program-prefix=x86_64-linux-gnu- --enable-shared --enable-linker-build-id
--libexecdir=/usr/lib --without-included-gettext --enable-threads=posix
--libdir=/usr/lib --enable-nls --enable-bootstrap --enable-clocale=gnu
--enable-libstdcxx-debug --enable-libstdcxx-time=yes
--with-default-libstdcxx-abi=new --enable-gnu-unique-object
--disable-vtable-verify --enable-plugin --enable-default-pie --with-system-zlib
--enable-libphobos-checking=release --with-target-system-zlib=auto
--enable-objc-gc=auto --enable-multiarch --disable-werror --with-arch-32=i686
--with-abi=m64 --with-multilib-list=m32,m64,mx32 --enable-multilib
--with-tune=generic
--enable-offload-targets=nvptx-none=/build/gcc-10-Km9U7s/gcc-10-10.2.1/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-10-Km9U7s/gcc-10-10.2.1/debian/tmp-gcn/usr,hsa
--without-cuda-driver --enable-checking=release --build=x86_64-linux-gnu
--host=x86_64-linux-gnu --target=x86_64-linux-gnu
--with-build-config=bootstrap-lto-lean --enable-link-mutex
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 10.2.1 20210110 (Debian 10.2.1-6)

Also tried with godbot: https://godbolt.org/z/xrKPj8 (without the static to
make the problem more visible)

This is extracted from:
https://sources.debian.org/src/db5.3/5.3.28+dfsg1-0.6/src/dbinc/shqueue.h/

^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug c/98880] [10/11 Regression] assignment unexpectedly optimized away with -O
  2021-01-29  9:14 [Bug c/98880] New: [10/11 Regression] assignment unexpectedly optimized away with -O gcc at jochen dot sprickerhof.de
@ 2021-01-29  9:18 ` jakub at gcc dot gnu.org
  2021-01-29  9:37 ` gcc at jochen dot sprickerhof.de
  2021-01-29 16:59 ` msebor at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: jakub at gcc dot gnu.org @ 2021-01-29  9:18 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98880

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
This is invalid code.
In both C and C++ pointer subtraction is only defined if both pointers point
into the same object (including the end of it).

^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug c/98880] [10/11 Regression] assignment unexpectedly optimized away with -O
  2021-01-29  9:14 [Bug c/98880] New: [10/11 Regression] assignment unexpectedly optimized away with -O gcc at jochen dot sprickerhof.de
  2021-01-29  9:18 ` [Bug c/98880] " jakub at gcc dot gnu.org
@ 2021-01-29  9:37 ` gcc at jochen dot sprickerhof.de
  2021-01-29 16:59 ` msebor at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: gcc at jochen dot sprickerhof.de @ 2021-01-29  9:37 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98880

Jochen Sprickerhof <gcc at jochen dot sprickerhof.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |INVALID
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #2 from Jochen Sprickerhof <gcc at jochen dot sprickerhof.de> ---
Thanks for the quick reply! Closing.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug c/98880] [10/11 Regression] assignment unexpectedly optimized away with -O
  2021-01-29  9:14 [Bug c/98880] New: [10/11 Regression] assignment unexpectedly optimized away with -O gcc at jochen dot sprickerhof.de
  2021-01-29  9:18 ` [Bug c/98880] " jakub at gcc dot gnu.org
  2021-01-29  9:37 ` gcc at jochen dot sprickerhof.de
@ 2021-01-29 16:59 ` msebor at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: msebor at gcc dot gnu.org @ 2021-01-29 16:59 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=98880

Martin Sebor <msebor at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |msebor at gcc dot gnu.org

--- Comment #3 from Martin Sebor <msebor at gcc dot gnu.org> ---
As a data point, the invalid pointer subtraction is diagnosed at -O2 with the
patch submitted in November (but not reviewed):
https://gcc.gnu.org/pipermail/gcc-patches/2020-November/558775.html

$ gcc -O2 -Wall pr98880.c 
pr98880.c: In function ‘main’:
pr98880.c:25:20: warning: subtracting pointers to distinct objects
[-Wpointer-compare=]
   25 |   list.slh_first = (ssize_t)((char *)&ele - (char *)&list);
      |                    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
pr98880.c:22:16: note: operand 1 points to ‘ele’ declared here
   22 |   struct sh_le ele;
      |                ^~~
pr98880.c:21:16: note: operand 2 points to ‘list’ declared here
   21 |   struct sh_lq list;
      |                ^~~~
In file included from pr98880.c:2:
In function ‘sh_l_discard’,
    inlined from ‘main’ at pr98880.c:27:3:
pr98880.c:17:13: warning: ‘ele.sle_next’ is used uninitialized
[-Wuninitialized]
   17 |   assert(ele->sle_next == -1);
      |          ~~~^~~~~~~~~~
pr98880.c: In function ‘main’:
pr98880.c:22:16: note: ‘ele’ declared here
   22 |   struct sh_le ele;
      |                ^~~

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2021-01-29 16:59 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-29  9:14 [Bug c/98880] New: [10/11 Regression] assignment unexpectedly optimized away with -O gcc at jochen dot sprickerhof.de
2021-01-29  9:18 ` [Bug c/98880] " jakub at gcc dot gnu.org
2021-01-29  9:37 ` gcc at jochen dot sprickerhof.de
2021-01-29 16:59 ` msebor 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).