[Bug middle-end/66881] New: Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[Bug middle-end/66881] New: Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[tkoeppe at google dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

            Bug ID: 66881
           Summary: Possibly inefficient std::atomic<int> codegen on x86
                    for simple arithmetic
           Product: gcc
           Version: 4.9.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: tkoeppe at google dot com
  Target Milestone: ---

Consider these two simple versions of addition:

  #include <atomic>

  std::atomic<int> x;
  int y;

  void f(int a) {
    x.store(x.load(std::memory_order_relaxed) + a, std::memory_order_relaxed);
  }

  void g(int a) {
    y += a;
  }

GCC generates the following assembly:

  f(int):
        mov     eax, DWORD PTR x[rip]
        add     edi, eax
        mov     DWORD PTR x[rip], edi
        ret

  g(int):
        add     DWORD PTR y[rip], edi
        ret

Now, it is clear to me that the correct atomic codegen for store() and load()
is "mov", as it appears here, but why aren't the two consecutive operations not
folded into a single add? Aren't the semantics and the memory ordering the
same? x86 says that (most) "reads" and "writes" are strongly ordered; doesn't
that apply to the read and write produced by "add", too?

(My original motivation came from a variant of this with floats, where the
non-atomic code executed noticeably faster, even though I would have expected
the two to produce the same machine code.)

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> ---
Dup of PR 50677?

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[ville.voutilainen at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

--- Comment #2 from Ville Voutilainen <ville.voutilainen at gmail dot com> ---
(In reply to H.J. Lu from comment #1)
> Dup of PR 50677?

Well, this report has no volatiles in it, so perhaps not?

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Ville Voutilainen from comment #2)
> (In reply to H.J. Lu from comment #1)
> > Dup of PR 50677?
> 
> Well, this report has no volatiles in it, so perhaps not?

Can you change init_recog_no_volatile to try it out?

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|FIXED                       |DUPLICATE

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> ---


*** This bug has been marked as a duplicate of bug 50677 ***

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

H.J. Lu <hjl.tools at gmail dot com> changed:

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

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> ---
Dup.

[Bug middle-end/66881] Possibly inefficient std::atomic<int> codegen on x86 for simple arithmetic

[jfb at chromium dot org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66881

JF Bastien <jfb at chromium dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jfb at chromium dot org

--- Comment #6 from JF Bastien <jfb at chromium dot org> ---
I don't think this is a duplicate of PR 50677: this issue uses two *separate*
atomic operations, which inherently aren't atomic, whereas PR 50677 uses a
single volatile operations which some (misguided) developers expect to be
atomic.

The compiler may decide to be nice to developers who have misguided
expectations from using volatile, but this example clearly uses separate
load/store operations. I believe Thomas is correct: the code for f() should
codegen the same code as g().