[Bug target/95076] New: Failure to optimize out stack alignment on function call of different type

[Bug target/95076] New: Failure to optimize out stack alignment on function call of different type

[gabravier at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

            Bug ID: 95076
           Summary: Failure to optimize out stack alignment on function
                    call of different type
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gabravier at gmail dot com
  Target Milestone: ---

long long f();

int g()
{
    return f();
}

With -O3, LLVM outputs :

g(): # @g()
  jmp f() # TAILCALL

GCC outputs : 

g():
  sub rsp, 8
  call f()
  add rsp, 8
  ret

[Bug target/95076] Failure to tail-call on function call of different return type

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Failure to optimize out     |Failure to tail-call on
                   |stack alignment on function |function call of different
                   |call of different type on   |return type
                   |x86                         |
                 CC|                            |hjl.tools at gmail dot com
             Target|                            |x86_64-*-* i?86-*-*

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC doesn't tail-call because the return types are not compatible.  With a call
it cannot optimize the stack adjustment because of the ABI.

Note I'm not sure whether the ABI allows %rax to contain "garbage" in the
upper half for a function returning in %eax.  So what LLVM does may be wrong.

[Bug target/95076] Failure to tail-call on function call of different return type

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Richard Biener from comment #1)
> GCC doesn't tail-call because the return types are not compatible.  With a
> call
> it cannot optimize the stack adjustment because of the ABI.
> 
> Note I'm not sure whether the ABI allows %rax to contain "garbage" in the
> upper half for a function returning in %eax.  So what LLVM does may be wrong.

Don't upper bits of %rax in GCC generate code also contain garbage?

[Bug target/95076] Failure to tail-call on function call of different return type

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2020-05-12
   Target Milestone|---                         |11.0
             Status|UNCONFIRMED                 |NEW
                 CC|                            |crazylht at gmail dot com
     Ever confirmed|0                           |1

[Bug target/95076] Failure to tail-call on function call of different return type

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> ---
This ia32 psABI return value discussion also applies here:

https://groups.google.com/forum/#!searchin/ia32-abi/return$20value%7Csort:date/ia32-abi/9H4BBrIdkmk/sjWw06ZPnS4J

[Bug target/95076] Failure to tail-call on function call of different return type

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

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

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

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Yeah, I believe it is safe in this case, but we would need the info whether the
conversion is ok or not already during the tailc pass and on many targets such
a conversion is not ok (e.g. if the target says that the value is zero-extended
into 64-bits or sign-extended, or 32-bit and 64-bit returned in different
register, whatever).
The question is if the existing target hooks are sufficient to query this or if
we need a new target hook.

[Bug target/95076] Failure to tail-call on function call of different return type

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.0                        |11.2

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 11.1 has been released, retargeting bugs to GCC 11.2.

[Bug target/95076] Failure to tail-call on function call of different return type

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95076

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|11.2                        |---