[Bug target/60104] New: load not folded into indirect branch on x86-64

[Bug target/60104] New: load not folded into indirect branch on x86-64

[dan433584 at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

            Bug ID: 60104
           Summary: load not folded into indirect branch on x86-64
           Product: gcc
           Version: 4.8.2
            Status: UNCONFIRMED
          Severity: enhancement
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: dan433584 at gmail dot com

Created attachment 32075
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=32075&action=edit
a C testcase

The attached testcase is a greatly reduced interpreter loop, containing a
simple load and indirect branch:

  goto *addresses[*pc++]

gcc 4.8.2 (as well as older versions) with -O2 produces the following x86-64
output:

  movq    addresses.1721(,%rax,8), %rax
  jmp    *%rax

Since the loaded value is not used after the branch, there's no need to hold it
in a register, so the load could be folded into the branch. This would improve
code size and instruction count.

[Bug target/60104] load not folded into indirect branch on x86-64

[ktietz at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

--- Comment #1 from Kai Tietz <ktietz at gcc dot gnu.org> ---
Author: ktietz
Date: Fri May 30 18:00:11 2014
New Revision: 211089

URL: http://gcc.gnu.org/viewcvs?rev=211089&root=gcc&view=rev
Log:
    PR target/60104
    * config/i386/i386.c (x86_output_mi_thunk): Add memory case
    for sibling-tail-calls.
    * config/i386/i386.md (sibcall_insn_operand): Add memory-constrain
    to its use.
    * config/i386/predicates.md (sibcall_memory_operand): New predicate.
    (sibcall_insn_operand): Add check for sibcall_memory_operand.

    PR target/60104
    * gcc.target/i386/sibcall-1.c: New test.
    * gcc.target/i386/sibcall-2.c: New test.
    * gcc.target/i386/sibcall-3.c: New test.
    * gcc.target/i386/sibcall-4.c: New test.


Added:
    trunk/gcc/testsuite/gcc.target/i386/sibcall-1.c
    trunk/gcc/testsuite/gcc.target/i386/sibcall-2.c
    trunk/gcc/testsuite/gcc.target/i386/sibcall-3.c
    trunk/gcc/testsuite/gcc.target/i386/sibcall-4.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/i386/i386.c
    trunk/gcc/config/i386/i386.md
    trunk/gcc/config/i386/predicates.md

[Bug target/60104] load not folded into indirect branch on x86-64

[fxcoudert at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> changed:

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

--- Comment #2 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
This patch caused PR61387.

[Bug target/60104] load not folded into indirect branch on x86-64

[fxcoudert at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

--- Comment #3 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
Testcase gcc.target/i386/sibcall-1.c fails on x86_64-apple-darwin14 (compiled
with "-O2 -m32"). Assembler code generated is:

    .section __TEXT,__text_cold,regular,pure_instructions
LCOLDB0:
    .text
LHOTB0:
    .align 4,0x90
    .globl _boo
_boo:
LFB0:
    call    ___x86.get_pc_thunk.ax
L1$pb:
    movl    L_foo$non_lazy_ptr-L1$pb(%eax), %eax
    movl    (%eax), %eax
    jmp    *%eax
LFE0:
    .section __TEXT,__text_cold,regular,pure_instructions
LCOLDE0:
    .text
LHOTE0:
    .section __TEXT,__textcoal_nt,coalesced,pure_instructions
    .weak_definition    ___x86.get_pc_thunk.ax
    .private_extern    ___x86.get_pc_thunk.ax
___x86.get_pc_thunk.ax:
LFB1:
    movl    (%esp), %eax
    ret
LFE1:
    .section __TEXT,__eh_frame,coalesced,no_toc+strip_static_syms+live_support
EH_frame1:
    .set L$set$0,LECIE1-LSCIE1
    .long L$set$0
LSCIE1:
    .long    0
    .byte    0x1
    .ascii "zR\0"
    .byte    0x1
    .byte    0x7c
    .byte    0x8
    .byte    0x1
    .byte    0x10
    .byte    0xc
    .byte    0x5
    .byte    0x4
    .byte    0x88
    .byte    0x1
    .align 2
LECIE1:
LSFDE1:
    .set L$set$1,LEFDE1-LASFDE1
    .long L$set$1
LASFDE1:
    .long    LASFDE1-EH_frame1
    .long    LFB0-.
    .set L$set$2,LFE0-LFB0
    .long L$set$2
    .byte    0
    .align 2
LEFDE1:
LSFDE3:
    .set L$set$3,LEFDE3-LASFDE3
    .long L$set$3
LASFDE3:
    .long    LASFDE3-EH_frame1
    .long    LFB1-.
    .set L$set$4,LFE1-LFB1
    .long L$set$4
    .byte    0
    .align 2
LEFDE3:
    .section __IMPORT,__pointers,non_lazy_symbol_pointers
L_foo$non_lazy_ptr:
    .indirect_symbol _foo
    .long    0
    .subsections_via_symbols

[Bug target/60104] load not folded into indirect branch on x86-64

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |FIXED
   Target Milestone|---                         |5.0

--- Comment #4 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Francois-Xavier Coudert from comment #3)
> Testcase gcc.target/i386/sibcall-1.c fails on x86_64-apple-darwin14
> (compiled with "-O2 -m32"). Assembler code generated is:

It is expected since this testcase always fails with PIC in 32-bit.

[Bug target/60104] load not folded into indirect branch on x86-64

[fxcoudert at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

--- Comment #5 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
Author: fxcoudert
Date: Sat Nov 15 19:46:14 2014
New Revision: 217611

URL: https://gcc.gnu.org/viewcvs?rev=217611&root=gcc&view=rev
Log:
    PR target/60104
    * gcc.target/i386/sibcall-1.c: Don't run on pic targets.

Modified:
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gcc.target/i386/sibcall-1.c

[Bug target/60104] load not folded into indirect branch on x86-64

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=60104

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hjl.tools at gmail dot com

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
*** Bug 53772 has been marked as a duplicate of this bug. ***