[Bug target/40811] New: unsigned int to float isn't vectorized

[Bug target/40811] New: unsigned int to float isn't vectorized

[hjl dot tools at gmail dot com]

[hjl@gnu-6 tmp]$ cat x.c
#define N 4
extern unsigned int u4[N] __attribute__ ((aligned(16)));
extern float f4[N] __attribute__ ((aligned(16)));
void
cvt_u4_f4()
{
  int j;
  for (j=0; j<N; j++)
    f4[j]=u4[j];
}
[hjl@gnu-6 tmp]$ /usr/gcc-4.5/bin/gcc -O3  x.c -S  
[hjl@gnu-6 tmp]$ cat x.s
        .file   "x.c"
        .text
        .p2align 4,,15
.globl cvt_u4_f4
        .type   cvt_u4_f4, @function
cvt_u4_f4:
.LFB0:
        .cfi_startproc
        mov     u4(%rip), %eax
        cvtsi2ssq       %rax, %xmm0
        mov     u4+4(%rip), %eax
        movss   %xmm0, f4(%rip)
        cvtsi2ssq       %rax, %xmm0
        mov     u4+8(%rip), %eax
        movss   %xmm0, f4+4(%rip)
        cvtsi2ssq       %rax, %xmm0
        mov     u4+12(%rip), %eax
        movss   %xmm0, f4+8(%rip)
        cvtsi2ssq       %rax, %xmm0
        movss   %xmm0, f4+12(%rip)
        ret
        .cfi_endproc
.LFE0:
        .size   cvt_u4_f4, .-cvt_u4_f4

Icc 11.1 generates:

        .globl cvt_u4_f4
cvt_u4_f4:
..B1.1:                         # Preds ..B1.0
..___tag_value_cvt_u4_f4.1:                                     #6.1
        cvtdq2ps  u4(%rip), %xmm2                               #9.11
        movaps    %xmm2, %xmm1                                  #9.11
        pxor      %xmm0, %xmm0                                  #9.11
        cmpltps   %xmm0, %xmm1                                  #9.11
        andps     _2il0floatpacket.1(%rip), %xmm1               #9.11
        addps     %xmm1, %xmm2                                  #9.11
        movaps    %xmm2, f4(%rip)                               #9.5
        ret                                                     #10.1
        .align    16,0x90
..___tag_value_cvt_u4_f4.2:                                     #
                                # LOE
# mark_end;
        .type   cvt_u4_f4,@function
        .size   cvt_u4_f4,.-cvt_u4_f4
        .data
# -- End  cvt_u4_f4
        .section .rodata, "a"
        .align 16
        .align 16
_2il0floatpacket.1:
        .long   0x4f800000,0x4f800000,0x4f800000,0x4f800000
        .type   _2il0floatpacket.1,@object
        .size   _2il0floatpacket.1,16


-- 
           Summary: unsigned int to float isn't vectorized
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: hjl dot tools at gmail dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #1 from ubizjak at gmail dot com  2009-07-20 23:26 -------
What about unsigned int -> double?


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2009-07-20 23:26:33
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[hjl dot tools at gmail dot com]

------- Comment #2 from hjl dot tools at gmail dot com  2009-07-21 02:52 -------
(In reply to comment #1)
> What about unsigned int -> double?
> 

We don't even support int -> double.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #3 from ubizjak at gmail dot com  2009-07-21 06:15 -------
(In reply to comment #2)
>
> We don't even support int -> double.

Yes, we do. Try:

#define N 16

extern int u4[N] __attribute__ ((aligned(16)));
extern double f4[N] __attribute__ ((aligned(16)));

void
cvt_u4_f4()
{
  int j;
  for (j=0; j<N; j++)
    f4[j]=u4[j];
}


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[uros at gcc dot gnu dot org]

------- Comment #4 from uros at gcc dot gnu dot org  2009-07-21 07:23 -------
Subject: Bug 40811

Author: uros
Date: Tue Jul 21 07:22:51 2009
New Revision: 149847

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=149847
Log:
        Backport from mainline:
        2009-04-29  Richard Guenther  <rguenther@suse.de>

        PR target/39943
        * config/i386/i386.c (ix86_vectorize_builtin_conversion): Only
        allow conversion to signed integers.

testsuite/ChangeLog:

        PR target/40811
        * gcc.target/i386/pr39943.c: New test.

        Backport from mainline:
        2009-04-29  Richard Guenther  <rguenther@suse.de>

        PR target/39943
        * lib/target-supports.exp (check_effective_target_vect_uintfloat_cvt):
        New.
        (check_effective_target_vect_floatuint_cvt): Likewise.
        * gcc.dg/vect/slp-10.c: Adjust.
        * gcc.dg/vect/slp-11.c: Adjust.
        * gcc.dg/vect/slp-12b.c: Adjust.
        * gcc.dg/vect/slp-33.c: Adjust.
        * gcc.c-torture/compile/pr39943.c: New testcase.


Added:
    branches/gcc-4_4-branch/gcc/testsuite/gcc.c-torture/compile/pr39943.c
    branches/gcc-4_4-branch/gcc/testsuite/gcc.target/i386/pr40811.c
Modified:
    branches/gcc-4_4-branch/gcc/ChangeLog
    branches/gcc-4_4-branch/gcc/config/i386/i386.c
    branches/gcc-4_4-branch/gcc/testsuite/ChangeLog
    branches/gcc-4_4-branch/gcc/testsuite/gcc.dg/vect/slp-10.c
    branches/gcc-4_4-branch/gcc/testsuite/gcc.dg/vect/slp-11.c
    branches/gcc-4_4-branch/gcc/testsuite/gcc.dg/vect/slp-12b.c
    branches/gcc-4_4-branch/gcc/testsuite/gcc.dg/vect/slp-33.c
    branches/gcc-4_4-branch/gcc/testsuite/lib/target-supports.exp


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #6 from ubizjak at gmail dot com  2009-07-21 10:04 -------
Taking a bug.


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |ubizjak at gmail dot com
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-07-20 23:26:33         |2009-07-21 10:04:16
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #5 from ubizjak at gmail dot com  2009-07-21 10:04 -------
(In reply to comment #4)
> Subject: Bug 40811
> 
> Author: uros
> Date: Tue Jul 21 07:22:51 2009
> New Revision: 149847

Sorry, wrong PR number.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #7 from ubizjak at gmail dot com  2009-07-21 11:32 -------
Created an attachment (id=18236)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18236&action=view)
implement vectorization of unsigned int -> float

This patch vectorizes unsigned int -> float conversion.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[uros at gcc dot gnu dot org]

------- Comment #8 from uros at gcc dot gnu dot org  2009-07-21 15:17 -------
Subject: Bug 40811

Author: uros
Date: Tue Jul 21 15:17:23 2009
New Revision: 149861

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=149861
Log:
        PR target/40811
        * config/i386/sse.md (sse2_cvtudq2ps): New expander.
        (enum ix86_builtins): Add IX86_BUILTIN_CVTUDQ2PS.
        (builtin_description): Add __builtin_ia32_cvtudq2ps.
        (ix86_vectorize_builtin_conversion): Handle IX86_BUILTIN_CVTUDQ2PS.

testsuite/ChangeLog:

        PR target/40811
        * lib/target-supports.exp (check_effective_target_vect_uintfloat_cvt):
        Add i?86 and x86_64 targets.
        * gcc.target/i386/vectorize7.c: New test.

        PR target/40809
        * gcc.target/i386/pr40809.c: New test.


Added:
    trunk/gcc/testsuite/gcc.target/i386/pr40809.c
    trunk/gcc/testsuite/gcc.target/i386/vectorize7.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/i386/i386.c
    trunk/gcc/config/i386/sse.md
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/lib/target-supports.exp


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811

[Bug target/40811] unsigned int to float isn't vectorized

[ubizjak at gmail dot com]

------- Comment #9 from ubizjak at gmail dot com  2009-07-21 15:36 -------
Fixed.

BTW: The patch to vectorize unsigned int -> double is at [1].

[1] http://gcc.gnu.org/ml/gcc-patches/2009-07/msg01170.html


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                URL|                            |http://gcc.gnu.org/ml/gcc-
                   |                            |patches/2009-
                   |                            |07/msg01164.html
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=40811