Re: [PATCH v2] Add vec_const_duplicate optab and TARGET_GEN_MEMSET_SCRATCH_RTX

[Hongtao Liu <crazylht@gmail.com>]

On Wed, Jun 2, 2021 at 7:07 AM H.J. Lu via Gcc-patches
<gcc-patches@gcc.gnu.org> wrote:
>
> On Tue, Jun 1, 2021 at 7:21 AM Jeff Law <jeffreyalaw@gmail.com> wrote:
> >
> >
> >
> > On 6/1/2021 7:29 AM, H.J. Lu via Gcc-patches wrote:
> > > On Tue, Jun 1, 2021 at 6:25 AM Richard Biener
> > > <richard.guenther@gmail.com> wrote:
> > >> On Tue, Jun 1, 2021 at 3:05 PM H.J. Lu <hjl.tools@gmail.com> wrote:
> > >>> On Mon, May 31, 2021 at 11:54:53PM -0600, Jeff Law wrote:
> > >>>>
> > >>>> On 5/31/2021 11:50 PM, Richard Sandiford wrote:
> > >>>>> "H.J. Lu via Gcc-patches" <gcc-patches@gcc.gnu.org> writes:
> > >>>>>> On Mon, May 31, 2021 at 06:32:04AM -0700, H.J. Lu wrote:
> > >>>>>>> On Mon, May 31, 2021 at 6:26 AM Richard Biener
> > >>>>>>> <richard.guenther@gmail.com> wrote:
> > >>>>>>>> On Mon, May 31, 2021 at 3:12 PM H.J. Lu <hjl.tools@gmail.com> wrote:
> > >>>>>>>>> On Mon, May 31, 2021 at 5:46 AM Richard Biener
> > >>>>>>>>> <richard.guenther@gmail.com> wrote:
> > >>>>>>>>>> On Mon, May 31, 2021 at 2:09 PM H.J. Lu <hjl.tools@gmail.com> wrote:
> > >>>>>>>>>>> On Wed, May 26, 2021 at 10:28:16AM +0200, Richard Biener wrote:
> > >>>>>>>>>>>>>>>    -- Target Hook: rtx TARGET_GEN_MEMSET_VALUE (rtx DATA, scalar_int_mode
> > >>>>>>>>>>>>>>>             MODE)
> > >>>>>>>>>>>>>>>        This function returns the RTL of a register containing
> > >>>>>>>>>>>>>>>        'GET_MODE_SIZE (MODE)' consecutive copies of the unsigned char
> > >>>>>>>>>>>>>>>        value given in the RTL register DATA.  For example, if MODE is 4
> > >>>>>>>>>>>>>>>        bytes wide, return the RTL for 0x01010101*DATA.
> > >>>>>>>>>>>>>> For this one I wonder if it should be an optab instead.  Couldn't you
> > >>>>>>>>>>>>>> use the existing vec_duplicate for this by using (paradoxical) subregs
> > >>>>>>>>>>>>>> like (subreg:TI (vec_duplicate:VnQI (subreg:VnQI (reg:QI ...)))?
> > >>>>>>>>>>>>> I tried.   It doesn't even work on x86.  See:
> > >>>>>>>>>>>>>
> > >>>>>>>>>>>>> https://gcc.gnu.org/pipermail/gcc-patches/2021-May/570661.html
> > >>>>>>>>>>>> Not sure what I should read from there...
> > >>>>>>>>>>>>
> > >>>>>>>>>>>>> There are special cases to subreg HI, SI and DI modes of TI mode in
> > >>>>>>>>>>>>> ix86_gen_memset_value_from_prev.   simplify_gen_subreg doesn't
> > >>>>>>>>>>>>> work here.   Each backend may need its own special handling.
> > >>>>>>>>>>>> OK, I guess I'm not (RTL) qualified enough to further review these parts,
> > >>>>>>>>>>>> sorry.  Since we're doing code generation the canonical way to communicate
> > >>>>>>>>>>>> with backends should be optabs, not some set of disconnected target hooks.
> > >>>>>>>>>>>> But as said, I probably don't know enough of RTL to see why it's the only way.
> > >>>>>>>>>>>>
> > >>>>>>>>>>>> Richard.
> > >>>>>>>>>>> Here is the patch to add optabs instead.  Does it look OK?
> > >>>>>>>>>>>
> > >>>>>>>>>>> Thanks.
> > >>>>>>>>>>>
> > >>>>>>>>>>> H.J.
> > >>>>>>>>>>> ---
> > >>>>>>>>>>> Add 2 optabs:
> > >>>>>>>>>>>
> > >>>>>>>>>>> 1. integer_extract: Extract lower bit value from the integer value in
> > >>>>>>>>>>> TImode, OImode or XImode.
> > >>>>>>>>>> That sounds very specific, esp. the restriction to {TI,OI,XI}mode.
> > >>>>>>>>>> It also sounds like it matches (subreg:{TI,OI,XI} (...) 0).  There are
> > >>>>>>>>>> existing target hooks verifying subreg validity - why's that not a good
> > >>>>>>>>>> fit here?  ISTR you say gen_lowpart () doesn't work (or was it
> > >>>>>>>>>> simplify_gen_subreg?), why's that so?
> > >>>>>>>>> {TI,OI,XI}mode are storage only integer types.   subreg doesn't work
> > >>>>>>>>> well on them.  I got
> > >>>>>>>>>
> > >>>>>>>>> [hjl@gnu-cfl-2 pieces]$ cat s2.i
> > >>>>>>>>> extern void *ops;
> > >>>>>>>>>
> > >>>>>>>>> void
> > >>>>>>>>> foo (int c)
> > >>>>>>>>> {
> > >>>>>>>>>     __builtin_memset (ops, c, 34);
> > >>>>>>>>> }
> > >>>>>>>>> [hjl@gnu-cfl-2 pieces]$ make s2.s
> > >>>>>>>>> /export/build/gnu/tools-build/gcc-gitlab-debug/build-x86_64-linux/gcc/xgcc
> > >>>>>>>>> -B/export/build/gnu/tools-build/gcc-gitlab-debug/build-x86_64-linux/gcc/
> > >>>>>>>>> -O2 -march=haswell -S s2.i
> > >>>>>>>>> during RTL pass: reload
> > >>>>>>>>> s2.i: In function ‘foo’:
> > >>>>>>>>> s2.i:7:1: internal compiler error: maximum number of generated reload
> > >>>>>>>>> insns per insn achieved (90)
> > >>>>>>>>>       7 | }
> > >>>>>>>>>         | ^
> > >>>>>>>>> 0x1050734 lra_constraints(bool)
> > >>>>>>>>> /export/gnu/import/git/gitlab/x86-gcc/gcc/lra-constraints.c:5091
> > >>>>>>>>> 0x1039536 lra(_IO_FILE*)
> > >>>>>>>>> /export/gnu/import/git/gitlab/x86-gcc/gcc/lra.c:2336
> > >>>>>>>>> 0xfe1140 do_reload
> > >>>>>>>>> /export/gnu/import/git/gitlab/x86-gcc/gcc/ira.c:5822
> > >>>>>>>>> 0xfe162e execute
> > >>>>>>>>> /export/gnu/import/git/gitlab/x86-gcc/gcc/ira.c:6008
> > >>>>>>>>> Please submit a full bug report,
> > >>>>>>>>> with preprocessed source if appropriate.
> > >>>>>>>>> Please include the complete backtrace with any bug report.
> > >>>>>>>>> See <https://gcc.gnu.org/bugs/> for instructions.
> > >>>>>>>>> make: *** [Makefile:32: s2.s] Error 1
> > >>>>>>>>> [hjl@gnu-cfl-2 pieces]$
> > >>>>>>>>>
> > >>>>>>>>> due to
> > >>>>>>>>>
> > >>>>>>>>> (insn 12 11 0 (set (mem:HI (plus:DI (reg/f:DI 84)
> > >>>>>>>>>                   (const_int 32 [0x20])) [0 MEM <char[1:34]> [(void
> > >>>>>>>>> *)ops.0_1]+32 S2 A8])
> > >>>>>>>>>           (subreg:HI (reg:OI 51 xmm15) 0)) "s2.i":6:3 -1
> > >>>>>>>>>        (nil))
> > >>>>>>>>>
> > >>>>>>>>> The new optab gives us
> > >>>>>>>>>
> > >>>>>>>>> (insn 12 11 13 2 (set (reg:TI 88)
> > >>>>>>>>>           (reg:TI 51 xmm15)) "s2.i":6:3 -1
> > >>>>>>>>>        (nil))
> > >>>>>>>>> (insn 13 12 14 2 (set (reg:SI 89)
> > >>>>>>>>>           (subreg:SI (reg:TI 88) 0)) "s2.i":6:3 -1
> > >>>>>>>>>        (nil))
> > >>>>>>>>> (insn 14 13 15 2 (set (reg:HI 87)
> > >>>>>>>>>           (subreg:HI (reg:SI 89) 0)) "s2.i":6:3 -1
> > >>>>>>>>>        (nil))
> > >>>>>>>> that looks odd to me - what's the final result after LRA?  I think
> > >>>>>>> I got:
> > >>>>>>>
> > >>>>>>> vmovd %edi, %xmm15
> > >>>>>>> movq ops(%rip), %rdx
> > >>>>>>> vpbroadcastb %xmm15, %ymm15
> > >>>>>>> vmovq %xmm15, %rax    <<<< move to GPR
> > >>>>>>> vmovdqu %ymm15, (%rdx)
> > >>>>>>> movw %ax, 32(%rdx)   <<<< subreg of GPR
> > >>>>>>> vzeroupper
> > >>>>>>> ret
> > >>>>>>>
> > >>>>>>>> we should see to make lowpart_subreg work on {XI,OI,TI}mode.
> > >>>>>>>> Only two steps should be necessary at most:
> > >>>>>>>> xmm -> gpr, grp -> subreg, or gpr -> subreg.  So the expander
> > >>>>>>>> code in memset should try to generate the subreg directly
> > >>>>>>> subreg didn't fail on x86 when I tried.
> > >>>>>>>
> > >>>>>>>> and if that fails, try a word_mode subreg followed by the subreg.
> > >>>>>>> I will try word_mode subreg.
> > >>>>>>>
> > >>>>>> Here is the v2 patch to use word_mode subreg.  For
> > >>>>>>
> > >>>>>> ---
> > >>>>>> extern void *ops;
> > >>>>>>
> > >>>>>> void
> > >>>>>> foo (int c)
> > >>>>>> {
> > >>>>>>     __builtin_memset (ops, 4, 32);
> > >>>>>> }
> > >>>>>> ---
> > >>>>>>
> > >>>>>> without vec_const_duplicate, I got
> > >>>>>>
> > >>>>>>    movl    $4, %eax
> > >>>>>>    movq    ops(%rip), %rdx
> > >>>>>>    movd    %eax, %xmm0
> > >>>>>>    punpcklbw       %xmm0, %xmm0
> > >>>>>>    punpcklwd       %xmm0, %xmm0
> > >>>>>>    pshufd  $0, %xmm0, %xmm0
> > >>>>>>    movups  %xmm0, (%rdx)
> > >>>>>>    movups  %xmm0, 16(%rdx)
> > >>>>>>    ret
> > >>>>>>
> > >>>>>> With vec_const_duplicate, I got
> > >>>>>>
> > >>>>>>    movq    ops(%rip), %rax
> > >>>>>>    movdqa  .LC0(%rip), %xmm0
> > >>>>>>    movups  %xmm0, (%rax)
> > >>>>>>    movups  %xmm0, 16(%rax)
> > >>>>>>    ret
> > >>>>>>
> > >>>>>> If vec_duplicate is allowed to fail, I don't need vec_const_duplicate.
> > >>>>> I don't understand why we need an optab for this though.  If the operand
> > >>>>> is constant then we should just be doing an ordinary move in which the
> > >>>>> source is a CONST_VECTOR.  It's then up to the move patterns to handle
> > >>>>> duplicated constants as efficiently as possible.  (Sorry if this was
> > >>>>> discussed upthread and I missed it.)
> > >>>> That's exactly the point I'm trying to get across as well.
> > >>>>
> > >>> This is what we do today.  But I'd like to generate
> > >>>
> > >>>          movl    $4, %eax
> > >>>          vpbroadcastb    %eax, %ymm15
> > >>>          movq    ops(%rip), %rax
> > >>>          vmovdqu %ymm15, (%rax)
> > >>>          vzeroupper
> > >>>          ret
> > >>>
> > >>> instead of
> > >>>
> > >>>          vmovdqa .LC0(%rip), %ymm15
> > >>>          movq    ops(%rip), %rax
> > >>>          vmovdqu %ymm15, (%rax)
> > >>>          vzeroupper
> > >>>          ret
> > >>>
> > >>> Do I need a vec_dup pattern for it?
> > >> I think we have special code sequences to materialize some
> > >> constant vectors already, we should be able to add to that, no?
> > > We can do that for all 0s and all 1s at the final codegen.   For
> > > other values, since we need a GPR, we can't do that.
> > You can catch them in your movxx expanders, you can create peep2
> > patterns that use available GPRs, etc.  I don't see a fundamental need
> > to to introduce new target macros or hooks to handle this stuff.  In
> > fact I've done both to handle a closely related issue on our port.
> >
>
> One problem of expanding TI/OI/XI moves to broadcast is that other
> RTL passes may change it.   For example, expander generates:
It could be handled in pass_data_constant_pool_broadcast which is
designed for avx512 embedding broadcast, but can also do such
transforming.

see
https://godbolt.org/z/8YGzqf938
>
> (insn 7 5 6 (set (reg:QI 85)
>         (const_int 12 [0xc]))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> -1
>      (nil))
>
> (insn 6 7 8 (set (reg:V16QI 84)
>         (vec_duplicate:V16QI (reg:QI 85)))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c
> ":9:3 5103 {*avx512vl_vec_dup_gprv16qi}
>      (nil))
>
> (insn 8 6 9 (set (subreg:V16QI (reg:TI 86) 0)
>         (reg:V16QI 84))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> -1
>      (nil))
>
> (insn 9 8 10 (set (mem:TI (reg/f:DI 83) [0 MEM <char[1:19]> [(void
> *)dst.0_1]+0 S16 A8])
>         (reg:TI 86))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> -1
>      (nil))
>
> combine turns it into:
>
> insn 9 6 10 2 (set (mem:TI (reg/f:DI 83 [ dst ]) [0 MEM <char[1:19]>
> [(void *)dst.0_1]+0 S16 A8])
>         (const_wide_int 0xc0c0c0c0c0c0c0c0c0c0c0c0c0c0c0c))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/
> i386/pr90773-17.c":9:3 73 {*movti_internal}
>      (nil))
>
> LRA tries:
>
> (insn 14 15 16 2 (set (reg:V16QI 20 xmm0 [89])
>         (vec_duplicate:V16QI (reg:QI 4 si [90])))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr907
> 73-17.c":9:3 5103 {*avx512vl_vec_dup_gprv16qi}
>      (nil))
> (insn 16 14 9 2 (set (reg:V16QI 1 dx)
>         (reg:V16QI 20 xmm0 [89]))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> 152
> 7 {movv16qi_internal}
>      (nil))
> (insn 9 16 10 2 (set (mem:TI (reg/f:DI 0 ax [orig:83 dst ] [83]) [0
> MEM <char[1:19]> [(void *)dst.0_1]+0 S16 A8])
>         (reg:TI 1 dx [88]))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> 73 {*movt
> i_internal}
>
> and fails:
>
> /export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c:
> In function ‘foo’:
> /export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c:10:1:
> error: insn does not satisfy its constraints:
> (insn 16 14 9 2 (set (reg:V16QI 1 dx)
>         (reg:V16QI 20 xmm0 [89]))
> "/export/gnu/import/git/gitlab/x86-gcc/gcc/testsuite/gcc.target/i386/pr90773-17.c":9:3
> 1527 {movv16qi_internal}
>      (nil))
>
> I want to hide
>
> (const_wide_int 0xc0c0c0c0c0c0c0c0c0c0c0c0c0c0c0c)
>
> from RTL passes.
>
>
> --
> H.J.



-- 
BR,
Hongtao