Re: [PATCH 06/13] [APX EGPR] Map reg/mem constraints in inline asm to non-EGPR constraint.

[Uros Bizjak <ubizjak@gmail.com>]

On Fri, Sep 1, 2023 at 11:10 AM Hongyu Wang <wwwhhhyyy333@gmail.com> wrote:
>
> Uros Bizjak via Gcc-patches <gcc-patches@gcc.gnu.org> 于2023年8月31日周四 18:01写道：
> >
> > On Thu, Aug 31, 2023 at 11:18 AM Jakub Jelinek via Gcc-patches
> > <gcc-patches@gcc.gnu.org> wrote:
> > >
> > > On Thu, Aug 31, 2023 at 04:20:17PM +0800, Hongyu Wang via Gcc-patches wrote:
> > > > From: Kong Lingling <lingling.kong@intel.com>
> > > >
> > > > In inline asm, we do not know if the insn can use EGPR, so disable EGPR
> > > > usage by default from mapping the common reg/mem constraint to non-EGPR
> > > > constraints. Use a flag mapx-inline-asm-use-gpr32 to enable EGPR usage
> > > > for inline asm.
> > > >
> > > > gcc/ChangeLog:
> > > >
> > > >       * config/i386/i386.cc (INCLUDE_STRING): Add include for
> > > >       ix86_md_asm_adjust.
> > > >       (ix86_md_asm_adjust): When APX EGPR enabled without specifying the
> > > >       target option, map reg/mem constraints to non-EGPR constraints.
> > > >       * config/i386/i386.opt: Add option mapx-inline-asm-use-gpr32.
> > > >
> > > > gcc/testsuite/ChangeLog:
> > > >
> > > >       * gcc.target/i386/apx-inline-gpr-norex2.c: New test.
> > > > ---
> > > >  gcc/config/i386/i386.cc                       |  44 +++++++
> > > >  gcc/config/i386/i386.opt                      |   5 +
> > > >  .../gcc.target/i386/apx-inline-gpr-norex2.c   | 107 ++++++++++++++++++
> > > >  3 files changed, 156 insertions(+)
> > > >  create mode 100644 gcc/testsuite/gcc.target/i386/apx-inline-gpr-norex2.c
> > > >
> > > > diff --git a/gcc/config/i386/i386.cc b/gcc/config/i386/i386.cc
> > > > index d26d9ab0d9d..9460ebbfda4 100644
> > > > --- a/gcc/config/i386/i386.cc
> > > > +++ b/gcc/config/i386/i386.cc
> > > > @@ -17,6 +17,7 @@ You should have received a copy of the GNU General Public License
> > > >  along with GCC; see the file COPYING3.  If not see
> > > >  <http://www.gnu.org/licenses/>.  */
> > > >
> > > > +#define INCLUDE_STRING
> > > >  #define IN_TARGET_CODE 1
> > > >
> > > >  #include "config.h"
> > > > @@ -23077,6 +23078,49 @@ ix86_md_asm_adjust (vec<rtx> &outputs, vec<rtx> & /*inputs*/,
> > > >    bool saw_asm_flag = false;
> > > >
> > > >    start_sequence ();
> > > > +  /* TODO: Here we just mapped the general r/m constraints to non-EGPR
> > > > +   constraints, will eventually map all the usable constraints in the future. */
> > >
> > > I think there should be some constraint which explicitly has all the 32
> > > GPRs, like there is one for just all 16 GPRs (h), so that regardless of
> > > -mapx-inline-asm-use-gpr32 one can be explicit what the inline asm wants.
> > >
> > > Also, what about the "g" constraint?  Shouldn't there be another for "g"
> > > without r16..r31?  What about the various other memory
> > > constraints ("<", "o", ...)?
> >
> > I think we should leave all existing constraints as they are, so "r"
> > covers only GPR16, "m" and "o" to only use GPR16. We can then
> > introduce "h" to instructions that have the ability to handle EGPR.
> > This would be somehow similar to the SSE -> AVX512F transition, where
> > we still have "x" for SSE16 and "v" was introduced as a separate
> > register class for EVEX SSE registers. This way, asm will be
> > compatible, when "r", "m", "o" and "g" are used. The new memory
> > constraint "Bt", should allow new registers, and should be added to
> > the constraint string as a separate constraint, and conditionally
> > enabled by relevant "isa" (AKA "enabled") attribute.
>
> The extended constraint can work for registers, but for memory it is more
> complicated.

Yes, unfortunately. The compiler assumes that an unchangeable register
class is used for BASE/INDEX registers. I have hit this limitation
when trying to implement memory support for instructions involving
8-bit high registers (%ah, %bh, %ch, %dh), which do not support REX
registers, also inside memory operand. (You can see the "hack" in e.g.
*extzvqi_mem_rex64" and corresponding peephole2 with the original
*extzvqi pattern). I am aware that dynamic insn-dependent BASE/INDEX
register class is the major limitation in the compiler, so perhaps the
strategy on how to override this limitation should be discussed with
the register allocator author first. Perhaps adding an insn attribute
to insn RTX pattern to specify different BASE/INDEX register sets can
be a better solution than passing insn RTX to the register allocator.

The above idea still does not solve the asm problem on how to select
correct BASE/INDEX register set for memory operands.

Uros.
>
> If we want to use new mem constraints that allow gpr32, then BASE/INDEX
> reg class still requires per-insn verification, so it means changes
> on all patterns with vm, and those SSE patterns on opcode map0/1. Also,
> several legacy insns that are promoted to EVEX encoding space need to be
> changed. The overall implementation could be 10 times larger than current,
> which would be quite hard for maintenance.
>
> >
> > Uros.
> >
> > > > +  if (TARGET_APX_EGPR && !ix86_apx_inline_asm_use_gpr32)
> > > > +    {
> > > > +      /* Map "r" constraint in inline asm to "h" that disallows r16-r31
> > > > +      and replace only r, exclude Br and Yr.  */
> > > > +      for (unsigned i = 0; i < constraints.length (); i++)
> > > > +     {
> > > > +       std::string *s = new std::string (constraints[i]);
> > >
> > > Doesn't this leak memory (all the time)?
> > > I must say I don't really understand why you need to use std::string here,
> > > but certainly it shouldn't leak.
> > >
> > > > +       size_t pos = s->find ('r');
> > > > +       while (pos != std::string::npos)
> > > > +         {
> > > > +           if (pos > 0
> > > > +               && (s->at (pos - 1) == 'Y' || s->at (pos - 1) == 'B'))
> > > > +             pos = s->find ('r', pos + 1);
> > > > +           else
> > > > +             {
> > > > +               s->replace (pos, 1, "h");
> > > > +               constraints[i] = (const char*) s->c_str ();
> > >
> > > Formatting (space before *).  The usual way for constraints is ggc_strdup on
> > > some string in a buffer.  Also, one could have several copies or r (or m, memory (doesn't
> > > that appear just in clobbers?  And that doesn't look like something that
> > > should be replaced), Bm, e.g. in various alternatives.  So, you
> > > need to change them all, not just the first hit.  "r,r,r,m" and the like.
> > > Normally, one would simply walk the constraint string, parsing the special
> > > letters (+, =, & etc.) and single letter constraints and 2 letter
> > > constraints using CONSTRAINT_LEN macro (tons of examples in GCC sources).
> > > Either do it in 2 passes, first one counts how long constraint string one
> > > will need after the adjustments (and whether to adjust something at all),
> > > then if needed XALLOCAVEC it and adjust in there, or say use a
> > > auto_vec<char, 32> for
> > > it.
> > >
> > > > +               break;
> > > > +             }
> > > > +         }
> > > > +     }
> > > > +      /* Also map "m/memory/Bm" constraint that may use GPR32, replace them with
> > > > +      "Bt/Bt/BT".  */
> > > > +      for (unsigned i = 0; i < constraints.length (); i++)
> > > > +     {
> > > > +       std::string *s = new std::string (constraints[i]);
> > > > +       size_t pos = s->find ("m");
> > > > +       size_t pos2 = s->find ("memory");
> > > > +       if (pos != std::string::npos)
> > > > +         {
> > > > +           if (pos > 0 && (s->at (pos - 1) == 'B'))
> > > > +               s->replace (pos - 1, 2, "BT");
> > > > +           else if (pos2 != std::string::npos)
> > > > +               s->replace (pos, 6, "Bt");
> > > > +           else
> > > > +               s->replace (pos, 1, "Bt");
> > >
> > > Formatting, the s->replace calls are indented too much.
> > >
> > >         Jakub
> > >