From: Uros Bizjak <ubizjak@gmail.com>
To: Roger Sayle <roger@nextmovesoftware.com>
Cc: gcc-patches@gcc.gnu.org
Subject: Re: [x86 PATCH] PR target/31985: Improve memory operand use with doubleword add.
Date: Wed, 7 Jun 2023 09:07:50 +0200 [thread overview]
Message-ID: <CAFULd4bpz5rqiHf-r5MG_=P=AmkNe5L+Dg0ZaBTa2OF8m7ZgpA@mail.gmail.com> (raw)
In-Reply-To: <CAFULd4Y5_b3tDkdvMWzq=T=197nnvkqLhvM9Noa_gA5CoW-53g@mail.gmail.com>
On Wed, Jun 7, 2023 at 8:32 AM Uros Bizjak <ubizjak@gmail.com> wrote:
>
> On Wed, Jun 7, 2023 at 1:05 AM Roger Sayle <roger@nextmovesoftware.com> wrote:
> >
> >
> > This patch addresses the last remaining issue with PR target/31985, that
> > GCC could make better use of memory addressing modes when implementing
> > double word addition. This is achieved by adding a define_insn_and_split
> > that combines an *add<dwi>3_doubleword with a *concat<mode><dwi>3, so
> > that the components of the concat can be used directly, without first
> > being loaded into a double word register.
> >
> > For test_c in the bugzilla PR:
> >
> > Before:
> > pushl %ebx
> > subl $16, %esp
> > movl 28(%esp), %eax
> > movl 36(%esp), %ecx
> > movl 32(%esp), %ebx
> > movl 24(%esp), %edx
> > addl %ecx, %eax
> > adcl %ebx, %edx
> > movl %eax, 8(%esp)
> > movl %edx, 12(%esp)
> > addl $16, %esp
> > popl %ebx
> > ret
> >
> > After:
> > test_c:
> > subl $20, %esp
> > movl 36(%esp), %eax
> > movl 32(%esp), %edx
> > addl 28(%esp), %eax
> > adcl 24(%esp), %edx
> > movl %eax, 8(%esp)
> > movl %edx, 12(%esp)
> > addl $20, %esp
> > ret
> >
> >
> > If this approach is considered acceptable, similar splitters can be
> > used for other doubleword operations.
>
> Yes, the approach looks good to me. But please take care not to
> interfere with STV transformations.
Hm, I have one second thought:
+(define_insn_and_split "*add<dwi>3_doubleword_concat"
+ [(set (match_operand:<DWI> 0 "register_operand" "=r")
+ (plus:<DWI>
+ (any_or_plus:<DWI>
+ (ashift:<DWI>
+ (zero_extend:<DWI>
+ (match_operand:DWIH 2 "nonimmediate_operand" "rm"))
+ (match_operand:<DWI> 3 "const_int_operand"))
+ (zero_extend:<DWI>
+ (match_operand:DWIH 4 "nonimmediate_operand" "rm")))
+ (match_operand:<DWI> 1 "register_operand" "0")))
+ (clobber (reg:CC FLAGS_REG))]
+ "INTVAL (operands[3]) == <MODE_SIZE> * BITS_PER_UNIT"
+ "#"
+ "&& reload_completed"
+ [(parallel [(set (reg:CCC FLAGS_REG)
+ (compare:CCC
+ (plus:DWIH (match_dup 1) (match_dup 4))
+ (match_dup 1)))
+ (set (match_dup 0)
+ (plus:DWIH (match_dup 1) (match_dup 4)))])
+ (parallel [(set (match_dup 5)
+ (plus:DWIH
+ (plus:DWIH
+ (ltu:DWIH (reg:CC FLAGS_REG) (const_int 0))
+ (match_dup 6))
+ (match_dup 2)))
+ (clobber (reg:CC FLAGS_REG))])]
+ "split_double_mode (<DWI>mode, &operands[0], 2, &operands[0], &operands[5]);")
The register allocator considers the instruction-to-be-split as one
instruction, so it can allocate output register to match an input
register (or a register that forms an input address), So, you have to
either add an early clobber to the output, or somehow prevent output
to clobber registers in the second pattern.
Uros.
>
> > This patch has been tested on x86_64-pc-linux-gnu with make bootstrap
> > and make -k check, both with and without --target_board=unix{-m32}
> > with no new failures. Ok for mainline?
> >
> > 2023-06-07 Roger Sayle <roger@nextmovesoftware.com>
> >
> > gcc/ChangeLog
> > PR target/31985
> > * config/i386/i386.md (*add<dwi>3_doubleword_concat): New
> > define_insn_and_split combine *add<dwi>3_doubleword with a
> > *concat<mode><dwi>3 for more efficient lowering after reload.
> >
> > gcc/testsuite/ChangeLog
> > PR target/31985
> > * gcc.target/i386/pr31985.c: New test case.
>
> OK.
>
> Thanks,
> Uros.
>
> >
> >
> > Roger
> > --
> >
next prev parent reply other threads:[~2023-06-07 7:08 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-06 23:05 Roger Sayle
2023-06-07 6:32 ` Uros Bizjak
2023-06-07 7:07 ` Uros Bizjak [this message]
2023-06-15 22:04 ` Roger Sayle
2023-06-16 7:13 ` Uros Bizjak
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAFULd4bpz5rqiHf-r5MG_=P=AmkNe5L+Dg0ZaBTa2OF8m7ZgpA@mail.gmail.com' \
--to=ubizjak@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=roger@nextmovesoftware.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).