From: Hongtao Liu <crazylht@gmail.com>
To: "H.J. Lu" <hjl.tools@gmail.com>
Cc: gcc-patches@gcc.gnu.org, Hongtao Liu <hongtao.liu@intel.com>,
Hongyu Wang <hongyu.wang@intel.com>
Subject: Re: [PATCH] i386: Mark XMM4-XMM6 as clobbered by encodekey128/encodekey256
Date: Wed, 28 Sep 2022 09:41:54 +0800 [thread overview]
Message-ID: <CAMZc-bx5vHO6j7Gag3Wp+5oCUMXLMtX=PkhsKqyemaXCbs6a4A@mail.gmail.com> (raw)
In-Reply-To: <20220927233454.144656-1-hjl.tools@gmail.com>
On Wed, Sep 28, 2022 at 7:35 AM H.J. Lu via Gcc-patches
<gcc-patches@gcc.gnu.org> wrote:
>
> encodekey128 and encodekey256 operations clear XMM4-XMM6. But it is
> documented that XMM4-XMM6 are reserved for future usages and software
> should not rely upon them being zeroed. Change encodekey128 and
Indeed. Ok for trunk and backport.
> encodekey256 to clobber XMM4-XMM6.
>
> gcc/
>
> PR target/107061
> * config/i386/predicates.md (encodekey128_operation): Check
> XMM4-XMM6 as clobbered.
> (encodekey256_operation): Likewise.
> * config/i386/sse.md (encodekey128u32): Clobber XMM4-XMM6.
> (encodekey256u32): Likewise.
>
> gcc/testsuite/
>
> PR target/107061
> * gcc.target/i386/keylocker-encodekey128.c: Don't check
> XMM4-XMM6.
> * gcc.target/i386/keylocker-encodekey256.c: Likewise.
> ---
> gcc/config/i386/predicates.md | 20 +++++++++----------
> gcc/config/i386/sse.md | 4 ++--
> .../gcc.target/i386/keylocker-encodekey128.c | 1 -
> .../gcc.target/i386/keylocker-encodekey256.c | 1 -
> 4 files changed, 12 insertions(+), 14 deletions(-)
>
> diff --git a/gcc/config/i386/predicates.md b/gcc/config/i386/predicates.md
> index 655eabf793b..c4141a96735 100644
> --- a/gcc/config/i386/predicates.md
> +++ b/gcc/config/i386/predicates.md
> @@ -2107,11 +2107,11 @@ (define_predicate "encodekey128_operation"
> for(i = 4; i < 7; i++)
> {
> elt = XVECEXP (op, 0, i);
> - if (GET_CODE (elt) != SET
> - || GET_CODE (SET_DEST (elt)) != REG
> - || GET_MODE (SET_DEST (elt)) != V2DImode
> - || REGNO (SET_DEST (elt)) != GET_SSE_REGNO (i)
> - || SET_SRC (elt) != CONST0_RTX (V2DImode))
> + if (GET_CODE (elt) != CLOBBER
> + || GET_MODE (elt) != VOIDmode
> + || GET_CODE (XEXP (elt, 0)) != REG
> + || GET_MODE (XEXP (elt, 0)) != V2DImode
> + || REGNO (XEXP (elt, 0)) != GET_SSE_REGNO (i))
> return false;
> }
>
> @@ -2157,11 +2157,11 @@ (define_predicate "encodekey256_operation"
> for(i = 4; i < 7; i++)
> {
> elt = XVECEXP (op, 0, i + 1);
> - if (GET_CODE (elt) != SET
> - || GET_CODE (SET_DEST (elt)) != REG
> - || GET_MODE (SET_DEST (elt)) != V2DImode
> - || REGNO (SET_DEST (elt)) != GET_SSE_REGNO (i)
> - || SET_SRC (elt) != CONST0_RTX (V2DImode))
> + if (GET_CODE (elt) != CLOBBER
> + || GET_MODE (elt) != VOIDmode
> + || GET_CODE (XEXP (elt, 0)) != REG
> + || GET_MODE (XEXP (elt, 0)) != V2DImode
> + || REGNO (XEXP (elt, 0)) != GET_SSE_REGNO (i))
> return false;
> }
>
> diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md
> index 5c189635124..076064f97e6 100644
> --- a/gcc/config/i386/sse.md
> +++ b/gcc/config/i386/sse.md
> @@ -29015,7 +29015,7 @@ (define_expand "encodekey128u32"
>
> for (i = 4; i < 7; i++)
> XVECEXP (operands[2], 0, i)
> - = gen_rtx_SET (xmm_regs[i], CONST0_RTX (V2DImode));
> + = gen_rtx_CLOBBER (VOIDmode, xmm_regs[i]);
>
> XVECEXP (operands[2], 0, 7)
> = gen_rtx_CLOBBER (VOIDmode, gen_rtx_REG (CCmode, FLAGS_REG));
> @@ -29072,7 +29072,7 @@ (define_expand "encodekey256u32"
>
> for (i = 4; i < 7; i++)
> XVECEXP (operands[2], 0, i + 1)
> - = gen_rtx_SET (xmm_regs[i], CONST0_RTX (V2DImode));
> + = gen_rtx_CLOBBER (VOIDmode, xmm_regs[i]);
>
> XVECEXP (operands[2], 0, 8)
> = gen_rtx_CLOBBER (VOIDmode, gen_rtx_REG (CCmode, FLAGS_REG));
> diff --git a/gcc/testsuite/gcc.target/i386/keylocker-encodekey128.c b/gcc/testsuite/gcc.target/i386/keylocker-encodekey128.c
> index 805e0628673..57fa9bdc831 100644
> --- a/gcc/testsuite/gcc.target/i386/keylocker-encodekey128.c
> +++ b/gcc/testsuite/gcc.target/i386/keylocker-encodekey128.c
> @@ -6,7 +6,6 @@
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm0,\[^\\n\\r\]*" } } */
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm1,\[^\\n\\r\]*16\[^\\n\\r\]*" } } */
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm2,\[^\\n\\r\]*32\[^\\n\\r\]*" } } */
> -/* { dg-final { scan-assembler "(?:movdqa|movaps)\[ \\t\]+\[^\\n\]*%xmm\[4-6\],\[^\\n\\r\]*" } } */
>
> #include <immintrin.h>
>
> diff --git a/gcc/testsuite/gcc.target/i386/keylocker-encodekey256.c b/gcc/testsuite/gcc.target/i386/keylocker-encodekey256.c
> index 26f04dcf014..a9398b4e7a2 100644
> --- a/gcc/testsuite/gcc.target/i386/keylocker-encodekey256.c
> +++ b/gcc/testsuite/gcc.target/i386/keylocker-encodekey256.c
> @@ -8,7 +8,6 @@
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm1,\[^\\n\\r\]*16\[^\\n\\r\]*" } } */
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm2,\[^\\n\\r\]*32\[^\\n\\r\]*" } } */
> /* { dg-final { scan-assembler "(?:movdqu|movups)\[ \\t\]+\[^\\n\]*%xmm3,\[^\\n\\r\]*48\[^\\n\\r\]*" } } */
> -/* { dg-final { scan-assembler "(?:movdqa|movaps)\[ \\t\]+\[^\\n\]*%xmm\[4-6\],\[^\\n\\r\]*" } } */
>
> #include <immintrin.h>
>
> --
> 2.37.3
>
--
BR,
Hongtao
prev parent reply other threads:[~2022-09-28 1:42 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-27 23:34 H.J. Lu
2022-09-28 1:41 ` Hongtao Liu [this message]
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='CAMZc-bx5vHO6j7Gag3Wp+5oCUMXLMtX=PkhsKqyemaXCbs6a4A@mail.gmail.com' \
--to=crazylht@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=hjl.tools@gmail.com \
--cc=hongtao.liu@intel.com \
--cc=hongyu.wang@intel.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).