From: "Jiang, Haochen" <haochen.jiang@intel.com>
To: Uros Bizjak <ubizjak@gmail.com>
Cc: "gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>,
"Liu, Hongtao" <hongtao.liu@intel.com>
Subject: RE: [PATCH] i386: Add syscall to enable AMX for latest kernels
Date: Tue, 21 Jun 2022 02:23:51 +0000 [thread overview]
Message-ID: <SA1PR11MB59467EBDB99027ABB4866B58ECB39@SA1PR11MB5946.namprd11.prod.outlook.com> (raw)
In-Reply-To: <CAFULd4aZneXR7mzBx5DB26Wa2F_S6ru86bThj+8NyvmLp2Lxwg@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 3262 bytes --]
> -----Original Message-----
> From: Uros Bizjak <ubizjak@gmail.com>
> Sent: Monday, June 20, 2022 10:54 PM
> To: Jiang, Haochen <haochen.jiang@intel.com>
> Cc: gcc-patches@gcc.gnu.org; Liu, Hongtao <hongtao.liu@intel.com>
> Subject: Re: [PATCH] i386: Add syscall to enable AMX for latest kernels
>
> On Mon, Jun 20, 2022 at 10:04 AM Haochen Jiang <haochen.jiang@intel.com>
> wrote:
> >
> > From: "Jiang, Haochen" <haochen.jiang@intel.com>
> >
> > Hi all,
> >
> > We need syscall to enable AMX for kernels>=5.4. It is missing in
> > current amx tests, which will cause test fail.
>
> So this new code is only valid for linux & co?
Thanks for reminding me for that, I only test on linux since the header file is only in linux.
Just updated a patch wrapping with a macro not to change the behavior on windows.
Regtested on x86_64-pc-linux-gnu.
Thx,
Haochen
>
> Uros.
>
> >
> > This patch aims to add them to fix this bug.
> >
> > BRs,
> > Haochen
> >
> > gcc/testsuite/ChangeLog:
> >
> > * gcc.target/i386/amx-check.h (request_perm_xtile_data):
> > New function to check if AMX is usable and enable AMX.
> > (main): Run test if AMX is usable.
> > ---
> > gcc/testsuite/gcc.target/i386/amx-check.h | 24
> > +++++++++++++++++++++++
> > 1 file changed, 24 insertions(+)
> >
> > diff --git a/gcc/testsuite/gcc.target/i386/amx-check.h
> > b/gcc/testsuite/gcc.target/i386/amx-check.h
> > index 434b0e59703..92ed8669304 100644
> > --- a/gcc/testsuite/gcc.target/i386/amx-check.h
> > +++ b/gcc/testsuite/gcc.target/i386/amx-check.h
> > @@ -4,11 +4,22 @@
> > #include <stdlib.h>
> > #include <string.h>
> > #include <stdint.h>
> > +#include <unistd.h>
> > +#include <sys/syscall.h>
> > #ifdef DEBUG
> > #include <stdio.h>
> > #endif
> > #include "cpuid.h"
> >
> > +#define XFEATURE_XTILECFG 17
> > +#define XFEATURE_XTILEDATA 18
> > +#define XFEATURE_MASK_XTILECFG (1 << XFEATURE_XTILECFG)
> > +#define XFEATURE_MASK_XTILEDATA (1 << XFEATURE_XTILEDATA)
> > +#define XFEATURE_MASK_XTILE (XFEATURE_MASK_XTILECFG |
> XFEATURE_MASK_XTILEDATA)
> > +
> > +#define ARCH_GET_XCOMP_PERM 0x1022
> > +#define ARCH_REQ_XCOMP_PERM 0x1023
> > +
> > /* TODO: The tmm emulation is temporary for current
> > AMX implementation with no tmm regclass, should
> > be changed in the future. */
> > @@ -44,6 +55,18 @@ typedef struct __tile
> > /* Stride (colum width in byte) used for tileload/store */ #define
> > _STRIDE 64
> >
> > +/* We need syscall to use amx functions */ int
> > +request_perm_xtile_data() {
> > + unsigned long bitmask;
> > +
> > + if (syscall (SYS_arch_prctl, ARCH_REQ_XCOMP_PERM,
> XFEATURE_XTILEDATA) ||
> > + syscall (SYS_arch_prctl, ARCH_GET_XCOMP_PERM, &bitmask))
> > + return 0;
> > +
> > + return (bitmask & XFEATURE_MASK_XTILE) != 0; }
> > +
> > /* Initialize tile config by setting all tmm size to 16x64 */ void
> > init_tile_config (__tilecfg_u *dst) { @@ -186,6 +209,7 @@ main ()
> > #ifdef AMX_BF16
> > && __builtin_cpu_supports ("amx-bf16") #endif
> > + && request_perm_xtile_data ()
> > )
> > {
> > DO_TEST ();
> > --
> > 2.18.2
> >
[-- Attachment #2: 0001-i386-Add-syscall-to-enable-AMX-for-latest-kernels.patch --]
[-- Type: application/octet-stream, Size: 2201 bytes --]
From 7b2ad89821e8c318f9c592a339389aa60325ad6b Mon Sep 17 00:00:00 2001
From: Haochen Jiang <haochen.jiang@intel.com>
Date: Thu, 16 Jun 2022 00:15:53 -0700
Subject: [PATCH] i386: Add syscall to enable AMX for latest kernels
gcc/testsuite/ChangeLog:
* gcc.target/i386/amx-check.h (request_perm_xtile_data):
New function to check if AMX is usable and enable AMX.
(main): Run test if AMX is usable.
---
gcc/testsuite/gcc.target/i386/amx-check.h | 30 +++++++++++++++++++++++
1 file changed, 30 insertions(+)
diff --git a/gcc/testsuite/gcc.target/i386/amx-check.h b/gcc/testsuite/gcc.target/i386/amx-check.h
index 434b0e59703..694b25ed2eb 100644
--- a/gcc/testsuite/gcc.target/i386/amx-check.h
+++ b/gcc/testsuite/gcc.target/i386/amx-check.h
@@ -4,11 +4,24 @@
#include <stdlib.h>
#include <string.h>
#include <stdint.h>
+#include <unistd.h>
+#ifdef __unix__
+#include <sys/syscall.h>
+#endif
#ifdef DEBUG
#include <stdio.h>
#endif
#include "cpuid.h"
+#define XFEATURE_XTILECFG 17
+#define XFEATURE_XTILEDATA 18
+#define XFEATURE_MASK_XTILECFG (1 << XFEATURE_XTILECFG)
+#define XFEATURE_MASK_XTILEDATA (1 << XFEATURE_XTILEDATA)
+#define XFEATURE_MASK_XTILE (XFEATURE_MASK_XTILECFG | XFEATURE_MASK_XTILEDATA)
+
+#define ARCH_GET_XCOMP_PERM 0x1022
+#define ARCH_REQ_XCOMP_PERM 0x1023
+
/* TODO: The tmm emulation is temporary for current
AMX implementation with no tmm regclass, should
be changed in the future. */
@@ -44,6 +57,20 @@ typedef struct __tile
/* Stride (colum width in byte) used for tileload/store */
#define _STRIDE 64
+#ifdef __unix__
+/* We need syscall to use amx functions */
+int request_perm_xtile_data()
+{
+ unsigned long bitmask;
+
+ if (syscall (SYS_arch_prctl, ARCH_REQ_XCOMP_PERM, XFEATURE_XTILEDATA) ||
+ syscall (SYS_arch_prctl, ARCH_GET_XCOMP_PERM, &bitmask))
+ return 0;
+
+ return (bitmask & XFEATURE_MASK_XTILE) != 0;
+}
+#endif
+
/* Initialize tile config by setting all tmm size to 16x64 */
void init_tile_config (__tilecfg_u *dst)
{
@@ -185,6 +212,9 @@ main ()
#endif
#ifdef AMX_BF16
&& __builtin_cpu_supports ("amx-bf16")
+#endif
+#ifdef __unix__
+ && request_perm_xtile_data ()
#endif
)
{
--
2.18.2
next prev parent reply other threads:[~2022-06-21 2:23 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-20 8:04 Haochen Jiang
2022-06-20 14:53 ` Uros Bizjak
2022-06-21 2:23 ` Jiang, Haochen [this message]
2022-06-21 7:06 ` Uros Bizjak
2022-06-21 7:41 ` Jiang, Haochen
2022-06-21 14:52 ` Uros Bizjak
2022-09-22 6:23 ` Jiang, Haochen
2022-09-22 6:23 ` Liu, Hongtao
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=SA1PR11MB59467EBDB99027ABB4866B58ECB39@SA1PR11MB5946.namprd11.prod.outlook.com \
--to=haochen.jiang@intel.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=hongtao.liu@intel.com \
--cc=ubizjak@gmail.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).