Patchwork [x86_64,BZ,#20139] Don't allow configure with not supporting AVX512 assembler w/o --disable-avx512.

login
register
mail settings
Submitter H.J. Lu
Date June 30, 2016, 12:23 p.m.
Message ID <CAMe9rOp3m68KQPFkDCu+ckETGGeQjEot=b7v0zY9g-bPPVk6eg@mail.gmail.com>
Download mbox | patch
Permalink /patch/13506/
State New
Headers show

Comments

H.J. Lu - June 30, 2016, 12:23 p.m.
On Thu, Jun 30, 2016 at 4:15 AM, Florian Weimer <fweimer@redhat.com> wrote:
> On 06/29/2016 11:45 PM, H.J. Lu wrote:
>>
>> Subject: [PATCH] Require binutils 2.24 to build x86-64 glibc
>
>
> Please put [BZ #20139] at the end of the first line.

What do you mean by that?  My ChangeLog entries are

                 [BZ #20139]
                 * config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
                 (HAVE_AVX512DQ_ASM_SUPPORT): This.

It looks normal to me.

>
>> If assembler doesn't support , _dl_runtime_resolve_avx is used
>> to save the first 8 vector registers, which only saves the lower 256
>> bits of vector register, for lazy binding.  When it is called on AVX512
>> platform, the upper 256 bits of ZMM registers are clobbered.  Parameters
>> passed in ZMM registers will be wrong when the function is called the
>> first time.  This patch requires binutils 2.24, whose assembler can store
>> and load ZMM registers, to build x86-64 glibc.  Since mathvec library
>> needs assembler support for AVX512DQ,  we disable mathvec if assembler
>> doesn't support AVX512DQ.
>
>
> The commit message seems to be inconsistent.  Should the “AVX512DQ” on the
> first line actually be “AVX512F”?
>

Currently we check assembler support for AVX512DQ before enabling AVX512.
If assembler doesn't support AVX512DQ, we won't save ZMM registers even
though AVX512F is sufficient.

Here is the patch with updated commit message, including only disabling
the AVX512DQ part of mathvec without AVX512DQ.

OK for master?
Florian Weimer - July 1, 2016, 6:13 a.m.
On 06/30/2016 02:23 PM, H.J. Lu wrote:
> On Thu, Jun 30, 2016 at 4:15 AM, Florian Weimer <fweimer@redhat.com> wrote:
>> On 06/29/2016 11:45 PM, H.J. Lu wrote:
>>>
>>> Subject: [PATCH] Require binutils 2.24 to build x86-64 glibc
>>
>>
>> Please put [BZ #20139] at the end of the first line.
>
> What do you mean by that?  My ChangeLog entries are
>
>                  [BZ #20139]
>                  * config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
>                  (HAVE_AVX512DQ_ASM_SUPPORT): This.
>
> It looks normal to me.

I meant at the end of the first line of the commit message.

> Here is the patch with updated commit message, including only disabling
> the AVX512DQ part of mathvec without AVX512DQ.
>
> OK for master?

I have no objections to the patch as such (but also see my other comment 
about aligning the XGETBV values with what the dynamic linker supports).

Thanks,
Florian
Andrew Senkevich - July 1, 2016, 12:28 p.m.
2016-07-01 9:13 GMT+03:00 Florian Weimer <fweimer@redhat.com>:
> On 06/30/2016 02:23 PM, H.J. Lu wrote:
>>
>> On Thu, Jun 30, 2016 at 4:15 AM, Florian Weimer <fweimer@redhat.com>
>> wrote:
>>>
>>> On 06/29/2016 11:45 PM, H.J. Lu wrote:
>>>>
>>>>
>>>> Subject: [PATCH] Require binutils 2.24 to build x86-64 glibc
>>>
>>>
>>>
>>> Please put [BZ #20139] at the end of the first line.
>>
>>
>> What do you mean by that?  My ChangeLog entries are
>>
>>                  [BZ #20139]
>>                  * config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
>>                  (HAVE_AVX512DQ_ASM_SUPPORT): This.
>>
>> It looks normal to me.
>
>
> I meant at the end of the first line of the commit message.
>
>> Here is the patch with updated commit message, including only disabling
>> the AVX512DQ part of mathvec without AVX512DQ.
>>
>> OK for master?
>
>
> I have no objections to the patch as such (but also see my other comment
> about aligning the XGETBV values with what the dynamic linker supports).

Have we decided how to fix it in 2.23 release (which contains this bug)?


--
WBR,
Andrew
Florian Weimer - July 1, 2016, 12:40 p.m.
On 07/01/2016 02:28 PM, Andrew Senkevich wrote:

>> I have no objections to the patch as such (but also see my other comment
>> about aligning the XGETBV values with what the dynamic linker supports).
>
> Have we decided how to fix it in 2.23 release (which contains this bug)?

Not to my knowledge.

Please also consider 
<https://sourceware.org/ml/libc-alpha/2016-06/msg01108.html>.

I would like to see a conceptual fix, not just papering over the current 
problem.

Thanks,
Florian
H.J. Lu - July 1, 2016, 12:56 p.m.
On Thu, Jun 30, 2016 at 11:13 PM, Florian Weimer <fweimer@redhat.com> wrote:
> On 06/30/2016 02:23 PM, H.J. Lu wrote:
>>
>> On Thu, Jun 30, 2016 at 4:15 AM, Florian Weimer <fweimer@redhat.com>
>> wrote:
>>>
>>> On 06/29/2016 11:45 PM, H.J. Lu wrote:
>>>>
>>>>
>>>> Subject: [PATCH] Require binutils 2.24 to build x86-64 glibc
>>>
>>>
>>>
>>> Please put [BZ #20139] at the end of the first line.
>>
>>
>> What do you mean by that?  My ChangeLog entries are
>>
>>                  [BZ #20139]
>>                  * config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
>>                  (HAVE_AVX512DQ_ASM_SUPPORT): This.
>>
>> It looks normal to me.
>
>
> I meant at the end of the first line of the commit message.

I will do that.

>> Here is the patch with updated commit message, including only disabling
>> the AVX512DQ part of mathvec without AVX512DQ.
>>
>> OK for master?
>
>
> I have no objections to the patch as such (but also see my other comment
> about aligning the XGETBV values with what the dynamic linker supports).
>

Yes, that was a mistake and we should have done that.
H.J. Lu - July 1, 2016, 12:57 p.m.
On Fri, Jul 1, 2016 at 5:40 AM, Florian Weimer <fweimer@redhat.com> wrote:
> On 07/01/2016 02:28 PM, Andrew Senkevich wrote:
>
>>> I have no objections to the patch as such (but also see my other comment
>>> about aligning the XGETBV values with what the dynamic linker supports).
>>
>>
>> Have we decided how to fix it in 2.23 release (which contains this bug)?
>
>
> Not to my knowledge.
>
> Please also consider
> <https://sourceware.org/ml/libc-alpha/2016-06/msg01108.html>.
>
> I would like to see a conceptual fix, not just papering over the current
> problem.
>

FWIW, glibc 2.23 requires binutils 2.24 for s390.

Patch

From f9bef2d38d8b41fc14e1bbcb7165bc8858d7bba6 Mon Sep 17 00:00:00 2001
From: "H.J. Lu" <hjl.tools@gmail.com>
Date: Wed, 29 Jun 2016 12:37:06 -0700
Subject: [PATCH] Require binutils 2.24 to build x86-64 glibc

Currently we check assembler support for AVX512DQ before enabling AVX512.
If assembler doesn't support AVX512DQ, _dl_runtime_resolve_avx is used
to save the first 8 vector registers, which only saves the lower 256
bits of vector register, for lazy binding.  When it is called on AVX512
platform, the upper 256 bits of ZMM registers are clobbered.  Parameters
passed in ZMM registers will be wrong when the function is called the
first time.  This patch requires binutils 2.24, whose assembler can store
and load ZMM registers with AVX512F, to build x86-64 glibc.  Since the
part of mathvec library needs assembler support for AVX512DQ, we disable
the AVX512DQ portion if assembler doesn't support AVX512DQ.

	[BZ #20139]
	* config.h.in (HAVE_AVX512_ASM_SUPPORT): Renamed to ...
	(HAVE_AVX512DQ_ASM_SUPPORT): This.
	* sysdeps/x86_64/configure.ac: Require assembler from binutils
	2.24 or above.
	(HAVE_AVX512_ASM_SUPPORT): Removed.
	(HAVE_AVX512DQ_ASM_SUPPORT): New.
	* sysdeps/x86_64/configure: Regenerated.
	* sysdeps/x86_64/dl-trampoline.S: Make HAVE_AVX512_ASM_SUPPORT
	check unconditional.
	* sysdeps/x86_64/multiarch/ifunc-impl-list.c: Likewise.
	* sysdeps/x86_64/multiarch/memcpy.S: Likewise.
	* sysdeps/x86_64/multiarch/memcpy_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memmove.S: Likewise.
	* sysdeps/x86_64/multiarch/memmove_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/mempcpy.S: Likewise.
	* sysdeps/x86_64/multiarch/mempcpy_chk.S: Likewise.
	* sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S:
	Likewise.
	* sysdeps/x86_64/multiarch/memset.S: Likewise.
	* sysdeps/x86_64/multiarch/memset_chk.S: Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S: Check
	HAVE_AVX512DQ_ASM_SUPPORT instead of HAVE_AVX512_ASM_SUPPORT.
	* sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx51:
	Likewise.
	* sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S:
	Likewise.
---
 config.h.in                                        |  4 +-
 sysdeps/x86_64/configure                           | 89 +++++++++++++++++++---
 sysdeps/x86_64/configure.ac                        | 25 +++---
 sysdeps/x86_64/dl-trampoline.S                     | 37 ++++-----
 .../x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S |  4 +-
 .../x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S |  4 +-
 .../x86_64/fpu/multiarch/svml_d_log8_core_avx512.S |  4 +-
 .../x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S |  4 +-
 .../x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S |  4 +-
 .../fpu/multiarch/svml_d_sincos8_core_avx512.S     |  4 +-
 .../fpu/multiarch/svml_s_cosf16_core_avx512.S      |  4 +-
 .../fpu/multiarch/svml_s_expf16_core_avx512.S      |  4 +-
 .../fpu/multiarch/svml_s_logf16_core_avx512.S      |  4 +-
 .../fpu/multiarch/svml_s_powf16_core_avx512.S      |  4 +-
 .../fpu/multiarch/svml_s_sincosf16_core_avx512.S   |  4 +-
 .../fpu/multiarch/svml_s_sinf16_core_avx512.S      |  4 +-
 sysdeps/x86_64/multiarch/ifunc-impl-list.c         | 16 ----
 sysdeps/x86_64/multiarch/memcpy.S                  |  2 -
 sysdeps/x86_64/multiarch/memcpy_chk.S              |  2 -
 .../multiarch/memmove-avx512-no-vzeroupper.S       |  2 +-
 .../multiarch/memmove-avx512-unaligned-erms.S      |  2 +-
 sysdeps/x86_64/multiarch/memmove.S                 |  2 -
 sysdeps/x86_64/multiarch/memmove_chk.S             |  2 -
 sysdeps/x86_64/multiarch/mempcpy.S                 |  2 -
 sysdeps/x86_64/multiarch/mempcpy_chk.S             |  2 -
 .../x86_64/multiarch/memset-avx512-no-vzeroupper.S |  2 +-
 .../multiarch/memset-avx512-unaligned-erms.S       |  2 +-
 sysdeps/x86_64/multiarch/memset.S                  |  2 -
 sysdeps/x86_64/multiarch/memset_chk.S              |  2 -
 29 files changed, 138 insertions(+), 105 deletions(-)

diff --git a/config.h.in b/config.h.in
index b28b513..856ef6a 100644
--- a/config.h.in
+++ b/config.h.in
@@ -67,8 +67,8 @@ 
 /* Define if compiler supports AVX512.  */
 #undef  HAVE_AVX512_SUPPORT
 
-/* Define if assembler supports AVX512.  */
-#undef  HAVE_AVX512_ASM_SUPPORT
+/* Define if assembler supports AVX512DQ.  */
+#undef  HAVE_AVX512DQ_ASM_SUPPORT
 
 /* Define if assembler supports vector instructions on S390.  */
 #undef  HAVE_S390_VX_ASM_SUPPORT
diff --git a/sysdeps/x86_64/configure b/sysdeps/x86_64/configure
index 88fbfe4..2d14c34 100644
--- a/sysdeps/x86_64/configure
+++ b/sysdeps/x86_64/configure
@@ -1,13 +1,76 @@ 
 # This file is generated from configure.ac by Autoconf.  DO NOT EDIT!
  # Local configure fragment for sysdeps/x86_64.
 
-{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for AVX512 support in assembler" >&5
-$as_echo_n "checking for AVX512 support in assembler... " >&6; }
-if ${libc_cv_asm_avx512+:} false; then :
+for ac_prog in $AS
+do
+  # Extract the first word of "$ac_prog", so it can be a program name with args.
+set dummy $ac_prog; ac_word=$2
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5
+$as_echo_n "checking for $ac_word... " >&6; }
+if ${ac_cv_prog_AS+:} false; then :
+  $as_echo_n "(cached) " >&6
+else
+  if test -n "$AS"; then
+  ac_cv_prog_AS="$AS" # Let the user override the test.
+else
+as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
+for as_dir in $PATH
+do
+  IFS=$as_save_IFS
+  test -z "$as_dir" && as_dir=.
+    for ac_exec_ext in '' $ac_executable_extensions; do
+  if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
+    ac_cv_prog_AS="$ac_prog"
+    $as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5
+    break 2
+  fi
+done
+  done
+IFS=$as_save_IFS
+
+fi
+fi
+AS=$ac_cv_prog_AS
+if test -n "$AS"; then
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: $AS" >&5
+$as_echo "$AS" >&6; }
+else
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+fi
+
+
+  test -n "$AS" && break
+done
+
+if test -z "$AS"; then
+  ac_verc_fail=yes
+else
+  # Found it, now check the version.
+  { $as_echo "$as_me:${as_lineno-$LINENO}: checking version of $AS" >&5
+$as_echo_n "checking version of $AS... " >&6; }
+  ac_prog_version=`$AS --version 2>&1 | sed -n 's/^.*GNU assembler.* \([0-9]*\.[0-9.]*\).*$/\1/p'`
+  case $ac_prog_version in
+    '') ac_prog_version="v. ?.??, bad"; ac_verc_fail=yes;;
+    2.2[4-9]*|2.[3-9][0-9]*|[3-9].*|[1-9][0-9]*)
+       ac_prog_version="$ac_prog_version, ok"; ac_verc_fail=no;;
+    *) ac_prog_version="$ac_prog_version, bad"; ac_verc_fail=yes;;
+
+  esac
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_prog_version" >&5
+$as_echo "$ac_prog_version" >&6; }
+fi
+if test $ac_verc_fail = yes; then
+  critic_missing="$critic_missing The program AS is required in version >= 2.24 for target x86_64."
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for AVX512DQ support in assembler" >&5
+$as_echo_n "checking for AVX512DQ support in assembler... " >&6; }
+if ${libc_cv_asm_avx512dq+:} false; then :
   $as_echo_n "(cached) " >&6
 else
   cat > conftest.s <<\EOF
-        vmovdqu64 %zmm0, (%rsp)
         vandpd (%rax), %zmm6, %zmm1
 EOF
 if { ac_try='${CC-cc} -c $ASFLAGS conftest.s 1>&5'
@@ -16,16 +79,16 @@  if { ac_try='${CC-cc} -c $ASFLAGS conftest.s 1>&5'
   ac_status=$?
   $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5
   test $ac_status = 0; }; }; then
-  libc_cv_asm_avx512=yes
+  libc_cv_asm_avx512dq=yes
 else
-  libc_cv_asm_avx512=no
+  libc_cv_asm_avx512dq=no
 fi
 rm -f conftest*
 fi
-{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $libc_cv_asm_avx512" >&5
-$as_echo "$libc_cv_asm_avx512" >&6; }
-if test $libc_cv_asm_avx512 = yes; then
-  $as_echo "#define HAVE_AVX512_ASM_SUPPORT 1" >>confdefs.h
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $libc_cv_asm_avx512dq" >&5
+$as_echo "$libc_cv_asm_avx512dq" >&6; }
+if test $libc_cv_asm_avx512dq = yes; then
+  $as_echo "#define HAVE_AVX512DQ_ASM_SUPPORT 1" >>confdefs.h
 
 fi
 
@@ -40,7 +103,7 @@  else
   ac_status=$?
   $as_echo "$as_me:${as_lineno-$LINENO}: \$? = $ac_status" >&5
   test $ac_status = 0; }; }; then :
-  libc_cv_cc_avx512=$libc_cv_asm_avx512
+  libc_cv_cc_avx512=$libc_cv_asm_avx512dq
 else
   libc_cv_cc_avx512=no
 fi
@@ -88,4 +151,6 @@  fi
 
 $as_echo "#define PI_STATIC_AND_HIDDEN 1" >>confdefs.h
 
-# work around problem with autoconf and empty lines at the end of files
+
+test -n "$critic_missing" && as_fn_error $? "
+*** $critic_missing" "$LINENO" 5
diff --git a/sysdeps/x86_64/configure.ac b/sysdeps/x86_64/configure.ac
index b39309e..7d8aaaf 100644
--- a/sysdeps/x86_64/configure.ac
+++ b/sysdeps/x86_64/configure.ac
@@ -1,25 +1,30 @@ 
 GLIBC_PROVIDES dnl See aclocal.m4 in the top level source directory.
 # Local configure fragment for sysdeps/x86_64.
 
-dnl Check if asm supports AVX512.
-AC_CACHE_CHECK(for AVX512 support in assembler, libc_cv_asm_avx512, [dnl
+dnl Accept as 2.24 or newer for AVX512 load and store.
+AC_CHECK_PROG_VER(AS, $AS, --version,
+		  [GNU assembler.* \([0-9]*\.[0-9.]*\)],
+		  [2.2[4-9]*|2.[3-9][0-9]*|[3-9].*|[1-9][0-9]*],
+		  critic_missing="$critic_missing The program AS is required in version >= 2.24 for target x86_64.")
+
+dnl Check if asm supports AVX512DQ.
+AC_CACHE_CHECK(for AVX512DQ support in assembler, libc_cv_asm_avx512dq, [dnl
 cat > conftest.s <<\EOF
-        vmovdqu64 %zmm0, (%rsp)
         vandpd (%rax), %zmm6, %zmm1
 EOF
 if AC_TRY_COMMAND(${CC-cc} -c $ASFLAGS conftest.s 1>&AS_MESSAGE_LOG_FD); then
-  libc_cv_asm_avx512=yes
+  libc_cv_asm_avx512dq=yes
 else
-  libc_cv_asm_avx512=no
+  libc_cv_asm_avx512dq=no
 fi
 rm -f conftest*])
-if test $libc_cv_asm_avx512 = yes; then
-  AC_DEFINE(HAVE_AVX512_ASM_SUPPORT)
+if test $libc_cv_asm_avx512dq = yes; then
+  AC_DEFINE(HAVE_AVX512DQ_ASM_SUPPORT)
 fi
 
 dnl Check if -mavx512f works.
 AC_CACHE_CHECK(for AVX512 support, libc_cv_cc_avx512, [dnl
-LIBC_TRY_CC_OPTION([-mavx512f], [libc_cv_cc_avx512=$libc_cv_asm_avx512], [libc_cv_cc_avx512=no])
+LIBC_TRY_CC_OPTION([-mavx512f], [libc_cv_cc_avx512=$libc_cv_asm_avx512dq], [libc_cv_cc_avx512=no])
 ])
 if test $libc_cv_cc_avx512 = yes; then
   AC_DEFINE(HAVE_AVX512_SUPPORT)
@@ -48,4 +53,6 @@  fi
 dnl It is always possible to access static and hidden symbols in an
 dnl position independent way.
 AC_DEFINE(PI_STATIC_AND_HIDDEN)
-# work around problem with autoconf and empty lines at the end of files
+
+test -n "$critic_missing" && AC_MSG_ERROR([
+*** $critic_missing])
diff --git a/sysdeps/x86_64/dl-trampoline.S b/sysdeps/x86_64/dl-trampoline.S
index 39b8771..12f1a5c 100644
--- a/sysdeps/x86_64/dl-trampoline.S
+++ b/sysdeps/x86_64/dl-trampoline.S
@@ -77,30 +77,23 @@ 
 
 #define RESTORE_AVX
 
-#ifdef HAVE_AVX512_ASM_SUPPORT
-# define VEC_SIZE		64
-# define VMOVA			vmovdqa64
-# if DL_RUNTIME_RESOLVE_REALIGN_STACK || VEC_SIZE <= DL_STACK_ALIGNMENT
-#  define VMOV			vmovdqa64
-# else
-#  define VMOV			vmovdqu64
-# endif
-# define VEC(i)			zmm##i
-# define _dl_runtime_resolve	_dl_runtime_resolve_avx512
-# define _dl_runtime_profile	_dl_runtime_profile_avx512
-# include "dl-trampoline.h"
-# undef _dl_runtime_resolve
-# undef _dl_runtime_profile
-# undef VEC
-# undef VMOV
-# undef VMOVA
-# undef VEC_SIZE
+#define VEC_SIZE		64
+#define VMOVA			vmovdqa64
+#if DL_RUNTIME_RESOLVE_REALIGN_STACK || VEC_SIZE <= DL_STACK_ALIGNMENT
+# define VMOV			vmovdqa64
 #else
-strong_alias (_dl_runtime_resolve_avx, _dl_runtime_resolve_avx512)
-	.hidden _dl_runtime_resolve_avx512
-strong_alias (_dl_runtime_profile_avx, _dl_runtime_profile_avx512)
-	.hidden _dl_runtime_profile_avx512
+# define VMOV			vmovdqu64
 #endif
+#define VEC(i)			zmm##i
+#define _dl_runtime_resolve	_dl_runtime_resolve_avx512
+#define _dl_runtime_profile	_dl_runtime_profile_avx512
+#include "dl-trampoline.h"
+#undef _dl_runtime_resolve
+#undef _dl_runtime_profile
+#undef VEC
+#undef VMOV
+#undef VMOVA
+#undef VEC_SIZE
 
 #define VEC_SIZE		32
 #define VMOVA			vmovdqa
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S
index 874bd80..91e92e9 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_cos8_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8v_cos_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_cos
 #else
 /*
@@ -236,7 +236,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN4v_cos
 END (_ZGVeN8v_cos_knl)
 
 ENTRY (_ZGVeN8v_cos_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_cos
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S
index 456792d..ea84091 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_exp8_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8v_exp_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_exp
 #else
 /*
@@ -238,7 +238,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN4v_exp
 END (_ZGVeN8v_exp_knl)
 
 ENTRY (_ZGVeN8v_exp_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_exp
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S
index 4c52a91..62854bb 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_log8_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8v_log_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_log
 #else
 /*
@@ -237,7 +237,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN4v_log
 END (_ZGVeN8v_log_knl)
 
 ENTRY (_ZGVeN8v_log_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_log
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S
index fd6a889..c6b6474 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_pow8_core_avx512.S
@@ -82,7 +82,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8vv_pow_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_ff _ZGVdN4vv_pow
 #else
         pushq     %rbp
@@ -409,7 +409,7 @@  WRAPPER_IMPL_AVX512_ff _ZGVdN4vv_pow
 END (_ZGVeN8vv_pow_knl)
 
 ENTRY (_ZGVeN8vv_pow_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_ff _ZGVdN4vv_pow
 #else
         pushq     %rbp
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S
index d3449e3..0bb2008 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_sin8_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8v_sin_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_sin
 #else
 /*
@@ -237,7 +237,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN4v_sin
 END (_ZGVeN8v_sin_knl)
 
 ENTRY (_ZGVeN8v_sin_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN4v_sin
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.S
index 1d9f426..8b10275 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_sincos8_core_avx512.S
@@ -37,7 +37,7 @@ 
 
 	.text
 ENTRY (_ZGVeN8vvv_sincos_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_fFF _ZGVdN4vvv_sincos
 #else
         pushq     %rbp
@@ -307,7 +307,7 @@  WRAPPER_IMPL_AVX512_fFF _ZGVdN4vvv_sincos
 END (_ZGVeN8vvv_sincos_knl)
 
 ENTRY (_ZGVeN8vvv_sincos_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_fFF _ZGVdN4vvv_sincos
 #else
         pushq     %rbp
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S
index b39ec3a..ca079a7 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_cosf16_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN16v_cosf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_cosf
 #else
 /*
@@ -239,7 +239,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN8v_cosf
 END (_ZGVeN16v_cosf_knl)
 
 ENTRY (_ZGVeN16v_cosf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_cosf
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S
index 44f61a2..18b8a5e 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_expf16_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN16v_expf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_expf
 #else
 /*
@@ -227,7 +227,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN8v_expf
 END (_ZGVeN16v_expf_knl)
 
 ENTRY (_ZGVeN16v_expf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_expf
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S
index 8d57e65..c714258 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_logf16_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY (_ZGVeN16v_logf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_logf
 #else
 /*
@@ -211,7 +211,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN8v_logf
 END (_ZGVeN16v_logf_knl)
 
 ENTRY (_ZGVeN16v_logf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_logf
 #else
 /*
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S
index 299e6ae..8b0c256 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_powf16_core_avx512.S
@@ -82,7 +82,7 @@ 
 
 	.text
 ENTRY (_ZGVeN16vv_powf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_ff _ZGVdN8vv_powf
 #else
         pushq     %rbp
@@ -359,7 +359,7 @@  WRAPPER_IMPL_AVX512_ff _ZGVdN8vv_powf
 END (_ZGVeN16vv_powf_knl)
 
 ENTRY (_ZGVeN16vv_powf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_ff _ZGVdN8vv_powf
 #else
         pushq     %rbp
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S
index e375de8..b3e8515 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_sincosf16_core_avx512.S
@@ -50,7 +50,7 @@ 
 
 	.text
 ENTRY (_ZGVeN16vvv_sincosf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_fFF _ZGVdN8vvv_sincosf
 #else
         pushq     %rbp
@@ -270,7 +270,7 @@  WRAPPER_IMPL_AVX512_fFF _ZGVdN8vvv_sincosf
 END (_ZGVeN16vvv_sincosf_knl)
 
 ENTRY (_ZGVeN16vvv_sincosf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512_fFF _ZGVdN8vvv_sincosf
 #else
         pushq     %rbp
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S
index 121714f..530d143 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_s_sinf16_core_avx512.S
@@ -22,7 +22,7 @@ 
 
 	.text
 ENTRY(_ZGVeN16v_sinf_knl)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_sinf
 #else
 /*
@@ -243,7 +243,7 @@  WRAPPER_IMPL_AVX512 _ZGVdN8v_sinf
 END(_ZGVeN16v_sinf_knl)
 
 ENTRY (_ZGVeN16v_sinf_skx)
-#ifndef HAVE_AVX512_ASM_SUPPORT
+#ifndef HAVE_AVX512DQ_ASM_SUPPORT
 WRAPPER_IMPL_AVX512 _ZGVdN8v_sinf
 #else
 /*
diff --git a/sysdeps/x86_64/multiarch/ifunc-impl-list.c b/sysdeps/x86_64/multiarch/ifunc-impl-list.c
index 449b046..a443e41 100644
--- a/sysdeps/x86_64/multiarch/ifunc-impl-list.c
+++ b/sysdeps/x86_64/multiarch/ifunc-impl-list.c
@@ -48,7 +48,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 
   /* Support sysdeps/x86_64/multiarch/memmove_chk.c.  */
   IFUNC_IMPL (i, name, __memmove_chk,
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, __memmove_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memmove_chk_avx512_no_vzeroupper)
@@ -58,7 +57,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, __memmove_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memmove_chk_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, __memmove_chk,
 			      HAS_ARCH_FEATURE (AVX_Usable),
 			      __memmove_chk_avx_unaligned)
@@ -84,7 +82,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, memmove,
 			      HAS_ARCH_FEATURE (AVX_Usable),
 			      __memmove_avx_unaligned_erms)
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, memmove,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memmove_avx512_no_vzeroupper)
@@ -94,7 +91,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, memmove,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memmove_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, memmove, HAS_CPU_FEATURE (SSSE3),
 			      __memmove_ssse3_back)
 	      IFUNC_IMPL_ADD (array, i, memmove, HAS_CPU_FEATURE (SSSE3),
@@ -117,7 +113,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, __memset_chk,
 			      HAS_ARCH_FEATURE (AVX2_Usable),
 			      __memset_chk_avx2_unaligned_erms)
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, __memset_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memset_chk_avx512_unaligned_erms)
@@ -127,7 +122,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, __memset_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memset_chk_avx512_no_vzeroupper)
-#endif
 	      )
 
   /* Support sysdeps/x86_64/multiarch/memset.S.  */
@@ -143,7 +137,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, memset,
 			      HAS_ARCH_FEATURE (AVX2_Usable),
 			      __memset_avx2_unaligned_erms)
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, memset,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memset_avx512_unaligned_erms)
@@ -153,7 +146,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, memset,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memset_avx512_no_vzeroupper)
-#endif
 	     )
 
   /* Support sysdeps/x86_64/multiarch/stpncpy.S.  */
@@ -311,7 +303,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 #ifdef SHARED
   /* Support sysdeps/x86_64/multiarch/memcpy_chk.S.  */
   IFUNC_IMPL (i, name, __memcpy_chk,
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, __memcpy_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memcpy_chk_avx512_no_vzeroupper)
@@ -321,7 +312,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, __memcpy_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memcpy_chk_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, __memcpy_chk,
 			      HAS_ARCH_FEATURE (AVX_Usable),
 			      __memcpy_chk_avx_unaligned)
@@ -351,7 +341,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 			      __memcpy_ssse3_back)
 	      IFUNC_IMPL_ADD (array, i, memcpy, HAS_CPU_FEATURE (SSSE3),
 			      __memcpy_ssse3)
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, memcpy,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memcpy_avx512_no_vzeroupper)
@@ -361,7 +350,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, memcpy,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __memcpy_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, memcpy, 1, __memcpy_sse2_unaligned)
 	      IFUNC_IMPL_ADD (array, i, memcpy, 1,
 			      __memcpy_sse2_unaligned_erms)
@@ -369,7 +357,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 
   /* Support sysdeps/x86_64/multiarch/mempcpy_chk.S.  */
   IFUNC_IMPL (i, name, __mempcpy_chk,
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, __mempcpy_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __mempcpy_chk_avx512_no_vzeroupper)
@@ -379,7 +366,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, __mempcpy_chk,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __mempcpy_chk_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, __mempcpy_chk,
 			      HAS_ARCH_FEATURE (AVX_Usable),
 			      __mempcpy_chk_avx_unaligned)
@@ -399,7 +385,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 
   /* Support sysdeps/x86_64/multiarch/mempcpy.S.  */
   IFUNC_IMPL (i, name, mempcpy,
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	      IFUNC_IMPL_ADD (array, i, mempcpy,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __mempcpy_avx512_no_vzeroupper)
@@ -409,7 +394,6 @@  __libc_ifunc_impl_list (const char *name, struct libc_ifunc_impl *array,
 	      IFUNC_IMPL_ADD (array, i, mempcpy,
 			      HAS_ARCH_FEATURE (AVX512F_Usable),
 			      __mempcpy_avx512_unaligned_erms)
-#endif
 	      IFUNC_IMPL_ADD (array, i, mempcpy,
 			      HAS_ARCH_FEATURE (AVX_Usable),
 			      __mempcpy_avx_unaligned)
diff --git a/sysdeps/x86_64/multiarch/memcpy.S b/sysdeps/x86_64/multiarch/memcpy.S
index f6771a4..4aec67d 100644
--- a/sysdeps/x86_64/multiarch/memcpy.S
+++ b/sysdeps/x86_64/multiarch/memcpy.S
@@ -29,7 +29,6 @@ 
 ENTRY(__new_memcpy)
 	.type	__new_memcpy, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__memcpy_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -40,7 +39,6 @@  ENTRY(__new_memcpy)
 	jnz	2f
 	lea	__memcpy_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__memcpy_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/memcpy_chk.S b/sysdeps/x86_64/multiarch/memcpy_chk.S
index 11f1310..9d92c8a 100644
--- a/sysdeps/x86_64/multiarch/memcpy_chk.S
+++ b/sysdeps/x86_64/multiarch/memcpy_chk.S
@@ -30,7 +30,6 @@ 
 ENTRY(__memcpy_chk)
 	.type	__memcpy_chk, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__memcpy_chk_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -41,7 +40,6 @@  ENTRY(__memcpy_chk)
 	jnz	2f
 	lea	__memcpy_chk_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__memcpy_chk_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S b/sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S
index 5b8ff57..664b74d 100644
--- a/sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S
+++ b/sysdeps/x86_64/multiarch/memmove-avx512-no-vzeroupper.S
@@ -18,7 +18,7 @@ 
 
 #include <sysdep.h>
 
-#if defined HAVE_AVX512_ASM_SUPPORT && IS_IN (libc)
+#if IS_IN (libc)
 
 # include "asm-syntax.h"
 
diff --git a/sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S b/sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S
index f9af6fd..aac1515 100644
--- a/sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S
+++ b/sysdeps/x86_64/multiarch/memmove-avx512-unaligned-erms.S
@@ -1,4 +1,4 @@ 
-#if defined HAVE_AVX512_ASM_SUPPORT && IS_IN (libc)
+#if IS_IN (libc)
 # define VEC_SIZE	64
 # define VEC(i)		zmm##i
 # define VMOVNT		vmovntdq
diff --git a/sysdeps/x86_64/multiarch/memmove.S b/sysdeps/x86_64/multiarch/memmove.S
index 25c3586..11ebf6e 100644
--- a/sysdeps/x86_64/multiarch/memmove.S
+++ b/sysdeps/x86_64/multiarch/memmove.S
@@ -27,7 +27,6 @@ 
 ENTRY(__libc_memmove)
 	.type	__libc_memmove, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__memmove_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -38,7 +37,6 @@  ENTRY(__libc_memmove)
 	jnz	2f
 	lea	__memmove_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__memmove_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/memmove_chk.S b/sysdeps/x86_64/multiarch/memmove_chk.S
index cd639b8..7f86120 100644
--- a/sysdeps/x86_64/multiarch/memmove_chk.S
+++ b/sysdeps/x86_64/multiarch/memmove_chk.S
@@ -29,7 +29,6 @@ 
 ENTRY(__memmove_chk)
 	.type	__memmove_chk, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__memmove_chk_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -40,7 +39,6 @@  ENTRY(__memmove_chk)
 	jnz	2f
 	lea	__memmove_chk_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__memmove_chk_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/mempcpy.S b/sysdeps/x86_64/multiarch/mempcpy.S
index f9c6df3..985fdfa 100644
--- a/sysdeps/x86_64/multiarch/mempcpy.S
+++ b/sysdeps/x86_64/multiarch/mempcpy.S
@@ -29,7 +29,6 @@ 
 ENTRY(__mempcpy)
 	.type	__mempcpy, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__mempcpy_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -40,7 +39,6 @@  ENTRY(__mempcpy)
 	jnz	2f
 	lea	__mempcpy_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__mempcpy_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/mempcpy_chk.S b/sysdeps/x86_64/multiarch/mempcpy_chk.S
index 80f460f..9e49f6f 100644
--- a/sysdeps/x86_64/multiarch/mempcpy_chk.S
+++ b/sysdeps/x86_64/multiarch/mempcpy_chk.S
@@ -30,7 +30,6 @@ 
 ENTRY(__mempcpy_chk)
 	.type	__mempcpy_chk, @gnu_indirect_function
 	LOAD_RTLD_GLOBAL_RO_RDX
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	1f
 	lea	__mempcpy_chk_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -41,7 +40,6 @@  ENTRY(__mempcpy_chk)
 	jnz	2f
 	lea	__mempcpy_chk_avx512_unaligned(%rip), %RAX_LP
 	ret
-# endif
 1:	lea	__mempcpy_chk_avx_unaligned(%rip), %RAX_LP
 	HAS_ARCH_FEATURE (AVX_Fast_Unaligned_Load)
 	jz	L(Fast_Unaligned_Load)
diff --git a/sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S b/sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S
index eab8c5a..9687df0 100644
--- a/sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S
+++ b/sysdeps/x86_64/multiarch/memset-avx512-no-vzeroupper.S
@@ -18,7 +18,7 @@ 
 
 #include <sysdep.h>
 
-#if defined HAVE_AVX512_ASM_SUPPORT && IS_IN (libc)
+#if IS_IN (libc)
 
 #include "asm-syntax.h"
 #ifndef MEMSET
diff --git a/sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S b/sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
index f1b3cb2..a5ec349 100644
--- a/sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
+++ b/sysdeps/x86_64/multiarch/memset-avx512-unaligned-erms.S
@@ -1,4 +1,4 @@ 
-#if defined HAVE_AVX512_ASM_SUPPORT && IS_IN (libc)
+#if IS_IN (libc)
 # define VEC_SIZE	64
 # define VEC(i)		zmm##i
 # define VMOVU		vmovdqu64
diff --git a/sysdeps/x86_64/multiarch/memset.S b/sysdeps/x86_64/multiarch/memset.S
index 4e52d8f..df11c16 100644
--- a/sysdeps/x86_64/multiarch/memset.S
+++ b/sysdeps/x86_64/multiarch/memset.S
@@ -38,7 +38,6 @@  ENTRY(memset)
 	jnz	L(AVX512F)
 	lea	__memset_avx2_unaligned(%rip), %RAX_LP
 L(AVX512F):
-# ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	2f
 	lea	__memset_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -48,7 +47,6 @@  L(AVX512F):
 	HAS_CPU_FEATURE (ERMS)
 	jnz	2f
 	lea	__memset_avx512_unaligned(%rip), %RAX_LP
-# endif
 2:	ret
 END(memset)
 #endif
diff --git a/sysdeps/x86_64/multiarch/memset_chk.S b/sysdeps/x86_64/multiarch/memset_chk.S
index 8517cfc..2efe6ed 100644
--- a/sysdeps/x86_64/multiarch/memset_chk.S
+++ b/sysdeps/x86_64/multiarch/memset_chk.S
@@ -38,7 +38,6 @@  ENTRY(__memset_chk)
 	jnz	L(AVX512F)
 	lea	__memset_chk_avx2_unaligned(%rip), %RAX_LP
 L(AVX512F):
-#ifdef HAVE_AVX512_ASM_SUPPORT
 	HAS_ARCH_FEATURE (AVX512F_Usable)
 	jz	2f
 	lea	__memset_chk_avx512_no_vzeroupper(%rip), %RAX_LP
@@ -48,7 +47,6 @@  L(AVX512F):
 	HAS_CPU_FEATURE (ERMS)
 	jnz	2f
 	lea	__memset_chk_avx512_unaligned(%rip), %RAX_LP
-#endif
 2:	ret
 END(__memset_chk)
 
-- 
2.7.4