powerpc: Add multiarch sqrtf128 for ppc64le
Commit Message
This patch creates ifunc for sqrtf128() to make use of new xssqrtqp
instruction for POWER9 when --enable-multi-arch and --with-cpu=power8
options are used on power9 system. This is achieved by explicitly
adding -mcpu=power9 flag for sqrtf128-power9.
2018-05-17 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
* sysdeps/powerpc/powerpc64/le/fpu/multiarch/Makefile: New file to
add w_sqrtf128-power9 and w_sqrtf128-ppc64le to libm-sysdep_routines.
* sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c:
New file.
* sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c:
Likewise.
* sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c: Likewise.
* sysdeps/powerpc/powerpc64/le/power9/fpu/e_sqrtf128.c: Remove file
as __ieee754_sqrtf128() for power9 is present already in
sysdeps/powerpc/fpu/math_private.h.
---
.../powerpc/powerpc64/le/fpu/multiarch/Makefile | 6 ++++
.../multiarch/w_sqrtf128-power9.c} | 21 ++++++------
.../le/fpu/multiarch/w_sqrtf128-ppc64le.c | 37 ++++++++++++++++++++++
.../powerpc64/le/fpu/multiarch/w_sqrtf128.c | 31 ++++++++++++++++++
4 files changed, 85 insertions(+), 10 deletions(-)
create mode 100644 sysdeps/powerpc/powerpc64/le/fpu/multiarch/Makefile
rename sysdeps/powerpc/powerpc64/le/{power9/fpu/e_sqrtf128.c => fpu/multiarch/w_sqrtf128-power9.c} (81%)
create mode 100644 sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c
create mode 100644 sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c
Comments
On Thu, 17 May 2018, Rajalakshmi Srinivasaraghavan wrote:
>This patch creates ifunc for sqrtf128() to make use of new xssqrtqp
>instruction for POWER9 when --enable-multi-arch and --with-cpu=power8
>options are used on power9 system. This is achieved by explicitly
>adding -mcpu=power9 flag for sqrtf128-power9.
>
>2018-05-17 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
>
> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/Makefile: New file to
> add w_sqrtf128-power9 and w_sqrtf128-ppc64le to libm-sysdep_routines.
> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c:
> New file.
> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c:
> Likewise.
> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c: Likewise.
This part looks good to me.
> * sysdeps/powerpc/powerpc64/le/power9/fpu/e_sqrtf128.c: Remove file
> as __ieee754_sqrtf128() for power9 is present already in
> sysdeps/powerpc/fpu/math_private.h.
This doesn't.
The function provided by e_sqrtf128.c (__sqrtf128_finite) is export by
GLIBC and accessible by users when they are building their code with
-ffast-math. When building glibc with --with-cpu=power8, removing this
file would be OK. However, when building glibc with --with-cpu=power9,
__sqrtf128_finite will use software emulation when it could use xssqrtqp.
Since math_private.h is not installed, it will not work for glibc users.
You patch is OK without this removal.
>+++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c
> (...)
>+#include <math-type-macros-float128.h>
> (...)
>+#undef __USE_WRAPPER_TEMPLATE
>+#define __USE_WRAPPER_TEMPLATE 1
Do you actually need this? What behaviour are you trying to override?
The float128 implementation always uses the wrapper templates.
>+++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c
> (...)
>+#include <math-type-macros-float128.h>
> (...)
>+#undef __USE_WRAPPER_TEMPLATE
>+#define __USE_WRAPPER_TEMPLATE 1
Likewise.
>+++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c
> (...)
>+libc_ifunc (__sqrtf128,
>+ (hwcap2 & PPC_FEATURE2_ARCH_3_00)
>+ ? __sqrtf128_power9
>+ : __sqrtf128_ppc64le);
>+declare_mgen_alias (__sqrt, sqrt)
OK.
On 05/29/2018 09:07 PM, Gabriel F. T. Gomes wrote:
> On Thu, 17 May 2018, Rajalakshmi Srinivasaraghavan wrote:
>
>> This patch creates ifunc for sqrtf128() to make use of new xssqrtqp
>> instruction for POWER9 when --enable-multi-arch and --with-cpu=power8
>> options are used on power9 system. This is achieved by explicitly
>> adding -mcpu=power9 flag for sqrtf128-power9.
>>
>> 2018-05-17 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
>>
>> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/Makefile: New file to
>> add w_sqrtf128-power9 and w_sqrtf128-ppc64le to libm-sysdep_routines.
>> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c:
>> New file.
>> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c:
>> Likewise.
>> * sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c: Likewise.
>
> This part looks good to me.
Thanks for the review.
>
>> * sysdeps/powerpc/powerpc64/le/power9/fpu/e_sqrtf128.c: Remove file
>> as __ieee754_sqrtf128() for power9 is present already in
>> sysdeps/powerpc/fpu/math_private.h.
>
> This doesn't.
>
> The function provided by e_sqrtf128.c (__sqrtf128_finite) is export by
> GLIBC and accessible by users when they are building their code with
> -ffast-math. When building glibc with --with-cpu=power8, removing this
> file would be OK. However, when building glibc with --with-cpu=power9,
> __sqrtf128_finite will use software emulation when it could use xssqrtqp.
> Since math_private.h is not installed, it will not work for glibc users.
>
> You patch is OK without this removal.
Good catch. Ack.
>
>> +++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c
>> (...)
>> +#include <math-type-macros-float128.h>
>> (...)
>> +#undef __USE_WRAPPER_TEMPLATE
>> +#define __USE_WRAPPER_TEMPLATE 1
>
> Do you actually need this? What behaviour are you trying to override?
> The float128 implementation always uses the wrapper templates.
Yes. This is not needed.
>
>> +++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-ppc64le.c
>> (...)
>> +#include <math-type-macros-float128.h>
>> (...)
>> +#undef __USE_WRAPPER_TEMPLATE
>> +#define __USE_WRAPPER_TEMPLATE 1
>
> Likewise.
>
>> +++ b/sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128.c
>> (...)
>> +libc_ifunc (__sqrtf128,
>> + (hwcap2 & PPC_FEATURE2_ARCH_3_00)
>> + ? __sqrtf128_power9
>> + : __sqrtf128_ppc64le);
>> +declare_mgen_alias (__sqrt, sqrt)
>
> OK.
Pushed as 2c93fce76a8e672abfdf7771c97be9ea49b7222b.
>
>
new file mode 100644
@@ -0,0 +1,6 @@
+ifeq ($(subdir),math)
+libm-sysdep_routines += w_sqrtf128-power9 w_sqrtf128-ppc64le
+
+CFLAGS-w_sqrtf128-ppc64le.c += -mfloat128
+CFLAGS-w_sqrtf128-power9.c += -mfloat128 -mcpu=power9
+endif
similarity index 81%
rename from sysdeps/powerpc/powerpc64/le/power9/fpu/e_sqrtf128.c
rename to sysdeps/powerpc/powerpc64/le/fpu/multiarch/w_sqrtf128-power9.c
@@ -1,6 +1,5 @@
/* POWER9 sqrt for _Float128
- Return sqrt(a)
- Copyright (C) 2017-2018 Free Software Foundation, Inc.
+ Copyright (C) 2018 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
@@ -26,11 +25,13 @@
License along with the GNU C Library; if not, see
<http://www.gnu.org/licenses/>. */
-__float128
-__ieee754_sqrtf128 (__float128 a)
-{
- __float128 z;
- asm ("xssqrtqp %0,%1" : "=v" (z) : "v" (a));
- return z;
-}
-strong_alias (__ieee754_sqrtf128, __sqrtf128_finite)
+#include <math-type-macros-float128.h>
+
+#define __sqrtf128 __sqrtf128_power9
+
+#undef __USE_WRAPPER_TEMPLATE
+#define __USE_WRAPPER_TEMPLATE 1
+#undef declare_mgen_alias
+#define declare_mgen_alias(a, b)
+
+#include <w_sqrt_template.c>
new file mode 100644
@@ -0,0 +1,37 @@
+/* PPC64LE sqrt for _Float128
+ Copyright (C) 2018 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ In addition to the permissions in the GNU Lesser General Public
+ License, the Free Software Foundation gives you unlimited
+ permission to link the compiled version of this file into
+ combinations with other programs, and to distribute those
+ combinations without any restriction coming from the use of this
+ file. (The Lesser General Public License restrictions do apply in
+ other respects; for example, they cover modification of the file,
+ and distribution when not linked into a combine executable.)
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <http://www.gnu.org/licenses/>. */
+
+#include <math-type-macros-float128.h>
+
+#define __sqrtf128 __sqrtf128_ppc64le
+
+#undef __USE_WRAPPER_TEMPLATE
+#define __USE_WRAPPER_TEMPLATE 1
+#undef declare_mgen_alias
+#define declare_mgen_alias(a, b)
+
+#include <w_sqrt_template.c>
new file mode 100644
@@ -0,0 +1,31 @@
+/* Multiple versions of __sqrtf128.
+ Copyright (C) 2018 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, see
+ <http://www.gnu.org/licenses/>. */
+
+#define NO_MATH_REDIRECT
+#include <math.h>
+#include "init-arch.h"
+#include <math-type-macros-float128.h>
+
+extern __typeof (__sqrtf128) __sqrtf128_ppc64le attribute_hidden;
+extern __typeof (__sqrtf128) __sqrtf128_power9 attribute_hidden;
+
+libc_ifunc (__sqrtf128,
+ (hwcap2 & PPC_FEATURE2_ARCH_3_00)
+ ? __sqrtf128_power9
+ : __sqrtf128_ppc64le);
+declare_mgen_alias (__sqrt, sqrt)