2014-06-20 20:00:00 +01:00
|
|
|
/* Copyright (C) 1995-1998 Eric Young (eay@cryptsoft.com)
|
|
|
|
* All rights reserved.
|
|
|
|
*
|
|
|
|
* This package is an SSL implementation written
|
|
|
|
* by Eric Young (eay@cryptsoft.com).
|
|
|
|
* The implementation was written so as to conform with Netscapes SSL.
|
|
|
|
*
|
|
|
|
* This library is free for commercial and non-commercial use as long as
|
|
|
|
* the following conditions are aheared to. The following conditions
|
|
|
|
* apply to all code found in this distribution, be it the RC4, RSA,
|
|
|
|
* lhash, DES, etc., code; not just the SSL code. The SSL documentation
|
|
|
|
* included with this distribution is covered by the same copyright terms
|
|
|
|
* except that the holder is Tim Hudson (tjh@cryptsoft.com).
|
|
|
|
*
|
|
|
|
* Copyright remains Eric Young's, and as such any Copyright notices in
|
|
|
|
* the code are not to be removed.
|
|
|
|
* If this package is used in a product, Eric Young should be given attribution
|
|
|
|
* as the author of the parts of the library used.
|
|
|
|
* This can be in the form of a textual message at program startup or
|
|
|
|
* in documentation (online or textual) provided with the package.
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
* 1. Redistributions of source code must retain the copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 3. All advertising materials mentioning features or use of this software
|
|
|
|
* must display the following acknowledgement:
|
|
|
|
* "This product includes cryptographic software written by
|
|
|
|
* Eric Young (eay@cryptsoft.com)"
|
|
|
|
* The word 'cryptographic' can be left out if the rouines from the library
|
|
|
|
* being used are not cryptographic related :-).
|
|
|
|
* 4. If you include any Windows specific code (or a derivative thereof) from
|
|
|
|
* the apps directory (application code) you must include an acknowledgement:
|
|
|
|
* "This product includes software written by Tim Hudson (tjh@cryptsoft.com)"
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY ERIC YOUNG ``AS IS'' AND
|
|
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
|
|
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
|
|
* SUCH DAMAGE.
|
|
|
|
*
|
|
|
|
* The licence and distribution terms for any publically available version or
|
|
|
|
* derivative of this code cannot be changed. i.e. this code cannot simply be
|
|
|
|
* copied and put under another distribution licence
|
|
|
|
* [including the GNU Public Licence.] */
|
|
|
|
|
|
|
|
#include <openssl/cpu.h>
|
|
|
|
|
|
|
|
|
2014-09-19 18:15:33 +01:00
|
|
|
#if !defined(OPENSSL_NO_ASM) && (defined(OPENSSL_X86) || defined(OPENSSL_X86_64))
|
2014-06-20 20:00:00 +01:00
|
|
|
|
|
|
|
#include <inttypes.h>
|
2015-01-31 01:08:37 +00:00
|
|
|
#include <stdio.h>
|
2016-03-16 23:53:34 +00:00
|
|
|
#include <stdlib.h>
|
2015-01-31 01:08:37 +00:00
|
|
|
#include <string.h>
|
2014-06-20 20:00:00 +01:00
|
|
|
|
2017-08-16 20:25:27 +01:00
|
|
|
#if defined(_MSC_VER)
|
2016-06-09 21:48:33 +01:00
|
|
|
OPENSSL_MSVC_PRAGMA(warning(push, 3))
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
#include <immintrin.h>
|
|
|
|
#include <intrin.h>
|
2016-06-09 21:48:33 +01:00
|
|
|
OPENSSL_MSVC_PRAGMA(warning(pop))
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
#endif
|
|
|
|
|
2016-03-19 00:28:36 +00:00
|
|
|
#include "internal.h"
|
|
|
|
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// OPENSSL_cpuid runs the cpuid instruction. |leaf| is passed in as EAX and ECX
|
|
|
|
// is set to zero. It writes EAX, EBX, ECX, and EDX to |*out_eax| through
|
|
|
|
// |*out_edx|.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
static void OPENSSL_cpuid(uint32_t *out_eax, uint32_t *out_ebx,
|
|
|
|
uint32_t *out_ecx, uint32_t *out_edx, uint32_t leaf) {
|
2017-08-16 20:25:27 +01:00
|
|
|
#if defined(_MSC_VER)
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
int tmp[4];
|
|
|
|
__cpuid(tmp, (int)leaf);
|
|
|
|
*out_eax = (uint32_t)tmp[0];
|
|
|
|
*out_ebx = (uint32_t)tmp[1];
|
|
|
|
*out_ecx = (uint32_t)tmp[2];
|
|
|
|
*out_edx = (uint32_t)tmp[3];
|
|
|
|
#elif defined(__pic__) && defined(OPENSSL_32_BIT)
|
2017-08-18 19:06:02 +01:00
|
|
|
// Inline assembly may not clobber the PIC register. For 32-bit, this is EBX.
|
|
|
|
// See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=47602.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
__asm__ volatile (
|
|
|
|
"xor %%ecx, %%ecx\n"
|
|
|
|
"mov %%ebx, %%edi\n"
|
|
|
|
"cpuid\n"
|
|
|
|
"xchg %%edi, %%ebx\n"
|
|
|
|
: "=a"(*out_eax), "=D"(*out_ebx), "=c"(*out_ecx), "=d"(*out_edx)
|
|
|
|
: "a"(leaf)
|
|
|
|
);
|
|
|
|
#else
|
|
|
|
__asm__ volatile (
|
|
|
|
"xor %%ecx, %%ecx\n"
|
|
|
|
"cpuid\n"
|
|
|
|
: "=a"(*out_eax), "=b"(*out_ebx), "=c"(*out_ecx), "=d"(*out_edx)
|
|
|
|
: "a"(leaf)
|
|
|
|
);
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// OPENSSL_xgetbv returns the value of an Intel Extended Control Register (XCR).
|
|
|
|
// Currently only XCR0 is defined by Intel so |xcr| should always be zero.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
static uint64_t OPENSSL_xgetbv(uint32_t xcr) {
|
2017-08-16 20:25:27 +01:00
|
|
|
#if defined(_MSC_VER)
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
return (uint64_t)_xgetbv(xcr);
|
|
|
|
#else
|
|
|
|
uint32_t eax, edx;
|
|
|
|
__asm__ volatile ("xgetbv" : "=a"(eax), "=d"(edx) : "c"(xcr));
|
|
|
|
return (((uint64_t)edx) << 32) | eax;
|
|
|
|
#endif
|
|
|
|
}
|
2014-06-20 20:00:00 +01:00
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// handle_cpu_env applies the value from |in| to the CPUID values in |out[0]|
|
|
|
|
// and |out[1]|. See the comment in |OPENSSL_cpuid_setup| about this.
|
2014-06-20 20:00:00 +01:00
|
|
|
static void handle_cpu_env(uint32_t *out, const char *in) {
|
|
|
|
const int invert = in[0] == '~';
|
|
|
|
uint64_t v;
|
|
|
|
|
2016-01-17 19:10:07 +00:00
|
|
|
if (!sscanf(in + invert, "%" PRIu64, &v)) {
|
2014-06-20 20:00:00 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (invert) {
|
|
|
|
out[0] &= ~v;
|
|
|
|
out[1] &= ~(v >> 32);
|
|
|
|
} else {
|
|
|
|
out[0] = v;
|
|
|
|
out[1] = v >> 32;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void OPENSSL_cpuid_setup(void) {
|
2017-08-18 19:06:02 +01:00
|
|
|
// Determine the vendor and maximum input value.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
uint32_t eax, ebx, ecx, edx;
|
|
|
|
OPENSSL_cpuid(&eax, &ebx, &ecx, &edx, 0);
|
2014-06-20 20:00:00 +01:00
|
|
|
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
uint32_t num_ids = eax;
|
|
|
|
|
|
|
|
int is_intel = ebx == 0x756e6547 /* Genu */ &&
|
|
|
|
edx == 0x49656e69 /* ineI */ &&
|
|
|
|
ecx == 0x6c65746e /* ntel */;
|
|
|
|
int is_amd = ebx == 0x68747541 /* Auth */ &&
|
|
|
|
edx == 0x69746e65 /* enti */ &&
|
|
|
|
ecx == 0x444d4163 /* cAMD */;
|
|
|
|
|
|
|
|
int has_amd_xop = 0;
|
|
|
|
if (is_amd) {
|
2017-08-18 19:06:02 +01:00
|
|
|
// AMD-specific logic.
|
|
|
|
// See http://developer.amd.com/wordpress/media/2012/10/254811.pdf
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
OPENSSL_cpuid(&eax, &ebx, &ecx, &edx, 0x80000000);
|
2015-07-12 17:32:46 +01:00
|
|
|
uint32_t num_extended_ids = eax;
|
|
|
|
if (num_extended_ids >= 0x80000001) {
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
OPENSSL_cpuid(&eax, &ebx, &ecx, &edx, 0x80000001);
|
2017-10-30 00:27:28 +00:00
|
|
|
if (ecx & (1u << 11)) {
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
has_amd_xop = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-04-19 21:34:35 +01:00
|
|
|
uint32_t extended_features[2] = {0};
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
if (num_ids >= 7) {
|
|
|
|
OPENSSL_cpuid(&eax, &ebx, &ecx, &edx, 7);
|
2018-04-19 21:34:35 +01:00
|
|
|
extended_features[0] = ebx;
|
|
|
|
extended_features[1] = ecx;
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
}
|
|
|
|
|
2015-07-12 17:32:46 +01:00
|
|
|
OPENSSL_cpuid(&eax, &ebx, &ecx, &edx, 1);
|
2014-06-20 20:00:00 +01:00
|
|
|
|
Assume hyper-threading-like vulnerabilities are always present.
It's not clear that CPUID will always report the correct value here,
especially for hyper-threading environments. It also isn't clear that
the assumptions made by AMD processors are correct and will always be
correct. It also seems likely that, if a code path is
security-sensitive w.r.t. SMT, it is probably also security-sensitive
w.r.t. other processor (mis)features. Finally, it isn't clear that all
dynamic analysis (fuzzing, SDE, etc.) is done separately for the cross
product of all CPU feature combinations * the value of this bit.
With all that in mind, instruct code sensitive to this bit to always
choose the more conservative path.
I only found one place that's sensitive to this bit, though I didn't
look too hard:
```
aes_nohw_cbc_encrypt:
[...]
leaq OPENSSL_ia32cap_P(%rip),%r10
mov (%r10), %r10d
[...]
bt \$28,%r10d
jc .Lcbc_slow_prologue
```
I didn't verify that the code in the HTT-enabled paths is any better
than the code in the HTT-disabled paths.
Change-Id: Ifd643e6a1301e5ca2174b84c344eb933d49e0067
Reviewed-on: https://boringssl-review.googlesource.com/c/33404
Reviewed-by: David Benjamin <davidben@google.com>
Reviewed-by: Adam Langley <agl@google.com>
Commit-Queue: Adam Langley <agl@google.com>
2018-11-29 22:40:37 +00:00
|
|
|
// Force the hyper-threading bit so that the more conservative path is always
|
|
|
|
// chosen.
|
|
|
|
edx |= 1u << 28;
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// Reserved bit #20 was historically repurposed to control the in-memory
|
|
|
|
// representation of RC4 state. Always set it to zero.
|
2017-10-30 00:27:28 +00:00
|
|
|
edx &= ~(1u << 20);
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// Reserved bit #30 is repurposed to signal an Intel CPU.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
if (is_intel) {
|
2017-10-30 00:27:28 +00:00
|
|
|
edx |= (1u << 30);
|
2017-07-25 21:59:58 +01:00
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// Clear the XSAVE bit on Knights Landing to mimic Silvermont. This enables
|
|
|
|
// some Silvermont-specific codepaths which perform better. See OpenSSL
|
|
|
|
// commit 64d92d74985ebb3d0be58a9718f9e080a14a8e7f.
|
2017-07-25 21:59:58 +01:00
|
|
|
if ((eax & 0x0fff0ff0) == 0x00050670 /* Knights Landing */ ||
|
|
|
|
(eax & 0x0fff0ff0) == 0x00080650 /* Knights Mill (per SDE) */) {
|
2017-10-30 00:27:28 +00:00
|
|
|
ecx &= ~(1u << 26);
|
2017-07-25 21:59:58 +01:00
|
|
|
}
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
} else {
|
2017-10-30 00:27:28 +00:00
|
|
|
edx &= ~(1u << 30);
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
}
|
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// The SDBG bit is repurposed to denote AMD XOP support.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
if (has_amd_xop) {
|
2017-10-30 00:27:28 +00:00
|
|
|
ecx |= (1u << 11);
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
} else {
|
2017-10-30 00:27:28 +00:00
|
|
|
ecx &= ~(1u << 11);
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
uint64_t xcr0 = 0;
|
2017-10-30 00:27:28 +00:00
|
|
|
if (ecx & (1u << 27)) {
|
2017-08-18 19:06:02 +01:00
|
|
|
// XCR0 may only be queried if the OSXSAVE bit is set.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
xcr0 = OPENSSL_xgetbv(0);
|
|
|
|
}
|
2017-08-18 19:06:02 +01:00
|
|
|
// See Intel manual, volume 1, section 14.3.
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
if ((xcr0 & 6) != 6) {
|
2017-08-18 19:06:02 +01:00
|
|
|
// YMM registers cannot be used.
|
2017-10-30 00:27:28 +00:00
|
|
|
ecx &= ~(1u << 28); // AVX
|
|
|
|
ecx &= ~(1u << 12); // FMA
|
|
|
|
ecx &= ~(1u << 11); // AMD XOP
|
2017-08-18 19:06:02 +01:00
|
|
|
// Clear AVX2 and AVX512* bits.
|
|
|
|
//
|
|
|
|
// TODO(davidben): Should bits 17 and 26-28 also be cleared? Upstream
|
|
|
|
// doesn't clear those.
|
2018-04-19 21:34:35 +01:00
|
|
|
extended_features[0] &=
|
2017-10-30 00:27:28 +00:00
|
|
|
~((1u << 5) | (1u << 16) | (1u << 21) | (1u << 30) | (1u << 31));
|
2017-07-25 22:10:07 +01:00
|
|
|
}
|
2017-08-18 19:06:02 +01:00
|
|
|
// See Intel manual, volume 1, section 15.2.
|
2017-07-25 22:10:07 +01:00
|
|
|
if ((xcr0 & 0xe6) != 0xe6) {
|
2017-08-18 19:06:02 +01:00
|
|
|
// Clear AVX512F. Note we don't touch other AVX512 extensions because they
|
|
|
|
// can be used with YMM.
|
2018-04-19 21:34:35 +01:00
|
|
|
extended_features[0] &= ~(1u << 16);
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
}
|
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// Disable ADX instructions on Knights Landing. See OpenSSL commit
|
|
|
|
// 64d92d74985ebb3d0be58a9718f9e080a14a8e7f.
|
2017-10-30 00:27:28 +00:00
|
|
|
if ((ecx & (1u << 26)) == 0) {
|
2018-04-19 21:34:35 +01:00
|
|
|
extended_features[0] &= ~(1u << 19);
|
2017-07-25 21:59:58 +01:00
|
|
|
}
|
|
|
|
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
OPENSSL_ia32cap_P[0] = edx;
|
|
|
|
OPENSSL_ia32cap_P[1] = ecx;
|
2018-04-19 21:34:35 +01:00
|
|
|
OPENSSL_ia32cap_P[2] = extended_features[0];
|
|
|
|
OPENSSL_ia32cap_P[3] = extended_features[1];
|
Replace cpuid assembly with C code.
Rather, take a leaf out of Chromium's book and use MSVC's __cpuid and
_xgetbv built-in, with an inline assembly emulated version for other
compilers.
This preserves the behavior of the original assembly with the following
differences:
- CPUs without cpuid aren't support. Chromium's base/cpu.cc doesn't
check, and SSE2 support is part of our baseline; the perlasm code
is always built with OPENSSL_IA32_SSE2.
- The clear_xmm block in cpu-x86-asm.pl is removed. This was used to
clear some XMM-using features if OSXSAVE was set but XCR0 reports the
OS doesn't use XSAVE to store SSE state. This wasn't present in the
x86_64 and seems wrong. Section 13.5.2 of the Intel manual, volume 1,
explicitly says SSE may still be used in this case; the OS may save
that state in FXSAVE instead. A side discussion on upstream's RT#2633
agrees.
- The old code ran some AMD CPUs through the "intel" codepath and some
went straight to "generic" after duplicating some, but not all, logic.
The AMD copy didn't clear some reserved bits and didn't query CPUID 7
for AVX2 support. This is moot since AMD CPUs today don't support
AVX2, but it seems they're expected to in the future?
- Setting bit 10 is dropped. This doesn't appear to be queried anywhere,
was 32-bit only, and seems a remnant of upstream's
14e21f863a3e3278bb8660ea9844e92e52e1f2f7.
Change-Id: I0548877c97e997f7beb25e15f3fea71c68a951d2
Reviewed-on: https://boringssl-review.googlesource.com/5434
Reviewed-by: Adam Langley <agl@google.com>
2015-07-11 14:47:14 +01:00
|
|
|
|
|
|
|
const char *env1, *env2;
|
2014-06-20 20:00:00 +01:00
|
|
|
env1 = getenv("OPENSSL_ia32cap");
|
|
|
|
if (env1 == NULL) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
// OPENSSL_ia32cap can contain zero, one or two values, separated with a ':'.
|
|
|
|
// Each value is a 64-bit, unsigned value which may start with "0x" to
|
|
|
|
// indicate a hex value. Prior to the 64-bit value, a '~' may be given.
|
|
|
|
//
|
|
|
|
// If '~' isn't present, then the value is taken as the result of the CPUID.
|
|
|
|
// Otherwise the value is inverted and ANDed with the probed CPUID result.
|
|
|
|
//
|
|
|
|
// The first value determines OPENSSL_ia32cap_P[0] and [1]. The second [2]
|
|
|
|
// and [3].
|
2014-06-20 20:00:00 +01:00
|
|
|
|
|
|
|
handle_cpu_env(&OPENSSL_ia32cap_P[0], env1);
|
|
|
|
env2 = strchr(env1, ':');
|
|
|
|
if (env2 != NULL) {
|
|
|
|
handle_cpu_env(&OPENSSL_ia32cap_P[2], env2 + 1);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-08-18 19:06:02 +01:00
|
|
|
#endif // !OPENSSL_NO_ASM && (OPENSSL_X86 || OPENSSL_X86_64)
|