在 2024-10-22 22:18, Jacek Caban 写道:
Style-wise, I would perhaps prefer a solution that uses fewer #ifdefs. We could for example change __mingw_has_sse to always return true when it's defined. It would limit optimization opportunities for the compiler, so perhaps we could have it in a header and make it a trivial inline in case __SSE__ is defined? Anyway, the current patch matches how similar things are done in mingw-w64, so I won't insist on that.Even with that patch, the original patch caching cpuid result would still be beneficial for cases when __SSE__ is not defined.
Here's an updated patch. -- Best regards, LIU Hao
From d54cc9e99b4f6130d6e6ac7ac4aaf077026fcb6d Mon Sep 17 00:00:00 2001
From: LIU Hao <lh_mo...@126.com>
Date: Tue, 22 Oct 2024 23:16:04 +0800
Subject: [PATCH] crt: Provide a common `__mingw_has_sse()`
This commit moves `__mingw_has_sse()` to a separate file, and makes its
result cacheable.
SSE is required for x86-64, and can be enabled for x86-32 by passing `-msse`
to GCC. On ARM and ARM64 it's always absent. A macro is defined if it would
be known to return a constant.
Co-authored-by: Morilli <molli.ben...@gmail.com>
Signed-off-by: LIU Hao <lh_mo...@126.com>
---
mingw-w64-crt/Makefile.am | 2 +-
mingw-w64-crt/include/internal.h | 8 ++++++
mingw-w64-crt/misc/__mingw_has_sse.c | 37 ++++++++++++++++++++++++++++
mingw-w64-crt/misc/feclearexcept.c | 31 ++---------------------
mingw-w64-crt/misc/fegetenv.c | 6 ++---
mingw-w64-crt/misc/fegetexceptflag.c | 10 +++-----
mingw-w64-crt/misc/fesetenv.c | 10 +++-----
mingw-w64-crt/misc/fesetexceptflag.c | 10 +++-----
mingw-w64-crt/misc/fesetround.c | 8 +++---
mingw-w64-crt/misc/fetestexcept.c | 9 +++----
10 files changed, 68 insertions(+), 63 deletions(-)
create mode 100644 mingw-w64-crt/misc/__mingw_has_sse.c
diff --git a/mingw-w64-crt/Makefile.am b/mingw-w64-crt/Makefile.am
index 7db657ed7..b5e23e084 100644
--- a/mingw-w64-crt/Makefile.am
+++ b/mingw-w64-crt/Makefile.am
@@ -1044,7 +1044,7 @@ src_libmingwex=\
misc/feclearexcept.c misc/fegetenv.c misc/fegetexceptflag.c
misc/fegetround.c misc/feholdexcept.c \
misc/feraiseexcept.c misc/fesetenv.c misc/fesetexceptflag.c
misc/fesetround.c misc/fetestexcept.c \
misc/feupdateenv.c misc/ftruncate.c misc/fwide.c
misc/getlogin.c misc/getopt.c \
- misc/gettimeofday.c \
+ misc/gettimeofday.c misc/__mingw_has_sse.c \
misc/mempcpy.c misc/mingw-aligned-malloc.c \
misc/mingw_matherr.c misc/mingw_mbwc_convert.c misc/mingw_usleep.c
misc/mingw_wcstod.c misc/mingw_wcstof.c \
misc/mingw_wcstold.c \
diff --git a/mingw-w64-crt/include/internal.h b/mingw-w64-crt/include/internal.h
index 3786ed7a1..752c5dc96 100644
--- a/mingw-w64-crt/include/internal.h
+++ b/mingw-w64-crt/include/internal.h
@@ -146,6 +146,14 @@ extern "C" {
PIMAGE_SECTION_HEADER __cdecl _FindPESection (PBYTE pImageBase, DWORD_PTR
rva);
BOOL __cdecl _IsNonwritableInCurrentImage (PBYTE pTarget);
+#if defined(__SSE__)
+# define __mingw_has_sse() 1
+#elif defined(__i386__)
+ int __mingw_has_sse(void);
+#else
+# define __mingw_has_sse() 0
+#endif
+
#ifdef __cplusplus
}
#endif
diff --git a/mingw-w64-crt/misc/__mingw_has_sse.c
b/mingw-w64-crt/misc/__mingw_has_sse.c
new file mode 100644
index 000000000..e8e8a7059
--- /dev/null
+++ b/mingw-w64-crt/misc/__mingw_has_sse.c
@@ -0,0 +1,37 @@
+/**
+ * This file has no copyright assigned and is placed in the Public Domain.
+ * This file is part of the mingw-w64 runtime package.
+ * No warranty is given; refer to the file DISCLAIMER.PD within this package.
+ */
+
+#if defined(__i386__)
+int __mingw_has_sse (void);
+static int __has_sse = -1;
+
+int __mingw_has_sse(void)
+{
+ int cpuInfo[4], infoType = 1;
+ int o_flag, n_flag;
+
+ if (__has_sse != -1)
+ return __has_sse;
+
+ __asm__ volatile ("pushfl\n\tpopl %0" : "=mr" (o_flag));
+ n_flag = o_flag ^ 0x200000;
+ __asm__ volatile ("pushl %0\n\tpopfl" : : "g" (n_flag));
+ __asm__ volatile ("pushfl\n\tpopl %0" : "=mr" (n_flag));
+ if (n_flag == o_flag)
+ {
+ __has_sse = 0;
+ return 0;
+ }
+
+ __asm__ __volatile__ (
+ "cpuid"
+ : "=a" (cpuInfo[0]), "=b" (cpuInfo[1]), "=c" (cpuInfo[2]),
+ "=d" (cpuInfo[3])
+ : "a" (infoType));
+ __has_sse = (cpuInfo[3] & 0x2000000);
+ return __has_sse;
+}
+#endif /* __i386__ */
diff --git a/mingw-w64-crt/misc/feclearexcept.c
b/mingw-w64-crt/misc/feclearexcept.c
index 673528b20..a5ab0f095 100644
--- a/mingw-w64-crt/misc/feclearexcept.c
+++ b/mingw-w64-crt/misc/feclearexcept.c
@@ -3,36 +3,9 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
-
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-int __mingw_has_sse (void);
-int __mingw_has_sse(void)
-{
- int cpuInfo[4],infoType = 1;
-
-#ifndef _WIN64
- int o_flag, n_flag;
-
- __asm__ volatile ("pushfl\n\tpopl %0" : "=mr" (o_flag));
- n_flag = o_flag ^ 0x200000;
- __asm__ volatile ("pushl %0\n\tpopfl" : : "g" (n_flag));
- __asm__ volatile ("pushfl\n\tpopl %0" : "=mr" (n_flag));
- if (n_flag == o_flag)
- return 0;
-#endif
-
- __asm__ __volatile__ (
- "cpuid"
- : "=a" (cpuInfo[0]), "=b" (cpuInfo[1]), "=c" (cpuInfo[2]),
- "=d" (cpuInfo[3])
- : "a" (infoType));
- if (cpuInfo[3] & 0x2000000)
- return 1;
- return 0;
-}
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <fenv.h>
+#include <internal.h>
/* 7.6.2.1
The feclearexcept function clears the supported exceptions
diff --git a/mingw-w64-crt/misc/fegetenv.c b/mingw-w64-crt/misc/fegetenv.c
index 31960816e..e17fd491c 100644
--- a/mingw-w64-crt/misc/fegetenv.c
+++ b/mingw-w64-crt/misc/fegetenv.c
@@ -3,11 +3,9 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <fenv.h>
+#include <internal.h>
/* 7.6.4.1
The fegetenv function stores the current floating-point environment
diff --git a/mingw-w64-crt/misc/fegetexceptflag.c
b/mingw-w64-crt/misc/fegetexceptflag.c
index eaa940a2a..b50ca6fc1 100644
--- a/mingw-w64-crt/misc/fegetexceptflag.c
+++ b/mingw-w64-crt/misc/fegetexceptflag.c
@@ -3,13 +3,11 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-extern int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <fenv.h>
+#include <internal.h>
-/* 7.6.2.2
+/* 7.6.2.2
The fegetexceptflag function stores an implementation-defined
representation of the exception flags indicated by the argument
excepts in the object pointed to by the argument flagp. */
@@ -32,7 +30,7 @@ int fegetexceptflag (fexcept_t * flagp, int excepts)
_mxcsr = 0;
if (__mingw_has_sse ())
__asm__ volatile ("stmxcsr %0" : "=m" (_mxcsr));
-
+
*flagp = (_mxcsr | _status) & excepts & FE_ALL_EXCEPT;
#endif /* defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__) */
return 0;
diff --git a/mingw-w64-crt/misc/fesetenv.c b/mingw-w64-crt/misc/fesetenv.c
index a3279bddd..76325ce37 100644
--- a/mingw-w64-crt/misc/fesetenv.c
+++ b/mingw-w64-crt/misc/fesetenv.c
@@ -3,13 +3,11 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
+
#include <_mingw.h>
#include <fenv.h>
#include <float.h>
-
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-extern int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <internal.h>
/* 7.6.4.3
The fesetenv function establishes the floating-point environment
@@ -28,7 +26,7 @@ int fesetenv (const fenv_t * envp)
{
#if defined(_ARM_) || defined(__arm__)
if (envp == FE_DFL_ENV)
- /* Use the choice made at app startup */
+ /* Use the choice made at app startup */
_fpreset();
else
__asm__ volatile ("fmxr FPSCR, %0" : : "r" (*envp));
@@ -59,7 +57,7 @@ int fesetenv (const fenv_t * envp)
(* __MINGW_IMP_SYMBOL(_fpreset))();
else if (envp == FE_DFL_ENV)
- /* Use the choice made at app startup */
+ /* Use the choice made at app startup */
_fpreset();
else
diff --git a/mingw-w64-crt/misc/fesetexceptflag.c
b/mingw-w64-crt/misc/fesetexceptflag.c
index 7e3361c6a..095dfea0b 100644
--- a/mingw-w64-crt/misc/fesetexceptflag.c
+++ b/mingw-w64-crt/misc/fesetexceptflag.c
@@ -3,11 +3,9 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-extern int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <fenv.h>
+#include <internal.h>
/* 7.6.2.4
The fesetexceptflag function sets the complete status for those
@@ -16,9 +14,9 @@ extern int __mingw_has_sse (void);
*flagp shall have been set by a previous call to fegetexceptflag
whose second argument represented at least those exceptions
represented by the argument excepts. This function does not raise
- exceptions, but only sets the state of the flags. */
+ exceptions, but only sets the state of the flags. */
-int fesetexceptflag (const fexcept_t * flagp, int excepts)
+int fesetexceptflag (const fexcept_t * flagp, int excepts)
{
fenv_t _env;
diff --git a/mingw-w64-crt/misc/fesetround.c b/mingw-w64-crt/misc/fesetround.c
index 624a9997d..e51f46cd7 100644
--- a/mingw-w64-crt/misc/fesetround.c
+++ b/mingw-w64-crt/misc/fesetround.c
@@ -3,11 +3,9 @@
* This file is part of the mingw-w64 runtime package.
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
+#include <fenv.h>
+#include <internal.h>
/* 7.6.3.2
The fesetround function establishes the rounding direction
@@ -42,7 +40,7 @@ int fesetround (int mode)
_cw &= ~0xc00;
_cw |= mode;
__asm__ volatile ("fldcw %0;" : : "m" (*&_cw));
-
+
if (__mingw_has_sse ())
{
int mxcsr;
diff --git a/mingw-w64-crt/misc/fetestexcept.c
b/mingw-w64-crt/misc/fetestexcept.c
index da073e6d0..04e75970f 100644
--- a/mingw-w64-crt/misc/fetestexcept.c
+++ b/mingw-w64-crt/misc/fetestexcept.c
@@ -4,13 +4,10 @@
* No warranty is given; refer to the file DISCLAIMER.PD within this package.
*/
-#include <fenv.h>
+#include <fenv.h>
+#include <internal.h>
-#if !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__))
-extern int __mingw_has_sse (void);
-#endif /* !(defined(_ARM_) || defined(__arm__) || defined(_ARM64_) ||
defined(__aarch64__)) */
-
-/* 7.6.2.5
+/* 7.6.2.5
The fetestexcept function determines which of a specified subset of
the exception flags are currently set. The excepts argument
specifies the exception flags to be queried.
--
2.47.0
OpenPGP_signature.asc
Description: OpenPGP digital signature
_______________________________________________ Mingw-w64-public mailing list Mingw-w64-public@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/mingw-w64-public
