Menu
▾
▴
[cadcdev-svn-commits] [SCM] branch master updated. v2.0.0-7-gca07b70
|
From: Lawrence S. <ljs...@us...> - 2013-06-06 15:27:39
|
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "A pseudo Operating System for the Dreamcast.".
The branch, master has been updated
via ca07b70bbfaf56e70e602b121480b0c4d7a2d6d0 (commit)
from d4332d01e6aff30c621c7a552af9e1ed04aad92d (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
commit ca07b70bbfaf56e70e602b121480b0c4d7a2d6d0
Author: Lawrence Sebald <ljs...@us...>
Date: Thu Jun 6 11:26:31 2013 -0400
Fix the Dreamcast fmath functions so they compile as C99 (while still allowing them to compile as C89).
-----------------------------------------------------------------------
Summary of changes:
doc/CHANGELOG | 1 +
kernel/arch/dreamcast/include/dc/fmath.h | 177 ++++++--------------
.../dreamcast/include/dc/{fmath.h => fmath_base.h} | 102 +-----------
kernel/arch/dreamcast/math/fmath.c | 15 ++-
4 files changed, 69 insertions(+), 226 deletions(-)
copy kernel/arch/dreamcast/include/dc/{fmath.h => fmath_base.h} (66%)
diff --git a/doc/CHANGELOG b/doc/CHANGELOG
index bbec776..e65d0e7 100644
--- a/doc/CHANGELOG
+++ b/doc/CHANGELOG
@@ -11,6 +11,7 @@ KallistiOS version 2.0.1 -----------------------------------------------
- DC Rearranged the SIP driver to get rid of the internal buffer [LS]
- *** Removed the GBA, ia32, and ps2 ports. If someone really wants them back
and wants to maintain them, let me know. [LS]
+- DC Fixed fmath functions so they compile in C99 mode [LS]
KallistiOS version 2.0.0 -----------------------------------------------
- DC Broadband Adapter driver fixes [Dan Potter == DP]
diff --git a/kernel/arch/dreamcast/include/dc/fmath.h b/kernel/arch/dreamcast/include/dc/fmath.h
index 5753f28..eeb0654 100644
--- a/kernel/arch/dreamcast/include/dc/fmath.h
+++ b/kernel/arch/dreamcast/include/dc/fmath.h
@@ -1,7 +1,8 @@
/* KallistiOS ##version##
dc/fmath.h
- (C)2001 Andrew Kieschnick
+ Copyright (C) 2001 Andrew Kieschnick
+ Copyright (C) 2013 Lawrence Sebald
*/
@@ -12,6 +13,7 @@
__BEGIN_DECLS
#include <arch/types.h>
+#include <dc/fmath_base.h>
/**
\file dc/fmath.h
@@ -19,133 +21,31 @@ __BEGIN_DECLS
\author Andrew Kieschnick
*/
-/** PI constant (if you don't want full math.h) */
-#define F_PI 3.1415926f
+/* Sigh... C99 treats inline stuff a lot differently than traditional GCC did,
+ so we need to take care of that... */
+#if __STDC_VERSION__ >= 199901L
+#define __FMINLINE static inline
+#elif __GNUC__
+#define __FMINLINE extern inline
+#else
+/* Uhm... I guess this is the best we can do? */
+#define __FMINLINE static
+#endif
-/** \cond */
-#define __fsin(x) \
- ({ float __value, __arg = (x), __scale = 10430.37835; \
- asm( "fmul %2,%1\n\t" \
- "ftrc %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fmov fr0,%0" \
- : "=f" (__value), "+&f" (__scale) \
- : "f" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-#define __fcos(x) \
- ({ float __value, __arg = (x), __scale = 10430.37835; \
- asm( "fmul %2,%1\n\t" \
- "ftrc %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fmov fr1,%0" \
- : "=f" (__value), "+&f" (__scale) \
- : "f" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-#define __ftan(x) \
- ({ float __value, __arg = (x), __scale = 10430.37835; \
- asm( "fmul %2,%1\n\t" \
- "ftrc %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fdiv fr1, fr0\n\t" \
- "fmov fr0,%0" \
- : "=f" (__value), "+&f" (__scale) \
- : "f" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-
-#define __fisin(x) \
- ({ float __value, __arg = (x); \
- asm( "lds %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fmov fr0,%0" \
- : "=f" (__value) \
- : "r" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-#define __ficos(x) \
- ({ float __value, __arg = (x); \
- asm( "lds %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fmov fr1,%0" \
- : "=f" (__value) \
- : "r" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-#define __fitan(x) \
- ({ float __value, __arg = (x); \
- asm( "lds %1,fpul\n\t" \
- "fsca fpul,dr0\n\t" \
- "fdiv fr1, fr0\n\t" \
- "fmov fr0,%0" \
- : "=f" (__value) \
- : "r" (__arg) \
- : "fpul", "fr0", "fr1"); \
- __value; })
-
-
-#define __fsqrt(x) \
- ({ float __arg = (x); \
- asm( "fsqrt %0\n\t" \
- : "=f" (__arg) : "0" (__arg)); \
- __arg; })
-
-#define __frsqrt(x) \
- ({ float __arg = (x); \
- asm( "fsrra %0\n\t" \
- : "=f" (__arg) : "0" (__arg)); \
- __arg; })
-
-/* Floating point inner product (dot product) */
-#define __fipr(x, y, z, w, a, b, c, d) ({ \
- register float __x __asm__("fr0") = (x); \
- register float __y __asm__("fr1") = (y); \
- register float __z __asm__("fr2") = (z); \
- register float __w __asm__("fr3") = (w); \
- register float __a __asm__("fr4") = (a); \
- register float __b __asm__("fr5") = (b); \
- register float __c __asm__("fr6") = (c); \
- register float __d __asm__("fr7") = (d); \
- __asm__ __volatile__( \
- "fipr fv4,fv0" \
- : "+f" (__w) \
- : "f" (__x), "f" (__y), "f" (__z), "f" (__w), \
- "f" (__a), "f" (__b), "f" (__c), "f" (__d) \
- ); \
- __w; })
-
-/* Floating point inner product w/self (square of vector magnitude) */
-#define __fipr_magnitude_sqr(x, y, z, w) ({ \
- register float __x __asm__("fr4") = (x); \
- register float __y __asm__("fr5") = (y); \
- register float __z __asm__("fr6") = (z); \
- register float __w __asm__("fr7") = (w); \
- __asm__ __volatile__( \
- "fipr fv4,fv4" \
- : "+f" (__w) \
- : "f" (__x), "f" (__y), "f" (__z), "f" (__w) \
- ); \
- __w; })
-
-/** \endcond */
-
-/** \return v1 dot v2 (inner product) */
-extern inline float fipr(float x, float y, float z, float w,
- float a, float b, float c, float d) {
+/**
+ \brief Floating point inner product.
+ \return v1 dot v2 (inner product)
+*/
+__FMINLINE float fipr(float x, float y, float z, float w,
+ float a, float b, float c, float d) {
return __fipr(x, y, z, w, a, b, c, d);
}
/**
- \brief Floating point inner product w/self (square of vector magnitude)
+ \brief Floating point inner product w/self (square of vector magnitude)
\return v1 dot v1 (square of magnitude)
*/
-extern inline float fipr_magnitude_sqr(float x, float y, float z, float w) {
+__FMINLINE float fipr_magnitude_sqr(float x, float y, float z, float w) {
return __fipr_magnitude_sqr(x, y, z, w);
}
@@ -154,7 +54,7 @@ extern inline float fipr_magnitude_sqr(float x, float y, float z, float w) {
\param r a floating point number between 0 and 2*PI
\return sin(r), where r is [0..2*PI]
*/
-extern inline float fsin(float r) {
+__FMINLINE float fsin(float r) {
return __fsin(r);
}
@@ -163,7 +63,7 @@ extern inline float fsin(float r) {
\param r a floating point number between 0 and 2*PI
\return cos(r), where r is [0..2*PI]
*/
-extern inline float fcos(float r) {
+__FMINLINE float fcos(float r) {
return __fcos(r);
}
@@ -172,7 +72,7 @@ extern inline float fcos(float r) {
\param r a floating point number between 0 and 2*PI
\return tan(r), where r is [0..2*PI]
*/
-extern inline float ftan(float r) {
+__FMINLINE float ftan(float r) {
return __ftan(r);
}
@@ -181,7 +81,7 @@ extern inline float ftan(float r) {
\param d an integer between 0 and 65535
\return sin(d), where d is [0..65535]
*/
-extern inline float fisin(int d) {
+__FMINLINE float fisin(int d) {
return __fisin(d);
}
@@ -190,7 +90,7 @@ extern inline float fisin(int d) {
\param d an integer between 0 and 65535
\return cos(d), where d is [0..65535]
*/
-extern inline float ficos(int d) {
+__FMINLINE float ficos(int d) {
return __ficos(d);
}
@@ -199,7 +99,7 @@ extern inline float ficos(int d) {
\param d an integer between 0 and 65535
\return tan(d), where d is [0..65535]
*/
-extern inline float fitan(int d) {
+__FMINLINE float fitan(int d) {
return __fitan(d);
}
@@ -207,17 +107,36 @@ extern inline float fitan(int d) {
\brief Floating point square root
\return sqrt(f)
*/
-extern inline float fsqrt(float f) {
+__FMINLINE float fsqrt(float f) {
return __fsqrt(f);
}
/**
\return 1.0f / sqrt(f)
*/
-extern inline float frsqrt(float f) {
+__FMINLINE float frsqrt(float f) {
return __frsqrt(f);
}
+/* Make sure we declare the non-inline versions for C99 and non-gcc. Why they'd
+ ever be needed, since they're inlined above, who knows? I guess in case
+ someone tries to take the address of one of them? */
+/** \cond */
+#if __STDC_VERSION__ >= 199901L || !defined(__GNUC__)
+extern float fipr(float x, float y, float z, float w, float a, float b, float c,
+ float d);
+extern float fipr_magnitude_sqr(float x, float y, float z, float w);
+extern float fsin(float r);
+extern float fcos(float r);
+extern float ftan(float r);
+extern float fisin(int d);
+extern float ficos(int d);
+extern float fitan(int d);
+extern float fsqrt(float f);
+extern float frsqrt(float f);
+#endif /* __STDC_VERSION__ >= 199901L || !defined(__GNUC__) */
+/** \endcond */
+
__END_DECLS
#endif /* __DC_FMATH_H */
diff --git a/kernel/arch/dreamcast/include/dc/fmath.h b/kernel/arch/dreamcast/include/dc/fmath_base.h
similarity index 66%
copy from kernel/arch/dreamcast/include/dc/fmath.h
copy to kernel/arch/dreamcast/include/dc/fmath_base.h
index 5753f28..f6272ac 100644
--- a/kernel/arch/dreamcast/include/dc/fmath.h
+++ b/kernel/arch/dreamcast/include/dc/fmath_base.h
@@ -1,21 +1,19 @@
/* KallistiOS ##version##
- dc/fmath.h
- (C)2001 Andrew Kieschnick
+ dc/fmath_base.h
+ Copyright (C) 2001 Andrew Kieschnick
*/
-#ifndef __DC_FMATH_H
-#define __DC_FMATH_H
+#ifndef __DC_FMATH_BASE_H
+#define __DC_FMATH_BASE_H
#include <sys/cdefs.h>
__BEGIN_DECLS
-#include <arch/types.h>
-
/**
- \file dc/fmath.h
- \brief Inline functions for the DC's special math instructions
+ \file dc/fmath_base.h
+ \brief Base definitions for the DC's special math instructions
\author Andrew Kieschnick
*/
@@ -135,90 +133,4 @@ __BEGIN_DECLS
/** \endcond */
-/** \return v1 dot v2 (inner product) */
-extern inline float fipr(float x, float y, float z, float w,
- float a, float b, float c, float d) {
- return __fipr(x, y, z, w, a, b, c, d);
-}
-
-/**
- \brief Floating point inner product w/self (square of vector magnitude)
- \return v1 dot v1 (square of magnitude)
-*/
-extern inline float fipr_magnitude_sqr(float x, float y, float z, float w) {
- return __fipr_magnitude_sqr(x, y, z, w);
-}
-
-/**
- \brief Floating point sine
- \param r a floating point number between 0 and 2*PI
- \return sin(r), where r is [0..2*PI]
-*/
-extern inline float fsin(float r) {
- return __fsin(r);
-}
-
-/**
- \brief Floating point cosine
- \param r a floating point number between 0 and 2*PI
- \return cos(r), where r is [0..2*PI]
-*/
-extern inline float fcos(float r) {
- return __fcos(r);
-}
-
-/**
- \brief Floating point tangent
- \param r a floating point number between 0 and 2*PI
- \return tan(r), where r is [0..2*PI]
-*/
-extern inline float ftan(float r) {
- return __ftan(r);
-}
-
-/**
- \brief Integer sine
- \param d an integer between 0 and 65535
- \return sin(d), where d is [0..65535]
-*/
-extern inline float fisin(int d) {
- return __fisin(d);
-}
-
-/**
- \brief Integer cosine
- \param d an integer between 0 and 65535
- \return cos(d), where d is [0..65535]
-*/
-extern inline float ficos(int d) {
- return __ficos(d);
-}
-
-/**
- \brief Integer tangent
- \param d an integer between 0 and 65535
- \return tan(d), where d is [0..65535]
-*/
-extern inline float fitan(int d) {
- return __fitan(d);
-}
-
-/**
- \brief Floating point square root
- \return sqrt(f)
-*/
-extern inline float fsqrt(float f) {
- return __fsqrt(f);
-}
-
-/**
- \return 1.0f / sqrt(f)
-*/
-extern inline float frsqrt(float f) {
- return __frsqrt(f);
-}
-
-__END_DECLS
-
-#endif /* __DC_FMATH_H */
-
+#endif /* !__DC_FMATH_BASE_H */
diff --git a/kernel/arch/dreamcast/math/fmath.c b/kernel/arch/dreamcast/math/fmath.c
index 24cb757..872603a 100644
--- a/kernel/arch/dreamcast/math/fmath.c
+++ b/kernel/arch/dreamcast/math/fmath.c
@@ -1,10 +1,21 @@
/* KallistiOS ##version##
dc/fmath.c
- (C)2001 Andrew Kieschnick
+ Copyright (C) 2001 Andrew Kieschnick
*/
-#include <dc/fmath.h>
+#include <dc/fmath_base.h>
+
+/* v1 dot v2 (inner product) */
+float fipr(float x, float y, float z, float w, float a, float b, float c,
+ float d) {
+ return __fipr(x, y, z, w, a, b, c, d);
+}
+
+/* v1 dot v1 (square of magnitude) */
+float fipr_magnitude_sqr(float x, float y, float z, float w) {
+ return __fipr_magnitude_sqr(x, y, z, w);
+}
/* Returns sin(r), where r is [0..2*PI] */
float fsin(float r) {
hooks/post-receive
--
A pseudo Operating System for the Dreamcast.
|
