X-Git-Url: https://git.libre-soc.org/?a=blobdiff_plain;f=src%2Fgallium%2Fauxiliary%2Futil%2Fu_math.c;h=c58af911be7400b043c7a3d38a94eaadfacb5e2a;hb=aa4386ebfed6a76badae9e1ecf783a372375f072;hp=2811475fa0017aeb721bcbe3e7a3ad0ab2c6184f;hpb=658b1bdb1cc5f9910be910dc156a2e81ed999756;p=mesa.git

diff --git a/src/gallium/auxiliary/util/u_math.c b/src/gallium/auxiliary/util/u_math.c
index 2811475fa00..c58af911be7 100644
--- a/src/gallium/auxiliary/util/u_math.c
+++ b/src/gallium/auxiliary/util/u_math.c
@@ -1,6 +1,6 @@
 /**************************************************************************
  * 
- * Copyright 2008 Tungsten Graphics, Inc., Cedar Park, Texas.
+ * Copyright 2008 VMware, Inc.
  * All Rights Reserved.
  * 
  * Permission is hereby granted, free of charge, to any person obtaining a
@@ -18,7 +18,7 @@
  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
  * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT.
- * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS BE LIABLE FOR
+ * IN NO EVENT SHALL VMWARE AND/OR ITS SUPPLIERS BE LIABLE FOR
  * ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
  * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
  * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
@@ -27,7 +27,16 @@
 
 
 
+#include "pipe/p_config.h"
 #include "util/u_math.h"
+#include "util/u_cpu_detect.h"
+
+#if defined(PIPE_ARCH_SSE)
+#include <xmmintrin.h>
+/* This is defined in pmmintrin.h, but it can only be included when -msse3 is
+ * used, so just define it here to avoid further. */
+#define _MM_DENORMALS_ZERO_MASK	0x0040
+#endif
 
 
 /** 2^x, for x in [-1.0, 1.0) */
@@ -39,7 +48,7 @@ init_pow2_table(void)
 {
    int i;
    for (i = 0; i < POW2_TABLE_SIZE; i++)
-      pow2_table[i] = (float) pow(2.0, (i - POW2_TABLE_OFFSET) / POW2_TABLE_SCALE);
+      pow2_table[i] = exp2f((i - POW2_TABLE_OFFSET) / POW2_TABLE_SCALE);
 }
 
 
@@ -70,4 +79,59 @@ util_init_math(void)
    }
 }
 
+/**
+ * Fetches the contents of the fpstate (mxcsr on x86) register.
+ *
+ * On platforms without support for it just returns 0.
+ */
+unsigned
+util_fpstate_get(void)
+{
+   unsigned mxcsr = 0;
+
+#if defined(PIPE_ARCH_SSE)
+   if (util_cpu_caps.has_sse) {
+      mxcsr = _mm_getcsr();
+   }
+#endif
+
+   return mxcsr;
+}
+
+/**
+ * Make sure that the fp treats the denormalized floating
+ * point numbers as zero.
+ *
+ * This is the behavior required by D3D10. OpenGL doesn't care.
+ */
+unsigned
+util_fpstate_set_denorms_to_zero(unsigned current_mxcsr)
+{
+#if defined(PIPE_ARCH_SSE)
+   if (util_cpu_caps.has_sse) {
+      /* Enable flush to zero mode */
+      current_mxcsr |= _MM_FLUSH_ZERO_MASK;
+      if (util_cpu_caps.has_daz) {
+         /* Enable denormals are zero mode */
+         current_mxcsr |= _MM_DENORMALS_ZERO_MASK;
+      }
+      util_fpstate_set(current_mxcsr);
+   }
+#endif
+   return current_mxcsr;
+}
 
+/**
+ * Set the state of the fpstate (mxcsr on x86) register.
+ *
+ * On platforms without support for it's a noop.
+ */
+void
+util_fpstate_set(unsigned mxcsr)
+{
+#if defined(PIPE_ARCH_SSE)
+   if (util_cpu_caps.has_sse) {
+      _mm_setcsr(mxcsr);
+   }
+#endif
+}