.text
.align 16
+.globl _mesa_x86_64_cpuid
+.hidden _mesa_x86_64_cpuid
+_mesa_x86_64_cpuid:
+ pushq %rbx
+ movl (%rdi), %eax
+ movl 8(%rdi), %ecx
+
+ cpuid
+
+ movl %ebx, 4(%rdi)
+ movl %eax, (%rdi)
+ movl %ecx, 8(%rdi)
+ movl %edx, 12(%rdi)
+ popq %rbx
+ ret
+.align 16
.globl _mesa_x86_64_transform_points4_general
+.hidden _mesa_x86_64_transform_points4_general
_mesa_x86_64_transform_points4_general:
/*
* rdi = dest
* rdx = source
*/
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
.byte 0x00, 0x00, 0x00, 0x00
.byte 0x00, 0x00, 0x00, 0x00
.byte 0x00, 0x00, 0x00, 0x00
-.float 0f+1.0
+.float 1.0
.text
.align 16
.globl _mesa_x86_64_transform_points4_3d
+.hidden _mesa_x86_64_transform_points4_3d
/*
* this is slower than _mesa_x86_64_transform_points4_general
* because it ensures that the last matrix row (or is it column?) is 0,0,0,1
movaps 16(%rax), %xmm10
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
.align 16
.globl _mesa_x86_64_transform_points4_identity
+.hidden _mesa_x86_64_transform_points4_identity
_mesa_x86_64_transform_points4_identity:
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
.align 16
-.globl _mesa_x86_64_transform_points4_3d_no_rot
-_mesa_x86_64_transform_points4_3d_no_rot:
+.globl _mesa_3dnow_transform_points4_3d_no_rot
+.hidden _mesa_3dnow_transform_points4_3d_no_rot
+_mesa_3dnow_transform_points4_3d_no_rot:
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
.align 16
-.globl _mesa_x86_64_transform_points4_perspective
-_mesa_x86_64_transform_points4_perspective:
+.globl _mesa_3dnow_transform_points4_perspective
+.hidden _mesa_3dnow_transform_points4_perspective
+_mesa_3dnow_transform_points4_perspective:
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
ret
.align 16
-.globl _mesa_x86_64_transform_points4_2d_no_rot
-_mesa_x86_64_transform_points4_2d_no_rot:
+.globl _mesa_3dnow_transform_points4_2d_no_rot
+.hidden _mesa_3dnow_transform_points4_2d_no_rot
+_mesa_3dnow_transform_points4_2d_no_rot:
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */
.align 16
-.globl _mesa_x86_64_transform_points4_2d
-_mesa_x86_64_transform_points4_2d:
+.globl _mesa_3dnow_transform_points4_2d
+.hidden _mesa_3dnow_transform_points4_2d
+_mesa_3dnow_transform_points4_2d:
movl V4F_COUNT(%rdx), %ecx /* count */
- movzx V4F_STRIDE(%rdx), %eax /* stride */
+ movzbl V4F_STRIDE(%rdx), %eax /* stride */
movl %ecx, V4F_COUNT(%rdi) /* set dest count */
movl $4, V4F_SIZE(%rdi) /* set dest size */