This situation can occur:

[oota-llvm.git] / lib / Target / X86 / README-SSE.txt

diff --git a/lib/Target/X86/README-SSE.txt b/lib/Target/X86/README-SSE.txt
index 1f5f2037cee3375ff6ff332544cf5d30e23cd265..ad28248ef735b459f157c187e212e96a72f49976 100644 (file)
--- a/lib/Target/X86/README-SSE.txt
+++ b/lib/Target/X86/README-SSE.txt
@@ -382,52 +382,6 @@ elements are fixed zeros.
 
 //===---------------------------------------------------------------------===//
 
-For this:
-
-#include <emmintrin.h>
-void test(__m128d *r, __m128d *A, double B) {
-  *r = _mm_loadl_pd(*A, &B);
-}
-
-We generates:
-
-       subl $12, %esp
-       movsd 24(%esp), %xmm0
-       movsd %xmm0, (%esp)
-       movl 20(%esp), %eax
-       movapd (%eax), %xmm0
-       movlpd (%esp), %xmm0
-       movl 16(%esp), %eax
-       movapd %xmm0, (%eax)
-       addl $12, %esp
-       ret
-
-icc generates:
-
-        movl      4(%esp), %edx                                 #3.6
-        movl      8(%esp), %eax                                 #3.6
-        movapd    (%eax), %xmm0                                 #4.22
-        movlpd    12(%esp), %xmm0                               #4.8
-        movapd    %xmm0, (%edx)                                 #4.3
-        ret                                                     #5.1
-
-So icc is smart enough to know that B is in memory so it doesn't load it and
-store it back to stack.
-
-This should be fixed by eliminating the llvm.x86.sse2.loadl.pd intrinsic, 
-lowering it to a load+insertelement instead.  Already match the load+shuffle 
-as movlpd, so this should be easy.  We already get optimal code for:
-
-define void @test2(<2 x double>* %r, <2 x double>* %A, double %B) {
-entry:
-       %tmp2 = load <2 x double>* %A, align 16
-       %tmp8 = insertelement <2 x double> %tmp2, double %B, i32 0
-       store <2 x double> %tmp8, <2 x double>* %r, align 16
-       ret void
-}
-
-//===---------------------------------------------------------------------===//
-
 __m128d test1( __m128d A, __m128d B) {
   return _mm_shuffle_pd(A, B, 0x3);
 }
@@ -505,46 +459,6 @@ nodes which are selected to max / min instructions that are marked commutable.
 
 //===---------------------------------------------------------------------===//
 
-We should compile this:
-#include <xmmintrin.h>
-typedef union {
-  int i[4];
-  float f[4];
-  __m128 v;
-} vector4_t;
-void swizzle (const void *a, vector4_t * b, vector4_t * c) {
-  b->v = _mm_loadl_pi (b->v, (__m64 *) a);
-  c->v = _mm_loadl_pi (c->v, ((__m64 *) a) + 1);
-}
-
-to:
-
-_swizzle:
-        movl    4(%esp), %eax
-        movl    8(%esp), %edx
-        movl    12(%esp), %ecx
-        movlps  (%eax), %xmm0
-        movlps  %xmm0, (%edx)
-        movlps  8(%eax), %xmm0
-        movlps  %xmm0, (%ecx)
-        ret
-
-not:
-
-swizzle:
-        movl 8(%esp), %eax
-        movaps (%eax), %xmm0
-        movl 4(%esp), %ecx
-        movlps (%ecx), %xmm0
-        movaps %xmm0, (%eax)
-        movl 12(%esp), %eax
-        movaps (%eax), %xmm0
-        movlps 8(%ecx), %xmm0
-        movaps %xmm0, (%eax)
-        ret
-
-//===---------------------------------------------------------------------===//
-
 We should materialize vector constants like "all ones" and "signbit" with 
 code like: