diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 1f8312f5457..de0d2cee708 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,16 @@
+2009-06-12  Joey Ye  <joey.ye@intel.com>
+
+	PR middle-end/39146
+	* cfgexpand.c (get_decl_align_unit): Update
+	max_used_stack_slot_alignment with align instead of
+	stack_alignment_needed.
+
+	* function.c (assign_stack_local_1): Update
+	max_used_stack_slot_alignment with alignment_in_bits instead
+	of stack_alignment_needed.
+	(locate_and_pad_parm): Don't update max_used_stack_slot_alignment
+	here.
+
 2009-06-12  Jakub Jelinek  <jakub@redhat.com>
 
 	* dwarf2out.c (last_var_location_insn): New variable.
@@ -8,7 +21,7 @@
 
 2009-06-11  Richard Henderson  <rth@redhat.com>
 
-        * alpha.c (alpha_expand_prologue): Add a REF_CFA_REGISTER
+	* alpha.c (alpha_expand_prologue): Add a REF_CFA_REGISTER
 	note when storing the frame pointer in a register.
 	(FRP): Don't redefine to nothing for epilogue.
 	(alpha_expand_epilogue): Mark register and sp restores.
diff --git a/gcc/cfgexpand.c b/gcc/cfgexpand.c
index ff2684ea91e..3346d2c80cc 100644
--- a/gcc/cfgexpand.c
+++ b/gcc/cfgexpand.c
@@ -564,8 +564,8 @@ get_decl_align_unit (tree decl)
      So here we only make sure stack_alignment_needed >= align.  */
   if (crtl->stack_alignment_needed < align)
     crtl->stack_alignment_needed = align;
-  if (crtl->max_used_stack_slot_alignment < crtl->stack_alignment_needed)
-    crtl->max_used_stack_slot_alignment = crtl->stack_alignment_needed;
+  if (crtl->max_used_stack_slot_alignment < align)
+    crtl->max_used_stack_slot_alignment = align;
 
   return align / BITS_PER_UNIT;
 }
diff --git a/gcc/function.c b/gcc/function.c
index ddf08b04602..82741779aa6 100644
--- a/gcc/function.c
+++ b/gcc/function.c
@@ -355,8 +355,8 @@ assign_stack_local_1 (enum machine_mode mode, HOST_WIDE_INT size,
 
   if (crtl->stack_alignment_needed < alignment_in_bits)
     crtl->stack_alignment_needed = alignment_in_bits;
-  if (crtl->max_used_stack_slot_alignment < crtl->stack_alignment_needed)
-    crtl->max_used_stack_slot_alignment = crtl->stack_alignment_needed;
+  if (crtl->max_used_stack_slot_alignment < alignment_in_bits)
+    crtl->max_used_stack_slot_alignment = alignment_in_bits;
 
   /* Calculate how many bytes the start of local variables is off from
      stack alignment.  */
@@ -3520,8 +3520,6 @@ locate_and_pad_parm (enum machine_mode passed_mode, tree type, int in_regs,
      calling function side.  */
   if (crtl->stack_alignment_needed < boundary)
     crtl->stack_alignment_needed = boundary;
-  if (crtl->max_used_stack_slot_alignment < crtl->stack_alignment_needed)
-    crtl->max_used_stack_slot_alignment = crtl->stack_alignment_needed;
   if (crtl->preferred_stack_boundary < boundary)
     crtl->preferred_stack_boundary = boundary;
 
diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index f64d23fddd6..d90eca58bca 100644
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,9 @@
+2009-06-12  Joey Ye  <joey.ye@intel.com>
+	    H.J. Lu  <hongjiu.lu@intel.com>
+
+	PR middle-end/39146
+	* gcc.target/i386/stackalign/pr39146.c: New.
+
 2009-06-11  H.J. Lu  <hongjiu.lu@intel.com>
 
 	* gcc.target/i386/crc32-1.c: New.
diff --git a/gcc/testsuite/gcc.target/i386/stackalign/pr39146.c b/gcc/testsuite/gcc.target/i386/stackalign/pr39146.c
new file mode 100644
index 00000000000..9ae5f03453e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/stackalign/pr39146.c
@@ -0,0 +1,13 @@
+/* { dg-do compile { target i?86-*-* x86_64-*-* } } */
+/* { dg-options "-O2 -mavx" } */
+
+typedef long long __m256i __attribute__ ((__vector_size__ (32), __may_alias__));
+
+
+__m256i
+bar (__m256i x)
+{
+  return x;
+}
+
+/* { dg-final { scan-assembler-not "and\[lq\]?\[^\\n\]*-32,\[^\\n\]*sp" } } */