Update aosp/master LLVM for rebase to r222494.

Change-Id: Ic787f5e0124df789bd26f3f24680f45e678eef2d
author: Stephen Hines <srhines@google.com> 2014-12-01 14:51:49 -0800
committer: Stephen Hines <srhines@google.com> 2014-12-02 16:08:10 -0800
commit: 37ed9c199ca639565f6ce88105f9e39e898d82d0 (patch)
tree: 8fb36d3910e3ee4c4e1b7422f4f017108efc52f5 /test/CodeGen/Mips
parent: d2327b22152ced7bc46dc629fc908959e8a52d03 (diff)
download: external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.zip
external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.tar.gz
external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.tar.bz2
72 files changed, 4220 insertions, 451 deletions
diff --git a/test/CodeGen/Mips/Fast-ISel/br1.ll b/test/CodeGen/Mips/Fast-ISel/br1.ll
new file mode 100644
index 0000000..579a77f
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/br1.ll
@@ -0,0 +1,34 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@b = global i32 1, align 4
+@i = global i32 0, align 4
+@.str = private unnamed_addr constant [5 x i8] c"%i \0A\00", align 1
+
+; Function Attrs: nounwind
+define void @br() #0 {
+entry:
+  %0 = load i32* @b, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %if.end, label %if.then
+
+if.then:                                          ; preds = %entry
+  store i32 6754, i32* @i, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+; FIXME: This instruction is redundant.
+; CHECK:  xor  $[[REG1:[0-9]+]], ${{[0-9]+}}, $zero
+; CHECK:  sltiu  $[[REG2:[0-9]+]], $[[REG1]], 1
+; CHECK:  bgtz  $[[REG2]], $BB[[BL:[0-9]+_[0-9]+]]
+; CHECK:  nop
+; CHECK:  addiu  ${{[0-9]+}}, $zero, 6754
+; CHECK:  sw  ${{[0-9]+}}, 0(${{[0-9]+}})
+; CHECK: $BB[[BL]]:
+
+}
+
+attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
diff --git a/test/CodeGen/Mips/Fast-ISel/callabi.ll b/test/CodeGen/Mips/Fast-ISel/callabi.ll
new file mode 100644
index 0000000..44b94bb
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/callabi.ll
@@ -0,0 +1,477 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32r2
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s -check-prefix=CHECK2
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s -check-prefix=CHECK2
+
+
+@c1 = global i8 -45, align 1
+@uc1 = global i8 27, align 1
+@s1 = global i16 -1789, align 2
+@us1 = global i16 1256, align 2
+
+; Function Attrs: nounwind
+define void @cxi() #0 {
+entry:
+; CHECK-LABEL:  cxi
+  call void @xi(i32 10)
+; CHECK-DAG:    addiu   $4, $zero, 10
+; CHECK-DAG:    lw      $25, %got(xi)(${{[0-9]+}})
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xi(i32) #1
+
+; Function Attrs: nounwind
+define void @cxii() #0 {
+entry:
+; CHECK-LABEL:  cxii
+  call void @xii(i32 746, i32 892)
+; CHECK-DAG:    addiu   $4, $zero, 746
+; CHECK-DAG:    addiu   $5, $zero, 892
+; CHECK-DAG:    lw      $25, %got(xii)(${{[0-9]+}})
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xii(i32, i32) #1
+
+; Function Attrs: nounwind
+define void @cxiii() #0 {
+entry:
+; CHECK-LABEL:  cxiii
+  call void @xiii(i32 88, i32 44, i32 11)
+; CHECK-DAG:    addiu   $4, $zero, 88
+; CHECK-DAG:    addiu   $5, $zero, 44
+; CHECK-DAG:    addiu   $6, $zero, 11
+; CHECK-DAG:    lw      $25, %got(xiii)(${{[0-9]+}})
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xiii(i32, i32, i32) #1
+
+; Function Attrs: nounwind
+define void @cxiiii() #0 {
+entry:
+; CHECK-LABEL:  cxiiii
+  call void @xiiii(i32 167, i32 320, i32 97, i32 14)
+; CHECK-DAG:    addiu   $4, $zero, 167
+; CHECK-DAG:    addiu   $5, $zero, 320
+; CHECK-DAG:    addiu   $6, $zero, 97
+; CHECK-DAG:    addiu   $7, $zero, 14
+; CHECK-DAG:    lw      $25, %got(xiiii)(${{[0-9]+}})
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xiiii(i32, i32, i32, i32) #1
+
+; Function Attrs: nounwind
+define void @cxiiiiconv() #0 {
+entry:
+; CHECK-LABEL: cxiiiiconv
+; mips32r2-LABEL:  cxiiiiconv
+; mips32-LABEL:  cxiiiiconv
+  %0 = load i8* @c1, align 1
+  %conv = sext i8 %0 to i32
+  %1 = load i8* @uc1, align 1
+  %conv1 = zext i8 %1 to i32
+  %2 = load i16* @s1, align 2
+  %conv2 = sext i16 %2 to i32
+  %3 = load i16* @us1, align 2
+  %conv3 = zext i16 %3 to i32
+  call void @xiiii(i32 %conv, i32 %conv1, i32 %conv2, i32 %conv3)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32r2:     addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32:       addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32r2-DAG:         lw      $[[REG_C1_ADDR:[0-9]+]], %got(c1)($[[REG_GP]])
+; mips32r2-DAG: lbu     $[[REG_C1:[0-9]+]], 0($[[REG_C1_ADDR]])
+; mips32r2-DAG  seb     $3, $[[REG_C1]]
+; mips32-DAG:   lw      $[[REG_C1_ADDR:[0-9]+]], %got(c1)($[[REG_GP]])
+; mips32-DAG:   lbu     $[[REG_C1:[0-9]+]], 0($[[REG_C1_ADDR]])
+; mips32-DAG:   sll     $[[REG_C1_1:[0-9]+]], $[[REG_C1]], 24
+; mips32-DAG:   sra     $4, $[[REG_C1_1]], 24
+; CHECK-DAG:    lw      $[[REG_UC1_ADDR:[0-9]+]], %got(uc1)($[[REG_GP]])
+; CHECK-DAG:    lbu     $[[REG_UC1:[0-9]+]], 0($[[REG_UC1_ADDR]])
+; FIXME andi is superfulous
+; CHECK-DAG:    andi    $5, $[[REG_UC1]], 255
+; mips32r2-DAG:         lw      $[[REG_S1_ADDR:[0-9]+]], %got(s1)($[[REG_GP]])
+; mips32r2-DAG: lhu     $[[REG_S1:[0-9]+]], 0($[[REG_S1_ADDR]])
+; mips32r2-DAG: seh     $6, $[[REG_S1]]
+; mips32-DAG:   lw      $[[REG_S1_ADDR:[0-9]+]], %got(s1)($[[REG_GP]])
+; mips32-DAG:   lhu     $[[REG_S1:[0-9]+]], 0($[[REG_S1_ADDR]])
+; mips32-DAG:   sll     $[[REG_S1_1:[0-9]+]], $[[REG_S1]], 16
+; mips32-DAG:   sra     $6, $[[REG_S1_1]], 16
+; CHECK-DAG:    lw      $[[REG_US1_ADDR:[0-9]+]], %got(us1)($[[REG_GP]])
+; CHECK-DAG:    lhu     $[[REG_US1:[0-9]+]], 0($[[REG_US1_ADDR]])
+; FIXME andi is superfulous
+; CHECK-DAG:    andi    $7, $[[REG_US1]], 65535
+; mips32r2:     jalr    $25
+; mips32r2:     jalr    $25
+; CHECK:        jalr    $25
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @cxf() #0 {
+entry:
+; CHECK-LABEL:  cxf
+  call void @xf(float 0x40BBC85560000000)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK:        lui     $[[REG_FPCONST_1:[0-9]+]], 17886
+; CHECK:        ori     $[[REG_FPCONST:[0-9]+]], $[[REG_FPCONST_1]], 17067
+; CHECK: mtc1   $[[REG_FPCONST]], $f12
+; CHECK:        lw      $25, %got(xf)($[[REG_GP]])
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xf(float) #1
+
+; Function Attrs: nounwind
+define void @cxff() #0 {
+entry:
+; CHECK-LABEL:  cxff
+  call void @xff(float 0x3FF74A6CA0000000, float 0x401A2C0840000000)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    lui     $[[REG_FPCONST_1:[0-9]+]], 16314
+; CHECK-DAG:    ori     $[[REG_FPCONST:[0-9]+]], $[[REG_FPCONST_1]], 21349
+; CHECK-DAG: mtc1       $[[REG_FPCONST]], $f12
+; CHECK-DAG:    lui     $[[REG_FPCONST_2:[0-9]+]], 16593
+; CHECK-DAG:    ori     $[[REG_FPCONST_3:[0-9]+]], $[[REG_FPCONST_2]], 24642
+; CHECK-DAG: mtc1       $[[REG_FPCONST_3]], $f14
+; CHECK:        lw      $25, %got(xff)($[[REG_GP]])
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xff(float, float) #1
+
+; Function Attrs: nounwind
+define void @cxfi() #0 {
+entry:
+; CHECK-LABEL: cxfi
+  call void @xfi(float 0x4013906240000000, i32 102)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    lui     $[[REG_FPCONST_1:[0-9]+]], 16540
+; CHECK-DAG:    ori     $[[REG_FPCONST:[0-9]+]], $[[REG_FPCONST_1]], 33554
+; CHECK-DAG: mtc1       $[[REG_FPCONST]], $f12
+; CHECK-DAG:    addiu   $5, $zero, 102
+; CHECK:        lw      $25, %got(xfi)($[[REG_GP]])
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xfi(float, i32) #1
+
+; Function Attrs: nounwind
+define void @cxfii() #0 {
+entry:
+; CHECK-LABEL: cxfii
+  call void @xfii(float 0x405EC7EE00000000, i32 9993, i32 10922)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    lui     $[[REG_FPCONST_1:[0-9]+]], 17142
+; CHECK-DAG:    ori     $[[REG_FPCONST:[0-9]+]], $[[REG_FPCONST_1]], 16240
+; CHECK-DAG: mtc1       $[[REG_FPCONST]], $f12
+; CHECK-DAG:    addiu   $5, $zero, 9993
+; CHECK-DAG:    addiu   $6, $zero, 10922
+; CHECK:        lw      $25, %got(xfii)($[[REG_GP]])
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xfii(float, i32, i32) #1
+
+; Function Attrs: nounwind
+define void @cxfiii() #0 {
+entry:
+; CHECK-LABEL: cxfiii
+  call void @xfiii(float 0x405C072B20000000, i32 3948, i32 89011, i32 111222)
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    lui     $[[REG_FPCONST_1:[0-9]+]], 17120
+; CHECK-DAG:    ori     $[[REG_FPCONST:[0-9]+]], $[[REG_FPCONST_1]], 14681
+; CHECK-DAG: mtc1       $[[REG_FPCONST]], $f12
+; CHECK-DAG:    addiu   $5, $zero, 3948
+; CHECK-DAG:    lui     $[[REG_I_1:[0-9]+]], 1
+; CHECK-DAG:    ori     $6, $[[REG_I_1]], 23475
+; CHECK-DAG:    lui     $[[REG_I_2:[0-9]+]], 1
+; CHECK-DAG:    ori     $7, $[[REG_I_2]], 45686
+; CHECK:        lw      $25, %got(xfiii)($[[REG_GP]])
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xfiii(float, i32, i32, i32) #1
+
+; Function Attrs: nounwind
+define void @cxd() #0 {
+entry:
+; mips32r2-LABEL: cxd:
+; mips32-LABEL: cxd:
+  call void @xd(double 5.994560e+02)
+; mips32:       addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32-DAG:   lui     $[[REG_FPCONST_1:[0-9]+]], 16514
+; mips32-DAG:   ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 48037
+; mips32-DAG:   lui     $[[REG_FPCONST_3:[0-9]+]], 58195
+; mips32-DAG:   ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 63439
+; mips32-DAG:    mtc1   $[[REG_FPCONST_4]], $f12
+; mips32-DAG:       mtc1        $[[REG_FPCONST_2]], $f13
+; mips32-DAG:   lw      $25, %got(xd)($[[REG_GP]])
+; mips32:       jalr    $25
+; mips32r2:     addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32r2-DAG: lui     $[[REG_FPCONST_1:[0-9]+]], 16514
+; mips32r2-DAG: ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 48037
+; mips32r2-DAG: lui     $[[REG_FPCONST_3:[0-9]+]], 58195
+; mips32r2-DAG: ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 63439
+; mips32r2-DAG: mtc1    $[[REG_FPCONST_4]], $f12
+; mips32r2-DAG: mthc1   $[[REG_FPCONST_2]], $f12
+; mips32r2-DAG: lw      $25, %got(xd)($[[REG_GP]])
+; mips32r2 :    jalr    $25
+  ret void
+}
+
+declare void @xd(double) #1
+
+; Function Attrs: nounwind
+define void @cxdd() #0 {
+; mips32r2-LABEL: cxdd:
+; mips32-LABEL: cxdd:
+entry:
+  call void @xdd(double 1.234980e+03, double 0x40F5B331F7CED917)
+; mips32:       addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32-DAG:   lui     $[[REG_FPCONST_1:[0-9]+]], 16531
+; mips32-DAG:   ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 19435
+; mips32-DAG:   lui     $[[REG_FPCONST_3:[0-9]+]], 34078
+; mips32-DAG:   ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 47186
+; mips32-DAG:   mtc1    $[[REG_FPCONST_4]], $f12
+; mips32-DAG:   mtc1    $[[REG_FPCONST_2]], $f13
+; mips32-DAG:   lui     $[[REG_FPCONST_1:[0-9]+]], 16629
+; mips32-DAG:   ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 45873
+; mips32-DAG:   lui     $[[REG_FPCONST_3:[0-9]+]], 63438
+; mips32-DAG:   ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 55575
+; mips32-DAG:   mtc1    $[[REG_FPCONST_4]], $f14
+; mips32-DAG:   mtc1    $[[REG_FPCONST_2]], $f15
+; mips32-DAG:   lw      $25, %got(xdd)($[[REG_GP]])
+; mips32:       jalr    $25
+; mips32r2:     addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; mips32r2-DAG: lui     $[[REG_FPCONST_1:[0-9]+]], 16531
+; mips32r2-DAG: ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 19435
+; mips32r2-DAG: lui     $[[REG_FPCONST_3:[0-9]+]], 34078
+; mips32r2-DAG: ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 47186
+; mips32r2-DAG: mtc1    $[[REG_FPCONST_4]], $f12
+; mips32r2-DAG: mthc1   $[[REG_FPCONST_2]], $f12
+; mips32r2-DAG: lui     $[[REG_FPCONST_1:[0-9]+]], 16629
+; mips32r2-DAG: ori     $[[REG_FPCONST_2:[0-9]+]], $[[REG_FPCONST_1]], 45873
+; mips32r2-DAG: lui     $[[REG_FPCONST_3:[0-9]+]], 63438
+; mips32r2-DAG: ori     $[[REG_FPCONST_4:[0-9]+]], $[[REG_FPCONST_3]], 55575
+; mips32r2-DAG: mtc1    $[[REG_FPCONST_4]], $f14
+; mips32r2-DAG: mthc1   $[[REG_FPCONST_2]], $f14
+; mips32r2-DAG: lw      $25, %got(xdd)($[[REG_GP]])
+; mips32r2 :    jalr    $25
+  ret void
+}
+
+declare void @xdd(double, double) #1
+
+; Function Attrs: nounwind
+define void @cxif() #0 {
+entry:
+; CHECK-LABEL: cxif:
+  call void @xif(i32 345, float 0x407BCE5A20000000)
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    addiu   $4, $zero, 345
+; CHECK-DAG:    lui     $[[REGF_1:[0-9]+]], 17374
+; CHECK-DAG:    ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 29393
+; CHECK-DAG:    mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK-DAG:    mfc1    $5, $f[[REGF_3]]
+; CHECK-DAG:    lw      $25, %got(xif)($[[REG_GP]])
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xif(i32, float) #1
+
+; Function Attrs: nounwind
+define void @cxiff() #0 {
+entry:
+; CHECK-LABEL: cxiff:
+; CHECK2-LABEL: cxiff:
+  call void @xiff(i32 12239, float 0x408EDB3340000000, float 0x4013FFE5C0000000)
+; We need to do the two floating point parameters in a separate
+; check because we can't control the ordering of parts of the sequence
+;;
+; CHECK:        addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK:        addiu   $4, $zero, 12239
+; CHECK2:       addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK2:       addiu   $4, $zero, 12239
+; CHECK:        lui     $[[REGF_1:[0-9]+]], 17526
+; CHECK:        ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 55706
+; CHECK:        mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK:        mfc1    $5, $f[[REGF_3]]
+; CHECK2:       lui     $[[REGF2_1:[0-9]+]], 16543
+; CHECK2:       ori     $[[REGF2_2:[0-9]+]], $[[REGF2_1]], 65326
+; CHECK2:       mtc1    $[[REGF2_2]], $f[[REGF2_3:[0-9]+]]
+; CHECK2:       mfc1    $6, $f[[REGF2_3]]
+; CHECK:        lw      $25, %got(xiff)($[[REG_GP]])
+; CHECK2:       lw      $25, %got(xiff)($[[REG_GP]])
+; CHECK:        jalr    $25
+; CHECK2:       jalr    $25
+  ret void
+}
+
+declare void @xiff(i32, float, float) #1
+
+; Function Attrs: nounwind
+define void @cxifi() #0 {
+entry:
+; CHECK: cxifi:
+  call void @xifi(i32 887, float 0x402277CEE0000000, i32 888)
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    addiu   $4, $zero, 887
+; CHECK-DAG:    lui     $[[REGF_1:[0-9]+]], 16659
+; CHECK-DAG:    ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 48759
+; CHECK-DAG:    mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK-DAG:    mfc1    $5, $f[[REGF_3]]
+; CHECk-DAG:    addiu   $6, $zero, 888
+; CHECK-DAG:    lw      $25, %got(xifi)($[[REG_GP]])
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xifi(i32, float, i32) #1
+
+; Function Attrs: nounwind
+define void @cxifif() #0 {
+entry:
+; CHECK: cxifif:
+; CHECK2: cxifif:
+  call void @xifif(i32 67774, float 0x408EE0FBE0000000, i32 9991, float 0x40B15C8CC0000000)
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    lui     $[[REGI:[0-9]+]], 1
+; CHECK-DAG:    ori     $4, $[[REGI]], 2238
+; CHECK-DAG:    lui     $[[REGF_1:[0-9]+]], 17527
+; CHECK-DAG:    ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 2015
+; CHECK-DAG:    mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK-DAG:    mfc1    $5, $f[[REGF_3]]
+; CHECk-DAG:    addiu   $6, $zero, 888
+; CHECK2:       lui     $[[REGF2_1:[0-9]+]], 17802
+; CHECK2:       ori     $[[REGF2_2:[0-9]+]], $[[REGF2_1]], 58470
+; CHECK2:       mtc1    $[[REGF2_2]], $f[[REGF2_3:[0-9]+]]
+; CHECK2:       mfc1    $7, $f[[REGF2_3]]
+; CHECK:        lw      $25, %got(xifif)($[[REG_GP]])
+; CHECK2:       lw      $25, %got(xifif)($[[REG_GP]])
+; CHECK2:       jalr    $25
+; CHECK:        jalr    $25
+
+  ret void
+}
+
+declare void @xifif(i32, float, i32, float) #1
+
+; Function Attrs: nounwind
+define void @cxiffi() #0 {
+entry:
+; CHECK-label: cxiffi:
+; CHECK2-label: cxiffi:
+  call void @xiffi(i32 45, float 0x3FF6666660000000, float 0x408F333340000000, i32 234)
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    addiu   $4, $zero, 45
+; CHECK2-DAG:   addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK2-DAG:   addiu   $4, $zero, 45
+; CHECK-DAG:    lui     $[[REGF_1:[0-9]+]], 16307
+; CHECK-DAG:    ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 13107
+; CHECK-DAG:    mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK-DAG:    mfc1    $5, $f[[REGF_3]]
+; CHECK2:       lui     $[[REGF2_1:[0-9]+]], 17529
+; CHECK2:       ori     $[[REGF2_2:[0-9]+]], $[[REGF2_1]], 39322
+; CHECK2:       mtc1    $[[REGF2_2]], $f[[REGF2_3:[0-9]+]]
+; CHECK2:       mfc1    $6, $f[[REGF2_3]]
+; CHECK-DAG:    lw      $25, %got(xiffi)($[[REG_GP]])
+; CHECK-DAG:    addiu   $7, $zero, 234
+; CHECK2-DAG:   lw      $25, %got(xiffi)($[[REG_GP]])
+; CHECK:        jalr    $25
+; CHECK2:       jalr    $25
+
+  ret void
+}
+
+declare void @xiffi(i32, float, float, i32) #1
+
+; Function Attrs: nounwind
+define void @cxifii() #0 {
+entry:
+; CHECK-DAG:    cxifii:
+  call void @xifii(i32 12239, float 0x408EDB3340000000, i32 998877, i32 1234)
+; CHECK-DAG:    addu    $[[REG_GP:[0-9]+]], ${{[0-9]+}}, ${{[0-9+]}}
+; CHECK-DAG:    addiu   $4, $zero, 12239
+; CHECK-DAG:    lui     $[[REGF_1:[0-9]+]], 17526
+; CHECK-DAG:    ori     $[[REGF_2:[0-9]+]], $[[REGF_1]], 55706
+; CHECK-DAG:    mtc1    $[[REGF_2]], $f[[REGF_3:[0-9]+]]
+; CHECK-DAG:    mfc1    $5, $f[[REGF_3]]
+; CHECK-DAG:    lui     $[[REGI2:[0-9]+]], 15
+; CHECK-DAG:    ori     $6, $[[REGI2]], 15837
+; CHECk-DAG:    addiu   $7, $zero, 1234
+; CHECK-DAG:    lw      $25, %got(xifii)($[[REG_GP]])
+; CHECK:        jalr    $25
+  ret void
+}
+
+declare void @xifii(i32, float, i32, i32) #1
+
+; FIXME: this function will not pass yet. 
+; Function Attrs: nounwind
+; define void @cxfid() #0 {
+;entry:
+;  call void @xfid(float 0x4013B851E0000000, i32 811123, double 0x40934BFF487FCB92)
+;  ret void
+;}
+
+declare void @xfid(float, i32, double) #1
+
+; Function Attrs: nounwind
+define void @g() #0 {
+entry:
+  call void @cxi()
+  call void @cxii()
+  call void @cxiii()
+  call void @cxiiii()
+  call void @cxiiiiconv()
+  call void @cxf()
+  call void @cxff()
+  call void @cxd()
+  call void @cxfi()
+  call void @cxfii()
+  call void @cxfiii()
+  call void @cxdd()
+  call void @cxif()
+  call void @cxiff()
+  call void @cxifi()
+  call void @cxifii()
+  call void @cxifif()
+  call void @cxiffi()
+  ret void
+}
+
+
+attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+attributes #1 = { "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+
+!llvm.ident = !{!0}
+
+!0 = metadata !{metadata !"clang version 3.6.0 (gitosis@dmz-portal.mips.com:clang 43992fe7b17de5553ac06d323cb80cc6723a9ae3) (gitosis@dmz-portal.mips.com:llvm.git 0834e6839eb170197c81bb02e916258d1527e312)"}
diff --git a/test/CodeGen/Mips/Fast-ISel/fpcmpa.ll b/test/CodeGen/Mips/Fast-ISel/fpcmpa.ll
new file mode 100644
index 0000000..c72b1e7
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/fpcmpa.ll
@@ -0,0 +1,254 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@f1 = common global float 0.000000e+00, align 4
+@f2 = common global float 0.000000e+00, align 4
+@b1 = common global i32 0, align 4
+@d1 = common global double 0.000000e+00, align 8
+@d2 = common global double 0.000000e+00, align 8
+
+; Function Attrs: nounwind
+define void @feq1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp oeq float %0, %1
+; CHECK-LABEL:  feq1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.eq.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @fne1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp une float %0, %1
+; CHECK-LABEL:  fne1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.eq.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @flt1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp olt float %0, %1
+; CHECK-LABEL:  flt1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.olt.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @fgt1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp ogt float %0, %1
+; CHECK-LABEL: fgt1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ule.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @fle1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp ole float %0, %1
+; CHECK-LABEL:  fle1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ole.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @fge1()  {
+entry:
+  %0 = load float* @f1, align 4
+  %1 = load float* @f2, align 4
+  %cmp = fcmp oge float %0, %1
+; CHECK-LABEL:  fge1:
+; CHECK-DAG:    lw      $[[REG_F2_GOT:[0-9]+]], %got(f2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_F1_GOT:[0-9]+]], %got(f1)(${{[0-9]+}})
+; CHECK-DAG:    lwc1    $f[[REG_F2:[0-9]+]], 0($[[REG_F2_GOT]])
+; CHECK-DAG:    lwc1    $f[[REG_F1:[0-9]+]], 0($[[REG_F1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ult.s  $f[[REG_F1]], $f[[REG_F2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @deq1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp oeq double %0, %1
+; CHECK-LABEL:  deq1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.eq.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @dne1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp une double %0, %1
+; CHECK-LABEL:  dne1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.eq.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @dlt1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp olt double %0, %1
+; CHECK-LABEL:  dlt1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.olt.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @dgt1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp ogt double %0, %1
+; CHECK-LABEL:  dgt1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ule.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @dle1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp ole double %0, %1
+; CHECK-LABEL:  dle1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ole.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movt  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @dge1()  {
+entry:
+  %0 = load double* @d1, align 8
+  %1 = load double* @d2, align 8
+  %cmp = fcmp oge double %0, %1
+; CHECK-LABEL:  dge1:
+; CHECK-DAG:    lw      $[[REG_D2_GOT:[0-9]+]], %got(d2)(${{[0-9]+}})
+; CHECK-DAG:    lw      $[[REG_D1_GOT:[0-9]+]], %got(d1)(${{[0-9]+}})
+; CHECK-DAG:    ldc1    $f[[REG_D2:[0-9]+]], 0($[[REG_D2_GOT]])
+; CHECK-DAG:    ldc1    $f[[REG_D1:[0-9]+]], 0($[[REG_D1_GOT]])
+; CHECK-DAG:    addiu   $[[REG_ZERO:[0-9]+]], $zero, 0
+; CHECK-DAG:    addiu   $[[REG_ONE:[0-9]+]], $zero, 1
+; CHECK:        c.ult.d  $f[[REG_D1]], $f[[REG_D2]]
+; CHECK:        movf  $[[REG_ZERO]], $[[REG_ONE]], $fcc0
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/Fast-ISel/fpext.ll b/test/CodeGen/Mips/Fast-ISel/fpext.ll
new file mode 100644
index 0000000..98aca75
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/fpext.ll
@@ -0,0 +1,21 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@f = global float 0x40147E6B80000000, align 4
+@d_f = common global double 0.000000e+00, align 8
+@.str = private unnamed_addr constant [6 x i8] c"%f  \0A\00", align 1
+
+; Function Attrs: nounwind
+define void @dv() #0 {
+entry:
+  %0 = load float* @f, align 4
+  %conv = fpext float %0 to double
+; CHECK: cvt.d.s  $f{{[0-9]+}}, $f{{[0-9]+}}
+  store double %conv, double* @d_f, align 8
+  ret void
+}
+
+
+attributes #1 = { nounwind }
diff --git a/test/CodeGen/Mips/Fast-ISel/fpintconv.ll b/test/CodeGen/Mips/Fast-ISel/fpintconv.ll
new file mode 100644
index 0000000..846726a
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/fpintconv.ll
@@ -0,0 +1,35 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+
+@f = global float 0x40D6E83280000000, align 4
+@d = global double 0x4132D68780000000, align 8
+@i_f = common global i32 0, align 4
+@i_d = common global i32 0, align 4
+@.str = private unnamed_addr constant [5 x i8] c"%i \0A\00", align 1
+
+; Function Attrs: nounwind
+define void @ifv() {
+entry:
+; CHECK-LABEL:   .ent  ifv
+  %0 = load float* @f, align 4
+  %conv = fptosi float %0 to i32
+; CHECK:   trunc.w.s  $f[[REG:[0-9]+]], $f{{[0-9]+}}
+; CHECK:   mfc1	${{[0-9]+}}, $f[[REG]]
+  store i32 %conv, i32* @i_f, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @idv() {
+entry:
+; CHECK-LABEL:   .ent  idv
+  %0 = load double* @d, align 8
+  %conv = fptosi double %0 to i32
+; CHECK:   trunc.w.d  $f[[REG:[0-9]+]], $f{{[0-9]+}}
+; CHECK:   mfc1	${{[0-9]+}}, $f[[REG]]
+  store i32 %conv, i32* @i_d, align 4
+  ret void
+}
diff --git a/test/CodeGen/Mips/Fast-ISel/fptrunc.ll b/test/CodeGen/Mips/Fast-ISel/fptrunc.ll
new file mode 100644
index 0000000..d843dee
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/fptrunc.ll
@@ -0,0 +1,20 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@d = global double 0x40147E6B74DF0446, align 8
+@f = common global float 0.000000e+00, align 4
+@.str = private unnamed_addr constant [6 x i8] c"%f  \0A\00", align 1
+
+; Function Attrs: nounwind
+define void @fv() #0 {
+entry:
+  %0 = load double* @d, align 8
+  %conv = fptrunc double %0 to float
+; CHECK: cvt.s.d  $f{{[0-9]+}}, $f{{[0-9]+}}
+  store float %conv, float* @f, align 4
+  ret void
+}
+
+attributes #1 = { nounwind }
diff --git a/test/CodeGen/Mips/Fast-ISel/icmpa.ll b/test/CodeGen/Mips/Fast-ISel/icmpa.ll
new file mode 100644
index 0000000..bd41a29
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/icmpa.ll
@@ -0,0 +1,210 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@c = global i32 4, align 4
+@d = global i32 9, align 4
+@uc = global i32 4, align 4
+@ud = global i32 9, align 4
+@b1 = common global i32 0, align 4
+
+; Function Attrs: nounwind
+define void @eq()  {
+entry:
+; CHECK-LABEL:  .ent  eq
+
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp eq i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:  xor  $[[REG1:[0-9]+]], $[[REG_C]], $[[REG_D]]
+; CHECK:  sltiu  $[[REG2:[0-9]+]], $[[REG1]], 1
+; FIXME: This instruction is redundant. The sltiu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG2]], 1
+
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @ne()  {
+entry:
+; CHECK-LABEL:  .ent  ne
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp ne i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:  xor  $[[REG1:[0-9]+]], $[[REG_C]], $[[REG_D]]
+; CHECK:  sltu  $[[REG2:[0-9]+]], $zero, $[[REG1]]
+; FIXME: This instruction is redundant. The sltu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG2]], 1
+
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @ugt()  {
+entry:
+; CHECK-LABEL:  .ent  ugt
+  %0 = load i32* @uc, align 4
+  %1 = load i32* @ud, align 4
+  %cmp = icmp ugt i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_UD_GOT:[0-9+]]], %got(ud)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UC_GOT:[0-9+]]], %got(uc)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UD:[0-9]+]], 0($[[REG_UD_GOT]])
+; CHECK-DAG:  lw	$[[REG_UC:[0-9]+]], 0($[[REG_UC_GOT]])
+; CHECK:  sltu  $[[REG1:[0-9]+]], $[[REG_UD]], $[[REG_UC]]
+; FIXME: This instruction is redundant. The sltu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 1
+
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @ult()  {
+entry:
+; CHECK-LABEL:  .ent  ult
+  %0 = load i32* @uc, align 4
+  %1 = load i32* @ud, align 4
+  %cmp = icmp ult i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_UD_GOT:[0-9+]]], %got(ud)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UC_GOT:[0-9+]]], %got(uc)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UD:[0-9]+]], 0($[[REG_UD_GOT]])
+; CHECK-DAG:  lw	$[[REG_UC:[0-9]+]], 0($[[REG_UC_GOT]])
+; CHECK:  sltu  $[[REG1:[0-9]+]], $[[REG_UC]], $[[REG_UD]]
+; FIXME: This instruction is redundant. The sltu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @uge()  {
+entry:
+; CHECK-LABEL:  .ent  uge
+  %0 = load i32* @uc, align 4
+  %1 = load i32* @ud, align 4
+  %cmp = icmp uge i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_UD_GOT:[0-9+]]], %got(ud)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UC_GOT:[0-9+]]], %got(uc)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UD:[0-9]+]], 0($[[REG_UD_GOT]])
+; CHECK-DAG:  lw	$[[REG_UC:[0-9]+]], 0($[[REG_UC_GOT]])
+; CHECK:  sltu  $[[REG1:[0-9]+]], $[[REG_UC]], $[[REG_UD]]
+; CHECK:  xori  $[[REG2:[0-9]+]], $[[REG1]], 1
+; FIXME: This instruction is redundant. The sltu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG2]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @ule()  {
+entry:
+; CHECK-LABEL:  .ent  ule
+  %0 = load i32* @uc, align 4
+  %1 = load i32* @ud, align 4
+  %cmp = icmp ule i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_UD_GOT:[0-9+]]], %got(ud)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UC_GOT:[0-9+]]], %got(uc)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_UD:[0-9]+]], 0($[[REG_UD_GOT]])
+; CHECK-DAG:  lw	$[[REG_UC:[0-9]+]], 0($[[REG_UC_GOT]])
+; CHECK:  sltu  $[[REG1:[0-9]+]], $[[REG_UD]], $[[REG_UC]]
+; CHECK:  xori  $[[REG2:[0-9]+]], $[[REG1]], 1
+; FIXME: This instruction is redundant. The sltu can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG2]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @sgt()  {
+entry:
+; CHECK-LABEL:  .ent sgt
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp sgt i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:  slt  $[[REG1:[0-9]+]], $[[REG_D]], $[[REG_C]]
+; FIXME: This instruction is redundant. The slt can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @slt()  {
+entry:
+; CHECK-LABEL:  .ent slt
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp slt i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:  slt  $[[REG1:[0-9]+]], $[[REG_C]], $[[REG_D]]
+; FIXME: This instruction is redundant. The slt can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @sge()  {
+entry:
+; CHECK-LABEL:  .ent sge
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp sge i32 %0, %1
+  %conv = zext i1 %cmp to i32
+  store i32 %conv, i32* @b1, align 4
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:  slt  $[[REG1:[0-9]+]], $[[REG_C]], $[[REG_D]]
+; CHECK:  xori  $[[REG2:[0-9]+]], $[[REG1]], 1
+; FIXME: This instruction is redundant. The slt can only produce 0 and 1.
+; CHECK:  andi  ${{[0-9]+}}, $[[REG2]], 1
+  ret void
+}
+
+; Function Attrs: nounwind
+define void @sle()  {
+entry:
+; CHECK-LABEL:  .ent sle
+  %0 = load i32* @c, align 4
+  %1 = load i32* @d, align 4
+  %cmp = icmp sle i32 %0, %1
+  %conv = zext i1 %cmp to i32
+; CHECK-DAG:  lw	$[[REG_D_GOT:[0-9+]]], %got(d)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_C_GOT:[0-9+]]], %got(c)(${{[0-9]+}})
+; CHECK-DAG:  lw	$[[REG_D:[0-9]+]], 0($[[REG_D_GOT]])
+; CHECK-DAG:  lw	$[[REG_C:[0-9]+]], 0($[[REG_C_GOT]])
+; CHECK:        slt     $[[REG1:[0-9]+]], $[[REG_D]], $[[REG_C]]
+; CHECK:        xori    $[[REG2:[0-9]+]], $[[REG1]], 1
+; FIXME: This instruction is redundant. The slt can only produce 0 and 1.
+; CHECK:        andi    ${{[0-9]+}}, $[[REG2]], 1
+  store i32 %conv, i32* @b1, align 4
+  ret void
+}
diff --git a/test/CodeGen/Mips/Fast-ISel/loadstore2.ll b/test/CodeGen/Mips/Fast-ISel/loadstore2.ll
index f113a0e..d84478b 100644
--- a/test/CodeGen/Mips/Fast-ISel/loadstore2.ll
+++ b/test/CodeGen/Mips/Fast-ISel/loadstore2.ll
@@ -6,6 +6,8 @@ target triple = "mips--linux-gnu"
 @c1 = common global i8 0, align 1
 ; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
 ; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
 
 @s2 = common global i16 0, align 2
 @s1 = common global i16 0, align 2
diff --git a/test/CodeGen/Mips/Fast-ISel/loadstoreconv.ll b/test/CodeGen/Mips/Fast-ISel/loadstoreconv.ll
new file mode 100644
index 0000000..f7f2c64
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/loadstoreconv.ll
@@ -0,0 +1,179 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32r2
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32
+
+@b2 = global i8 0, align 1
+@b1 = global i8 1, align 1
+@uc1 = global i8 0, align 1
+@uc2 = global i8 -1, align 1
+@sc1 = global i8 -128, align 1
+@sc2 = global i8 127, align 1
+@ss1 = global i16 -32768, align 2
+@ss2 = global i16 32767, align 2
+@us1 = global i16 0, align 2
+@us2 = global i16 -1, align 2
+@ssi = global i16 0, align 2
+@ssj = global i16 0, align 2
+@i = global i32 0, align 4
+@j = global i32 0, align 4
+@.str = private unnamed_addr constant [4 x i8] c"%i\0A\00", align 1
+@.str1 = private unnamed_addr constant [7 x i8] c"%i %i\0A\00", align 1
+
+; Function Attrs: nounwind
+define void @_Z3b_iv()  {
+entry:
+; CHECK-LABEL:   .ent  _Z3b_iv
+  %0 = load i8* @b1, align 1
+  %tobool = trunc i8 %0 to i1
+  %frombool = zext i1 %tobool to i8
+  store i8 %frombool, i8* @b2, align 1
+  %1 = load i8* @b2, align 1
+  %tobool1 = trunc i8 %1 to i1
+  %conv = zext i1 %tobool1 to i32
+  store i32 %conv, i32* @i, align 4
+; CHECK:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; CHECK:  andi  $[[REG2:[0-9]+]], $[[REG1]], 1
+; CHECK:  sb  $[[REG2]], 0(${{[0-9]+}})
+
+
+
+  ret void
+; CHECK:   .end  _Z3b_iv
+}
+
+; Function Attrs: nounwind
+define void @_Z4uc_iv()  {
+entry:
+; CHECK-LABEL:  .ent  _Z4uc_iv
+
+  %0 = load i8* @uc1, align 1
+  %conv = zext i8 %0 to i32
+  store i32 %conv, i32* @i, align 4
+  %1 = load i8* @uc2, align 1
+  %conv1 = zext i8 %1 to i32
+; CHECK:   lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 255
+
+  store i32 %conv1, i32* @j, align 4
+  ret void
+; CHECK:  .end  _Z4uc_iv
+
+}
+
+; Function Attrs: nounwind
+define void @_Z4sc_iv()  {
+entry:
+; mips32r2-LABEL:  .ent  _Z4sc_iv
+; mips32-LABEL:  .ent  _Z4sc_iv
+
+  %0 = load i8* @sc1, align 1
+  %conv = sext i8 %0 to i32
+  store i32 %conv, i32* @i, align 4
+  %1 = load i8* @sc2, align 1
+  %conv1 = sext i8 %1 to i32
+  store i32 %conv1, i32* @j, align 4
+; mips32r2:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32r2:  seb  ${{[0-9]+}}, $[[REG1]]
+; mips32:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32:    sll  $[[REG2:[0-9]+]], $[[REG1]], 24
+; mips32:    sra  ${{[0-9]+}}, $[[REG2]], 24
+
+  ret void
+; CHECK:  .end  _Z4sc_iv
+}
+
+; Function Attrs: nounwind
+define void @_Z4us_iv()  {
+entry:
+; CHECK-LABEL:  .ent  _Z4us_iv
+  %0 = load i16* @us1, align 2
+  %conv = zext i16 %0 to i32
+  store i32 %conv, i32* @i, align 4
+  %1 = load i16* @us2, align 2
+  %conv1 = zext i16 %1 to i32
+  store i32 %conv1, i32* @j, align 4
+  ret void
+; CHECK:  lhu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 65535
+; CHECK:  .end  _Z4us_iv
+}
+
+; Function Attrs: nounwind
+define void @_Z4ss_iv()  {
+entry:
+; mips32r2-LABEL:  .ent  _Z4ss_iv
+; mips32=LABEL:  .ent  _Z4ss_iv
+
+  %0 = load i16* @ss1, align 2
+  %conv = sext i16 %0 to i32
+  store i32 %conv, i32* @i, align 4
+  %1 = load i16* @ss2, align 2
+  %conv1 = sext i16 %1 to i32
+  store i32 %conv1, i32* @j, align 4
+; mips32r2:  lhu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32r2:  seh  ${{[0-9]+}}, $[[REG1]]
+; mips32:    lhu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32:    sll  $[[REG2:[0-9]+]], $[[REG1]], 16
+; mips32:    sra  ${{[0-9]+}}, $[[REG2]], 16
+
+  ret void
+; CHECK:  .end  _Z4ss_iv
+}
+
+; Function Attrs: nounwind
+define void @_Z4b_ssv()  {
+entry:
+; CHECK-LABEL:  .ent  _Z4b_ssv
+  %0 = load i8* @b2, align 1
+  %tobool = trunc i8 %0 to i1
+  %conv = zext i1 %tobool to i16
+  store i16 %conv, i16* @ssi, align 2
+  ret void
+; CHECK:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 1
+; CHECK:  .end  _Z4b_ssv
+}
+
+; Function Attrs: nounwind
+define void @_Z5uc_ssv()  {
+entry:
+; CHECK-LABEL:  .ent  _Z5uc_ssv
+  %0 = load i8* @uc1, align 1
+  %conv = zext i8 %0 to i16
+  store i16 %conv, i16* @ssi, align 2
+  %1 = load i8* @uc2, align 1
+  %conv1 = zext i8 %1 to i16
+; CHECK:   lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; CHECK:  andi  ${{[0-9]+}}, $[[REG1]], 255
+
+  store i16 %conv1, i16* @ssj, align 2
+  ret void
+; CHECK:  .end  _Z5uc_ssv
+}
+
+; Function Attrs: nounwind
+define void @_Z5sc_ssv()  {
+entry:
+; mips32r2-LABEL:  .ent  _Z5sc_ssv
+; mips32-LABEL:  .ent  _Z5sc_ssv
+  %0 = load i8* @sc1, align 1
+  %conv = sext i8 %0 to i16
+  store i16 %conv, i16* @ssi, align 2
+  %1 = load i8* @sc2, align 1
+  %conv1 = sext i8 %1 to i16
+  store i16 %conv1, i16* @ssj, align 2
+; mips32r2:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32r2:  seb  ${{[0-9]+}}, $[[REG1]]
+; mips32:  lbu  $[[REG1:[0-9]+]], 0(${{[0-9]+}})
+; mips32:    sll  $[[REG2:[0-9]+]], $[[REG1]], 24
+; mips32:    sra  ${{[0-9]+}}, $[[REG2]], 24
+
+  ret void
+; CHECK:  .end  _Z5sc_ssv
+}
+
diff --git a/test/CodeGen/Mips/Fast-ISel/loadstrconst.ll b/test/CodeGen/Mips/Fast-ISel/loadstrconst.ll
new file mode 100644
index 0000000..93cf4c1
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/loadstrconst.ll
@@ -0,0 +1,21 @@
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
+
+@.str = private unnamed_addr constant [6 x i8] c"hello\00", align 1
+@s = common global i8* null, align 4
+
+; Function Attrs: nounwind
+define void @foo() #0 {
+entry:
+  store i8* getelementptr inbounds ([6 x i8]* @.str, i32 0, i32 0), i8** @s, align 4
+  ret void
+; CHECK:        .ent    foo
+; CHECK:        lw      $[[REG1:[0-9]+]], %got($.str)(${{[0-9]+}})
+; CHECK:        addiu   ${{[0-9]+}}, $[[REG1]], %lo($.str)
+
+}
+
+attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+
diff --git a/test/CodeGen/Mips/Fast-ISel/nullvoid.ll b/test/CodeGen/Mips/Fast-ISel/nullvoid.ll
index eeaff87..c847561 100644
--- a/test/CodeGen/Mips/Fast-ISel/nullvoid.ll
+++ b/test/CodeGen/Mips/Fast-ISel/nullvoid.ll
@@ -1,5 +1,7 @@
 ; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
 ; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
 
 ; Function Attrs: nounwind
 define void @foo() {
diff --git a/test/CodeGen/Mips/Fast-ISel/shift.ll b/test/CodeGen/Mips/Fast-ISel/shift.ll
new file mode 100644
index 0000000..18fd5ac
--- /dev/null
+++ b/test/CodeGen/Mips/Fast-ISel/shift.ll
@@ -0,0 +1,24 @@
+; RUN: llc -march=mipsel -mcpu=mips32r2 -O1 -fast-isel=true -mips-fast-isel -filetype=obj %s -o - \
+; RUN:   | llvm-objdump -arch mipsel -mcpu=mips32r2 -d - | FileCheck %s
+
+; This test checks that encoding for srl is correct when fast-isel for mips32r2 is used.
+
+%struct.s = type { [4 x i8], i32 }
+
+define i32 @main() nounwind uwtable {
+entry:
+  %foo = alloca %struct.s, align 4
+  %0 = bitcast %struct.s* %foo to i32*
+  %bf.load = load i32* %0, align 4
+  %bf.lshr = lshr i32 %bf.load, 2
+  %cmp = icmp ne i32 %bf.lshr, 2
+  br i1 %cmp, label %if.then, label %if.end
+
+if.then:
+  unreachable
+
+if.end:
+  ret i32 0
+}
+
+; CHECK: srl    ${{[0-9]+}}, ${{[0-9]+}}, {{[0-9]+}}
diff --git a/test/CodeGen/Mips/Fast-ISel/simplestore.ll b/test/CodeGen/Mips/Fast-ISel/simplestore.ll
index 5d52481..83e3f3f 100644
--- a/test/CodeGen/Mips/Fast-ISel/simplestore.ll
+++ b/test/CodeGen/Mips/Fast-ISel/simplestore.ll
@@ -1,5 +1,7 @@
 ; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
 ; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
 
 @abcd = external global i32
 
diff --git a/test/CodeGen/Mips/Fast-ISel/simplestorefp1.ll b/test/CodeGen/Mips/Fast-ISel/simplestorefp1.ll
index 6759c01..74723ae 100644
--- a/test/CodeGen/Mips/Fast-ISel/simplestorefp1.ll
+++ b/test/CodeGen/Mips/Fast-ISel/simplestorefp1.ll
@@ -1,5 +1,11 @@
 ; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s 
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
 ; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32r2 
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s -check-prefix=mips32
 
 @f = common global float 0.000000e+00, align 4
 @de = common global double 0.000000e+00, align 8
@@ -23,15 +29,25 @@ entry:
 define void @d1() #0 {
 entry:
   store double 1.234567e+00, double* @de, align 8
-; CHECK:  .ent  d1
-; CHECK:  lui  $[[REG1a:[0-9]+]], 16371
-; CHECK:  ori  $[[REG2a:[0-9]+]], $[[REG1a]], 49353
-; CHECK:  lui  $[[REG1b:[0-9]+]], 21403
-; CHECK:  ori  $[[REG2b:[0-9]+]], $[[REG1b]], 34951
-; CHECK:  mtc1  $[[REG2b]], $f[[REG3:[0-9]+]]
-; CHECK:  mthc1  $[[REG2a]], $f[[REG3]]
-; CHECK:  sdc1  $f[[REG3]], 0(${{[0-9]+}})
-; CHECK:  .end  d1
+; mip32r2:  .ent  d1
+; mips32r2:  lui  $[[REG1a:[0-9]+]], 16371
+; mips32r2:  ori  $[[REG2a:[0-9]+]], $[[REG1a]], 49353
+; mips32r2:  lui  $[[REG1b:[0-9]+]], 21403
+; mips32r2:  ori  $[[REG2b:[0-9]+]], $[[REG1b]], 34951
+; mips32r2:  mtc1  $[[REG2b]], $f[[REG3:[0-9]+]]
+; mips32r2:  mthc1  $[[REG2a]], $f[[REG3]]
+; mips32r2:  sdc1  $f[[REG3]], 0(${{[0-9]+}})
+; mips32r2:  .end  d1
+; mips32:  .ent  d1
+; mips32:  lui  $[[REG1a:[0-9]+]], 16371
+; mips32:  ori  $[[REG2a:[0-9]+]], $[[REG1a]], 49353
+; mips32:  lui  $[[REG1b:[0-9]+]], 21403
+; mips32:  ori  $[[REG2b:[0-9]+]], $[[REG1b]], 34951
+; mips32:  mtc1  $[[REG2b]], $f[[REG3:[0-9]+]]
+; mips32:  mtc1  $[[REG2a]], $f{{[0-9]+}}
+; mips32:  sdc1  $f[[REG3]], 0(${{[0-9]+}})
+; mips32:  .end  d1
+
   ret void
 }
 
diff --git a/test/CodeGen/Mips/Fast-ISel/simplestorei.ll b/test/CodeGen/Mips/Fast-ISel/simplestorei.ll
index 7d2c8e7..128e1de 100644
--- a/test/CodeGen/Mips/Fast-ISel/simplestorei.ll
+++ b/test/CodeGen/Mips/Fast-ISel/simplestorei.ll
@@ -1,5 +1,7 @@
 ; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32r2 \
 ; RUN:     < %s | FileCheck %s
+; RUN: llc -march=mipsel -relocation-model=pic -O0 -mips-fast-isel -fast-isel-abort -mcpu=mips32 \
+; RUN:     < %s | FileCheck %s
 
 @ijk = external global i32
 
diff --git a/test/CodeGen/Mips/abicalls.ll b/test/CodeGen/Mips/abicalls.ll
index 6fa33aa..7edc3e2 100644
--- a/test/CodeGen/Mips/abicalls.ll
+++ b/test/CodeGen/Mips/abicalls.ll
@@ -1,16 +1,11 @@
-; 
-; When the assembler is ready a .s file for it will
-; be created.
+; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 -relocation-model=static %s -o - | FileCheck -check-prefix=ABICALLS -check-prefix=STATIC %s
+; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 %s -o - | FileCheck -check-prefix=ABICALLS -check-prefix=PIC %s
+; RUN: llc -filetype=asm -mtriple mips64el-unknown-linux -mcpu=mips4 -relocation-model=static %s -o - | FileCheck -check-prefix=ABICALLS -check-prefix=PIC %s
+; RUN: llc -filetype=asm -mtriple mips64el-unknown-linux -mcpu=mips64 -relocation-model=static %s -o - | FileCheck -check-prefix=ABICALLS -check-prefix=PIC %s
 
-; Note that EF_MIPS_CPIC is set by -mabicalls which is the default on Linux
-; TODO need to support -mno-abicalls
+; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 -mattr noabicalls -relocation-model=static %s -o - | FileCheck -implicit-check-not='.abicalls' -implicit-check-not='pic0' %s
 
-; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 -relocation-model=static %s -o - | FileCheck -check-prefix=CHECK-STATIC %s
-; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 %s -o - | FileCheck -check-prefix=CHECK-PIC %s
-; RUN: llc -filetype=asm -mtriple mips64el-unknown-linux -mcpu=mips4 -relocation-model=static %s -o - | FileCheck -check-prefix=CHECK-PIC %s
-; RUN: llc -filetype=asm -mtriple mips64el-unknown-linux -mcpu=mips64 -relocation-model=static %s -o - | FileCheck -check-prefix=CHECK-PIC %s
+; ABICALLS: .abicalls
 
-; CHECK-STATIC: .abicalls
-; CHECK-STATIC-NEXT: pic0
-; CHECK-PIC: .abicalls
-; CHECK-PIC-NOT: pic0
+; STATIC: pic0
+; PIC-NOT: pic0
diff --git a/test/CodeGen/Mips/abiflags-xx.ll b/test/CodeGen/Mips/abiflags-xx.ll
index b8aa071..c461012 100644
--- a/test/CodeGen/Mips/abiflags-xx.ll
+++ b/test/CodeGen/Mips/abiflags-xx.ll
@@ -1,5 +1,4 @@
 ; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips32 -mattr=fpxx %s -o - | FileCheck %s
-; XFAIL: *
 
 ; CHECK: .nan    legacy
 ; CHECK: .module fp=xx
diff --git a/test/CodeGen/Mips/abiflags32.ll b/test/CodeGen/Mips/abiflags32.ll
index 093964f..e32d4a5 100644
--- a/test/CodeGen/Mips/abiflags32.ll
+++ b/test/CodeGen/Mips/abiflags32.ll
@@ -3,10 +3,15 @@
 ; RUN: llc -filetype=asm -mtriple mipsel-unknown-linux -mcpu=mips64 -mattr=-n64,n32 %s -o - | FileCheck  -check-prefix=CHECK-64n %s
 
 ; CHECK: .nan    legacy
-; CHECK: .module fp=32
+; We don't emit '.module fp=32' for compatibility with binutils 2.24 which
+; doesn't accept .module.
+; CHECK-NOT: .module fp=32
 
 ; CHECK-64: .nan    legacy
+; We do emit '.module fp=64' though since it contradicts the default value.
 ; CHECK-64: .module fp=64
 
 ; CHECK-64n: .nan    legacy
-; CHECK-64n: .module fp=64
+; We don't emit '.module fp=64' for compatibility with binutils 2.24 which
+; doesn't accept .module.
+; CHECK-64n-NOT: .module fp=64
diff --git a/test/CodeGen/Mips/atomic.ll b/test/CodeGen/Mips/atomic.ll
index 066d42c..78fd829 100644
--- a/test/CodeGen/Mips/atomic.ll
+++ b/test/CodeGen/Mips/atomic.ll
@@ -8,11 +8,11 @@
 
 ; Keep one big-endian check so that we don't reduce testing, but don't add more
 ; since endianness doesn't affect the body of the atomic operations.
-; RUN: llc -march=mips   --disable-machine-licm -mcpu=mips32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=MIPS32-ANY -check-prefix=CHECK-EB
+; RUN: llc -march=mips   --disable-machine-licm -mcpu=mips32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=MIPS32-ANY -check-prefix=NO-SEB-SEH -check-prefix=CHECK-EB
 
 @x = common global i32 0, align 4
 
-define i32 @AtomicLoadAdd32(i32 %incr) nounwind {
+define i32 @AtomicLoadAdd32(i32 signext %incr) nounwind {
 entry:
   %0 = atomicrmw add i32* @x, i32 %incr monotonic
   ret i32 %0
@@ -29,7 +29,7 @@ entry:
 ; ALL:           beqz    $[[R2]], $[[BB0]]
 }
 
-define i32 @AtomicLoadNand32(i32 %incr) nounwind {
+define i32 @AtomicLoadNand32(i32 signext %incr) nounwind {
 entry:
   %0 = atomicrmw nand i32* @x, i32 %incr monotonic
   ret i32 %0
@@ -47,7 +47,7 @@ entry:
 ; ALL:           beqz    $[[R2]], $[[BB0]]
 }
 
-define i32 @AtomicSwap32(i32 %newval) nounwind {
+define i32 @AtomicSwap32(i32 signext %newval) nounwind {
 entry:
   %newval.addr = alloca i32, align 4
   store i32 %newval, i32* %newval.addr, align 4
@@ -66,7 +66,7 @@ entry:
 ; ALL:           beqz    $[[R2]], $[[BB0]]
 }
 
-define i32 @AtomicCmpSwap32(i32 %oldval, i32 %newval) nounwind {
+define i32 @AtomicCmpSwap32(i32 signext %oldval, i32 signext %newval) nounwind {
 entry:
   %newval.addr = alloca i32, align 4
   store i32 %newval, i32* %newval.addr, align 4
@@ -246,6 +246,7 @@ entry:
 ; NO-SEB-SEH:    sra     $2, $[[R17]], 24
 
 ; HAS-SEB-SEH:   seb     $2, $[[R16]]
+
 }
 
 define signext i8 @AtomicCmpSwap8(i8 signext %oldval, i8 signext %newval) nounwind {
@@ -292,6 +293,49 @@ entry:
 ; HAS-SEB-SEH:   seb     $2, $[[R17]]
 }
 
+define i1 @AtomicCmpSwapRes8(i8* %ptr, i8 signext %oldval, i8 signext %newval) nounwind {
+entry:
+  %0 = cmpxchg i8* %ptr, i8 %oldval, i8 %newval monotonic monotonic
+  %1 = extractvalue { i8, i1 } %0, 1
+  ret i1 %1
+; ALL-LABEL: AtomicCmpSwapRes8
+
+; ALL:           addiu   $[[R1:[0-9]+]], $zero, -4
+; ALL:           and     $[[R2:[0-9]+]], $4, $[[R1]]
+; ALL:           andi    $[[R3:[0-9]+]], $4, 3
+; CHECK-EL:      sll     $[[R5:[0-9]+]], $[[R3]], 3
+; CHECK-EB:      xori    $[[R4:[0-9]+]], $[[R3]], 3
+; CHECK-EB:      sll     $[[R5:[0-9]+]], $[[R4]], 3
+; ALL:           ori     $[[R6:[0-9]+]], $zero, 255
+; ALL:           sllv    $[[R7:[0-9]+]], $[[R6]], $[[R5]]
+; ALL:           nor     $[[R8:[0-9]+]], $zero, $[[R7]]
+; ALL:           andi    $[[R9:[0-9]+]], $5, 255
+; ALL:           sllv    $[[R10:[0-9]+]], $[[R9]], $[[R5]]
+; ALL:           andi    $[[R11:[0-9]+]], $6, 255
+; ALL:           sllv    $[[R12:[0-9]+]], $[[R11]], $[[R5]]
+
+; ALL:       $[[BB0:[A-Z_0-9]+]]:
+; ALL:           ll      $[[R13:[0-9]+]], 0($[[R2]])
+; ALL:           and     $[[R14:[0-9]+]], $[[R13]], $[[R7]]
+; ALL:           bne     $[[R14]], $[[R10]], $[[BB1:[A-Z_0-9]+]]
+
+; ALL:           and     $[[R15:[0-9]+]], $[[R13]], $[[R8]]
+; ALL:           or      $[[R16:[0-9]+]], $[[R15]], $[[R12]]
+; ALL:           sc      $[[R16]], 0($[[R2]])
+; ALL:           beqz    $[[R16]], $[[BB0]]
+
+; ALL:       $[[BB1]]:
+; ALL:           srlv    $[[R17:[0-9]+]], $[[R14]], $[[R5]]
+
+; NO-SEB-SEH:    sll     $[[R18:[0-9]+]], $[[R17]], 24
+; NO-SEB-SEH:    sra     $[[R19:[0-9]+]], $[[R18]], 24
+
+; HAS-SEB-SEH:   seb     $[[R19:[0-9]+]], $[[R17]]
+
+; ALL:           xor     $[[R20:[0-9]+]], $[[R19]], $5
+; ALL:           sltiu   $2, $[[R20]], 1
+}
+
 ; Check one i16 so that we cover the seh sign extend
 @z = common global i16 0, align 1
 
@@ -337,7 +381,7 @@ entry:
 
 @countsint = common global i32 0, align 4
 
-define i32 @CheckSync(i32 %v) nounwind noinline {
+define i32 @CheckSync(i32 signext %v) nounwind noinline {
 entry:
   %0 = atomicrmw add i32* @countsint, i32 %v seq_cst
   ret i32 %0 
@@ -371,7 +415,7 @@ entry:
 
 ; Check that MIPS32R6 has the correct offset range.
 ; FIXME: At the moment, we don't seem to do addr+offset for any atomic load/store.
-define i32 @AtomicLoadAdd32_OffGt9Bit(i32 %incr) nounwind {
+define i32 @AtomicLoadAdd32_OffGt9Bit(i32 signext %incr) nounwind {
 entry:
   %0 = atomicrmw add i32* getelementptr(i32* @x, i32 256), i32 %incr monotonic
   ret i32 %0
diff --git a/test/CodeGen/Mips/bswap.ll b/test/CodeGen/Mips/bswap.ll
index 812eef1..f182e65 100644
--- a/test/CodeGen/Mips/bswap.ll
+++ b/test/CodeGen/Mips/bswap.ll
@@ -2,7 +2,7 @@
 ; RUN: llc  < %s -march=mips64el -mcpu=mips64r2 | FileCheck %s -check-prefix=MIPS64
 ; RUN: llc  < %s -mtriple=mipsel-linux-gnu -march=mipsel -mcpu=mips32r2 -mattr=+mips16 | FileCheck %s -check-prefix=MIPS16
 
-define i32 @bswap32(i32 %x) nounwind readnone {
+define i32 @bswap32(i32 signext %x) nounwind readnone {
 entry:
 ; MIPS32-LABEL: bswap32:
 ; MIPS32: wsbh $[[R0:[0-9]+]]
@@ -29,7 +29,7 @@ entry:
   ret i32 %or.3
 }
 
-define i64 @bswap64(i64 %x) nounwind readnone {
+define i64 @bswap64(i64 signext %x) nounwind readnone {
 entry:
 ; MIPS32-LABEL: bswap64:
 ; MIPS32: wsbh $[[R0:[0-9]+]]
@@ -72,24 +72,24 @@ entry:
 define <4 x i32> @bswapv4i32(<4 x i32> %x) nounwind readnone {
 entry:
 ; MIPS32-LABEL: bswapv4i32:
-; MIPS32: wsbh $[[R0:[0-9]+]]
-; MIPS32: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS32: wsbh $[[R0:[0-9]+]]
-; MIPS32: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS32: wsbh $[[R0:[0-9]+]]
-; MIPS32: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS32: wsbh $[[R0:[0-9]+]]
-; MIPS32: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS32-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS32-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS32-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS32-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS32-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS32-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS32-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS32-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
 
 ; MIPS64-LABEL: bswapv4i32:
-; MIPS64: wsbh $[[R0:[0-9]+]]
-; MIPS64: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS64: wsbh $[[R0:[0-9]+]]
-; MIPS64: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS64: wsbh $[[R0:[0-9]+]]
-; MIPS64: rotr ${{[0-9]+}}, $[[R0]], 16
-; MIPS64: wsbh $[[R0:[0-9]+]]
-; MIPS64: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS64-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS64-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS64-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS64-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS64-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS64-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
+; MIPS64-DAG: wsbh $[[R0:[0-9]+]]
+; MIPS64-DAG: rotr ${{[0-9]+}}, $[[R0]], 16
 
 ; Don't bother with a MIPS16 version. It's just bswap32 repeated four times and
 ; would be very long
diff --git a/test/CodeGen/Mips/buildpairextractelementf64.ll b/test/CodeGen/Mips/buildpairextractelementf64.ll
index 88d1d07..7682a98 100644
--- a/test/CodeGen/Mips/buildpairextractelementf64.ll
+++ b/test/CodeGen/Mips/buildpairextractelementf64.ll
@@ -1,15 +1,19 @@
-; RUN: llc -march=mipsel < %s | FileCheck %s -check-prefix=FP32 -check-prefix=CHECK
-; RUN: llc -march=mips  < %s | FileCheck %s -check-prefix=FP32 -check-prefix=CHECK
-; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+fp64 < %s | FileCheck %s -check-prefix=FP64 -check-prefix=CHECK
-; RUN: llc -march=mips -mcpu=mips32r2 -mattr=+fp64 < %s | FileCheck %s -check-prefix=FP64 -check-prefix=CHECK
+; RUN: llc -march=mipsel < %s | FileCheck %s -check-prefix=NO-MFHC1 -check-prefix=ALL
+; RUN: llc -march=mips  < %s | FileCheck %s -check-prefix=NO-MFHC1 -check-prefix=ALL
+; RUN: llc -march=mipsel -mcpu=mips32r2 < %s | FileCheck %s -check-prefix=HAS-MFHC1 -check-prefix=ALL
+; RUN: llc -march=mips -mcpu=mips32r2 < %s | FileCheck %s -check-prefix=HAS-MFHC1 -check-prefix=ALL
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+fp64 < %s | FileCheck %s -check-prefix=HAS-MFHC1 -check-prefix=ALL
+; RUN: llc -march=mips -mcpu=mips32r2 -mattr=+fp64 < %s | FileCheck %s -check-prefix=HAS-MFHC1 -check-prefix=ALL
 
 @a = external global i32
 
-; CHECK-LABEL: f:
-; FP32: mtc1
-; FP32: mtc1
-; FP64-DAG: mtc1
-; FP64-DAG: mthc1
+; ALL-LABEL: f:
+
+; NO-MFHC1: mtc1
+; NO-MFHC1: mtc1
+
+; HAS-MFHC1-DAG: mtc1
+; HAS-MFHC1-DAG: mthc1
 
 define double @f(i32 %a1, double %d) nounwind {
 entry:
@@ -18,11 +22,13 @@ entry:
   ret double %add
 }
 
-; CHECK-LABEL: f3:
-; FP32: mfc1
-; FP32: mfc1
-; FP64-DAG: mfc1
-; FP64-DAG: mfhc1
+; ALL-LABEL: f3:
+
+; NO-MFHC1: mfc1
+; NO-MFHC1: mfc1
+
+; HAS-MFHC1-DAG: mfc1
+; HAS-MFHC1-DAG: mfhc1
 
 define void @f3(double %d, i32 %a1) nounwind {
 entry:
diff --git a/test/CodeGen/Mips/cconv/arguments-float.ll b/test/CodeGen/Mips/cconv/arguments-float.ll
index e2119ec..14a3baa 100644
--- a/test/CodeGen/Mips/cconv/arguments-float.ll
+++ b/test/CodeGen/Mips/cconv/arguments-float.ll
@@ -69,26 +69,26 @@ entry:
 ; O32-DAG:           sw [[R4]], 28([[R2]])
 ; NEW-DAG:           sd $6, 24([[R2]])
 
-; O32-DAG:           lw [[R3:\$[0-9]+]], 32($sp)
-; O32-DAG:           lw [[R4:\$[0-9]+]], 36($sp)
+; O32-DAG:           lw [[R3:\$([0-9]+|gp)]], 32($sp)
+; O32-DAG:           lw [[R4:\$([0-9]+|gp)]], 36($sp)
 ; O32-DAG:           sw [[R3]], 32([[R2]])
 ; O32-DAG:           sw [[R4]], 36([[R2]])
 ; NEW-DAG:           sd $7, 32([[R2]])
 
-; O32-DAG:           lw [[R3:\$[0-9]+]], 40($sp)
-; O32-DAG:           lw [[R4:\$[0-9]+]], 44($sp)
+; O32-DAG:           lw [[R3:\$([0-9]+|gp)]], 40($sp)
+; O32-DAG:           lw [[R4:\$([0-9]+|gp)]], 44($sp)
 ; O32-DAG:           sw [[R3]], 40([[R2]])
 ; O32-DAG:           sw [[R4]], 44([[R2]])
 ; NEW-DAG:           sd $8, 40([[R2]])
 
-; O32-DAG:           lw [[R3:\$[0-9]+]], 48($sp)
-; O32-DAG:           lw [[R4:\$[0-9]+]], 52($sp)
+; O32-DAG:           lw [[R3:\$([0-9]+|gp)]], 48($sp)
+; O32-DAG:           lw [[R4:\$([0-9]+|gp)]], 52($sp)
 ; O32-DAG:           sw [[R3]], 48([[R2]])
 ; O32-DAG:           sw [[R4]], 52([[R2]])
 ; NEW-DAG:           sd $9, 48([[R2]])
 
-; O32-DAG:           lw [[R3:\$[0-9]+]], 56($sp)
-; O32-DAG:           lw [[R4:\$[0-9]+]], 60($sp)
+; O32-DAG:           lw [[R3:\$([0-9]+|gp)]], 56($sp)
+; O32-DAG:           lw [[R4:\$([0-9]+|gp)]], 60($sp)
 ; O32-DAG:           sw [[R3]], 56([[R2]])
 ; O32-DAG:           sw [[R4]], 60([[R2]])
 ; NEW-DAG:           sd $10, 56([[R2]])
@@ -135,8 +135,8 @@ entry:
 ; SYM64-DAG:           ld [[R2:\$[0-9]]], %got_disp(floats)(
 
 ; The first four arguments are the same in O32/N32/N64.
-; The first argument isn't floating point so floating point registers are not
-; used.
+; The first argument is floating point but soft-float is enabled so floating
+; point registers are not used.
 ; MD00305 and GCC disagree on this one. MD00305 says that floats are treated
 ; as 8-byte aligned and occupy two slots on O32. GCC is treating them as 4-byte
 ; aligned and occupying one slot. We'll use GCC's definition.
@@ -195,7 +195,7 @@ entry:
 ; O32-DAG:           sw $7, 12([[R2]])
 ; NEW-DAG:           sd $5, 8([[R2]])
 
-define void @float_arg2(i8 %a, float %b) nounwind {
+define void @float_arg2(i8 signext %a, float %b) nounwind {
 entry:
         %0 = getelementptr [11 x i8]* @bytes, i32 0, i32 1
         store volatile i8 %a, i8* %0
diff --git a/test/CodeGen/Mips/cconv/arguments-hard-float-varargs.ll b/test/CodeGen/Mips/cconv/arguments-hard-float-varargs.ll
index aadf7d1..70ccf14 100644
--- a/test/CodeGen/Mips/cconv/arguments-hard-float-varargs.ll
+++ b/test/CodeGen/Mips/cconv/arguments-hard-float-varargs.ll
@@ -4,11 +4,11 @@
 ; RUN-TODO: llc -march=mips64 -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
 ; RUN-TODO: llc -march=mips64el -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
 
-; RUN: llc -march=mips64 -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=N32 --check-prefix=NEW %s
-; RUN: llc -march=mips64el -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=N32 --check-prefix=NEW %s
+; RUN: llc -march=mips64 -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=N32 --check-prefix=NEW --check-prefix=NEWBE %s
+; RUN: llc -march=mips64el -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=N32 --check-prefix=NEW --check-prefix=NEWLE %s
 
-; RUN: llc -march=mips64 -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM64 --check-prefix=N64 --check-prefix=NEW %s
-; RUN: llc -march=mips64el -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM64 --check-prefix=N64 --check-prefix=NEW %s
+; RUN: llc -march=mips64 -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM64 --check-prefix=N64 --check-prefix=NEW --check-prefix=NEWBE %s
+; RUN: llc -march=mips64el -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM64 --check-prefix=N64 --check-prefix=NEW --check-prefix=NEWLE %s
 
 ; Test the effect of varargs on floating point types in the non-variable part
 ; of the argument list as specified by section 2 of the MIPSpro N32 Handbook.
@@ -34,6 +34,7 @@ entry:
         %b = va_arg i8** %ap, double
         %1 = getelementptr [11 x double]* @doubles, i32 0, i32 2
         store volatile double %b, double* %1
+        call void @llvm.va_end(i8* %ap2)
         ret void
 }
 
@@ -98,6 +99,7 @@ entry:
         %b = va_arg i8** %ap, float
         %1 = getelementptr [11 x float]* @floats, i32 0, i32 2
         store volatile float %b, float* %1
+        call void @llvm.va_end(i8* %ap2)
         ret void
 }
 
@@ -140,16 +142,18 @@ entry:
 ; Increment the pointer then get the varargs arg
 ; LLVM will rebind the load to the stack pointer instead of the varargs pointer
 ; during lowering. This is fine and doesn't change the behaviour.
-; N32/N64 is using ori instead of addiu/daddiu but (although odd) this is fine
-; since the stack is always aligned.
+; Also, in big-endian mode the offset must be increased by 4 to retrieve the
+; correct half of the argument slot.
+;
 ; O32-DAG:           addiu [[VAPTR]], [[VAPTR]], 4
 ; O32-DAG:           sw [[VAPTR]], 4($sp)
-; N32-DAG:           ori [[VAPTR]], [[VAPTR]], 4
+; N32-DAG:           addiu [[VAPTR]], [[VAPTR]], 8
 ; N32-DAG:           sw [[VAPTR]], 4($sp)
-; N64-DAG:           ori [[VAPTR]], [[VAPTR]], 4
+; N64-DAG:           daddiu [[VAPTR]], [[VAPTR]], 8
 ; N64-DAG:           sd [[VAPTR]], 0($sp)
 ; O32-DAG:           lwc1 [[FTMP1:\$f[0-9]+]], 12($sp)
-; NEW-DAG:           lwc1 [[FTMP1:\$f[0-9]+]], 8($sp)
+; NEWLE-DAG:         lwc1 [[FTMP1:\$f[0-9]+]], 8($sp)
+; NEWBE-DAG:         lwc1 [[FTMP1:\$f[0-9]+]], 12($sp)
 ; ALL-DAG:           swc1 [[FTMP1]], 8([[R2]])
 
 declare void @llvm.va_start(i8*)
diff --git a/test/CodeGen/Mips/cconv/arguments-varargs.ll b/test/CodeGen/Mips/cconv/arguments-varargs.ll
new file mode 100644
index 0000000..adacda5
--- /dev/null
+++ b/test/CodeGen/Mips/cconv/arguments-varargs.ll
@@ -0,0 +1,1104 @@
+; RUN: llc -mtriple=mips-linux -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=O32 --check-prefix=O32-BE %s
+; RUN: llc -mtriple=mipsel-linux -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=O32 --check-prefix=O32-LE %s
+
+; RUN-TODO: llc -march=mips64 -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=O32 %s
+; RUN-TODO: llc -march=mips64el -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=O32 %s
+
+; RUN: llc -mtriple=mips64-linux -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=NEW --check-prefix=N32 --check-prefix=NEW-BE %s
+; RUN: llc -mtriple=mips64el-linux -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=NEW --check-prefix=N32 --check-prefix=NEW-LE %s
+
+; RUN: llc -march=mips64 -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=NEW --check-prefix=N64 --check-prefix=NEW-BE %s
+; RUN: llc -march=mips64el -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=NEW --check-prefix=N64 --check-prefix=NEW-LE %s
+
+@hwords = global [3 x i16] zeroinitializer, align 1
+@words  = global [3 x i32] zeroinitializer, align 1
+@dwords = global [3 x i64] zeroinitializer, align 1
+
+define void @fn_i16_dotdotdot_i16(i16 %a, ...) {
+entry:
+; ALL-LABEL: fn_i16_dotdotdot_i16:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(hwords)(
+
+; ALL-DAG:       sh [[ARG1]], 2([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sh [[ARG2]], 4([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i16
+  %e1 = getelementptr [3 x i16]* @hwords, i32 0, i32 1
+  store volatile i16 %arg1, i16* %e1, align 2
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i16
+  %e2 = getelementptr [3 x i16]* @hwords, i32 0, i32 2
+  store volatile i16 %arg2, i16* %e2, align 2
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i16_dotdotdot_i32(i16 %a, ...) {
+entry:
+; ALL-LABEL: fn_i16_dotdotdot_i32:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(words)(
+
+; ALL-DAG:       sw [[ARG1]], 4([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sw [[ARG2]], 8([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i32
+  %e1 = getelementptr [3 x i32]* @words, i32 0, i32 1
+  store volatile i32 %arg1, i32* %e1, align 4
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i32
+  %e2 = getelementptr [3 x i32]* @words, i32 0, i32 2
+  store volatile i32 %arg2, i32* %e2, align 4
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i16_dotdotdot_i64(i16 %a, ...) {
+entry:
+; ALL-LABEL: fn_i16_dotdotdot_i64:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]] (and realign pointer for O32)
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       addiu [[VA_TMP1:\$[0-9]+]], $zero, -8
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion and copy it to the global.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 8([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 12([[GV]])
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(dwords)(
+; NEW-DAG:       ld [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-DAG:       sd [[ARG1]], 8([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; FIXME: We're still aligned from the last one but CodeGen doesn't spot that.
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion and copy it to the global.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 16([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 20([[GV]])
+
+; NEW-DAG:       ld [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-DAG:       sd [[ARG2]], 16([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i64
+  %e1 = getelementptr [3 x i64]* @dwords, i32 0, i32 1
+  store volatile i64 %arg1, i64* %e1, align 8
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i64
+  %e2 = getelementptr [3 x i64]* @dwords, i32 0, i32 2
+  store volatile i64 %arg2, i64* %e2, align 8
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i32_dotdotdot_i16(i32 %a, ...) {
+entry:
+; ALL-LABEL: fn_i32_dotdotdot_i16:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(hwords)(
+
+; ALL-DAG:       sh [[ARG1]], 2([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sh [[ARG2]], 4([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i16
+  %e1 = getelementptr [3 x i16]* @hwords, i32 0, i32 1
+  store volatile i16 %arg1, i16* %e1, align 2
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i16
+  %e2 = getelementptr [3 x i16]* @hwords, i32 0, i32 2
+  store volatile i16 %arg2, i16* %e2, align 2
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i32_dotdotdot_i32(i32 %a, ...) {
+entry:
+; ALL-LABEL: fn_i32_dotdotdot_i32:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(words)(
+
+; ALL-DAG:       sw [[ARG1]], 4([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sw [[ARG2]], 8([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i32
+  %e1 = getelementptr [3 x i32]* @words, i32 0, i32 1
+  store volatile i32 %arg1, i32* %e1, align 4
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i32
+  %e2 = getelementptr [3 x i32]* @words, i32 0, i32 2
+  store volatile i32 %arg2, i32* %e2, align 4
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i32_dotdotdot_i64(i32 %a, ...) {
+entry:
+; ALL-LABEL: fn_i32_dotdotdot_i64:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+; O32-DAG:       sw $5, 12([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 12 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the 4 byte slot for the first
+; fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 12
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]] (and realign pointer for O32)
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       addiu [[VA_TMP1:\$[0-9]+]], $zero, -8
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion and copy it to the global.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 8([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 12([[GV]])
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(dwords)(
+; NEW-DAG:       ld [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-DAG:       sd [[ARG1]], 8([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; FIXME: We're still aligned from the last one but CodeGen doesn't spot that.
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion and copy it to the global.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 16([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 20([[GV]])
+
+; NEW-DAG:       ld [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-DAG:       sd [[ARG2]], 16([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i64
+  %e1 = getelementptr [3 x i64]* @dwords, i32 0, i32 1
+  store volatile i64 %arg1, i64* %e1, align 8
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i64
+  %e2 = getelementptr [3 x i64]* @dwords, i32 0, i32 2
+  store volatile i64 %arg2, i64* %e2, align 8
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i64_dotdotdot_i16(i64 %a, ...) {
+entry:
+; ALL-LABEL: fn_i64_dotdotdot_i16:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 16 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the two 4 byte slots for the
+; first fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 16
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(hwords)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(hwords)(
+
+; ALL-DAG:       sh [[ARG1]], 2([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sh [[ARG2]], 4([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i16
+  %e1 = getelementptr [3 x i16]* @hwords, i32 0, i32 1
+  store volatile i16 %arg1, i16* %e1, align 2
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i16
+  %e2 = getelementptr [3 x i16]* @hwords, i32 0, i32 2
+  store volatile i16 %arg2, i16* %e2, align 2
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i64_dotdotdot_i32(i64 %a, ...) {
+entry:
+; ALL-LABEL: fn_i64_dotdotdot_i32:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 16 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the two 4 byte slots for the
+; first fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 16
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]]
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-BE-DAG:    lw [[ARG1:\$[0-9]+]], 4([[VA]])
+
+; Copy the arg to the global
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(words)
+
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(words)(
+
+; ALL-DAG:       sw [[ARG1]], 4([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+
+; NEW-LE-DAG:    lw [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-BE-DAG:    lw [[ARG2:\$[0-9]+]], 4([[VA2]])
+
+; Copy the arg to the global
+; ALL-DAG:       sw [[ARG2]], 8([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i32
+  %e1 = getelementptr [3 x i32]* @words, i32 0, i32 1
+  store volatile i32 %arg1, i32* %e1, align 4
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i32
+  %e2 = getelementptr [3 x i32]* @words, i32 0, i32 2
+  store volatile i32 %arg2, i32* %e2, align 4
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+define void @fn_i64_dotdotdot_i64(i64 %a, ...) {
+entry:
+; ALL-LABEL: fn_i64_dotdotdot_i64:
+
+; Set up the stack with an 8-byte local area. N32/N64 must also make room for
+; the argument save area (56 bytes).
+; O32:           addiu  [[SP:\$sp]], $sp, -8
+; N32:           addiu  [[SP:\$sp]], $sp, -64
+; N64:           daddiu  [[SP:\$sp]], $sp, -64
+
+; Save variable argument portion on the stack
+; O32-DAG:       sw $7, 20([[SP]])
+; O32-DAG:       sw $6, 16([[SP]])
+
+; NEW-DAG:       sd $11, 56([[SP]])
+; NEW-DAG:       sd $10, 48([[SP]])
+; NEW-DAG:       sd $9, 40([[SP]])
+; NEW-DAG:       sd $8, 32([[SP]])
+; NEW-DAG:       sd $7, 24([[SP]])
+; NEW-DAG:       sd $6, 16([[SP]])
+; NEW-DAG:       sd $5, 8([[SP]])
+
+; Initialize variable argument pointer.
+; For O32, the offset is 16 due to the 4 bytes used to store local variables,
+; 4 bytes padding to maintain stack alignment, and the two 4 byte slots for the
+; first fixed argument.
+; For N32/N64, it is only 8 since the fixed arguments do not reserve stack
+; space.
+; O32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 16
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; N32-DAG:       addiu [[VA:\$[0-9]+]], [[SP]], 8
+; N32-DAG:       sw [[VA]], 0([[SP]])
+
+; N64-DAG:       daddiu [[VA:\$[0-9]+]], [[SP]], 8
+; N64-DAG:       sd [[VA]], 0([[SP]])
+
+; Store [[VA]]
+; O32-DAG:       sw [[VA]], 0([[SP]])
+
+; ALL: # ANCHOR1
+
+; Increment [[VA]] (and realign pointer for O32)
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       addiu [[VA_TMP1:\$[0-9]+]], $zero, -8
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N64-DAG:       ld [[VA:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 8
+; N64-DAG:       sd [[VA2]], 0([[SP]])
+
+; Load the first argument from the variable portion and copy it to the global.
+; This has used the stack pointer directly rather than the [[VA]] we just set
+; up.
+; Big-endian mode for N32/N64 must add an additional 4 to the offset due to byte
+; order.
+; O32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 8([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG1:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG1]], 12([[GV]])
+
+; N32-DAG:       addiu [[GV:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
+; N64-DAG:       ld [[GV:\$[0-9]+]], %got_disp(dwords)(
+; NEW-DAG:       ld [[ARG1:\$[0-9]+]], 0([[VA]])
+; NEW-DAG:       sd [[ARG1]], 8([[GV]])
+
+; ALL: # ANCHOR2
+
+; Increment [[VA]] again.
+; FIXME: We're still aligned from the last one but CodeGen doesn't spot that.
+; O32:           lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA_TMP0:\$[0-9]+]], [[VA]], 7
+; O32-DAG:       and   [[VA_TMP2:\$[0-9]+]], [[VA_TMP0]], [[VA_TMP1]]
+; O32-DAG:       ori   [[VA2:\$[0-9]+]], [[VA_TMP2]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+
+; N32-DAG:       lw [[VA2:\$[0-9]+]], 0([[SP]])
+; N32-DAG:       addiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N32-DAG:       sw [[VA3]], 0([[SP]])
+
+; N64-DAG:       ld [[VA2:\$[0-9]+]], 0([[SP]])
+; N64-DAG:       daddiu [[VA3:\$[0-9]+]], [[VA2]], 8
+; N64-DAG:       sd [[VA3]], 0([[SP]])
+
+; Load the second argument from the variable portion and copy it to the global.
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 16([[GV]])
+; O32-DAG:       lw [[VA:\$[0-9]+]], 0([[SP]])
+; O32-DAG:       addiu [[VA2:\$[0-9]+]], [[VA]], 4
+; O32-DAG:       sw [[VA2]], 0([[SP]])
+; O32-DAG:       lw [[ARG2:\$[0-9]+]], 0([[VA]])
+; O32-DAG:       sw [[ARG2]], 20([[GV]])
+
+; NEW-DAG:       ld [[ARG2:\$[0-9]+]], 0([[VA2]])
+; NEW-DAG:       sd [[ARG2]], 16([[GV]])
+
+  %ap = alloca i8*, align 8
+  %ap2 = bitcast i8** %ap to i8*
+  call void @llvm.va_start(i8* %ap2)
+
+  call void asm sideeffect "# ANCHOR1", ""()
+  %arg1 = va_arg i8** %ap, i64
+  %e1 = getelementptr [3 x i64]* @dwords, i32 0, i32 1
+  store volatile i64 %arg1, i64* %e1, align 8
+
+  call void asm sideeffect "# ANCHOR2", ""()
+  %arg2 = va_arg i8** %ap, i64
+  %e2 = getelementptr [3 x i64]* @dwords, i32 0, i32 2
+  store volatile i64 %arg2, i64* %e2, align 8
+
+  call void @llvm.va_end(i8* %ap2)
+
+  ret void
+}
+
+declare void @llvm.va_start(i8*)
+declare void @llvm.va_end(i8*)
diff --git a/test/CodeGen/Mips/cconv/arguments.ll b/test/CodeGen/Mips/cconv/arguments.ll
index 8fe29f3..43da604 100644
--- a/test/CodeGen/Mips/cconv/arguments.ll
+++ b/test/CodeGen/Mips/cconv/arguments.ll
@@ -1,5 +1,5 @@
-; RUN: llc -march=mips -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 --check-prefix=O32BE %s
-; RUN: llc -march=mipsel -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 --check-prefix=O32LE %s
+; RUN: llc -march=mips -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
+; RUN: llc -march=mipsel -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
 
 ; RUN-TODO: llc -march=mips64 -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
 ; RUN-TODO: llc -march=mips64el -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=SYM32 --check-prefix=O32 %s
@@ -23,8 +23,10 @@
 @floats = global [11 x float] zeroinitializer
 @doubles = global [11 x double] zeroinitializer
 
-define void @align_to_arg_slots(i8 %a, i8 %b, i8 %c, i8 %d, i8 %e, i8 %f, i8 %g,
-                                i8 %h, i8 %i, i8 %j) nounwind {
+define void @align_to_arg_slots(i8 signext %a, i8 signext %b, i8 signext %c,
+                                i8 signext %d, i8 signext %e, i8 signext %f,
+                                i8 signext %g, i8 signext %h, i8 signext %i,
+                                i8 signext %j) nounwind {
 entry:
         %0 = getelementptr [11 x i8]* @bytes, i32 0, i32 1
         store volatile i8 %a, i8* %0
@@ -53,7 +55,7 @@ entry:
 ; We won't test the way the global address is calculated in this test. This is
 ; just to get the register number for the other checks.
 ; SYM32-DAG:           addiu [[R1:\$[0-9]+]], ${{[0-9]+}}, %lo(bytes)
-; SYM64-DAG:           ld [[R1:\$[0-9]]], %got_disp(bytes)(
+; SYM64-DAG:           ld [[R1:\$[0-9]+]], %got_disp(bytes)(
 
 ; The first four arguments are the same in O32/N32/N64
 ; ALL-DAG:           sb $4, 1([[R1]])
@@ -82,15 +84,16 @@ entry:
 ; increase by 4 for O32 and 8 for N32/N64.
 ; O32-DAG:           lw [[R3:\$[0-9]+]], 32($sp)
 ; O32-DAG:           sb [[R3]], 9([[R1]])
-; NEW-DAG:           lw [[R3:\$[0-9]+]], 0($sp)
+; NEW-DAG:           ld [[R3:\$[0-9]+]], 0($sp)
 ; NEW-DAG:           sb [[R3]], 9([[R1]])
 ; O32-DAG:           lw [[R3:\$[0-9]+]], 36($sp)
 ; O32-DAG:           sb [[R3]], 10([[R1]])
-; NEW-DAG:           lw [[R3:\$[0-9]+]], 8($sp)
+; NEW-DAG:           ld [[R3:\$[0-9]+]], 8($sp)
 ; NEW-DAG:           sb [[R3]], 10([[R1]])
 
-define void @slot_skipping(i8 %a, i64 %b, i8 %c, i8 %d,
-                           i8 %e, i8 %f, i8 %g, i64 %i, i8 %j) nounwind {
+define void @slot_skipping(i8 signext %a, i64 signext %b, i8 signext %c,
+                           i8 signext %d, i8 signext %e, i8 signext %f,
+                           i8 signext %g, i64 signext %i, i8 signext %j) nounwind {
 entry:
         %0 = getelementptr [11 x i8]* @bytes, i32 0, i32 1
         store volatile i8 %a, i8* %0
@@ -117,9 +120,9 @@ entry:
 ; We won't test the way the global address is calculated in this test. This is
 ; just to get the register number for the other checks.
 ; SYM32-DAG:           addiu [[R1:\$[0-9]+]], ${{[0-9]+}}, %lo(bytes)
-; SYM64-DAG:           ld [[R1:\$[0-9]]], %got_disp(bytes)(
+; SYM64-DAG:           ld [[R1:\$[0-9]+]], %got_disp(bytes)(
 ; SYM32-DAG:           addiu [[R2:\$[0-9]+]], ${{[0-9]+}}, %lo(dwords)
-; SYM64-DAG:           ld [[R2:\$[0-9]]], %got_disp(dwords)(
+; SYM64-DAG:           ld [[R2:\$[0-9]+]], %got_disp(dwords)(
 
 ; The first argument is the same in O32/N32/N64.
 ; ALL-DAG:           sb $4, 1([[R1]])
@@ -137,8 +140,7 @@ entry:
 ; It's not clear why O32 uses lbu for this argument, but it's not wrong so we'll
 ; accept it for now. The only IR difference is that this argument has
 ; anyext from i8 and align 8 on it.
-; O32LE-DAG:           lbu [[R3:\$[0-9]+]], 16($sp)
-; O32BE-DAG:           lbu [[R3:\$[0-9]+]], 19($sp)
+; O32-DAG:           lw [[R3:\$[0-9]+]], 16($sp)
 ; O32-DAG:           sb [[R3]], 2([[R1]])
 ; NEW-DAG:           sb $6, 2([[R1]])
 ; O32-DAG:           lw [[R3:\$[0-9]+]], 20($sp)
@@ -166,5 +168,5 @@ entry:
 ; increase by 4 for O32 and 8 for N32/N64.
 ; O32-DAG:           lw [[R3:\$[0-9]+]], 48($sp)
 ; O32-DAG:           sb [[R3]], 7([[R1]])
-; NEW-DAG:           lw [[R3:\$[0-9]+]], 0($sp)
+; NEW-DAG:           ld [[R3:\$[0-9]+]], 0($sp)
 ; NEW-DAG:           sb [[R3]], 7([[R1]])
diff --git a/test/CodeGen/Mips/cconv/return-float.ll b/test/CodeGen/Mips/cconv/return-float.ll
index 28cf83d..d1a5e4f 100644
--- a/test/CodeGen/Mips/cconv/return-float.ll
+++ b/test/CodeGen/Mips/cconv/return-float.ll
@@ -30,7 +30,7 @@ entry:
 ; O32-DAG:           lw $2, %lo(float)([[R1]])
 ; N32-DAG:           lui [[R1:\$[0-9]+]], %hi(float)
 ; N32-DAG:           lw $2, %lo(float)([[R1]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(float)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(float)(
 ; N64-DAG:           lw $2, 0([[R1]])
 
 define double @retdouble() nounwind {
@@ -44,5 +44,5 @@ entry:
 ; O32-DAG:           addiu [[R2:\$[0-9]+]], [[R1]], %lo(double)
 ; O32-DAG:           lw $3, 4([[R2]])
 ; N32-DAG:           ld $2, %lo(double)([[R1:\$[0-9]+]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(double)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(double)(
 ; N64-DAG:           ld $2, 0([[R1]])
diff --git a/test/CodeGen/Mips/cconv/return-hard-float.ll b/test/CodeGen/Mips/cconv/return-hard-float.ll
index 371b3a5..123b499 100644
--- a/test/CodeGen/Mips/cconv/return-hard-float.ll
+++ b/test/CodeGen/Mips/cconv/return-hard-float.ll
@@ -10,6 +10,9 @@
 ; RUN: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 %s
 ; RUN: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 %s
 
+; RUN: llc -mtriple=mips-linux-gnu -relocation-model=static -mattr=+o32,+fp64 < %s | FileCheck --check-prefix=ALL --check-prefix=032FP64 %s
+; RUN: llc -mtriple=mipsel-linux-gnu -relocation-model=static -mattr=+o32,+fp64 < %s | FileCheck --check-prefix=ALL --check-prefix=032FP64 %s
+
 ; Test the float returns for all ABI's and byte orders as specified by
 ; section 5 of MD00305 (MIPS ABIs Described).
 
@@ -30,7 +33,7 @@ entry:
 ; O32-DAG:           lwc1 $f0, %lo(float)([[R1]])
 ; N32-DAG:           lui [[R1:\$[0-9]+]], %hi(float)
 ; N32-DAG:           lwc1 $f0, %lo(float)([[R1]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(float)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(float)(
 ; N64-DAG:           lwc1 $f0, 0([[R1]])
 
 define double @retdouble() nounwind {
@@ -42,5 +45,15 @@ entry:
 ; ALL-LABEL: retdouble:
 ; O32-DAG:           ldc1 $f0, %lo(double)([[R1:\$[0-9]+]])
 ; N32-DAG:           ldc1 $f0, %lo(double)([[R1:\$[0-9]+]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(double)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(double)(
 ; N64-DAG:           ldc1 $f0, 0([[R1]])
+
+define { double, double } @retComplexDouble() #0 {
+  %retval = alloca { double, double }, align 8
+  %1 = load { double, double }* %retval
+  ret { double, double } %1
+}
+
+; ALL-LABEL: retComplexDouble:
+; 032FP64-DAG:      ldc1     $f0, 0($sp)
+; 032FP64-DAG:      ldc1     $f2, 8($sp)
diff --git a/test/CodeGen/Mips/cconv/return-hard-struct-f128.ll b/test/CodeGen/Mips/cconv/return-hard-struct-f128.ll
new file mode 100644
index 0000000..2e84477
--- /dev/null
+++ b/test/CodeGen/Mips/cconv/return-hard-struct-f128.ll
@@ -0,0 +1,36 @@
+; RUN: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=N32 %s
+; RUN: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=N32 %s
+
+; RUN: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 %s
+; RUN: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 %s
+
+; Test return of {fp128} agrees with de-facto N32/N64 ABI.
+
+@struct_fp128 = global {fp128} zeroinitializer
+
+define inreg {fp128} @ret_struct_fp128() nounwind {
+entry:
+        %0 = load volatile {fp128}* @struct_fp128
+        ret {fp128} %0
+}
+
+; ALL-LABEL: ret_struct_fp128:
+
+; O32 generates different IR so we don't test it here. It returns the struct
+; indirectly.
+
+; Contrary to the N32/N64 ABI documentation, a struct containing a long double
+; is returned in $f0, and $f1 instead of the usual $f0, and $f2. This is to
+; match the de facto ABI as implemented by GCC.
+; N32-DAG:        lui [[R1:\$[0-9]+]], %hi(struct_fp128)
+; N32-DAG:        ld  [[R2:\$[0-9]+]], %lo(struct_fp128)([[R1]])
+; N32-DAG:        dmtc1 [[R2]], $f0
+; N32-DAG:        addiu [[R3:\$[0-9]+]], [[R1]], %lo(struct_fp128)
+; N32-DAG:        ld  [[R4:\$[0-9]+]], 8([[R3]])
+; N32-DAG:        dmtc1 [[R4]], $f1
+
+; N64-DAG:        ld  [[R1:\$[0-9]+]], %got_disp(struct_fp128)($1)
+; N64-DAG:        ld  [[R2:\$[0-9]+]], 0([[R1]])
+; N64-DAG:        dmtc1 [[R2]], $f0
+; N64-DAG:        ld  [[R4:\$[0-9]+]], 8([[R1]])
+; N64-DAG:        dmtc1 [[R4]], $f1
diff --git a/test/CodeGen/Mips/cconv/return-struct.ll b/test/CodeGen/Mips/cconv/return-struct.ll
new file mode 100644
index 0000000..11a8cf0
--- /dev/null
+++ b/test/CodeGen/Mips/cconv/return-struct.ll
@@ -0,0 +1,232 @@
+; RUN: llc -mtriple=mips-linux-gnu -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=O32 --check-prefix=O32-BE %s
+; RUN: llc -mtriple=mipsel-linux-gnu -relocation-model=static < %s | FileCheck --check-prefix=ALL --check-prefix=O32 --check-prefix=O32-LE %s
+
+; RUN-TODO: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=O32 %s
+; RUN-TODO: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+o32 < %s | FileCheck --check-prefix=ALL --check-prefix=O32 %s
+
+; RUN: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=N32 --check-prefix=N32-BE %s
+; RUN: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+n32 < %s | FileCheck --check-prefix=ALL --check-prefix=N32 --check-prefix=N32-LE %s
+
+; RUN: llc -mtriple=mips64-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 --check-prefix=N64-BE %s
+; RUN: llc -mtriple=mips64el-linux-gnu -relocation-model=static -mattr=-n64,+n64 < %s | FileCheck --check-prefix=ALL --check-prefix=N64 --check-prefix=N64-LE %s
+
+; Test struct returns for all ABI's and byte orders.
+
+@struct_byte = global {i8} zeroinitializer
+@struct_2byte = global {i8,i8} zeroinitializer
+@struct_3xi16 = global {[3 x i16]} zeroinitializer
+@struct_6xi32 = global {[6 x i32]} zeroinitializer
+@struct_128xi16 = global {[128 x i16]} zeroinitializer
+
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
+
+define inreg {i8} @ret_struct_i8() nounwind {
+entry:
+        %0 = load volatile {i8}* @struct_byte
+        ret {i8} %0
+}
+
+; ALL-LABEL: ret_struct_i8:
+; O32-DAG:           lui [[R1:\$[0-9]+]], %hi(struct_byte)
+; O32-DAG:           lbu $2, %lo(struct_byte)([[R1]])
+
+; N32-LE-DAG:        lui [[R1:\$[0-9]+]], %hi(struct_byte)
+; N32-LE-DAG:        lb $2, %lo(struct_byte)([[R1]])
+
+; N32-BE-DAG:        lui [[R1:\$[0-9]+]], %hi(struct_byte)
+; N32-BE-DAG:        lb [[R2:\$[0-9]+]], %lo(struct_byte)([[R1]])
+; N32-BE-DAG:        dsll $2, [[R2]], 56
+
+; N64-LE-DAG:        ld  [[R1:\$[0-9]+]], %got_disp(struct_byte)($1)
+; N64-LE-DAG:        lb $2, 0([[R1]])
+
+; N64-BE-DAG:        ld  [[R1:\$[0-9]+]], %got_disp(struct_byte)($1)
+; N64-BE-DAG:        lb [[R2:\$[0-9]+]], 0([[R1]])
+; N64-BE-DAG:        dsll $2, [[R2]], 56
+
+; This test is based on the way clang currently lowers {i8,i8} to {i16}.
+; FIXME: It should probably work for without any lowering too but this doesn't
+;        work as expected. Each member gets mapped to a register rather than
+;        packed into a single register.
+define inreg {i16} @ret_struct_i16() nounwind {
+entry:
+        %retval = alloca {i8,i8}, align 1
+        %0 = bitcast {i8,i8}* %retval to i8*
+        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds ({i8,i8}* @struct_2byte, i32 0, i32 0), i64 2, i32 1, i1 false)
+        %1 = bitcast {i8,i8}* %retval to {i16}*
+        %2 = load volatile {i16}* %1
+        ret {i16} %2
+}
+
+; ALL-LABEL: ret_struct_i16:
+; O32-DAG:           lui [[R1:\$[0-9]+]], %hi(struct_2byte)
+; O32-DAG:           lhu [[R2:\$[0-9]+]], %lo(struct_2byte)([[R1]])
+; O32-DAG:           sh  [[R2]], 0([[SP:\$sp]])
+; O32-DAG:           lhu $2, 0([[SP:\$sp]])
+
+; N32-LE-DAG:        lui [[R1:\$[0-9]+]], %hi(struct_2byte)
+; N32-LE-DAG:        lhu [[R2:\$[0-9]+]], %lo(struct_2byte)([[R1]])
+; N32-LE-DAG:        sh  [[R2]], 8([[SP:\$sp]])
+; N32-LE-DAG:        lh  $2, 8([[SP:\$sp]])
+
+; N32-BE-DAG:        lui [[R1:\$[0-9]+]], %hi(struct_2byte)
+; N32-BE-DAG:        lhu [[R2:\$[0-9]+]], %lo(struct_2byte)([[R1]])
+; N32-BE-DAG:        sh  [[R2]], 8([[SP:\$sp]])
+; N32-BE-DAG:        lh  [[R3:\$[0-9]+]], 8([[SP:\$sp]])
+; N32-BE-DAG:        dsll $2, [[R3]], 48
+
+; N64-LE-DAG:        ld  [[R1:\$[0-9]+]], %got_disp(struct_2byte)($1)
+; N64-LE-DAG:        lhu [[R2:\$[0-9]+]], 0([[R1]])
+; N64-LE-DAG:        sh  [[R2]], 8([[SP:\$sp]])
+; N64-LE-DAG:        lh  $2, 8([[SP:\$sp]])
+
+; N64-BE-DAG:        ld  [[R1:\$[0-9]+]], %got_disp(struct_2byte)($1)
+; N64-BE-DAG:        lhu [[R2:\$[0-9]+]], 0([[R1]])
+; N64-BE-DAG:        sh  [[R2]], 8([[SP:\$sp]])
+; N64-BE-DAG:        lh  [[R3:\$[0-9]+]], 8([[SP:\$sp]])
+; N64-BE-DAG:        dsll $2, [[R3]], 48
+
+; Ensure that structures bigger than 32-bits but smaller than 64-bits are
+; also returned in the upper bits on big endian targets. Previously, these were
+; missed by the CCPromoteToType and the shift didn't happen.
+define inreg {i48} @ret_struct_3xi16() nounwind {
+entry:
+        %0 = load volatile i48* bitcast ({[3 x i16]}* @struct_3xi16 to i48*), align 2
+        %1 = insertvalue {i48} undef, i48 %0, 0
+        ret {i48} %1
+}
+
+; ALL-LABEL: ret_struct_3xi16:
+
+; O32-BE-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_3xi16)
+; O32-BE-DAG:        addiu [[PTR_LO:\$[0-9]+]], [[PTR_HI]], %lo(struct_3xi16)
+; O32-BE-DAG:        lhu [[R1:\$[0-9]+]], 4([[PTR_LO]])
+; O32-BE-DAG:        lw [[R2:\$[0-9]+]], %lo(struct_3xi16)([[PTR_HI]])
+; O32-BE-DAG:        sll [[R3:\$[0-9]+]], [[R2]], 16
+; O32-BE-DAG:        or  $3, [[R1]], [[R3]]
+; O32-BE-DAG:        srl $2, [[R2]], 16
+
+; O32-LE-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_3xi16)
+; O32-LE-DAG:        addiu [[PTR_LO:\$[0-9]+]], [[PTR_HI]], %lo(struct_3xi16)
+; O32-LE-DAG:        lhu $3, 4([[PTR_LO]])
+; O32-LE-DAG:        lw $2, %lo(struct_3xi16)([[PTR_HI]])
+
+; N32-LE-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_3xi16)
+; N32-LE-DAG:        addiu [[PTR_LO:\$[0-9]+]], [[PTR_HI]], %lo(struct_3xi16)
+; N32-LE-DAG:        lh [[R1:\$[0-9]+]], 4([[PTR_LO]])
+; N32-LE-DAG:        lwu [[R2:\$[0-9]+]], %lo(struct_3xi16)([[PTR_HI]])
+; N32-LE-DAG:        dsll [[R3:\$[0-9]+]], [[R1]], 32
+; N32-LE-DAG:        or $2, [[R2]], [[R3]]
+
+; N32-BE-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_3xi16)
+; N32-BE-DAG:        addiu [[PTR_LO:\$[0-9]+]], [[PTR_HI]], %lo(struct_3xi16)
+; N32-BE-DAG:        lw [[R1:\$[0-9]+]], %lo(struct_3xi16)([[PTR_HI]])
+; N32-BE-DAG:        dsll [[R2:\$[0-9]+]], [[R1]], 16
+; N32-BE-DAG:        lhu [[R3:\$[0-9]+]], 4([[PTR_LO]])
+; N32-BE-DAG:        or [[R4:\$[0-9]+]], [[R3]], [[R2]]
+; N32-BE-DAG:        dsll $2, [[R4]], 16
+
+; N64-LE-DAG:        ld  [[PTR:\$[0-9]+]], %got_disp(struct_3xi16)($1)
+; N64-LE-DAG:        lh [[R1:\$[0-9]+]], 4([[PTR]])
+; N64-LE-DAG:        lwu [[R2:\$[0-9]+]], 0([[PTR]])
+; N64-LE-DAG:        dsll [[R3:\$[0-9]+]], [[R1]], 32
+; N64-LE-DAG:        or $2, [[R2]], [[R3]]
+
+; N64-BE-DAG:        ld  [[PTR:\$[0-9]+]], %got_disp(struct_3xi16)($1)
+; N64-BE-DAG:        lw [[R1:\$[0-9]+]], 0([[PTR]])
+; N64-BE-DAG:        dsll [[R2:\$[0-9]+]], [[R1]], 16
+; N64-BE-DAG:        lhu [[R3:\$[0-9]+]], 4([[PTR]])
+; N64-BE-DAG:        or [[R4:\$[0-9]+]], [[R3]], [[R2]]
+; N32-BE-DAG:        dsll $2, [[R4]], 16
+
+; Ensure that large structures (>128-bit) are returned indirectly.
+; We pick an extremely large structure so we don't have to match inlined memcpy's.
+define void @ret_struct_128xi16({[128 x i16]}* sret %returnval) {
+entry:
+        %0 = bitcast {[128 x i16]}* %returnval to i8*
+        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ({[128 x i16]}* @struct_128xi16 to i8*), i64 256, i32 2, i1 false)
+        ret void
+}
+
+; ALL-LABEL: ret_struct_128xi16:
+
+; sret pointer is already in $4
+; O32-DAG:        lui [[PTR:\$[0-9]+]], %hi(struct_128xi16)
+; O32-DAG:        addiu $5, [[PTR]], %lo(struct_128xi16)
+; O32:            jal memcpy
+
+; sret pointer is already in $4
+; N32-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_128xi16)
+; N32-DAG:        addiu [[PTR:\$[0-9]+]], [[PTR_HI]], %lo(struct_128xi16)
+; FIXME: This signext isn't necessary. Like integers, pointers are
+;        but unlike integers, pointers cannot have the signext attribute.
+; N32-DAG:        sll $5, [[PTR]], 0
+; N32:            jal memcpy
+
+; sret pointer is already in $4
+; N64-DAG:        ld $5, %got_disp(struct_128xi16)(
+; N64-DAG:        ld $25, %call16(memcpy)(
+; N64:            jalr $25
+
+; Ensure that large structures (>128-bit) are returned indirectly.
+; This will generate inlined memcpy's anyway so pick the smallest large
+; structure
+; This time we let the backend lower the sret argument.
+define {[6 x i32]} @ret_struct_6xi32() {
+entry:
+        %0 = load volatile {[6 x i32]}* @struct_6xi32, align 2
+        ret {[6 x i32]} %0
+}
+
+; ALL-LABEL: ret_struct_6xi32:
+
+; sret pointer is already in $4
+; O32-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_6xi32)
+; O32-DAG:        addiu [[PTR:\$[0-9]+]], [[PTR_HI]], %lo(struct_6xi32)
+; O32-DAG:        lw [[T0:\$[0-9]+]], %lo(struct_6xi32)([[PTR]])
+; O32-DAG:        lw [[T1:\$[0-9]+]], 4([[PTR]])
+; O32-DAG:        lw [[T2:\$[0-9]+]], 8([[PTR]])
+; O32-DAG:        lw [[T3:\$[0-9]+]], 12([[PTR]])
+; O32-DAG:        lw [[T4:\$[0-9]+]], 16([[PTR]])
+; O32-DAG:        lw [[T5:\$[0-9]+]], 20([[PTR]])
+; O32-DAG:        sw [[T0]], 0($4)
+; O32-DAG:        sw [[T1]], 4($4)
+; O32-DAG:        sw [[T2]], 8($4)
+; O32-DAG:        sw [[T3]], 12($4)
+; O32-DAG:        sw [[T4]], 16($4)
+; O32-DAG:        sw [[T5]], 20($4)
+
+; FIXME: This signext isn't necessary. Like integers, pointers are
+;        but unlike integers, pointers cannot have the signext attribute.
+;        In this case we don't have anywhere to put the signext either since
+;        the sret argument is invented by the backend.
+; N32-DAG:        sll [[RET_PTR:\$[0-9]+]], $4, 0
+; N32-DAG:        lui [[PTR_HI:\$[0-9]+]], %hi(struct_6xi32)
+; N32-DAG:        addiu [[PTR:\$[0-9]+]], [[PTR_HI]], %lo(struct_6xi32)
+; N32-DAG:        lw [[T0:\$[0-9]+]], %lo(struct_6xi32)([[PTR]])
+; N32-DAG:        lw [[T1:\$[0-9]+]], 4([[PTR]])
+; N32-DAG:        lw [[T2:\$[0-9]+]], 8([[PTR]])
+; N32-DAG:        lw [[T3:\$[0-9]+]], 12([[PTR]])
+; N32-DAG:        lw [[T4:\$[0-9]+]], 16([[PTR]])
+; N32-DAG:        lw [[T5:\$[0-9]+]], 20([[PTR]])
+; N32-DAG:        sw [[T0]], 0([[RET_PTR]])
+; N32-DAG:        sw [[T1]], 4([[RET_PTR]])
+; N32-DAG:        sw [[T2]], 8([[RET_PTR]])
+; N32-DAG:        sw [[T3]], 12([[RET_PTR]])
+; N32-DAG:        sw [[T4]], 16([[RET_PTR]])
+; N32-DAG:        sw [[T5]], 20([[RET_PTR]])
+
+; sret pointer is already in $4
+; N64-DAG:        ld [[PTR:\$[0-9]+]], %got_disp(struct_6xi32)(
+; N64-DAG:        lw [[T0:\$[0-9]+]], 0([[PTR]])
+; N64-DAG:        lw [[T1:\$[0-9]+]], 4([[PTR]])
+; N64-DAG:        lw [[T2:\$[0-9]+]], 8([[PTR]])
+; N64-DAG:        lw [[T3:\$[0-9]+]], 12([[PTR]])
+; N64-DAG:        lw [[T4:\$[0-9]+]], 16([[PTR]])
+; N64-DAG:        lw [[T5:\$[0-9]+]], 20([[PTR]])
+; N64-DAG:        sw [[T0]], 0($4)
+; N64-DAG:        sw [[T1]], 4($4)
+; N64-DAG:        sw [[T2]], 8($4)
+; N64-DAG:        sw [[T3]], 12($4)
+; N64-DAG:        sw [[T4]], 16($4)
+; N64-DAG:        sw [[T5]], 20($4)
diff --git a/test/CodeGen/Mips/cconv/return.ll b/test/CodeGen/Mips/cconv/return.ll
index 76ce5e4..63f9b5f 100644
--- a/test/CodeGen/Mips/cconv/return.ll
+++ b/test/CodeGen/Mips/cconv/return.ll
@@ -33,7 +33,7 @@ entry:
 ; O32-DAG:           lbu $2, %lo(byte)([[R1]])
 ; N32-DAG:           lui [[R1:\$[0-9]+]], %hi(byte)
 ; N32-DAG:           lbu $2, %lo(byte)([[R1]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(byte)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(byte)(
 ; N64-DAG:           lbu $2, 0([[R1]])
 
 define i32 @reti32() nounwind {
@@ -47,7 +47,7 @@ entry:
 ; O32-DAG:           lw $2, %lo(word)([[R1]])
 ; N32-DAG:           lui [[R1:\$[0-9]+]], %hi(word)
 ; N32-DAG:           lw $2, %lo(word)([[R1]])
-; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(word)($1)
+; N64-DAG:           ld  [[R1:\$[0-9]+]], %got_disp(word)(
 ; N64-DAG:           lw $2, 0([[R1]])
 
 define i64 @reti64() nounwind {
diff --git a/test/CodeGen/Mips/cfi_offset.ll b/test/CodeGen/Mips/cfi_offset.ll
new file mode 100644
index 0000000..e23855b
--- /dev/null
+++ b/test/CodeGen/Mips/cfi_offset.ll
@@ -0,0 +1,41 @@
+; RUN: llc -march=mips -mattr=+o32 < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EB
+; RUN: llc -march=mipsel -mattr=+o32 < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EL
+; RUN: llc -march=mips -mattr=+o32,+fpxx < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EB
+; RUN: llc -march=mipsel -mattr=+o32,+fpxx < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EL
+; RUN: llc -march=mips -mattr=+o32,+fp64 < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EB
+; RUN: llc -march=mipsel -mattr=+o32,+fp64 < %s | FileCheck %s --check-prefix=CHECK --check-prefix=CHECK-EL
+
+@var = global double 0.0
+
+declare void @foo(...)
+
+define void @bar() {
+
+; CHECK-LABEL:  bar:
+
+; CHECK:  .cfi_def_cfa_offset 40
+; CHECK:  sdc1  $f22, 32($sp)
+; CHECK:  sdc1  $f20, 24($sp)
+; CHECK:  sw  $ra, 20($sp)
+; CHECK:  sw  $16, 16($sp)
+
+; CHECK-EB:  .cfi_offset 55, -8
+; CHECK-EB:  .cfi_offset 54, -4
+; CHECK-EB:  .cfi_offset 53, -16
+; CHECK-EB:  .cfi_offset 52, -12
+
+; CHECK-EL:  .cfi_offset 54, -8
+; CHECK-EL:  .cfi_offset 55, -4
+; CHECK-EL:  .cfi_offset 52, -16
+; CHECK-EL:  .cfi_offset 53, -12
+
+; CHECK:  .cfi_offset 31, -20
+; CHECK:  .cfi_offset 16, -24
+
+    %val1 = load volatile double* @var
+    %val2 = load volatile double* @var
+    call void (...)* @foo() nounwind
+    store volatile double %val1, double* @var
+    store volatile double %val2, double* @var
+    ret void
+}
diff --git a/test/CodeGen/Mips/cmov.ll b/test/CodeGen/Mips/cmov.ll
index 999bdb4..b12c2df 100644
--- a/test/CodeGen/Mips/cmov.ll
+++ b/test/CodeGen/Mips/cmov.ll
@@ -38,7 +38,7 @@
 ; 64-CMP-DAG:   or $[[T2:[0-9]+]], $[[T0]], $[[T1]]
 ; 64-CMP-DAG:   ld $2, 0($[[T2]])
 
-define i32* @cmov1(i32 %s) nounwind readonly {
+define i32* @cmov1(i32 signext %s) nounwind readonly {
 entry:
   %tobool = icmp ne i32 %s, 0
   %tmp1 = load i32** @i3, align 4
@@ -78,7 +78,7 @@ entry:
 ; 64-CMP-DAG:   or $[[T2:[0-9]+]], $[[T0]], $[[T1]]
 ; 64-CMP-DAG:   lw $2, 0($[[T2]])
 
-define i32 @cmov2(i32 %s) nounwind readonly {
+define i32 @cmov2(i32 signext %s) nounwind readonly {
 entry:
   %tobool = icmp ne i32 %s, 0
   %tmp1 = load i32* @c, align 4
@@ -109,13 +109,46 @@ entry:
 ; 64-CMP-DAG:   selnez $[[T1:[0-9]+]], $6, $[[CC]]
 ; 64-CMP-DAG:   or $2, $[[T0]], $[[T1]]
 
-define i32 @cmov3(i32 %a, i32 %b, i32 %c) nounwind readnone {
+define i32 @cmov3(i32 signext %a, i32 signext %b, i32 signext %c) nounwind readnone {
 entry:
   %cmp = icmp eq i32 %a, 234
   %cond = select i1 %cmp, i32 %b, i32 %c
   ret i32 %cond
 }
 
+; ALL-LABEL: cmov3_ne:
+
+; We won't check the result register since we can't know if the move is first
+; or last. We do know it will be either one of two registers so we can at least
+; check that.
+
+; FIXME: Use xori instead of addiu+xor.
+; 32-CMOV:      addiu $[[R0:[0-9]+]], $zero, 234
+; 32-CMOV:      xor $[[R1:[0-9]+]], $4, $[[R0]]
+; 32-CMOV:      movn ${{[26]}}, $5, $[[R1]]
+
+; 32-CMP-DAG:   xori $[[CC:[0-9]+]], $4, 234
+; 32-CMP-DAG:   selnez $[[T0:[0-9]+]], $5, $[[CC]]
+; 32-CMP-DAG:   seleqz $[[T1:[0-9]+]], $6, $[[CC]]
+; 32-CMP-DAG:   or $2, $[[T0]], $[[T1]]
+
+; FIXME: Use xori instead of addiu+xor.
+; 64-CMOV:      addiu $[[R0:[0-9]+]], $zero, 234
+; 64-CMOV:      xor $[[R1:[0-9]+]], $4, $[[R0]]
+; 64-CMOV:      movn ${{[26]}}, $5, $[[R1]]
+
+; 64-CMP-DAG:   xori $[[CC:[0-9]+]], $4, 234
+; 64-CMP-DAG:   selnez $[[T0:[0-9]+]], $5, $[[CC]]
+; 64-CMP-DAG:   seleqz $[[T1:[0-9]+]], $6, $[[CC]]
+; 64-CMP-DAG:   or $2, $[[T0]], $[[T1]]
+
+define i32 @cmov3_ne(i32 signext %a, i32 signext %b, i32 signext %c) nounwind readnone {
+entry:
+  %cmp = icmp ne i32 %a, 234
+  %cond = select i1 %cmp, i32 %b, i32 %c
+  ret i32 %cond
+}
+
 ; ALL-LABEL: cmov4:
 
 ; We won't check the result register since we can't know if the move is first
@@ -146,13 +179,54 @@ entry:
 ; 64-CMP-DAG:  selnez $[[T1:[0-9]+]], $6, $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i64 @cmov4(i32 %a, i64 %b, i64 %c) nounwind readnone {
+define i64 @cmov4(i32 signext %a, i64 %b, i64 %c) nounwind readnone {
 entry:
   %cmp = icmp eq i32 %a, 234
   %cond = select i1 %cmp, i64 %b, i64 %c
   ret i64 %cond
 }
 
+; ALL-LABEL: cmov4_ne:
+
+; We won't check the result register since we can't know if the move is first
+; or last. We do know it will be one of two registers so we can at least check
+; that.
+
+; FIXME: Use xori instead of addiu+xor.
+; 32-CMOV-DAG: addiu $[[R0:[0-9]+]], $zero, 234
+; 32-CMOV-DAG: xor $[[R1:[0-9]+]], $4, $[[R0]]
+; 32-CMOV-DAG: lw $[[R2:2]], 16($sp)
+; 32-CMOV-DAG: lw $[[R3:3]], 20($sp)
+; 32-CMOV-DAG: movn $[[R2]], $6, $[[R1]]
+; 32-CMOV-DAG: movn $[[R3]], $7, $[[R1]]
+
+; 32-CMP-DAG:  xori $[[R0:[0-9]+]], $4, 234
+; 32-CMP-DAG:  lw $[[R1:[0-9]+]], 16($sp)
+; 32-CMP-DAG:  lw $[[R2:[0-9]+]], 20($sp)
+; 32-CMP-DAG:  selnez $[[T0:[0-9]+]], $6, $[[R0]]
+; 32-CMP-DAG:  selnez $[[T1:[0-9]+]], $7, $[[R0]]
+; 32-CMP-DAG:  seleqz $[[T2:[0-9]+]], $[[R1]], $[[R0]]
+; 32-CMP-DAG:  seleqz $[[T3:[0-9]+]], $[[R2]], $[[R0]]
+; 32-CMP-DAG:  or $2, $[[T0]], $[[T2]]
+; 32-CMP-DAG:  or $3, $[[T1]], $[[T3]]
+
+; FIXME: Use xori instead of addiu+xor.
+; 64-CMOV: addiu $[[R0:[0-9]+]], $zero, 234
+; 64-CMOV: xor $[[R1:[0-9]+]], $4, $[[R0]]
+; 64-CMOV: movn ${{[26]}}, $5, $[[R1]]
+
+; 64-CMP-DAG:  xori $[[R0:[0-9]+]], $4, 234
+; 64-CMP-DAG:  selnez $[[T0:[0-9]+]], $5, $[[R0]]
+; 64-CMP-DAG:  seleqz $[[T1:[0-9]+]], $6, $[[R0]]
+; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
+
+define i64 @cmov4_ne(i32 signext %a, i64 %b, i64 %c) nounwind readnone {
+entry:
+  %cmp = icmp ne i32 %a, 234
+  %cond = select i1 %cmp, i64 %b, i64 %c
+  ret i64 %cond
+}
+
 ; slti and conditional move.
 ;
 ; Check that, pattern
@@ -189,7 +263,7 @@ entry:
 ; 64-CMP-DAG:  selnez $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @slti0(i32 %a) {
+define i32 @slti0(i32 signext %a) {
 entry:
   %cmp = icmp sgt i32 %a, 32766
   %cond = select i1 %cmp, i32 3, i32 5
@@ -228,7 +302,7 @@ entry:
 ; 64-CMP-DAG:  seleqz $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @slti1(i32 %a) {
+define i32 @slti1(i32 signext %a) {
 entry:
   %cmp = icmp sgt i32 %a, 32767
   %cond = select i1 %cmp, i32 7, i32 5
@@ -263,7 +337,7 @@ entry:
 ; 64-CMP-DAG:  selnez $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @slti2(i32 %a) {
+define i32 @slti2(i32 signext %a) {
 entry:
   %cmp = icmp sgt i32 %a, -32769
   %cond = select i1 %cmp, i32 3, i32 5
@@ -306,7 +380,7 @@ entry:
 ; 64-CMP-DAG:  seleqz $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @slti3(i32 %a) {
+define i32 @slti3(i32 signext %a) {
 entry:
   %cmp = icmp sgt i32 %a, -32770
   %cond = select i1 %cmp, i32 3, i32 5
@@ -493,7 +567,7 @@ entry:
 ; 64-CMP-DAG:  selnez $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @sltiu0(i32 %a) {
+define i32 @sltiu0(i32 signext %a) {
 entry:
   %cmp = icmp ugt i32 %a, 32766
   %cond = select i1 %cmp, i32 3, i32 5
@@ -532,7 +606,7 @@ entry:
 ; 64-CMP-DAG:  seleqz $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @sltiu1(i32 %a) {
+define i32 @sltiu1(i32 signext %a) {
 entry:
   %cmp = icmp ugt i32 %a, 32767
   %cond = select i1 %cmp, i32 7, i32 5
@@ -567,7 +641,7 @@ entry:
 ; 64-CMP-DAG:  selnez $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @sltiu2(i32 %a) {
+define i32 @sltiu2(i32 signext %a) {
 entry:
   %cmp = icmp ugt i32 %a, -32769
   %cond = select i1 %cmp, i32 3, i32 5
@@ -610,7 +684,7 @@ entry:
 ; 64-CMP-DAG:  seleqz $[[T1:[0-9]+]], $[[I5]], $[[R0]]
 ; 64-CMP-DAG:  or $2, $[[T0]], $[[T1]]
 
-define i32 @sltiu3(i32 %a) {
+define i32 @sltiu3(i32 signext %a) {
 entry:
   %cmp = icmp ugt i32 %a, -32770
   %cond = select i1 %cmp, i32 3, i32 5
@@ -623,7 +697,7 @@ entry:
 ; doesn't generate conditional moves
 ; for constant operands whose difference is |1|
 
-define i32 @slti4(i32 %a) nounwind readnone {
+define i32 @slti4(i32 signext %a) nounwind readnone {
   %1 = icmp slt i32 %a, 7
   %2 = select i1 %1, i32 4, i32 3
   ret i32 %2
@@ -649,7 +723,7 @@ define i32 @slti4(i32 %a) nounwind readnone {
 ; 64-CMP-NOT:  seleqz
 ; 64-CMP-NOT:  selnez
 
-define i32 @slti5(i32 %a) nounwind readnone {
+define i32 @slti5(i32 signext %a) nounwind readnone {
   %1 = icmp slt i32 %a, 7
   %2 = select i1 %1, i32 -3, i32 -4
   ret i32 %2
@@ -675,7 +749,7 @@ define i32 @slti5(i32 %a) nounwind readnone {
 ; 64-CMP-NOT:  seleqz
 ; 64-CMP-NOT:  selnez
 
-define i32 @slti6(i32 %a) nounwind readnone {
+define i32 @slti6(i32 signext %a) nounwind readnone {
   %1 = icmp slt i32 %a, 7
   %2 = select i1 %1, i32 3, i32 4
   ret i32 %2
@@ -683,24 +757,9 @@ define i32 @slti6(i32 %a) nounwind readnone {
 
 ; ALL-LABEL: slti6:
 
-; 32-CMOV-DAG: slti [[R1:\$[0-9]+]], $4, 7
-; 32-CMOV-DAG: xori [[R1]], [[R1]], 1
-; 32-CMOV-DAG: addiu [[R2:\$[0-9]+]], [[R1]], 3
-; 32-CMOV-NOT: movn
-
-; 32-CMP-DAG:  slti [[R1:\$[0-9]+]], $4, 7
-; 32-CMP-DAG:  xori [[R1]], [[R1]], 1
-; 32-CMP-DAG:  addiu [[R2:\$[0-9]+]], [[R1]], 3
-; 32-CMP-NOT:  seleqz
-; 32-CMP-NOT:  selnez
-
-; 64-CMOV-DAG: slti [[R1:\$[0-9]+]], $4, 7
-; 64-CMOV-DAG: xori [[R1]], [[R1]], 1
-; 64-CMOV-DAG: addiu [[R2:\$[0-9]+]], [[R1]], 3
-; 64-CMOV-NOT: movn
-
-; 64-CMP-DAG:  slti [[R1:\$[0-9]+]], $4, 7
-; 64-CMP-DAG:  xori [[R1]], [[R1]], 1
-; 64-CMP-DAG:  addiu [[R2:\$[0-9]+]], [[R1]], 3
-; 64-CMP-NOT:  seleqz
-; 64-CMP-NOT:  selnez
+; ALL-DAG: addiu [[R1:\$[0-9]+]], $zero, 6
+; ALL-DAG: slt [[R1]], [[R1]], $4
+; ALL-DAG: addiu [[R2:\$[0-9]+]], [[R1]], 3
+; ALL-NOT: movn
+; ALL-NOT:  seleqz
+; ALL-NOT:  selnez
diff --git a/test/CodeGen/Mips/const-mult.ll b/test/CodeGen/Mips/const-mult.ll
index 1862021..60b2a88 100644
--- a/test/CodeGen/Mips/const-mult.ll
+++ b/test/CodeGen/Mips/const-mult.ll
@@ -5,7 +5,7 @@
 ; CHECK: sll $[[R0:[0-9]+]], $4, 2
 ; CHECK: addu ${{[0-9]+}}, $[[R0]], $4
 
-define i32 @mul5_32(i32 %a) {
+define i32 @mul5_32(i32 signext %a) {
 entry:
   %mul = mul nsw i32 %a, 5
   ret i32 %mul
@@ -17,7 +17,7 @@ entry:
 ; CHECK-DAG: sll $[[R2:[0-9]+]], $4, 5
 ; CHECK:     subu ${{[0-9]+}}, $[[R2]], $[[R1]]
 
-define i32 @mul27_32(i32 %a) {
+define i32 @mul27_32(i32 signext %a) {
 entry:
   %mul = mul nsw i32 %a, 27
   ret i32 %mul
@@ -29,7 +29,7 @@ entry:
 ; CHECK-DAG: sll $[[R2:[0-9]+]], $4, 31
 ; CHECK:     addu ${{[0-9]+}}, $[[R2]], $[[R1]]
 
-define i32 @muln2147483643_32(i32 %a) {
+define i32 @muln2147483643_32(i32 signext %a) {
 entry:
   %mul = mul nsw i32 %a, -2147483643
   ret i32 %mul
@@ -41,7 +41,7 @@ entry:
 ; CHECK64-DAG: dsll $[[R2:[0-9]+]], $4, 63
 ; CHECK64:     daddu ${{[0-9]+}}, $[[R2]], $[[R1]]
 
-define i64 @muln9223372036854775805_64(i64 %a) {
+define i64 @muln9223372036854775805_64(i64 signext %a) {
 entry:
   %mul = mul nsw i64 %a, -9223372036854775805
   ret i64 %mul
diff --git a/test/CodeGen/Mips/countleading.ll b/test/CodeGen/Mips/countleading.ll
index 6e63cff..b7aad04 100644
--- a/test/CodeGen/Mips/countleading.ll
+++ b/test/CodeGen/Mips/countleading.ll
@@ -11,7 +11,7 @@
 ;   MIPS32-GT-R1 - MIPS64r1 and above (does not include MIPS64's)
 ;   MIPS64-GT-R1 - MIPS64r1 and above
 
-define i32 @ctlz_i32(i32 %X) nounwind readnone {
+define i32 @ctlz_i32(i32 signext %X) nounwind readnone {
 entry:
 ; ALL-LABEL: ctlz_i32:
 
@@ -27,7 +27,7 @@ entry:
 
 declare i32 @llvm.ctlz.i32(i32, i1) nounwind readnone
 
-define i32 @ctlo_i32(i32 %X) nounwind readnone {
+define i32 @ctlo_i32(i32 signext %X) nounwind readnone {
 entry:
 ; ALL-LABEL: ctlo_i32:
 
diff --git a/test/CodeGen/Mips/ctlz-v.ll b/test/CodeGen/Mips/ctlz-v.ll
index 270f404..3d580e5 100644
--- a/test/CodeGen/Mips/ctlz-v.ll
+++ b/test/CodeGen/Mips/ctlz-v.ll
@@ -6,12 +6,12 @@ declare <2 x i32> @llvm.ctlz.v2i32(<2 x i32>, i1)
 define <2 x i32> @ctlzv2i32(<2 x i32> %x) {
 entry:
 ; MIPS32: clz     $2, $4
-; MIPS32: jr      $ra
 ; MIPS32: clz     $3, $5
 
-; MIPS64: clz     $2, $4
-; MIPS64: jr      $ra
-; MIPS64: clz     $3, $5
+; MIPS64-DAG: sll $[[A0:[0-9]+]], $4, 0
+; MIPS64-DAG: clz $2, $[[A0]]
+; MIPS64-DAG: sll $[[A1:[0-9]+]], $5, 0
+; MIPS64-DAG: clz $3, $[[A1]]
 
   %ret = call <2 x i32> @llvm.ctlz.v2i32(<2 x i32> %x, i1 true)
   ret <2 x i32> %ret
diff --git a/test/CodeGen/Mips/cttz-v.ll b/test/CodeGen/Mips/cttz-v.ll
index 9470441..85f69f9 100644
--- a/test/CodeGen/Mips/cttz-v.ll
+++ b/test/CodeGen/Mips/cttz-v.ll
@@ -18,14 +18,16 @@ entry:
 ; MIPS32-DAG: jr      $ra
 ; MIPS32-DAG: subu    $3, $[[R4]], $[[R8]]
 
-; MIPS64-DAG: addiu   $[[R0:[0-9]+]], $4, -1
-; MIPS64-DAG: not     $[[R1:[0-9]+]], $4
+; MIPS64-DAG: sll     $[[A0:[0-9]+]], $4, 0
+; MIPS64-DAG: addiu   $[[R0:[0-9]+]], $[[A0]], -1
+; MIPS64-DAG: not     $[[R1:[0-9]+]], $[[A0]]
 ; MIPS64-DAG: and     $[[R2:[0-9]+]], $[[R1]], $[[R0]]
 ; MIPS64-DAG: clz     $[[R3:[0-9]+]], $[[R2]]
 ; MIPS64-DAG: addiu   $[[R4:[0-9]+]], $zero, 32
 ; MIPS64-DAG: subu    $2, $[[R4]], $[[R3]]
-; MIPS64-DAG: addiu   $[[R5:[0-9]+]], $5, -1
-; MIPS64-DAG: not     $[[R6:[0-9]+]], $5
+; MIPS64-DAG: sll     $[[A1:[0-9]+]], $5, 0
+; MIPS64-DAG: addiu   $[[R5:[0-9]+]], $[[A1]], -1
+; MIPS64-DAG: not     $[[R6:[0-9]+]], $[[A1]]
 ; MIPS64-DAG: and     $[[R7:[0-9]+]], $[[R6]], $[[R5]]
 ; MIPS64-DAG: clz     $[[R8:[0-9]+]], $[[R7]]
 ; MIPS64-DAG: jr      $ra
diff --git a/test/CodeGen/Mips/divrem.ll b/test/CodeGen/Mips/divrem.ll
index 97f8360..a9cfe0f 100644
--- a/test/CodeGen/Mips/divrem.ll
+++ b/test/CodeGen/Mips/divrem.ll
@@ -27,7 +27,7 @@
 @g0 = common global i32 0, align 4
 @g1 = common global i32 0, align 4
 
-define i32 @sdiv1(i32 %a0, i32 %a1) nounwind readnone {
+define i32 @sdiv1(i32 signext %a0, i32 signext %a1) nounwind readnone {
 entry:
 ; ALL-LABEL: sdiv1:
 
@@ -54,7 +54,7 @@ entry:
   ret i32 %div
 }
 
-define i32 @srem1(i32 %a0, i32 %a1) nounwind readnone {
+define i32 @srem1(i32 signext %a0, i32 signext %a1) nounwind readnone {
 entry:
 ; ALL-LABEL: srem1:
 
@@ -81,7 +81,7 @@ entry:
   ret i32 %rem
 }
 
-define i32 @udiv1(i32 %a0, i32 %a1) nounwind readnone {
+define i32 @udiv1(i32 zeroext %a0, i32 zeroext %a1) nounwind readnone {
 entry:
 ; ALL-LABEL: udiv1:
 
@@ -107,7 +107,7 @@ entry:
   ret i32 %div
 }
 
-define i32 @urem1(i32 %a0, i32 %a1) nounwind readnone {
+define i32 @urem1(i32 zeroext %a0, i32 zeroext %a1) nounwind readnone {
 entry:
 ; ALL-LABEL: urem1:
 
@@ -134,7 +134,7 @@ entry:
   ret i32 %rem
 }
 
-define i32 @sdivrem1(i32 %a0, i32 %a1, i32* nocapture %r) nounwind {
+define i32 @sdivrem1(i32 signext %a0, i32 signext %a1, i32* nocapture %r) nounwind {
 entry:
 ; ALL-LABEL: sdivrem1:
 
@@ -175,7 +175,7 @@ entry:
   ret i32 %div
 }
 
-define i32 @udivrem1(i32 %a0, i32 %a1, i32* nocapture %r) nounwind {
+define i32 @udivrem1(i32 zeroext %a0, i32 zeroext %a1, i32* nocapture %r) nounwind {
 entry:
 ; ALL-LABEL: udivrem1:
 
diff --git a/test/CodeGen/Mips/ehframe-indirect.ll b/test/CodeGen/Mips/ehframe-indirect.ll
index e78497a..b4efb40 100644
--- a/test/CodeGen/Mips/ehframe-indirect.ll
+++ b/test/CodeGen/Mips/ehframe-indirect.ll
@@ -1,5 +1,7 @@
-; RUN: llc -mtriple=mipsel-linux-gnu < %s | FileCheck %s
-; RUN: llc -mtriple=mipsel-linux-android < %s | FileCheck %s
+; RUN: llc -mtriple=mipsel-linux-gnu < %s | FileCheck  -check-prefix=CHECK32 %s
+; RUN: llc -mtriple=mipsel-linux-android < %s | FileCheck -check-prefix=CHECK32 %s
+; RUN: llc -mtriple=mips64el-linux-gnu < %s | FileCheck  -check-prefix=CHECK64 %s
+; RUN: llc -mtriple=mips64el-linux-android < %s | FileCheck -check-prefix=CHECK64 %s
 
 define i32 @main() {
 ; CHECK: .cfi_startproc
@@ -27,8 +29,11 @@ declare void @foo()
 ; CHECK: .hidden DW.ref.__gxx_personality_v0
 ; CHECK: .weak DW.ref.__gxx_personality_v0
 ; CHECK: .section .data.DW.ref.__gxx_personality_v0,"aGw",@progbits,DW.ref.__gxx_personality_v0,comdat
-; CHECK: .align 2
+; CHECK32: .align 2
+; CHECK64: .align 3
 ; CHECK: .type DW.ref.__gxx_personality_v0,@object
-; CHECK: .size DW.ref.__gxx_personality_v0, 4
+; CHECK32: .size DW.ref.__gxx_personality_v0, 4
+; CHECK64: .size DW.ref.__gxx_personality_v0, 8
 ; CHECK: DW.ref.__gxx_personality_v0:
-; CHECK: .4byte __gxx_personality_v0
+; CHECK32: .4byte __gxx_personality_v0
+; CHECK64: .8byte __gxx_personality_v0
diff --git a/test/CodeGen/Mips/fastcc.ll b/test/CodeGen/Mips/fastcc.ll
index 8ee7af8..6b022c5 100644
--- a/test/CodeGen/Mips/fastcc.ll
+++ b/test/CodeGen/Mips/fastcc.ll
@@ -1,6 +1,8 @@
 ; RUN: llc  < %s -march=mipsel | FileCheck %s 
 ; RUN: llc  < %s -mtriple=mipsel-none-nacl-gnu \
 ; RUN:  | FileCheck %s -check-prefix=CHECK-NACL
+; RUN: llc  < %s -march=mipsel -mcpu=mips32 -mattr=+nooddspreg | FileCheck %s -check-prefix=NOODDSPREG
+; RUN: llc  < %s -march=mipsel -mcpu=mips32r2 -mattr=+fp64,+nooddspreg | FileCheck %s -check-prefix=FP64-NOODDSPREG
 
 
 @gi0 = external global i32
@@ -80,6 +82,9 @@
 @g15 = external global i32
 @g16 = external global i32
 
+@fa = common global [11 x float] zeroinitializer, align 4
+@da = common global [11 x double] zeroinitializer, align 8
+
 define void @caller0() nounwind {
 entry:
 ; CHECK: caller0
@@ -264,3 +269,164 @@ entry:
   ret void
 }
 
+define void @caller2() {
+entry:
+
+; NOODDSPREG-LABEL:  caller2:
+
+; Check that first 10 arguments are passed in even float registers
+; f0, f2, ... , f18. Check that 11th argument is passed on stack.
+
+; NOODDSPREG-DAG:    lw      $[[R0:[0-9]+]], %got(fa)(${{[0-9]+|gp}})
+; NOODDSPREG-DAG:    lwc1    $f0, 0($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f2, 4($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f4, 8($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f6, 12($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f8, 16($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f10, 20($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f12, 24($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f14, 28($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f16, 32($[[R0]])
+; NOODDSPREG-DAG:    lwc1    $f18, 36($[[R0]])
+
+; NOODDSPREG-DAG:    lwc1    $[[F0:f[0-9]*[02468]]], 40($[[R0]])
+; NOODDSPREG-DAG:    swc1    $[[F0]], 0($sp)
+
+  %0 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 0), align 4
+  %1 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 1), align 4
+  %2 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 2), align 4
+  %3 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 3), align 4
+  %4 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 4), align 4
+  %5 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 5), align 4
+  %6 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 6), align 4
+  %7 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 7), align 4
+  %8 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 8), align 4
+  %9 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 9), align 4
+  %10 = load float* getelementptr ([11 x float]* @fa, i32 0, i32 10), align 4
+  tail call fastcc void @callee2(float %0, float %1, float %2, float %3,
+                                 float %4, float %5, float %6, float %7,
+                                 float %8, float %9, float %10)
+  ret void
+}
+
+define fastcc void @callee2(float %a0, float %a1, float %a2, float %a3,
+                            float %a4, float %a5, float %a6, float %a7,
+                            float %a8, float %a9, float %a10) {
+entry:
+
+; NOODDSPREG-LABEL:  callee2:
+
+; NOODDSPREG:        addiu   $sp, $sp, -[[OFFSET:[0-9]+]]
+
+; Check that first 10 arguments are received in even float registers
+; f0, f2, ... , f18. Check that 11th argument is received on stack.
+
+; NOODDSPREG-DAG:    lw      $[[R0:[0-9]+]], %got(fa)(${{[0-9]+|gp}})
+; NOODDSPREG-DAG:    swc1    $f0, 0($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f2, 4($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f4, 8($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f6, 12($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f8, 16($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f10, 20($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f12, 24($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f14, 28($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f16, 32($[[R0]])
+; NOODDSPREG-DAG:    swc1    $f18, 36($[[R0]])
+
+; NOODDSPREG-DAG:    lwc1    $[[F0:f[0-9]*[02468]]], [[OFFSET]]($sp)
+; NOODDSPREG-DAG:    swc1    $[[F0]], 40($[[R0]])
+
+  store float %a0, float* getelementptr ([11 x float]* @fa, i32 0, i32 0), align 4
+  store float %a1, float* getelementptr ([11 x float]* @fa, i32 0, i32 1), align 4
+  store float %a2, float* getelementptr ([11 x float]* @fa, i32 0, i32 2), align 4
+  store float %a3, float* getelementptr ([11 x float]* @fa, i32 0, i32 3), align 4
+  store float %a4, float* getelementptr ([11 x float]* @fa, i32 0, i32 4), align 4
+  store float %a5, float* getelementptr ([11 x float]* @fa, i32 0, i32 5), align 4
+  store float %a6, float* getelementptr ([11 x float]* @fa, i32 0, i32 6), align 4
+  store float %a7, float* getelementptr ([11 x float]* @fa, i32 0, i32 7), align 4
+  store float %a8, float* getelementptr ([11 x float]* @fa, i32 0, i32 8), align 4
+  store float %a9, float* getelementptr ([11 x float]* @fa, i32 0, i32 9), align 4
+  store float %a10, float* getelementptr ([11 x float]* @fa, i32 0, i32 10), align 4
+  ret void
+}
+
+define void @caller3() {
+entry:
+
+; FP64-NOODDSPREG-LABEL:  caller3:
+
+; Check that first 10 arguments are passed in even float registers
+; f0, f2, ... , f18. Check that 11th argument is passed on stack.
+
+; FP64-NOODDSPREG-DAG:    lw      $[[R0:[0-9]+]], %got(da)(${{[0-9]+|gp}})
+; FP64-NOODDSPREG-DAG:    ldc1    $f0, 0($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f2, 8($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f4, 16($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f6, 24($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f8, 32($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f10, 40($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f12, 48($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f14, 56($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f16, 64($[[R0]])
+; FP64-NOODDSPREG-DAG:    ldc1    $f18, 72($[[R0]])
+
+; FP64-NOODDSPREG-DAG:    ldc1    $[[F0:f[0-9]*[02468]]], 80($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $[[F0]], 0($sp)
+
+  %0 = load double* getelementptr ([11 x double]* @da, i32 0, i32 0), align 8
+  %1 = load double* getelementptr ([11 x double]* @da, i32 0, i32 1), align 8
+  %2 = load double* getelementptr ([11 x double]* @da, i32 0, i32 2), align 8
+  %3 = load double* getelementptr ([11 x double]* @da, i32 0, i32 3), align 8
+  %4 = load double* getelementptr ([11 x double]* @da, i32 0, i32 4), align 8
+  %5 = load double* getelementptr ([11 x double]* @da, i32 0, i32 5), align 8
+  %6 = load double* getelementptr ([11 x double]* @da, i32 0, i32 6), align 8
+  %7 = load double* getelementptr ([11 x double]* @da, i32 0, i32 7), align 8
+  %8 = load double* getelementptr ([11 x double]* @da, i32 0, i32 8), align 8
+  %9 = load double* getelementptr ([11 x double]* @da, i32 0, i32 9), align 8
+  %10 = load double* getelementptr ([11 x double]* @da, i32 0, i32 10), align 8
+  tail call fastcc void @callee3(double %0, double %1, double %2, double %3,
+                                 double %4, double %5, double %6, double %7,
+                                 double %8, double %9, double %10)
+  ret void
+}
+
+define fastcc void @callee3(double %a0, double %a1, double %a2, double %a3,
+                            double %a4, double %a5, double %a6, double %a7,
+                            double %a8, double %a9, double %a10) {
+entry:
+
+; FP64-NOODDSPREG-LABEL:  callee3:
+
+; FP64-NOODDSPREG:        addiu   $sp, $sp, -[[OFFSET:[0-9]+]]
+
+; Check that first 10 arguments are received in even float registers
+; f0, f2, ... , f18. Check that 11th argument is received on stack.
+
+; FP64-NOODDSPREG-DAG:    lw      $[[R0:[0-9]+]], %got(da)(${{[0-9]+|gp}})
+; FP64-NOODDSPREG-DAG:    sdc1    $f0, 0($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f2, 8($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f4, 16($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f6, 24($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f8, 32($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f10, 40($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f12, 48($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f14, 56($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f16, 64($[[R0]])
+; FP64-NOODDSPREG-DAG:    sdc1    $f18, 72($[[R0]])
+
+; FP64-NOODDSPREG-DAG:    ldc1    $[[F0:f[0-9]*[02468]]], [[OFFSET]]($sp)
+; FP64-NOODDSPREG-DAG:    sdc1    $[[F0]], 80($[[R0]])
+
+  store double %a0, double* getelementptr ([11 x double]* @da, i32 0, i32 0), align 8
+  store double %a1, double* getelementptr ([11 x double]* @da, i32 0, i32 1), align 8
+  store double %a2, double* getelementptr ([11 x double]* @da, i32 0, i32 2), align 8
+  store double %a3, double* getelementptr ([11 x double]* @da, i32 0, i32 3), align 8
+  store double %a4, double* getelementptr ([11 x double]* @da, i32 0, i32 4), align 8
+  store double %a5, double* getelementptr ([11 x double]* @da, i32 0, i32 5), align 8
+  store double %a6, double* getelementptr ([11 x double]* @da, i32 0, i32 6), align 8
+  store double %a7, double* getelementptr ([11 x double]* @da, i32 0, i32 7), align 8
+  store double %a8, double* getelementptr ([11 x double]* @da, i32 0, i32 8), align 8
+  store double %a9, double* getelementptr ([11 x double]* @da, i32 0, i32 9), align 8
+  store double %a10, double* getelementptr ([11 x double]* @da, i32 0, i32 10), align 8
+  ret void
+}
diff --git a/test/CodeGen/Mips/fp16instrinsmc.ll b/test/CodeGen/Mips/fp16instrinsmc.ll
index 7ced36c..84d3814 100644
--- a/test/CodeGen/Mips/fp16instrinsmc.ll
+++ b/test/CodeGen/Mips/fp16instrinsmc.ll
@@ -385,7 +385,7 @@ entry:
 ; Function Attrs: nounwind
 declare double @exp2(double) #0
 
-attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="true" }
+attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind readnone "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="true" }
 attributes #2 = { nounwind readnone }
 attributes #3 = { nounwind }
diff --git a/test/CodeGen/Mips/fp64a.ll b/test/CodeGen/Mips/fp64a.ll
new file mode 100644
index 0000000..fadce5c
--- /dev/null
+++ b/test/CodeGen/Mips/fp64a.ll
@@ -0,0 +1,161 @@
+; Test that the FP64A ABI performs double precision moves via a spill/reload.
+; The requirement is really that odd-numbered double precision registers do not
+; use mfc1/mtc1 to move the bottom 32-bits (because the hardware will redirect
+; this to the top 32-bits of the even register) but we have to make the decision
+; before register allocation so we do this for all double-precision values.
+
+; We don't test MIPS32r1 since support for 64-bit coprocessors (such as a 64-bit
+; FPU) on a 32-bit architecture was added in MIPS32r2.
+; FIXME: We currently don't test that attempting to use FP64 on MIPS32r1 is an
+;        error either. This is because a large number of CodeGen tests are
+;        incorrectly using this case. We should fix those test cases then add
+;        this check here.
+
+; RUN: llc -march=mips -mcpu=mips32r2 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-NO-FP64A-BE
+; RUN: llc -march=mips -mcpu=mips32r2 -mattr=fp64,nooddspreg < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-FP64A
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-NO-FP64A-LE
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=fp64,nooddspreg < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-FP64A
+
+; RUN: llc -march=mips64 -mcpu=mips64 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=64-NO-FP64A
+; RUN: not llc -march=mips64 -mcpu=mips64 -mattr=fp64,nooddspreg < %s 2>&1 | FileCheck %s -check-prefix=64-FP64A
+; RUN: llc -march=mips64el -mcpu=mips64 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=64-NO-FP64A
+; RUN: not llc -march=mips64el -mcpu=mips64 -mattr=fp64,nooddspreg < %s 2>&1 | FileCheck %s -check-prefix=64-FP64A
+
+; 64-FP64A: LLVM ERROR: -mattr=+nooddspreg requires the O32 ABI.
+
+declare double @dbl();
+
+define double @call1(double %d, ...) {
+  ret double %d
+
+; ALL-LABEL:            call1:
+
+; 32R2-NO-FP64A-LE-NOT:     addiu   $sp, $sp
+; 32R2-NO-FP64A-LE:         mtc1    $4, $f0
+; 32R2-NO-FP64A-LE:         mthc1   $5, $f0
+
+; 32R2-NO-FP64A-BE-NOT:     addiu   $sp, $sp
+; 32R2-NO-FP64A-BE:         mtc1    $5, $f0
+; 32R2-NO-FP64A-BE:         mthc1   $4, $f0
+
+; 32R2-FP64A:               addiu   $sp, $sp, -8
+; 32R2-FP64A:               sw      $4, 0($sp)
+; 32R2-FP64A:               sw      $5, 4($sp)
+; 32R2-FP64A:               ldc1    $f0, 0($sp)
+
+; 64-NO-FP64A:              daddiu  $sp, $sp, -64
+; 64-NO-FP64A:              mov.d   $f0, $f12
+}
+
+define double @call2(i32 %i, double %d) {
+  ret double %d
+
+; ALL-LABEL:        call2:
+
+; 32R2-NO-FP64A-LE:     mtc1    $6, $f0
+; 32R2-NO-FP64A-LE:     mthc1   $7, $f0
+
+; 32R2-NO-FP64A-BE:     mtc1    $7, $f0
+; 32R2-NO-FP64A-BE:     mthc1   $6, $f0
+
+; 32R2-FP64A:           addiu   $sp, $sp, -8
+; 32R2-FP64A:           sw      $6, 0($sp)
+; 32R2-FP64A:           sw      $7, 4($sp)
+; 32R2-FP64A:           ldc1    $f0, 0($sp)
+
+; 64-NO-FP64A-NOT:      daddiu  $sp, $sp
+; 64-NO-FP64A:          mov.d   $f0, $f13
+}
+
+define double @call3(float %f1, float %f2, double %d) {
+  ret double %d
+
+; ALL-LABEL:        call3:
+
+; 32R2-NO-FP64A-LE:     mtc1    $6, $f0
+; 32R2-NO-FP64A-LE:     mthc1   $7, $f0
+
+; 32R2-NO-FP64A-BE:     mtc1    $7, $f0
+; 32R2-NO-FP64A-BE:     mthc1   $6, $f0
+
+; 32R2-FP64A:           addiu   $sp, $sp, -8
+; 32R2-FP64A:           sw      $6, 0($sp)
+; 32R2-FP64A:           sw      $7, 4($sp)
+; 32R2-FP64A:           ldc1    $f0, 0($sp)
+
+; 64-NO-FP64A-NOT:      daddiu  $sp, $sp
+; 64-NO-FP64A:          mov.d   $f0, $f14
+}
+
+define double @call4(float %f, double %d, ...) {
+  ret double %d
+
+; ALL-LABEL:        call4:
+
+; 32R2-NO-FP64A-LE:     mtc1    $6, $f0
+; 32R2-NO-FP64A-LE:     mthc1   $7, $f0
+
+; 32R2-NO-FP64A-BE:     mtc1    $7, $f0
+; 32R2-NO-FP64A-BE:     mthc1   $6, $f0
+
+; 32R2-FP64A:           addiu   $sp, $sp, -8
+; 32R2-FP64A:           sw      $6, 0($sp)
+; 32R2-FP64A:           sw      $7, 4($sp)
+; 32R2-FP64A:           ldc1    $f0, 0($sp)
+
+; 64-NO-FP64A:          daddiu  $sp, $sp, -48
+; 64-NO-FP64A:          mov.d   $f0, $f13
+}
+
+define double @call5(double %a, double %b, ...) {
+  %1 = fsub double %a, %b
+  ret double %1
+
+; ALL-LABEL:            call5:
+
+; 32R2-NO-FP64A-LE-DAG:     mtc1    $4, $[[T0:f[0-9]+]]
+; 32R2-NO-FP64A-LE-DAG:     mthc1   $5, $[[T0:f[0-9]+]]
+; 32R2-NO-FP64A-LE-DAG:     mtc1    $6, $[[T1:f[0-9]+]]
+; 32R2-NO-FP64A-LE-DAG:     mthc1   $7, $[[T1:f[0-9]+]]
+; 32R2-NO-FP64A-LE:         sub.d   $f0, $[[T0]], $[[T1]]
+
+; 32R2-NO-FP64A-BE-DAG:     mtc1    $5, $[[T0:f[0-9]+]]
+; 32R2-NO-FP64A-BE-DAG:     mthc1   $4, $[[T0:f[0-9]+]]
+; 32R2-NO-FP64A-BE-DAG:     mtc1    $7, $[[T1:f[0-9]+]]
+; 32R2-NO-FP64A-BE-DAG:     mthc1   $6, $[[T1:f[0-9]+]]
+; 32R2-NO-FP64A-BE:         sub.d   $f0, $[[T0]], $[[T1]]
+
+; 32R2-FP64A:               addiu   $sp, $sp, -8
+; 32R2-FP64A:               sw      $6, 0($sp)
+; 32R2-FP64A:               sw      $7, 4($sp)
+; 32R2-FP64A:               ldc1    $[[T1:f[0-9]+]], 0($sp)
+; 32R2-FP64A:               sw      $4, 0($sp)
+; 32R2-FP64A:               sw      $5, 4($sp)
+; 32R2-FP64A:               ldc1    $[[T0:f[0-9]+]], 0($sp)
+; 32R2-FP64A:               sub.d   $f0, $[[T0]], $[[T1]]
+
+; 64-NO-FP64A:              sub.d   $f0, $f12, $f13
+}
+
+define double @move_from(double %d) {
+  %1 = call double @dbl()
+  %2 = call double @call2(i32 0, double %1)
+  ret double %2
+
+; ALL-LABEL:        move_from:
+
+; 32R2-NO-FP64A-LE-DAG: mfc1    $6, $f0
+; 32R2-NO-FP64A-LE-DAG: mfhc1   $7, $f0
+
+; 32R2-NO-FP64A-BE-DAG: mfc1    $7, $f0
+; 32R2-NO-FP64A-BE-DAG: mfhc1   $6, $f0
+
+; 32R2-FP64A:           addiu   $sp, $sp, -32
+; 32R2-FP64A:           sdc1    $f0, 16($sp)
+; 32R2-FP64A:           lw      $6, 16($sp)
+; FIXME: This store is redundant
+; 32R2-FP64A:           sdc1    $f0, 16($sp)
+; 32R2-FP64A:           lw      $7, 20($sp)
+
+; 64-NO-FP64A:          mov.d   $f13, $f0
+}
diff --git a/test/CodeGen/Mips/fpxx.ll b/test/CodeGen/Mips/fpxx.ll
new file mode 100644
index 0000000..7e2ed22
--- /dev/null
+++ b/test/CodeGen/Mips/fpxx.ll
@@ -0,0 +1,221 @@
+; RUN: llc -march=mipsel -mcpu=mips32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=32-NOFPXX
+; RUN: llc -march=mipsel -mcpu=mips32 -mattr=fpxx < %s | FileCheck %s -check-prefix=ALL -check-prefix=32-FPXX
+
+; RUN: llc -march=mipsel -mcpu=mips32r2 < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-NOFPXX
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=fpxx < %s | FileCheck %s -check-prefix=ALL -check-prefix=32R2-FPXX
+
+; RUN: llc -march=mips64 -mcpu=mips4 < %s | FileCheck %s -check-prefix=ALL -check-prefix=4-NOFPXX
+; RUN: not llc -march=mips64 -mcpu=mips4 -mattr=fpxx < %s 2>&1 | FileCheck %s -check-prefix=4-FPXX
+
+; RUN: llc -march=mips64 -mcpu=mips64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=64-NOFPXX
+; RUN: not llc -march=mips64 -mcpu=mips64 -mattr=fpxx < %s 2>&1 | FileCheck %s -check-prefix=64-FPXX
+
+; RUN-TODO: llc -march=mips64 -mcpu=mips4 -mattr=-n64,+o32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=4-O32-NOFPXX
+; RUN-TODO: llc -march=mips64 -mcpu=mips4 -mattr=-n64,+o32 -mattr=fpxx < %s | FileCheck %s -check-prefix=ALL -check-prefix=4-O32-FPXX
+
+; RUN-TODO: llc -march=mips64 -mcpu=mips64 -mattr=-n64,+o32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=64-O32-NOFPXX
+; RUN-TODO: llc -march=mips64 -mcpu=mips64 -mattr=-n64,+o32 -mattr=fpxx < %s | FileCheck %s -check-prefix=ALL -check-prefix=64-O32-FPXX
+
+declare double @dbl();
+
+; 4-FPXX:  LLVM ERROR: FPXX is not permitted for the N32/N64 ABI's.
+; 64-FPXX: LLVM ERROR: FPXX is not permitted for the N32/N64 ABI's.
+
+define double @test1(double %d, ...) {
+  ret double %d
+
+; ALL-LABEL: test1:
+
+; 32-NOFPXX:     mtc1    $4, $f0
+; 32-NOFPXX:     mtc1    $5, $f1
+
+; 32-FPXX:       addiu   $sp, $sp, -8
+; 32-FPXX:       sw      $4, 0($sp)
+; 32-FPXX:       sw      $5, 4($sp)
+; 32-FPXX:       ldc1    $f0, 0($sp)
+
+; 32R2-NOFPXX:   mtc1    $4, $f0
+; 32R2-NOFPXX:   mthc1   $5, $f0
+
+; 32R2-FPXX:     mtc1    $4, $f0
+; 32R2-FPXX:     mthc1   $5, $f0
+
+; floats/doubles are not passed in integer registers for n64, so dmtc1 is not used.
+; 4-NOFPXX:      mov.d   $f0, $f12
+
+; 64-NOFPXX:     mov.d   $f0, $f12
+}
+
+define double @test2(i32 %i, double %d) {
+  ret double %d
+
+; ALL-LABEL: test2:
+
+; 32-NOFPXX:     mtc1    $6, $f0
+; 32-NOFPXX:     mtc1    $7, $f1
+
+; 32-FPXX:       addiu   $sp, $sp, -8
+; 32-FPXX:       sw      $6, 0($sp)
+; 32-FPXX:       sw      $7, 4($sp)
+; 32-FPXX:       ldc1    $f0, 0($sp)
+
+; 32R2-NOFPXX:   mtc1    $6, $f0
+; 32R2-NOFPXX:   mthc1   $7, $f0
+
+; 32R2-FPXX:     mtc1    $6, $f0
+; 32R2-FPXX:     mthc1   $7, $f0
+
+; 4-NOFPXX:      mov.d   $f0, $f13
+
+; 64-NOFPXX:     mov.d   $f0, $f13
+}
+
+define double @test3(float %f1, float %f2, double %d) {
+  ret double %d
+
+; ALL-LABEL: test3:
+
+; 32-NOFPXX:     mtc1    $6, $f0
+; 32-NOFPXX:     mtc1    $7, $f1
+
+; 32-FPXX:       addiu   $sp, $sp, -8
+; 32-FPXX:       sw      $6, 0($sp)
+; 32-FPXX:       sw      $7, 4($sp)
+; 32-FPXX:       ldc1    $f0, 0($sp)
+
+; 32R2-NOFPXX:   mtc1    $6, $f0
+; 32R2-NOFPXX:   mthc1   $7, $f0
+
+; 32R2-FPXX:     mtc1    $6, $f0
+; 32R2-FPXX:     mthc1   $7, $f0
+
+; 4-NOFPXX:      mov.d   $f0, $f14
+
+; 64-NOFPXX:     mov.d   $f0, $f14
+}
+
+define double @test4(float %f, double %d, ...) {
+  ret double %d
+
+; ALL-LABEL: test4:
+
+; 32-NOFPXX:     mtc1    $6, $f0
+; 32-NOFPXX:     mtc1    $7, $f1
+
+; 32-FPXX:       addiu   $sp, $sp, -8
+; 32-FPXX:       sw      $6, 0($sp)
+; 32-FPXX:       sw      $7, 4($sp)
+; 32-FPXX:       ldc1    $f0, 0($sp)
+
+; 32R2-NOFPXX:   mtc1    $6, $f0
+; 32R2-NOFPXX:   mthc1   $7, $f0
+
+; 32R2-FPXX:     mtc1    $6, $f0
+; 32R2-FPXX:     mthc1   $7, $f0
+
+; 4-NOFPXX:      mov.d   $f0, $f13
+
+; 64-NOFPXX:     mov.d   $f0, $f13
+}
+
+define double @test5() {
+  ret double 0.000000e+00
+
+; ALL-LABEL: test5:
+
+; 32-NOFPXX:     mtc1    $zero, $f0
+; 32-NOFPXX:     mtc1    $zero, $f1
+
+; 32-FPXX:       addiu   $sp, $sp, -8
+; 32-FPXX:       sw      $zero, 0($sp)
+; 32-FPXX:       sw      $zero, 4($sp)
+; 32-FPXX:       ldc1    $f0, 0($sp)
+
+; 32R2-NOFPXX:   mtc1    $zero, $f0
+; 32R2-NOFPXX:   mthc1   $zero, $f0
+
+; 32R2-FPXX:     mtc1    $zero, $f0
+; 32R2-FPXX:     mthc1   $zero, $f0
+
+; 4-NOFPXX:      dmtc1 $zero, $f0
+
+; 64-NOFPXX:     dmtc1 $zero, $f0
+}
+
+define double @test6(double %a, double %b, ...) {
+  %1 = fsub double %a, %b
+  ret double %1
+
+; ALL-LABEL:     test6:
+
+; 32-NOFPXX-DAG:     mtc1    $4, $[[T0:f[0-9]+]]
+; 32-NOFPXX-DAG:     mtc1    $5, ${{f[0-9]*[13579]}}
+; 32-NOFPXX-DAG:     mtc1    $6, $[[T1:f[0-9]+]]
+; 32-NOFPXX-DAG:     mtc1    $7, ${{f[0-9]*[13579]}}
+; 32-NOFPXX:         sub.d   $f0, $[[T0]], $[[T1]]
+
+; 32-FPXX:           addiu   $sp, $sp, -8
+; 32-FPXX:           sw      $6, 0($sp)
+; 32-FPXX:           sw      $7, 4($sp)
+; 32-FPXX:           ldc1    $[[T1:f[0-9]+]], 0($sp)
+; 32-FPXX:           sw      $4, 0($sp)
+; 32-FPXX:           sw      $5, 4($sp)
+; 32-FPXX:           ldc1    $[[T0:f[0-9]+]], 0($sp)
+; 32-FPXX:           sub.d   $f0, $[[T0]], $[[T1]]
+
+; 32R2-NOFPXX-DAG:   mtc1    $4, $[[T0:f[0-9]+]]
+; 32R2-NOFPXX-DAG:   mthc1   $5, $[[T0]]
+; 32R2-NOFPXX-DAG:   mtc1    $6, $[[T1:f[0-9]+]]
+; 32R2-NOFPXX-DAG:   mthc1   $7, $[[T1]]
+; 32R2-NOFPXX:       sub.d   $f0, $[[T0]], $[[T1]]
+
+; 32R2-FPXX-DAG:     mtc1    $4, $[[T0:f[0-9]+]]
+; 32R2-FPXX-DAG:     mthc1   $5, $[[T0]]
+; 32R2-FPXX-DAG:     mtc1    $6, $[[T1:f[0-9]+]]
+; 32R2-FPXX-DAG:     mthc1   $7, $[[T1]]
+; 32R2-FPXX:         sub.d   $f0, $[[T0]], $[[T1]]
+
+; floats/doubles are not passed in integer registers for n64, so dmtc1 is not used.
+; 4-NOFPXX:          sub.d   $f0, $f12, $f13
+
+; floats/doubles are not passed in integer registers for n64, so dmtc1 is not used.
+; 64-NOFPXX:         sub.d   $f0, $f12, $f13
+}
+
+define double @move_from1(double %d) {
+  %1 = call double @dbl()
+  %2 = call double @test2(i32 0, double %1)
+  ret double %2
+
+; ALL-LABEL:   move_from1:
+
+; 32-NOFPXX-DAG:   mfc1    $6, $f0
+; 32-NOFPXX-DAG:   mfc1    $7, $f1
+
+; 32-FPXX:         addiu   $sp, $sp, -32
+; 32-FPXX:         sdc1    $f0, 16($sp)
+; 32-FPXX:         lw      $6, 16($sp)
+; FIXME: This store is redundant
+; 32-FPXX:         sdc1    $f0, 16($sp)
+; 32-FPXX:         lw      $7, 20($sp)
+
+; 32R2-NOFPXX-DAG: mfc1    $6, $f0
+; 32R2-NOFPXX-DAG: mfhc1   $7, $f0
+
+; 32R2-FPXX-DAG:   mfc1    $6, $f0
+; 32R2-FPXX-DAG:   mfhc1   $7, $f0
+
+; floats/doubles are not passed in integer registers for n64, so dmfc1 is not used.
+; We can't use inline assembly to force a copy either because trying to force
+; a copy to a GPR this way fails with ; "couldn't allocate input reg for
+; constraint 'r'". It therefore seems impossible to test the generation of dmfc1
+; in a simple test.
+; 4-NOFPXX:        mov.d   $f13, $f0
+
+; floats/doubles are not passed in integer registers for n64, so dmfc1 is not used.
+; We can't use inline assembly to force a copy either because trying to force
+; a copy to a GPR this way fails with ; "couldn't allocate input reg for
+; constraint 'r'". It therefore seems impossible to test the generation of dmfc1
+; in a simple test.
+; 64-NOFPXX:       mov.d   $f13, $f0
+}
diff --git a/test/CodeGen/Mips/gpreg-lazy-binding.ll b/test/CodeGen/Mips/gpreg-lazy-binding.ll
index 88e596b..3a636d8 100644
--- a/test/CodeGen/Mips/gpreg-lazy-binding.ll
+++ b/test/CodeGen/Mips/gpreg-lazy-binding.ll
@@ -25,3 +25,11 @@ entry:
   ret void
 }
 
+define void @no_lazy(void (i32)* %pf) {
+
+; CHECK-LABEL:  no_lazy
+; CHECK-NOT:    gp_disp
+
+  tail call void %pf(i32 1)
+  ret void
+}
diff --git a/test/CodeGen/Mips/hfptrcall.ll b/test/CodeGen/Mips/hfptrcall.ll
index 9df8d90..683952d 100644
--- a/test/CodeGen/Mips/hfptrcall.ll
+++ b/test/CodeGen/Mips/hfptrcall.ll
@@ -118,8 +118,8 @@ entry:
 
 declare i32 @printf(i8*, ...) #1
 
-attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="true" }
-attributes #1 = { "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="true" }
+attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="false" }
+attributes #1 = { "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="false" }
 
 
 
diff --git a/test/CodeGen/Mips/init-array.ll b/test/CodeGen/Mips/init-array.ll
index f96ce26..1ca182d 100644
--- a/test/CodeGen/Mips/init-array.ll
+++ b/test/CodeGen/Mips/init-array.ll
@@ -1,4 +1,4 @@
-; RUN: llc -mtriple mipsel-unknown-linux -use-init-array < %s | FileCheck  %s
+; RUN: llc -mtriple mipsel-unknown-linux < %s | FileCheck  %s
 
 target triple = "mipsel-unknown-linux"
 
diff --git a/test/CodeGen/Mips/inlineasm-operand-code.ll b/test/CodeGen/Mips/inlineasm-operand-code.ll
index 6512851..3d9dec7 100644
--- a/test/CodeGen/Mips/inlineasm-operand-code.ll
+++ b/test/CodeGen/Mips/inlineasm-operand-code.ll
@@ -65,6 +65,33 @@ entry:
 ;CHECK_LITTLE_32:    addiu ${{[0-9]+}},${{[0-9]+}},$0
 ;CHECK_LITTLE_32:    #NO_APP
   tail call i32 asm sideeffect "addiu $0,$1,${2:z}", "=r,r,I"(i32 7, i32 0) nounwind
+
+; z with non-zero and the "r"(register) and "J"(integer zero) constraints
+;CHECK_LITTLE_32:    #APP
+;CHECK_LITTLE_32:    mtc0 ${{[1-9][0-9]?}}, ${{[0-9]+}}
+;CHECK_LITTLE_32:    #NO_APP
+  call void asm sideeffect "mtc0 ${0:z}, $$12", "Jr"(i32 7) nounwind
+
+; z with zero and the "r"(register) and "J"(integer zero) constraints
+;CHECK_LITTLE_32:    #APP
+;CHECK_LITTLE_32:    mtc0 $0, ${{[0-9]+}}
+;CHECK_LITTLE_32:    #NO_APP
+  call void asm sideeffect "mtc0 ${0:z}, $$12", "Jr"(i32 0) nounwind
+
+; z with non-zero and just the "r"(register) constraint
+;CHECK_LITTLE_32:    #APP
+;CHECK_LITTLE_32:    mtc0 ${{[1-9][0-9]?}}, ${{[0-9]+}}
+;CHECK_LITTLE_32:    #NO_APP
+  call void asm sideeffect "mtc0 ${0:z}, $$12", "r"(i32 7) nounwind
+
+; z with zero and just the "r"(register) constraint
+; FIXME: Check for $0, instead of other registers.
+;        We should be using $0 directly in this case, not real registers.
+;        When the materialization of 0 gets fixed, this test will fail.
+;CHECK_LITTLE_32:    #APP
+;CHECK_LITTLE_32:    mtc0 ${{[1-9][0-9]?}}, ${{[0-9]+}}
+;CHECK_LITTLE_32:    #NO_APP
+  call void asm sideeffect "mtc0 ${0:z}, $$12", "r"(i32 0) nounwind
   ret i32 0
 }
 
diff --git a/test/CodeGen/Mips/llvm-ir/mul.ll b/test/CodeGen/Mips/llvm-ir/mul.ll
new file mode 100644
index 0000000..1674124
--- /dev/null
+++ b/test/CodeGen/Mips/llvm-ir/mul.ll
@@ -0,0 +1,181 @@
+; RUN: llc < %s -march=mips -mcpu=mips2 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=M2
+; RUN: llc < %s -march=mips -mcpu=mips32 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=32R1-R2 -check-prefix=32R1
+; RUN: llc < %s -march=mips -mcpu=mips32r2 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=32R1-R2 -check-prefix=32R2
+; RUN: llc < %s -march=mips -mcpu=mips32r6 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=32R6
+; RUN: llc < %s -march=mips64 -mcpu=mips4 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=M4
+; RUN: llc < %s -march=mips64 -mcpu=mips64 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=64R1-R2
+; RUN: llc < %s -march=mips64 -mcpu=mips64r2 | FileCheck %s \
+; RUN:    -check-prefix=ALL -check-prefix=64R1-R2
+; RUN: llc < %s -march=mips64 -mcpu=mips64r6 | FileCheck %s \
+; RUN:     -check-prefix=ALL -check-prefix=64R6
+
+define signext i1 @mul_i1(i1 signext %a, i1 signext %b) {
+entry:
+; ALL-LABEL: mul_i1:
+
+  ; M2:         mult    $4, $5
+  ; M2:         mflo    $[[T0:[0-9]+]]
+  ; M2:         sll     $[[T0]], $[[T0]], 31
+  ; M2:         sra     $2, $[[T0]], 31
+
+  ; 32R1-R2:    mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R1-R2:    sll     $[[T0]], $[[T0]], 31
+  ; 32R1-R2:    sra     $2, $[[T0]], 31
+
+  ; 32R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R6:       sll     $[[T0]], $[[T0]], 31
+  ; 32R6:       sra     $2, $[[T0]], 31
+
+  ; M4:         mult    $4, $5
+  ; M4:         mflo    $[[T0:[0-9]+]]
+  ; M4:         sll     $[[T0]], $[[T0]], 31
+  ; M4:         sra     $2, $[[T0]], 31
+
+  ; 64R1-R2:    mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R1-R2:    sll     $[[T0]], $[[T0]], 31
+  ; 64R1-R2:    sra     $2, $[[T0]], 31
+
+  ; 64R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R6:       sll     $[[T0]], $[[T0]], 31
+  ; 64R6:       sra     $2, $[[T0]], 31
+
+  %r = mul i1 %a, %b
+  ret i1 %r
+}
+
+define signext i8 @mul_i8(i8 signext %a, i8 signext %b) {
+entry:
+; ALL-LABEL: mul_i8:
+
+  ; M2:         mult    $4, $5
+  ; M2:         mflo    $[[T0:[0-9]+]]
+  ; M2:         sll     $[[T0]], $[[T0]], 24
+  ; M2:         sra     $2, $[[T0]], 24
+
+  ; 32R1:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R1:       sll     $[[T0]], $[[T0]], 24
+  ; 32R1:       sra     $2, $[[T0]], 24
+
+  ; 32R2:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R2:       seb     $2, $[[T0]]
+
+  ; 32R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R6:       seb     $2, $[[T0]]
+
+  ; M4:         mult    $4, $5
+  ; M4:         mflo    $[[T0:[0-9]+]]
+  ; M4:         sll     $[[T0]], $[[T0]], 24
+  ; M4:         sra     $2, $[[T0]], 24
+
+  ; 64R1:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R1:       sll     $[[T0]], $[[T0]], 24
+  ; 64R1:       sra     $2, $[[T0]], 24
+
+  ; 64R2:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R2:       seb     $2, $[[T0]]
+
+  ; 64R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R6:       seb     $2, $[[T0]]
+  %r = mul i8 %a, %b
+  ret i8 %r
+}
+
+define signext i16 @mul_i16(i16 signext %a, i16 signext %b) {
+entry:
+; ALL-LABEL: mul_i16:
+
+  ; M2:         mult    $4, $5
+  ; M2:         mflo    $[[T0:[0-9]+]]
+  ; M2:         sll     $[[T0]], $[[T0]], 16
+  ; M2:         sra     $2, $[[T0]], 16
+
+  ; 32R1:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R1:       sll     $[[T0]], $[[T0]], 16
+  ; 32R1:       sra     $2, $[[T0]], 16
+
+  ; 32R2:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R2:       seh     $2, $[[T0]]
+
+  ; 32R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 32R6:       seh     $2, $[[T0]]
+
+  ; M4:         mult    $4, $5
+  ; M4:         mflo    $[[T0:[0-9]+]]
+  ; M4:         sll     $[[T0]], $[[T0]], 16
+  ; M4:         sra     $2, $[[T0]], 16
+
+  ; 64R1:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R1:       sll     $[[T0]], $[[T0]], 16
+  ; 64R1:       sra     $2, $[[T0]], 16
+
+  ; 64R2:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R2:       seh     $2, $[[T0]]
+
+  ; 64R6:       mul     $[[T0:[0-9]+]], $4, $5
+  ; 64R6:       seh     $2, $[[T0]]
+  %r = mul i16 %a, %b
+  ret i16 %r
+}
+
+define signext i32 @mul_i32(i32 signext %a, i32 signext %b) {
+entry:
+; ALL-LABEL: mul_i32:
+
+  ; M2:         mult    $4, $5
+  ; M2:         mflo    $2
+
+  ; 32R1-R2:    mul     $2, $4, $5
+  ; 32R6:       mul     $2, $4, $5
+
+  ; 64R1-R2:    mul     $2, $4, $5
+  ; 64R6:       mul     $2, $4, $5
+  %r = mul i32 %a, %b
+  ret i32 %r
+}
+
+define signext i64 @mul_i64(i64 signext %a, i64 signext %b) {
+entry:
+; ALL-LABEL: mul_i64:
+
+  ; M2:         mult    $4, $7
+  ; M2:         mflo    $[[T0:[0-9]+]]
+  ; M2:         mult    $5, $6
+  ; M2:         mflo    $[[T1:[0-9]+]]
+  ; M2:         multu   $5, $7
+  ; M2:         mflo    $3
+  ; M2:         mfhi    $4
+  ; M2:         addu    $[[T2:[0-9]+]], $4, $[[T1]]
+  ; M2:         addu    $2, $[[T2]], $[[T0]]
+
+  ; 32R1-R2:    multu   $5, $7
+  ; 32R1-R2:    mflo    $3
+  ; 32R1-R2:    mfhi    $[[T0:[0-9]+]]
+  ; 32R1-R2:    mul     $[[T1:[0-9]+]], $4, $7
+  ; 32R1-R2:    mul     $[[T2:[0-9]+]], $5, $6
+  ; 32R1-R2:    addu    $[[T0]], $[[T0]], $[[T2:[0-9]+]]
+  ; 32R1-R2:    addu    $2, $[[T0]], $[[T1]]
+
+  ; 32R6:       mul     $[[T0:[0-9]+]], $5, $6
+  ; 32R6:       muhu    $[[T1:[0-9]+]], $5, $7
+  ; 32R6:       addu    $[[T0]], $[[T1]], $[[T0]]
+  ; 32R6:       mul     $[[T2:[0-9]+]], $4, $7
+  ; 32R6:       addu    $2, $[[T0]], $[[T2]]
+  ; 32R6:       mul     $3, $5, $7
+
+  ; M4:         dmult   $4, $5
+  ; M4:         mflo    $2
+
+  ; 64R1-R2:    dmult   $4, $5
+  ; 64R1-R2:    mflo    $2
+
+  ; 64R6:       dmul    $2, $4, $5
+
+  %r = mul i64 %a, %b
+  ret i64 %r
+}
diff --git a/test/CodeGen/Mips/load-store-left-right.ll b/test/CodeGen/Mips/load-store-left-right.ll
index a3f5ebf..f6d0e8d 100644
--- a/test/CodeGen/Mips/load-store-left-right.ll
+++ b/test/CodeGen/Mips/load-store-left-right.ll
@@ -47,7 +47,7 @@ entry:
   ret i32 %0
 }
 
-define void @store_SI(i32 %a) nounwind {
+define void @store_SI(i32 signext %a) nounwind {
 entry:
 ; ALL-LABEL: store_SI:
 
@@ -201,7 +201,7 @@ entry:
   ret void
 }
 
-define void @store_SI_trunc_from_i64(i32 %a) nounwind {
+define void @store_SI_trunc_from_i64(i32 signext %a) nounwind {
 entry:
 ; ALL-LABEL: store_SI_trunc_from_i64:
 
diff --git a/test/CodeGen/Mips/longbranch.ll b/test/CodeGen/Mips/longbranch.ll
index a403744..b9b52be 100644
--- a/test/CodeGen/Mips/longbranch.ll
+++ b/test/CodeGen/Mips/longbranch.ll
@@ -13,7 +13,7 @@
 
 @x = external global i32
 
-define void @test1(i32 %s) {
+define void @test1(i32 signext %s) {
 entry:
   %cmp = icmp eq i32 %s, 0
   br i1 %cmp, label %end, label %then
diff --git a/test/CodeGen/Mips/madd-msub.ll b/test/CodeGen/Mips/madd-msub.ll
index 8222967..b0c3ff6 100644
--- a/test/CodeGen/Mips/madd-msub.ll
+++ b/test/CodeGen/Mips/madd-msub.ll
@@ -76,26 +76,14 @@ entry:
 ; 32R6-DAG:      muhu $[[T3:[0-9]+]], ${{[45]}}, ${{[45]}}
 ; 32R6-DAG:      addu $2, $[[T3]], $[[T2]]
 
-; 64-DAG:        dsll $[[T0:[0-9]+]], $4, 32
-; 64-DAG:        dsrl $[[T1:[0-9]+]], $[[T0]], 32
-; 64-DAG:        dsll $[[T2:[0-9]+]], $5, 32
-; 64-DAG:        dsrl $[[T3:[0-9]+]], $[[T2]], 32
-; 64-DAG:        d[[m:m]]ult $[[T3]], $[[T1]]
-; 64-DAG:        [[m]]flo $[[T4:[0-9]+]]
-; 64-DAG:        dsll $[[T5:[0-9]+]], $6, 32
-; 64-DAG:        dsrl $[[T6:[0-9]+]], $[[T5]], 32
-; 64-DAG:        daddu $2, $[[T4]], $[[T6]]
-
-; 64R6-DAG:      dsll $[[T0:[0-9]+]], $4, 32
-; 64R6-DAG:      dsrl $[[T1:[0-9]+]], $[[T0]], 32
-; 64R6-DAG:      dsll $[[T2:[0-9]+]], $5, 32
-; 64R6-DAG:      dsrl $[[T3:[0-9]+]], $[[T2]], 32
-; 64R6-DAG:      dmul $[[T4:[0-9]+]], $[[T3]], $[[T1]]
-; 64R6-DAG:      dsll $[[T5:[0-9]+]], $6, 32
-; 64R6-DAG:      dsrl $[[T6:[0-9]+]], $[[T5]], 32
-; 64R6-DAG:      daddu $2, $[[T4]], $[[T6]]
-
-define i64 @madd2(i32 %a, i32 %b, i32 %c) nounwind readnone {
+; 64-DAG:        d[[m:m]]ult $5, $4
+; 64-DAG:        [[m]]flo $[[T0:[0-9]+]]
+; 64-DAG:        daddu $2, $[[T0]], $6
+
+; 64R6-DAG:      dmul $[[T0:[0-9]+]], $5, $4
+; 64R6-DAG:      daddu $2, $[[T0]], $6
+
+define i64 @madd2(i32 zeroext %a, i32 zeroext %b, i32 zeroext %c) nounwind readnone {
 entry:
   %conv = zext i32 %a to i64
   %conv2 = zext i32 %b to i64
@@ -214,26 +202,14 @@ entry:
 ; 32R6-DAG:      negu $2, $[[T3]]
 ; 32R6-DAG:      subu $3, $6, $[[T1]]
 
-; 64-DAG:        dsll $[[T0:[0-9]+]], $4, 32
-; 64-DAG:        dsrl $[[T1:[0-9]+]], $[[T0]], 32
-; 64-DAG:        dsll $[[T2:[0-9]+]], $5, 32
-; 64-DAG:        dsrl $[[T3:[0-9]+]], $[[T2]], 32
-; 64-DAG:        d[[m:m]]ult $[[T3]], $[[T1]]
-; 64-DAG:        [[m]]flo $[[T4:[0-9]+]]
-; 64-DAG:        dsll $[[T5:[0-9]+]], $6, 32
-; 64-DAG:        dsrl $[[T6:[0-9]+]], $[[T5]], 32
-; 64-DAG:        dsubu $2, $[[T6]], $[[T4]]
-
-; 64R6-DAG:      dsll $[[T0:[0-9]+]], $4, 32
-; 64R6-DAG:      dsrl $[[T1:[0-9]+]], $[[T0]], 32
-; 64R6-DAG:      dsll $[[T2:[0-9]+]], $5, 32
-; 64R6-DAG:      dsrl $[[T3:[0-9]+]], $[[T2]], 32
-; 64R6-DAG:      dmul $[[T4:[0-9]+]], $[[T3]], $[[T1]]
-; 64R6-DAG:      dsll $[[T5:[0-9]+]], $6, 32
-; 64R6-DAG:      dsrl $[[T6:[0-9]+]], $[[T5]], 32
-; 64R6-DAG:      dsubu $2, $[[T6]], $[[T4]]
-
-define i64 @msub2(i32 %a, i32 %b, i32 %c) nounwind readnone {
+; 64-DAG:        d[[m:m]]ult $5, $4
+; 64-DAG:        [[m]]flo $[[T0:[0-9]+]]
+; 64-DAG:        dsubu $2, $6, $[[T0]]
+
+; 64R6-DAG:      dmul $[[T0:[0-9]+]], $5, $4
+; 64R6-DAG:      dsubu $2, $6, $[[T0]]
+
+define i64 @msub2(i32 zeroext %a, i32 zeroext %b, i32 zeroext %c) nounwind readnone {
 entry:
   %conv = zext i32 %c to i64
   %conv2 = zext i32 %a to i64
diff --git a/test/CodeGen/Mips/micromips-addiu.ll b/test/CodeGen/Mips/micromips-addiu.ll
new file mode 100644
index 0000000..c5bee34
--- /dev/null
+++ b/test/CodeGen/Mips/micromips-addiu.ll
@@ -0,0 +1,32 @@
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+micromips \
+; RUN:   -relocation-model=pic -O3 < %s | FileCheck %s
+
+@x = global i32 65504, align 4
+@y = global i32 60929, align 4
+@z = global i32 60929, align 4
+@.str = private unnamed_addr constant [7 x i8] c"%08x \0A\00", align 1
+
+define i32 @main() nounwind {
+entry:
+  %0 = load i32* @x, align 4
+  %addiu1 = add i32 %0, -7
+  %call1 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds
+                                  ([7 x i8]* @.str, i32 0, i32 0), i32 %addiu1)
+
+  %1 = load i32* @y, align 4
+  %addiu2 = add i32 %1, 55
+  %call2 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds
+                                  ([7 x i8]* @.str, i32 0, i32 0), i32 %addiu2)
+
+  %2 = load i32* @z, align 4
+  %addiu3 = add i32 %2, 24
+  %call3 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds
+                                  ([7 x i8]* @.str, i32 0, i32 0), i32 %addiu3)
+  ret i32 0
+}
+
+declare i32 @printf(i8*, ...)
+
+; CHECK: addius5  ${{[0-9]+}}, -7
+; CHECK: addiu    ${{[0-9]+}}, ${{[0-9]+}}, 55
+; CHECK: addiur2  ${{[2-7]|16|17}}, ${{[2-7]|16|17}}, 24
diff --git a/test/CodeGen/Mips/micromips-andi.ll b/test/CodeGen/Mips/micromips-andi.ll
new file mode 100644
index 0000000..b82d2b0
--- /dev/null
+++ b/test/CodeGen/Mips/micromips-andi.ll
@@ -0,0 +1,25 @@
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+micromips \
+; RUN:   -relocation-model=pic -O3 < %s | FileCheck %s
+
+@x = global i32 65504, align 4
+@y = global i32 60929, align 4
+@.str = private unnamed_addr constant [7 x i8] c"%08x \0A\00", align 1
+
+define i32 @main() nounwind {
+entry:
+  %0 = load i32* @x, align 4
+  %and1 = and i32 %0, 4
+  %call1 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds
+                                  ([7 x i8]* @.str, i32 0, i32 0), i32 %and1)
+
+  %1 = load i32* @y, align 4
+  %and2 = and i32 %1, 5
+  %call2 = call i32 (i8*, ...)* @printf(i8* getelementptr inbounds
+                                  ([7 x i8]* @.str, i32 0, i32 0), i32 %and2)
+  ret i32 0
+}
+
+declare i32 @printf(i8*, ...)
+
+; CHECK: andi16 ${{[2-7]|16|17}}, ${{[2-7]|16|17}}
+; CHECK: andi   ${{[0-9]+}}, ${{[0-9]+}}
diff --git a/test/CodeGen/Mips/micromips-delay-slot.ll b/test/CodeGen/Mips/micromips-delay-slot.ll
new file mode 100644
index 0000000..4bab97a
--- /dev/null
+++ b/test/CodeGen/Mips/micromips-delay-slot.ll
@@ -0,0 +1,18 @@
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+micromips \
+; RUN:   -relocation-model=pic -O3 < %s | FileCheck %s
+
+; Function Attrs: nounwind uwtable
+define i32 @foo(i32 %a) #0 {
+entry:
+  %a.addr = alloca i32, align 4
+  store i32 %a, i32* %a.addr, align 4
+  %0 = load i32* %a.addr, align 4
+  %shl = shl i32 %0, 2
+  %call = call i32 @bar(i32 %shl)
+  ret i32 %call
+}
+
+declare i32 @bar(i32) #1
+
+; CHECK: nop
+
diff --git a/test/CodeGen/Mips/micromips-rdhwr-directives.ll b/test/CodeGen/Mips/micromips-rdhwr-directives.ll
new file mode 100644
index 0000000..af40a87
--- /dev/null
+++ b/test/CodeGen/Mips/micromips-rdhwr-directives.ll
@@ -0,0 +1,15 @@
+; RUN: llc -march=mipsel -mcpu=mips32 -relocation-model=static < %s \
+; RUN:   -mattr=+micromips | FileCheck %s
+
+@a = external thread_local global i32
+
+define i32 @foo() nounwind readonly {
+entry:
+; CHECK: .set  push
+; CHECK: .set  mips32r2
+; CHECK: rdhwr
+; CHECK: .set  pop
+
+  %0 = load i32* @a, align 4
+  ret i32 %0
+}
diff --git a/test/CodeGen/Mips/micromips-shift.ll b/test/CodeGen/Mips/micromips-shift.ll
new file mode 100644
index 0000000..8215010
--- /dev/null
+++ b/test/CodeGen/Mips/micromips-shift.ll
@@ -0,0 +1,44 @@
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+micromips \
+; RUN:   -relocation-model=pic -O3 < %s | FileCheck %s
+
+@a = global i32 10, align 4
+@b = global i32 0, align 4
+@c = global i32 10, align 4
+@d = global i32 0, align 4
+
+define i32 @shift_left() nounwind {
+entry:
+  %0 = load i32* @a, align 4
+  %shl = shl i32 %0, 4
+  store i32 %shl, i32* @b, align 4
+
+  %1 = load i32* @c, align 4
+  %shl1 = shl i32 %1, 10
+  store i32 %shl1, i32* @d, align 4
+
+  ret i32 0
+}
+
+; CHECK: sll16  ${{[2-7]|16|17}}, ${{[2-7]|16|17}}, {{[0-7]}}
+; CHECK: sll    ${{[0-9]+}}, ${{[0-9]+}}, {{[0-9]+}}
+
+@i = global i32 10654, align 4
+@j = global i32 0, align 4
+@m = global i32 10, align 4
+@n = global i32 0, align 4
+
+define i32 @shift_right() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %shr = lshr i32 %0, 4
+  store i32 %shr, i32* @j, align 4
+
+  %1 = load i32* @m, align 4
+  %shr1 = lshr i32 %1, 10
+  store i32 %shr1, i32* @n, align 4
+
+  ret i32 0
+}
+
+; CHECK: srl16  ${{[2-7]|16|17}}, ${{[2-7]|16|17}}, {{[0-7]}}
+; CHECK: srl    ${{[0-9]+}}, ${{[0-9]+}}, {{[0-9]+}}
diff --git a/test/CodeGen/Mips/mips16-hf-attr-2.ll b/test/CodeGen/Mips/mips16-hf-attr-2.ll
new file mode 100644
index 0000000..60c6eaa
--- /dev/null
+++ b/test/CodeGen/Mips/mips16-hf-attr-2.ll
@@ -0,0 +1,45 @@
+; Check that stubs generation for mips16 hard-float mode does not depend
+; on the function 'use-soft-float' attribute's value.
+; RUN: llc -mtriple=mipsel-linux-gnu -march=mipsel \
+; RUN:     -mcpu=mips16 -relocation-model=pic < %s | FileCheck %s
+
+define void @bar_sf() #1 {
+; CHECK: bar_sf:
+entry:
+  %call1 = call float @foo(float 1.000000e+00)
+; CHECK: lw $3, %call16(foo)($2)
+; CHECK-NOT: lw $5, %got(__mips16_call_stub_sf_1)($3)
+  ret void
+}
+
+define void @bar_hf() #0 {
+; CHECK: bar_hf:
+entry:
+  %call1 = call float @foo(float 1.000000e+00)
+; CHECK: lw $2, %call16(foo)($3)
+; CHECK: lw $5, %got(__mips16_call_stub_sf_1)($3)
+  ret void
+}
+
+declare float @foo(float) #2
+
+attributes #0 = {
+  nounwind
+  "less-precise-fpmad"="false" "no-frame-pointer-elim"="true"
+  "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false"
+  "no-nans-fp-math"="false" "stack-protector-buffer-size"="8"
+  "unsafe-fp-math"="false" "use-soft-float"="false"
+}
+attributes #1 = {
+  nounwind
+  "less-precise-fpmad"="false" "no-frame-pointer-elim"="true"
+  "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false"
+  "no-nans-fp-math"="false" "stack-protector-buffer-size"="8"
+  "unsafe-fp-math"="false" "use-soft-float"="true"
+}
+attributes #2 = {
+  "less-precise-fpmad"="false" "no-frame-pointer-elim"="true"
+  "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false"
+  "no-nans-fp-math"="false" "stack-protector-buffer-size"="8"
+  "unsafe-fp-math"="false" "use-soft-float"="true"
+}
diff --git a/test/CodeGen/Mips/mips16-hf-attr.ll b/test/CodeGen/Mips/mips16-hf-attr.ll
index d9ad629..c6ad442 100644
--- a/test/CodeGen/Mips/mips16-hf-attr.ll
+++ b/test/CodeGen/Mips/mips16-hf-attr.ll
@@ -3,8 +3,8 @@
 ; RUN: llc -mtriple=mipsel-linux-gnu -march=mipsel \
 ; RUN:     -mcpu=mips16 -relocation-model=pic < %s | FileCheck %s
 
-define void @bar_sf() #0 {
-; CHECK: bar_sf:
+define void @bar_hf() #0 {
+; CHECK: bar_hf:
 entry:
   %call1 = call float @foo(float 1.000000e+00)
 ; CHECK: lw $2, %call16(foo)($3)
@@ -12,12 +12,12 @@ entry:
   ret void
 }
 
-define void @bar_hf() #1 {
-; CHECK: bar_hf:
+define void @bar_sf() #1 {
+; CHECK: bar_sf:
 entry:
   %call1 = call float @foo(float 1.000000e+00)
-; CHECK: lw $2, %call16(foo)($3)
-; CHECK: lw $5, %got(__mips16_call_stub_sf_1)($3)
+; CHECK: lw $3, %call16(foo)($2)
+; CHECK-NOT: lw $5, %got(__mips16_call_stub_sf_1)($3)
   ret void
 }
 
diff --git a/test/CodeGen/Mips/mips64-f128.ll b/test/CodeGen/Mips/mips64-f128.ll
index 7f7d515..6987d4a 100644
--- a/test/CodeGen/Mips/mips64-f128.ll
+++ b/test/CodeGen/Mips/mips64-f128.ll
@@ -114,7 +114,7 @@ entry:
 ; ALL-LABEL: conv_LD_UInt:
 ; ALL: ld $25, %call16(__floatunsitf)
 
-define fp128 @conv_LD_UInt(i32 %a) {
+define fp128 @conv_LD_UInt(i32 signext %a) {
 entry:
   %conv = uitofp i32 %a to fp128
   ret fp128 %conv
@@ -545,7 +545,7 @@ entry:
 
 ; ALL-LABEL: load_LD_float:
 ; ALL: ld   $[[R0:[0-9]+]], %got_disp(gf1)
-; ALL: lw   $4, 0($[[R0]])
+; ALL: lwu  $4, 0($[[R0]])
 ; ALL: ld   $25, %call16(__extendsftf2)
 ; ALL: jalr $25
 
@@ -635,7 +635,7 @@ entry:
 ; CMP_CC_FMT-DAG: selnez $[[NE2:[0-9]+]], $7, $[[CC]]
 ; CMP_CC_FMT-DAG: or $4, $[[NE2]], $[[EQ2]]
 
-define fp128 @select_LD(i32 %a, i64, fp128 %b, fp128 %c) {
+define fp128 @select_LD(i32 signext %a, i64, fp128 %b, fp128 %c) {
 entry:
   %tobool = icmp ne i32 %a, 0
   %cond = select i1 %tobool, fp128 %b, fp128 %c
diff --git a/test/CodeGen/Mips/mips64-sret.ll b/test/CodeGen/Mips/mips64-sret.ll
index 7a52c3d..ed494e9 100644
--- a/test/CodeGen/Mips/mips64-sret.ll
+++ b/test/CodeGen/Mips/mips64-sret.ll
@@ -11,7 +11,7 @@ entry:
   ret void
 }
 
-define void @bar(i32 %v, i32* noalias sret %agg.result) nounwind {
+define void @bar(i32 signext %v, i32* noalias sret %agg.result) nounwind {
 entry:
 ; CHECK-LABEL: bar:
 ; CHECK: sw $4, 0($5)
diff --git a/test/CodeGen/Mips/mno-ldc1-sdc1.ll b/test/CodeGen/Mips/mno-ldc1-sdc1.ll
index 244b03d..db653ea 100644
--- a/test/CodeGen/Mips/mno-ldc1-sdc1.ll
+++ b/test/CodeGen/Mips/mno-ldc1-sdc1.ll
@@ -123,7 +123,7 @@ entry:
 ; 32R1-LE-PIC-DAG:    sw $[[R1]], 4(${{[0-9]+}})
 
 ; 32R2-LE-PIC-DAG:    mfc1 $[[R0:[0-9]+]], $f12
-; 32R2-LE-PIC-DAG:    mfc1 $[[R1:[0-9]+]], $f13
+; 32R2-LE-PIC-DAG:    mfhc1 $[[R1:[0-9]+]], $f12
 ; 32R2-LE-PIC-DAG:    sw $[[R0]], 0(${{[0-9]+}})
 ; 32R2-LE-PIC-DAG:    sw $[[R1]], 4(${{[0-9]+}})
 
@@ -140,7 +140,7 @@ entry:
 ; 32R1-LE-STATIC-DAG: sw $[[R1]], 4($[[R3]])
 
 ; 32R2-LE-STATIC-DAG: mfc1 $[[R0:[0-9]+]], $f12
-; 32R2-LE-STATIC-DAG: mfc1 $[[R1:[0-9]+]], $f13
+; 32R2-LE-STATIC-DAG: mfhc1 $[[R1:[0-9]+]], $f12
 ; 32R2-LE-STATIC-DAG: lui $[[R2:[0-9]+]], %hi(g0)
 ; 32R2-LE-STATIC-DAG: sw $[[R0]], %lo(g0)($[[R2]])
 ; 32R2-LE-STATIC-DAG: addiu $[[R3:[0-9]+]], $[[R2]], %lo(g0)
@@ -159,7 +159,7 @@ entry:
 ; 32R1-BE-PIC-DAG:    sw $[[R0]], 4(${{[0-9]+}})
 
 ; 32R2-BE-PIC-DAG:    mfc1 $[[R0:[0-9]+]], $f12
-; 32R2-BE-PIC-DAG:    mfc1 $[[R1:[0-9]+]], $f13
+; 32R2-BE-PIC-DAG:    mfhc1 $[[R1:[0-9]+]], $f12
 ; 32R2-BE-PIC-DAG:    sw $[[R1]], 0(${{[0-9]+}})
 ; 32R2-BE-PIC-DAG:    sw $[[R0]], 4(${{[0-9]+}})
 
@@ -225,7 +225,7 @@ entry:
 ; 32R1-DAG:      sw $[[R1]], 4(${{[0-9]+}})
 
 ; 32R2-DAG:      mfc1 $[[R0:[0-9]+]], $f12
-; 32R2-DAG:      mfc1 $[[R1:[0-9]+]], $f13
+; 32R2-DAG:      mfhc1 $[[R1:[0-9]+]], $f12
 ; 32R2-DAG:      sw $[[R0]], 0(${{[0-9]+}})
 ; 32R2-DAG:      sw $[[R1]], 4(${{[0-9]+}})
 
diff --git a/test/CodeGen/Mips/msa/arithmetic_float.ll b/test/CodeGen/Mips/msa/arithmetic_float.ll
index 86e57ac..9aae284 100644
--- a/test/CodeGen/Mips/msa/arithmetic_float.ll
+++ b/test/CodeGen/Mips/msa/arithmetic_float.ll
@@ -276,8 +276,8 @@ define void @fexp2_v4f32_2(<4 x float>* %c, <4 x float>* %a) nounwind {
   ; CHECK-DAG: ld.w [[R1:\$w[0-9]+]], 0($5)
   %2 = tail call <4 x float> @llvm.exp2.v4f32 (<4 x float> %1)
   %3 = fmul <4 x float> <float 2.0, float 2.0, float 2.0, float 2.0>, %2
-  ; CHECK-DAG: lui [[R3:\$[0-9]+]], 16384
-  ; CHECK-DAG: fill.w [[R4:\$w[0-9]+]], [[R3]]
+  ; CHECK-DAG: ldi.w [[R3:\$w[0-9]+]], 1
+  ; CHECK-DAG: ffint_u.w [[R4:\$w[0-9]+]], [[R3]]
   ; CHECK-DAG: fexp2.w [[R5:\$w[0-9]+]], [[R4]], [[R1]]
   store <4 x float> %3, <4 x float>* %c
   ; CHECK-DAG: st.w [[R5]], 0($4)
@@ -287,16 +287,14 @@ define void @fexp2_v4f32_2(<4 x float>* %c, <4 x float>* %a) nounwind {
 }
 
 define void @fexp2_v2f64_2(<2 x double>* %c, <2 x double>* %a) nounwind {
-  ; CHECK:      .8byte 4611686018427387904
-  ; CHECK-NEXT: .8byte 4611686018427387904
   ; CHECK: fexp2_v2f64_2:
 
   %1 = load <2 x double>* %a
   ; CHECK-DAG: ld.d [[R1:\$w[0-9]+]], 0($5)
   %2 = tail call <2 x double> @llvm.exp2.v2f64 (<2 x double> %1)
   %3 = fmul <2 x double> <double 2.0, double 2.0>, %2
-  ; CHECK-DAG: addiu [[G_PTR:\$[0-9]+]], {{.*}}, %lo($
-  ; CHECK-DAG: ld.d [[R3:\$w[0-9]+]], 0([[G_PTR]])
+  ; CHECK-DAG: ldi.d [[R2:\$w[0-9]+]], 1
+  ; CHECK-DAG: ffint_u.d [[R3:\$w[0-9]+]], [[R2]]
   ; CHECK-DAG: fexp2.d [[R4:\$w[0-9]+]], [[R3]], [[R1]]
   store <2 x double> %3, <2 x double>* %c
   ; CHECK-DAG: st.d [[R4]], 0($4)
diff --git a/test/CodeGen/Mips/msa/frameindex.ll b/test/CodeGen/Mips/msa/frameindex.ll
index 07e67bf..ebec465 100644
--- a/test/CodeGen/Mips/msa/frameindex.ll
+++ b/test/CodeGen/Mips/msa/frameindex.ll
@@ -36,10 +36,10 @@ define void @loadstore_v16i8_just_over_simm10() nounwind {
   %2 = alloca [497 x i8] ; Push the frame just over 512 bytes
 
   %3 = load volatile <16 x i8>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 512
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 512
   ; MIPS32-AE: ld.b [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <16 x i8> %3, <16 x i8>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 512
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 512
   ; MIPS32-AE: st.b [[R1]], 0([[BASE]])
 
   ret void
@@ -53,12 +53,12 @@ define void @loadstore_v16i8_just_under_simm16() nounwind {
   %2 = alloca [32752 x i8] ; Push the frame right up to 32768 bytes
 
   %3 = load volatile <16 x i8>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.b [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <16 x i8> %3, <16 x i8>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.b [[R1]], 0([[BASE]])
 
   ret void
@@ -72,12 +72,12 @@ define void @loadstore_v16i8_just_over_simm16() nounwind {
   %2 = alloca [32753 x i8] ; Push the frame just over 32768 bytes
 
   %3 = load volatile <16 x i8>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.b [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <16 x i8> %3, <16 x i8>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.b [[R1]], 0([[BASE]])
 
   ret void
@@ -107,10 +107,10 @@ define void @loadstore_v8i16_unaligned() nounwind {
   %5 = getelementptr [2 x <8 x i16>]* %4, i32 0, i32 0
 
   %6 = load volatile <8 x i16>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: ld.h [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <8 x i16> %6, <8 x i16>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: st.h [[R1]], 0([[BASE]])
 
   ret void
@@ -139,10 +139,10 @@ define void @loadstore_v8i16_just_over_simm10() nounwind {
   %2 = alloca [1009 x i8] ; Push the frame just over 1024 bytes
 
   %3 = load volatile <8 x i16>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1024
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1024
   ; MIPS32-AE: ld.h [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <8 x i16> %3, <8 x i16>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1024
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1024
   ; MIPS32-AE: st.h [[R1]], 0([[BASE]])
 
   ret void
@@ -156,12 +156,12 @@ define void @loadstore_v8i16_just_under_simm16() nounwind {
   %2 = alloca [32752 x i8] ; Push the frame right up to 32768 bytes
 
   %3 = load volatile <8 x i16>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.h [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <8 x i16> %3, <8 x i16>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.h [[R1]], 0([[BASE]])
 
   ret void
@@ -175,12 +175,12 @@ define void @loadstore_v8i16_just_over_simm16() nounwind {
   %2 = alloca [32753 x i8] ; Push the frame just over 32768 bytes
 
   %3 = load volatile <8 x i16>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.h [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <8 x i16> %3, <8 x i16>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.h [[R1]], 0([[BASE]])
 
   ret void
@@ -210,10 +210,10 @@ define void @loadstore_v4i32_unaligned() nounwind {
   %5 = getelementptr [2 x <4 x i32>]* %4, i32 0, i32 0
 
   %6 = load volatile <4 x i32>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: ld.w [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <4 x i32> %6, <4 x i32>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: st.w [[R1]], 0([[BASE]])
 
   ret void
@@ -242,10 +242,10 @@ define void @loadstore_v4i32_just_over_simm10() nounwind {
   %2 = alloca [2033 x i8] ; Push the frame just over 2048 bytes
 
   %3 = load volatile <4 x i32>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 2048
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 2048
   ; MIPS32-AE: ld.w [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <4 x i32> %3, <4 x i32>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 2048
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 2048
   ; MIPS32-AE: st.w [[R1]], 0([[BASE]])
 
   ret void
@@ -259,12 +259,12 @@ define void @loadstore_v4i32_just_under_simm16() nounwind {
   %2 = alloca [32752 x i8] ; Push the frame right up to 32768 bytes
 
   %3 = load volatile <4 x i32>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.w [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <4 x i32> %3, <4 x i32>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.w [[R1]], 0([[BASE]])
 
   ret void
@@ -278,12 +278,12 @@ define void @loadstore_v4i32_just_over_simm16() nounwind {
   %2 = alloca [32753 x i8] ; Push the frame just over 32768 bytes
 
   %3 = load volatile <4 x i32>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.w [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <4 x i32> %3, <4 x i32>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.w [[R1]], 0([[BASE]])
 
   ret void
@@ -313,10 +313,10 @@ define void @loadstore_v2i64_unaligned() nounwind {
   %5 = getelementptr [2 x <2 x i64>]* %4, i32 0, i32 0
 
   %6 = load volatile <2 x i64>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: ld.d [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <2 x i64> %6, <2 x i64>* %5
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 1
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 1
   ; MIPS32-AE: st.d [[R1]], 0([[BASE]])
 
   ret void
@@ -345,10 +345,10 @@ define void @loadstore_v2i64_just_over_simm10() nounwind {
   %2 = alloca [4081 x i8] ; Push the frame just over 4096 bytes
 
   %3 = load volatile <2 x i64>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 4096
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 4096
   ; MIPS32-AE: ld.d [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <2 x i64> %3, <2 x i64>* %1
-  ; MIPS32-AE: addiu [[BASE:\$[0-9]+]], $sp, 4096
+  ; MIPS32-AE: addiu [[BASE:\$([0-9]+|gp)]], $sp, 4096
   ; MIPS32-AE: st.d [[R1]], 0([[BASE]])
 
   ret void
@@ -362,12 +362,12 @@ define void @loadstore_v2i64_just_under_simm16() nounwind {
   %2 = alloca [32752 x i8] ; Push the frame right up to 32768 bytes
 
   %3 = load volatile <2 x i64>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.d [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <2 x i64> %3, <2 x i64>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.d [[R1]], 0([[BASE]])
 
   ret void
@@ -381,12 +381,12 @@ define void @loadstore_v2i64_just_over_simm16() nounwind {
   %2 = alloca [32753 x i8] ; Push the frame just over 32768 bytes
 
   %3 = load volatile <2 x i64>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: ld.d [[R1:\$w[0-9]+]], 0([[BASE]])
   store volatile <2 x i64> %3, <2 x i64>* %1
-  ; MIPS32-AE: ori [[R2:\$[0-9]+]], $zero, 32768
-  ; MIPS32-AE: addu [[BASE:\$[0-9]+]], $sp, [[R2]]
+  ; MIPS32-AE: ori [[R2:\$([0-9]+|gp)]], $zero, 32768
+  ; MIPS32-AE: addu [[BASE:\$([0-9]+|gp)]], $sp, [[R2]]
   ; MIPS32-AE: st.d [[R1]], 0([[BASE]])
 
   ret void
diff --git a/test/CodeGen/Mips/no-odd-spreg.ll b/test/CodeGen/Mips/no-odd-spreg.ll
index b42ed6a..572e940 100644
--- a/test/CodeGen/Mips/no-odd-spreg.ll
+++ b/test/CodeGen/Mips/no-odd-spreg.ll
@@ -1,10 +1,14 @@
-; RUN: llc -march=mipsel -mcpu=mips32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=ODDSPREG
+; RUN: llc -march=mipsel -mcpu=mips32 < %s | FileCheck %s -check-prefix=ALL -check-prefix=ODDSPREG -check-prefix=ODDSPREG-NO-EMIT
 ; RUN: llc -march=mipsel -mcpu=mips32 -mattr=+nooddspreg < %s | FileCheck %s -check-prefix=ALL -check-prefix=NOODDSPREG
-; RUN: llc -march=mipsel -mcpu=mips32r6 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=ODDSPREG
+; RUN: llc -march=mipsel -mcpu=mips32r6 -mattr=fp64 < %s | FileCheck %s -check-prefix=ALL -check-prefix=ODDSPREG -check-prefix=ODDSPREG-NO-EMIT
 ; RUN: llc -march=mipsel -mcpu=mips32r6 -mattr=fp64,+nooddspreg < %s | FileCheck %s -check-prefix=ALL -check-prefix=NOODDSPREG
+; RUN: llc -march=mipsel -mcpu=mips32r6 -mattr=fpxx,-nooddspreg < %s | FileCheck %s -check-prefix=ALL -check-prefix=ODDSPREG -check-prefix=ODDSPREG-EMIT
 
-; ODDSPREG:       .module oddspreg
-; NOODDSPREG:     .module nooddspreg
+; We don't emit a directive unless we need to. This is to support versions of
+; GAS which do not support the directive.
+; ODDSPREG-EMIT:        .module oddspreg
+; ODDSPREG-NO-EMIT-NOT: .module oddspreg
+; NOODDSPREG:           .module nooddspreg
 
 define float @two_floats(float %a) {
 entry:
diff --git a/test/CodeGen/Mips/nomips16.ll b/test/CodeGen/Mips/nomips16.ll
index 0affb16..5f7d74e 100644
--- a/test/CodeGen/Mips/nomips16.ll
+++ b/test/CodeGen/Mips/nomips16.ll
@@ -33,6 +33,6 @@ entry:
 ; CHECK: 	.end	nofoo
 
 
-attributes #0 = { nounwind "less-precise-fpmad"="false" "mips16" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="true" }
-attributes #1 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "nomips16" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="true" }
+attributes #0 = { nounwind "less-precise-fpmad"="false" "mips16" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+attributes #1 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "nomips16" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 
diff --git a/test/CodeGen/Mips/o32_cc.ll b/test/CodeGen/Mips/o32_cc.ll
index 08e5aab..c28f9ab 100644
--- a/test/CodeGen/Mips/o32_cc.ll
+++ b/test/CodeGen/Mips/o32_cc.ll
@@ -1,12 +1,13 @@
-; RUN: llc -march=mipsel < %s | FileCheck %s
-; RUN: llc -march=mipsel -mattr=+fp64 < %s | FileCheck %s
-; RUN: llc -march=mipsel < %s | FileCheck -check-prefix=FP32EL %s
-; RUN: llc -march=mipsel -mattr=+fp64 < %s | FileCheck -check-prefix=FP64EL %s
+; RUN: llc -march=mipsel < %s | FileCheck -check-prefix=ALL %s
+; RUN: llc -march=mipsel -mattr=+fp64 < %s | FileCheck -check-prefix=ALL %s
+; RUN: llc -march=mipsel -mcpu=mips32 < %s | FileCheck -check-prefix=ALL -check-prefix=NO-MFHC1 %s
+; RUN: llc -march=mipsel -mcpu=mips32r2              < %s | FileCheck -check-prefix=ALL -check-prefix=HAS-MFHC1 %s
+; RUN: llc -march=mipsel -mcpu=mips32r2 -mattr=+fp64 < %s | FileCheck -check-prefix=ALL -check-prefix=HAS-MFHC1 %s
 
 ; $f12, $f14
-; CHECK-LABEL: testlowercall0:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: ldc1 $f14, %lo
+; ALL-LABEL: testlowercall0:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       ldc1 $f14, %lo
 define void @testlowercall0() nounwind {
 entry:
   tail call void @f0(double 5.000000e+00, double 6.000000e+00) nounwind
@@ -16,9 +17,9 @@ entry:
 declare void @f0(double, double)
 
 ; $f12, $f14
-; CHECK-LABEL: testlowercall1:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
+; ALL-LABEL: testlowercall1:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       lwc1 $f14, %lo
 define void @testlowercall1() nounwind {
 entry:
   tail call void @f1(float 8.000000e+00, float 9.000000e+00) nounwind
@@ -28,9 +29,9 @@ entry:
 declare void @f1(float, float)
 
 ; $f12, $f14
-; CHECK-LABEL: testlowercall2:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: ldc1 $f14, %lo
+; ALL-LABEL: testlowercall2:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       ldc1 $f14, %lo
 define void @testlowercall2() nounwind {
 entry:
   tail call void @f2(float 8.000000e+00, double 6.000000e+00) nounwind
@@ -40,9 +41,9 @@ entry:
 declare void @f2(float, double)
 
 ; $f12, $f14
-; CHECK-LABEL: testlowercall3:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
+; ALL-LABEL: testlowercall3:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       lwc1 $f14, %lo
 define void @testlowercall3() nounwind {
 entry:
   tail call void @f3(double 5.000000e+00, float 9.000000e+00) nounwind
@@ -52,11 +53,11 @@ entry:
 declare void @f3(double, float)
 
 ; $4, $5, $6, $7
-; CHECK-LABEL: testlowercall4:
-; CHECK-DAG: addiu $4, $zero, 12
-; CHECK-DAG: addiu $5, $zero, 13
-; CHECK-DAG: addiu $6, $zero, 14
-; CHECK-DAG: addiu $7, $zero, 15
+; ALL-LABEL: testlowercall4:
+; ALL-DAG:       addiu $4, $zero, 12
+; ALL-DAG:       addiu $5, $zero, 13
+; ALL-DAG:       addiu $6, $zero, 14
+; ALL-DAG:       addiu $7, $zero, 15
 define void @testlowercall4() nounwind {
 entry:
   tail call void @f4(i32 12, i32 13, i32 14, i32 15) nounwind
@@ -66,11 +67,11 @@ entry:
 declare void @f4(i32, i32, i32, i32)
 
 ; $f12, $6, stack
-; CHECK-LABEL: testlowercall5:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: addiu $6, $zero, 23
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 16($sp)
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 20($sp)
+; ALL-LABEL: testlowercall5:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       addiu $6, $zero, 23
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 16($sp)
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 20($sp)
 define void @testlowercall5() nounwind {
 entry:
   tail call void @f5(double 1.500000e+01, i32 23, double 1.700000e+01) nounwind
@@ -80,10 +81,10 @@ entry:
 declare void @f5(double, i32, double)
 
 ; $f12, $6, $7
-; CHECK-LABEL: testlowercall6:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: addiu $6, $zero, 33
-; CHECK-DAG: addiu $7, $zero, 24
+; ALL-LABEL: testlowercall6:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       addiu $6, $zero, 33
+; ALL-DAG:       addiu $7, $zero, 24
 define void @testlowercall6() nounwind {
 entry:
   tail call void @f6(double 2.500000e+01, i32 33, i32 24) nounwind
@@ -93,10 +94,10 @@ entry:
 declare void @f6(double, i32, i32)
 
 ; $f12, $5, $6
-; CHECK-LABEL: testlowercall7:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: addiu $5, $zero, 43
-; CHECK-DAG: addiu $6, $zero, 34
+; ALL-LABEL: testlowercall7:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       addiu $5, $zero, 43
+; ALL-DAG:       addiu $6, $zero, 34
 define void @testlowercall7() nounwind {
 entry:
   tail call void @f7(float 1.800000e+01, i32 43, i32 34) nounwind
@@ -106,12 +107,12 @@ entry:
 declare void @f7(float, i32, i32)
 
 ; $4, $5, $6, stack
-; CHECK-LABEL: testlowercall8:
-; CHECK-DAG: addiu $4, $zero, 22
-; CHECK-DAG: addiu $5, $zero, 53
-; CHECK-DAG: addiu $6, $zero, 44
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 16($sp)
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 20($sp)
+; ALL-LABEL: testlowercall8:
+; ALL-DAG:       addiu $4, $zero, 22
+; ALL-DAG:       addiu $5, $zero, 53
+; ALL-DAG:       addiu $6, $zero, 44
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 16($sp)
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 20($sp)
 define void @testlowercall8() nounwind {
 entry:
   tail call void @f8(i32 22, i32 53, i32 44, double 4.000000e+00) nounwind
@@ -121,11 +122,11 @@ entry:
 declare void @f8(i32, i32, i32, double)
 
 ; $4, $5, $6, $7
-; CHECK-LABEL: testlowercall9:
-; CHECK-DAG: addiu $4, $zero, 32
-; CHECK-DAG: addiu $5, $zero, 63
-; CHECK-DAG: addiu $6, $zero, 54
-; CHECK-DAG: lui $7, 16688
+; ALL-LABEL: testlowercall9:
+; ALL-DAG:       addiu $4, $zero, 32
+; ALL-DAG:       addiu $5, $zero, 63
+; ALL-DAG:       addiu $6, $zero, 54
+; ALL-DAG:       lui $7, 16688
 define void @testlowercall9() nounwind {
 entry:
   tail call void @f9(i32 32, i32 63, i32 54, float 1.100000e+01) nounwind
@@ -135,15 +136,16 @@ entry:
 declare void @f9(i32, i32, i32, float)
 
 ; $4, $5, ($6, $7)
-; CHECK-LABEL: testlowercall10:
-; CHECK-DAG: addiu $4, $zero, 42
-; CHECK-DAG: addiu $5, $zero, 73
-; FP32EL-LABEL: testlowercall10:
-; FP32EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP32EL-DAG: mfc1 $7, $f{{[0-9]+}}
-; FP64EL-LABEL: testlowercall10:
-; FP64EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP64EL-DAG: mfhc1 $7, $f{{[0-9]+}}
+; ALL-LABEL: testlowercall10:
+
+; ALL-DAG:       addiu $4, $zero, 42
+; ALL-DAG:       addiu $5, $zero, 73
+
+; NO-MFHC1-DAG:  mfc1 $6, $f{{[0-9]+}}
+; NO-MFHC1-DAG:  mfc1 $7, $f{{[0-9]+}}
+
+; HAS-MFHC1-DAG: mfc1 $6, $f{{[0-9]+}}
+; HAS-MFHC1-DAG: mfhc1 $7, $f{{[0-9]+}}
 define void @testlowercall10() nounwind {
 entry:
   tail call void @f10(i32 42, i32 73, double 2.700000e+01) nounwind
@@ -153,14 +155,14 @@ entry:
 declare void @f10(i32, i32, double)
 
 ; $4, ($6, $7)
-; CHECK-LABEL: testlowercall11:
-; CHECK-DAG: addiu $4, $zero, 52
-; FP32EL-LABEL: testlowercall11:
-; FP32EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP32EL-DAG: mfc1 $7, $f{{[0-9]+}}
-; FP64EL-LABEL: testlowercall11:
-; FP64EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP64EL-DAG: mfhc1 $7, $f{{[0-9]+}}
+; ALL-LABEL: testlowercall11:
+; ALL-DAG:       addiu $4, $zero, 52
+
+; NO-MFHC1-DAG:  mfc1 $6, $f{{[0-9]+}}
+; NO-MFHC1-DAG:  mfc1 $7, $f{{[0-9]+}}
+
+; HAS-MFHC1-DAG: mfc1 $6, $f{{[0-9]+}}
+; HAS-MFHC1-DAG: mfhc1 $7, $f{{[0-9]+}}
 define void @testlowercall11() nounwind {
 entry:
   tail call void @f11(i32 52, double 1.600000e+01) nounwind
@@ -170,11 +172,11 @@ entry:
 declare void @f11(i32, double)
 
 ; $f12, $f14, $6, $7
-; CHECK-LABEL: testlowercall12:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
-; CHECK-DAG: lui $6, 16672
-; CHECK-DAG: lui $7, 16808
+; ALL-LABEL: testlowercall12:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       lwc1 $f14, %lo
+; ALL-DAG:       lui $6, 16672
+; ALL-DAG:       lui $7, 16808
 define void @testlowercall12() nounwind {
 entry:
   tail call void @f12(float 2.800000e+01, float 1.900000e+01, float 1.000000e+01, float 2.100000e+01) nounwind
@@ -184,11 +186,11 @@ entry:
 declare void @f12(float, float, float, float)
 
 ; $f12, $5, $6, $7
-; CHECK-LABEL: testlowercall13:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: addiu $5, $zero, 83
-; CHECK-DAG: lui $6, 16800
-; CHECK-DAG: addiu $7, $zero, 25
+; ALL-LABEL: testlowercall13:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       addiu $5, $zero, 83
+; ALL-DAG:       lui $6, 16800
+; ALL-DAG:       addiu $7, $zero, 25
 define void @testlowercall13() nounwind {
 entry:
   tail call void @f13(float 3.800000e+01, i32 83, float 2.000000e+01, i32 25) nounwind
@@ -199,10 +201,10 @@ entry:
 declare void @f13(float, i32, float, i32)
 
 ; $f12, $f14, $7
-; CHECK-LABEL: testlowercall14:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
-; CHECK-DAG: lui $7, 16880
+; ALL-LABEL: testlowercall14:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       lwc1 $f14, %lo
+; ALL-DAG:       lui $7, 16880
 define void @testlowercall14() nounwind {
 entry:
   tail call void @f14(double 3.500000e+01, float 2.900000e+01, float 3.000000e+01) nounwind
@@ -212,15 +214,15 @@ entry:
 declare void @f14(double, float, float)
 
 ; $f12, $f14, ($6, $7)
-; CHECK-LABEL: testlowercall15:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
-; FP32EL-LABEL: testlowercall15:
-; FP32EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP32EL-DAG: mfc1 $7, $f{{[0-9]+}}
-; FP64EL-LABEL: testlowercall15:
-; FP64EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP64EL-DAG: mfhc1 $7, $f{{[0-9]+}}
+; ALL-LABEL: testlowercall15:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       lwc1 $f14, %lo
+
+; NO-MFHC1-DAG:  mfc1 $6, $f{{[0-9]+}}
+; NO-MFHC1-DAG:  mfc1 $7, $f{{[0-9]+}}
+
+; HAS-MFHC1-DAG: mfc1 $6, $f{{[0-9]+}}
+; HAS-MFHC1-DAG: mfhc1 $7, $f{{[0-9]+}}
 define void @testlowercall15() nounwind {
 entry:
   tail call void @f15(float 4.800000e+01, float 3.900000e+01, double 3.700000e+01) nounwind
@@ -230,11 +232,11 @@ entry:
 declare void @f15(float, float, double)
 
 ; $4, $5, $6, $7
-; CHECK-LABEL: testlowercall16:
-; CHECK-DAG: addiu $4, $zero, 62
-; CHECK-DAG: lui $5, 16964
-; CHECK-DAG: addiu $6, $zero, 64
-; CHECK-DAG: lui $7, 16888
+; ALL-LABEL: testlowercall16:
+; ALL-DAG:       addiu $4, $zero, 62
+; ALL-DAG:       lui $5, 16964
+; ALL-DAG:       addiu $6, $zero, 64
+; ALL-DAG:       lui $7, 16888
 define void @testlowercall16() nounwind {
 entry:
   tail call void @f16(i32 62, float 4.900000e+01, i32 64, float 3.100000e+01) nounwind
@@ -244,11 +246,11 @@ entry:
 declare void @f16(i32, float, i32, float)
 
 ; $4, $5, $6, $7
-; CHECK-LABEL: testlowercall17:
-; CHECK-DAG: addiu $4, $zero, 72
-; CHECK-DAG: lui $5, 17004
-; CHECK-DAG: addiu $6, $zero, 74
-; CHECK-DAG: addiu $7, $zero, 35
+; ALL-LABEL: testlowercall17:
+; ALL-DAG:       addiu $4, $zero, 72
+; ALL-DAG:       lui $5, 17004
+; ALL-DAG:       addiu $6, $zero, 74
+; ALL-DAG:       addiu $7, $zero, 35
 define void @testlowercall17() nounwind {
 entry:
   tail call void @f17(i32 72, float 5.900000e+01, i32 74, i32 35) nounwind
@@ -258,11 +260,11 @@ entry:
 declare void @f17(i32, float, i32, i32)
 
 ; $4, $5, $6, $7
-; CHECK-LABEL: testlowercall18:
-; CHECK-DAG: addiu $4, $zero, 82
-; CHECK-DAG: addiu $5, $zero, 93
-; CHECK-DAG: lui $6, 16928
-; CHECK-DAG: addiu $7, $zero, 45
+; ALL-LABEL: testlowercall18:
+; ALL-DAG:       addiu $4, $zero, 82
+; ALL-DAG:       addiu $5, $zero, 93
+; ALL-DAG:       lui $6, 16928
+; ALL-DAG:       addiu $7, $zero, 45
 define void @testlowercall18() nounwind {
 entry:
   tail call void @f18(i32 82, i32 93, float 4.000000e+01, i32 45) nounwind
@@ -273,16 +275,16 @@ declare void @f18(i32, i32, float, i32)
 
 
 ; $4, ($6, $7), stack
-; CHECK-LABEL: testlowercall20:
-; CHECK-DAG: addiu $4, $zero, 92
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 16($sp)
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 20($sp)
-; FP32EL-LABEL: testlowercall20:
-; FP32EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP32EL-DAG: mfc1 $7, $f{{[0-9]+}}
-; FP64EL-LABEL: testlowercall20:
-; FP64EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP64EL-DAG: mfhc1 $7, $f{{[0-9]+}}
+; ALL-LABEL: testlowercall20:
+; ALL-DAG:       addiu $4, $zero, 92
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 16($sp)
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 20($sp)
+
+; NO-MFHC1-DAG:  mfc1 $6, $f{{[0-9]+}}
+; NO-MFHC1-DAG:  mfc1 $7, $f{{[0-9]+}}
+
+; HAS-MFHC1-DAG: mfc1 $6, $f{{[0-9]+}}
+; HAS-MFHC1-DAG: mfhc1 $7, $f{{[0-9]+}}
 define void @testlowercall20() nounwind {
 entry:
   tail call void @f20(i32 92, double 2.600000e+01, double 4.700000e+01) nounwind
@@ -292,9 +294,9 @@ entry:
 declare void @f20(i32, double, double)
 
 ; $f12, $5
-; CHECK-LABEL: testlowercall21:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: addiu $5, $zero, 103
+; ALL-LABEL: testlowercall21:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       addiu $5, $zero, 103
 define void @testlowercall21() nounwind {
 entry:
   tail call void @f21(float 5.800000e+01, i32 103) nounwind
@@ -304,15 +306,15 @@ entry:
 declare void @f21(float, i32)
 
 ; $f12, $5, ($6, $7)
-; CHECK-LABEL: testlowercall22:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: addiu $5, $zero, 113
-; FP32EL-LABEL: testlowercall22:
-; FP32EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP32EL-DAG: mfc1 $7, $f{{[0-9]+}}
-; FP64EL-LABEL: testlowercall22:
-; FP64EL-DAG: mfc1 $6, $f{{[0-9]+}}
-; FP64EL-DAG: mfhc1 $7, $f{{[0-9]+}}
+; ALL-LABEL: testlowercall22:
+; ALL-DAG:       lwc1 $f12, %lo
+; ALL-DAG:       addiu $5, $zero, 113
+
+; NO-MFHC1-DAG:  mfc1 $6, $f{{[0-9]+}}
+; NO-MFHC1-DAG:  mfc1 $7, $f{{[0-9]+}}
+
+; HAS-MFHC1-DAG: mfc1 $6, $f{{[0-9]+}}
+; HAS-MFHC1-DAG: mfhc1 $7, $f{{[0-9]+}}
 define void @testlowercall22() nounwind {
 entry:
   tail call void @f22(float 6.800000e+01, i32 113, double 5.700000e+01) nounwind
@@ -322,9 +324,9 @@ entry:
 declare void @f22(float, i32, double)
 
 ; $f12, f6
-; CHECK-LABEL: testlowercall23:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: addiu $6, $zero, 123
+; ALL-LABEL: testlowercall23:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       addiu $6, $zero, 123
 define void @testlowercall23() nounwind {
 entry:
   tail call void @f23(double 4.500000e+01, i32 123) nounwind
@@ -334,11 +336,11 @@ entry:
 declare void @f23(double, i32)
 
 ; $f12,$6, stack
-; CHECK-LABEL: testlowercall24:
-; CHECK-DAG: ldc1 $f12, %lo
-; CHECK-DAG: addiu $6, $zero, 133
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 16($sp)
-; CHECK-DAG: sw ${{[a-z0-9]+}}, 20($sp)
+; ALL-LABEL: testlowercall24:
+; ALL-DAG:       ldc1 $f12, %lo
+; ALL-DAG:       addiu $6, $zero, 133
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 16($sp)
+; ALL-DAG:       sw ${{[a-z0-9]+}}, 20($sp)
 define void @testlowercall24() nounwind {
 entry:
   tail call void @f24(double 5.500000e+01, i32 133, double 6.700000e+01) nounwind
@@ -347,19 +349,19 @@ entry:
 
 declare void @f24(double, i32, double)
 
-; CHECK-LABEL: testlowercall25:
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: lwc1 $f14, %lo
-; CHECK-DAG: lui $6
-; CHECK-DAG: lui $7
-; CHECK-DAG: lwc1 $f12, %lo
-; CHECK-DAG: addiu $5, $zero, 83
-; CHECK-DAG: lui $6
-; CHECK-DAG: addiu $7, $zero, 25
-; CHECK-DAG: addiu $4, $zero, 82
-; CHECK-DAG: addiu $5, $zero, 93
-; CHECK-DAG: lui $6
-; CHECK-DAG: addiu $7, $zero, 45
+; ALL-LABEL: testlowercall25:
+; ALL-DAG:      lwc1 $f12, %lo
+; ALL-DAG:      lwc1 $f14, %lo
+; ALL-DAG:      lui $6
+; ALL-DAG:      lui $7
+; ALL-DAG:      lwc1 $f12, %lo
+; ALL-DAG:      addiu $5, $zero, 83
+; ALL-DAG:      lui $6
+; ALL-DAG:      addiu $7, $zero, 25
+; ALL-DAG:      addiu $4, $zero, 82
+; ALL-DAG:      addiu $5, $zero, 93
+; ALL-DAG:      lui $6
+; ALL-DAG:      addiu $7, $zero, 45
 define void @testlowercall25() nounwind {
 entry:
   tail call void @f12(float 2.800000e+01, float 1.900000e+01, float 1.000000e+01, float 2.100000e+01) nounwind
diff --git a/test/CodeGen/Mips/octeon_popcnt.ll b/test/CodeGen/Mips/octeon_popcnt.ll
index 52c37f6..3432b39 100644
--- a/test/CodeGen/Mips/octeon_popcnt.ll
+++ b/test/CodeGen/Mips/octeon_popcnt.ll
@@ -6,7 +6,7 @@ define i8 @cnt8(i8 %x) nounwind readnone {
   ret i8 %cnt
 ; OCTEON-LABEL: cnt8:
 ; OCTEON: jr   $ra
-; OCTEON: pop  $2, $1
+; OCTEON: pop  $2, [[R1:\$[0-9]+]]
 ; MIPS64-LABEL: cnt8:
 ; MIPS64-NOT: pop
 }
@@ -16,12 +16,12 @@ define i16 @cnt16(i16 %x) nounwind readnone {
   ret i16 %cnt
 ; OCTEON-LABEL: cnt16:
 ; OCTEON: jr   $ra
-; OCTEON: pop  $2, $1
+; OCTEON: pop  $2, [[R1:\$[0-9]+]]
 ; MIPS64-LABEL: cnt16:
 ; MIPS64-NOT: pop
 }
 
-define i32 @cnt32(i32 %x) nounwind readnone {
+define i32 @cnt32(i32 zeroext %x) nounwind readnone {
   %cnt = tail call i32 @llvm.ctpop.i32(i32 %x)
   ret i32 %cnt
 ; OCTEON-LABEL: cnt32:
diff --git a/test/CodeGen/Mips/prevent-hoisting.ll b/test/CodeGen/Mips/prevent-hoisting.ll
index da665c2..210fe3b 100644
--- a/test/CodeGen/Mips/prevent-hoisting.ll
+++ b/test/CodeGen/Mips/prevent-hoisting.ll
@@ -10,16 +10,19 @@
 
 ; CHECK-LABEL: readLumaCoeff8x8_CABAC
 
-; The check for "addiu" instruction is added so that we can match the correct "b" instruction.
+; The check for first "addiu" instruction is added so that we can match the correct "b" instruction.
 ; CHECK:           addiu ${{[0-9]+}}, $zero, -1
 ; CHECK:           b $[[BB0:BB[0-9_]+]]
+; CHECK-NEXT:      addiu ${{[0-9]+}}, $zero, 0
 
-; Check that sll instruction that writes to $1 starts basic block.
-; CHECK:       {{BB[0-9_#]+}}: 
+; Check that at the start of a fallthrough block there is a instruction that writes to $1.
+; CHECK-NEXT:  {{BB[0-9_#]+}}: 
+; CHECK-NEXT:      lw      $[[R1:[0-9]+]], %got(assignSE2partition)($[[R2:[0-9]+]])
 ; CHECK-NEXT:      sll $1, $[[R0:[0-9]+]], 4
 
-; Check that identical sll instruction starts another basic block.
+; Check that identical instructions are at the start of a target block.
 ; CHECK:       [[BB0]]:
+; CHECK-NEXT:      lw      $[[R1]], %got(assignSE2partition)($[[R2]])
 ; CHECK-NEXT:      sll $1, $[[R0]], 4
 
 
diff --git a/test/CodeGen/Mips/select.ll b/test/CodeGen/Mips/select.ll
index eb2198b..d6e1826 100644
--- a/test/CodeGen/Mips/select.ll
+++ b/test/CodeGen/Mips/select.ll
@@ -8,7 +8,7 @@
 @d2 = external global double
 @d3 = external global double
 
-define i32 @i32_icmp_ne_i32_val(i32 %s, i32 %f0, i32 %f1) nounwind readnone {
+define i32 @i32_icmp_ne_i32_val(i32 signext %s, i32 signext %f0, i32 signext %f1) nounwind readnone {
 entry:
 ; ALL-LABEL: i32_icmp_ne_i32_val:
 
@@ -37,7 +37,7 @@ entry:
   ret i32 %cond
 }
 
-define i64 @i32_icmp_ne_i64_val(i32 %s, i64 %f0, i64 %f1) nounwind readnone {
+define i64 @i32_icmp_ne_i64_val(i32 signext %s, i64 %f0, i64 %f1) nounwind readnone {
 entry:
 ; ALL-LABEL: i32_icmp_ne_i64_val:
 
@@ -128,7 +128,7 @@ entry:
   ret i64 %cond
 }
 
-define float @i32_icmp_ne_f32_val(i32 %s, float %f0, float %f1) nounwind readnone {
+define float @i32_icmp_ne_f32_val(i32 signext %s, float %f0, float %f1) nounwind readnone {
 entry:
 ; ALL-LABEL: i32_icmp_ne_f32_val:
 
@@ -161,7 +161,7 @@ entry:
   ret float %cond
 }
 
-define double @i32_icmp_ne_f64_val(i32 %s, double %f0, double %f1) nounwind readnone {
+define double @i32_icmp_ne_f64_val(i32 signext %s, double %f0, double %f1) nounwind readnone {
 entry:
 ; ALL-LABEL: i32_icmp_ne_f64_val:
 
@@ -496,7 +496,7 @@ entry:
   ret float %cond
 }
 
-define i32 @f32_fcmp_oeq_i32_val(i32 %f0, i32 %f1, float %f2, float %f3) nounwind readnone {
+define i32 @f32_fcmp_oeq_i32_val(i32 signext %f0, i32 signext %f1, float %f2, float %f3) nounwind readnone {
 entry:
 ; ALL-LABEL: f32_fcmp_oeq_i32_val:
 
@@ -541,7 +541,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @f32_fcmp_olt_i32_val(i32 %f0, i32 %f1, float %f2, float %f3) nounwind readnone {
+define i32 @f32_fcmp_olt_i32_val(i32 signext %f0, i32 signext %f1, float %f2, float %f3) nounwind readnone {
 entry:
 ; ALL-LABEL: f32_fcmp_olt_i32_val:
 
@@ -585,7 +585,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @f32_fcmp_ogt_i32_val(i32 %f0, i32 %f1, float %f2, float %f3) nounwind readnone {
+define i32 @f32_fcmp_ogt_i32_val(i32 signext %f0, i32 signext %f1, float %f2, float %f3) nounwind readnone {
 entry:
 ; ALL-LABEL: f32_fcmp_ogt_i32_val:
 
@@ -630,7 +630,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @f64_fcmp_oeq_i32_val(i32 %f0, i32 %f1) nounwind readonly {
+define i32 @f64_fcmp_oeq_i32_val(i32 signext %f0, i32 signext %f1) nounwind readonly {
 entry:
 ; ALL-LABEL: f64_fcmp_oeq_i32_val:
 
@@ -707,7 +707,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @f64_fcmp_olt_i32_val(i32 %f0, i32 %f1) nounwind readonly {
+define i32 @f64_fcmp_olt_i32_val(i32 signext %f0, i32 signext %f1) nounwind readonly {
 entry:
 ; ALL-LABEL: f64_fcmp_olt_i32_val:
 
@@ -784,7 +784,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @f64_fcmp_ogt_i32_val(i32 %f0, i32 %f1) nounwind readonly {
+define i32 @f64_fcmp_ogt_i32_val(i32 signext %f0, i32 signext %f1) nounwind readonly {
 entry:
 ; ALL-LABEL: f64_fcmp_ogt_i32_val:
 
diff --git a/test/CodeGen/Mips/seleq.ll b/test/CodeGen/Mips/seleq.ll
index 190baad..9af422f 100644
--- a/test/CodeGen/Mips/seleq.ll
+++ b/test/CodeGen/Mips/seleq.ll
@@ -10,7 +10,7 @@
 @z3 = common global i32 0, align 4
 @z4 = common global i32 0, align 4
 
-define void @calc_seleq() nounwind "target-cpu"="mips32" "target-features"="+o32,+mips32" {
+define void @calc_seleq() nounwind {
 entry:
   %0 = load i32* @a, align 4
   %1 = load i32* @b, align 4
diff --git a/test/CodeGen/Mips/small-section-reserve-gp.ll b/test/CodeGen/Mips/small-section-reserve-gp.ll
index 03503fb..cbf0681 100644
--- a/test/CodeGen/Mips/small-section-reserve-gp.ll
+++ b/test/CodeGen/Mips/small-section-reserve-gp.ll
@@ -1,4 +1,4 @@
-; RUN: llc -mtriple=mipsel-sde-elf -march=mipsel -relocation-model=static < %s \
+; RUN: llc -mtriple=mipsel-sde-elf -march=mipsel -relocation-model=static -mattr=+noabicalls -mgpopt < %s \
 ; RUN: | FileCheck %s
 
 @i = internal unnamed_addr global i32 0, align 4
diff --git a/test/CodeGen/Mips/start-asm-file.ll b/test/CodeGen/Mips/start-asm-file.ll
index 8872464..9dc501c 100644
--- a/test/CodeGen/Mips/start-asm-file.ll
+++ b/test/CodeGen/Mips/start-asm-file.ll
@@ -1,7 +1,4 @@
 ; Check the emission of directives at the start of an asm file.
-; This test is XFAILED until we fix the emission of '.option pic0' on
-; N32. At the moment we check if subtarget is Mips64 when we should be
-; checking the Subtarget's ABI.
 
 ; ### O32 ABI ###
 ; RUN: llc -filetype=asm -mtriple mips-unknown-linux -mcpu=mips32 \
diff --git a/test/CodeGen/Mips/zeroreg.ll b/test/CodeGen/Mips/zeroreg.ll
index a1b6cb0..c766d3b 100644
--- a/test/CodeGen/Mips/zeroreg.ll
+++ b/test/CodeGen/Mips/zeroreg.ll
@@ -8,7 +8,7 @@
 
 @g1 = external global i32
 
-define i32 @sel_icmp_nez_i32_z0(i32 %s) nounwind readonly {
+define i32 @sel_icmp_nez_i32_z0(i32 signext %s) nounwind readonly {
 entry:
 ; ALL-LABEL: sel_icmp_nez_i32_z0:
 
@@ -30,7 +30,7 @@ entry:
   ret i32 %cond
 }
 
-define i32 @sel_icmp_nez_i32_z1(i32 %s) nounwind readonly {
+define i32 @sel_icmp_nez_i32_z1(i32 signext %s) nounwind readonly {
 entry:
 ; ALL-LABEL: sel_icmp_nez_i32_z1:
author	Stephen Hines <srhines@google.com>	2014-12-01 14:51:49 -0800
committer	Stephen Hines <srhines@google.com>	2014-12-02 16:08:10 -0800
commit	37ed9c199ca639565f6ce88105f9e39e898d82d0 (patch)
tree	8fb36d3910e3ee4c4e1b7422f4f017108efc52f5 /test/CodeGen/Mips
parent	d2327b22152ced7bc46dc629fc908959e8a52d03 (diff)
download	external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.zip external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.tar.gz external_llvm-37ed9c199ca639565f6ce88105f9e39e898d82d0.tar.bz2