Move the segmented stack switch to a function attribute

This removes the -segmented-stacks command line flag in favor of a
per-function "split-stack" attribute.

Patch by Luqman Aden and Alex Crichton!

llvm-svn: 205997
diff --git a/llvm/test/CodeGen/ARM/debug-segmented-stacks.ll b/llvm/test/CodeGen/ARM/debug-segmented-stacks.ll
index b0dc467..e866b4e 100644
--- a/llvm/test/CodeGen/ARM/debug-segmented-stacks.ll
+++ b/llvm/test/CodeGen/ARM/debug-segmented-stacks.ll
@@ -1,11 +1,11 @@
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -verify-machineinstrs -filetype=asm | FileCheck %s -check-prefix=ARM-linux
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -verify-machineinstrs -filetype=asm | FileCheck %s -check-prefix=ARM-linux
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -filetype=obj
 
 !llvm.dbg.cu = !{!0}
 !llvm.module.flags = !{!9, !10}
 !llvm.ident = !{!11}
 
-define void @test_basic() {
+define void @test_basic() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
 	ret void
@@ -78,3 +78,5 @@
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/ARM/segmented-stacks-dynamic.ll b/llvm/test/CodeGen/ARM/segmented-stacks-dynamic.ll
index 13b5bcf..77c62de 100644
--- a/llvm/test/CodeGen/ARM/segmented-stacks-dynamic.ll
+++ b/llvm/test/CodeGen/ARM/segmented-stacks-dynamic.ll
@@ -1,12 +1,12 @@
-; RUN: llc < %s -mtriple=arm-linux-androideabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=ARM-android
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=ARM-linux
-; RUN: llc < %s -mtriple=arm-linux-androideabi -segmented-stacks -filetype=obj
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=arm-linux-androideabi -verify-machineinstrs | FileCheck %s -check-prefix=ARM-android
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -verify-machineinstrs | FileCheck %s -check-prefix=ARM-linux
+; RUN: llc < %s -mtriple=arm-linux-androideabi -filetype=obj
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -filetype=obj
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define i32 @test_basic(i32 %l) {
+define i32 @test_basic(i32 %l) #0 {
         %mem = alloca i32, i32 %l
         call void @dummy_use (i32* %mem, i32 %l)
         %terminate = icmp eq i32 %l, 0
@@ -60,3 +60,5 @@
 ; ARM-android:      pop     {r4, r5}
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/ARM/segmented-stacks.ll b/llvm/test/CodeGen/ARM/segmented-stacks.ll
index 5eff633..a7804b9 100644
--- a/llvm/test/CodeGen/ARM/segmented-stacks.ll
+++ b/llvm/test/CodeGen/ARM/segmented-stacks.ll
@@ -1,15 +1,15 @@
-; RUN: llc < %s -mtriple=arm-linux-androideabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=ARM-android
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=ARM-linux
+; RUN: llc < %s -mtriple=arm-linux-androideabi -verify-machineinstrs | FileCheck %s -check-prefix=ARM-android
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -verify-machineinstrs | FileCheck %s -check-prefix=ARM-linux
 
 ; We used to crash with filetype=obj
-; RUN: llc < %s -mtriple=arm-linux-androideabi -segmented-stacks -filetype=obj
-; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=arm-linux-androideabi -filetype=obj
+; RUN: llc < %s -mtriple=arm-linux-unknown-gnueabi -filetype=obj
 
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define void @test_basic() {
+define void @test_basic() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
 	ret void
@@ -54,7 +54,7 @@
 
 }
 
-define i32 @test_nested(i32 * nest %closure, i32 %other) {
+define i32 @test_nested(i32 * nest %closure, i32 %other) #0 {
        %addend = load i32 * %closure
        %result = add i32 %other, %addend
        ret i32 %result
@@ -99,7 +99,7 @@
 
 }
 
-define void @test_large() {
+define void @test_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -144,7 +144,7 @@
 
 }
 
-define fastcc void @test_fastcc() {
+define fastcc void @test_fastcc() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
         ret void
@@ -189,7 +189,7 @@
 
 }
 
-define fastcc void @test_fastcc_large() {
+define fastcc void @test_fastcc_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -233,3 +233,5 @@
 ; ARM-android:      pop     {r4, r5}
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/Thumb/segmented-stacks-dynamic.ll b/llvm/test/CodeGen/Thumb/segmented-stacks-dynamic.ll
index 067c07b..5d51f40 100644
--- a/llvm/test/CodeGen/Thumb/segmented-stacks-dynamic.ll
+++ b/llvm/test/CodeGen/Thumb/segmented-stacks-dynamic.ll
@@ -1,12 +1,12 @@
-; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-linux
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
-; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -segmented-stacks -filetype=obj
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-linux
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
+; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -filetype=obj
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -filetype=obj
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define i32 @test_basic(i32 %l) {
+define i32 @test_basic(i32 %l) #0 {
         %mem = alloca i32, i32 %l
         call void @dummy_use (i32* %mem, i32 %l)
         %terminate = icmp eq i32 %l, 0
@@ -61,3 +61,5 @@
 ; Thumb-android:      pop {r4, r5}
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/Thumb/segmented-stacks.ll b/llvm/test/CodeGen/Thumb/segmented-stacks.ll
index 5649b00..89043ec 100644
--- a/llvm/test/CodeGen/Thumb/segmented-stacks.ll
+++ b/llvm/test/CodeGen/Thumb/segmented-stacks.ll
@@ -1,13 +1,13 @@
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
-; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-linux
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -segmented-stacks -filetype=obj
-; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
+; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-linux
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -filetype=obj
+; RUN: llc < %s -mtriple=thumb-linux-unknown-gnueabi -filetype=obj
 
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define void @test_basic() {
+define void @test_basic() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
 	ret void
@@ -54,7 +54,7 @@
 
 }
 
-define i32 @test_nested(i32 * nest %closure, i32 %other) {
+define i32 @test_nested(i32 * nest %closure, i32 %other) #0 {
        %addend = load i32 * %closure
        %result = add i32 %other, %addend
        ret i32 %result
@@ -101,7 +101,7 @@
 
 }
 
-define void @test_large() {
+define void @test_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -150,7 +150,7 @@
 
 }
 
-define fastcc void @test_fastcc() {
+define fastcc void @test_fastcc() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
         ret void
@@ -197,7 +197,7 @@
 
 }
 
-define fastcc void @test_fastcc_large() {
+define fastcc void @test_fastcc_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -245,3 +245,5 @@
 ; Thumb-linux:      pop     {r4, r5}
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/Thumb2/segmented-stacks.ll b/llvm/test/CodeGen/Thumb2/segmented-stacks.ll
index 602fc84..38bf915 100644
--- a/llvm/test/CodeGen/Thumb2/segmented-stacks.ll
+++ b/llvm/test/CodeGen/Thumb2/segmented-stacks.ll
@@ -1,11 +1,11 @@
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -march=thumb -mcpu=arm1156t2-s -mattr=+thumb2 -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
-; RUN: llc < %s -mtriple=thumb-linux-androideabi -march=thumb -mcpu=arm1156t2-s -mattr=+thumb2 -segmented-stacks -filetype=obj
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -march=thumb -mcpu=arm1156t2-s -mattr=+thumb2 -verify-machineinstrs | FileCheck %s -check-prefix=Thumb-android
+; RUN: llc < %s -mtriple=thumb-linux-androideabi -march=thumb -mcpu=arm1156t2-s -mattr=+thumb2 -filetype=obj
 
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define void @test_basic() {
+define void @test_basic() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
 	ret void
@@ -30,3 +30,5 @@
 ; Thumb-android:      pop     {r4, r5}
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/X86/segmented-stacks-dynamic.ll b/llvm/test/CodeGen/X86/segmented-stacks-dynamic.ll
index e170762..b82be41 100644
--- a/llvm/test/CodeGen/X86/segmented-stacks-dynamic.ll
+++ b/llvm/test/CodeGen/X86/segmented-stacks-dynamic.ll
@@ -1,12 +1,12 @@
-; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X32
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux  -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X64
-; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux -segmented-stacks -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -verify-machineinstrs | FileCheck %s -check-prefix=X32
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux -verify-machineinstrs | FileCheck %s -check-prefix=X64
+; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux -filetype=obj
 
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define i32 @test_basic(i32 %l) {
+define i32 @test_basic(i32 %l) #0 {
         %mem = alloca i32, i32 %l
         call void @dummy_use (i32* %mem, i32 %l)
         %terminate = icmp eq i32 %l, 0
@@ -62,3 +62,5 @@
 ; X64:      movq %rax, %rdi
 
 }
+
+attributes #0 = { "split-stack" }
diff --git a/llvm/test/CodeGen/X86/segmented-stacks.ll b/llvm/test/CodeGen/X86/segmented-stacks.ll
index c02152b..8089f20 100644
--- a/llvm/test/CodeGen/X86/segmented-stacks.ll
+++ b/llvm/test/CodeGen/X86/segmented-stacks.ll
@@ -1,23 +1,23 @@
-; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X32-Linux
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux  -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X64-Linux
-; RUN: llc < %s -mcpu=generic -mtriple=i686-darwin -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X32-Darwin
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-darwin -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X64-Darwin
-; RUN: llc < %s -mcpu=generic -mtriple=i686-mingw32 -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X32-MinGW
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-freebsd -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X64-FreeBSD
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-mingw32 -segmented-stacks -verify-machineinstrs | FileCheck %s -check-prefix=X64-MinGW
+; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -verify-machineinstrs | FileCheck %s -check-prefix=X32-Linux
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux  -verify-machineinstrs | FileCheck %s -check-prefix=X64-Linux
+; RUN: llc < %s -mcpu=generic -mtriple=i686-darwin -verify-machineinstrs | FileCheck %s -check-prefix=X32-Darwin
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-darwin -verify-machineinstrs | FileCheck %s -check-prefix=X64-Darwin
+; RUN: llc < %s -mcpu=generic -mtriple=i686-mingw32 -verify-machineinstrs | FileCheck %s -check-prefix=X32-MinGW
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-freebsd -verify-machineinstrs | FileCheck %s -check-prefix=X64-FreeBSD
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-mingw32 -verify-machineinstrs | FileCheck %s -check-prefix=X64-MinGW
 
 ; We used to crash with filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=i686-darwin -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-darwin -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=i686-mingw32 -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-freebsd -segmented-stacks -filetype=obj
-; RUN: llc < %s -mcpu=generic -mtriple=x86_64-mingw32 -segmented-stacks -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=i686-linux -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-linux -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=i686-darwin -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-darwin -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=i686-mingw32 -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-freebsd -filetype=obj
+; RUN: llc < %s -mcpu=generic -mtriple=x86_64-mingw32 -filetype=obj
 
-; RUN: not llc < %s -mcpu=generic -mtriple=x86_64-solaris -segmented-stacks 2> %t.log
+; RUN: not llc < %s -mcpu=generic -mtriple=x86_64-solaris 2> %t.log
 ; RUN: FileCheck %s -input-file=%t.log -check-prefix=X64-Solaris
-; RUN: not llc < %s -mcpu=generic -mtriple=i686-freebsd -segmented-stacks 2> %t.log
+; RUN: not llc < %s -mcpu=generic -mtriple=i686-freebsd 2> %t.log
 ; RUN: FileCheck %s -input-file=%t.log -check-prefix=X32-FreeBSD
 
 ; X64-Solaris: Segmented stacks not supported on this platform
@@ -26,7 +26,7 @@
 ; Just to prevent the alloca from being optimized away
 declare void @dummy_use(i32*, i32)
 
-define void @test_basic() {
+define void @test_basic() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
 	ret void
@@ -104,7 +104,7 @@
 
 }
 
-define i32 @test_nested(i32 * nest %closure, i32 %other) {
+define i32 @test_nested(i32 * nest %closure, i32 %other) #0 {
        %addend = load i32 * %closure
        %result = add i32 %other, %addend
        ret i32 %result
@@ -177,7 +177,7 @@
 
 }
 
-define void @test_large() {
+define void @test_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -249,7 +249,7 @@
 
 }
 
-define fastcc void @test_fastcc() {
+define fastcc void @test_fastcc() #0 {
         %mem = alloca i32, i32 10
         call void @dummy_use (i32* %mem, i32 10)
         ret void
@@ -327,7 +327,7 @@
 
 }
 
-define fastcc void @test_fastcc_large() {
+define fastcc void @test_fastcc_large() #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 0)
         ret void
@@ -412,7 +412,7 @@
 
 }
 
-define fastcc void @test_fastcc_large_with_ecx_arg(i32 %a) {
+define fastcc void @test_fastcc_large_with_ecx_arg(i32 %a) #0 {
         %mem = alloca i32, i32 10000
         call void @dummy_use (i32* %mem, i32 %a)
         ret void
@@ -434,3 +434,5 @@
 ; X32-Darwin-NEXT: ret
 
 }
+
+attributes #0 = { "split-stack" }