[bpf] add big- and host- endian support

Summary:
-march=bpf    -> host endian
-march=bpf_le -> little endian
-match=bpf_be -> big endian

Test Plan:
v1 was tested by IBM s390 guys and appears to be working there.
It bit rots too fast here.

Reviewers: chandlerc, tstellarAMD

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D10177

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@239071 91177308-0d34-0410-b5e6-96231b3b80d8

14 files changed, 109 insertions, 18 deletions

diff --git a/test/CodeGen/BPF/alu8.ll b/test/CodeGen/BPF/alu8.ll
index 0233225f81b..21083b1aaf8 100644
--- a/test/CodeGen/BPF/alu8.ll
+++ b/test/CodeGen/BPF/alu8.ll
@@ -1,5 +1,4 @@
-; RUN: llc -march=bpf -show-mc-encoding < %s | FileCheck %s
-; test little endian only for now
+; RUN: llc -march=bpf_le -show-mc-encoding < %s | FileCheck %s
 
 define i8 @mov(i8 %a, i8 %b) nounwind {
 ; CHECK-LABEL: mov:
diff --git a/test/CodeGen/BPF/atomics.ll b/test/CodeGen/BPF/atomics.ll
index 2f9730dddde..b6cb65f3aea 100644
--- a/test/CodeGen/BPF/atomics.ll
+++ b/test/CodeGen/BPF/atomics.ll
@@ -1,5 +1,4 @@
-; RUN: llc < %s -march=bpf -verify-machineinstrs -show-mc-encoding | FileCheck %s
-; test little endian only for now
+; RUN: llc < %s -march=bpf_le -verify-machineinstrs -show-mc-encoding | FileCheck %s
 
 ; CHECK-LABEL: test_load_add_32
 ; CHECK: xadd32
diff --git a/test/CodeGen/BPF/basictest.ll b/test/CodeGen/BPF/basictest.ll
index 2a2d49878a6..b1c1f808160 100644
--- a/test/CodeGen/BPF/basictest.ll
+++ b/test/CodeGen/BPF/basictest.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 define i32 @test0(i32 %X) {
   %tmp.1 = add i32 %X, 1
diff --git a/test/CodeGen/BPF/cc_args.ll b/test/CodeGen/BPF/cc_args.ll
index 5085fe5684e..42ac830c1ba 100644
--- a/test/CodeGen/BPF/cc_args.ll
+++ b/test/CodeGen/BPF/cc_args.ll
@@ -1,5 +1,4 @@
-; RUN: llc < %s -march=bpf -show-mc-encoding | FileCheck %s
-; test little endian only for now
+; RUN: llc < %s -march=bpf_le -show-mc-encoding | FileCheck %s
 
 define void @test() #0 {
 entry:
diff --git a/test/CodeGen/BPF/cc_args_be.ll b/test/CodeGen/BPF/cc_args_be.ll
new file mode 100644
index 00000000000..d5a6539aa5e
--- /dev/null
+++ b/test/CodeGen/BPF/cc_args_be.ll
@@ -0,0 +1,96 @@
+; RUN: llc < %s -march=bpf_be -show-mc-encoding | FileCheck %s
+; test big endian
+
+define void @test() #0 {
+entry:
+; CHECK: test:
+
+; CHECK: mov  r1, 123 # encoding: [0xb7,0x10,0x00,0x00,0x00,0x00,0x00,0x7b]
+; CHECK: call f_i16
+  call void @f_i16(i16 123)
+
+; CHECK: mov  r1, 12345678 # encoding: [0xb7,0x10,0x00,0x00,0x00,0xbc,0x61,0x4e]
+; CHECK: call f_i32
+  call void @f_i32(i32 12345678)
+
+; CHECK: ld_64 r1, 72623859790382856 # encoding: [0x18,0x10,0x00,0x00,0x05,0x06,0x07,0x08,0x00,0x00,0x00,0x00,0x01,0x02,0x03,0x04]
+; CHECK: call f_i64
+  call void @f_i64(i64 72623859790382856)
+
+; CHECK: mov  r1, 1234
+; CHECK: mov  r2, 5678
+; CHECK: call f_i32_i32
+  call void @f_i32_i32(i32 1234, i32 5678)
+
+; CHECK: mov  r1, 2
+; CHECK: mov  r2, 3
+; CHECK: mov  r3, 4
+; CHECK: call f_i16_i32_i16
+  call void @f_i16_i32_i16(i16 2, i32 3, i16 4)
+
+; CHECK: mov  r1, 5
+; CHECK: ld_64 r2, 7262385979038285
+; CHECK: mov  r3, 6
+; CHECK: call f_i16_i64_i16
+  call void @f_i16_i64_i16(i16 5, i64 7262385979038285, i16 6)
+
+  ret void
+}
+
+@g_i16 = common global i16 0, align 2
+@g_i32 = common global i32 0, align 2
+@g_i64 = common global i64 0, align 4
+
+define void @f_i16(i16 %a) #0 {
+; CHECK: f_i16:
+; CHECK: sth 0(r2), r1 # encoding: [0x6b,0x21,0x00,0x00,0x00,0x00,0x00,0x00]
+  store volatile i16 %a, i16* @g_i16, align 2
+  ret void
+}
+
+define void @f_i32(i32 %a) #0 {
+; CHECK: f_i32:
+; CHECK: sth 2(r2), r1 # encoding: [0x6b,0x21,0x00,0x02,0x00,0x00,0x00,0x00]
+; CHECK: sth 0(r2), r1 # encoding: [0x6b,0x21,0x00,0x00,0x00,0x00,0x00,0x00]
+  store volatile i32 %a, i32* @g_i32, align 2
+  ret void
+}
+
+define void @f_i64(i64 %a) #0 {
+; CHECK: f_i64:
+; CHECK: stw 4(r2), r1 # encoding: [0x63,0x21,0x00,0x04,0x00,0x00,0x00,0x00]
+; CHECK: stw 0(r2), r1
+  store volatile i64 %a, i64* @g_i64, align 2
+  ret void
+}
+
+define void @f_i32_i32(i32 %a, i32 %b) #0 {
+; CHECK: f_i32_i32:
+; CHECK: stw 0(r3), r1
+  store volatile i32 %a, i32* @g_i32, align 4
+; CHECK: stw 0(r3), r2
+  store volatile i32 %b, i32* @g_i32, align 4
+  ret void
+}
+
+define void @f_i16_i32_i16(i16 %a, i32 %b, i16 %c) #0 {
+; CHECK: f_i16_i32_i16:
+; CHECK: sth 0(r4), r1
+  store volatile i16 %a, i16* @g_i16, align 2
+; CHECK: stw 0(r1), r2
+  store volatile i32 %b, i32* @g_i32, align 4
+; CHECK: sth 0(r4), r3
+  store volatile i16 %c, i16* @g_i16, align 2
+  ret void
+}
+
+define void @f_i16_i64_i16(i16 %a, i64 %b, i16 %c) #0 {
+; CHECK: f_i16_i64_i16:
+; CHECK: sth 0(r4), r1
+  store volatile i16 %a, i16* @g_i16, align 2
+; CHECK: std 0(r1), r2 # encoding: [0x7b,0x12,0x00,0x00,0x00,0x00,0x00,0x00]
+  store volatile i64 %b, i64* @g_i64, align 8
+; CHECK: sth 0(r4), r3
+  store volatile i16 %c, i16* @g_i16, align 2
+  ret void
+}
diff --git a/test/CodeGen/BPF/cc_ret.ll b/test/CodeGen/BPF/cc_ret.ll
index e32b17bcc61..ea23bdebd1d 100644
--- a/test/CodeGen/BPF/cc_ret.ll
+++ b/test/CodeGen/BPF/cc_ret.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 define void @test() #0 {
 entry:
diff --git a/test/CodeGen/BPF/ex1.ll b/test/CodeGen/BPF/ex1.ll
index be038e9a3d8..abf068ca9a3 100644
--- a/test/CodeGen/BPF/ex1.ll
+++ b/test/CodeGen/BPF/ex1.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 %struct.bpf_context = type { i64, i64, i64, i64, i64, i64, i64 }
 %struct.sk_buff = type { i64, i64, i64, i64, i64, i64, i64 }
diff --git a/test/CodeGen/BPF/intrinsics.ll b/test/CodeGen/BPF/intrinsics.ll
index 98b57deb7c8..586269751de 100644
--- a/test/CodeGen/BPF/intrinsics.ll
+++ b/test/CodeGen/BPF/intrinsics.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf -show-mc-encoding | FileCheck %s
+; RUN: llc < %s -march=bpf_le -show-mc-encoding | FileCheck %s
 
 ; Function Attrs: nounwind uwtable
 define i32 @ld_b(i64 %foo, i64* nocapture %bar, i8* %ctx, i8* %ctx2) #0 {
diff --git a/test/CodeGen/BPF/load.ll b/test/CodeGen/BPF/load.ll
index 03fb17c965b..9787379c281 100644
--- a/test/CodeGen/BPF/load.ll
+++ b/test/CodeGen/BPF/load.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 define i16 @am1(i16* %a) nounwind {
   %1 = load i16, i16* %a
diff --git a/test/CodeGen/BPF/loops.ll b/test/CodeGen/BPF/loops.ll
index 4798d78842c..7a4d5dbcb34 100644
--- a/test/CodeGen/BPF/loops.ll
+++ b/test/CodeGen/BPF/loops.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 define zeroext i16 @add(i16* nocapture %a, i16 zeroext %n) nounwind readonly {
 entry:
diff --git a/test/CodeGen/BPF/sanity.ll b/test/CodeGen/BPF/sanity.ll
index 09a6b65d085..b45f5a1186e 100644
--- a/test/CodeGen/BPF/sanity.ll
+++ b/test/CodeGen/BPF/sanity.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=bpf | FileCheck %s
+; RUN: llc < %s -march=bpf_le | FileCheck %s
 
 @foo_printf.fmt = private unnamed_addr constant [9 x i8] c"hello  \0A\00", align 1
 
diff --git a/test/CodeGen/BPF/setcc.ll b/test/CodeGen/BPF/setcc.ll
index eabb6c9bf2d..6198197d244 100644
--- a/test/CodeGen/BPF/setcc.ll
+++ b/test/CodeGen/BPF/setcc.ll
@@ -1,4 +1,4 @@
-; RUN: llc -march=bpf < %s | FileCheck %s
+; RUN: llc -march=bpf_le < %s | FileCheck %s
 
 define i16 @sccweqand(i16 %a, i16 %b) nounwind {
   %t1 = and i16 %a, %b
diff --git a/test/CodeGen/BPF/shifts.ll b/test/CodeGen/BPF/shifts.ll
index 898ae2d4612..77d02d76c05 100644
--- a/test/CodeGen/BPF/shifts.ll
+++ b/test/CodeGen/BPF/shifts.ll
@@ -1,5 +1,4 @@
-; RUN: llc < %s -march=bpf -show-mc-encoding | FileCheck %s
-; test little endian only for now
+; RUN: llc < %s -march=bpf_le -show-mc-encoding | FileCheck %s
 
 define zeroext i8 @lshr8(i8 zeroext %a, i8 zeroext %cnt) nounwind readnone {
 entry:
diff --git a/test/CodeGen/BPF/sockex2.ll b/test/CodeGen/BPF/sockex2.ll
index 6ae5e1c8d6b..99458a965a9 100644
--- a/test/CodeGen/BPF/sockex2.ll
+++ b/test/CodeGen/BPF/sockex2.ll
@@ -1,5 +1,4 @@
-; RUN: llc < %s -march=bpf -show-mc-encoding | FileCheck %s
-; test little endian only for now
+; RUN: llc < %s -march=bpf_le -show-mc-encoding | FileCheck %s
 
 %struct.bpf_map_def = type { i32, i32, i32, i32 }
 %struct.sk_buff = type opaque