[AArch64] Use LL for 64-bit arguments

[samparker]

Summary

The ACLE states that 64-bit crc32, wsr, rsr and rbit operands are uint64_t so we should have the clang builtin match this description - which is what we already do for AArch32.

[rnk]

include/clang/Basic/BuiltinsAArch64.def

54 ↗

(On Diff #182268)

If we can't change the signature on Linux (I don't see any reason why couldn't, though) we can use the 'W' builtin code to indicate that it's long on LP64 and long long on LLP64.

[efriedma]

While you're here, can you also fix __builtin_arm_rbit64, __builtin_arm_rsr64, and __builtin_arm_wsr64?

include/clang/Basic/BuiltinsAArch64.def

54 ↗

(On Diff #182268)

Changing the signature would be fine, I think (we don't expect people to use these directly, and even if they did it wouldn't make any practical difference outside of obscure edge cases). But probably better to match the ACLE signature for the sake of clarity.

[samparker]

Updated wsr, rsr and rbit

[efriedma]

LGTM

test/CodeGen/builtins-arm64.c

2 ↗

(On Diff #182481)

"-S" is redundant.

[phosek]

This broke our kernel build which uses PRIx64 to print the return value of __builtin_arm_rsr64. This is correct because according to ACLE, the return type of that builtin should be uint64_t. Problem is that on AArch64, unsigned long is equivalent to uint64_t, not unsigned long long (at least on Linux and other platforms like Fuchsia).

The reason this wasn't caught by the test is because that test isn't strict enough. I've updated the test in D57210. I believe this change should be reverted because it's wrong and I created revert in D57209. I think the correct solution might be what rnk suggested in his comment, that is to use W instead of LL. I plan on landing my revert if I don't hear back by EOD because this is currently blocking our kernel build with the latest Clang.

[samparker]

Changed the builtins to use W instead of LL. I've also updated the tests, adding a test for rbitl.

[efriedma]

LGTM

cfe/trunk/include/clang/Basic/BuiltinsAArch64.def

 BUILTIN(__builtin_arm_ldrex, "v.", "t")
 BUILTIN(__builtin_arm_ldaex, "v.", "t")
 BUILTIN(__builtin_arm_strex, "i.", "t")
 BUILTIN(__builtin_arm_stlex, "i.", "t")
 BUILTIN(__builtin_arm_clrex, "v", "")
 
 // Bit manipulation
 BUILTIN(__builtin_arm_rbit, "UiUi", "nc")
-BUILTIN(__builtin_arm_rbit64, "LUiLUi", "nc")
+BUILTIN(__builtin_arm_rbit64, "LLUiLLUi", "nc")
 
 // HINT
 BUILTIN(__builtin_arm_nop, "v", "")
 BUILTIN(__builtin_arm_yield, "v", "")
 BUILTIN(__builtin_arm_wfe, "v", "")
 BUILTIN(__builtin_arm_wfi, "v", "")
 BUILTIN(__builtin_arm_sev, "v", "")
 BUILTIN(__builtin_arm_sevl, "v", "")
 
 // CRC32
 BUILTIN(__builtin_arm_crc32b, "UiUiUc", "nc")
 BUILTIN(__builtin_arm_crc32cb, "UiUiUc", "nc")
 BUILTIN(__builtin_arm_crc32h, "UiUiUs", "nc")
 BUILTIN(__builtin_arm_crc32ch, "UiUiUs", "nc")
 BUILTIN(__builtin_arm_crc32w, "UiUiUi", "nc")
 BUILTIN(__builtin_arm_crc32cw, "UiUiUi", "nc")
-BUILTIN(__builtin_arm_crc32d, "UiUiLUi", "nc")
-BUILTIN(__builtin_arm_crc32cd, "UiUiLUi", "nc")
+BUILTIN(__builtin_arm_crc32d, "UiUiLLUi", "nc")
+BUILTIN(__builtin_arm_crc32cd, "UiUiLLUi", "nc")
 
 // Memory barrier
 BUILTIN(__builtin_arm_dmb, "vUi", "nc")
 BUILTIN(__builtin_arm_dsb, "vUi", "nc")
 BUILTIN(__builtin_arm_isb, "vUi", "nc")
 
 // Prefetch
 BUILTIN(__builtin_arm_prefetch, "vvC*UiUiUiUi", "nc")
 
 // System Registers
 BUILTIN(__builtin_arm_rsr, "UicC*", "nc")
-BUILTIN(__builtin_arm_rsr64, "LUicC*", "nc")
+BUILTIN(__builtin_arm_rsr64, "LLUicC*", "nc")
 BUILTIN(__builtin_arm_rsrp, "v*cC*", "nc")
 BUILTIN(__builtin_arm_wsr, "vcC*Ui", "nc")
-BUILTIN(__builtin_arm_wsr64, "vcC*LUi", "nc")
+BUILTIN(__builtin_arm_wsr64, "vcC*LLUi", "nc")
 BUILTIN(__builtin_arm_wsrp, "vcC*vC*", "nc")
 
 // MSVC
 LANGBUILTIN(__dmb, "vUi", "nc", ALL_MS_LANGUAGES)
 LANGBUILTIN(__dsb, "vUi", "nc", ALL_MS_LANGUAGES)
 LANGBUILTIN(__isb, "vUi", "nc", ALL_MS_LANGUAGES)
 LANGBUILTIN(__yield, "v", "",   ALL_MS_LANGUAGES)
 LANGBUILTIN(__wfe,   "v", "",   ALL_MS_LANGUAGES)

cfe/trunk/test/CodeGen/arm64-crc32.c

 // REQUIRES: aarch64-registered-target
 // RUN: %clang_cc1 -triple arm64-none-linux-gnu \
 // RUN:  -disable-O0-optnone -S -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-windows \
+// RUN:  -disable-O0-optnone -S -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
+#include <stdint.h>
 
-int crc32b(int a, char b)
+uint32_t crc32b(uint32_t a, uint8_t b)
 {
         return __builtin_arm_crc32b(a,b);
 // CHECK: [[T0:%[0-9]+]] = zext i8 %b to i32
 // CHECK: call i32 @llvm.aarch64.crc32b(i32 %a, i32 [[T0]])
 }
 
-int crc32cb(int a, char b)
+uint32_t crc32cb(uint32_t a, uint8_t b)
 {
         return __builtin_arm_crc32cb(a,b);
 // CHECK: [[T0:%[0-9]+]] = zext i8 %b to i32
 // CHECK: call i32 @llvm.aarch64.crc32cb(i32 %a, i32 [[T0]])
 }
 
-int crc32h(int a, short b)
+uint32_t crc32h(uint32_t a, uint16_t b)
 {
         return __builtin_arm_crc32h(a,b);
 // CHECK: [[T0:%[0-9]+]] = zext i16 %b to i32
 // CHECK: call i32 @llvm.aarch64.crc32h(i32 %a, i32 [[T0]])
 }
 
-int crc32ch(int a, short b)
+uint32_t crc32ch(uint32_t a, uint16_t b)
 {
         return __builtin_arm_crc32ch(a,b);
 // CHECK: [[T0:%[0-9]+]] = zext i16 %b to i32
 // CHECK: call i32 @llvm.aarch64.crc32ch(i32 %a, i32 [[T0]])
 }
 
-int crc32w(int a, int b)
+uint32_t crc32w(uint32_t a, uint32_t b)
 {
         return __builtin_arm_crc32w(a,b);
 // CHECK: call i32 @llvm.aarch64.crc32w(i32 %a, i32 %b)
 }
 
-int crc32cw(int a, int b)
+uint32_t crc32cw(uint32_t a, uint32_t b)
 {
         return __builtin_arm_crc32cw(a,b);
 // CHECK: call i32 @llvm.aarch64.crc32cw(i32 %a, i32 %b)
 }
 
-int crc32d(int a, long b)
+uint32_t crc32d(uint32_t a, uint64_t b)
 {
         return __builtin_arm_crc32d(a,b);
 // CHECK: call i32 @llvm.aarch64.crc32x(i32 %a, i64 %b)
 }
 
-int crc32cd(int a, long b)
+uint32_t crc32cd(uint32_t a, uint64_t b)
 {
         return __builtin_arm_crc32cd(a,b);
 // CHECK: call i32 @llvm.aarch64.crc32cx(i32 %a, i64 %b)
 }

cfe/trunk/test/CodeGen/builtins-arm64.c

 // RUN: %clang_cc1 -triple arm64-unknown-linux -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-windows -disable-O0-optnone -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
+#include <stdint.h>
 
 void f0(void *a, void *b) {
   __clear_cache(a,b);
 // CHECK: call {{.*}} @__clear_cache
 }
 
 void *tp (void) {
   return __builtin_thread_pointer ();

 }
 
 unsigned rsr() {
   // CHECK: [[V0:[%A-Za-z0-9.]+]] = call i64 @llvm.read_register.i64(metadata ![[M0:[0-9]]])
   // CHECK-NEXT: trunc i64 [[V0]] to i32
   return __builtin_arm_rsr("1:2:3:4:5");
 }
 
-unsigned long rsr64() {
+uint64_t rsr64() {
   // CHECK: call i64 @llvm.read_register.i64(metadata ![[M0:[0-9]]])
   return __builtin_arm_rsr64("1:2:3:4:5");
 }
 
 void *rsrp() {
   // CHECK: [[V0:[%A-Za-z0-9.]+]] = call i64 @llvm.read_register.i64(metadata ![[M0:[0-9]]])
   // CHECK-NEXT: inttoptr i64 [[V0]] to i8*
   return __builtin_arm_rsrp("1:2:3:4:5");
 }
 
 void wsr(unsigned v) {
   // CHECK: [[V0:[%A-Za-z0-9.]+]] = zext i32 %v to i64
   // CHECK-NEXT: call void @llvm.write_register.i64(metadata ![[M0:[0-9]]], i64 [[V0]])
   __builtin_arm_wsr("1:2:3:4:5", v);
 }
 
-void wsr64(unsigned long v) {
+void wsr64(uint64_t v) {
   // CHECK: call void @llvm.write_register.i64(metadata ![[M0:[0-9]]], i64 %v)
   __builtin_arm_wsr64("1:2:3:4:5", v);
 }
 
 void wsrp(void *v) {
   // CHECK: [[V0:[%A-Za-z0-9.]+]] = ptrtoint i8* %v to i64
   // CHECK-NEXT: call void @llvm.write_register.i64(metadata ![[M0:[0-9]]], i64 [[V0]])
   __builtin_arm_wsrp("1:2:3:4:5", v);
 }
 
 // CHECK: ![[M0]] = !{!"1:2:3:4:5"}