[BPF] Make -mcpu=v3 as the default #107008
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
clang
|
@llvm/pr-subscribers-llvm-binary-utilities Author: None (yonghong-song) ChangesBefore llvm20, (void)__sync_fetch_and_add(...) always generates locked xadd insns. In linux kernel upstream discussion [1], it is found that for arm64 architecture, the original semantics of (void)__sync_fetch_and_add(...), i.e., __atomic_fetch_add(...), is preferred in order for jit to emit proper native barrier insns. In llvm commits [2] and [3], (void)__sync_fetch_and_add(...) will generate the following insns:
To ensure proper barrier semantics for (void)__sync_fetch_and_add(...), cpu v3/v4 is recommended. This patch enables cpu=v3 as the default cpu version. For users wanting to use cpu v1, -mcpu=v1 needs to be explicitly added to clang/llc command line. [1] https://lore.kernel.org/bpf/ZqqiQQWRnz7H93Hc@google.com/T/#mb68d67bc8f39e35a0c3db52468b9de59b79f021f Patch is 51.46 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/107008.diff 81 Files Affected:
diff --git a/clang/lib/Basic/Targets/BPF.cpp b/clang/lib/Basic/Targets/BPF.cpp
index b5ba11a3bdca9d..7171afe975a8dc 100644
--- a/clang/lib/Basic/Targets/BPF.cpp
+++ b/clang/lib/Basic/Targets/BPF.cpp
@@ -38,16 +38,22 @@ void BPFTargetInfo::getTargetDefines(const LangOptions &Opts,
Builder.defineMacro("__BPF_FEATURE_ADDR_SPACE_CAST");
- if (CPU.empty() || CPU == "generic" || CPU == "v1") {
+ if (CPU == "generic" || CPU == "v1") {
Builder.defineMacro("__BPF_CPU_VERSION__", "1");
return;
}
- std::string CpuVerNumStr = CPU.substr(1);
- Builder.defineMacro("__BPF_CPU_VERSION__", CpuVerNumStr);
+ int CpuVerNum;
+ if (CPU.empty()) {
+ Builder.defineMacro("__BPF_CPU_VERSION__", "3");
+ CpuVerNum = 3;
+ } else {
+ std::string CpuVerNumStr = CPU.substr(1);
+ Builder.defineMacro("__BPF_CPU_VERSION__", CpuVerNumStr);
+ CpuVerNum = std::stoi(CpuVerNumStr);
+ }
Builder.defineMacro("__BPF_FEATURE_MAY_GOTO");
- int CpuVerNum = std::stoi(CpuVerNumStr);
if (CpuVerNum >= 2)
Builder.defineMacro("__BPF_FEATURE_JMP_EXT");
diff --git a/clang/test/Preprocessor/bpf-predefined-macros.c b/clang/test/Preprocessor/bpf-predefined-macros.c
index 246cbfa2d6ab7d..360b9330b75fef 100644
--- a/clang/test/Preprocessor/bpf-predefined-macros.c
+++ b/clang/test/Preprocessor/bpf-predefined-macros.c
@@ -1,5 +1,5 @@
-// RUN: %clang -E -target bpfel -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
-// RUN: %clang -E -target bpfeb -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
+// RUN: %clang -E -target bpfel -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
+// RUN: %clang -E -target bpfeb -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
// RUN: %clang -E -target bpfel -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V1 %s
// RUN: %clang -E -target bpfel -mcpu=v2 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V2 %s
// RUN: %clang -E -target bpfel -mcpu=v3 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V3 %s
diff --git a/llvm/lib/Target/BPF/BPFSubtarget.cpp b/llvm/lib/Target/BPF/BPFSubtarget.cpp
index 9a8e42f3237114..7ffa150d3b07ef 100644
--- a/llvm/lib/Target/BPF/BPFSubtarget.cpp
+++ b/llvm/lib/Target/BPF/BPFSubtarget.cpp
@@ -73,7 +73,7 @@ void BPFSubtarget::initSubtargetFeatures(StringRef CPU, StringRef FS) {
HasJmpExt = true;
return;
}
- if (CPU == "v3") {
+ if (CPU.empty() || CPU == "v3") {
HasJmpExt = true;
HasJmp32 = true;
HasAlu32 = true;
diff --git a/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll b/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
index e8a4f81a0240ea..a60d39e5d6c1e3 100644
--- a/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
+++ b/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
@@ -1,4 +1,4 @@
-; RUN: llc -O2 -march=bpfel -mattr=+alu32 < %s | FileCheck %s
+; RUN: llc -O2 -march=bpfel -mcpu=v1 -mattr=+alu32 < %s | FileCheck %s
;
; unsigned int select_cc_32 (unsigned a, unsigned b, int c, int d)
; {
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
index 63a0945edf152a..b8b7a0b8e7c2d3 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
@@ -1,6 +1,6 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
-; RUN: opt -O2 -S < %s | llc -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: opt -O2 -S < %s | llc -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; struct s {
; unsigned long long f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
index 33462198eab15a..4cf0a13f31e80d 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
@@ -1,6 +1,6 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
-; RUN: opt -O2 -S < %s | llc -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: opt -O2 -S < %s | llc -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; struct s {
; unsigned long long f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
index 8f344f2df9d70e..cdcd7e6f5e3744 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
; Source code:
; struct s {
; char f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
index 903fcb2b30a0c0..dd7f1c7a626452 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
; Source code:
; struct s {
; char f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
index a8280579ea509f..525f38de765b20 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
index 39c02da24da42c..11235b55790f1a 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; char a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
index f63a3ee4de7036..e3382d658ca5e5 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1[10][10]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
index cab5474145b46b..69872db3bed0bd 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
index 2c56c0d1d704b4..90706e9b4b98f3 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
index 020b53238767c4..22970401e88849 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1[10][10]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
index f4180c0d0de351..503a26c391acf9 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
index 342a9b70291964..0327f1a0b5e891 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
index 13c79563d4fcb4..2a92d088dc6b68 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; short a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
index 7790273a4978d3..6e62bb308b5801 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
index f41c6a788f6e7c..77ea26a1ff9b3f 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; char a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
index 2ede30701e95b3..556f69f491aae4 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { char a1 [5][5]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
index 9242183d15a33a..2741050b58107c 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
index d5fbc657d4ba7c..b71bbf3de868bc 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; enum A { AA = -1, AB = 0, }; /* signed */
; enum B { BA = 0, BB = 1, }; /* unsigned */
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
index 5a2a1b7031c335..5caea9751b7908 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; enum A { AA = -1, AB = 0, };
; enum B { BA = 0, BB = 1, };
diff --git a/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll b/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
index dbfc4d2f99fc56..9998c985e2aeeb 100644
--- a/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
+++ b/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
@@ -1,5 +1,5 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK %s
; Source code:
; struct data_t {
; int d1;
diff --git a/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll b/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
index 7980c02bde4282..499e368983f3a5 100644
--- a/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
+++ b/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 ...
[truncated]
|
|
@llvm/pr-subscribers-clang Author: None (yonghong-song) ChangesBefore llvm20, (void)__sync_fetch_and_add(...) always generates locked xadd insns. In linux kernel upstream discussion [1], it is found that for arm64 architecture, the original semantics of (void)__sync_fetch_and_add(...), i.e., __atomic_fetch_add(...), is preferred in order for jit to emit proper native barrier insns. In llvm commits [2] and [3], (void)__sync_fetch_and_add(...) will generate the following insns:
To ensure proper barrier semantics for (void)__sync_fetch_and_add(...), cpu v3/v4 is recommended. This patch enables cpu=v3 as the default cpu version. For users wanting to use cpu v1, -mcpu=v1 needs to be explicitly added to clang/llc command line. [1] https://lore.kernel.org/bpf/ZqqiQQWRnz7H93Hc@google.com/T/#mb68d67bc8f39e35a0c3db52468b9de59b79f021f Patch is 51.46 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/107008.diff 81 Files Affected:
diff --git a/clang/lib/Basic/Targets/BPF.cpp b/clang/lib/Basic/Targets/BPF.cpp
index b5ba11a3bdca9d..7171afe975a8dc 100644
--- a/clang/lib/Basic/Targets/BPF.cpp
+++ b/clang/lib/Basic/Targets/BPF.cpp
@@ -38,16 +38,22 @@ void BPFTargetInfo::getTargetDefines(const LangOptions &Opts,
Builder.defineMacro("__BPF_FEATURE_ADDR_SPACE_CAST");
- if (CPU.empty() || CPU == "generic" || CPU == "v1") {
+ if (CPU == "generic" || CPU == "v1") {
Builder.defineMacro("__BPF_CPU_VERSION__", "1");
return;
}
- std::string CpuVerNumStr = CPU.substr(1);
- Builder.defineMacro("__BPF_CPU_VERSION__", CpuVerNumStr);
+ int CpuVerNum;
+ if (CPU.empty()) {
+ Builder.defineMacro("__BPF_CPU_VERSION__", "3");
+ CpuVerNum = 3;
+ } else {
+ std::string CpuVerNumStr = CPU.substr(1);
+ Builder.defineMacro("__BPF_CPU_VERSION__", CpuVerNumStr);
+ CpuVerNum = std::stoi(CpuVerNumStr);
+ }
Builder.defineMacro("__BPF_FEATURE_MAY_GOTO");
- int CpuVerNum = std::stoi(CpuVerNumStr);
if (CpuVerNum >= 2)
Builder.defineMacro("__BPF_FEATURE_JMP_EXT");
diff --git a/clang/test/Preprocessor/bpf-predefined-macros.c b/clang/test/Preprocessor/bpf-predefined-macros.c
index 246cbfa2d6ab7d..360b9330b75fef 100644
--- a/clang/test/Preprocessor/bpf-predefined-macros.c
+++ b/clang/test/Preprocessor/bpf-predefined-macros.c
@@ -1,5 +1,5 @@
-// RUN: %clang -E -target bpfel -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
-// RUN: %clang -E -target bpfeb -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
+// RUN: %clang -E -target bpfel -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
+// RUN: %clang -E -target bpfeb -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_NO %s
// RUN: %clang -E -target bpfel -mcpu=v1 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V1 %s
// RUN: %clang -E -target bpfel -mcpu=v2 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V2 %s
// RUN: %clang -E -target bpfel -mcpu=v3 -x c -o - %s | FileCheck -check-prefix=CHECK -check-prefix=CPU_V3 %s
diff --git a/llvm/lib/Target/BPF/BPFSubtarget.cpp b/llvm/lib/Target/BPF/BPFSubtarget.cpp
index 9a8e42f3237114..7ffa150d3b07ef 100644
--- a/llvm/lib/Target/BPF/BPFSubtarget.cpp
+++ b/llvm/lib/Target/BPF/BPFSubtarget.cpp
@@ -73,7 +73,7 @@ void BPFSubtarget::initSubtargetFeatures(StringRef CPU, StringRef FS) {
HasJmpExt = true;
return;
}
- if (CPU == "v3") {
+ if (CPU.empty() || CPU == "v3") {
HasJmpExt = true;
HasJmp32 = true;
HasAlu32 = true;
diff --git a/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll b/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
index e8a4f81a0240ea..a60d39e5d6c1e3 100644
--- a/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
+++ b/llvm/test/CodeGen/BPF/32-bit-subreg-cond-select.ll
@@ -1,4 +1,4 @@
-; RUN: llc -O2 -march=bpfel -mattr=+alu32 < %s | FileCheck %s
+; RUN: llc -O2 -march=bpfel -mcpu=v1 -mattr=+alu32 < %s | FileCheck %s
;
; unsigned int select_cc_32 (unsigned a, unsigned b, int c, int d)
; {
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
index 63a0945edf152a..b8b7a0b8e7c2d3 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1-bpfeb.ll
@@ -1,6 +1,6 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
-; RUN: opt -O2 -S < %s | llc -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: opt -O2 -S < %s | llc -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; struct s {
; unsigned long long f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
index 33462198eab15a..4cf0a13f31e80d 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-1.ll
@@ -1,6 +1,6 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 3
-; RUN: opt -O2 -S < %s | llc -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: opt -O2 -S < %s | llc -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: opt -O2 -S < %s | llc -mcpu=v1 -mattr=+alu32 -filetype=asm | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; struct s {
; unsigned long long f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
index 8f344f2df9d70e..cdcd7e6f5e3744 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2-bpfeb.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
; Source code:
; struct s {
; char f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
index 903fcb2b30a0c0..dd7f1c7a626452 100644
--- a/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/field-reloc-bitfield-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
; Source code:
; struct s {
; char f1;
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
index a8280579ea509f..525f38de765b20 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
index 39c02da24da42c..11235b55790f1a 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; char a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
index f63a3ee4de7036..e3382d658ca5e5 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-byte-size-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1[10][10]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
index cab5474145b46b..69872db3bed0bd 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
index 2c56c0d1d704b4..90706e9b4b98f3 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
index 020b53238767c4..22970401e88849 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-existence-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1[10][10]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
index f4180c0d0de351..503a26c391acf9 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1-bpfeb.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EB,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
index 342a9b70291964..0327f1a0b5e891 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-EL,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
index 13c79563d4fcb4..2a92d088dc6b68 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-lshift-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; short a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
index 7790273a4978d3..6e62bb308b5801 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1:7; int a2:4; int a3:5; int a4:16;} __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
index f41c6a788f6e7c..77ea26a1ff9b3f 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { int a1; char a2; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
index 2ede30701e95b3..556f69f491aae4 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-rshift-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef struct s1 { char a1 [5][5]; } __s1;
; union u1 { int b1; __s1 b2; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
index 9242183d15a33a..2741050b58107c 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-1.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; typedef unsigned __uint;
; struct s1 { int a1; __uint a2:9; __uint a3:4; };
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
index d5fbc657d4ba7c..b71bbf3de868bc 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-2.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; enum A { AA = -1, AB = 0, }; /* signed */
; enum B { BA = 0, BB = 1, }; /* unsigned */
diff --git a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
index 5a2a1b7031c335..5caea9751b7908 100644
--- a/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
+++ b/llvm/test/CodeGen/BPF/CORE/intrinsic-fieldinfo-signedness-3.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
+; RUN: llc -mcpu=v1 -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
; Source code:
; enum A { AA = -1, AB = 0, };
; enum B { BA = 0, BB = 1, };
diff --git a/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll b/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
index dbfc4d2f99fc56..9998c985e2aeeb 100644
--- a/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
+++ b/llvm/test/CodeGen/BPF/CORE/no-narrow-load.ll
@@ -1,5 +1,5 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK %s
+; RUN: llc -mcpu=v1 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK %s
; Source code:
; struct data_t {
; int d1;
diff --git a/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll b/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
index 7980c02bde4282..499e368983f3a5 100644
--- a/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
+++ b/llvm/test/CodeGen/BPF/CORE/offset-reloc-end-load.ll
@@ -1,6 +1,6 @@
; RUN: opt -O2 %s | llvm-dis > %t1
-; RUN: llc -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU64 %s
-; RUN: llc -mattr=+alu32 -filetype=asm -o - %t1 | FileCheck -check-prefixes=CHECK,CHECK-ALU32 %s
+; RUN: llc -mcpu=v1 ...
[truncated]
|
eddyz87
reviewed
Sep 2, 2024
eddyz87
approved these changes
Sep 2, 2024
4ast
approved these changes
Sep 3, 2024
Before llvm20, (void)__sync_fetch_and_add(...) always generates locked xadd insns. In linux kernel upstream discussion [1], it is found that for arm64 architecture, the original semantics of (void)__sync_fetch_and_add(...), i.e., __atomic_fetch_add(...), is preferred in order for jit to emit proper native barrier insns. In llvm commits [2] and [3], (void)__sync_fetch_and_add(...) will generate the following insns: - for cpu v1/v2: locked xadd insns to keep backward compatibility - for cpu v3/v4: __atomic_fetch_add() insns To ensure proper barrier semantics for (void)__sync_fetch_and_add(...), cpu v3/v4 is recommended. This patch enables cpu=v3 as the default cpu version. For users wanting to use cpu v1, -mcpu=v1 needs to be explicitly added to clang/llc command line. [1] https://lore.kernel.org/bpf/[email protected]/T/#mb68d67bc8f39e35a0c3db52468b9de59b79f021f [2] llvm#101428 [3] llvm#106494
yonghong-song
force-pushed
the
bpf-default-v3
branch
from
8fcbb0a to
670fa2f
Compare
yonghong-song
pushed a commit
to iovisor/bcc
that referenced
this pull request
Sep 4, 2024
Upstream patch llvm/llvm-project#107008 makes cpu=v3 as the default. Previously cpu=v1 as the default for bcc. Let us mark cpu=v1 explicitly for bcc for now to keep it backword compatible.
yonghong-song
pushed a commit
to iovisor/bcc
that referenced
this pull request
Sep 4, 2024
Upstream patch llvm/llvm-project#107008 makes cpu=v3 as the default. Previously cpu=v1 as the default for bcc. Let us mark cpu=v1 explicitly for bcc for now to keep it backword compatible.
dkruces
pushed a commit
to dkruces/bcc
that referenced
this pull request
Nov 28, 2024
Upstream patch llvm/llvm-project#107008 makes cpu=v3 as the default. Previously cpu=v1 as the default for bcc. Let us mark cpu=v1 explicitly for bcc for now to keep it backword compatible.
|
Hey, I didn't see this change on the LLVM 20.1.0 Release Notes - it would be nice if you could add those in the future as a heads up. Anyway, I got here after tracing a regression in eBPF codegen introduced in Clang 20 which results in dramatically larger kernel verifier states (and eventually causes the verifier to give up with sufficiently large functions). The culprit does indeed seem to be BPF v3 - my guess is that branching according to 32-bit registers provides less refutability for the verifier which ends up carrying more possible states as a result, but I didn't dive too deep into this. FYI |
|
Thanks @yuvald-sweet-security I am not sure whether I can add default v1->v3 in llvm20 or not. But let me check anyway. Also, if the verification failure can be easily reproduced, could you submit the test case, we can do some analysis to see why and maybe verifier could be improved due to this. Thanks! |
The specific case I was researching is in proprietary code, but I can see eBPF verifier/codegen degradations caused by moving from v1 to v3 in many public repos, e.g. try to compile Tracee (https://github.com/aquasecurity/tracee): Note that e.g. the As I said I haven't had the time to research this deeply, but I think that all cases are the result of kernel verifier getting less refutability from 32-bit jump instructions introduced in v3. But I might be wrong. |
|
@4ast perhaps you know who in LLVM can help with adding a "Changes to the BPF Backend" section to the release notes, similar to how other backends have their own sections? |
|
@yuvald-sweet-security Could you share which kernel you are used for above testing? |
yonghong-song
pushed a commit
to yonghong-song/llvm-project
that referenced
this pull request
Mar 17, 2025
The pull request [1] changed bpf default cpu from -mcpu=v1 to -mcpu=v3 in clang20. Recently in [1], Yuval Deutscher suggested to add an entry to clang20 ReleaseNotes so users can easily find the change from documentation. [1] llvm#107008
The llvm patch link to add release note in llvm20: #131691 |
yonghong-song
pushed a commit
to yonghong-song/llvm-project
that referenced
this pull request
Mar 18, 2025
The pull request [1] changed bpf default cpu from -mcpu=v1 to -mcpu=v3 in clang20. Recently in [1], Yuval Deutscher suggested to add an entry to clang20 ReleaseNotes so users can easily find the change from documentation. [1] llvm#107008
yonghong-song
pushed a commit
to yonghong-song/llvm-project
that referenced
this pull request
Mar 18, 2025
The pull request [1] changed bpf default cpu from -mcpu=v1 to -mcpu=v3 in clang20. Recently in [1], Yuval Deutscher suggested to add an entry to clang20 ReleaseNotes so users can easily find the change from documentation. [1] llvm#107008
dkruces
pushed a commit
to dkruces/bcc
that referenced
this pull request
Mar 18, 2025
Upstream patch llvm/llvm-project#107008 makes cpu=v3 as the default. Previously cpu=v1 as the default for bcc. Let us mark cpu=v1 explicitly for bcc for now to keep it backword compatible.
tstellar
pushed a commit
that referenced
this pull request
Mar 18, 2025
The pull request [1] changed bpf default cpu from -mcpu=v1 to -mcpu=v3 in clang20. Recently in [1], Yuval Deutscher suggested to add an entry to clang20 ReleaseNotes so users can easily find the change from documentation. [1] #107008 Co-authored-by: Yonghong Song <[email protected]>
I ran this on my host machine which is Windows with WSL, kernel |
Thank you very much :) |
Thanks @yuvald-sweet-security I will take a look. |
|
@yuvald-sweet-security I did one example with your above example, I can reproduce the issue. The v3 failure reason is actually due to your have inline asm code which prevents compiler from doing a good job for v3. I did the following change to the source The barrier_var(len) (i.e., asm volatile("" : "+r"(len))) is only needed for cpu=v1 but also works for cpu=v3. The original asm code prevents a good compiler transformation at cpu=v3. I suggest you try to replace asm code with proper C code and if necessary barrier_var. |
|
@yonghong-song thank you for taking the time to help with this issue and providing your suggestions, it’s greatly appreciated. I am also glad to hear that asm barriers are no longer necessary, as they caused quite some trouble for me in the past. However, I've encountered a few issues while attempting to apply this as a solution to the problem with moving to mpcu v3.
As you can see the v3 codegen causes the verifier to consume about 4 times the number of instructions when verifying it, and while it's not such a big issue for this particular function, it can be an issue for larger functions which are already close to the verifier's 1 million instructions limit as this can cause them to go over it.
thank you again for your thoughtful advice and support |
llvm-sync bot
pushed a commit
to arm/arm-toolchain
that referenced
this pull request
May 15, 2025
The pull request [1] changed bpf default cpu from -mcpu=v1 to -mcpu=v3 in clang20. Recently in [1], Yuval Deutscher suggested to add an entry to clang20 ReleaseNotes so users can easily find the change from documentation. [1] llvm/llvm-project#107008 Co-authored-by: Yonghong Song <[email protected]>
intel-lab-lkp
pushed a commit
to intel-lab-lkp/linux
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]>
kernel-patches-daemon-bpf-rc bot
pushed a commit
to kernel-patches/bpf-rc
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]>
kernel-patches-daemon-bpf bot
pushed a commit
to kernel-patches/bpf
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]>
kernel-patches-daemon-bpf-rc bot
pushed a commit
to kernel-patches/bpf-rc
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]>
kernel-patches-daemon-bpf bot
pushed a commit
to kernel-patches/bpf
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]>
kernel-patches-daemon-bpf bot
pushed a commit
to kernel-patches/bpf
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Alexei Starovoitov <[email protected]>
kernel-patches-daemon-bpf-rc bot
pushed a commit
to kernel-patches/bpf-rc
that referenced
this pull request
Jun 12, 2025
The default cpu version is changed from v1 to v3 in llvm version 20. See [1] for more detailed reasoning. Update bpf_devel_QA.rst so developers can find such information easily. [1] llvm/llvm-project#107008 Signed-off-by: Yonghong Song <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Alexei Starovoitov <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Before llvm20, (void)__sync_fetch_and_add(...) always generates locked xadd insns. In linux kernel upstream discussion [1], it is found that for arm64 architecture, the original semantics of (void)__sync_fetch_and_add(...), i.e., __atomic_fetch_add(...), is preferred in order for jit to emit proper native barrier insns.
In llvm commits [2] and [3], (void)__sync_fetch_and_add(...) will generate the following insns:
To ensure proper barrier semantics for (void)__sync_fetch_and_add(...), cpu v3/v4 is recommended.
This patch enables cpu=v3 as the default cpu version. For users wanting to use cpu v1, -mcpu=v1 needs to be explicitly added to clang/llc command line.
[1] https://lore.kernel.org/bpf/[email protected]/T/#mb68d67bc8f39e35a0c3db52468b9de59b79f021f
[2] #101428
[3] #106494