Rebased #2570 (#2614)

* Add ARC files

* Added ARC files for successful compilation

* Refactor ARC files

* Add ARC c/cs tests

* Add ARC python test/bindings

* Add ARC to CI/CD

* Avoid omitting parameter names

* Update cs files

* Fix ARC bugs

* Update ARC python bindings

* Refactor and update ARC test files

* Add detail flag to arc test

* Fix ARC test problems

* Fix ARCMapping compile error

* Replace __CHAR_BIT__ to CHAR_BIT

* Add credits and ARC info

* Update ARC to match the latest next

* Python formatting

* Remove asserts on 'Unknown condition code passed'

* Add ARC to some more documentation

* Add ARC to Targets constants

* Add ARC support to llvm-tblgen

* Replace asserts & add Expr handling

* Check DecodeGPR32RegisterClass return value

* Fix fieldFromInstruction patch

* Refactor ARC

* Reformat python files

* Fix incorrect import

* Update inc files and insn names

* Update python files

* Disable AArch64 Linux wheel build due to #2615.

---------

Co-authored-by: R33v0LT <[email protected]>

Author: Rot127

.github/labeler.yml

@@ -48,6 +48,13 @@ Alpha:
   - tests/MC/Alpha/**
   - tests/details/alpha.yaml
 
+ARC:
+- arch/ARC/*
+- cstool/cstool_arc.c
+- include/capstone/arc.h
+- suite/MC/ARC/*
+- tests/test_arc.c
+
 BPF:
   - arch/BPF/**
   - cstool/cstool_bpf.c

.github/workflows/auto-sync.yml

@@ -55,7 +55,7 @@ jobs:
           cd vendor/llvm_root
           mkdir build
           cd build
-          cmake -G Ninja -DCMAKE_BUILD_TYPE=Debug ../llvm
+          cmake -G Ninja -DCMAKE_BUILD_TYPE=Debug -DLLVM_EXPERIMENTAL_TARGETS_TO_BUILD=ARC ../llvm
           cmake --build . --target llvm-tblgen --config Debug
           cd ../../../
 
@@ -74,6 +74,7 @@ jobs:
         run: |
           ./src/autosync/ASUpdater.py -d -a AArch64 -s IncGen
           ./src/autosync/ASUpdater.py -d -a Alpha -s IncGen
+          ./src/autosync/ASUpdater.py -d -a ARC -s IncGen
           ./src/autosync/ASUpdater.py -d -a ARM -s IncGen
           ./src/autosync/ASUpdater.py -d -a PPC -s IncGen
           ./src/autosync/ASUpdater.py -d -a LoongArch -s IncGen
@@ -93,6 +94,7 @@ jobs:
       - name: CppTranslator - Test translation
         run: |
           ./src/autosync/ASUpdater.py --ci -d -a AArch64 -s Translate
+          ./src/autosync/ASUpdater.py --ci -d -a ARC -s Translate
           ./src/autosync/ASUpdater.py --ci -d -a ARM -s Translate
           ./src/autosync/ASUpdater.py --ci -d -a PPC -s Translate
           ./src/autosync/ASUpdater.py --ci -d -a LoongArch -s Translate

.github/workflows/build-wheels-publish.yml

@@ -52,9 +52,11 @@ jobs:
           # x86_64 - musllinux
           - { os: ubuntu-latest, arch: x86_64, cibw_build: 'cp38-musllinux* cp313-musllinux*', cibw_skip: '' }
           # aarch64 - manylinux
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp38-manylinux* cp313-manylinux*', cibw_skip: '' }
+          # Disabled for now due to: https://github.com/capstone-engine/capstone/issues/2615
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp38-manylinux* cp313-manylinux*', cibw_skip: '' }
           # aarch64 - musllinux
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp38-musllinux* cp313-musllinux*', cibw_skip: '' }
+          # Disabled for now due to: https://github.com/capstone-engine/capstone/issues/2615
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp38-musllinux* cp313-musllinux*', cibw_skip: '' }
           # macos - x86_64
           - { os: macos-13, arch: x86_64, cibw_build: 'cp38* cp313*', cibw_skip: '' }
           # macos - arm64
@@ -162,15 +164,17 @@ jobs:
           - { os: ubuntu-latest, arch: x86_64, cibw_build: 'cp311-musllinux*', cibw_skip: '' }
           - { os: ubuntu-latest, arch: x86_64, cibw_build: 'cp312-musllinux*', cibw_skip: '' }
           # aarch64 - manylinux
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp39-manylinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp310-manylinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp311-manylinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp312-manylinux*', cibw_skip: '' }
+          # Disabled for now due to: https://github.com/capstone-engine/capstone/issues/2615
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp39-manylinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp310-manylinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp311-manylinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp312-manylinux*', cibw_skip: '' }
           # aarch64 - musllinux
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp39-musllinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp310-musllinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp311-musllinux*', cibw_skip: '' }
-          - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp312-musllinux*', cibw_skip: '' }
+          # Disabled for now due to: https://github.com/capstone-engine/capstone/issues/2615
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp39-musllinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp310-musllinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp311-musllinux*', cibw_skip: '' }
+          # - { os: ubuntu-latest, arch: aarch64, cibw_build: 'cp312-musllinux*', cibw_skip: '' }
           # macos - x86_64
           - { os: macos-13, arch: x86_64, cibw_build: 'cp*', cibw_skip: '*36* *37* *38* *313*' }
           # macos - arm64

.gitignore

@@ -79,6 +79,7 @@ tests/test_riscv
 tests/test_sh
 tests/test_alpha
 tests/test_hppa
+tests/test_arc
 
 # regress binaries
 suite/regress/invalid_read_in_print_operand

BUILDING.md

@@ -44,6 +44,7 @@ By default all are enabled.
 - `CAPSTONE_ARM_SUPPORT`: Support ARM.
 - `CAPSTONE_AARCH64_SUPPORT`: Support AARCH64.
 - `CAPSTONE_ALPHA_SUPPORT`: Support Alpha.
+- `CAPSTONE_ARC_SUPPORT`: Support ARC.
 - `CAPSTONE_HPPA_SUPPORT`: Support HPPA.
 - `CAPSTONE_LOONGARCH_SUPPORT`: Support LoongArch.
 - `CAPSTONE_M680X_SUPPORT`: Support M680X.

CMakeLists.txt

@@ -105,8 +105,8 @@ if(APPLE AND NOT CAPSTONE_BUILD_MACOS_THIN)
     set(CMAKE_OSX_ARCHITECTURES "x86_64;arm64")
 endif()
 
-set(SUPPORTED_ARCHITECTURES ARM AARCH64 M68K MIPS PPC SPARC SYSTEMZ XCORE X86 TMS320C64X M680X EVM MOS65XX WASM BPF RISCV SH TRICORE ALPHA HPPA LOONGARCH XTENSA)
-set(SUPPORTED_ARCHITECTURE_LABELS ARM AARCH64 M68K MIPS PowerPC Sparc SystemZ XCore x86 TMS320C64x M680x EVM MOS65XX WASM BPF RISCV SH TriCore Alpha HPPA LoongArch Xtensa)
+set(SUPPORTED_ARCHITECTURES ARM AARCH64 M68K MIPS PPC SPARC SYSTEMZ XCORE X86 TMS320C64X M680X EVM MOS65XX WASM BPF RISCV SH TRICORE ALPHA HPPA LOONGARCH XTENSA ARC)
+set(SUPPORTED_ARCHITECTURE_LABELS ARM AARCH64 M68K MIPS PowerPC Sparc SystemZ XCore x86 TMS320C64x M680x EVM MOS65XX WASM BPF RISCV SH TriCore Alpha HPPA LoongArch Xtensa ARC)
 
 # If building for OSX it's best to allow CMake to handle building both architectures
 if(APPLE AND NOT CAPSTONE_BUILD_MACOS_THIN)
@@ -232,6 +232,7 @@ set(HEADERS_COMMON
     include/capstone/hppa.h
     include/capstone/loongarch.h
     include/capstone/xtensa.h
+    include/capstone/arc.h
 )
 
 ## architecture support
@@ -716,6 +717,22 @@ if(CAPSTONE_XTENSA_SUPPORT)
     )
 endif()
 
+if (CAPSTONE_ARC_SUPPORT)
+    add_definitions(-DCAPSTONE_HAS_ARC)
+    set(SOURCES_ARC
+            arch/ARC/ARCDisassembler.c
+            arch/ARC/ARCInstPrinter.c
+            arch/ARC/ARCMapping.c
+            arch/ARC/ARCModule.c
+            )
+    set(HEADERS_ARC
+            arch/ARC/ARCInstPrinter.h
+            arch/ARC/ARCMapping.h
+            arch/ARC/ARCModule.h
+            arch/ARC/ARCLinkage.h
+            )
+endif ()
+
 if (CAPSTONE_OSXKERNEL_SUPPORT)
     add_definitions(-DCAPSTONE_HAS_OSXKERNEL)
 endif()
@@ -744,6 +761,7 @@ set(ALL_SOURCES
     ${SOURCES_HPPA}
     ${SOURCES_LOONGARCH}
     ${SOURCES_XTENSA}
+    ${SOURCES_ARC}
 )
 
 set(ALL_HEADERS
@@ -771,6 +789,7 @@ set(ALL_HEADERS
     ${HEADERS_HPPA}
     ${HEADERS_LOONGARCH}
     ${HEADERS_XTENSA}
+    ${HEADERS_ARC}
 )
 
 ## properties
@@ -839,6 +858,7 @@ source_group("Source\\Alpha" FILES ${SOURCES_ALPHA})
 source_group("Source\\HPPA" FILES ${SOURCES_HPPA})
 source_group("Source\\LoongArch" FILES ${SOURCES_LOONGARCH})
 source_group("Source\\Xtensa" FILES ${SOURCES_XTENSA})
+source_group("Source\\ARC" FILES ${SOURCES_ARC})
 
 source_group("Include\\Common" FILES ${HEADERS_COMMON})
 source_group("Include\\Engine" FILES ${HEADERS_ENGINE})
@@ -864,6 +884,7 @@ source_group("Include\\Alpha" FILES ${HEADERS_ALPHA})
 source_group("Include\\HPPA" FILES ${HEADERS_HPPA})
 source_group("Include\\LoongArch" FILES ${HEADERS_LOONGARCH})
 source_group("Include\\Xtensa" FILES ${HEADERS_XTENSA})
+source_group("Include\\ARC" FILES ${HEADERS_ARC})
 
 ## installation
 if(CAPSTONE_INSTALL)

COMPILE_MAKE.TXT

@@ -94,6 +94,7 @@ Capstone requires no prerequisite packages, so it is easy to compile & install.
 	/usr/include/capstone/arm.h
 	/usr/include/capstone/arm64.h
 	/usr/include/capstone/alpha.h
+	/usr/include/capstone/arc.h
 	/usr/include/capstone/bpf.h
 	/usr/include/capstone/capstone.h
 	/usr/include/capstone/evm.h

CREDITS.TXT

@@ -88,6 +88,6 @@ fanfuqiang & citypw & porto703 : RISCV architecture.
 Josh "blacktop" Maine: Arm64 architecture improvements.
 Finn Wilkinson: AArch64 update to Armv9.2-a (SME + SVE2 support)
 Billow & Sidneyp: TriCore architecture.
-Dmitry Sibirtsev: Alpha & HPPA architecture.
+Dmitry Sibirtsev: Alpha, HPPA, ARC architecture.
 Jiajie Chen & Yanglin Xun: LoongArch architecture.
 Billow: Xtensa architecture.

Makefile

@@ -365,11 +365,21 @@ ifneq (,$(findstring xtensa,$(CAPSTONE_ARCHS)))
 	LIBOBJ_XTENSA += $(LIBSRC_XTENSA:%.c=$(OBJDIR)/%.o)
 endif
 
+DEP_ARC =
+DEP_ARC += $(wildcard arch/ARC/ARC*.inc)
+
+LIBOBJ_ARC =
+ifneq (,$(findstring arc,$(CAPSTONE_ARCHS)))
+	CFLAGS += -DCAPSTONE_HAS_ARC
+	LIBSRC_ARC += $(wildcard arch/ARC/ARC*.c)
+	LIBOBJ_ARC += $(LIBSRC_ARC:%.c=$(OBJDIR)/%.o)
+endif
+
 LIBOBJ =
 LIBOBJ += $(OBJDIR)/cs.o $(OBJDIR)/utils.o $(OBJDIR)/SStream.o $(OBJDIR)/MCInstrDesc.o $(OBJDIR)/MCRegisterInfo.o $(OBJDIR)/MCInst.o $(OBJDIR)/MCInstPrinter.o $(OBJDIR)/Mapping.o
 LIBOBJ += $(LIBOBJ_ARM) $(LIBOBJ_AARCH64) $(LIBOBJ_M68K) $(LIBOBJ_MIPS) $(LIBOBJ_PPC) $(LIBOBJ_RISCV) $(LIBOBJ_SPARC) $(LIBOBJ_SYSZ) $(LIBOBJ_SH)
 LIBOBJ += $(LIBOBJ_X86) $(LIBOBJ_XCORE) $(LIBOBJ_TMS320C64X) $(LIBOBJ_M680X) $(LIBOBJ_EVM) $(LIBOBJ_MOS65XX) $(LIBOBJ_WASM) $(LIBOBJ_BPF)
-LIBOBJ += $(LIBOBJ_TRICORE) $(LIBOBJ_ALPHA) $(LIBOBJ_HPPA) $(LIBOBJ_LOONGARCH) $(LIBOBJ_XTENSA)
+LIBOBJ += $(LIBOBJ_TRICORE) $(LIBOBJ_ALPHA) $(LIBOBJ_HPPA) $(LIBOBJ_LOONGARCH) $(LIBOBJ_XTENSA) $(LIBOBJ_ARC)
 
 
 ifeq ($(PKG_EXTRA),)
@@ -503,6 +513,7 @@ $(LIBOBJ_ALPHA): $(DEP_ALPHA)
 $(LIBOBJ_HPPA): $(DEP_HPPA)
 $(LIBOBJ_LOONGARCH): $(DEP_LOONGARCH)
 $(LIBOBJ_XTENSA): $(DEP_XTENSA)
+$(LIBOBJ_ARC): $(DEP_ARC)
 
 ifeq ($(CAPSTONE_STATIC),yes)
 $(ARCHIVE): $(LIBOBJ)

Mapping.c

@@ -354,6 +354,7 @@ DEFINE_get_detail_op(riscv, RISCV);
 DEFINE_get_detail_op(systemz, SystemZ);
 DEFINE_get_detail_op(xtensa, Xtensa);
 DEFINE_get_detail_op(bpf, BPF);
+DEFINE_get_detail_op(arc, ARC);
 
 /// Returns true if for this architecture the
 /// alias operands should be filled.

Mapping.h

@@ -146,6 +146,7 @@ DECL_get_detail_op(riscv, RISCV);
 DECL_get_detail_op(systemz, SystemZ);
 DECL_get_detail_op(xtensa, Xtensa);
 DECL_get_detail_op(bpf, BPF);
+DECL_get_detail_op(arc, ARC);
 
 /// Increments the detail->arch.op_count by one.
 #define DEFINE_inc_detail_op_count(arch, ARCH) \
@@ -185,6 +186,8 @@ DEFINE_inc_detail_op_count(xtensa, Xtensa);
 DEFINE_dec_detail_op_count(xtensa, Xtensa);
 DEFINE_inc_detail_op_count(bpf, BPF);
 DEFINE_dec_detail_op_count(bpf, BPF);
+DEFINE_inc_detail_op_count(arc, ARC);
+DEFINE_dec_detail_op_count(arc, ARC);
 
 /// Returns true if a memory operand is currently edited.
 static inline bool doing_mem(const MCInst *MI)
@@ -215,6 +218,7 @@ DEFINE_get_arch_detail(hppa, HPPA);
 DEFINE_get_arch_detail(loongarch, LoongArch);
 DEFINE_get_arch_detail(mips, Mips);
 DEFINE_get_arch_detail(riscv, RISCV);
+DEFINE_get_arch_detail(arc, ARC);
 DEFINE_get_arch_detail(systemz, SystemZ);
 DEFINE_get_arch_detail(xtensa, Xtensa);
 DEFINE_get_arch_detail(bpf, BPF);
@@ -235,6 +239,7 @@ DEFINE_check_safe_inc(RISCV, RISCV);
 DEFINE_check_safe_inc(SystemZ, SYSTEMZ);
 DEFINE_check_safe_inc(Mips, MIPS);
 DEFINE_check_safe_inc(BPF, BPF);
+DEFINE_check_safe_inc(ARC, ARC);
 
 static inline bool detail_is_set(const MCInst *MI)
 {

MathExtras.h

@@ -493,4 +493,12 @@ static inline uint32_t get_insn_bit(uint32_t insn, uint8_t bit)
 	return get_insn_field(insn, bit, bit);
 }
 
+/// \brief Create a bitmask with the N right-most bits set to 1, and all other
+/// bits set to 0.  Only unsigned types are allowed.
+static inline uint32_t maskTrailingOnes32(uint32_t N) 
+{
+	const unsigned Bits = CHAR_BIT * sizeof(uint32_t);
+	return N == 0 ? 0 : (((uint32_t) -1) >> (Bits - N));
+}
+
 #endif

README.md

@@ -16,7 +16,7 @@ disasm engine for binary analysis and reversing in the security community.
 Created by Nguyen Anh Quynh, then developed and maintained by a small community,
 Capstone offers some unparalleled features:
 
-- Support multiple hardware architectures: ARM, AArch64, Alpha, BPF, Ethereum VM,
+- Support multiple hardware architectures: ARM, AArch64, Alpha, ARC, BPF, Ethereum VM,
   LoongArch, HP PA-RISC (HPPA), M68K, M680X, Mips, MOS65XX, PPC, RISC-V(rv32G/rv64G), SH,
   Sparc, SystemZ, TMS320C64X, TriCore, Webassembly, XCore and X86 (16, 32, 64), Xtensa.