llvm-project/clang/docs/ClangCommandLineReference.rst

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

4196 lines
103 KiB
ReStructuredText
Raw Normal View History

..
-------------------------------------------------------------------
NOTE: This file is automatically generated by running clang-tblgen
-gen-opt-docs. Do not edit this file by hand!!
-------------------------------------------------------------------
=====================================
Clang command line argument reference
=====================================
.. contents::
:local:
Introduction
============
This page lists the command line arguments currently supported by the
GCC-compatible ``clang`` and ``clang++`` drivers.
.. program:: clang
.. option:: -B<prefix>, --prefix <arg>, --prefix=<arg>
Search $prefix$file for executables, libraries, and data files. If $prefix is a directory, search $prefix/$file
.. option:: -F<arg>
Add directory to framework include search path
.. option:: -ObjC
Treat source input files as Objective-C inputs
.. program:: clang1
.. option:: -ObjC++
.. program:: clang
Treat source input files as Objective-C++ inputs
.. option:: -Qn, -fno-ident
Do not emit metadata containing compiler name and version
.. option:: -Qunused-arguments
Don't emit warning for unused driver arguments
.. option:: -Qy, -fident
Emit metadata containing compiler name and version
.. option:: -Wa,<arg>,<arg2>...
Pass the comma separated arguments in <arg> to the assembler
.. option:: -Wlarge-by-value-copy=<arg>
.. option:: -Xarch\_<arg1> <arg2>
.. program:: clang1
.. option:: -Xarch\_device <arg>
.. program:: clang
Pass <arg> to the CUDA/HIP device compilation
.. program:: clang2
.. option:: -Xarch\_host <arg>
.. program:: clang
Pass <arg> to the CUDA/HIP host compilation
.. option:: -Xcuda-fatbinary <arg>
Pass <arg> to fatbinary invocation
.. option:: -Xcuda-ptxas <arg>
Pass <arg> to the ptxas assembler
.. option:: -Xopenmp-target <arg>
Pass <arg> to the target offloading toolchain.
.. program:: clang1
.. option:: -Xopenmp-target=<triple> <arg>
.. program:: clang
Pass <arg> to the target offloading toolchain identified by <triple>.
.. option:: -Z<arg>
.. option:: -a<arg>, --profile-blocks
.. option:: -all\_load
.. option:: -allowable\_client <arg>
.. option:: --analyze
Run the static analyzer
.. option:: --analyzer-no-default-checks
.. option:: --analyzer-output<arg>
Static analyzer report output format (html\|plist\|plist-multi-file\|plist-html\|sarif\|sarif-html\|text).
.. option:: -arch <arg>
.. program:: clang1
.. option:: -arch\_errors\_fatal
.. program:: clang
.. program:: clang2
.. option:: -arch\_only <arg>
.. program:: clang
.. option:: -arcmt-migrate-emit-errors
Emit ARC errors even if the migrator can fix them
.. option:: -arcmt-migrate-report-output <arg>
Output path for the plist report
.. option:: --autocomplete=<arg>
.. option:: -bind\_at\_load
.. option:: -bundle
.. program:: clang1
.. option:: -bundle\_loader <arg>
.. program:: clang
.. option:: -client\_name<arg>
.. option:: -compatibility\_version<arg>
.. option:: --config <arg>
Specifies configuration file
.. option:: --constant-cfstrings
.. option:: --cuda-compile-host-device
Compile CUDA code for both host and device (default). Has no effect on non-CUDA compilations.
.. option:: --cuda-device-only
Compile CUDA code for device only
.. option:: --cuda-host-only
Compile CUDA code for host only. Has no effect on non-CUDA compilations.
.. option:: --cuda-include-ptx=<arg>, --no-cuda-include-ptx=<arg>
Include PTX for the following GPU architecture (e.g. sm\_35) or 'all'. May be specified more than once.
.. option:: --cuda-noopt-device-debug, --no-cuda-noopt-device-debug
Enable device-side debug info generation. Disables ptxas optimizations.
.. option:: -cuid=<arg>
An ID for compilation unit, which should be the same for the same compilation unit but different for different compilation units. It is used to externalize device-side static variables for single source offloading languages CUDA and HIP so that they can be accessed by the host code of the same compilation unit.
.. option:: -current\_version<arg>
.. option:: -dead\_strip
.. option:: -dependency-dot <arg>
Filename to write DOT-formatted header dependencies to
.. option:: -dependency-file <arg>
Filename (or -) to write dependency output to
.. option:: -dsym-dir<dir>
Directory to output dSYM's (if any) to
.. option:: -dumpmachine
.. option:: -dumpversion
.. option:: --dyld-prefix=<arg>, --dyld-prefix <arg>
.. option:: -dylib\_file <arg>
.. option:: -dylinker
.. program:: clang1
.. option:: -dylinker\_install\_name<arg>
.. program:: clang
.. option:: -dynamic
.. option:: -dynamiclib
.. option:: -emit-ast
Emit Clang AST files for source inputs
.. option:: --emit-static-lib
Enable linker job to emit a static library.
.. option:: -enable-trivial-auto-var-init-zero-knowing-it-will-be-removed-from-clang
Trivial automatic variable initialization to zero is only here for benchmarks, it'll eventually be removed, and I'm OK with that because I'm only using it to benchmark
.. option:: --end-no-unused-arguments
Start emitting warnings for unused driver arguments
.. option:: -exported\_symbols\_list <arg>
.. option:: -faligned-new=<arg>
.. option:: -fautomatic
.. option:: -ffixed-r19
Reserve register r19 (Hexagon only)
.. option:: -fgpu-flush-denormals-to-zero, -fcuda-flush-denormals-to-zero, -fno-gpu-flush-denormals-to-zero
Flush denormal floating point values to zero in CUDA/HIP device mode.
.. option:: -fheinous-gnu-extensions
.. option:: -flat\_namespace
.. option:: -fopenmp-targets=<arg1>,<arg2>...
Specify comma-separated list of triples OpenMP offloading targets to be supported
.. option:: -force\_cpusubtype\_ALL
.. program:: clang1
.. option:: -force\_flat\_namespace
.. program:: clang
.. program:: clang2
.. option:: -force\_load <arg>
.. program:: clang
.. option:: -fplugin-arg-<name>-<arg>
Pass <arg> to plugin <name>
.. option:: -framework <arg>
.. option:: -frtlib-add-rpath, -fno-rtlib-add-rpath
Add -rpath with architecture-specific resource directory to the linker flags
.. option:: -fsanitize-system-ignorelist=<arg>, -fsanitize-system-blacklist=<arg>
Path to system ignorelist file for sanitizers
.. option:: -fshow-skipped-includes
#include files may be "skipped" due to include guard optimization
or #pragma once. This flag makes -H show also such includes.
.. option:: -fsystem-module
Build this module as a system module. Only used with -emit-module
.. option:: -fuse-cuid=<arg>
Method to generate ID's for compilation units for single source offloading languages CUDA and HIP: 'hash' (ID's generated by hashing file path and command line options) \| 'random' (ID's generated as random numbers) \| 'none' (disabled). Default is 'hash'. This option will be overridden by option '-cuid=\[ID\]' if it is specified.
.. option:: --gcc-toolchain=<arg>
Search for GCC installation in the specified directory on targets which commonly use GCC. The directory usually contains 'lib{,32,64}/gcc{,-cross}/$triple' and 'include'. If specified, sysroot is skipped for GCC detection. Note: executables (e.g. ld) used by the compiler are not overridden by the selected GCC installation
.. option:: -gcodeview
Generate CodeView debug information
.. option:: -gcodeview-ghash, -gno-codeview-ghash
Emit type record hashes in a .debug$H section
.. option:: --gpu-instrument-lib=<arg>
Instrument device library for HIP, which is a LLVM bitcode containing \_\_cyg\_profile\_func\_enter and \_\_cyg\_profile\_func\_exit
.. option:: --gpu-max-threads-per-block=<arg>
Default max threads per block for kernel launch bounds for HIP
.. option:: -headerpad\_max\_install\_names<arg>
.. option:: -help, --help
Display available options
.. option:: --help-hidden
Display help for hidden options
.. option:: --hip-link
Link clang-offload-bundler bundles for HIP
.. option:: --hip-version=<arg>
HIP version in the format of major.minor.patch
.. option:: -ibuiltininc
Enable builtin #include directories even when -nostdinc is used before or after -ibuiltininc. Using -nobuiltininc after the option disables it
.. option:: -image\_base <arg>
.. option:: -index-header-map
Make the next included directory (-I or -F) an indexer header map
.. option:: -init <arg>
.. option:: -install\_name <arg>
.. option:: -interface-stub-version=<arg>
.. option:: -keep\_private\_externs
.. option:: -lazy\_framework <arg>
.. program:: clang1
.. option:: -lazy\_library <arg>
.. program:: clang
.. option:: -mbig-endian, -EB
.. option:: -mbranch-protection=<arg>
Enforce targets of indirect branches and function returns
.. option:: -menable-unsafe-fp-math
Allow unsafe floating-point math optimizations which may decrease precision
.. option:: -mharden-sls=<arg>
Select straight-line speculation hardening scope
.. option:: --migrate
Run the migrator
.. option:: -mios-simulator-version-min=<arg>, -miphonesimulator-version-min=<arg>
.. option:: -mlinker-version=<arg>
.. option:: -mlittle-endian, -EL
.. option:: -mllvm <arg>
Additional arguments to forward to LLVM's option processing
.. option:: -module-dependency-dir <arg>
Directory to dump module dependencies to
.. option:: -mtvos-simulator-version-min=<arg>, -mappletvsimulator-version-min=<arg>
.. option:: -multi\_module
.. option:: -multiply\_defined <arg>
.. program:: clang1
.. option:: -multiply\_defined\_unused <arg>
.. program:: clang
.. option:: -mwatchos-simulator-version-min=<arg>, -mwatchsimulator-version-min=<arg>
.. option:: --no-cuda-version-check
Don't error out if the detected version of the CUDA install is too low for the requested CUDA gpu architecture.
.. option:: -no-integrated-cpp, --no-integrated-cpp
.. option:: -no\_dead\_strip\_inits\_and\_terms
.. option:: -nobuiltininc
Disable builtin #include directories
.. option:: -nodefaultlibs
.. option:: -nofixprebinding
.. option:: -nogpuinc, -nocudainc
Do not add include paths for CUDA/HIP and do not include the default CUDA/HIP wrapper headers
.. option:: -nogpulib, -nocudalib
Do not link device library for CUDA/HIP device compilation
.. option:: -nohipwrapperinc
Do not include the default HIP wrapper headers and include paths
.. option:: -nolibc
.. option:: -nomultidefs
.. option:: -nopie, -no-pie
.. option:: -noprebind
.. option:: -noprofilelib
.. option:: -noseglinkedit
.. option:: -nostdinc, --no-standard-includes
.. program:: clang1
.. option:: -nostdinc++
.. program:: clang
Disable standard #include directories for the C++ standard library
.. option:: -nostdlib++
.. option:: -nostdlibinc
.. option:: -o<file>, --output <arg>, --output=<arg>
Write output to <file>
.. option:: -objcmt-allowlist-dir-path=<arg>, -objcmt-white-list-dir-path=<arg>, -objcmt-whitelist-dir-path=<arg>
Only modify files with a filename contained in the provided directory path
.. option:: -objcmt-atomic-property
Make migration to 'atomic' properties
.. option:: -objcmt-migrate-all
Enable migration to modern ObjC
.. option:: -objcmt-migrate-annotation
Enable migration to property and method annotations
.. option:: -objcmt-migrate-designated-init
Enable migration to infer NS\_DESIGNATED\_INITIALIZER for initializer methods
.. option:: -objcmt-migrate-instancetype
Enable migration to infer instancetype for method result type
.. option:: -objcmt-migrate-literals
Enable migration to modern ObjC literals
.. option:: -objcmt-migrate-ns-macros
Enable migration to NS\_ENUM/NS\_OPTIONS macros
.. option:: -objcmt-migrate-property
Enable migration to modern ObjC property
.. option:: -objcmt-migrate-property-dot-syntax
Enable migration of setter/getter messages to property-dot syntax
.. option:: -objcmt-migrate-protocol-conformance
Enable migration to add protocol conformance on classes
.. option:: -objcmt-migrate-readonly-property
Enable migration to modern ObjC readonly property
.. option:: -objcmt-migrate-readwrite-property
Enable migration to modern ObjC readwrite property
.. option:: -objcmt-migrate-subscripting
Enable migration to modern ObjC subscripting
.. option:: -objcmt-ns-nonatomic-iosonly
Enable migration to use NS\_NONATOMIC\_IOSONLY macro for setting property's 'atomic' attribute
.. option:: -objcmt-returns-innerpointer-property
Enable migration to annotate property with NS\_RETURNS\_INNER\_POINTER
.. option:: -object
.. option:: -object-file-name=<file>, -object-file-name <arg>
Set the output <file> for debug infos
.. option:: --offload-arch=<arg>, --cuda-gpu-arch=<arg>, --no-offload-arch=<arg>
CUDA offloading device architecture (e.g. sm\_35), or HIP offloading target ID in the form of a device architecture followed by target ID features delimited by a colon. Each target ID feature is a pre-defined string followed by a plus or minus sign (e.g. gfx908:xnack+:sramecc-). May be specified more than once.
.. option:: --offload=<arg1>,<arg2>...
Specify comma-separated list of offloading target triples (CUDA and HIP only)
.. option:: -p, --profile
.. option:: -pagezero\_size<arg>
.. option:: -pg
Enable mcount instrumentation
.. option:: -pipe, --pipe
Use pipes between commands, when possible
.. option:: -prebind
.. program:: clang1
.. option:: -prebind\_all\_twolevel\_modules
.. program:: clang
.. option:: -preload
.. option:: --print-diagnostic-categories
.. option:: -print-effective-triple, --print-effective-triple
Print the effective target triple
.. option:: -print-file-name=<file>, --print-file-name=<file>, --print-file-name <arg>
Print the full library path of <file>
.. option:: -print-ivar-layout
Enable Objective-C Ivar layout bitmap print trace
.. option:: -print-libgcc-file-name, --print-libgcc-file-name
Print the library path for the currently used compiler runtime library ("libgcc.a" or "libclang\_rt.builtins.\*.a")
.. option:: -print-multi-directory, --print-multi-directory
.. option:: -print-multi-lib, --print-multi-lib
.. option:: -print-multiarch, --print-multiarch
Print the multiarch target triple
.. option:: -print-prog-name=<name>, --print-prog-name=<name>, --print-prog-name <arg>
Print the full program path of <name>
.. option:: -print-resource-dir, --print-resource-dir
Print the resource directory pathname
.. option:: -print-rocm-search-dirs, --print-rocm-search-dirs
Print the paths used for finding ROCm installation
.. option:: -print-runtime-dir, --print-runtime-dir
Print the directory pathname containing clangs runtime libraries
.. option:: -print-search-dirs, --print-search-dirs
Print the paths used for finding libraries and programs
.. option:: -print-target-triple, --print-target-triple
Print the normalized target triple
.. option:: -print-targets, --print-targets
Print the registered targets
.. option:: -private\_bundle
.. option:: -pthread, -no-pthread
Support POSIX threads in generated code
.. option:: -pthreads
.. option:: -read\_only\_relocs <arg>
.. option:: -relocatable-pch, --relocatable-pch
Whether to build a relocatable precompiled header
.. option:: -remap
.. option:: -rewrite-legacy-objc
Rewrite Legacy Objective-C source to C++
.. option:: -rtlib=<arg>, --rtlib=<arg>, --rtlib <arg>
Compiler runtime library to use
.. option:: -save-stats=<arg>, --save-stats=<arg>, -save-stats (equivalent to -save-stats=cwd), --save-stats (equivalent to -save-stats=cwd)
Save llvm statistics.
.. option:: -save-temps=<arg>, --save-temps=<arg>, -save-temps (equivalent to -save-temps=cwd), --save-temps (equivalent to -save-temps=cwd)
Save intermediate compilation results.
.. option:: -sectalign <arg1> <arg2> <arg3>
.. option:: -sectcreate <arg1> <arg2> <arg3>
.. option:: -sectobjectsymbols <arg1> <arg2>
.. option:: -sectorder <arg1> <arg2> <arg3>
.. option:: -seg1addr<arg>
.. option:: -seg\_addr\_table <arg>
.. program:: clang1
.. option:: -seg\_addr\_table\_filename <arg>
.. program:: clang
.. option:: -segaddr <arg1> <arg2>
.. option:: -segcreate <arg1> <arg2> <arg3>
.. option:: -seglinkedit
.. option:: -segprot <arg1> <arg2> <arg3>
.. option:: -segs\_read\_<arg>
.. program:: clang1
.. option:: -segs\_read\_only\_addr <arg>
.. program:: clang
.. program:: clang2
.. option:: -segs\_read\_write\_addr <arg>
.. program:: clang
.. option:: -serialize-diagnostics <arg>, --serialize-diagnostics <arg>
Serialize compiler diagnostics to a file
.. option:: -shared-libgcc
.. option:: -shared-libsan, -shared-libasan
Dynamically link the sanitizer runtime
.. option:: -single\_module
.. option:: --start-no-unused-arguments
Don't emit warnings about unused arguments for the following arguments
.. option:: -static-libgcc
.. option:: -static-libsan
Statically link the sanitizer runtime
.. option:: -static-libstdc++
.. option:: -static-openmp
Use the static host OpenMP runtime while linking.
.. option:: -std-default=<arg>
.. option:: -stdlib=<arg>, --stdlib=<arg>, --stdlib <arg>
C++ standard library to use
.. option:: -sub\_library<arg>
.. program:: clang1
.. option:: -sub\_umbrella<arg>
.. program:: clang
.. option:: --sysroot=<arg>, --sysroot <arg>
.. option:: --target-help
.. option:: --target=<arg>, -target <arg>
Generate code for the given target
.. option:: -time
Time individual commands
.. option:: -traditional, --traditional
.. option:: -traditional-cpp, --traditional-cpp
Enable some traditional CPP emulation
.. option:: -twolevel\_namespace
.. program:: clang1
.. option:: -twolevel\_namespace\_hints
.. program:: clang
.. option:: -umbrella <arg>
.. option:: -unexported\_symbols\_list <arg>
.. option:: -unwindlib=<arg>, --unwindlib=<arg>
Unwind library to use
.. option:: -v, --verbose
Show commands to run and use verbose output
.. option:: --verify-debug-info
Verify the binary representation of debug output
.. option:: --version
Print version information
.. option:: -w, --no-warnings
Suppress all warnings
.. option:: -weak-l<arg>
.. option:: -weak\_framework <arg>
.. program:: clang1
.. option:: -weak\_library <arg>
.. program:: clang
.. program:: clang2
.. option:: -weak\_reference\_mismatches <arg>
.. program:: clang
.. option:: -whatsloaded
.. option:: -why\_load, -whyload
.. option:: -working-directory<arg>, -working-directory=<arg>
Resolve file paths relative to the specified directory
.. option:: -x<language>, --language <arg>, --language=<arg>
Treat subsequent input files as having type <language>
.. option:: -y<arg>
Actions
=======
The action to perform on the input.
.. option:: -E, --preprocess
Only run the preprocessor
.. option:: -S, --assemble
Only run preprocess and compilation steps
.. option:: -c, --compile
Only run preprocess, compile, and assemble steps
.. option:: -emit-interface-stubs
Generate Interface Stub Files.
.. option:: -emit-llvm
Use the LLVM representation for assembler and object files
.. option:: -emit-merged-ifs
Generate Interface Stub Files, emit merged text not binary.
.. option:: -extract-api
Extract API information
.. option:: -fopenmp-new-driver, fno-openmp-new-driver
Use the new driver for OpenMP offloading.
.. option:: -fsyntax-only
.. option:: -module-file-info
Provide information about a particular module file
.. option:: --precompile
Only precompile the input
.. option:: -rewrite-objc
Rewrite Objective-C source to C++
.. option:: -verify-pch
Load and verify that a pre-compiled header file is not stale
Compilation flags
=================
Flags controlling the behavior of Clang during compilation. These flags have
no effect during actions that do not perform compilation.
.. option:: -Xassembler <arg>
Pass <arg> to the assembler
.. option:: -Xclang <arg>
Pass <arg> to the clang compiler
.. option:: -ansi, --ansi
.. option:: -fc++-abi=<arg>
C++ ABI to use. This will override the target C++ ABI.
.. option:: -fclang-abi-compat=<version>
Attempt to match the ABI of Clang <version>
.. option:: -fcomment-block-commands=<arg>,<arg2>...
Treat each comma separated argument in <arg> as a documentation comment block command
.. option:: -fcomplete-member-pointers, -fno-complete-member-pointers
Require member pointer base types to be complete if they would be significant under the Microsoft ABI
.. option:: -fcrash-diagnostics-dir=<dir>
Put crash-report files in <dir>
.. option:: -fdeclspec, -fno-declspec
Allow \_\_declspec as a keyword
.. option:: -fdepfile-entry=<arg>
.. option:: -fdiagnostics-fixit-info, -fno-diagnostics-fixit-info
.. option:: -fdiagnostics-format=<arg>
.. option:: -fdiagnostics-parseable-fixits
Print fix-its in machine parseable form
.. option:: -fdiagnostics-print-source-range-info
Print source range spans in numeric form
.. option:: -fdiagnostics-show-category=<arg>
.. option:: -fdiscard-value-names, -fno-discard-value-names
Discard value names in LLVM IR
.. option:: -fexperimental-relative-c++-abi-vtables, -fno-experimental-relative-c++-abi-vtables
Use the experimental C++ class ABI for classes with virtual tables
.. option:: -fexperimental-strict-floating-point
Enables experimental strict floating point in LLVM.
.. option:: -ffine-grained-bitfield-accesses, -fno-fine-grained-bitfield-accesses
Use separate accesses for consecutive bitfield runs with legal widths and alignments.
.. option:: -fglobal-isel, -fexperimental-isel, -fno-global-isel
Enables the global instruction selector
.. option:: -finline-functions, -fno-inline-functions
Inline suitable functions
.. option:: -finline-hint-functions
Inline functions which are (explicitly or implicitly) marked inline
.. option:: -flegacy-pass-manager, -fno-legacy-pass-manager
Use the legacy pass manager in LLVM (deprecated, to be removed in a future release)
.. option:: -fno-crash-diagnostics
Disable auto-generation of preprocessed source files and a script for reproduction during a clang crash
.. option:: -fno-sanitize-ignorelist, -fno-sanitize-blacklist
Don't use ignorelist file for sanitizers
.. option:: -fparse-all-comments
.. option:: -frecord-command-line, -fno-record-command-line, -frecord-gcc-switches
.. option:: -fsanitize-address-destructor=<arg>
Set destructor type used in ASan instrumentation
.. option:: -fsanitize-address-field-padding=<arg>
Level of field padding for AddressSanitizer
.. option:: -fsanitize-address-globals-dead-stripping
Enable linker dead stripping of globals in AddressSanitizer
.. option:: -fsanitize-address-outline-instrumentation, -fno-sanitize-address-outline-instrumentation
Always generate function calls for address sanitizer instrumentation
.. option:: -fsanitize-address-poison-custom-array-cookie, -fno-sanitize-address-poison-custom-array-cookie
Enable poisoning array cookies when using custom operator new\[\] in AddressSanitizer
.. option:: -fsanitize-address-use-after-return=<mode>
Select the mode of detecting stack use-after-return in AddressSanitizer: never \| runtime (default) \| always
.. option:: -fsanitize-address-use-after-scope, -fno-sanitize-address-use-after-scope
Enable use-after-scope detection in AddressSanitizer
.. option:: -fsanitize-address-use-odr-indicator, -fno-sanitize-address-use-odr-indicator
Enable ODR indicator globals to avoid false ODR violation reports in partially sanitized programs at the cost of an increase in binary size
.. option:: -fsanitize-cfi-canonical-jump-tables, -fno-sanitize-cfi-canonical-jump-tables
Make the jump table addresses canonical in the symbol table
.. option:: -fsanitize-cfi-cross-dso, -fno-sanitize-cfi-cross-dso
Enable control flow integrity (CFI) checks for cross-DSO calls.
.. option:: -fsanitize-cfi-icall-generalize-pointers
Generalize pointers in CFI indirect call type signature checks
.. option:: -fsanitize-coverage-allowlist=<arg>, -fsanitize-coverage-whitelist=<arg>
Restrict sanitizer coverage instrumentation exclusively to modules and functions that match the provided special case list, except the blocked ones
.. option:: -fsanitize-coverage-ignorelist=<arg>, -fsanitize-coverage-blacklist=<arg>
Disable sanitizer coverage instrumentation for modules and functions that match the provided special case list, even the allowed ones
.. option:: -fsanitize-coverage=<arg1>,<arg2>..., -fno-sanitize-coverage=<arg1>,<arg2>...
Specify the type of coverage instrumentation for Sanitizers
.. option:: -fsanitize-hwaddress-abi=<arg>
Select the HWAddressSanitizer ABI to target (interceptor or platform, default interceptor). This option is currently unused.
.. option:: -fsanitize-hwaddress-experimental-aliasing, -fno-sanitize-hwaddress-experimental-aliasing
Enable aliasing mode in HWAddressSanitizer
.. option:: -fsanitize-ignorelist=<arg>, -fsanitize-blacklist=<arg>
Path to ignorelist file for sanitizers
.. option:: -fsanitize-link-c++-runtime, -fno-sanitize-link-c++-runtime
.. option:: -fsanitize-link-runtime, -fno-sanitize-link-runtime
.. option:: -fsanitize-memory-track-origins, -fno-sanitize-memory-track-origins
Enable origins tracking in MemorySanitizer
.. program:: clang1
.. option:: -fsanitize-memory-track-origins=<arg>
.. program:: clang
Enable origins tracking in MemorySanitizer
.. option:: -fsanitize-memory-use-after-dtor, -fno-sanitize-memory-use-after-dtor
Enable use-after-destroy detection in MemorySanitizer
.. option:: -fsanitize-minimal-runtime, -fno-sanitize-minimal-runtime
.. option:: -fsanitize-recover=<arg1>,<arg2>..., -fno-sanitize-recover=<arg1>,<arg2>..., -fsanitize-recover (equivalent to -fsanitize-recover=all)
Enable recovery for specified sanitizers
.. option:: -fsanitize-stats, -fno-sanitize-stats
Enable sanitizer statistics gathering.
.. option:: -fsanitize-thread-atomics, -fno-sanitize-thread-atomics
Enable atomic operations instrumentation in ThreadSanitizer (default)
.. option:: -fsanitize-thread-func-entry-exit, -fno-sanitize-thread-func-entry-exit
Enable function entry/exit instrumentation in ThreadSanitizer (default)
.. option:: -fsanitize-thread-memory-access, -fno-sanitize-thread-memory-access
Enable memory access instrumentation in ThreadSanitizer (default)
.. option:: -fsanitize-trap=<arg1>,<arg2>..., -fno-sanitize-trap=<arg1>,<arg2>..., -fsanitize-trap (equivalent to -fsanitize-trap=all), -fsanitize-undefined-trap-on-error (equivalent to -fsanitize-trap=undefined)
Enable trapping for specified sanitizers
.. option:: -fsanitize-undefined-strip-path-components=<number>
Strip (or keep only, if negative) a given number of path components when emitting check metadata.
.. option:: -fsanitize=<check>,<arg2>..., -fno-sanitize=<arg1>,<arg2>...
Turn on runtime checks for various forms of undefined or suspicious behavior. See user manual for available checks
.. option:: -moutline, -mno-outline
Enable function outlining (AArch64 only)
.. option:: -moutline-atomics, -mno-outline-atomics
Generate local calls to out-of-line atomic operations
.. option:: --param <arg>, --param=<arg>
.. option:: -print-supported-cpus, --print-supported-cpus, -mcpu=?, -mtune=?
Print supported cpu models for the given target (if target is not specified, it will print the supported cpus for the default target)
.. option:: -std=<arg>, --std=<arg>, --std <arg>
Language standard to compile for
Preprocessor flags
~~~~~~~~~~~~~~~~~~
Flags controlling the behavior of the Clang preprocessor.
.. option:: -C, --comments
Include comments in preprocessed output
.. option:: -CC, --comments-in-macros
Include comments from within macros in preprocessed output
.. option:: -D<macro>=<value>, --define-macro <arg>, --define-macro=<arg>
Define <macro> to <value> (or 1 if <value> omitted)
.. option:: -H, --trace-includes
Show header includes and nesting depth
.. option:: -P, --no-line-commands
Disable linemarker output in -E mode
.. option:: -U<macro>, --undefine-macro <arg>, --undefine-macro=<arg>
Undefine macro <macro>
.. option:: -Wp,<arg>,<arg2>...
Pass the comma separated arguments in <arg> to the preprocessor
.. option:: -Xpreprocessor <arg>
Pass <arg> to the preprocessor
Include path management
-----------------------
Flags controlling how ``#include``\s are resolved to files.
2020-08-02 06:54:11 +08:00
.. option:: -I<dir>, --include-directory <arg>, --include-directory=<arg>
Add directory to include search path. For C++ inputs, if
there are multiple -I options, these directories are searched
in the order they are given before the standard system directories
are searched. If the same directory is in the SYSTEM include search
paths, for example if also specified with -isystem, the -I option
will be ignored
.. option:: -I-, --include-barrier
Restrict all prior -I flags to double-quoted inclusion and remove current directory from include path
.. option:: --amdgpu-arch-tool=<arg>
Tool used for detecting AMD GPU arch in the system.
.. option:: --cuda-path-ignore-env
Ignore environment variables to detect CUDA installation
.. option:: --cuda-path=<arg>
CUDA installation path
.. option:: -cxx-isystem<directory>
Add directory to the C++ SYSTEM include search path
.. option:: -fbuild-session-file=<file>
Use the last modification time of <file> as the build session timestamp
.. option:: -fbuild-session-timestamp=<time since Epoch in seconds>
Time when the current build session started
.. option:: -fmodule-file=\[<name>=\]<file>
Specify the mapping of module name to precompiled module file, or load a module file if name is omitted.
.. option:: -fmodules-cache-path=<directory>
Specify the module cache path
.. option:: -fmodules-disable-diagnostic-validation
Disable validation of the diagnostic options when loading the module
.. option:: -fmodules-prune-after=<seconds>
Specify the interval (in seconds) after which a module file will be considered unused
.. option:: -fmodules-prune-interval=<seconds>
Specify the interval (in seconds) between attempts to prune the module cache
.. option:: -fmodules-user-build-path <directory>
Specify the module user build path
.. option:: -fmodules-validate-once-per-build-session
Don't verify input files for the modules if the module has been successfully validated or loaded during this build session
.. option:: -fmodules-validate-system-headers, -fno-modules-validate-system-headers
Validate the system headers that a module depends on when loading the module
.. option:: -fprebuilt-module-path=<directory>
Specify the prebuilt module path
.. option:: --hip-path=<arg>
HIP runtime installation path, used for finding HIP version and adding HIP include path.
.. option:: -idirafter<arg>, --include-directory-after <arg>, --include-directory-after=<arg>
Add directory to AFTER include search path
.. option:: -iframework<arg>
Add directory to SYSTEM framework search path
.. option:: -iframeworkwithsysroot<directory>
Add directory to SYSTEM framework search path, absolute paths are relative to -isysroot
.. option:: -imacros<file>, --imacros<file>, --imacros=<arg>
Include macros from file before parsing
.. option:: -include<file>, --include<file>, --include=<arg>
Include file before parsing
.. option:: -include-pch <file>
Include precompiled header file
.. option:: -iprefix<dir>, --include-prefix <arg>, --include-prefix=<arg>
Set the -iwithprefix/-iwithprefixbefore prefix
.. option:: -iquote<directory>
Add directory to QUOTE include search path
.. option:: -isysroot<dir>
Set the system root directory (usually /)
.. option:: -isystem<directory>
Add directory to SYSTEM include search path
.. option:: -isystem-after<directory>
Add directory to end of the SYSTEM include search path
.. option:: -ivfsoverlay<arg>
Overlay the virtual filesystem described by file over the real file system
.. option:: -iwithprefix<dir>, --include-with-prefix <arg>, --include-with-prefix-after <arg>, --include-with-prefix-after=<arg>, --include-with-prefix=<arg>
Set directory to SYSTEM include search path with prefix
.. option:: -iwithprefixbefore<dir>, --include-with-prefix-before <arg>, --include-with-prefix-before=<arg>
Set directory to include search path with prefix
.. option:: -iwithsysroot<directory>
Add directory to SYSTEM include search path, absolute paths are relative to -isysroot
.. option:: --libomptarget-amdgpu-bc-path=<arg>, --libomptarget-amdgcn-bc-path=<arg>
Path to libomptarget-amdgcn bitcode library
.. option:: --libomptarget-nvptx-bc-path=<arg>
Path to libomptarget-nvptx bitcode library
.. option:: --ptxas-path=<arg>
Path to ptxas (used for compiling CUDA code)
.. option:: --rocm-path=<arg>
ROCm installation path, used for finding and automatically linking required bitcode libraries.
.. program:: clang1
.. option:: -stdlib++-isystem<directory>
.. program:: clang
Use directory as the C++ standard library include path
.. option:: --system-header-prefix=<prefix>, --no-system-header-prefix=<prefix>, --system-header-prefix <arg>
Treat all #include paths starting with <prefix> as including a system header.
Dependency file generation
--------------------------
Flags controlling generation of a dependency file for ``make``-like build
systems.
.. option:: -M, --dependencies
Like -MD, but also implies -E and writes to stdout by default
.. option:: -MD, --write-dependencies
Write a depfile containing user and system headers
.. option:: -MF<file>
Write depfile output from -MMD, -MD, -MM, or -M to <file>
.. option:: -MG, --print-missing-file-dependencies
Add missing headers to depfile
.. option:: -MJ<arg>
Write a compilation database entry per input
.. option:: -MM, --user-dependencies
Like -MMD, but also implies -E and writes to stdout by default
.. option:: -MMD, --write-user-dependencies
Write a depfile containing user headers
.. option:: -MP
Create phony target for each dependency (other than main file)
.. option:: -MQ<arg>
Specify name of main file output to quote in depfile
.. option:: -MT<arg>
Specify name of main file output in depfile
.. option:: -MV
Use NMake/Jom format for the depfile
Dumping preprocessor state
--------------------------
Flags allowing the state of the preprocessor to be dumped in various ways.
.. option:: -d
.. program:: clang1
.. option:: -d<arg>
.. program:: clang
.. option:: -dD
Print macro definitions in -E mode in addition to normal output
.. option:: -dI
Print include directives in -E mode in addition to normal output
.. option:: -dM
Print macro definitions in -E mode instead of normal output
Diagnostic flags
~~~~~~~~~~~~~~~~
Flags controlling which warnings, errors, and remarks Clang will generate.
See the :doc:`full list of warning and remark flags <DiagnosticsReference>`.
.. option:: -R<remark>
Enable the specified remark
.. option:: -Rpass-analysis=<arg>
Report transformation analysis from optimization passes whose name matches the given POSIX regular expression
.. option:: -Rpass-missed=<arg>
Report missed transformations by optimization passes whose name matches the given POSIX regular expression
.. option:: -Rpass=<arg>
Report transformations performed by optimization passes whose name matches the given POSIX regular expression
.. option:: -W<warning>, --extra-warnings, --warn-<arg>, --warn-=<arg>
Enable the specified warning
.. option:: -Wdeprecated, -Wno-deprecated
Enable warnings for deprecated constructs and define \_\_DEPRECATED
.. option:: -Wframe-larger-than=<arg>, -Wframe-larger-than
.. option:: -Wnonportable-cfstrings<arg>, -Wno-nonportable-cfstrings<arg>
Target-independent compilation options
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. option:: -fPIC, -fno-PIC
.. option:: -fPIE, -fno-PIE
.. option:: -faccess-control, -fno-access-control
.. option:: -faddrsig, -fno-addrsig
Emit an address-significance table
.. option:: -falign-functions, -fno-align-functions
.. program:: clang1
.. option:: -falign-functions=<arg>
.. program:: clang
.. option:: -falign-loops=<N>
N must be a power of two. Align loops to the boundary
.. program:: clang1
.. option:: -faligned-allocation, -faligned-new, -fno-aligned-allocation
.. program:: clang
Enable C++17 aligned allocation functions
.. option:: -fallow-editor-placeholders, -fno-allow-editor-placeholders
Treat editor placeholders as valid source code
.. option:: -fallow-unsupported
.. option:: -faltivec, -fno-altivec
.. option:: -faltivec-src-compat=<arg>
Source-level compatibility for Altivec vectors (for PowerPC targets). This includes results of vector comparison (scalar for 'xl', vector for 'gcc') as well as behavior when initializing with a scalar (splatting for 'xl', element zero only for 'gcc'). For 'mixed', the compatibility is as 'gcc' for 'vector bool/vector pixel' and as 'xl' for other types. Current default is 'mixed'.
.. option:: -fansi-escape-codes
Use ANSI escape codes for diagnostics
.. option:: -fapple-kext, -findirect-virtual-calls, -fterminated-vtables
Use Apple's kernel extensions ABI
.. option:: -fapple-link-rtlib
Force linking the clang builtins runtime library
.. option:: -fapple-pragma-pack, -fno-apple-pragma-pack
Enable Apple gcc-compatible #pragma pack handling
.. option:: -fapplication-extension, -fno-application-extension
Restrict code to those available for App Extensions
.. option:: -fapprox-func, -fno-approx-func
Allow certain math function calls to be replaced with an approximately equivalent calculation
.. option:: -fasm, -fno-asm
.. option:: -fasm-blocks, -fno-asm-blocks
.. option:: -fassociative-math, -fno-associative-math
.. option:: -fassume-sane-operator-new, -fno-assume-sane-operator-new
.. option:: -fast
.. option:: -fastcp
.. option:: -fastf
.. option:: -fasync-exceptions, -fno-async-exceptions
Enable EH Asynchronous exceptions
.. option:: -fasynchronous-unwind-tables, -fno-asynchronous-unwind-tables
.. option:: -fautolink, -fno-autolink
.. option:: -fbasic-block-sections=<arg>
Generate labels for each basic block or place each basic block or a subset of basic blocks in its own section.
.. option:: -fbinutils-version=<major.minor>
Produced object files can use all ELF features supported by this binutils version and newer. If -fno-integrated-as is specified, the generated assembly will consider GNU as support. 'none' means that all ELF features can be used, regardless of binutils support. Defaults to 2.26.
.. option:: -fblocks, -fno-blocks
Enable the 'blocks' language feature
.. option:: -fbootclasspath=<arg>, --bootclasspath <arg>, --bootclasspath=<arg>
.. option:: -fborland-extensions, -fno-borland-extensions
Accept non-standard constructs supported by the Borland compiler
.. option:: -fbracket-depth=<arg>
.. option:: -fbuiltin, -fno-builtin
.. option:: -fbuiltin-module-map
Load the clang builtins module map file.
.. program:: clang1
.. option:: -fc++-static-destructors, -fno-c++-static-destructors
.. program:: clang
.. option:: -fcaret-diagnostics, -fno-caret-diagnostics
.. option:: -fcf-protection=<arg>, -fcf-protection (equivalent to -fcf-protection=full)
Instrument control-flow architecture protection. Options: return, branch, full, none.
.. option:: -fcf-runtime-abi=<arg>
.. option:: -fchar8\_t, -fno-char8\_t
Enable C++ builtin type char8\_t
.. option:: -fclasspath=<arg>, --CLASSPATH <arg>, --CLASSPATH=<arg>, --classpath <arg>, --classpath=<arg>
.. option:: -fcolor-diagnostics, -fno-color-diagnostics
Enable colors in diagnostics
.. option:: -fcommon, -fno-common
Place uninitialized global variables in a common block
.. option:: -fcompile-resource=<arg>, --resource <arg>, --resource=<arg>
.. option:: -fconstant-cfstrings, -fno-constant-cfstrings
.. option:: -fconstant-string-class=<arg>
.. option:: -fconstexpr-backtrace-limit=<arg>
.. option:: -fconstexpr-depth=<arg>
.. option:: -fconstexpr-steps=<arg>
.. option:: -fconvergent-functions
Assume functions may be convergent
.. option:: -fcoroutines-ts, -fno-coroutines-ts
Enable support for the C++ Coroutines TS
.. option:: -fcoverage-compilation-dir=<arg>
The compilation directory to embed in the coverage mapping.
.. option:: -fcoverage-mapping, -fno-coverage-mapping
Generate coverage mapping to enable code coverage analysis
.. option:: -fcoverage-prefix-map=<arg>
remap file source paths in coverage mapping
.. option:: -fcreate-profile
.. option:: -fcs-profile-generate
Generate instrumented code to collect context sensitive execution counts into default.profraw (overridden by LLVM\_PROFILE\_FILE env var)
.. program:: clang1
.. option:: -fcs-profile-generate=<directory>
.. program:: clang
Generate instrumented code to collect context sensitive execution counts into <directory>/default.profraw (overridden by LLVM\_PROFILE\_FILE env var)
.. option:: -fcuda-approx-transcendentals, -fno-cuda-approx-transcendentals
Use approximate transcendental functions
.. option:: -fcuda-short-ptr, -fno-cuda-short-ptr
Use 32-bit pointers for accessing const/local/shared address spaces
.. option:: -fcxx-exceptions, -fno-cxx-exceptions
Enable C++ exceptions
.. option:: -fcxx-modules, -fno-cxx-modules
Enable modules for C++
.. option:: -fdata-sections, -fno-data-sections
Place each data in its own section
.. option:: -fdebug-compilation-dir=<arg>, -fdebug-compilation-dir <arg>
The compilation directory to embed in the debug info
.. option:: -fdebug-default-version=<arg>
Default DWARF version to use, if a -g option caused DWARF debug info to be produced
.. option:: -fdebug-info-for-profiling, -fno-debug-info-for-profiling
Emit extra debug info to make sample profile more accurate
.. option:: -fdebug-macro, -fno-debug-macro
Emit macro debug information
.. option:: -fdebug-pass-arguments
.. option:: -fdebug-pass-structure
.. option:: -fdebug-prefix-map=<arg>
remap file source paths in debug info
.. option:: -fdebug-ranges-base-address, -fno-debug-ranges-base-address
Use DWARF base address selection entries in .debug\_ranges
.. option:: -fdebug-types-section, -fno-debug-types-section
Place debug types in their own section (ELF Only)
.. option:: -fdelayed-template-parsing, -fno-delayed-template-parsing
Parse templated function definitions at the end of the translation unit
.. option:: -fdelete-null-pointer-checks, -fno-delete-null-pointer-checks
Treat usage of null pointers as undefined behavior (default)
.. option:: -fdenormal-fp-math=<arg>
.. option:: -fdiagnostics-absolute-paths
Print absolute paths in diagnostics
.. option:: -fdiagnostics-color, -fno-diagnostics-color
.. program:: clang1
.. option:: -fdiagnostics-color=<arg>
.. program:: clang
.. option:: -fdiagnostics-hotness-threshold=<value>
Prevent optimization remarks from being output if they do not have at least this profile count. Use 'auto' to apply the threshold from profile summary
.. option:: -fdiagnostics-show-hotness, -fno-diagnostics-show-hotness
Enable profile hotness information in diagnostic line
.. option:: -fdiagnostics-show-note-include-stack, -fno-diagnostics-show-note-include-stack
Display include stacks for diagnostic notes
.. option:: -fdiagnostics-show-option, -fno-diagnostics-show-option
Print option name with mappable diagnostics
.. option:: -fdiagnostics-show-template-tree
Print a template comparison tree for differing templates
.. option:: -fdigraphs, -fno-digraphs
Enable alternative token representations '<:', ':>', '<%', '%>', '%:', '%:%:' (default)
.. option:: -fdirect-access-external-data, -fno-direct-access-external-data
Don't use GOT indirection to reference external data symbols
.. option:: -fdollars-in-identifiers, -fno-dollars-in-identifiers
Allow '$' in identifiers
.. option:: -fdouble-square-bracket-attributes, -fno-double-square-bracket-attributes
Enable '\[\[\]\]' attributes in all C and C++ language modes
.. option:: -fdwarf-directory-asm, -fno-dwarf-directory-asm
.. option:: -fdwarf-exceptions
Use DWARF style exceptions
.. option:: -felide-constructors, -fno-elide-constructors
.. option:: -feliminate-unused-debug-symbols, -fno-eliminate-unused-debug-symbols
.. option:: -feliminate-unused-debug-types, -fno-eliminate-unused-debug-types
Do not emit debug info for defined but unused types
.. option:: -fembed-bitcode=<option>, -fembed-bitcode (equivalent to -fembed-bitcode=all), -fembed-bitcode-marker (equivalent to -fembed-bitcode=marker)
Embed LLVM bitcode (option: off, all, bitcode, marker)
.. option:: -fembed-offload-object=<arg>
Embed Offloading device-side binary into host object file as a section.
.. option:: -femit-all-decls
Emit all declarations, even if unused
.. option:: -femulated-tls, -fno-emulated-tls
Use emutls functions to access thread\_local variables
.. option:: -fenable-matrix
Enable matrix data type and related builtin functions
.. option:: -fencoding=<arg>, --encoding <arg>, --encoding=<arg>
.. option:: -ferror-limit=<arg>
.. option:: -fescaping-block-tail-calls, -fno-escaping-block-tail-calls
.. option:: -fexceptions, -fno-exceptions
Enable support for exception handling
.. option:: -fexec-charset=<arg>
.. option:: -fexperimental-new-constant-interpreter
Enable the experimental new constant interpreter
.. option:: -fextdirs=<arg>, --extdirs <arg>, --extdirs=<arg>
.. option:: -fextend-arguments=<arg>
Controls how scalar integer arguments are extended in calls to unprototyped and varargs functions
.. option:: -ffast-math, -fno-fast-math
Allow aggressive, lossy floating-point optimizations
.. option:: -ffile-compilation-dir=<arg>
The compilation directory to embed in the debug info and coverage mapping.
.. option:: -ffile-prefix-map=<arg>
remap file source paths in debug info, predefined preprocessor macros and \_\_builtin\_FILE()
.. option:: -ffinite-loops, -fno-finite-loops
Assume all loops are finite.
.. option:: -ffinite-math-only, -fno-finite-math-only
.. option:: -ffixed-point, -fno-fixed-point
Enable fixed point types
.. option:: -ffor-scope, -fno-for-scope
.. option:: -fforce-dwarf-frame, -fno-force-dwarf-frame
Always emit a debug frame section
.. option:: -fforce-emit-vtables, -fno-force-emit-vtables
Emits more virtual tables to improve devirtualization
.. option:: -fforce-enable-int128, -fno-force-enable-int128
Enable support for int128\_t type
.. option:: -ffp-contract=<arg>
Form fused FP ops (e.g. FMAs): fast (fuses across statements disregarding pragmas) \| on (only fuses in the same statement unless dictated by pragmas) \| off (never fuses) \| fast-honor-pragmas (fuses across statements unless diectated by pragmas). Default is 'fast' for CUDA, 'fast-honor-pragmas' for HIP, and 'on' otherwise.
.. option:: -ffp-exception-behavior=<arg>
Specifies the exception behavior of floating-point operations.
.. option:: -ffp-model=<arg>
Controls the semantics of floating-point calculations.
.. option:: -ffreestanding
Assert that the compilation takes place in a freestanding environment
.. option:: -ffunction-sections, -fno-function-sections
Place each function in its own section
.. option:: -fgnu-inline-asm, -fno-gnu-inline-asm
.. option:: -fgnu-keywords, -fno-gnu-keywords
Allow GNU-extension keywords regardless of language standard
.. option:: -fgnu-runtime
Generate output compatible with the standard GNU Objective-C runtime
.. option:: -fgnu89-inline, -fno-gnu89-inline
Use the gnu89 inline semantics
.. option:: -fgnuc-version=<arg>
Sets various macros to claim compatibility with the given GCC version (default is 4.2.1)
.. option:: -fgpu-allow-device-init, -fno-gpu-allow-device-init
Allow device side init function in HIP (experimental)
.. option:: -fgpu-defer-diag, -fno-gpu-defer-diag
Defer host/device related diagnostic messages for CUDA/HIP
.. option:: -fgpu-rdc, -fcuda-rdc, -fno-gpu-rdc
Generate relocatable device code, also known as separate compilation mode
.. option:: -fgpu-sanitize, -fno-gpu-sanitize
Enable sanitizer for AMDGPU target
.. option:: -fhip-fp32-correctly-rounded-divide-sqrt, -fno-hip-fp32-correctly-rounded-divide-sqrt
Specify that single precision floating-point divide and sqrt used in the program source are correctly rounded (HIP device compilation only)
.. option:: -fhip-new-launch-api, -fno-hip-new-launch-api
Use new kernel launching API for HIP
.. option:: -fhonor-infinities, -fhonor-infinites, -fno-honor-infinities
.. option:: -fhonor-nans, -fno-honor-nans
.. option:: -fhosted
.. option:: -fignore-exceptions
Enable support for ignoring exception handling constructs
.. option:: -fimplicit-module-maps, -fmodule-maps, -fno-implicit-module-maps
Implicitly search the file system for module map files.
.. option:: -fimplicit-modules, -fno-implicit-modules
.. option:: -finput-charset=<arg>
Specify the default character set for source files
.. option:: -finstrument-function-entry-bare
Instrument function entry only, after inlining, without arguments to the instrumentation call
.. option:: -finstrument-functions
Generate calls to instrument function entry and exit
.. option:: -finstrument-functions-after-inlining
Like -finstrument-functions, but insert the calls after inlining
.. option:: -fintegrated-as, -fno-integrated-as, -integrated-as
Enable the integrated assembler
.. option:: -fintegrated-cc1, -fno-integrated-cc1
Run cc1 in-process
.. option:: -fjump-tables, -fno-jump-tables
Use jump tables for lowering switches
.. option:: -fkeep-static-consts, -fno-keep-static-consts
Keep static const variables if unused
.. option:: -flax-vector-conversions=<arg>, -flax-vector-conversions (equivalent to -flax-vector-conversions=integer), -fno-lax-vector-conversions (equivalent to -flax-vector-conversions=none)
Enable implicit vector bit-casts
.. option:: -flimited-precision=<arg>
.. option:: -flto-jobs=<arg>
Controls the backend parallelism of -flto=thin (default of 0 means the number of threads will be derived from the number of CPUs detected)
.. option:: -flto=<arg>, -flto (equivalent to -flto=full), -flto=auto (equivalent to -flto=full), -flto=jobserver (equivalent to -flto=full)
Set LTO mode to either 'full' or 'thin'
.. option:: -fmacro-backtrace-limit=<arg>
.. option:: -fmacro-prefix-map=<arg>
remap file source paths in predefined preprocessor macros and \_\_builtin\_FILE()
.. option:: -fmath-errno, -fno-math-errno
Require math functions to indicate errors by setting errno
.. option:: -fmax-tokens=<arg>
Max total number of preprocessed tokens for -Wmax-tokens.
.. option:: -fmax-type-align=<arg>
Specify the maximum alignment to enforce on pointers lacking an explicit alignment
.. option:: -fmemory-profile, -fno-memory-profile
Enable heap memory profiling
.. program:: clang1
.. option:: -fmemory-profile=<directory>
.. program:: clang
Enable heap memory profiling and dump results into <directory>
.. option:: -fmerge-all-constants, -fno-merge-all-constants
Allow merging of constants
.. option:: -fmessage-length=<arg>
Format message diagnostics so that they fit within N columns
.. option:: -fminimize-whitespace, -fno-minimize-whitespace
Minimize whitespace when emitting preprocessor output
.. option:: -fmodule-file-deps, -fno-module-file-deps
.. option:: -fmodule-map-file=<file>
Load this module map file
.. option:: -fmodule-name=<name>, -fmodule-implementation-of <arg>
Specify the name of the module to build
.. option:: -fmodules, -fno-modules
Enable the 'modules' language feature
.. option:: -fmodules-decluse, -fno-modules-decluse
Require declaration of modules used within a module
.. option:: -fmodules-ignore-macro=<arg>
Ignore the definition of the given macro when building and loading modules
.. option:: -fmodules-search-all, -fno-modules-search-all
Search even non-imported modules to resolve references
.. option:: -fmodules-strict-decluse
Like -fmodules-decluse but requires all headers to be in modules
.. option:: -fmodules-ts
Enable support for the C++ Modules TS
.. option:: -fmodules-validate-input-files-content
Validate PCM input files based on content if mtime differs
.. option:: -fms-compatibility, -fno-ms-compatibility
Enable full Microsoft Visual C++ compatibility
.. option:: -fms-compatibility-version=<arg>
Dot-separated value representing the Microsoft compiler version number to report in \_MSC\_VER (0 = don't define it (default))
.. option:: -fms-extensions, -fno-ms-extensions
Accept some non-standard constructs supported by the Microsoft compiler
.. option:: -fms-hotpatch
Ensure that all functions can be hotpatched at runtime
.. option:: -fms-memptr-rep=<arg>
.. option:: -fms-volatile
.. option:: -fmsc-version=<arg>
Microsoft compiler version number to report in \_MSC\_VER (0 = don't define it (default))
.. option:: -fmudflap
.. option:: -fmudflapth
.. option:: -fnested-functions
.. option:: -fnew-alignment=<align>, -fnew-alignment <arg>
Specifies the largest alignment guaranteed by '::operator new(size\_t)'
.. option:: -fnew-infallible, -fno-new-infallible
[clang] Add support for optional flag -fnew-infallible to restrict exception propagation The declaration for the global new function in C++ is generated in the compiler front-end. When examining exception propagation, we found that this is the largest root throw site propagator requiring unwind code to be generated for callers up the stack. Allowing this to be handled immediately with termination stops upward propagation and leads to significantly less landing pads generated. This in turns leads to a performance and .text size win. With `-fnew-infallible` this annotates the declaration with `throw()` and `__attribute__((returns_nonnull))`. `throw()` allows the compiler to assume exceptions do not propagate out of new and eliminate it as a root throw site. Note that the definition of global new is user-replaceable so users should ensure that the one used follows these semantics. Measuring internally, we're seeing at 0.5% CPU win in one of our large internal FB workload. Measuring on clang self-build (cd0a1226b50081e86eb75a89d01e8782423971a0) we get: thinlto/ "dwarfehprepare.NumCleanupLandingPadsRemaining": 153494, "dwarfehprepare.NumNoUnwind": 26309, thinlto_newinfallible/ "dwarfehprepare.NumCleanupLandingPadsRemaining": 143660, "dwarfehprepare.NumNoUnwind": 28744, a 1-143660/153494 = 6.4% reduction in landing pads and a 28744/26309 = 9.3% increase in the number of nounwind functions. Testing: ninja check-all new test case to make sure these attributes are added correctly to global new. Reviewed By: urnathan Differential Revision: https://reviews.llvm.org/D105225
2021-08-03 06:44:10 +08:00
Enable treating throwing global C++ operator new as always returning valid memory (annotates with \_\_attribute\_\_((returns\_nonnull)) and throw()). This is detectable in source.
[clang] Add support for optional flag -fnew-infallible to restrict exception propagation The declaration for the global new function in C++ is generated in the compiler front-end. When examining exception propagation, we found that this is the largest root throw site propagator requiring unwind code to be generated for callers up the stack. Allowing this to be handled immediately with termination stops upward propagation and leads to significantly less landing pads generated. This in turns leads to a performance and .text size win. With `-fnew-infallible` this annotates the declaration with `throw()` and `__attribute__((returns_nonnull))`. `throw()` allows the compiler to assume exceptions do not propagate out of new and eliminate it as a root throw site. Note that the definition of global new is user-replaceable so users should ensure that the one used follows these semantics. Measuring internally, we're seeing at 0.5% CPU win in one of our large internal FB workload. Measuring on clang self-build (cd0a1226b50081e86eb75a89d01e8782423971a0) we get: thinlto/ "dwarfehprepare.NumCleanupLandingPadsRemaining": 153494, "dwarfehprepare.NumNoUnwind": 26309, thinlto_newinfallible/ "dwarfehprepare.NumCleanupLandingPadsRemaining": 143660, "dwarfehprepare.NumNoUnwind": 28744, a 1-143660/153494 = 6.4% reduction in landing pads and a 28744/26309 = 9.3% increase in the number of nounwind functions. Testing: ninja check-all new test case to make sure these attributes are added correctly to global new. Reviewed By: urnathan Differential Revision: https://reviews.llvm.org/D105225
2021-08-03 06:44:10 +08:00
.. option:: -fnext-runtime
.. option:: -fno-builtin-<arg>
Disable implicit builtin knowledge of a specific function
.. option:: -fno-elide-type
Do not elide types when printing diagnostics
.. option:: -fno-max-type-align
.. option:: -fno-strict-modules-decluse
.. option:: -fno-temp-file
Directly create compilation output files. This may lead to incorrect incremental builds if the compiler crashes
.. option:: -fno-working-directory
.. option:: -fno\_modules-validate-input-files-content
.. program:: clang1
.. option:: -fno\_pch-validate-input-files-content
.. program:: clang
.. option:: -fnoxray-link-deps
.. option:: -fobjc-abi-version=<arg>
.. option:: -fobjc-arc, -fno-objc-arc
Synthesize retain and release calls for Objective-C pointers
.. option:: -fobjc-arc-exceptions, -fno-objc-arc-exceptions
Use EH-safe code when synthesizing retains and releases in -fobjc-arc
.. option:: -fobjc-convert-messages-to-runtime-calls, -fno-objc-convert-messages-to-runtime-calls
.. option:: -fobjc-disable-direct-methods-for-testing
Ignore attribute objc\_direct so that direct methods can be tested
.. option:: -fobjc-encode-cxx-class-template-spec, -fno-objc-encode-cxx-class-template-spec
Fully encode c++ class template specialization
.. option:: -fobjc-exceptions, -fno-objc-exceptions
Enable Objective-C exceptions
.. option:: -fobjc-infer-related-result-type, -fno-objc-infer-related-result-type
.. option:: -fobjc-legacy-dispatch, -fno-objc-legacy-dispatch
.. option:: -fobjc-link-runtime
.. option:: -fobjc-nonfragile-abi, -fno-objc-nonfragile-abi
.. option:: -fobjc-nonfragile-abi-version=<arg>
.. option:: -fobjc-runtime=<arg>
Specify the target Objective-C runtime kind and version
.. option:: -fobjc-sender-dependent-dispatch
.. option:: -fobjc-weak, -fno-objc-weak
Enable ARC-style weak references in Objective-C
.. option:: -foffload-lto=<arg>, -foffload-lto (equivalent to -foffload-lto=full)
Set LTO mode to either 'full' or 'thin' for offload compilation
.. option:: -fomit-frame-pointer, -fno-omit-frame-pointer
.. option:: -fopenmp, -fno-openmp
Parse OpenMP pragmas and generate parallel code.
.. option:: -fopenmp-extensions, -fno-openmp-extensions
Enable all Clang extensions for OpenMP directives and clauses
.. option:: -fopenmp-implicit-rpath, -fno-openmp-implicit-rpath
Set rpath on OpenMP executables
.. option:: -fopenmp-simd, -fno-openmp-simd
Emit OpenMP code only for SIMD-based constructs.
.. option:: -fopenmp-target-debug, -fno-openmp-target-debug
Enable debugging in the OpenMP offloading device RTL
.. option:: -fopenmp-offload-mandatory
Indicate that offloading to the device is mandatory and do not generate host-fallback code.
.. option:: -fopenmp-version=<arg>
[OpenMP][OpenACC] Implement `ompx_hold` map type modifier extension in Clang (1/2) This patch implements Clang support for an original OpenMP extension we have developed to support OpenACC: the `ompx_hold` map type modifier. The next patch in this series, D106510, implements OpenMP runtime support. Consider the following example: ``` #pragma omp target data map(ompx_hold, tofrom: x) // holds onto mapping of x { foo(); // might have map(delete: x) #pragma omp target map(present, alloc: x) // x is guaranteed to be present printf("%d\n", x); } ``` The `ompx_hold` map type modifier above specifies that the `target data` directive holds onto the mapping for `x` throughout the associated region regardless of any `target exit data` directives executed during the call to `foo`. Thus, the presence assertion for `x` at the enclosed `target` construct cannot fail. (As usual, the standard OpenMP reference count for `x` must also reach zero before the data is unmapped.) Justification for inclusion in Clang and LLVM's OpenMP runtime: * The `ompx_hold` modifier supports OpenACC functionality (structured reference count) that cannot be achieved in standard OpenMP, as of 5.1. * The runtime implementation for `ompx_hold` (next patch) will thus be used by Flang's OpenACC support. * The Clang implementation for `ompx_hold` (this patch) as well as the runtime implementation are required for the Clang OpenACC support being developed as part of the ECP Clacc project, which translates OpenACC to OpenMP at the directive AST level. These patches are the first step in upstreaming OpenACC functionality from Clacc. * The Clang implementation for `ompx_hold` is also used by the tests in the runtime implementation. That syntactic support makes the tests more readable than low-level runtime calls can. Moreover, upstream Flang and Clang do not yet support OpenACC syntax sufficiently for writing the tests. * More generally, the Clang implementation enables a clean separation of concerns between OpenACC and OpenMP development in LLVM. That is, LLVM's OpenMP developers can discuss, modify, and debug LLVM's extended OpenMP implementation and test suite without directly considering OpenACC's language and execution model, which can be handled by LLVM's OpenACC developers. * OpenMP users might find the `ompx_hold` modifier useful, as in the above example. See new documentation introduced by this patch in `openmp/docs` for more detail on the functionality of this extension and its relationship with OpenACC. For example, it explains how the runtime must support two reference counts, as specified by OpenACC. Clang recognizes `ompx_hold` unless `-fno-openmp-extensions`, a new command-line option introduced by this patch, is specified. Reviewed By: ABataev, jdoerfert, protze.joachim, grokos Differential Revision: https://reviews.llvm.org/D106509
2021-09-01 03:17:07 +08:00
Set OpenMP version (e.g. 45 for OpenMP 4.5, 50 for OpenMP 5.0). Default value is 50.
[OpenMP][OpenACC] Implement `ompx_hold` map type modifier extension in Clang (1/2) This patch implements Clang support for an original OpenMP extension we have developed to support OpenACC: the `ompx_hold` map type modifier. The next patch in this series, D106510, implements OpenMP runtime support. Consider the following example: ``` #pragma omp target data map(ompx_hold, tofrom: x) // holds onto mapping of x { foo(); // might have map(delete: x) #pragma omp target map(present, alloc: x) // x is guaranteed to be present printf("%d\n", x); } ``` The `ompx_hold` map type modifier above specifies that the `target data` directive holds onto the mapping for `x` throughout the associated region regardless of any `target exit data` directives executed during the call to `foo`. Thus, the presence assertion for `x` at the enclosed `target` construct cannot fail. (As usual, the standard OpenMP reference count for `x` must also reach zero before the data is unmapped.) Justification for inclusion in Clang and LLVM's OpenMP runtime: * The `ompx_hold` modifier supports OpenACC functionality (structured reference count) that cannot be achieved in standard OpenMP, as of 5.1. * The runtime implementation for `ompx_hold` (next patch) will thus be used by Flang's OpenACC support. * The Clang implementation for `ompx_hold` (this patch) as well as the runtime implementation are required for the Clang OpenACC support being developed as part of the ECP Clacc project, which translates OpenACC to OpenMP at the directive AST level. These patches are the first step in upstreaming OpenACC functionality from Clacc. * The Clang implementation for `ompx_hold` is also used by the tests in the runtime implementation. That syntactic support makes the tests more readable than low-level runtime calls can. Moreover, upstream Flang and Clang do not yet support OpenACC syntax sufficiently for writing the tests. * More generally, the Clang implementation enables a clean separation of concerns between OpenACC and OpenMP development in LLVM. That is, LLVM's OpenMP developers can discuss, modify, and debug LLVM's extended OpenMP implementation and test suite without directly considering OpenACC's language and execution model, which can be handled by LLVM's OpenACC developers. * OpenMP users might find the `ompx_hold` modifier useful, as in the above example. See new documentation introduced by this patch in `openmp/docs` for more detail on the functionality of this extension and its relationship with OpenACC. For example, it explains how the runtime must support two reference counts, as specified by OpenACC. Clang recognizes `ompx_hold` unless `-fno-openmp-extensions`, a new command-line option introduced by this patch, is specified. Reviewed By: ABataev, jdoerfert, protze.joachim, grokos Differential Revision: https://reviews.llvm.org/D106509
2021-09-01 03:17:07 +08:00
.. program:: clang1
.. option:: -fopenmp=<arg>
.. program:: clang
.. option:: -foperator-arrow-depth=<arg>
.. option:: -foperator-names, -fno-operator-names
.. option:: -foptimization-record-file=<file>
Specify the output name of the file containing the optimization remarks. Implies -fsave-optimization-record. On Darwin platforms, this cannot be used with multiple -arch <arch> options.
.. option:: -foptimization-record-passes=<regex>
Only include passes which match a specified regular expression in the generated optimization record (by default, include all passes)
.. option:: -foptimize-sibling-calls, -fno-optimize-sibling-calls
.. option:: -forder-file-instrumentation
Generate instrumented code to collect order file into default.profraw file (overridden by '=' form of option or LLVM\_PROFILE\_FILE env var)
.. option:: -foutput-class-dir=<arg>, --output-class-directory <arg>, --output-class-directory=<arg>
.. option:: -fpack-struct, -fno-pack-struct
.. program:: clang1
.. option:: -fpack-struct=<arg>
.. program:: clang
Specify the default maximum struct packing alignment
.. option:: -fpascal-strings, -fno-pascal-strings, -mpascal-strings
Recognize and construct Pascal-style string literals
.. option:: -fpass-plugin=<dsopath>
Load pass plugin from a dynamic shared object file (only with new pass manager).
.. option:: -fpatchable-function-entry=<N,M>
Generate M NOPs before function entry and N-M NOPs after function entry
.. option:: -fpcc-struct-return
Override the default ABI to return all structs on the stack
.. option:: -fpch-codegen, -fno-pch-codegen
Generate code for uses of this PCH that assumes an explicit object file will be built for the PCH
.. option:: -fpch-debuginfo, -fno-pch-debuginfo
Generate debug info for types in an object file built from this PCH and do not generate them elsewhere
.. option:: -fpch-instantiate-templates, -fno-pch-instantiate-templates
Instantiate templates already while building a PCH
.. option:: -fpch-preprocess
.. option:: -fpch-validate-input-files-content
Validate PCH input files based on content if mtime differs
.. option:: -fpic, -fno-pic
.. option:: -fpie, -fno-pie
.. option:: -fplt, -fno-plt
.. option:: -fplugin=<dsopath>
Load the named plugin (dynamic shared object)
.. option:: -fprebuilt-implicit-modules, -fno-prebuilt-implicit-modules
Look up implicit modules in the prebuilt module path
.. option:: -fpreserve-as-comments, -fno-preserve-as-comments
.. option:: -fproc-stat-report<arg>
Print subprocess statistics
.. program:: clang1
.. option:: -fproc-stat-report=<arg>
.. program:: clang
Save subprocess statistics to the given file
.. option:: -fprofile-arcs, -fno-profile-arcs
.. option:: -fprofile-dir=<arg>
.. option:: -fprofile-exclude-files=<arg>
Instrument only functions from files where names don't match all the regexes separated by a semi-colon
.. option:: -fprofile-filter-files=<arg>
Instrument only functions from files where names match any regex separated by a semi-colon
.. option:: -fprofile-generate, -fno-profile-generate
Generate instrumented code to collect execution counts into default.profraw (overridden by LLVM\_PROFILE\_FILE env var)
.. program:: clang1
.. option:: -fprofile-generate=<directory>
.. program:: clang
Generate instrumented code to collect execution counts into <directory>/default.profraw (overridden by LLVM\_PROFILE\_FILE env var)
.. option:: -fprofile-instr-generate, -fno-profile-instr-generate
Generate instrumented code to collect execution counts into default.profraw file (overridden by '=' form of option or LLVM\_PROFILE\_FILE env var)
.. program:: clang1
.. option:: -fprofile-instr-generate=<file>
.. program:: clang
Generate instrumented code to collect execution counts into <file> (overridden by LLVM\_PROFILE\_FILE env var)
.. option:: -fprofile-instr-use, -fno-profile-instr-use, -fprofile-use
.. program:: clang1
.. option:: -fprofile-instr-use=<arg>
.. program:: clang
Use instrumentation data for profile-guided optimization
.. option:: -fprofile-list=<arg>
Filename defining the list of functions/files to instrument
.. option:: -fprofile-remapping-file=<file>
Use the remappings described in <file> to match the profile data against names in the program
.. option:: -fprofile-sample-accurate, -fauto-profile-accurate, -fno-profile-sample-accurate
Specifies that the sample profile is accurate. If the sample
profile is accurate, callsites without profile samples are marked
as cold. Otherwise, treat callsites without profile samples as if
we have no profile
.. option:: -fprofile-sample-use, -fauto-profile, -fno-profile-sample-use
.. program:: clang1
.. option:: -fprofile-sample-use=<arg>, -fauto-profile=<arg>
.. program:: clang
Enable sample-based profile guided optimizations
.. option:: -fprofile-update=<method>
Set update method of profile counters (atomic,prefer-atomic,single)
.. program:: clang1
.. option:: -fprofile-use=<pathname>
.. program:: clang
Use instrumentation data for profile-guided optimization. If pathname is a directory, it reads from <pathname>/default.profdata. Otherwise, it reads from file <pathname>.
.. option:: -fprotect-parens, -fno-protect-parens
Determines whether the optimizer honors parentheses when floating-point expressions are evaluated
.. option:: -fpseudo-probe-for-profiling, -fno-pseudo-probe-for-profiling
Emit pseudo probes for sample profiling
.. option:: -freciprocal-math, -fno-reciprocal-math
Allow division operations to be reassociated
.. option:: -freg-struct-return
Override the default ABI to return small structs in registers
.. option:: -fregister-global-dtors-with-atexit, -fno-register-global-dtors-with-atexit
Use atexit or \_\_cxa\_atexit to register global destructors
.. option:: -frelaxed-template-template-args, -fno-relaxed-template-template-args
Enable C++17 relaxed template template argument matching
.. option:: -freroll-loops, -fno-reroll-loops
Turn on loop reroller
.. option:: -fretain-comments-from-system-headers
.. option:: -frewrite-imports, -fno-rewrite-imports
.. option:: -frewrite-includes, -fno-rewrite-includes
.. option:: -frewrite-map-file=<arg>
.. option:: -fropi, -fno-ropi
Generate read-only position independent code (ARM only)
.. option:: -frounding-math, -fno-rounding-math
.. option:: -frtti, -fno-rtti
.. option:: -frtti-data, -fno-rtti-data
.. option:: -frwpi, -fno-rwpi
Generate read-write position independent code (ARM only)
.. option:: -fsanitize-memory-param-retval, -fno-sanitize-memory-param-retval
Enable detection of uninitialized parameters and return values
.. option:: -fsave-optimization-record, -fno-save-optimization-record
Generate a YAML optimization record file
.. program:: clang1
.. option:: -fsave-optimization-record=<format>
.. program:: clang
Generate an optimization record file in a specific format
.. option:: -fseh-exceptions
Use SEH style exceptions
.. option:: -fsemantic-interposition, -fno-semantic-interposition
.. option:: -fshort-enums, -fno-short-enums
Allocate to an enum type only as many bytes as it needs for the declared range of possible values
.. option:: -fshort-wchar, -fno-short-wchar
Force wchar\_t to be a short unsigned int
.. option:: -fshow-column, -fno-show-column
.. option:: -fshow-overloads=<arg>
Which overload candidates to show when overload resolution fails: best\|all; defaults to all
.. option:: -fshow-source-location, -fno-show-source-location
.. option:: -fsignaling-math, -fno-signaling-math
.. option:: -fsigned-bitfields
.. option:: -fsigned-char, -fno-signed-char, --signed-char
char is signed
.. option:: -fsigned-zeros, -fno-signed-zeros
.. option:: -fsized-deallocation, -fno-sized-deallocation
Enable C++14 sized global deallocation functions
.. option:: -fsjlj-exceptions
Use SjLj style exceptions
.. option:: -fslp-vectorize, -fno-slp-vectorize, -ftree-slp-vectorize
Enable the superword-level parallelism vectorization passes
.. option:: -fspell-checking, -fno-spell-checking
.. option:: -fspell-checking-limit=<arg>
.. option:: -fsplit-dwarf-inlining, -fno-split-dwarf-inlining
Provide minimal debug info in the object/executable to facilitate online symbolication/stack traces in the absence of .dwo/.dwp files when using Split DWARF
.. option:: -fsplit-lto-unit, -fno-split-lto-unit
Enables splitting of the LTO unit
.. option:: -fsplit-machine-functions, -fno-split-machine-functions
Enable late function splitting using profile information (x86 ELF)
.. option:: -fsplit-stack, -fno-split-stack
Use segmented stack
.. option:: -fstack-clash-protection, -fno-stack-clash-protection
Enable stack clash protection
.. option:: -fstack-protector, -fno-stack-protector
Enable stack protectors for some functions vulnerable to stack smashing. This uses a loose heuristic which considers functions vulnerable if they contain a char (or 8bit integer) array or constant sized calls to alloca , which are of greater size than ssp-buffer-size (default: 8 bytes). All variable sized calls to alloca are considered vulnerable. A function with a stack protector has a guard value added to the stack frame that is checked on function exit. The guard value must be positioned in the stack frame such that a buffer overflow from a vulnerable variable will overwrite the guard value before overwriting the function's return address. The reference stack guard value is stored in a global variable.
.. option:: -fstack-protector-all
Enable stack protectors for all functions
.. option:: -fstack-protector-strong
Enable stack protectors for some functions vulnerable to stack smashing. Compared to -fstack-protector, this uses a stronger heuristic that includes functions containing arrays of any size (and any type), as well as any calls to alloca or the taking of an address from a local variable
.. option:: -fstack-size-section, -fno-stack-size-section
Emit section containing metadata on function stack sizes
.. option:: -fstack-usage
Emit .su file containing information on function stack sizes
.. option:: -fstandalone-debug, -fno-limit-debug-info, -fno-standalone-debug
Emit full debug info for all types used by the program
.. option:: -fstrict-aliasing, -fno-strict-aliasing
.. option:: -fstrict-enums, -fno-strict-enums
Enable optimizations based on the strict definition of an enum's value range
.. option:: -fstrict-float-cast-overflow, -fno-strict-float-cast-overflow
Assume that overflowing float-to-int casts are undefined (default)
.. option:: -fstrict-overflow, -fno-strict-overflow
.. option:: -fstrict-return, -fno-strict-return
.. option:: -fstrict-vtable-pointers, -fno-strict-vtable-pointers
Enable optimizations based on the strict rules for overwriting polymorphic C++ objects
.. option:: -fstruct-path-tbaa, -fno-struct-path-tbaa
.. option:: -fswift-async-fp=<option>
Control emission of Swift async extended frame info (option: auto, always, never)
.. option:: -fsymbol-partition=<arg>
.. option:: -ftabstop=<arg>
.. option:: -ftemplate-backtrace-limit=<arg>
.. option:: -ftemplate-depth-<arg>
.. option:: -ftemplate-depth=<arg>
.. option:: -ftest-coverage, -fno-test-coverage
.. option:: -fthin-link-bitcode=<arg>
Write minimized bitcode to <file> for the ThinLTO thin link only
.. option:: -fthinlto-index=<arg>
Perform ThinLTO importing using provided function summary index
.. option:: -fthreadsafe-statics, -fno-threadsafe-statics
.. option:: -ftime-report
.. program:: clang1
.. option:: -ftime-report=<arg>
.. program:: clang
(For new pass manager) "per-pass": one report for each pass; "per-pass-run": one report for each pass invocation
.. option:: -ftime-trace
Turn on time profiler. Generates JSON file based on output filename. Results
can be analyzed with chrome://tracing or `Speedscope App
<https://www.speedscope.app>`_ for flamegraph visualization.
.. option:: -ftime-trace-granularity=<arg>
Minimum time granularity (in microseconds) traced by time profiler
.. option:: -ftls-model=<arg>
.. option:: -ftrap-function=<arg>
Issue call to specified function rather than a trap instruction
.. option:: -ftrapping-math, -fno-trapping-math
.. option:: -ftrapv
Trap on integer overflow
.. option:: -ftrapv-handler <arg>
.. program:: clang1
.. option:: -ftrapv-handler=<function name>
.. program:: clang
Specify the function to be called on overflow
.. option:: -ftrigraphs, -fno-trigraphs, -trigraphs, --trigraphs
Process trigraph sequences
.. option:: -ftrivial-auto-var-init-stop-after=<arg>
Stop initializing trivial automatic stack variables after the specified number of instances
.. option:: -ftrivial-auto-var-init=<arg>
Initialize trivial automatic stack variables: uninitialized (default) \| pattern
.. option:: -funique-basic-block-section-names, -fno-unique-basic-block-section-names
Use unique names for basic block sections (ELF Only)
.. option:: -funique-internal-linkage-names, -fno-unique-internal-linkage-names
Uniqueify Internal Linkage Symbol Names by appending the MD5 hash of the module path
.. option:: -funique-section-names, -fno-unique-section-names
.. option:: -funit-at-a-time, -fno-unit-at-a-time
.. option:: -funroll-loops, -fno-unroll-loops
Turn on loop unroller
.. option:: -funsafe-math-optimizations, -fno-unsafe-math-optimizations
.. option:: -funsigned-bitfields
.. option:: -funsigned-char, -fno-unsigned-char, --unsigned-char
.. option:: -funwind-tables, -fno-unwind-tables
.. option:: -fuse-cxa-atexit, -fno-use-cxa-atexit
.. option:: -fuse-init-array, -fno-use-init-array
.. option:: -fuse-ld=<arg>
.. option:: -fuse-line-directives, -fno-use-line-directives
Use #line in preprocessed output
.. option:: -fvalidate-ast-input-files-content
Compute and store the hash of input files used to build an AST. Files with mismatching mtime's are considered valid if both contents is identical
.. option:: -fveclib=<arg>
Use the given vector functions library
.. option:: -fvectorize, -fno-vectorize, -ftree-vectorize
Enable the loop vectorization passes
.. option:: -fverbose-asm, -dA, -fno-verbose-asm
Generate verbose assembly output
.. option:: -fvirtual-function-elimination, -fno-virtual-function-elimination
Enables dead virtual function elimination optimization. Requires -flto=full
.. option:: -fvisibility-dllexport=<arg>
The visibility for dllexport definitions \[-fvisibility-from-dllstorageclass\]
.. option:: -fvisibility-externs-dllimport=<arg>
The visibility for dllimport external declarations \[-fvisibility-from-dllstorageclass\]
.. option:: -fvisibility-externs-nodllstorageclass=<arg>
The visibility for external declarations without an explicit DLL dllstorageclass \[-fvisibility-from-dllstorageclass\]
.. option:: -fvisibility-from-dllstorageclass, -fno-visibility-from-dllstorageclass
Set the visibility of symbols in the generated code from their DLL storage class
.. option:: -fvisibility-global-new-delete-hidden
Give global C++ operator new and delete declarations hidden visibility
.. option:: -fvisibility-inlines-hidden, -fno-visibility-inlines-hidden
Give inline C++ member functions hidden visibility by default
.. option:: -fvisibility-inlines-hidden-static-local-var, -fno-visibility-inlines-hidden-static-local-var
When -fvisibility-inlines-hidden is enabled, static variables in inline C++ member functions will also be given hidden visibility by default
.. option:: -fvisibility-ms-compat
Give global types 'default' visibility and global functions and variables 'hidden' visibility by default
.. option:: -fvisibility-nodllstorageclass=<arg>
The visibility for defintiions without an explicit DLL export class \[-fvisibility-from-dllstorageclass\]
.. option:: -fvisibility=<arg>
Set the default symbol visibility for all global declarations
.. option:: -fwasm-exceptions
Use WebAssembly style exceptions
.. option:: -fwhole-program-vtables, -fno-whole-program-vtables
Enables whole-program vtable optimization. Requires -flto
.. option:: -fwrapv, -fno-wrapv
Treat signed integer overflow as two's complement
.. option:: -fwritable-strings
Store string literals as writable data
.. option:: -fxl-pragma-pack, -fno-xl-pragma-pack
Enable IBM XL #pragma pack handling
.. option:: -fxray-always-emit-customevents, -fno-xray-always-emit-customevents
Always emit \_\_xray\_customevent(...) calls even if the containing function is not always instrumented
.. option:: -fxray-always-emit-typedevents, -fno-xray-always-emit-typedevents
Always emit \_\_xray\_typedevent(...) calls even if the containing function is not always instrumented
.. option:: -fxray-always-instrument=<arg>
DEPRECATED: Filename defining the whitelist for imbuing the 'always instrument' XRay attribute.
.. option:: -fxray-attr-list=<arg>
Filename defining the list of functions/types for imbuing XRay attributes.
.. option:: -fxray-function-groups=<arg>
Only instrument 1 of N groups
.. option:: -fxray-function-index, -fno-xray-function-index
.. option:: -fxray-ignore-loops, -fno-xray-ignore-loops
Don't instrument functions with loops unless they also meet the minimum function size
.. option:: -fxray-instruction-threshold<arg>
.. program:: clang1
.. option:: -fxray-instruction-threshold=<arg>
.. program:: clang
Sets the minimum function size to instrument with XRay
.. option:: -fxray-instrument, -fno-xray-instrument
Generate XRay instrumentation sleds on function entry and exit
.. option:: -fxray-instrumentation-bundle=<arg>
Select which XRay instrumentation points to emit. Options: all, none, function-entry, function-exit, function, custom. Default is 'all'. 'function' includes both 'function-entry' and 'function-exit'.
.. option:: -fxray-link-deps
Tells clang to add the link dependencies for XRay.
.. option:: -fxray-modes=<arg>
List of modes to link in by default into XRay instrumented binaries.
.. option:: -fxray-never-instrument=<arg>
DEPRECATED: Filename defining the whitelist for imbuing the 'never instrument' XRay attribute.
.. option:: -fxray-selected-function-group=<arg>
When using -fxray-function-groups, select which group of functions to instrument. Valid range is 0 to fxray-function-groups - 1
.. option:: -fzero-call-used-regs=<arg>
Clear call-used registers upon function return.
.. option:: -fzero-initialized-in-bss, -fno-zero-initialized-in-bss
.. option:: -fzvector, -fno-zvector, -mzvector
Enable System z vector language extension
.. option:: --gpu-bundle-output, --no-gpu-bundle-output
Bundle output files of HIP device compilation
.. option:: -pedantic, --pedantic, -no-pedantic, --no-pedantic
Warn on language extensions
.. option:: -pedantic-errors, --pedantic-errors
OpenCL flags
------------
.. option:: -cl-denorms-are-zero
OpenCL only. Allow denormals to be flushed to zero.
.. option:: -cl-fast-relaxed-math
OpenCL only. Sets -cl-finite-math-only and -cl-unsafe-math-optimizations, and defines \_\_FAST\_RELAXED\_MATH\_\_.
.. option:: -cl-finite-math-only
OpenCL only. Allow floating-point optimizations that assume arguments and results are not NaNs or +-Inf.
.. option:: -cl-fp32-correctly-rounded-divide-sqrt
OpenCL only. Specify that single precision floating-point divide and sqrt used in the program source are correctly rounded.
.. option:: -cl-kernel-arg-info
OpenCL only. Generate kernel argument metadata.
.. option:: -cl-mad-enable
OpenCL only. Allow use of less precise MAD computations in the generated binary.
.. option:: -cl-no-signed-zeros
OpenCL only. Allow use of less precise no signed zeros computations in the generated binary.
.. option:: -cl-no-stdinc
OpenCL only. Disables all standard includes containing non-native compiler types and functions.
.. option:: -cl-opt-disable
OpenCL only. This option disables all optimizations. By default optimizations are enabled.
.. option:: -cl-single-precision-constant
OpenCL only. Treat double precision floating-point constant as single precision constant.
.. option:: -cl-std=<arg>
OpenCL language standard to compile for.
.. option:: -cl-strict-aliasing
OpenCL only. This option is added for compatibility with OpenCL 1.0.
.. option:: -cl-uniform-work-group-size
OpenCL only. Defines that the global work-size be a multiple of the work-group size specified to clEnqueueNDRangeKernel
.. option:: -cl-unsafe-math-optimizations
OpenCL only. Allow unsafe floating-point optimizations. Also implies -cl-no-signed-zeros and -cl-mad-enable.
SYCL flags
----------
.. option:: -fsycl, -fno-sycl
Enables SYCL kernels compilation for device
.. option:: -sycl-std=<arg>
SYCL language standard to compile for.
Target-dependent compilation options
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.. option:: -G<size>, -G=<arg>, -msmall-data-limit=<arg>, -msmall-data-threshold=<arg>
Put objects of at most <size> bytes into small data section (MIPS / Hexagon)
.. option:: -ffixed-x1
Reserve the x1 register (AArch64/RISC-V only)
.. option:: -ffixed-x10
Reserve the x10 register (AArch64/RISC-V only)
.. option:: -ffixed-x11
Reserve the x11 register (AArch64/RISC-V only)
.. option:: -ffixed-x12
Reserve the x12 register (AArch64/RISC-V only)
.. option:: -ffixed-x13
Reserve the x13 register (AArch64/RISC-V only)
.. option:: -ffixed-x14
Reserve the x14 register (AArch64/RISC-V only)
.. option:: -ffixed-x15
Reserve the x15 register (AArch64/RISC-V only)
.. option:: -ffixed-x16
Reserve the x16 register (AArch64/RISC-V only)
.. option:: -ffixed-x17
Reserve the x17 register (AArch64/RISC-V only)
.. option:: -ffixed-x18
Reserve the x18 register (AArch64/RISC-V only)
.. option:: -ffixed-x19
Reserve the x19 register (AArch64/RISC-V only)
.. option:: -ffixed-x2
Reserve the x2 register (AArch64/RISC-V only)
.. option:: -ffixed-x20
Reserve the x20 register (AArch64/RISC-V only)
.. option:: -ffixed-x21
Reserve the x21 register (AArch64/RISC-V only)
.. option:: -ffixed-x22
Reserve the x22 register (AArch64/RISC-V only)
.. option:: -ffixed-x23
Reserve the x23 register (AArch64/RISC-V only)
.. option:: -ffixed-x24
Reserve the x24 register (AArch64/RISC-V only)
.. option:: -ffixed-x25
Reserve the x25 register (AArch64/RISC-V only)
.. option:: -ffixed-x26
Reserve the x26 register (AArch64/RISC-V only)
.. option:: -ffixed-x27
Reserve the x27 register (AArch64/RISC-V only)
.. option:: -ffixed-x28
Reserve the x28 register (AArch64/RISC-V only)
.. option:: -ffixed-x29
Reserve the x29 register (AArch64/RISC-V only)
.. option:: -ffixed-x3
Reserve the x3 register (AArch64/RISC-V only)
.. option:: -ffixed-x30
Reserve the x30 register (AArch64/RISC-V only)
.. option:: -ffixed-x31
Reserve the x31 register (AArch64/RISC-V only)
.. option:: -ffixed-x4
Reserve the x4 register (AArch64/RISC-V only)
.. option:: -ffixed-x5
Reserve the x5 register (AArch64/RISC-V only)
.. option:: -ffixed-x6
Reserve the x6 register (AArch64/RISC-V only)
.. option:: -ffixed-x7
Reserve the x7 register (AArch64/RISC-V only)
.. option:: -ffixed-x8
Reserve the x8 register (AArch64/RISC-V only)
.. option:: -ffixed-x9
Reserve the x9 register (AArch64/RISC-V only)
.. option:: -ffuchsia-api-level=<arg>
Set Fuchsia API level
.. option:: -inline-asm=<arg>
.. option:: -m16
.. option:: -m32
.. option:: -m64
.. option:: -mabi=<arg>
.. program:: clang1
.. option:: -mabi=vec-default
.. program:: clang
Enable the default Altivec ABI on AIX (AIX only). Uses only volatile vector registers.
.. program:: clang2
.. option:: -mabi=vec-extabi
.. program:: clang
Enable the extended Altivec ABI on AIX (AIX only). Uses volatile and nonvolatile vector registers
.. option:: -maix-struct-return
Return all structs in memory (PPC32 only)
.. option:: -malign-branch-boundary=<arg>
Specify the boundary's size to align branches
.. option:: -malign-branch=<arg1>,<arg2>...
Specify types of branches to align
.. option:: -malign-double
Align doubles to two words in structs (x86 only)
.. option:: -mamdgpu-ieee, -mno-amdgpu-ieee
Sets the IEEE bit in the expected default floating point mode register. Floating point opcodes that support exception flag gathering quiet and propagate signaling NaN inputs per IEEE 754-2008. This option changes the ABI. (AMDGPU only)
.. option:: -march=<arg>
.. option:: -masm=<arg>
.. option:: -mbackchain, -mno-backchain
Link stack frames through backchain on System Z
.. option:: -mbranches-within-32B-boundaries
Align selected branches (fused, jcc, jmp) within 32-byte boundary
.. option:: -mcmodel=<arg>, -mcmodel=medany (equivalent to -mcmodel=medium), -mcmodel=medlow (equivalent to -mcmodel=small)
.. option:: -mcode-object-v3, -mno-code-object-v3
Legacy option to specify code object ABI V3 (AMDGPU only)
.. option:: -mcode-object-version=<arg>
Specify code object ABI version. Allowed values are 2, 3, 4, and 5. Defaults to 4. (AMDGPU only)
.. option:: -mconsole<arg>
.. program:: clang1
.. option:: -mcpu=<arg>, -mv5 (equivalent to -mcpu=hexagonv5), -mv55 (equivalent to -mcpu=hexagonv55), -mv60 (equivalent to -mcpu=hexagonv60), -mv62 (equivalent to -mcpu=hexagonv62), -mv65 (equivalent to -mcpu=hexagonv65), -mv66 (equivalent to -mcpu=hexagonv66), -mv67 (equivalent to -mcpu=hexagonv67), -mv67t (equivalent to -mcpu=hexagonv67t), -mv68 (equivalent to -mcpu=hexagonv68), -mv69 (equivalent to -mcpu=hexagonv69)
.. program:: clang
.. option:: -mcrc, -mno-crc
Allow use of CRC instructions (ARM/Mips only)
.. option:: -mdefault-build-attributes<arg>, -mno-default-build-attributes<arg>
.. option:: -mdll<arg>
.. option:: -mdouble=<arg>
Force double to be 32 bits or 64 bits
.. option:: -mdynamic-no-pic<arg>
.. option:: -meabi <arg>
Set EABI type, e.g. 4, 5 or gnu (default depends on triple)
.. option:: -menable-experimental-extensions
Enable use of experimental RISC-V extensions.
.. option:: -mfentry
Insert calls to fentry at function entry (x86/SystemZ only)
.. option:: -mfloat-abi=<arg>
.. option:: -mfpmath=<arg>
.. option:: -mfpu=<arg>
.. option:: -mgeneral-regs-only
Generate code which only uses the general purpose registers (AArch64/x86 only)
.. option:: -mglobal-merge, -mno-global-merge
Enable merging of globals
.. option:: -mhard-float
.. option:: -mhwdiv=<arg>, --mhwdiv <arg>, --mhwdiv=<arg>
.. option:: -mhwmult=<arg>
.. option:: -miamcu, -mno-iamcu
Use Intel MCU ABI
.. option:: -mibt-seal
Optimize fcf-protection=branch/full (requires LTO).
.. option:: -mignore-xcoff-visibility
Not emit the visibility attribute for asm in AIX OS or give all symbols 'unspecified' visibility in XCOFF object file
.. option:: -mimplicit-float, -mno-implicit-float
.. option:: -mimplicit-it=<arg>
.. option:: -mincremental-linker-compatible, -mno-incremental-linker-compatible
(integrated-as) Emit an object file which can be used with an incremental linker
.. option:: -miphoneos-version-min=<arg>, -mios-version-min=<arg>
.. option:: -mkernel
.. option:: -mlong-calls, -mno-long-calls
Generate branches with extended addressability, usually via indirect jumps.
.. option:: -mlvi-cfi, -mno-lvi-cfi
Enable only control-flow mitigations for Load Value Injection (LVI)
.. option:: -mlvi-hardening, -mno-lvi-hardening
Enable all mitigations for Load Value Injection (LVI)
.. option:: -mmacosx-version-min=<arg>, -mmacos-version-min=<arg>
Set Mac OS X deployment target
.. option:: -mmcu=<arg>
.. option:: -mms-bitfields, -mno-ms-bitfields
Set the default structure layout to be compatible with the Microsoft compiler standard
.. option:: -mnop-mcount
Generate mcount/\_\_fentry\_\_ calls as nops. To activate they need to be patched in.
.. option:: -momit-leaf-frame-pointer, -mno-omit-leaf-frame-pointer
Omit frame pointer setup for leaf functions
.. option:: -moslib=<arg>
.. option:: -mpacked-stack, -mno-packed-stack
Use packed stack layout (SystemZ only).
.. option:: -mpad-max-prefix-size=<arg>
Specify maximum number of prefixes to use for padding
.. option:: -mprefer-vector-width=<arg>
Specifies preferred vector width for auto-vectorization. Defaults to 'none' which allows target specific decisions.
.. option:: -mqdsp6-compat
Enable hexagon-qdsp6 backward compatibility
.. option:: -mrecip
.. program:: clang1
.. option:: -mrecip=<arg1>,<arg2>...
.. program:: clang
.. option:: -mrecord-mcount
Generate a \_\_mcount\_loc section entry for each \_\_fentry\_\_ call.
.. option:: -mred-zone, -mno-red-zone
.. option:: -mregparm=<arg>
.. option:: -mrelax, -mno-relax
Enable linker relaxation
.. option:: -mrelax-all, -mno-relax-all
(integrated-as) Relax all machine instructions
.. option:: -mretpoline, -mno-retpoline
.. option:: -mrtd, -mno-rtd
Make StdCall calling convention the default
.. option:: -mseses, -mno-seses
Enable speculative execution side effect suppression (SESES). Includes LVI control flow integrity mitigations
.. option:: -msign-return-address=<arg>
Select return address signing scope
.. option:: -msim
.. option:: -mskip-rax-setup, -mno-skip-rax-setup
Skip setting up RAX register when passing variable arguments (x86 only)
.. option:: -msoft-float, -mno-soft-float
Use software floating point
.. option:: -mspeculative-load-hardening, -mno-speculative-load-hardening
.. option:: -mstack-alignment=<arg>
Set the stack alignment
.. option:: -mstack-arg-probe, -mno-stack-arg-probe
Enable stack probes
.. option:: -mstack-probe-size=<arg>
Set the stack probe size
.. option:: -mstack-protector-guard-offset=<arg>
Use the given offset for addressing the stack-protector guard
.. option:: -mstack-protector-guard-reg=<arg>
Use the given reg for addressing the stack-protector guard
.. option:: -mstack-protector-guard=<arg>
Use the given guard (global, tls) for addressing the stack-protector guard
.. option:: -mstackrealign, -mno-stackrealign
Force realign the stack at entry to every function
.. option:: -msvr4-struct-return
Return small structs in registers (PPC32 only)
.. option:: -mtargetos=<arg>
Set the deployment target to be the specified OS and OS version
.. option:: -mthread-model <arg>
The thread model to use, e.g. posix, single (posix by default)
.. option:: -mthreads<arg>
.. option:: -mthumb, -mno-thumb
.. option:: -mtls-direct-seg-refs, -mno-tls-direct-seg-refs
Enable direct TLS access through segment registers (default)
.. option:: -mtls-size=<arg>
Specify bit size of immediate TLS offsets (AArch64 ELF only): 12 (for 4KB) \| 24 (for 16MB, default) \| 32 (for 4GB) \| 48 (for 256TB, needs -mcmodel=large)
.. program:: clang1
.. option:: -mtune=<arg>
.. program:: clang
Only supported on X86 and RISC-V. Otherwise accepted for compatibility with GCC.
.. option:: -mtvos-version-min=<arg>, -mappletvos-version-min=<arg>
.. option:: -municode<arg>
.. option:: -munsafe-fp-atomics, -mno-unsafe-fp-atomics
Enable unsafe floating point atomic instructions (AMDGPU only)
.. option:: -mvx, -mno-vx
.. option:: -mwarn-nonportable-cfstrings, -mno-warn-nonportable-cfstrings
.. option:: -mwatchos-version-min=<arg>
.. option:: -mwavefrontsize64, -mno-wavefrontsize64
Specify wavefront size 64 mode (AMDGPU only)
.. option:: -mwindows<arg>
.. option:: -mx32
AARCH64
-------
.. option:: -fcall-saved-x10
Make the x10 register call-saved (AArch64 only)
.. option:: -fcall-saved-x11
Make the x11 register call-saved (AArch64 only)
.. option:: -fcall-saved-x12
Make the x12 register call-saved (AArch64 only)
.. option:: -fcall-saved-x13
Make the x13 register call-saved (AArch64 only)
.. option:: -fcall-saved-x14
Make the x14 register call-saved (AArch64 only)
.. option:: -fcall-saved-x15
Make the x15 register call-saved (AArch64 only)
.. option:: -fcall-saved-x18
Make the x18 register call-saved (AArch64 only)
.. option:: -fcall-saved-x8
Make the x8 register call-saved (AArch64 only)
.. option:: -fcall-saved-x9
Make the x9 register call-saved (AArch64 only)
.. option:: -mfix-cortex-a53-835769, -mno-fix-cortex-a53-835769
Workaround Cortex-A53 erratum 835769 (AArch64 only)
.. option:: -mmark-bti-property
Add .note.gnu.property with BTI to assembly files (AArch64 only)
.. option:: -msve-vector-bits=<arg>
Specify the size in bits of an SVE vector register. Defaults to the vector length agnostic value of "scalable". (AArch64 only)
.. option:: -mvscale-max=<arg>
Specify the vscale maximum. Defaults to the vector length agnostic value of "0". (AArch64 only)
.. option:: -mvscale-min=<arg>
Specify the vscale minimum. Defaults to "1". (AArch64 only)
AMDGPU
------
.. option:: -mcumode, -mno-cumode
Specify CU wavefront execution mode (AMDGPU only)
.. option:: -mtgsplit, -mno-tgsplit
Enable threadgroup split execution mode (AMDGPU only)
ARM
---
.. option:: -faapcs-bitfield-load
Follows the AAPCS standard that all volatile bit-field write generates at least one load. (ARM only).
.. option:: -faapcs-bitfield-width, -fno-aapcs-bitfield-width
Follow the AAPCS standard requirement stating that volatile bit-field width is dictated by the field container type. (ARM only).
.. option:: -ffixed-r9
Reserve the r9 register (ARM only)
.. option:: -mcmse
Allow use of CMSE (Armv8-M Security Extensions)
.. option:: -mexecute-only, -mno-execute-only, -mpure-code
Disallow generation of data access to code sections (ARM only)
.. option:: -mfix-cmse-cve-2021-35465, -mno-fix-cmse-cve-2021-35465
Work around VLLDM erratum CVE-2021-35465 (ARM only)
.. option:: -mno-bti-at-return-twice
Do not add a BTI instruction after a setjmp or other return-twice construct (AArch32/AArch64 only)
.. option:: -mno-movt
Disallow use of movt/movw pairs (ARM only)
.. option:: -mno-neg-immediates
Disallow converting instructions with negative immediates to their negation or inversion.
.. option:: -mnocrc
Disallow use of CRC instructions (ARM only)
.. option:: -mrestrict-it, -mno-restrict-it
Disallow generation of complex IT blocks.
.. option:: -mtp=<arg>
Thread pointer access method (AArch32/AArch64 only)
.. option:: -munaligned-access, -mno-unaligned-access
Allow memory accesses to be unaligned (AArch32/AArch64 only)
Hexagon
-------
.. option:: -mhvx-ieee-fp, -mno-hvx-ieee-fp
Enable Hexagon HVX IEEE floating-point
.. option:: -mieee-rnd-near
.. option:: -mmemops, -mno-memops
Enable generation of memop instructions
.. option:: -mnvj, -mno-nvj
Enable generation of new-value jumps
.. option:: -mnvs, -mno-nvs
Enable generation of new-value stores
.. option:: -mpackets, -mno-packets
Enable generation of instruction packets
Hexagon
-------
.. option:: -mhvx, -mno-hvx
Enable Hexagon Vector eXtensions
.. option:: -mhvx-length=<arg>
Set Hexagon Vector Length
.. option:: -mhvx-qfloat, -mno-hvx-qfloat
Enable Hexagon HVX QFloat instructions
.. program:: clang1
.. option:: -mhvx=<arg>
.. program:: clang
Enable Hexagon Vector eXtensions
M68k
----
.. option:: -ffixed-a0
Reserve the a0 register (M68k only)
.. option:: -ffixed-a1
Reserve the a1 register (M68k only)
.. option:: -ffixed-a2
Reserve the a2 register (M68k only)
.. option:: -ffixed-a3
Reserve the a3 register (M68k only)
.. option:: -ffixed-a4
Reserve the a4 register (M68k only)
.. option:: -ffixed-a5
Reserve the a5 register (M68k only)
.. option:: -ffixed-a6
Reserve the a6 register (M68k only)
.. option:: -ffixed-d0
Reserve the d0 register (M68k only)
.. option:: -ffixed-d1
Reserve the d1 register (M68k only)
.. option:: -ffixed-d2
Reserve the d2 register (M68k only)
.. option:: -ffixed-d3
Reserve the d3 register (M68k only)
.. option:: -ffixed-d4
Reserve the d4 register (M68k only)
.. option:: -ffixed-d5
Reserve the d5 register (M68k only)
.. option:: -ffixed-d6
Reserve the d6 register (M68k only)
.. option:: -ffixed-d7
Reserve the d7 register (M68k only)
.. option:: -m68000
.. option:: -m68010
.. option:: -m68020
.. option:: -m68030
.. option:: -m68040
.. option:: -m68060
MIPS
----
.. option:: -mabicalls, -mno-abicalls
Enable SVR4-style position-independent code (Mips only)
.. option:: -mabs=<arg>
.. option:: -mcheck-zero-division, -mno-check-zero-division
.. option:: -mcompact-branches=<arg>
.. option:: -mdouble-float
.. option:: -mdsp, -mno-dsp
.. option:: -mdspr2, -mno-dspr2
.. option:: -membedded-data, -mno-embedded-data
Place constants in the .rodata section instead of the .sdata section even if they meet the -G <size> threshold (MIPS)
.. option:: -mextern-sdata, -mno-extern-sdata
Assume that externally defined data is in the small data if it meets the -G <size> threshold (MIPS)
.. option:: -mfix4300
.. option:: -mfp32
Use 32-bit floating point registers (MIPS only)
.. option:: -mfp64
Use 64-bit floating point registers (MIPS only)
.. option:: -mginv, -mno-ginv
.. option:: -mgpopt, -mno-gpopt
Use GP relative accesses for symbols known to be in a small data section (MIPS)
.. option:: -mindirect-jump=<arg>
Change indirect jump instructions to inhibit speculation
.. option:: -mips16
.. option:: -mldc1-sdc1, -mno-ldc1-sdc1
.. option:: -mlocal-sdata, -mno-local-sdata
Extend the -G behaviour to object local data (MIPS)
.. option:: -mmadd4, -mno-madd4
Enable the generation of 4-operand madd.s, madd.d and related instructions.
.. option:: -mmicromips, -mno-micromips
.. option:: -mmsa, -mno-msa
Enable MSA ASE (MIPS only)
.. option:: -mmt, -mno-mt
Enable MT ASE (MIPS only)
.. option:: -mnan=<arg>
.. option:: -mno-mips16
.. option:: -msingle-float
.. option:: -mvirt, -mno-virt
.. option:: -mxgot, -mno-xgot
PowerPC
-------
.. option:: -maltivec, -mno-altivec
.. option:: -mcmpb, -mno-cmpb
.. option:: -mcrbits, -mno-crbits
.. option:: -mcrypto, -mno-crypto
.. option:: -mdirect-move, -mno-direct-move
.. option:: -mefpu2
.. option:: -mfloat128, -mno-float128
.. option:: -mfprnd, -mno-fprnd
.. option:: -mhtm, -mno-htm
.. option:: -minvariant-function-descriptors, -mno-invariant-function-descriptors
.. option:: -misel, -mno-isel
.. option:: -mlongcall, -mno-longcall
.. option:: -mmfocrf, -mmfcrf, -mno-mfocrf
.. option:: -mmma, -mno-mma
.. option:: -mpaired-vector-memops, -mno-paired-vector-memops
.. option:: -mpcrel, -mno-pcrel
.. option:: -mpopcntd, -mno-popcntd
.. option:: -mpower10-vector, -mno-power10-vector
.. option:: -mpower8-vector, -mno-power8-vector
.. option:: -mpower9-vector, -mno-power9-vector
.. option:: -mprefixed, -mno-prefixed
.. option:: -mprivileged
.. option:: -mrop-protect
.. option:: -msecure-plt
.. option:: -mspe, -mno-spe
.. option:: -mvsx, -mno-vsx
WebAssembly
-----------
.. option:: -matomics, -mno-atomics
.. option:: -mbulk-memory, -mno-bulk-memory
.. option:: -mexception-handling, -mno-exception-handling
.. option:: -mmultivalue, -mno-multivalue
.. option:: -mmutable-globals, -mno-mutable-globals
.. option:: -mnontrapping-fptoint, -mno-nontrapping-fptoint
.. option:: -mreference-types, -mno-reference-types
.. option:: -mrelaxed-simd, -mno-relaxed-simd
.. option:: -msign-ext, -mno-sign-ext
.. option:: -msimd128, -mno-simd128
.. option:: -mtail-call, -mno-tail-call
.. option:: -mextended-const, -mno-extended-const
WebAssembly Driver
------------------
.. option:: -mexec-model=<arg>
Execution model (WebAssembly only)
X86
---
.. option:: -m3dnow, -mno-3dnow
.. option:: -m3dnowa, -mno-3dnowa
.. option:: -madx, -mno-adx
.. option:: -maes, -mno-aes
.. option:: -mamx-bf16, -mno-amx-bf16
.. option:: -mamx-int8, -mno-amx-int8
.. option:: -mamx-tile, -mno-amx-tile
.. option:: -mavx, -mno-avx
.. option:: -mavx2, -mno-avx2
.. option:: -mavx512bf16, -mno-avx512bf16
.. option:: -mavx512bitalg, -mno-avx512bitalg
.. option:: -mavx512bw, -mno-avx512bw
.. option:: -mavx512cd, -mno-avx512cd
.. option:: -mavx512dq, -mno-avx512dq
.. option:: -mavx512er, -mno-avx512er
.. option:: -mavx512f, -mno-avx512f
.. option:: -mavx512fp16, -mno-avx512fp16
.. option:: -mavx512ifma, -mno-avx512ifma
.. option:: -mavx512pf, -mno-avx512pf
.. option:: -mavx512vbmi, -mno-avx512vbmi
.. option:: -mavx512vbmi2, -mno-avx512vbmi2
.. option:: -mavx512vl, -mno-avx512vl
.. option:: -mavx512vnni, -mno-avx512vnni
.. option:: -mavx512vp2intersect, -mno-avx512vp2intersect
.. option:: -mavx512vpopcntdq, -mno-avx512vpopcntdq
.. option:: -mavxvnni, -mno-avxvnni
.. option:: -mbmi, -mno-bmi
.. option:: -mbmi2, -mno-bmi2
.. option:: -mcldemote, -mno-cldemote
.. option:: -mclflushopt, -mno-clflushopt
.. option:: -mclwb, -mno-clwb
.. option:: -mclzero, -mno-clzero
.. option:: -mcrc32, -mno-crc32
.. option:: -mcx16, -mno-cx16
.. option:: -menqcmd, -mno-enqcmd
.. option:: -mf16c, -mno-f16c
.. option:: -mfma, -mno-fma
.. option:: -mfma4, -mno-fma4
.. option:: -mfsgsbase, -mno-fsgsbase
.. option:: -mfxsr, -mno-fxsr
.. option:: -mgfni, -mno-gfni
.. option:: -mhreset, -mno-hreset
.. option:: -minvpcid, -mno-invpcid
.. option:: -mkl, -mno-kl
.. option:: -mlwp, -mno-lwp
.. option:: -mlzcnt, -mno-lzcnt
.. option:: -mmmx, -mno-mmx
.. option:: -mmovbe, -mno-movbe
.. option:: -mmovdir64b, -mno-movdir64b
.. option:: -mmovdiri, -mno-movdiri
.. option:: -mmwaitx, -mno-mwaitx
.. option:: -mpclmul, -mno-pclmul
.. option:: -mpconfig, -mno-pconfig
.. option:: -mpku, -mno-pku
.. option:: -mpopcnt, -mno-popcnt
.. option:: -mprefetchwt1, -mno-prefetchwt1
.. option:: -mprfchw, -mno-prfchw
.. option:: -mptwrite, -mno-ptwrite
.. option:: -mrdpid, -mno-rdpid
.. option:: -mrdrnd, -mno-rdrnd
.. option:: -mrdseed, -mno-rdseed
.. option:: -mretpoline-external-thunk, -mno-retpoline-external-thunk
.. option:: -mrtm, -mno-rtm
.. option:: -msahf, -mno-sahf
.. option:: -mserialize, -mno-serialize
.. option:: -msgx, -mno-sgx
.. option:: -msha, -mno-sha
.. option:: -mshstk, -mno-shstk
.. option:: -msse, -mno-sse
.. option:: -msse2, -mno-sse2
.. option:: -msse3, -mno-sse3
.. option:: -msse4.1, -mno-sse4.1
.. program:: clang1
.. option:: -msse4.2, -mno-sse4.2, -msse4
.. program:: clang
.. option:: -msse4a, -mno-sse4a
.. option:: -mssse3, -mno-ssse3
.. option:: -mtbm, -mno-tbm
.. option:: -mtsxldtrk, -mno-tsxldtrk
.. option:: -muintr, -mno-uintr
.. option:: -mvaes, -mno-vaes
.. option:: -mvpclmulqdq, -mno-vpclmulqdq
.. option:: -mvzeroupper, -mno-vzeroupper
.. option:: -mwaitpkg, -mno-waitpkg
.. option:: -mwbnoinvd, -mno-wbnoinvd
.. option:: -mwidekl, -mno-widekl
.. option:: -mx87, -m80387, -mno-x87
.. option:: -mxop, -mno-xop
.. option:: -mxsave, -mno-xsave
.. option:: -mxsavec, -mno-xsavec
.. option:: -mxsaveopt, -mno-xsaveopt
.. option:: -mxsaves, -mno-xsaves
RISCV
-----
.. option:: -msave-restore, -mno-save-restore
Enable using library calls for save and restore
Long double flags
-----------------
Selects the long double implementation
.. option:: -mlong-double-128
Force long double to be 128 bits
.. option:: -mlong-double-64
Force long double to be 64 bits
.. option:: -mlong-double-80
Force long double to be 80 bits, padded to 128 bits for storage
Optimization level
~~~~~~~~~~~~~~~~~~
Flags controlling how much optimization should be performed.
.. option:: -O<arg>, -O (equivalent to -O1), --optimize, --optimize=<arg>
.. option:: -Ofast<arg>
Debug information generation
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Flags controlling how much and what kind of debug information should be
generated.
Kind and level of debug information
-----------------------------------
.. option:: -g, --debug, --debug=<arg>
Generate source-level debug information
.. option:: -gdwarf
Generate source-level debug information with the default dwarf version
.. option:: -gdwarf-2
Generate source-level debug information with dwarf version 2
.. option:: -gdwarf-3
Generate source-level debug information with dwarf version 3
.. option:: -gdwarf-4
Generate source-level debug information with dwarf version 4
.. option:: -gdwarf-5
Generate source-level debug information with dwarf version 5
.. option:: -gdwarf32
Enables DWARF32 format for ELF binaries, if debug information emission is enabled.
.. option:: -gdwarf64
Enables DWARF64 format for ELF binaries, if debug information emission is enabled.
.. option:: -gfull
.. option:: -ginline-line-tables, -gno-inline-line-tables
.. option:: -gused
Debug level
___________
.. option:: -g0
.. option:: -g2
.. option:: -g3
.. option:: -ggdb0
.. option:: -ggdb1
.. option:: -ggdb2
.. option:: -ggdb3
.. option:: -gline-directives-only
Emit debug line info directives only
.. option:: -gline-tables-only, -g1, -gmlt
Emit debug line number tables only
.. option:: -gmodules
Generate debug info with external references to clang modules or precompiled headers
Debugger to tune debug information for
______________________________________
.. option:: -gdbx
.. option:: -ggdb
.. option:: -glldb
.. option:: -gsce
Debug information flags
-----------------------
.. option:: -gcolumn-info, -gno-column-info
.. option:: -gdwarf-aranges
.. option:: -gembed-source, -gno-embed-source
Embed source text in DWARF debug sections
.. option:: -ggnu-pubnames, -gno-gnu-pubnames
.. option:: -gpubnames, -gno-pubnames
.. option:: -grecord-command-line, -gno-record-command-line, -grecord-gcc-switches
.. option:: -gsimple-template-names, -gno-simple-template-names
.. option:: -gsplit-dwarf, -gno-split-dwarf
.. program:: clang1
.. option:: -gsplit-dwarf=<arg>
.. program:: clang
Set DWARF fission mode to either 'split' or 'single'
.. option:: -gstrict-dwarf, -gno-strict-dwarf
.. option:: -gz=<arg>, -gz (equivalent to -gz=zlib)
DWARF debug sections compression type
Static analyzer flags
=====================
Flags controlling the behavior of the Clang Static Analyzer.
.. option:: -Xanalyzer <arg>
Pass <arg> to the static analyzer
Fortran compilation flags
=========================
Flags that will be passed onto the ``gfortran`` compiler when Clang is given
a Fortran input.
.. option:: -A<arg>, --assert <arg>, --assert=<arg>
.. option:: -A-<arg>
.. option:: -faggressive-function-elimination, -fno-aggressive-function-elimination
.. option:: -falign-commons, -fno-align-commons
.. option:: -fall-intrinsics, -fno-all-intrinsics
.. option:: -fbacktrace, -fno-backtrace
.. option:: -fblas-matmul-limit=<arg>
.. option:: -fbounds-check, -fno-bounds-check
.. option:: -fcheck-array-temporaries, -fno-check-array-temporaries
.. option:: -fcheck=<arg>
.. option:: -fcoarray=<arg>
.. option:: -fconvert=<arg>
.. option:: -fcray-pointer, -fno-cray-pointer
.. option:: -fd-lines-as-code, -fno-d-lines-as-code
.. option:: -fd-lines-as-comments, -fno-d-lines-as-comments
.. option:: -fdollar-ok, -fno-dollar-ok
.. option:: -fdump-fortran-optimized, -fno-dump-fortran-optimized
.. option:: -fdump-fortran-original, -fno-dump-fortran-original
.. option:: -fdump-parse-tree, -fno-dump-parse-tree
.. option:: -fexternal-blas, -fno-external-blas
.. option:: -ff2c, -fno-f2c
.. option:: -ffpe-trap=<arg>
.. option:: -ffree-line-length-<arg>
.. option:: -ffrontend-optimize, -fno-frontend-optimize
.. option:: -finit-character=<arg>
.. option:: -finit-integer=<arg>
.. option:: -finit-local-zero, -fno-init-local-zero
.. option:: -finit-logical=<arg>
.. option:: -finit-real=<arg>
.. option:: -finteger-4-integer-8, -fno-integer-4-integer-8
.. option:: -fmax-array-constructor=<arg>
.. option:: -fmax-errors=<arg>
.. option:: -fmax-identifier-length, -fno-max-identifier-length
.. option:: -fmax-stack-var-size=<arg>
.. option:: -fmax-subrecord-length=<arg>
.. option:: -fmodule-private, -fno-module-private
.. option:: -fpack-derived, -fno-pack-derived
.. option:: -frange-check, -fno-range-check
.. option:: -freal-4-real-10, -fno-real-4-real-10
.. option:: -freal-4-real-16, -fno-real-4-real-16
.. option:: -freal-4-real-8, -fno-real-4-real-8
.. option:: -freal-8-real-10, -fno-real-8-real-10
.. option:: -freal-8-real-16, -fno-real-8-real-16
.. option:: -freal-8-real-4, -fno-real-8-real-4
.. option:: -frealloc-lhs, -fno-realloc-lhs
.. option:: -frecord-marker=<arg>
.. option:: -frecursive, -fno-recursive
.. option:: -frepack-arrays, -fno-repack-arrays
.. option:: -fsecond-underscore, -fno-second-underscore
.. option:: -fsign-zero, -fno-sign-zero
.. option:: -fstack-arrays, -fno-stack-arrays
.. option:: -funderscoring, -fno-underscoring
.. option:: -fwhole-file, -fno-whole-file
.. option:: -imultilib <arg>
.. option:: -static-libgfortran
Linker flags
============
Flags that are passed on to the linker
.. option:: -L<dir>, --library-directory <arg>, --library-directory=<arg>
Add directory to library search path
.. option:: -Mach
.. option:: -T<script>
Specify <script> as linker script
.. option:: -Tbss<addr>
Set starting address of BSS to <addr>
.. option:: -Tdata<addr>
Set starting address of DATA to <addr>
.. option:: -Ttext<addr>
Set starting address of TEXT to <addr>
.. option:: -Wl,<arg>,<arg2>...
Pass the comma separated arguments in <arg> to the linker
.. option:: -X
.. option:: -Xlinker <arg>, --for-linker <arg>, --for-linker=<arg>
Pass <arg> to the linker
.. program:: clang1
.. option:: -Z
.. program:: clang
.. option:: -b<arg>
Pass -b <arg> to the linker on AIX (only).
.. option:: -coverage, --coverage
.. option:: -e<arg>, --entry
.. option:: -filelist <arg>
.. option:: --hip-device-lib=<arg>
HIP device library
.. option:: --hipspv-pass-plugin=<dsopath>
path to a pass plugin for HIP to SPIR-V passes.
.. option:: -l<arg>
.. option:: --ld-path=<arg>
.. option:: -nostartfiles
.. program:: clang1
.. option:: -nostdlib, --no-standard-libraries
.. program:: clang
.. option:: -pie
.. option:: -r
.. option:: -rdynamic
.. option:: --rocm-device-lib-path=<arg>, --hip-device-lib-path=<arg>
ROCm device library path. Alternative to rocm-path.
.. option:: -rpath <arg>
.. option:: -s
.. option:: -shared, --shared
.. option:: -specs=<arg>, --specs=<arg>
.. option:: -static, --static
.. option:: -static-pie
.. option:: -t
.. option:: -u<arg>, --force-link <arg>, --force-link=<arg>
.. option:: -undef
undef all system defines
.. option:: -undefined<arg>, --no-undefined
.. option:: -z <arg>
Pass -z <arg> to the linker