Contents
- 1. Introduction
- 2. Compilation Phases
- 3. The CUDA Compilation Trajectory
-
4. NVCC Command Options
- 4.1. Command Option Types and Notation
-
4.2. Command Option Description
-
4.2.1. File and Path Specifications
- 4.2.1.1.
--output-file file(-o) - 4.2.1.2.
--objdir-as-tempdir(-objtemp) - 4.2.1.3.
--pre-include file,...(-include) - 4.2.1.4.
--library library,...(-l) - 4.2.1.5.
--define-macro def,...(-D) - 4.2.1.6.
--undefine-macro def,...(-U) - 4.2.1.7.
--include-path path,...(-I) - 4.2.1.8.
--system-include path,...(-isystem) - 4.2.1.9.
--library-path path,...(-L) - 4.2.1.10.
--output-directory directory(-odir) - 4.2.1.11.
--dependency-output file(-MF) - 4.2.1.12.
--generate-dependency-targets(-MP) - 4.2.1.13.
--compiler-bindir directory(-ccbin) - 4.2.1.14.
--allow-unsupported-compiler(-allow-unsupported-compiler) - 4.2.1.15.
--archiver-binary executable(-arbin) - 4.2.1.16.
--cudart{none|shared|static} (-cudart) - 4.2.1.17.
--cudadevrt{none|static} (-cudadevrt) - 4.2.1.18.
--libdevice-directory directory(-ldir) - 4.2.1.19.
--target-directory string(-target-dir)
- 4.2.1.1.
-
4.2.2. Options for Specifying the Compilation Phase
- 4.2.2.1.
--link(-link) - 4.2.2.2.
--lib(-lib) - 4.2.2.3.
--device-link(-dlink) - 4.2.2.4.
--device-c(-dc) - 4.2.2.5.
--device-w(-dw) - 4.2.2.6.
--cuda(-cuda) - 4.2.2.7.
--compile(-c) - 4.2.2.8.
--fatbin(-fatbin) - 4.2.2.9.
--cubin(-cubin) - 4.2.2.10.
--ptx(-ptx) - 4.2.2.11.
--preprocess(-E) - 4.2.2.12.
--generate-dependencies(-M) - 4.2.2.13.
--generate-nonsystem-dependencies(-MM) - 4.2.2.14.
--generate-dependencies-with-compile(-MD) - 4.2.2.15.
--generate-nonsystem-dependencies-with-compile(-MMD) - 4.2.2.16.
--optix-ir(-optix-ir) - 4.2.2.17.
--run(-run)
- 4.2.2.1.
-
4.2.3. Options for Specifying Behavior of Compiler/Linker
- 4.2.3.1.
--profile(-pg) - 4.2.3.2.
--debug(-g) - 4.2.3.3.
--device-debug(-G) - 4.2.3.4.
--extensible-whole-program(-ewp) - 4.2.3.5.
--no-compress(-no-compress) - 4.2.3.6.
--generate-line-info(-lineinfo) - 4.2.3.7.
--optimization-info kind,...(-opt-info) - 4.2.3.8.
--optimize level(-O) - 4.2.3.9.
--dopt kind(-dopt) - 4.2.3.10.
--dlink-time-opt(-dlto) - 4.2.3.11.
--gen-opt-lto(-gen-opt-lto) - 4.2.3.12.
--split-compile number(-split-compile) - 4.2.3.13.
--split-compile-extended number(-split-compile-extended) - 4.2.3.14.
--ftemplate-backtrace-limit limit(-ftemplate-backtrace-limit) - 4.2.3.15.
--ftemplate-depth limit(-ftemplate-depth) - 4.2.3.16.
--no-exceptions(-noeh) - 4.2.3.17.
--shared(-shared) - 4.2.3.18.
--x{c|c++|cu} (-x) - 4.2.3.19.
--std{c++03|c++11|c++14|c++17|c++20} (-std) - 4.2.3.20.
--no-host-device-initializer-list(-nohdinitlist) - 4.2.3.21.
--expt-relaxed-constexpr(-expt-relaxed-constexpr) - 4.2.3.22.
--extended-lambda(-extended-lambda) - 4.2.3.23.
--expt-extended-lambda(-expt-extended-lambda) - 4.2.3.24.
--machine{64} (-m) - 4.2.3.25.
--m64(-m64) - 4.2.3.26.
--host-linker-script{use-lcs|gen-lcs} (-hls) - 4.2.3.27.
--augment-host-linker-script(-aug-hls) - 4.2.3.28.
--host-relocatable-link(-r)
- 4.2.3.1.
- 4.2.4. Options for Passing Specific Phase Options
-
4.2.5. Options for Guiding the Compiler Driver
- 4.2.5.1.
--forward-unknown-to-host-compiler(-forward-unknown-to-host-compiler) - 4.2.5.2.
--forward-unknown-to-host-linker(-forward-unknown-to-host-linker) - 4.2.5.3.
--forward-unknown-opts(-forward-unknown-opts) - 4.2.5.4.
--forward-slash-prefix-opts(-forward-slash-prefix-opts) - 4.2.5.5.
--dont-use-profile(-noprof) - 4.2.5.6.
--threads number(-t) - 4.2.5.7.
--dryrun(-dryrun) - 4.2.5.8.
--verbose(-v) - 4.2.5.9.
--keep(-keep) - 4.2.5.10.
--keep-dir directory(-keep-dir) - 4.2.5.11.
--save-temps(-save-temps) - 4.2.5.12.
--clean-targets(-clean) - 4.2.5.13.
--run-args arguments,...(-run-args) - 4.2.5.14.
--use-local-env(-use-local-env) - 4.2.5.15.
--input-drive-prefix prefix(-idp) - 4.2.5.16.
--dependency-drive-prefix prefix(-ddp) - 4.2.5.17.
--drive-prefix prefix(-dp) - 4.2.5.18.
--dependency-target-name target(-MT) - 4.2.5.19.
--no-align-double - 4.2.5.20.
--no-device-link(-nodlink) - 4.2.5.21.
--allow-unsupported-compiler(-allow-unsupported-compiler)
- 4.2.5.1.
- 4.2.6. Options for Steering CUDA Compilation
-
4.2.7. Options for Steering GPU Code Generation
- 4.2.7.1.
--gpu-architecture(-arch) - 4.2.7.2.
--gpu-code code,...(-code) - 4.2.7.3.
--generate-code specification(-gencode) - 4.2.7.4.
--relocatable-device-code{true|false} (-rdc) - 4.2.7.5.
--entries entry,...(-e) - 4.2.7.6.
--maxrregcount amount(-maxrregcount) - 4.2.7.7.
--use_fast_math(-use_fast_math) - 4.2.7.8.
--ftz{true|false} (-ftz) - 4.2.7.9.
--prec-div{true|false} (-prec-div) - 4.2.7.10.
--prec-sqrt{true|false} (-prec-sqrt) - 4.2.7.11.
--fmad{true|false} (-fmad) - 4.2.7.12.
--extra-device-vectorization(-extra-device-vectorization) - 4.2.7.13.
--compile-as-tools-patch(-astoolspatch) - 4.2.7.14.
--keep-device-functions(-keep-device-functions) - 4.2.7.15.
--jump-table-density percentage(-jtd)
- 4.2.7.1.
-
4.2.8. Generic Tool Options
- 4.2.8.1.
--disable-warnings(-w) - 4.2.8.2.
--source-in-ptx(-src-in-ptx) - 4.2.8.3.
--restrict(-restrict) - 4.2.8.4.
--Wno-deprecated-gpu-targets(-Wno-deprecated-gpu-targets) - 4.2.8.5.
--Wno-deprecated-declarations(-Wno-deprecated-declarations) - 4.2.8.6.
--Wreorder(-Wreorder) - 4.2.8.7.
--Wdefault-stream-launch(-Wdefault-stream-launch) - 4.2.8.8.
--Wmissing-launch-bounds(-Wmissing-launch-bounds) - 4.2.8.9.
--Wext-lambda-captures-this(-Wext-lambda-captures-this) - 4.2.8.10.
--Werror kind,...(-Werror) - 4.2.8.11.
--display-error-number(-err-no) - 4.2.8.12.
--no-display-error-number(-no-err-no) - 4.2.8.13.
--diag-error errNum,...(-diag-error) - 4.2.8.14.
--diag-suppress errNum,...(-diag-suppress) - 4.2.8.15.
--diag-warn errNum,...(-diag-warn) - 4.2.8.16.
--resource-usage(-res-usage) - 4.2.8.17.
--device-stack-protector{true|false} (-device-stack-protector) - 4.2.8.18.
--help(-h) - 4.2.8.19.
--version(-V) - 4.2.8.20.
--options-file file,...(-optf) - 4.2.8.21.
--time filename(-time) - 4.2.8.22.
--qpp-config config(-qpp-config) - 4.2.8.23.
--list-gpu-code(-code-ls) - 4.2.8.24.
--list-gpu-arch(-arch-ls)
- 4.2.8.1.
-
4.2.9. Phase Options
-
4.2.9.1. Ptxas Options
- 4.2.9.1.1.
--allow-expensive-optimizations(-allow-expensive-optimizations) - 4.2.9.1.2.
--compile-only(-c) - 4.2.9.1.3.
--def-load-cache(-dlcm) - 4.2.9.1.4.
--def-store-cache(-dscm) - 4.2.9.1.5.
--device-debug(-g) - 4.2.9.1.6.
--disable-optimizer-constants(-disable-optimizer-consts) - 4.2.9.1.7.
--entry entry,...(-e) - 4.2.9.1.8.
--fmad(-fmad) - 4.2.9.1.9.
--force-load-cache(-flcm) - 4.2.9.1.10.
--force-store-cache(-fscm) - 4.2.9.1.11.
--generate-line-info(-lineinfo) - 4.2.9.1.12.
--gpu-name gpuname(-arch) - 4.2.9.1.13.
--help(-h) - 4.2.9.1.14.
--machine(-m) - 4.2.9.1.15.
--maxrregcount amount(-maxrregcount) - 4.2.9.1.16.
--opt-level N(-O) - 4.2.9.1.17.
--options-file file,...(-optf) - 4.2.9.1.18.
--position-independent-code(-pic) - 4.2.9.1.19.
--preserve-relocs(-preserve-relocs) - 4.2.9.1.20.
--sp-bound-check(-sp-bound-check) - 4.2.9.1.21.
--suppress-async-bulk-multicast-advisory-warning(-suppress-async-bulk-multicast-advisory-warning) - 4.2.9.1.22.
--verbose(-v) - 4.2.9.1.23.
--version(-V) - 4.2.9.1.24.
--warning-as-error(-Werror) - 4.2.9.1.25.
--warn-on-double-precision-use(-warn-double-usage) - 4.2.9.1.26.
--warn-on-local-memory-usage(-warn-lmem-usage) - 4.2.9.1.27.
--warn-on-spills(-warn-spills) - 4.2.9.1.28.
--compile-as-tools-patch(-astoolspatch) - 4.2.9.1.29.
--maxntid(-maxntid) - 4.2.9.1.30.
--minnctapersm(-minnctapersm) - 4.2.9.1.31.
--override-directive-values(-override-directive-values) - 4.2.9.1.32.
--make-errors-visible-at-exit(-make-errors-visible-at-exit) - 4.2.9.1.33.
--oFast-compile(-Ofc) - 4.2.9.1.34.
--device-stack-protector{true|false} (-device-stack-protector)
- 4.2.9.1.1.
-
4.2.9.2. NVLINK Options
- 4.2.9.2.1.
--disable-warnings(-w) - 4.2.9.2.2.
--preserve-relocs(-preserve-relocs) - 4.2.9.2.3.
--verbose(-v) - 4.2.9.2.4.
--warning-as-error(-Werror) - 4.2.9.2.5.
--suppress-arch-warning(-suppress-arch-warning) - 4.2.9.2.6.
--suppress-stack-size-warning(-suppress-stack-size-warning) - 4.2.9.2.7.
--dump-callgraph(-dump-callgraph) - 4.2.9.2.8.
--dump-callgraph-no-demangle(-dump-callgraph-no-demangle) - 4.2.9.2.9.
--Xptxas(-Xptxas) - 4.2.9.2.10.
--cpu-arch(-cpu-arch) - 4.2.9.2.11.
--extra-warnings(-extrawarn) - 4.2.9.2.12.
--gen-host-linker-script(-ghls) - 4.2.9.2.13.
--ignore-host-info(-ignore-host-info) - 4.2.9.2.14.
--keep-system-libraries(-keep-system-libraries) - 4.2.9.2.15.
--kernels-used(-kernels-used) - 4.2.9.2.16.
--options-file(-optf) - 4.2.9.2.17.
--report-arch(-report-arch) - 4.2.9.2.18.
--suppress-debug-info(-suppress-debug-info) - 4.2.9.2.19.
--variables-used(-variables used) - 4.2.9.2.20.
--device-stack-protector{true|false} (-device-stack-protector)
- 4.2.9.2.1.
-
4.2.9.1. Ptxas Options
-
4.2.1. File and Path Specifications
- 4.3. NVCC Environment Variables
- 5. GPU Compilation
- 6. Using Separate Compilation in CUDA
- 7. Miscellaneous NVCC Usage
- 8. Notices