kefir-cc1(1)

📦
version 0.5.0 · 2 versions
, in alpine 3.24

💅 kefir-cc1(1) — General Commands Manual 💅

okay so. bestie. meet kefir-cc1, a C17/C23 language compiler — and the whole personality of this page is that it's the compiler core with no driver attached. Like, it's the main character before the entourage shows up. ✨

🌸 SYNOPSIS 🌸

kefir-cc1 [options] [file]

🦋 DESCRIPTION 🦋

kefir-cc1 is a C17/C23 language compiler — but without a driver, iykyk. That means you feed it stuff directly, no hand-holding.

And if you don't hand it an input file at all? It just reads from stdin. no cap, it will literally wait there for you to type. 🫶

✨ OPTIONS ✨

so here's the menu. every single flag, no skipping, we don't do that here 💅

  • --c17-standard — use the C17 language standard.
  • --c23-standard — use the C23 language standard.
  • --pp-timestamp timestamp — override the preprocessor timestamp. yes, you get to be the clock now ⏰
  • --json-errors — print errors in JSON format to stderr.
  • --tabular-errors — print errors in tabular format to stderr. tables!! for your errors!!
  • --target-profile profile — generate code for the specified target (peep the TARGET subsection 🎯).
  • --source-id identifier — force it to use the source file identifier you provide.
  • --extension-lib libpath — load an extension library.
  • --quote-include-dir dir — add a directory to the include search path exclusively for quoted includes. exclusive!! 👑
  • --system-include-dir dir — add a directory to the include search path and mark it as a system include path (used for dependency output).
  • --embed-dir directory — add a directory to the preprocessor embed search path.
  • --debug-info — include debug information in the produced output. so your debugger can actually find things 😭
  • --optimizer-max-inline-depth depth — maximum depth of function inlining (5 by default).
  • --optimizer-max-inlines-per-func inlines — maximum number of inlined functions (10 by default).
  • --enable-lowering — perform target-specific SSA lowering — and heads up, this is mandatory for bit-precise integer support — [default].
  • --disable-lowering — skip target-specific SSA lowering.
  • --preprocessor-assembly-mode — enable assembly preprocessing mode (default for .S files). In this mode, leading '$' characters are not treated as part of an identifier.
  • --preprocessor-normal-mode — disable assembly preprocessing mode (default).
  • --dump-dependencies — output a make rule describing the dependencies of the source file.
  • --system-dependencies — include dependencies from system include directories in the make rule (used with the --dump-dependencies flag) [default: on].
  • --no-system-dependencies — do not include dependencies from system include directories in the make rule (used with the --dump-dependencies flag).
  • --add-phony-targets — generate phony make targets for each dependency (used with the --dump-dependencies flag) [default: off].
  • --no-add-phony-targets — do not generate phony make targets for each dependency (used with --dump-dependencies).
  • --dependency-target target — override the default make rule target (used with the --dump-dependencies flag).
  • --dependency-output filename — output the make rule into the specified file (used with the --dump-dependencies flag).
  • --unsigned-char — let the type char be unsigned. By default it's target-specific.
  • --signed-char — let the type char be signed. By default it's target-specific.
  • --codegen-tentative-common — place tentative definitions into the common section (equivalent to applying the common attribute to all variables).
  • --codegen-tentative-no-common — do not place tentative definitions into the common section [default].
  • --codegen-visibility-default — keep default visibility for all symbols [default].
  • --codegen-visibility-hidden — make all symbols hidden unless an attribute says otherwise.
  • --codegen-visibility-protected — make all symbols protected unless an attribute says otherwise.
  • --codegen-visibility-internal — make all symbols internal unless an attribute says otherwise.
  • --feature-[name] — enable a compiler feature (see the FEATURES subsection).
  • --no-feature-[name] — disable a compiler feature (see the FEATURES subsection).
  • --internal-[flag] — enable a compiler internal flag (see the INTERNALS subsection).
  • --no-internal-[flag] — disable a compiler internal flag (see the INTERNALS subsection).
  • --codegen-[option] — enable an option for the code generator (see the CODEGEN subsection).
  • --no-codegen-[option] — disable an option for the code generator (see the CODEGEN subsection).
  • --optimizer-pipeline spec — optimizer pipeline specification (see the OPTIMIZER subsection).
  • --precise-bitfield-load-store — load and store bitfields at byte boundaries [default: on]. Gives you extra correctness for structures allocated close to a boundary with unmapped memory pages at the expense of less efficient bitfield handling. Does not change bitfield layout.
  • --no-precise-bitfield-load-store — load and store bitfields at machine word boundaries [default: off]. ⚠️ may cause failures for structures allocated close to boundaries with unmapped memory pages. Does not change bitfield layout.
  • --declare-atomic-support — declare support of atomics (might require explicit linking of a software atomic library such as libatomic or libcompiler_rt).
  • --no-declare-atomic-support — declare the absence of atomic support by defining the __STDC_NO_ATOMICS__ macro.
  • --optimize-stack-frame — optimize stack frame layout by reusing space for variables with disjoint lifetimes. Might increase compilation time and memory use on very large functions. [default: on].
  • --no-optimize-stack-frame — do not optimize stack frame layout, and allocate each variable separately.

🎀 FEATURES 🎀

these are the supported features — use them with the --feature-[feature-name] and --no-feature-[feature-name] flags. heads up: all disabled by default. 🙈

  • fail-on-attributes — fail if __attribute__((...)) is encountered.
  • missing-function-return-type — permit function definitions with a missing return type.
  • designated-init-colons — permit the "fieldname:" syntax in designated initializers.
  • labels-as-values — permit label-addressing with the && operator.
  • non-strict-qualifiers — disable strict qualifier checks for pointers.
  • signed-enums — force all enums to have a signed integral type.
  • implicit-function-decl — permit implicit function declarations at the use-site.
  • empty-structs — permit empty structure/union definitions.
  • ext-pointer-arithmetics — permit pointer arithmetics with function and void pointers.
  • missing-braces-subobj — permit missing braces for subobject initialization with a scalar.
  • statement-expressions — enable statement expressions.
  • omitted-conditional-operand — permit omission of the middle ternary expression operand.
  • int-to-pointer — permit conversion of any integral type to a pointer.
  • permissive-pointer-conv — permit conversions between any pointer types.
  • named-macro-vararg — permit named macro variable arguments.
  • include-next — permit the include_next preprocessor directive.
  • fail-on-assembly — disable support of inline assembly.
  • va-args-comma-concat — enable special processing for the ", ##__VA_ARGS" case in the preprocessor.
  • switch-case-ranges — enable support for ranges in switch cases.
  • designator-subscript-ranges — enable support for ranges in designator subscripts.

🌈 CODEGEN 🌈

these are the supported code generator options — used with the --codegen-[option] and --no-codegen-[option] flags:

  • emulated-tls — use emulated TLS [disabled by default, enabled on openbsd platform].
  • tls-common — generate a common section for thread-local storage if the respective attribute is on [enabled for linux and netbsd].
  • pic — generate position-independent code.
  • omit-frame-pointer — omit the frame pointer in leaf functions that don't need it.
  • valgrind-compatible-x87 — replace x87 opcodes not supported by Valgrind with more expensive alternatives [default: on].
  • syntax=SYNTAX — produce assembly output with the specified syntax — x86_64-intel_noprefix, x86_64-intel_prefix, x86_64-att (default).
  • details=DETAILS-SPEC — augment assembly output with internal code generator details in comments. DETAILS-SPEC can be: vasm (virtual assembly), vasm+regs (virtual assembly and register allocations), devasm (devirtualized assembly).
  • pipeline=PIPELINE-SPEC — code generator transformation pipeline specification (not supported for native-* targets), made up of comma separated names of pipeline passes:
    • noop — no-operation pass.
    • amd64-drop-virtual — non-functional virtual instruction elimination for amd64 targets.
    • amd64-peephole — peephole optimizations for amd64 targets.

🔥 OPTIMIZER 🔥

the optimizer pipeline specification is made up of comma separated names of optimization pipeline passes:

  • noop — no-operation pass.
  • phi-propagate — wherever possible, substitute phi nodes of the SSA representation by actual references.
  • mem2reg — pull function local variables into registers.
  • op-simplify — general code simplification.
  • constant-fold — folding constant expressions.
  • gvn — global value numbering pass to eliminate redundant arithmetic and bitwise operations.
  • local-alloc-sink — moving local variable allocations closer to their actual uses (does not affect effective variable lifetimes).
  • dead-code-elimination — eliminating dead code, blocks and phi links.
  • block-merge — remove unnecessary jump instructions and merge the respective blocks.
  • dead-alloc — remove dead local variable allocations.
  • inline-func — perform function inlining.
  • tail-calls — perform tail call optimization.
  • lowering — perform target-specific lowering of the SSA representation. Mandatory for bit-precise integer support.

🎯 TARGET 🎯

supported compilation targets — yes, there's one, and it's iconic:

  • amd64-sysv-gas — produce optimized GNU As-compatible assembly adhering to the System-V AMD64 ABI [default].

💀 EXIT STATUS 💀

normally kefir-cc1 exits with exit code 0. 🫶

but if there are any errors in any of the compilation stages, all further compilation is aborted and a non-zero exit code is returned. 😭

⭐ STANDARDS ⭐

kefir-cc1 implements the C17 language standard as specified in its final draft, and the C23 language standard (except _Decimal support) as specified in the first draft of C2Y. periodt. 👑

🎀 NOTES 🎀

okay bestie, housekeeping time — the boring-but-actually-important stuff 💅

Licensing tea: the kefir-cc1 compiler is licensed under the terms of the GNU GPLv3 license. periodt. that's the whole vibe. ✨

Who's behind this icon: kefir-cc1 is developed and maintained by Jevgenij Protopopov (legally: Jevgenijs Protopopovs) 🫶 like, put some respect on the name fr fr 👑

Found a bug? 😭 please report bugs found in kefir-cc1 to jevgenij@protopopov.lv. do NOT suffer in silence, bestie — that's not the move 🙈

Where to find it: kefir-cc1 is available online at https://sr.ht/~jprotopopov/kefir/ with read-only mirrors at https://git.protopopov.lv/kefir and https://codeberg.org/jprotopopov/kefir 🌈 go follow the lore iykyk

and that's a wrap on the whole page, bestie 🎀 this cutie comes to you from the Debian fam, dated September 2025, and yes — it's literally kefir-cc1(1) 💖 slay, go compile something iconic ✨🦋
📦 where this page comes from