LLI(1)

📦

version 14.0.6 · 8 versions

💅 LLI — manual page for `lli` 14 💅

okay bestie, sit down, hydrate, because this is lli(1) from the User Commands section of the manual and she is LONG. Like, four-parts long. This is part one. We're doing this together, no skimming, no skipping, every single flag is getting its moment. 🫶

✨ NAME ✨

lli — manual page for lli 14

🦋 DESCRIPTION 🦋

OVERVIEW: llvm interpreter & dynamic compiler

USAGE:

lli [options] <input bitcode> <program arguments>...

OPTIONS:

🎀 Color Options 🎀

--color — Use colors in output (default=autodetect). yes she will make it pretty, and yes she figures out whether your terminal can handle it all by herself ✨

👑 General options 👑

Okay this is the big one. Every single flag. Buckle up, bestie. 💖

-O=<char> — Optimization level. [-O0, -O1, -O2, or -O3] (default = '-O2')
--aarch64-neon-syntax=<value> — Choose style of NEON code to emit from AArch64 backend:
- =generic — Emit generic NEON assembly
- =apple — Emit Apple-style NEON assembly
--aarch64-use-aa — Enable the use of AA during codegen.
--abort-on-max-devirt-iterations-reached — Abort when the max iterations for devirtualization CGSCC repeat pass is reached
--addrsig — Emit an address-significance table
--align-loops=<uint> — Default alignment for loops
--allow-ginsert-as-artifact — Allow G_INSERT to be considered an artifact. Hack around AMDGPU test infinite loops. (a hack, she said it herself, we respect the honesty)
--amdgpu-bypass-slow-div — Skip 64-bit divide for dynamic 32-bit values
--amdgpu-disable-loop-alignment — Do not align and prefetch loops
--amdgpu-disable-power-sched — Disable scheduling to minimize mAI power bursts
--amdgpu-dpp-combine — Enable DPP combiner
--amdgpu-dump-hsa-metadata — Dump AMDGPU HSA Metadata
--amdgpu-enable-flat-scratch — Use flat scratch instructions
--amdgpu-enable-merge-m0 — Merge and hoist M0 initializations
--amdgpu-promote-alloca-to-vector-limit=<uint> — Maximum byte size to consider promote alloca to vector
--amdgpu-sdwa-peephole — Enable SDWA peepholer
--amdgpu-use-aa-in-codegen — Enable the use of AA during codegen.
--amdgpu-verify-hsa-metadata — Verify AMDGPU HSA Metadata
--amdgpu-vgpr-index-mode — Use GPR indexing mode instead of movrel for vector indexing
--arm-add-build-attributes — (no description given, she's simply there, mysterious queen energy 💅)
--arm-implicit-it=<value> — Allow conditional instructions outdside of an IT block (yes, "outdside", we keep the typo, we do not touch the artifacts)
- =always — Accept in both ISAs, emit implicit ITs in Thumb
- =never — Warn in ARM, reject in Thumb
- =arm — Accept in ARM, reject in Thumb
- =thumb — Warn in ARM, emit implicit ITs in Thumb
--asm-show-inst — Emit internal instruction representation to assembly file
--atomic-counter-update-promoted — Do counter update using atomic fetch add for promoted counters only
--atomic-first-counter — Use atomic fetch add for first counter in a function (usually the entry counter)
--basic-block-sections=<all | <function list (file)> | labels | none> — Emit basic blocks into separate sections
--bounds-checking-single-trap — Use one trap block per function
--cfg-hide-cold-paths=<number> — Hide blocks with relative frequency below the given value
--cfg-hide-deoptimize-paths — (description not included in the original, so no guessing from me, bestie)
--cfg-hide-unreachable-paths — (also mysterious, also iconic)
--code-model=<value> — Choose code model
- =tiny — Tiny code model
- =small — Small code model
- =kernel — Kernel code model
- =medium — Medium code model
- =large — Large code model
--compile-threads=<uint> — Choose the number of compile threads (jit-kind=orc-lazy only)
--cost-kind=<value> — Target cost kind
- =throughput — Reciprocal throughput
- =latency — Instruction latency
- =code-size — Code size
- =size-latency — Code size and latency
--data-sections — Emit data into separate sections
--debug-entry-values — Enable debug info for the debug entry values.
--debug-info-correlate — Use debug info to correlate profiles.
--debugger-tune=<value> — Tune debug info for a particular debugger
- =gdb — gdb
- =lldb — lldb
- =dbx — dbx
- =sce — SCE targets (e.g. PS4)
--debugify-level=<value> — Kind of debug info to add
- =locations — Locations only
- =location+variables — Locations and Variables
--debugify-quiet — Suppress verbose debugify output
--denormal-fp-math=<value> — Select which denormal numbers the code is permitted to require
- =ieee — IEEE 754 denormal numbers
- =preserve-sign — the sign of a flushed-to-zero number is preserved in the sign of 0
- =positive-zero — denormals are flushed to positive zero
--denormal-fp-math-f32=<value> — Select which denormal numbers the code is permitted to require for float
- =ieee — IEEE 754 denormal numbers
- =preserve-sign — the sign of a flushed-to-zero number is preserved in the sign of 0
- =positive-zero — denormals are flushed to positive zero
--disable-i2p-p2i-opt — Disables inttoptr/ptrtoint roundtrip optimization
--disable-lazy-compilation — Disable JIT lazy compilation
--disable-promote-alloca-to-lds — Disable promote alloca to LDS
--disable-promote-alloca-to-vector — Disable promote alloca to vector
--disable-tail-calls — Never emit tail calls
--dlopen=<string> — Dynamic libraries to load before linking
--do-counter-promotion — Do counter register promotion
--dot-cfg-mssa=<file name for generated dot file> — file name for generated dot file
--dwarf-version=<int> — Dwarf version
--dwarf64 — Generate debugging info in the 64-bit DWARF format
--emit-call-site-info — Emit call site debug information, if debug information is enabled.
--emscripten-cxx-exceptions-allowed=<string> — The list of function names in which Emscripten-style exception handling is enabled (see emscripten EMSCRIPTEN_CATCHING_ALLOWED options)
--emulated-tls — Use emulated TLS model
--enable-cache-manager — Use cache manager to save/load modules
--enable-cse-in-irtranslator — Should enable CSE in irtranslator
--enable-cse-in-legalizer — Should enable CSE in Legalizer
--enable-emscripten-cxx-exceptions — WebAssembly Emscripten-style exception handling
--enable-emscripten-sjlj — WebAssembly Emscripten-style setjmp/longjmp handling
--enable-gvn-hoist — Enable the GVN hoisting pass (default = off)
--enable-gvn-memdep — (no description, just vibes 🦋)
--enable-gvn-sink — Enable the GVN sinking pass (default = off)
--enable-load-in-loop-pre — (no description given)
--enable-load-pre — (no description given)
--enable-loop-simplifycfg-term-folding — (no description given)
--enable-name-compression — Enable name/filename string compression
--enable-no-infs-fp-math — Enable FP math optimizations that assume no +-Infs
--enable-no-nans-fp-math — Enable FP math optimizations that assume no NaNs
--enable-no-signed-zeros-fp-math — Enable FP math optimizations that assume the sign of 0 is insignificant
--enable-no-trapping-fp-math — Enable setting the FP exceptions build attribute not to use exceptions
--enable-split-backedge-in-load-pre — (no description given)
--enable-unsafe-fp-math — Enable optimizations that may decrease FP precision
--entry-function=<function> — Specify the entry function (default = 'main') of the executable
--exception-model=<value> — exception model
- =default — default exception handling model
- =dwarf — DWARF-like CFI based exception handling
- =sjlj — SjLj exception handling
- =arm — ARM EHABI exceptions

and that's the end of part one, bestie!! do NOT close the tab, there's more, we're so not done 💜

=wineh — the Windows exception model 🪟
- =wasm — WebAssembly exception handling 🕸️

--experimental-debug-variable-locations — use the shiny new experimental value-tracking variable locations. it's giving beta energy, bestie 🙈
--extra-archive=<input archive> — extra archive files to be loaded
--extra-module=<input bitcode> — extra modules to be loaded
--extra-object=<input object> — extra object files to be loaded
--fake-argv0=<executable> — override the 'argv[0]' value that gets passed into the executing program. catfishing the process, fr fr 🎭
--fatal-warnings — treat warnings as errors. no chill mode activated 💀
--filetype=<value> — choose a file type (not all types are supported by all targets, so check before you commit 💅):
- =asm — emit an assembly ('.s') file
- =obj — emit a native object ('.o') file
- =null — emit nothing, purely for performance testing. the void. the emptiness. iconic ✨
--float-abi=<value> — choose your float ABI type
- =default — target default float ABI type
- =soft — soft float ABI (implied by -soft-float)
- =hard — hard float ABI (uses FP registers)
--force-dwarf-frame-section — always emit a debug frame section. every single time. no skipping 🌸
--force-interpreter — force interpretation: disable JIT 🚫
--fp-contract=<value> — enable aggressive formation of fused FP ops
- =fast — fuse FP ops whenever profitable
- =on — only fuse 'blessed' FP ops.
- =off — only fuse FP ops when the result won't be affected.
--frame-pointer=<value> — specify frame pointer elimination optimization
- =all — disable frame pointer elimination
- =non-leaf — disable frame pointer elimination for non-leaf frame
- =none — enable frame pointer elimination
--fs-profile-debug-bw-threshold=<uint> — only show the debug message if the source branch weight is greater than this value.
--fs-profile-debug-prob-diff-threshold=<uint> — only show the debug message if the branch probility is greater than this value (in percentage). (yes, "probility", we're quoting the original, no cap) 😭
--function-sections — emit functions into separate sections
--generate-merged-base-profiles — when generating nested context-sensitive profiles, always generate an extra base profile for a function with all its context profiles merged into it.
--gpsize=<uint> — Global Pointer Addressing Size. the default size is 8.
--hash-based-counter-split — rename the counter variable of a comdat function based on cfg hash
--hot-cold-split — enable the hot-cold splitting pass 🔥🧊
--ignore-xcoff-visibility — don't emit the visibility attribute for asm in AIX OS, or give all symbols 'unspecified' visibility in the XCOFF object file
--import-all-index — import all external functions in index.
--incremental-linker-compatible — when used with filetype=obj, emit an object file which can be used with an incremental linker
--instcombine-code-sinking — enable code sinking 🫧
--instcombine-guard-widening-window=<uint> — how wide an instruction window to bypass looking for another guard
--instcombine-max-iterations=<uint> — limit the maximum number of instruction combining iterations
--instcombine-max-num-phis=<uint> — maximum number of phis to handle in intptr/ptrint folding
--instcombine-maxarray-size=<uint> — maximum array size considered when doing a combine
--instcombine-negator-enabled — should we attempt to sink negations?
--instcombine-negator-max-depth=<uint> — what is the maximal lookup depth when trying to check for viability of negation sinking.
--instrprof-atomic-counter-update-all — make all profile counter updates atomic (for testing only) ⚛️
--internalize-public-api-file=<filename> — a file containing a list of symbol names to preserve
--internalize-public-api-list=<list> — a list of symbol names to preserve
--iterative-counter-promotion — allow counter promotion across the whole loop nest.
--jd=<string> — specifies the JITDylib to be used for any subsequent -extra-module arguments.
--jit-kind=<value> — choose the underlying JIT kind.
- =mcjit — MCJIT
- =orc — Orc JIT
- =orc-lazy — Orc-based lazy JIT. 💤
--jit-linker=<value> — choose the dynamic linker/loader.
- =default — default for platform and JIT-kind
- =rtdyld — RuntimeDyld
- =jitlink — Orc-specific linker
--load=<pluginfilename> — load the specified plugin 🔌
--lto-embed-bitcode=<value> — embed LLVM bitcode in object files produced by LTO
- =none — do not embed
- =optimized — embed after all optimization passes
- =post-merge-pre-opt — embed post merge, but before optimizations
--lto-pass-remarks-filter=<regex> — only record optimization remarks from passes whose names match the given regular expression
--lto-pass-remarks-format=<format> — the format used for serializing remarks (default: YAML)
--lto-pass-remarks-output=<filename> — output filename for pass remarks
--march=<string> — architecture to generate code for (see --version)
--matrix-default-layout=<value> — sets the default matrix layout
- =column-major — use column-major layout
- =row-major — use row-major layout
--mattr=<a1,+a2,-a3,...> — target specific attributes (-mattr=help for details)
--max-counter-promotions=<int> — max number of allowed counter promotions
--max-counter-promotions-per-loop=<uint> — max number of counter promotions per loop, so we don't increase register pressure too much
--mc-relax-all — when used with filetype=obj, relax all fixups in the emitted object file
--mcabac — tbd. that's it. that's the whole description. mysterious queen behavior 👑
--mcjit-remote-process=<filename> — specify the filename of the process to launch for remote MCJIT execution. if none is specified, remote execution will be simulated in-process.
--mcpu=<cpu-name> — target a specific cpu type (-mcpu=help for details)
--meabi=<value> — set EABI type (default depends on triple):
- =default — triple default EABI version
- =4 — EABI version 4
- =5 — EABI version 5
- =gnu — EABI GNU
--merror-missing-parenthesis — error for missing parenthesis around predicate registers
--merror-noncontigious-register — error for register names that aren't contigious (again with the spelling, we're keeping it exactly as written bestie) 🙈
--mhvx — enable Hexagon Vector eXtensions
--mhvx=<value> — enable Hexagon Vector eXtensions
- =v60 — build for HVX v60
- =v62 — build for HVX v62
- =v65 — build for HVX v65
- =v66 — build for HVX v66
- =v67 — build for HVX v67
- =v68 — build for HVX v68
- =v69 — build for HVX v69
--mips-compact-branches=<value> — MIPS Specific: Compact branch policy.
- =never — do not use compact branches if possible.
- =optimal — use compact branches where appropriate (default).
- =always — always use compact branches if possible.
--mips16-constant-islands — enable mips16 constant islands.
--mips16-hard-float — enable mips16 hard float.
--mir-strip-debugify-only — should mir-strip-debug only strip debug info from debugified modules by default
--mno-compound — disable looking for compound instructions for Hexagon
--mno-fixup — disable fixing up resolved relocations for Hexagon
--mno-ldc1-sdc1 — expand double precision loads and stores to their single precision counterparts
--mno-pairing — disable looking for duplex instructions for Hexagon
--mtriple=<string> — override target triple for module
--mwarn-missing-parenthesis — warn for missing parenthesis around predicate registers
--mwarn-noncontigious-register — warn for register names that arent contigious
--mwarn-sign-mismatch — bestie, this one warns for mismatching a signed and unsigned value. it's giving “type safety said no” 💅
--no-deprecated-warn — Suppress all deprecated warnings. the deprecated stuff can still exist, it just stops yelling at you about it. lowkey peaceful 🫶
--no-discriminators — Disable generation of discriminator information. bye bye discriminators, not today 💀
--no-process-syms — Do not resolve lli process symbols in JIT'd code. no cap ✨
--no-type-check — Suppress type errors (Wasm). only Wasm, iykyk 🌸
--no-warn — Suppress all warnings. silence is golden, periodt 👑
--no-xray-index — Don't emit xray_fn_idx section. that section? not invited 💅
--nozero-initialized-in-bss — Don't place zero-initialized symbols into bss section. they're not moving into bss today 🙈
--nvptx-sched4reg — NVPTX Specific: schedule for register pressue. (yes, “pressue” is literally how it's spelled in the original, I'm not fixing it 💀) NVPTX only, bestie.
--object-cache-dir=<string> — Directory to store cached object files (must be user writable). so like, make sure you can write there or it's gonna be a whole thing 😭
--opaque-pointers — Use opaque pointers. slay ✨
--per-module-lazy — Performs lazy compilation on whole module boundaries rather than individual functions. it's giving big-picture lazy, not tiny-function lazy 🦋
--poison-checking-function-local — Check that returns are non-poison (for testing). for testing only, bestie 🧚
--print-pipeline-passes — Print a '-passes' compatible string describing the pipeline (best-effort only). best-effort only, so don't come for it if it's not perfect 💖
--r600-ir-structurize — Use StructurizeCFG IR pass. R600 said structure, periodt 🌈
--rdf-dump — no description in the original, just an empty lil space. mysterious main character energy 💀
--rdf-limit=<uint> — also no description, so... vibes only. it takes a <uint> though ✨
--relax-elf-relocations — Emit GOTPCRELX/REX_GOTPCRELX instead of GOTPCREL on x86-64 ELF. x86-64 ELF specific, no cap 🔥
--relocation-model=<value> — Choose relocation model. the options are giving choose-your-own-adventure:
- =static — Non-relocatable code
- =pic — Fully relocatable, position independent code
- =dynamic-no-pic — Relocatable external references, non-relocatable code
- =ropi — Code and read-only data relocatable, accessed PC-relative
- =rwpi — Read-write data relocatable, accessed relative to static base
- =ropi-rwpi — Combination of ropi and rwpi
--remote-mcjit — Execute MCJIT'ed code in a separate process. separate process, separate vibes 💅
--runtime-counter-relocation — Enable relocating counters at runtime. runtime relocation, slay ✨
--safepoint-ir-verifier-print-only — no description again, just the option name and a dream ✨
--sample-profile-check-record-coverage=<N> — Emit a warning if less than N% of records in the input profile are matched to the IR. so if coverage is low, it's gonna tattle 👀
--sample-profile-check-sample-coverage=<N> — Emit a warning if less than N% of samples in the input profile are matched to the IR. same energy but for samples 💖
--sample-profile-max-propagate-iterations=<uint> — Maximum number of iterations to go through when propagating sample block/edge weights through the CFG. it's giving max effort 🔥
--skip-ret-exit-block — Suppress counter promotion if exit blocks contain ret. if there's a ret, promotion is not happening 🙅‍♀️
--soft-float — Generate software floating point library calls. software float, iconic 🌸
--speculative-counter-promotion-max-exiting=<uint> — The max number of exiting blocks of a loop to allow speculative counter promotion. so like, how many exits before we stop being speculative? 🤔
--speculative-counter-promotion-to-loop — When the option is false, if the target block is in a loop, the promotion will be disallowed unless the promoted counter update can be further/iteratively promoted into an acyclic region. read that twice, bestie, it's a whole sentence 😭
--split-machine-functions — Split out cold basic blocks from machine functions based on profile information. cold blocks get their own lil section, periodt 🧊
--stack-size-section — Emit a section containing stack size metadata. stack size metadata said hello 👋
--stack-symbol-ordering — Order local stack symbols. alphabetical? no, just ordered. don't ask me, I just write the vibe 💅
--stackrealign — Force align the stack to the minimum alignment. alignment forced, no negotiation 💀
--strict-dwarf — use strict dwarf. lowercase “use” and all, iconic ✨
--summary-file=<string> — The summary file to use for function importing. it's the file that says who gets imported 🎀
--swift-async-fp=<value> — Determine when the Swift async frame pointer should be set. choose your fighter:
- =auto — Determine based on deployment target
- =always — Always set the bit
- =never — Never set the bit
--tail-predication=<value> — MVE tail-predication pass options. the menu is:
- =disabled — Don't tail-predicate loops
- =enabled-no-reductions — Enable tail-predication, but not for reduction loops
- =enabled — Enable tail-predication, including reduction loops
- =force-enabled-no-reductions — Enable tail-predication, but not for reduction loops, and force this which might be unsafe
- =force-enabled — Enable tail-predication, including reduction loops, and force this which might be unsafe

the force ones might be unsafe, so be careful bestie ⚠️

--tailcallopt — Turn fastcc calls into tail calls by (potentially) changing ABI. ABI might change, so heads up 🫶
--thinlto-assume-merged — Assume the input has already undergone ThinLTO function importing and the other pre-optimization pipeline changes. it's assuming, okay? 🙈
--thread-entry=<string> — calls the given entry-point on a new thread (jit-kind=orc-lazy only). only for jit-kind=orc-lazy, iykyk 🧵
--thread-model=<value> — Choose threading model. pick your thread vibe:
- =posix — POSIX thread model
- =single — Single thread model
--threads=<int> — no description, just a number of threads and a dream 💀
--tls-size=<uint> — Bit size of immediate TLS offsets. TLS offsets said bit size ✨
--unique-basic-block-section-names — Give unique names to every basic block section. uniqueness slay 💅
--unique-section-names — Give unique names to every section. every section gets a unique name, no duplicates 👑
--use-ctors — Use .ctors instead of .init_array. old school? maybe. iconic? yes 🌈
--vec-extabi — Enable the AIX Extended Altivec ABI. AIX extended altivec ABI, periodt 💖
--verify-region-info — Verify region info (time consuming). time consuming, so maybe grab a snack 🍓
--vp-counters-per-site=<number> — The average number of profile counters allocated per value profiling site. average, bestie ✨
--vp-static-alloc — Do static counter allocation for value profiler. static allocation, no cap 💅
--wasm-enable-eh — WebAssembly exception handling. Wasm EH said hi 👋
--wasm-enable-sjlj — WebAssembly setjmp/longjmp handling. setjmp/longjmp but make it Wasm 🦋
--x86-align-branch=<string> — Specify types of branches to align (plus separated list of types):
- jcc — indicates conditional jumps
- fused — indicates fused conditional jumps
- jmp — indicates direct unconditional jumps
- call — indicates direct and indirect calls
- ret — indicates rets
- indirect — indicates indirect unconditional jumps
--x86-align-branch-boundary=<uint> — Control how the assembler should align branches with NOP. If the boundary's size is not 0, it should be a power of 2 and no less than 32. Branches will be aligned to prevent from being across or against the boundary of specified size. The default value 0 does not align branches. so default 0 = no alignment, and if you set it, power of 2, at least 32. got it 🔥
--x86-branches-within-32B-boundaries — Align selected instructions to mitigate negative performance impact of Intel's micro code update for errata skx102. May break assumptions about labels corresponding to particular instructions, and should be used with caution. ⚠️ caution bestie, it can break label assumptions.
--x86-pad-max-prefix-size=<uint> — Maximum number of prefixes to use for padding. max prefixes, periodt 💅
--xcoff-traceback-table — Emit the XCOFF traceback table. XCOFF traceback table said hello 👑

🌈 Generic Options 🌈

--help — Display available options (--help-hidden for more). so meta, bestie ✨
--help-list — Display list of available options (--help-list-hidden for more). list girlies rise 💖
--version — Display the version of this program. version check, slay 👑

🦋 Polly Options: Configure the polly loop optimizer 🦋

--polly — Enable the polly optimizer (with -O1, -O2 or -O3). only with those optimization levels, iykyk 🔥
--polly-2nd-level-tiling — Enable a 2nd level loop of loop tiling. second level tiling, obsessed 💅
--polly-ast-print-accesses — Print memory access functions. memory access functions said print me 🎀
--polly-context=<isl parameter set> — Provide additional constraints on the context parameters. context constraints, periodt ✨
--polly-dce-precise-steps=<int> — The number of precise steps between two approximating iterations. (A value of -1 schedules another approximation stage before the actual dead code elimination. (yes, the paren is open, iconic 💀)
--polly-delicm-max-ops=<int> — Maximum number of isl operations to invest for lifetime analysis; 0=no limit. zero means no limit, so go off 🫶
--polly-detect-full-functions — Allow the detection of full functions. full functions detected, slay 🌈
--polly-dump-after — drops a dump of the module after Polly has done its thing ✨ into a file whose name gets the suffix "-after" glued on the end 💅
--polly-dump-after-file=<string> — same energy as above, but you pick the file, bestie — dump the post-Polly module straight into the <string> path you hand it 🎀
--polly-dump-before — dump the module before Polly touches anything, and it lands in a file suffixed with "-before" 🌸
--polly-dump-before-file=<string> — pre-Polly module, your choice of destination — write it to <string> 🦋
--polly-enable-simplify — simplifies the SCoP after optimizations have run 💖
--polly-ignore-func=<string> — skips any function matching a regex. You can pass multiple regexes separated by commas. Scop detection will ignore every function that matches ANY of the regexes you gave it 🙈
--polly-isl-arg=<argument> — an option handed straight through to ISL
--polly-on-isl-error-abort — the moment an isl error shows up, we abort. no chill, no recovery 💀
--polly-only-func=<string> — only run on functions matching a regex. Multiple regexes = comma separated. Scop detection will run on all functions that match ANY of the provided regexes ⭐
--polly-only-region=<identifier> — only run on certain regions. (The provided identifier must appear in the name of the region's entry block 👑
--polly-only-scop-detection — scop detection only, bestie — no other optimizations ride along 🧚
--polly-optimized-scops — Polly: dumps the polyhedral description of Scops optimized with the isl scheduling optimizer, and the set of post-scheduling transformations is applied on the schedule tree 🌈
--polly-parallel — generates thread parallel code (isl codegen only) 🔥
--polly-parallel-force — forces generation of thread parallel code and ignores any cost model while doing it 💅
--polly-pattern-matching-based-opts — optimizations based on pattern matching ✨
--polly-postopts — applies post-rescheduling optimizations like tiling. Heads up: this one requires -polly-reschedule ⚠️
--polly-pragma-based-opts — applies user-directed transformation from metadata 🎀
--polly-pragma-ignore-depcheck — skips the dependency check for pragma-based transformations 😳
--polly-process-unprofitable — processes scops that are unlikely to benefit from Polly optimizations. Chaotic? maybe. Iconic? absolutely 👑
--polly-register-tiling — enables register tiling 🍓
--polly-report — prints info about what Polly's been up to, like a little gossip session 💖
--polly-reschedule — optimizes SCoPs using ISL 🌟
--polly-show — highlights the code regions that will be optimized in a (CFG BBs and LLVM-IR instructions) 🦋
--polly-show-only — highlights the code regions that will be optimized in a (CFG only BBs) 🌸
--polly-stmt-granularity=<value> — the algorithm used for splitting basic blocks into multiple statements. Pick your fighter:
- =bb — one statement per basic block
- =scalar-indep — scalar independence heuristic
- =store — store-level granularity
--polly-target=<value> — the hardware to target:
- =cpu — generate CPU code
--polly-tiling — enables loop tiling 🧩
--polly-vectorizer=<value> — picks the vectorization strategy:
- =none — No Vectorization
- =polly — Polly internal vectorizer
- =stripmine — strip-mine outer loops for the loop-vectorizer to trigger

and that's the whole Polly buffet, bestie 🫶 go forth and optimize, you absolute legend.

this page is lli version 14, dated May 2025 — so if you're reading this, shoutout to the version, shoutout to the date, they ate 💅✨

📦 where this page comes from

ubuntu questing: llvm-14-runtime 1:14.0.6-20ubuntu3
💖 girlified by deepseek-ai/DeepSeek-V4.1-Flash on 2026-10-13: a language model's rewrite of the original, under its terms