Next: Instrumentation Options, Previous: Debugging Options, Up: Invoking GCC [Contents][Index]
These options control various sorts of optimizations.
Without any optimization option, the compiler’s goal is to reduce the cost of compilation and to make debugging produce the expected results. Statements are independent: if you stop the program with a breakpoint between statements, you can then assign a new value to any variable or change the program counter to any other statement in the function and get exactly the results you expect from the source code.
Turning on optimization flags makes the compiler attempt to improve the performance and/or code size at the expense of compilation time and possibly the ability to debug the program.
The compiler performs optimization based on the knowledge it has of the program. Compiling multiple files at once to a single output file mode allows the compiler to use information gained from all of the files when compiling each of them.
Not all optimizations are controlled directly by a flag. Only optimizations that have a flag are listed in this section.
Most optimizations are only enabled if an -O level is set on the command line. Otherwise they are disabled, even if individual optimization flags are specified.
Depending on the target and how GCC was configured, a slightly different set of optimizations may be enabled at each -O level than those listed here. You can invoke GCC with -Q --help=optimizers to find out the exact set of optimizations that are enabled at each level. See Overall Options, for examples.
-O-O1Optimize. Optimizing compilation takes somewhat more time, and a lot more memory for a large function.
With -O, the compiler tries to reduce code size and execution time, without performing any optimizations that take a great deal of compilation time.
-O turns on the following optimization flags:
-fauto-inc-dec -fbranch-count-reg -fcombine-stack-adjustments -fcompare-elim -fcprop-registers -fdce -fdefer-pop -fdelayed-branch -fdse -fforward-propagate -fguess-branch-probability -fif-conversion2 -fif-conversion -finline-functions-called-once -fipa-pure-const -fipa-profile -fipa-reference -fmerge-constants -fmove-loop-invariants -freorder-blocks -fshrink-wrap -fsplit-wide-types -fssa-backprop -fssa-phiopt -ftree-bit-ccp -ftree-ccp -ftree-ch -ftree-coalesce-vars -ftree-copy-prop -ftree-dce -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre -ftree-phiprop -ftree-sink -ftree-slsr -ftree-sra -ftree-pta -ftree-ter -funit-at-a-time
-O also turns on -fomit-frame-pointer on machines where doing so does not interfere with debugging.
-O2Optimize even more. GCC performs nearly all supported optimizations that do not involve a space-speed tradeoff. As compared to -O, this option increases both compilation time and the performance of the generated code.
-O2 turns on all optimization flags specified by -O. It also turns on the following optimization flags:
-fthread-jumps -falign-functions -falign-jumps -falign-loops -falign-labels -fcaller-saves -fcrossjumping -fcse-follow-jumps -fcse-skip-blocks -fdelete-null-pointer-checks -fdevirtualize -fdevirtualize-speculatively -fexpensive-optimizations -fgcse -fgcse-lm -fhoist-adjacent-loads -finline-small-functions -findirect-inlining -fipa-cp -fipa-cp-alignment -fipa-sra -fipa-icf -fisolate-erroneous-paths-dereference -flra-remat -foptimize-sibling-calls -foptimize-strlen -fpartial-inlining -fpeephole2 -freorder-blocks-algorithm=stc -freorder-blocks-and-partition -freorder-functions -frerun-cse-after-loop -fsched-interblock -fsched-spec -fschedule-insns -fschedule-insns2 -fstrict-aliasing -fstrict-overflow -ftree-builtin-call-dce -ftree-switch-conversion -ftree-tail-merge -ftree-pre -ftree-vrp -fipa-ra
Please note the warning under -fgcse about invoking -O2 on programs that use computed gotos.
-O3Optimize yet more. -O3 turns on all optimizations specified by -O2 and also turns on the -finline-functions, -funswitch-loops, -fpredictive-commoning, -fgcse-after-reload, -ftree-loop-vectorize, -ftree-loop-distribute-patterns, -fsplit-paths -ftree-slp-vectorize, -fvect-cost-model, -ftree-partial-pre and -fipa-cp-clone options.
-O0Reduce compilation time and make debugging produce the expected results. This is the default.
-OsOptimize for size. -Os enables all -O2 optimizations that do not typically increase code size. It also performs further optimizations designed to reduce code size.
-Os disables the following optimization flags:
-falign-functions -falign-jumps -falign-loops -falign-labels -freorder-blocks -freorder-blocks-algorithm=stc -freorder-blocks-and-partition -fprefetch-loop-arrays
-OfastDisregard strict standards compliance. -Ofast enables all -O3 optimizations. It also enables optimizations that are not valid for all standard-compliant programs. It turns on -ffast-math and the Fortran-specific -fno-protect-parens and -fstack-arrays.
-OgOptimize debugging experience. -Og enables optimizations that do not interfere with debugging. It should be the optimization level of choice for the standard edit-compile-debug cycle, offering a reasonable level of optimization while maintaining fast compilation and a good debugging experience.
If you use multiple -O options, with or without level numbers, the last such option is the one that is effective.
Options of the form -fflag specify machine-independent flags. Most flags have both positive and negative forms; the negative form of -ffoo is -fno-foo. In the table below, only one of the forms is listed—the one you typically use. You can figure out the other form by either removing ‘no-’ or adding it.
The following options control specific optimizations. They are either activated by -O options or are related to ones that are. You can use the following flags in the rare cases when “fine-tuning” of optimizations to be performed is desired.
-fno-defer-popAlways pop the arguments to each function call as soon as that function returns. For machines that must pop arguments after a function call, the compiler normally lets arguments accumulate on the stack for several function calls and pops them all at once.
Disabled at levels -O, -O2, -O3, -Os.
-fforward-propagatePerform a forward propagation pass on RTL. The pass tries to combine two instructions and checks if the result can be simplified. If loop unrolling is active, two passes are performed and the second is scheduled after loop unrolling.
This option is enabled by default at optimization levels -O, -O2, -O3, -Os.
-ffp-contract=style-ffp-contract=off disables floating-point expression contraction. -ffp-contract=fast enables floating-point expression contraction such as forming of fused multiply-add operations if the target has native support for them. -ffp-contract=on enables floating-point expression contraction if allowed by the language standard. This is currently not implemented and treated equal to -ffp-contract=off.
The default is -ffp-contract=fast.
-fomit-frame-pointerDon’t keep the frame pointer in a register for functions that don’t need one. This avoids the instructions to save, set up and restore frame pointers; it also makes an extra register available in many functions. It also makes debugging impossible on some machines.
On some machines, such as the VAX, this flag has no effect, because
the standard calling sequence automatically handles the frame pointer
and nothing is saved by pretending it doesn’t exist. The
machine-description macro FRAME_POINTER_REQUIRED controls
whether a target machine supports this flag. See Register
Usage in GNU Compiler Collection (GCC) Internals.
The default setting (when not optimizing for size) for 32-bit GNU/Linux x86 and 32-bit Darwin x86 targets is -fomit-frame-pointer. You can configure GCC with the --enable-frame-pointer configure option to change the default.
Enabled at levels -O, -O2, -O3, -Os.
-foptimize-sibling-callsOptimize sibling and tail recursive calls.
Enabled at levels -O2, -O3, -Os.
-foptimize-strlenOptimize various standard C string functions (e.g. strlen,
strchr or strcpy) and
their _FORTIFY_SOURCE counterparts into faster alternatives.
Enabled at levels -O2, -O3.
-fno-inlineDo not expand any functions inline apart from those marked with
the always_inline attribute. This is the default when not
optimizing.
Single functions can be exempted from inlining by marking them
with the noinline attribute.
-finline-small-functionsIntegrate functions into their callers when their body is smaller than expected function call code (so overall size of program gets smaller). The compiler heuristically decides which functions are simple enough to be worth integrating in this way. This inlining applies to all functions, even those not declared inline.
Enabled at level -O2.
-findirect-inliningInline also indirect calls that are discovered to be known at compile time thanks to previous inlining. This option has any effect only when inlining itself is turned on by the -finline-functions or -finline-small-functions options.
Enabled at level -O2.
-finline-functionsConsider all functions for inlining, even if they are not declared inline. The compiler heuristically decides which functions are worth integrating in this way.
If all calls to a given function are integrated, and the function is
declared static, then the function is normally not output as
assembler code in its own right.
Enabled at level -O3.
-finline-functions-called-onceConsider all static functions called once for inlining into their
caller even if they are not marked inline. If a call to a given
function is integrated, then the function is not output as assembler code
in its own right.
Enabled at levels -O1, -O2, -O3 and -Os.
-fearly-inliningInline functions marked by always_inline and functions whose body seems
smaller than the function call overhead early before doing
-fprofile-generate instrumentation and real inlining pass. Doing so
makes profiling significantly cheaper and usually inlining faster on programs
having large chains of nested wrapper functions.
Enabled by default.
-fipa-sraPerform interprocedural scalar replacement of aggregates, removal of unused parameters and replacement of parameters passed by reference by parameters passed by value.
Enabled at levels -O2, -O3 and -Os.
-finline-limit=nBy default, GCC limits the size of functions that can be inlined. This flag allows coarse control of this limit. n is the size of functions that can be inlined in number of pseudo instructions.
Inlining is actually controlled by a number of parameters, which may be specified individually by using --param name=value. The -finline-limit=n option sets some of these parameters as follows:
max-inline-insns-singleis set to n/2.
max-inline-insns-autois set to n/2.
See below for a documentation of the individual parameters controlling inlining and for the defaults of these parameters.
Note: there may be no value to -finline-limit that results in default behavior.
Note: pseudo instruction represents, in this particular context, an abstract measurement of function’s size. In no way does it represent a count of assembly instructions and as such its exact meaning might change from one release to an another.
-fno-keep-inline-dllexportThis is a more fine-grained version of -fkeep-inline-functions,
which applies only to functions that are declared using the dllexport
attribute or declspec (See Declaring Attributes of
Functions.)
-fkeep-inline-functionsIn C, emit static functions that are declared inline
into the object file, even if the function has been inlined into all
of its callers. This switch does not affect functions using the
extern inline extension in GNU C90. In C++, emit any and all
inline functions into the object file.
-fkeep-static-functionsEmit static functions into the object file, even if the function
is never used.
-fkeep-static-constsEmit variables declared static const when optimization isn’t turned
on, even if the variables aren’t referenced.
GCC enables this option by default. If you want to force the compiler to check if a variable is referenced, regardless of whether or not optimization is turned on, use the -fno-keep-static-consts option.
-fmerge-constantsAttempt to merge identical constants (string constants and floating-point constants) across compilation units.
This option is the default for optimized compilation if the assembler and linker support it. Use -fno-merge-constants to inhibit this behavior.
Enabled at levels -O, -O2, -O3, -Os.
-fmerge-all-constantsAttempt to merge identical constants and identical variables.
This option implies -fmerge-constants. In addition to -fmerge-constants this considers e.g. even constant initialized arrays or initialized constant variables with integral or floating-point types. Languages like C or C++ require each variable, including multiple instances of the same variable in recursive calls, to have distinct locations, so using this option results in non-conforming behavior.
-fmodulo-schedPerform swing modulo scheduling immediately before the first scheduling pass. This pass looks at innermost loops and reorders their instructions by overlapping different iterations.
-fmodulo-sched-allow-regmovesPerform more aggressive SMS-based modulo scheduling with register moves allowed. By setting this flag certain anti-dependences edges are deleted, which triggers the generation of reg-moves based on the life-range analysis. This option is effective only with -fmodulo-sched enabled.
-fno-branch-count-regAvoid running a pass scanning for opportunities to use “decrement and branch” instructions on a count register instead of generating sequences of instructions that decrement a register, compare it against zero, and then branch based upon the result. This option is only meaningful on architectures that support such instructions, which include x86, PowerPC, IA-64 and S/390. Note that the -fno-branch-count-reg option doesn’t remove the decrement and branch instructions from the generated instruction stream introduced by other optimization passes.
Enabled by default at -O1 and higher.
The default is -fbranch-count-reg.
-fno-function-cseDo not put function addresses in registers; make each instruction that calls a constant function contain the function’s address explicitly.
This option results in less efficient code, but some strange hacks that alter the assembler output may be confused by the optimizations performed when this option is not used.
The default is -ffunction-cse
-fno-zero-initialized-in-bssIf the target supports a BSS section, GCC by default puts variables that are initialized to zero into BSS. This can save space in the resulting code.
This option turns off this behavior because some programs explicitly rely on variables going to the data section—e.g., so that the resulting executable can find the beginning of that section and/or make assumptions based on that.
The default is -fzero-initialized-in-bss.
-fthread-jumpsPerform optimizations that check to see if a jump branches to a location where another comparison subsumed by the first is found. If so, the first branch is redirected to either the destination of the second branch or a point immediately following it, depending on whether the condition is known to be true or false.
Enabled at levels -O2, -O3, -Os.
-fsplit-wide-typesWhen using a type that occupies multiple registers, such as long
long on a 32-bit system, split the registers apart and allocate them
independently. This normally generates better code for those types,
but may make debugging more difficult.
Enabled at levels -O, -O2, -O3, -Os.
-fcse-follow-jumpsIn common subexpression elimination (CSE), scan through jump instructions
when the target of the jump is not reached by any other path. For
example, when CSE encounters an if statement with an
else clause, CSE follows the jump when the condition
tested is false.
Enabled at levels -O2, -O3, -Os.
-fcse-skip-blocksThis is similar to -fcse-follow-jumps, but causes CSE to
follow jumps that conditionally skip over blocks. When CSE
encounters a simple if statement with no else clause,
-fcse-skip-blocks causes CSE to follow the jump around the
body of the if.
Enabled at levels -O2, -O3, -Os.
-frerun-cse-after-loopRe-run common subexpression elimination after loop optimizations are performed.
Enabled at levels -O2, -O3, -Os.
-fgcsePerform a global common subexpression elimination pass. This pass also performs global constant and copy propagation.
Note: When compiling a program using computed gotos, a GCC extension, you may get better run-time performance if you disable the global common subexpression elimination pass by adding -fno-gcse to the command line.
Enabled at levels -O2, -O3, -Os.
-fgcse-lmWhen -fgcse-lm is enabled, global common subexpression elimination attempts to move loads that are only killed by stores into themselves. This allows a loop containing a load/store sequence to be changed to a load outside the loop, and a copy/store within the loop.
Enabled by default when -fgcse is enabled.
-fgcse-smWhen -fgcse-sm is enabled, a store motion pass is run after global common subexpression elimination. This pass attempts to move stores out of loops. When used in conjunction with -fgcse-lm, loops containing a load/store sequence can be changed to a load before the loop and a store after the loop.
Not enabled at any optimization level.
-fgcse-lasWhen -fgcse-las is enabled, the global common subexpression elimination pass eliminates redundant loads that come after stores to the same memory location (both partial and full redundancies).
Not enabled at any optimization level.
-fgcse-after-reloadWhen -fgcse-after-reload is enabled, a redundant load elimination pass is performed after reload. The purpose of this pass is to clean up redundant spilling.
-faggressive-loop-optimizationsThis option tells the loop optimizer to use language constraints to derive bounds for the number of iterations of a loop. This assumes that loop code does not invoke undefined behavior by for example causing signed integer overflows or out-of-bound array accesses. The bounds for the number of iterations of a loop are used to guide loop unrolling and peeling and loop exit test optimizations. This option is enabled by default.
-funsafe-loop-optimizationsThis option tells the loop optimizer to assume that loop indices do not overflow, and that loops with nontrivial exit condition are not infinite. This enables a wider range of loop optimizations even if the loop optimizer itself cannot prove that these assumptions are valid. If you use -Wunsafe-loop-optimizations, the compiler warns you if it finds this kind of loop.
-funconstrained-commonsThis option tells the compiler that variables declared in common blocks (e.g. Fortran) may later be overridden with longer trailing arrays. This prevents certain optimizations that depend on knowing the array bounds.
-fcrossjumpingPerform cross-jumping transformation. This transformation unifies equivalent code and saves code size. The resulting code may or may not perform better than without cross-jumping.
Enabled at levels -O2, -O3, -Os.
-fauto-inc-decCombine increments or decrements of addresses with memory accesses. This pass is always skipped on architectures that do not have instructions to support this. Enabled by default at -O and higher on architectures that support this.
-fdcePerform dead code elimination (DCE) on RTL. Enabled by default at -O and higher.
-fdsePerform dead store elimination (DSE) on RTL. Enabled by default at -O and higher.
-fif-conversionAttempt to transform conditional jumps into branch-less equivalents. This includes use of conditional moves, min, max, set flags and abs instructions, and some tricks doable by standard arithmetics. The use of conditional execution on chips where it is available is controlled by -fif-conversion2.
Enabled at levels -O, -O2, -O3, -Os.
-fif-conversion2Use conditional execution (where available) to transform conditional jumps into branch-less equivalents.
Enabled at levels -O, -O2, -O3, -Os.
-fdeclone-ctor-dtorThe C++ ABI requires multiple entry points for constructors and destructors: one for a base subobject, one for a complete object, and one for a virtual destructor that calls operator delete afterwards. For a hierarchy with virtual bases, the base and complete variants are clones, which means two copies of the function. With this option, the base and complete variants are changed to be thunks that call a common implementation.
Enabled by -Os.
-fdelete-null-pointer-checksAssume that programs cannot safely dereference null pointers, and that no code or data element resides at address zero. This option enables simple constant folding optimizations at all optimization levels. In addition, other optimization passes in GCC use this flag to control global dataflow analyses that eliminate useless checks for null pointers; these assume that a memory access to address zero always results in a trap, so that if a pointer is checked after it has already been dereferenced, it cannot be null.
Note however that in some environments this assumption is not true. Use -fno-delete-null-pointer-checks to disable this optimization for programs that depend on that behavior.
This option is enabled by default on most targets. On Nios II ELF, it defaults to off. On AVR and CR16, this option is completely disabled.
Passes that use the dataflow information are enabled independently at different optimization levels.
-fdevirtualizeAttempt to convert calls to virtual functions to direct calls. This is done both within a procedure and interprocedurally as part of indirect inlining (-findirect-inlining) and interprocedural constant propagation (-fipa-cp). Enabled at levels -O2, -O3, -Os.
-fdevirtualize-speculativelyAttempt to convert calls to virtual functions to speculative direct calls. Based on the analysis of the type inheritance graph, determine for a given call the set of likely targets. If the set is small, preferably of size 1, change the call into a conditional deciding between direct and indirect calls. The speculative calls enable more optimizations, such as inlining. When they seem useless after further optimization, they are converted back into original form.
-fdevirtualize-at-ltransStream extra information needed for aggressive devirtualization when running the link-time optimizer in local transformation mode. This option enables more devirtualization but significantly increases the size of streamed data. For this reason it is disabled by default.
-fexpensive-optimizationsPerform a number of minor optimizations that are relatively expensive.
Enabled at levels -O2, -O3, -Os.
-freeAttempt to remove redundant extension instructions. This is especially helpful for the x86-64 architecture, which implicitly zero-extends in 64-bit registers after writing to their lower 32-bit half.
Enabled for Alpha, AArch64 and x86 at levels -O2, -O3, -Os.
-fno-lifetime-dseIn C++ the value of an object is only affected by changes within its lifetime: when the constructor begins, the object has an indeterminate value, and any changes during the lifetime of the object are dead when the object is destroyed. Normally dead store elimination will take advantage of this; if your code relies on the value of the object storage persisting beyond the lifetime of the object, you can use this flag to disable this optimization. To preserve stores before the constructor starts (e.g. because your operator new clears the object storage) but still treat the object as dead after the destructor you, can use -flifetime-dse=1. The default behavior can be explicitly selected with -flifetime-dse=2. -flifetime-dse=0 is equivalent to -fno-lifetime-dse.
-flive-range-shrinkageAttempt to decrease register pressure through register live range shrinkage. This is helpful for fast processors with small or moderate size register sets.
-fira-algorithm=algorithmUse the specified coloring algorithm for the integrated register allocator. The algorithm argument can be ‘priority’, which specifies Chow’s priority coloring, or ‘CB’, which specifies Chaitin-Briggs coloring. Chaitin-Briggs coloring is not implemented for all architectures, but for those targets that do support it, it is the default because it generates better code.
-fira-region=regionUse specified regions for the integrated register allocator. The region argument should be one of the following:
Use all loops as register allocation regions. This can give the best results for machines with a small and/or irregular register set.
Use all loops except for loops with small register pressure as the regions. This value usually gives the best results in most cases and for most architectures, and is enabled by default when compiling with optimization for speed (-O, -O2, …).
Use all functions as a single region. This typically results in the smallest code size, and is enabled by default for -Os or -O0.
-fira-hoist-pressureUse IRA to evaluate register pressure in the code hoisting pass for decisions to hoist expressions. This option usually results in smaller code, but it can slow the compiler down.
This option is enabled at level -Os for all targets.
-fira-loop-pressureUse IRA to evaluate register pressure in loops for decisions to move loop invariants. This option usually results in generation of faster and smaller code on machines with large register files (>= 32 registers), but it can slow the compiler down.
This option is enabled at level -O3 for some targets.
-fno-ira-share-save-slotsDisable sharing of stack slots used for saving call-used hard registers living through a call. Each hard register gets a separate stack slot, and as a result function stack frames are larger.
-fno-ira-share-spill-slotsDisable sharing of stack slots allocated for pseudo-registers. Each pseudo-register that does not get a hard register gets a separate stack slot, and as a result function stack frames are larger.
-flra-rematEnable CFG-sensitive rematerialization in LRA. Instead of loading values of spilled pseudos, LRA tries to rematerialize (recalculate) values if it is profitable.
Enabled at levels -O2, -O3, -Os.
-fdelayed-branchIf supported for the target machine, attempt to reorder instructions to exploit instruction slots available after delayed branch instructions.
Enabled at levels -O, -O2, -O3, -Os.
-fschedule-insnsIf supported for the target machine, attempt to reorder instructions to eliminate execution stalls due to required data being unavailable. This helps machines that have slow floating point or memory load instructions by allowing other instructions to be issued until the result of the load or floating-point instruction is required.
Enabled at levels -O2, -O3.
-fschedule-insns2Similar to -fschedule-insns, but requests an additional pass of instruction scheduling after register allocation has been done. This is especially useful on machines with a relatively small number of registers and where memory load instructions take more than one cycle.
Enabled at levels -O2, -O3, -Os.
-fno-sched-interblockDon’t schedule instructions across basic blocks. This is normally enabled by default when scheduling before register allocation, i.e. with -fschedule-insns or at -O2 or higher.
-fno-sched-specDon’t allow speculative motion of non-load instructions. This is normally enabled by default when scheduling before register allocation, i.e. with -fschedule-insns or at -O2 or higher.
-fsched-pressureEnable register pressure sensitive insn scheduling before register allocation. This only makes sense when scheduling before register allocation is enabled, i.e. with -fschedule-insns or at -O2 or higher. Usage of this option can improve the generated code and decrease its size by preventing register pressure increase above the number of available hard registers and subsequent spills in register allocation.
-fsched-spec-loadAllow speculative motion of some load instructions. This only makes sense when scheduling before register allocation, i.e. with -fschedule-insns or at -O2 or higher.
-fsched-spec-load-dangerousAllow speculative motion of more load instructions. This only makes sense when scheduling before register allocation, i.e. with -fschedule-insns or at -O2 or higher.
-fsched-stalled-insns-fsched-stalled-insns=nDefine how many insns (if any) can be moved prematurely from the queue of stalled insns into the ready list during the second scheduling pass. -fno-sched-stalled-insns means that no insns are moved prematurely, -fsched-stalled-insns=0 means there is no limit on how many queued insns can be moved prematurely. -fsched-stalled-insns without a value is equivalent to -fsched-stalled-insns=1.
-fsched-stalled-insns-dep-fsched-stalled-insns-dep=nDefine how many insn groups (cycles) are examined for a dependency on a stalled insn that is a candidate for premature removal from the queue of stalled insns. This has an effect only during the second scheduling pass, and only if -fsched-stalled-insns is used. -fno-sched-stalled-insns-dep is equivalent to -fsched-stalled-insns-dep=0. -fsched-stalled-insns-dep without a value is equivalent to -fsched-stalled-insns-dep=1.
-fsched2-use-superblocksWhen scheduling after register allocation, use superblock scheduling. This allows motion across basic block boundaries, resulting in faster schedules. This option is experimental, as not all machine descriptions used by GCC model the CPU closely enough to avoid unreliable results from the algorithm.
This only makes sense when scheduling after register allocation, i.e. with -fschedule-insns2 or at -O2 or higher.
-fsched-group-heuristicEnable the group heuristic in the scheduler. This heuristic favors the instruction that belongs to a schedule group. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-fsched-critical-path-heuristicEnable the critical-path heuristic in the scheduler. This heuristic favors instructions on the critical path. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-fsched-spec-insn-heuristicEnable the speculative instruction heuristic in the scheduler. This heuristic favors speculative instructions with greater dependency weakness. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-fsched-rank-heuristicEnable the rank heuristic in the scheduler. This heuristic favors the instruction belonging to a basic block with greater size or frequency. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-fsched-last-insn-heuristicEnable the last-instruction heuristic in the scheduler. This heuristic favors the instruction that is less dependent on the last instruction scheduled. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-fsched-dep-count-heuristicEnable the dependent-count heuristic in the scheduler. This heuristic favors the instruction that has more instructions depending on it. This is enabled by default when scheduling is enabled, i.e. with -fschedule-insns or -fschedule-insns2 or at -O2 or higher.
-freschedule-modulo-scheduled-loopsModulo scheduling is performed before traditional scheduling. If a loop is modulo scheduled, later scheduling passes may change its schedule. Use this option to control that behavior.
-fselective-schedulingSchedule instructions using selective scheduling algorithm. Selective scheduling runs instead of the first scheduler pass.
-fselective-scheduling2Schedule instructions using selective scheduling algorithm. Selective scheduling runs instead of the second scheduler pass.
-fsel-sched-pipeliningEnable software pipelining of innermost loops during selective scheduling. This option has no effect unless one of -fselective-scheduling or -fselective-scheduling2 is turned on.
-fsel-sched-pipelining-outer-loopsWhen pipelining loops during selective scheduling, also pipeline outer loops. This option has no effect unless -fsel-sched-pipelining is turned on.
-fsemantic-interpositionSome object formats, like ELF, allow interposing of symbols by the dynamic linker. This means that for symbols exported from the DSO, the compiler cannot perform interprocedural propagation, inlining and other optimizations in anticipation that the function or variable in question may change. While this feature is useful, for example, to rewrite memory allocation functions by a debugging implementation, it is expensive in the terms of code quality. With -fno-semantic-interposition the compiler assumes that if interposition happens for functions the overwriting function will have precisely the same semantics (and side effects). Similarly if interposition happens for variables, the constructor of the variable will be the same. The flag has no effect for functions explicitly declared inline (where it is never allowed for interposition to change semantics) and for symbols explicitly declared weak.
-fshrink-wrapEmit function prologues only before parts of the function that need it, rather than at the top of the function. This flag is enabled by default at -O and higher.
-fcaller-savesEnable allocation of values to registers that are clobbered by function calls, by emitting extra instructions to save and restore the registers around such calls. Such allocation is done only when it seems to result in better code.
This option is always enabled by default on certain machines, usually those which have no call-preserved registers to use instead.
Enabled at levels -O2, -O3, -Os.
-fcombine-stack-adjustmentsTracks stack adjustments (pushes and pops) and stack memory references and then tries to find ways to combine them.
Enabled by default at -O1 and higher.
-fipa-raUse caller save registers for allocation if those registers are not used by any called function. In that case it is not necessary to save and restore them around calls. This is only possible if called functions are part of same compilation unit as current function and they are compiled before it.
Enabled at levels -O2, -O3, -Os.
-fconserve-stackAttempt to minimize stack usage. The compiler attempts to use less stack space, even if that makes the program slower. This option implies setting the large-stack-frame parameter to 100 and the large-stack-frame-growth parameter to 400.
-ftree-reassocPerform reassociation on trees. This flag is enabled by default at -O and higher.
-ftree-prePerform partial redundancy elimination (PRE) on trees. This flag is enabled by default at -O2 and -O3.
-ftree-partial-preMake partial redundancy elimination (PRE) more aggressive. This flag is enabled by default at -O3.
-ftree-forwpropPerform forward propagation on trees. This flag is enabled by default at -O and higher.
-ftree-frePerform full redundancy elimination (FRE) on trees. The difference between FRE and PRE is that FRE only considers expressions that are computed on all paths leading to the redundant computation. This analysis is faster than PRE, though it exposes fewer redundancies. This flag is enabled by default at -O and higher.
-ftree-phipropPerform hoisting of loads from conditional pointers on trees. This pass is enabled by default at -O and higher.
-fhoist-adjacent-loadsSpeculatively hoist loads from both branches of an if-then-else if the loads are from adjacent locations in the same structure and the target architecture has a conditional move instruction. This flag is enabled by default at -O2 and higher.
-ftree-copy-propPerform copy propagation on trees. This pass eliminates unnecessary copy operations. This flag is enabled by default at -O and higher.
-fipa-pure-constDiscover which functions are pure or constant. Enabled by default at -O and higher.
-fipa-referenceDiscover which static variables do not escape the compilation unit. Enabled by default at -O and higher.
-fipa-ptaPerform interprocedural pointer analysis and interprocedural modification and reference analysis. This option can cause excessive memory and compile-time usage on large compilation units. It is not enabled by default at any optimization level.
-fipa-profilePerform interprocedural profile propagation. The functions called only from
cold functions are marked as cold. Also functions executed once (such as
cold, noreturn, static constructors or destructors) are identified. Cold
functions and loop less parts of functions executed once are then optimized for
size.
Enabled by default at -O and higher.
-fipa-cpPerform interprocedural constant propagation. This optimization analyzes the program to determine when values passed to functions are constants and then optimizes accordingly. This optimization can substantially increase performance if the application has constants passed to functions. This flag is enabled by default at -O2, -Os and -O3.
-fipa-cp-clonePerform function cloning to make interprocedural constant propagation stronger. When enabled, interprocedural constant propagation performs function cloning when externally visible function can be called with constant arguments. Because this optimization can create multiple copies of functions, it may significantly increase code size (see --param ipcp-unit-growth=value). This flag is enabled by default at -O3.
-fipa-cp-alignmentWhen enabled, this optimization propagates alignment of function parameters to support better vectorization and string operations.
This flag is enabled by default at -O2 and -Os. It requires that -fipa-cp is enabled.
-fipa-icfPerform Identical Code Folding for functions and read-only variables. The optimization reduces code size and may disturb unwind stacks by replacing a function by equivalent one with a different name. The optimization works more effectively with link time optimization enabled.
Nevertheless the behavior is similar to Gold Linker ICF optimization, GCC ICF works on different levels and thus the optimizations are not same - there are equivalences that are found only by GCC and equivalences found only by Gold.
This flag is enabled by default at -O2 and -Os.
-fisolate-erroneous-paths-dereferenceDetect paths that trigger erroneous or undefined behavior due to dereferencing a null pointer. Isolate those paths from the main control flow and turn the statement with erroneous or undefined behavior into a trap. This flag is enabled by default at -O2 and higher and depends on -fdelete-null-pointer-checks also being enabled.
-fisolate-erroneous-paths-attributeDetect paths that trigger erroneous or undefined behavior due a null value
being used in a way forbidden by a returns_nonnull or nonnull
attribute. Isolate those paths from the main control flow and turn the
statement with erroneous or undefined behavior into a trap. This is not
currently enabled, but may be enabled by -O2 in the future.
-ftree-sinkPerform forward store motion on trees. This flag is enabled by default at -O and higher.
-ftree-bit-ccpPerform sparse conditional bit constant propagation on trees and propagate pointer alignment information. This pass only operates on local scalar variables and is enabled by default at -O and higher. It requires that -ftree-ccp is enabled.
-ftree-ccpPerform sparse conditional constant propagation (CCP) on trees. This pass only operates on local scalar variables and is enabled by default at -O and higher.
-fssa-backpropPropagate information about uses of a value up the definition chain in order to simplify the definitions. For example, this pass strips sign operations if the sign of a value never matters. The flag is enabled by default at -O and higher.
-fssa-phioptPerform pattern matching on SSA PHI nodes to optimize conditional code. This pass is enabled by default at -O and higher.
-ftree-switch-conversionPerform conversion of simple initializations in a switch to initializations from a scalar array. This flag is enabled by default at -O2 and higher.
-ftree-tail-mergeLook for identical code sequences. When found, replace one with a jump to the other. This optimization is known as tail merging or cross jumping. This flag is enabled by default at -O2 and higher. The compilation time in this pass can be limited using max-tail-merge-comparisons parameter and max-tail-merge-iterations parameter.
-ftree-dcePerform dead code elimination (DCE) on trees. This flag is enabled by default at -O and higher.
-ftree-builtin-call-dcePerform conditional dead code elimination (DCE) for calls to built-in functions
that may set errno but are otherwise side-effect free. This flag is
enabled by default at -O2 and higher if -Os is not also
specified.
-ftree-dominator-optsPerform a variety of simple scalar cleanups (constant/copy propagation, redundancy elimination, range propagation and expression simplification) based on a dominator tree traversal. This also performs jump threading (to reduce jumps to jumps). This flag is enabled by default at -O and higher.
-ftree-dsePerform dead store elimination (DSE) on trees. A dead store is a store into a memory location that is later overwritten by another store without any intervening loads. In this case the earlier store can be deleted. This flag is enabled by default at -O and higher.
-ftree-chPerform loop header copying on trees. This is beneficial since it increases effectiveness of code motion optimizations. It also saves one jump. This flag is enabled by default at -O and higher. It is not enabled for -Os, since it usually increases code size.
-ftree-loop-optimizePerform loop optimizations on trees. This flag is enabled by default at -O and higher.
-ftree-loop-linear-floop-interchange-floop-strip-mine-floop-block-floop-unroll-and-jamPerform loop nest optimizations. Same as -floop-nest-optimize. To use this code transformation, GCC has to be configured with --with-isl to enable the Graphite loop transformation infrastructure.
-fgraphite-identityEnable the identity transformation for graphite. For every SCoP we generate the polyhedral representation and transform it back to gimple. Using -fgraphite-identity we can check the costs or benefits of the GIMPLE -> GRAPHITE -> GIMPLE transformation. Some minimal optimizations are also performed by the code generator isl, like index splitting and dead code elimination in loops.
-floop-nest-optimizeEnable the isl based loop nest optimizer. This is a generic loop nest optimizer based on the Pluto optimization algorithms. It calculates a loop structure optimized for data-locality and parallelism. This option is experimental.
-floop-parallelize-allUse the Graphite data dependence analysis to identify loops that can be parallelized. Parallelize all the loops that can be analyzed to not contain loop carried dependences without checking that it is profitable to parallelize the loops.
-ftree-coalesce-varsWhile transforming the program out of the SSA representation, attempt to reduce copying by coalescing versions of different user-defined variables, instead of just compiler temporaries. This may severely limit the ability to debug an optimized program compiled with -fno-var-tracking-assignments. In the negated form, this flag prevents SSA coalescing of user variables. This option is enabled by default if optimization is enabled, and it does very little otherwise.
-ftree-loop-if-convertAttempt to transform conditional jumps in the innermost loops to branch-less equivalents. The intent is to remove control-flow from the innermost loops in order to improve the ability of the vectorization pass to handle these loops. This is enabled by default if vectorization is enabled.
-ftree-loop-if-convert-storesAttempt to also if-convert conditional jumps containing memory writes. This transformation can be unsafe for multi-threaded programs as it transforms conditional memory writes into unconditional memory writes. For example,
for (i = 0; i < N; i++)
if (cond)
A[i] = expr;
is transformed to
for (i = 0; i < N; i++) A[i] = cond ? expr : A[i];
potentially producing data races.
-ftree-loop-distributionPerform loop distribution. This flag can improve cache performance on big loop bodies and allow further loop optimizations, like parallelization or vectorization, to take place. For example, the loop
DO I = 1, N A(I) = B(I) + C D(I) = E(I) * F ENDDO
is transformed to
DO I = 1, N A(I) = B(I) + C ENDDO DO I = 1, N D(I) = E(I) * F ENDDO
-ftree-loop-distribute-patternsPerform loop distribution of patterns that can be code generated with calls to a library. This flag is enabled by default at -O3.
This pass distributes the initialization loops and generates a call to memset zero. For example, the loop
DO I = 1, N A(I) = 0 B(I) = A(I) + I ENDDO
is transformed to
DO I = 1, N A(I) = 0 ENDDO DO I = 1, N B(I) = A(I) + I ENDDO
and the initialization loop is transformed into a call to memset zero.
-ftree-loop-imPerform loop invariant motion on trees. This pass moves only invariants that are hard to handle at RTL level (function calls, operations that expand to nontrivial sequences of insns). With -funswitch-loops it also moves operands of conditions that are invariant out of the loop, so that we can use just trivial invariantness analysis in loop unswitching. The pass also includes store motion.
-ftree-loop-ivcanonCreate a canonical counter for number of iterations in loops for which determining number of iterations requires complicated analysis. Later optimizations then may determine the number easily. Useful especially in connection with unrolling.
-fivoptsPerform induction variable optimizations (strength reduction, induction variable merging and induction variable elimination) on trees.
-ftree-parallelize-loops=nParallelize loops, i.e., split their iteration space to run in n threads. This is only possible for loops whose iterations are independent and can be arbitrarily reordered. The optimization is only profitable on multiprocessor machines, for loops that are CPU-intensive, rather than constrained e.g. by memory bandwidth. This option implies -pthread, and thus is only supported on targets that have support for -pthread.
-ftree-ptaPerform function-local points-to analysis on trees. This flag is enabled by default at -O and higher.
-ftree-sraPerform scalar replacement of aggregates. This pass replaces structure references with scalars to prevent committing structures to memory too early. This flag is enabled by default at -O and higher.
-ftree-terPerform temporary expression replacement during the SSA->normal phase. Single use/single def temporaries are replaced at their use location with their defining expression. This results in non-GIMPLE code, but gives the expanders much more complex trees to work on resulting in better RTL generation. This is enabled by default at -O and higher.
-ftree-slsrPerform straight-line strength reduction on trees. This recognizes related expressions involving multiplications and replaces them by less expensive calculations when possible. This is enabled by default at -O and higher.
-ftree-vectorizePerform vectorization on trees. This flag enables -ftree-loop-vectorize and -ftree-slp-vectorize if not explicitly specified.
-ftree-loop-vectorizePerform loop vectorization on trees. This flag is enabled by default at -O3 and when -ftree-vectorize is enabled.
-ftree-slp-vectorizePerform basic block vectorization on trees. This flag is enabled by default at -O3 and when -ftree-vectorize is enabled.
-fvect-cost-model=modelAlter the cost model used for vectorization. The model argument should be one of ‘unlimited’, ‘dynamic’ or ‘cheap’. With the ‘unlimited’ model the vectorized code-path is assumed to be profitable while with the ‘dynamic’ model a runtime check guards the vectorized code-path to enable it only for iteration counts that will likely execute faster than when executing the original scalar loop. The ‘cheap’ model disables vectorization of loops where doing so would be cost prohibitive for example due to required runtime checks for data dependence or alignment but otherwise is equal to the ‘dynamic’ model. The default cost model depends on other optimization flags and is either ‘dynamic’ or ‘cheap’.
-fsimd-cost-model=modelAlter the cost model used for vectorization of loops marked with the OpenMP or Cilk Plus simd directive. The model argument should be one of ‘unlimited’, ‘dynamic’, ‘cheap’. All values of model have the same meaning as described in -fvect-cost-model and by default a cost model defined with -fvect-cost-model is used.
-ftree-vrpPerform Value Range Propagation on trees. This is similar to the constant propagation pass, but instead of values, ranges of values are propagated. This allows the optimizers to remove unnecessary range checks like array bound checks and null pointer checks. This is enabled by default at -O2 and higher. Null pointer check elimination is only done if -fdelete-null-pointer-checks is enabled.
-fsplit-pathsSplit paths leading to loop backedges. This can improve dead code elimination and common subexpression elimination. This is enabled by default at -O2 and above.
-fsplit-ivs-in-unrollerEnables expression of values of induction variables in later iterations of the unrolled loop using the value in the first iteration. This breaks long dependency chains, thus improving efficiency of the scheduling passes.
A combination of -fweb and CSE is often sufficient to obtain the same effect. However, that is not reliable in cases where the loop body is more complicated than a single basic block. It also does not work at all on some architectures due to restrictions in the CSE pass.
This optimization is enabled by default.
-fvariable-expansion-in-unrollerWith this option, the compiler creates multiple copies of some local variables when unrolling a loop, which can result in superior code.
-fpartial-inliningInline parts of functions. This option has any effect only when inlining itself is turned on by the -finline-functions or -finline-small-functions options.
Enabled at level -O2.
-fpredictive-commoningPerform predictive commoning optimization, i.e., reusing computations (especially memory loads and stores) performed in previous iterations of loops.
This option is enabled at level -O3.
-fprefetch-loop-arraysIf supported by the target machine, generate instructions to prefetch memory to improve the performance of loops that access large arrays.
This option may generate better or worse code; results are highly dependent on the structure of loops within the source code.
Disabled at level -Os.
-fno-peephole-fno-peephole2Disable any machine-specific peephole optimizations. The difference between -fno-peephole and -fno-peephole2 is in how they are implemented in the compiler; some targets use one, some use the other, a few use both.
-fpeephole is enabled by default. -fpeephole2 enabled at levels -O2, -O3, -Os.
-fno-guess-branch-probabilityDo not guess branch probabilities using heuristics.
GCC uses heuristics to guess branch probabilities if they are
not provided by profiling feedback (-fprofile-arcs). These
heuristics are based on the control flow graph. If some branch probabilities
are specified by __builtin_expect, then the heuristics are
used to guess branch probabilities for the rest of the control flow graph,
taking the __builtin_expect info into account. The interactions
between the heuristics and __builtin_expect can be complex, and in
some cases, it may be useful to disable the heuristics so that the effects
of __builtin_expect are easier to understand.
The default is -fguess-branch-probability at levels -O, -O2, -O3, -Os.
-freorder-blocksReorder basic blocks in the compiled function in order to reduce number of taken branches and improve code locality.
Enabled at levels -O, -O2, -O3, -Os.
-freorder-blocks-algorithm=algorithmUse the specified algorithm for basic block reordering. The algorithm argument can be ‘simple’, which does not increase code size (except sometimes due to secondary effects like alignment), or ‘stc’, the “software trace cache” algorithm, which tries to put all often executed code together, minimizing the number of branches executed by making extra copies of code.
The default is ‘simple’ at levels -O, -Os, and ‘stc’ at levels -O2, -O3.
-freorder-blocks-and-partitionIn addition to reordering basic blocks in the compiled function, in order to reduce number of taken branches, partitions hot and cold basic blocks into separate sections of the assembly and .o files, to improve paging and cache locality performance.
This optimization is automatically turned off in the presence of exception handling, for linkonce sections, for functions with a user-defined section attribute and on any architecture that does not support named sections.
Enabled for x86 at levels -O2, -O3.
-freorder-functionsReorder functions in the object file in order to
improve code locality. This is implemented by using special
subsections .text.hot for most frequently executed functions and
.text.unlikely for unlikely executed functions. Reordering is done by
the linker so object file format must support named sections and linker must
place them in a reasonable way.
Also profile feedback must be available to make this option effective. See -fprofile-arcs for details.
Enabled at levels -O2, -O3, -Os.
-fstrict-aliasingAllow the compiler to assume the strictest aliasing rules applicable to
the language being compiled. For C (and C++), this activates
optimizations based on the type of expressions. In particular, an
object of one type is assumed never to reside at the same address as an
object of a different type, unless the types are almost the same. For
example, an unsigned int can alias an int, but not a
void* or a double. A character type may alias any other
type.
Pay special attention to code like this:
union a_union {
int i;
double d;
};
int f() {
union a_union t;
t.d = 3.0;
return t.i;
}
The practice of reading from a different union member than the one most recently written to (called “type-punning”) is common. Even with -fstrict-aliasing, type-punning is allowed, provided the memory is accessed through the union type. So, the code above works as expected. See Structures unions enumerations and bit-fields implementation. However, this code might not:
int f() {
union a_union t;
int* ip;
t.d = 3.0;
ip = &t.i;
return *ip;
}
Similarly, access by taking the address, casting the resulting pointer and dereferencing the result has undefined behavior, even if the cast uses a union type, e.g.:
int f() {
double d = 3.0;
return ((union a_union *) &d)->i;
}
The -fstrict-aliasing option is enabled at levels -O2, -O3, -Os.
-fstrict-overflowAllow the compiler to assume strict signed overflow rules, depending
on the language being compiled. For C (and C++) this means that
overflow when doing arithmetic with signed numbers is undefined, which
means that the compiler may assume that it does not happen. This
permits various optimizations. For example, the compiler assumes
that an expression like i + 10 > i is always true for
signed i. This assumption is only valid if signed overflow is
undefined, as the expression is false if i + 10 overflows when
using twos complement arithmetic. When this option is in effect any
attempt to determine whether an operation on signed numbers
overflows must be written carefully to not actually involve overflow.
This option also allows the compiler to assume strict pointer
semantics: given a pointer to an object, if adding an offset to that
pointer does not produce a pointer to the same object, the addition is
undefined. This permits the compiler to conclude that p + u >
p is always true for a pointer p and unsigned integer
u. This assumption is only valid because pointer wraparound is
undefined, as the expression is false if p + u overflows using
twos complement arithmetic.
See also the -fwrapv option. Using -fwrapv means that integer signed overflow is fully defined: it wraps. When -fwrapv is used, there is no difference between -fstrict-overflow and -fno-strict-overflow for integers. With -fwrapv certain types of overflow are permitted. For example, if the compiler gets an overflow when doing arithmetic on constants, the overflowed value can still be used with -fwrapv, but not otherwise.
The -fstrict-overflow option is enabled at levels -O2, -O3, -Os.
-falign-functions-falign-functions=nAlign the start of functions to the next power-of-two greater than n, skipping up to n bytes. For instance, -falign-functions=32 aligns functions to the next 32-byte boundary, but -falign-functions=24 aligns to the next 32-byte boundary only if this can be done by skipping 23 bytes or less.
-fno-align-functions and -falign-functions=1 are equivalent and mean that functions are not aligned.
Some assemblers only support this flag when n is a power of two; in that case, it is rounded up.
If n is not specified or is zero, use a machine-dependent default.
Enabled at levels -O2, -O3.
-falign-labels-falign-labels=nAlign all branch targets to a power-of-two boundary, skipping up to n bytes like -falign-functions. This option can easily make code slower, because it must insert dummy operations for when the branch target is reached in the usual flow of the code.
-fno-align-labels and -falign-labels=1 are equivalent and mean that labels are not aligned.
If -falign-loops or -falign-jumps are applicable and are greater than this value, then their values are used instead.
If n is not specified or is zero, use a machine-dependent default which is very likely to be ‘1’, meaning no alignment.
Enabled at levels -O2, -O3.
-falign-loops-falign-loops=nAlign loops to a power-of-two boundary, skipping up to n bytes like -falign-functions. If the loops are executed many times, this makes up for any execution of the dummy operations.
-fno-align-loops and -falign-loops=1 are equivalent and mean that loops are not aligned.
If n is not specified or is zero, use a machine-dependent default.
Enabled at levels -O2, -O3.
-falign-jumps-falign-jumps=nAlign branch targets to a power-of-two boundary, for branch targets where the targets can only be reached by jumping, skipping up to n bytes like -falign-functions. In this case, no dummy operations need be executed.
-fno-align-jumps and -falign-jumps=1 are equivalent and mean that loops are not aligned.
If n is not specified or is zero, use a machine-dependent default.
Enabled at levels -O2, -O3.
-funit-at-a-timeThis option is left for compatibility reasons. -funit-at-a-time has no effect, while -fno-unit-at-a-time implies -fno-toplevel-reorder and -fno-section-anchors.
Enabled by default.
-fno-toplevel-reorderDo not reorder top-level functions, variables, and asm
statements. Output them in the same order that they appear in the
input file. When this option is used, unreferenced static variables
are not removed. This option is intended to support existing code
that relies on a particular ordering. For new code, it is better to
use attributes when possible.
Enabled at level -O0. When disabled explicitly, it also implies -fno-section-anchors, which is otherwise enabled at -O0 on some targets.
-fwebConstructs webs as commonly used for register allocation purposes and assign each web individual pseudo register. This allows the register allocation pass to operate on pseudos directly, but also strengthens several other optimization passes, such as CSE, loop optimizer and trivial dead code remover. It can, however, make debugging impossible, since variables no longer stay in a “home register”.
Enabled by default with -funroll-loops.
-fwhole-programAssume that the current compilation unit represents the whole program being
compiled. All public functions and variables with the exception of main
and those merged by attribute externally_visible become static functions
and in effect are optimized more aggressively by interprocedural optimizers.
This option should not be used in combination with -flto. Instead relying on a linker plugin should provide safer and more precise information.
-flto[=n]This option runs the standard link-time optimizer. When invoked with source code, it generates GIMPLE (one of GCC’s internal representations) and writes it to special ELF sections in the object file. When the object files are linked together, all the function bodies are read from these ELF sections and instantiated as if they had been part of the same translation unit.
To use the link-time optimizer, -flto and optimization options should be specified at compile time and during the final link. It is recommended that you compile all the files participating in the same link with the same options and also specify those options at link time. For example:
gcc -c -O2 -flto foo.c gcc -c -O2 -flto bar.c gcc -o myprog -flto -O2 foo.o bar.o
The first two invocations to GCC save a bytecode representation of GIMPLE into special ELF sections inside foo.o and bar.o. The final invocation reads the GIMPLE bytecode from foo.o and bar.o, merges the two files into a single internal image, and compiles the result as usual. Since both foo.o and bar.o are merged into a single image, this causes all the interprocedural analyses and optimizations in GCC to work across the two files as if they were a single one. This means, for example, that the inliner is able to inline functions in bar.o into functions in foo.o and vice-versa.
Another (simpler) way to enable link-time optimization is:
gcc -o myprog -flto -O2 foo.c bar.c
The above generates bytecode for foo.c and bar.c, merges them together into a single GIMPLE representation and optimizes them as usual to produce myprog.
The only important thing to keep in mind is that to enable link-time optimizations you need to use the GCC driver to perform the link step. GCC then automatically performs link-time optimization if any of the objects involved were compiled with the -flto command-line option. You generally should specify the optimization options to be used for link-time optimization though GCC tries to be clever at guessing an optimization level to use from the options used at compile time if you fail to specify one at link time. You can always override the automatic decision to do link-time optimization at link time by passing -fno-lto to the link command.
To make whole program optimization effective, it is necessary to make certain whole program assumptions. The compiler needs to know what functions and variables can be accessed by libraries and runtime outside of the link-time optimized unit. When supported by the linker, the linker plugin (see -fuse-linker-plugin) passes information to the compiler about used and externally visible symbols. When the linker plugin is not available, -fwhole-program should be used to allow the compiler to make these assumptions, which leads to more aggressive optimization decisions.
When -fuse-linker-plugin is not enabled, when a file is compiled with -flto, the generated object file is larger than a regular object file because it contains GIMPLE bytecodes and the usual final code (see -ffat-lto-objects. This means that object files with LTO information can be linked as normal object files; if -fno-lto is passed to the linker, no interprocedural optimizations are applied. Note that when -fno-fat-lto-objects is enabled the compile stage is faster but you cannot perform a regular, non-LTO link on them.
Additionally, the optimization flags used to compile individual files are not necessarily related to those used at link time. For instance,
gcc -c -O0 -ffat-lto-objects -flto foo.c gcc -c -O0 -ffat-lto-objects -flto bar.c gcc -o myprog -O3 foo.o bar.o
This produces individual object files with unoptimized assembler code, but the resulting binary myprog is optimized at -O3. If, instead, the final binary is generated with -fno-lto, then myprog is not optimized.
When producing the final binary, GCC only applies link-time optimizations to those files that contain bytecode. Therefore, you can mix and match object files and libraries with GIMPLE bytecodes and final object code. GCC automatically selects which files to optimize in LTO mode and which files to link without further processing.
There are some code generation flags preserved by GCC when generating bytecodes, as they need to be used during the final link stage. Generally options specified at link time override those specified at compile time.
If you do not specify an optimization level option -O at link time, then GCC uses the highest optimization level used when compiling the object files.
Currently, the following options and their settings are taken from the first object file that explicitly specifies them: -fPIC, -fpic, -fpie, -fcommon, -fexceptions, -fnon-call-exceptions, -fgnu-tm and all the -m target flags.
Certain ABI-changing flags are required to match in all compilation units, and trying to override this at link time with a conflicting value is ignored. This includes options such as -freg-struct-return and -fpcc-struct-return.
Other options such as -ffp-contract, -fno-strict-overflow, -fwrapv, -fno-trapv or -fno-strict-aliasing are passed through to the link stage and merged conservatively for conflicting translation units. Specifically -fno-strict-overflow, -fwrapv and -fno-trapv take precedence; and for example -ffp-contract=off takes precedence over -ffp-contract=fast. You can override them at link time.
If LTO encounters objects with C linkage declared with incompatible types in separate translation units to be linked together (undefined behavior according to ISO C99 6.2.7), a non-fatal diagnostic may be issued. The behavior is still undefined at run time. Similar diagnostics may be raised for other languages.
Another feature of LTO is that it is possible to apply interprocedural optimizations on files written in different languages:
gcc -c -flto foo.c g++ -c -flto bar.cc gfortran -c -flto baz.f90 g++ -o myprog -flto -O3 foo.o bar.o baz.o -lgfortran
Notice that the final link is done with g++ to get the C++
runtime libraries and -lgfortran is added to get the Fortran
runtime libraries. In general, when mixing languages in LTO mode, you
should use the same link command options as when mixing languages in a
regular (non-LTO) compilation.
If object files containing GIMPLE bytecode are stored in a library archive, say
libfoo.a, it is possible to extract and use them in an LTO link if you
are using a linker with plugin support. To create static libraries suitable
for LTO, use gcc-ar and gcc-ranlib instead of ar
and ranlib;
to show the symbols of object files with GIMPLE bytecode, use
gcc-nm. Those commands require that ar, ranlib
and nm have been compiled with plugin support. At link time, use the the
flag -fuse-linker-plugin to ensure that the library participates in
the LTO optimization process:
gcc -o myprog -O2 -flto -fuse-linker-plugin a.o b.o -lfoo
With the linker plugin enabled, the linker extracts the needed GIMPLE files from libfoo.a and passes them on to the running GCC to make them part of the aggregated GIMPLE image to be optimized.
If you are not using a linker with plugin support and/or do not enable the linker plugin, then the objects inside libfoo.a are extracted and linked as usual, but they do not participate in the LTO optimization process. In order to make a static library suitable for both LTO optimization and usual linkage, compile its object files with -flto -ffat-lto-objects.
Link-time optimizations do not require the presence of the whole program to operate. If the program does not require any symbols to be exported, it is possible to combine -flto and -fwhole-program to allow the interprocedural optimizers to use more aggressive assumptions which may lead to improved optimization opportunities. Use of -fwhole-program is not needed when linker plugin is active (see -fuse-linker-plugin).
The current implementation of LTO makes no attempt to generate bytecode that is portable between different types of hosts. The bytecode files are versioned and there is a strict version check, so bytecode files generated in one version of GCC do not work with an older or newer version of GCC.
Link-time optimization does not work well with generation of debugging information. Combining -flto with -g is currently experimental and expected to produce unexpected results.
If you specify the optional n, the optimization and code
generation done at link time is executed in parallel using n
parallel jobs by utilizing an installed make program. The
environment variable MAKE may be used to override the program
used. The default value for n is 1.
You can also specify -flto=jobserver to use GNU make’s
job server mode to determine the number of parallel jobs. This
is useful when the Makefile calling GCC is already executing in parallel.
You must prepend a ‘+’ to the command recipe in the parent Makefile
for this to work. This option likely only works if MAKE is
GNU make.
-flto-partition=algSpecify the partitioning algorithm used by the link-time optimizer. The value is either ‘1to1’ to specify a partitioning mirroring the original source files or ‘balanced’ to specify partitioning into equally sized chunks (whenever possible) or ‘max’ to create new partition for every symbol where possible. Specifying ‘none’ as an algorithm disables partitioning and streaming completely. The default value is ‘balanced’. While ‘1to1’ can be used as an workaround for various code ordering issues, the ‘max’ partitioning is intended for internal testing only. The value ‘one’ specifies that exactly one partition should be used while the value ‘none’ bypasses partitioning and executes the link-time optimization step directly from the WPA phase.
-flto-odr-type-mergingEnable streaming of mangled types names of C++ types and their unification at link time. This increases size of LTO object files, but enables diagnostics about One Definition Rule violations.
-flto-compression-level=nThis option specifies the level of compression used for intermediate language written to LTO object files, and is only meaningful in conjunction with LTO mode (-flto). Valid values are 0 (no compression) to 9 (maximum compression). Values outside this range are clamped to either 0 or 9. If the option is not given, a default balanced compression setting is used.
-fuse-linker-pluginEnables the use of a linker plugin during link-time optimization. This option relies on plugin support in the linker, which is available in gold or in GNU ld 2.21 or newer.
This option enables the extraction of object files with GIMPLE bytecode out of library archives. This improves the quality of optimization by exposing more code to the link-time optimizer. This information specifies what symbols can be accessed externally (by non-LTO object or during dynamic linking). Resulting code quality improvements on binaries (and shared libraries that use hidden visibility) are similar to -fwhole-program. See -flto for a description of the effect of this flag and how to use it.
This option is enabled by default when LTO support in GCC is enabled and GCC was configured for use with a linker supporting plugins (GNU ld 2.21 or newer or gold).
-ffat-lto-objectsFat LTO objects are object files that contain both the intermediate language and the object code. This makes them usable for both LTO linking and normal linking. This option is effective only when compiling with -flto and is ignored at link time.
-fno-fat-lto-objects improves compilation time over plain LTO, but
requires the complete toolchain to be aware of LTO. It requires a linker with
linker plugin support for basic functionality. Additionally,
nm, ar and ranlib
need to support linker plugins to allow a full-featured build environment
(capable of building static libraries etc). GCC provides the gcc-ar,
gcc-nm, gcc-ranlib wrappers to pass the right options
to these tools. With non fat LTO makefiles need to be modified to use them.
The default is -fno-fat-lto-objects on targets with linker plugin support.
-fcompare-elimAfter register allocation and post-register allocation instruction splitting, identify arithmetic instructions that compute processor flags similar to a comparison operation based on that arithmetic. If possible, eliminate the explicit comparison operation.
This pass only applies to certain targets that cannot explicitly represent the comparison operation before register allocation is complete.
Enabled at levels -O, -O2, -O3, -Os.
-fcprop-registersAfter register allocation and post-register allocation instruction splitting, perform a copy-propagation pass to try to reduce scheduling dependencies and occasionally eliminate the copy.
Enabled at levels -O, -O2, -O3, -Os.
-fprofile-correctionProfiles collected using an instrumented binary for multi-threaded programs may be inconsistent due to missed counter updates. When this option is specified, GCC uses heuristics to correct or smooth out such inconsistencies. By default, GCC emits an error message when an inconsistent profile is detected.
-fprofile-use-fprofile-use=pathEnable profile feedback-directed optimizations, and the following optimizations which are generally profitable only with profile feedback available: -fbranch-probabilities, -fvpt, -funroll-loops, -fpeel-loops, -ftracer, -ftree-vectorize, and ftree-loop-distribute-patterns.
Before you can use this option, you must first generate profiling information. See Optimize Options, for information about the -fprofile-generate option.
By default, GCC emits an error message if the feedback profiles do not match the source code. This error can be turned into a warning by using -Wcoverage-mismatch. Note this may result in poorly optimized code.
If path is specified, GCC looks at the path to find the profile feedback data files. See -fprofile-dir.
-fauto-profile-fauto-profile=pathEnable sampling-based feedback-directed optimizations, and the following optimizations which are generally profitable only with profile feedback available: -fbranch-probabilities, -fvpt, -funroll-loops, -fpeel-loops, -ftracer, -ftree-vectorize, -finline-functions, -fipa-cp, -fipa-cp-clone, -fpredictive-commoning, -funswitch-loops, -fgcse-after-reload, and -ftree-loop-distribute-patterns.
path is the name of a file containing AutoFDO profile information. If omitted, it defaults to fbdata.afdo in the current directory.
Producing an AutoFDO profile data file requires running your program
with the perf utility on a supported GNU/Linux target system.
For more information, see https://perf.wiki.kernel.org/.
E.g.
perf record -e br_inst_retired:near_taken -b -o perf.data \
-- your_program
Then use the create_gcov tool to convert the raw profile data
to a format that can be used by GCC. You must also supply the
unstripped binary for your program to this tool.
See https://github.com/google/autofdo.
E.g.
create_gcov --binary=your_program.unstripped --profile=perf.data \
--gcov=profile.afdo
The following options control compiler behavior regarding floating-point arithmetic. These options trade off between speed and correctness. All must be specifically enabled.
-ffloat-storeDo not store floating-point variables in registers, and inhibit other options that might change whether a floating-point value is taken from a register or memory.
This option prevents undesirable excess precision on machines such as
the 68000 where the floating registers (of the 68881) keep more
precision than a double is supposed to have. Similarly for the
x86 architecture. For most programs, the excess precision does only
good, but a few programs rely on the precise definition of IEEE floating
point. Use -ffloat-store for such programs, after modifying
them to store all pertinent intermediate computations into variables.
-fexcess-precision=styleThis option allows further control over excess precision on machines
where floating-point registers have more precision than the IEEE
float and double types and the processor does not
support operations rounding to those types. By default,
-fexcess-precision=fast is in effect; this means that
operations are carried out in the precision of the registers and that
it is unpredictable when rounding to the types specified in the source
code takes place. When compiling C, if
-fexcess-precision=standard is specified then excess
precision follows the rules specified in ISO C99; in particular,
both casts and assignments cause values to be rounded to their
semantic types (whereas -ffloat-store only affects
assignments). This option is enabled by default for C if a strict
conformance option such as -std=c99 is used.
-fexcess-precision=standard is not implemented for languages other than C, and has no effect if -funsafe-math-optimizations or -ffast-math is specified. On the x86, it also has no effect if -mfpmath=sse or -mfpmath=sse+387 is specified; in the former case, IEEE semantics apply without excess precision, and in the latter, rounding is unpredictable.
-ffast-mathSets the options -fno-math-errno, -funsafe-math-optimizations, -ffinite-math-only, -fno-rounding-math, -fno-signaling-nans and -fcx-limited-range.
This option causes the preprocessor macro __FAST_MATH__ to be defined.
This option is not turned on by any -O option besides -Ofast since it can result in incorrect output for programs that depend on an exact implementation of IEEE or ISO rules/specifications for math functions. It may, however, yield faster code for programs that do not require the guarantees of these specifications.
-fno-math-errnoDo not set errno after calling math functions that are executed
with a single instruction, e.g., sqrt. A program that relies on
IEEE exceptions for math error handling may want to use this flag
for speed while maintaining IEEE arithmetic compatibility.
This option is not turned on by any -O option since it can result in incorrect output for programs that depend on an exact implementation of IEEE or ISO rules/specifications for math functions. It may, however, yield faster code for programs that do not require the guarantees of these specifications.
The default is -fmath-errno.
On Darwin systems, the math library never sets errno. There is
therefore no reason for the compiler to consider the possibility that
it might, and -fno-math-errno is the default.
-funsafe-math-optimizationsAllow optimizations for floating-point arithmetic that (a) assume that arguments and results are valid and (b) may violate IEEE or ANSI standards. When used at link time, it may include libraries or startup files that change the default FPU control word or other similar optimizations.
This option is not turned on by any -O option since it can result in incorrect output for programs that depend on an exact implementation of IEEE or ISO rules/specifications for math functions. It may, however, yield faster code for programs that do not require the guarantees of these specifications. Enables -fno-signed-zeros, -fno-trapping-math, -fassociative-math and -freciprocal-math.
The default is -fno-unsafe-math-optimizations.
-fassociative-mathAllow re-association of operands in series of floating-point operations.
This violates the ISO C and C++ language standard by possibly changing
computation result. NOTE: re-ordering may change the sign of zero as
well as ignore NaNs and inhibit or create underflow or overflow (and
thus cannot be used on code that relies on rounding behavior like
(x + 2**52) - 2**52. May also reorder floating-point comparisons
and thus may not be used when ordered comparisons are required.
This option requires that both -fno-signed-zeros and
-fno-trapping-math be in effect. Moreover, it doesn’t make
much sense with -frounding-math. For Fortran the option
is automatically enabled when both -fno-signed-zeros and
-fno-trapping-math are in effect.
The default is -fno-associative-math.
-freciprocal-mathAllow the reciprocal of a value to be used instead of dividing by
the value if this enables optimizations. For example x / y
can be replaced with x * (1/y), which is useful if (1/y)
is subject to common subexpression elimination. Note that this loses
precision and increases the number of flops operating on the value.
The default is -fno-reciprocal-math.
-ffinite-math-onlyAllow optimizations for floating-point arithmetic that assume that arguments and results are not NaNs or +-Infs.
This option is not turned on by any -O option since it can result in incorrect output for programs that depend on an exact implementation of IEEE or ISO rules/specifications for math functions. It may, however, yield faster code for programs that do not require the guarantees of these specifications.
The default is -fno-finite-math-only.
-fno-signed-zerosAllow optimizations for floating-point arithmetic that ignore the signedness of zero. IEEE arithmetic specifies the behavior of distinct +0.0 and -0.0 values, which then prohibits simplification of expressions such as x+0.0 or 0.0*x (even with -ffinite-math-only). This option implies that the sign of a zero result isn’t significant.
The default is -fsigned-zeros.
-fno-trapping-mathCompile code assuming that floating-point operations cannot generate user-visible traps. These traps include division by zero, overflow, underflow, inexact result and invalid operation. This option requires that -fno-signaling-nans be in effect. Setting this option may allow faster code if one relies on “non-stop” IEEE arithmetic, for example.
This option should never be turned on by any -O option since it can result in incorrect output for programs that depend on an exact implementation of IEEE or ISO rules/specifications for math functions.
The default is -ftrapping-math.
-frounding-mathDisable transformations and optimizations that assume default floating-point rounding behavior. This is round-to-zero for all floating point to integer conversions, and round-to-nearest for all other arithmetic truncations. This option should be specified for programs that change the FP rounding mode dynamically, or that may be executed with a non-default rounding mode. This option disables constant folding of floating-point expressions at compile time (which may be affected by rounding mode) and arithmetic transformations that are unsafe in the presence of sign-dependent rounding modes.
The default is -fno-rounding-math.
This option is experimental and does not currently guarantee to
disable all GCC optimizations that are affected by rounding mode.
Future versions of GCC may provide finer control of this setting
using C99’s FENV_ACCESS pragma. This command-line option
will be used to specify the default state for FENV_ACCESS.
-fsignaling-nansCompile code assuming that IEEE signaling NaNs may generate user-visible traps during floating-point operations. Setting this option disables optimizations that may change the number of exceptions visible with signaling NaNs. This option implies -ftrapping-math.
This option causes the preprocessor macro __SUPPORT_SNAN__ to
be defined.
The default is -fno-signaling-nans.
This option is experimental and does not currently guarantee to disable all GCC optimizations that affect signaling NaN behavior.
-fsingle-precision-constantTreat floating-point constants as single precision instead of implicitly converting them to double-precision constants.
-fcx-limited-rangeWhen enabled, this option states that a range reduction step is not
needed when performing complex division. Also, there is no checking
whether the result of a complex multiplication or division is NaN
+ I*NaN, with an attempt to rescue the situation in that case. The
default is -fno-cx-limited-range, but is enabled by
-ffast-math.
This option controls the default setting of the ISO C99
CX_LIMITED_RANGE pragma. Nevertheless, the option applies to
all languages.
-fcx-fortran-rulesComplex multiplication and division follow Fortran rules. Range
reduction is done as part of complex division, but there is no checking
whether the result of a complex multiplication or division is NaN
+ I*NaN, with an attempt to rescue the situation in that case.
The default is -fno-cx-fortran-rules.
The following options control optimizations that may improve performance, but are not enabled by any -O options. This section includes experimental options that may produce broken code.
-fbranch-probabilitiesAfter running a program compiled with -fprofile-arcs (see Instrumentation Options), you can compile it a second time using -fbranch-probabilities, to improve optimizations based on the number of times each branch was taken. When a program compiled with -fprofile-arcs exits, it saves arc execution counts to a file called sourcename.gcda for each source file. The information in this data file is very dependent on the structure of the generated code, so you must use the same source code and the same optimization options for both compilations.
With -fbranch-probabilities, GCC puts a ‘REG_BR_PROB’ note on each ‘JUMP_INSN’ and ‘CALL_INSN’. These can be used to improve optimization. Currently, they are only used in one place: in reorg.c, instead of guessing which path a branch is most likely to take, the ‘REG_BR_PROB’ values are used to exactly determine which path is taken more often.
-fprofile-valuesIf combined with -fprofile-arcs, it adds code so that some data about values of expressions in the program is gathered.
With -fbranch-probabilities, it reads back the data gathered from profiling values of expressions for usage in optimizations.
Enabled with -fprofile-generate and -fprofile-use.
-fprofile-reorder-functionsFunction reordering based on profile instrumentation collects first time of execution of a function and orders these functions in ascending order.
Enabled with -fprofile-use.
-fvptIf combined with -fprofile-arcs, this option instructs the compiler to add code to gather information about values of expressions.
With -fbranch-probabilities, it reads back the data gathered and actually performs the optimizations based on them. Currently the optimizations include specialization of division operations using the knowledge about the value of the denominator.
-frename-registersAttempt to avoid false dependencies in scheduled code by making use of registers left over after register allocation. This optimization most benefits processors with lots of registers. Depending on the debug information format adopted by the target, however, it can make debugging impossible, since variables no longer stay in a “home register”.
Enabled by default with -funroll-loops and -fpeel-loops.
-fschedule-fusionPerforms a target dependent pass over the instruction stream to schedule instructions of same type together because target machine can execute them more efficiently if they are adjacent to each other in the instruction flow.
Enabled at levels -O2, -O3, -Os.
-ftracerPerform tail duplication to enlarge superblock size. This transformation simplifies the control flow of the function allowing other optimizations to do a better job.
Enabled with -fprofile-use.
-funroll-loopsUnroll loops whose number of iterations can be determined at compile time or upon entry to the loop. -funroll-loops implies -frerun-cse-after-loop, -fweb and -frename-registers. It also turns on complete loop peeling (i.e. complete removal of loops with a small constant number of iterations). This option makes code larger, and may or may not make it run faster.
Enabled with -fprofile-use.
-funroll-all-loopsUnroll all loops, even if their number of iterations is uncertain when the loop is entered. This usually makes programs run more slowly. -funroll-all-loops implies the same options as -funroll-loops.
-fpeel-loopsPeels loops for which there is enough information that they do not roll much (from profile feedback). It also turns on complete loop peeling (i.e. complete removal of loops with small constant number of iterations).
Enabled with -fprofile-use.
-fmove-loop-invariantsEnables the loop invariant motion pass in the RTL loop optimizer. Enabled at level -O1
-funswitch-loopsMove branches with loop invariant conditions out of the loop, with duplicates of the loop on both branches (modified according to result of the condition).
-ffunction-sections-fdata-sectionsPlace each function or data item into its own section in the output file if the target supports arbitrary sections. The name of the function or the name of the data item determines the section’s name in the output file.
Use these options on systems where the linker can perform optimizations to improve locality of reference in the instruction space. Most systems using the ELF object format and SPARC processors running Solaris 2 have linkers with such optimizations. AIX may have these optimizations in the future.
Only use these options when there are significant benefits from doing
so. When you specify these options, the assembler and linker
create larger object and executable files and are also slower.
You cannot use gprof on all systems if you
specify this option, and you may have problems with debugging if
you specify both this option and -g.
-fbranch-target-load-optimizePerform branch target register load optimization before prologue / epilogue threading. The use of target registers can typically be exposed only during reload, thus hoisting loads out of loops and doing inter-block scheduling needs a separate optimization pass.
-fbranch-target-load-optimize2Perform branch target register load optimization after prologue / epilogue threading.
-fbtr-bb-exclusiveWhen performing branch target register load optimization, don’t reuse branch target registers within any basic block.
-fstdarg-optOptimize the prologue of variadic argument functions with respect to usage of those arguments.
-fsection-anchorsTry to reduce the number of symbolic address calculations by using shared “anchor” symbols to address nearby objects. This transformation can help to reduce the number of GOT entries and GOT accesses on some targets.
For example, the implementation of the following function foo:
static int a, b, c;
int foo (void) { return a + b + c; }
usually calculates the addresses of all three variables, but if you compile it with -fsection-anchors, it accesses the variables from a common anchor point instead. The effect is similar to the following pseudocode (which isn’t valid C):
int foo (void)
{
register int *xr = &x;
return xr[&a - &x] + xr[&b - &x] + xr[&c - &x];
}
Not all targets support this option.
--param name=valueIn some places, GCC uses various constants to control the amount of optimization that is done. For example, GCC does not inline functions that contain more than a certain number of instructions. You can control some of these constants on the command line using the --param option.
The names of specific parameters, and the meaning of the values, are tied to the internals of the compiler, and are subject to change without notice in future releases.
In each case, the value is an integer. The allowable choices for name are:
predictable-branch-outcomeWhen branch is predicted to be taken with probability lower than this threshold (in percent), then it is considered well predictable. The default is 10.
max-rtl-if-conversion-insnsRTL if-conversion tries to remove conditional branches around a block and replace them with conditionally executed instructions. This parameter gives the maximum number of instructions in a block which should be considered for if-conversion. The default is 10, though the compiler will also use other heuristics to decide whether if-conversion is likely to be profitable.
max-crossjump-edgesThe maximum number of incoming edges to consider for cross-jumping. The algorithm used by -fcrossjumping is O(N^2) in the number of edges incoming to each block. Increasing values mean more aggressive optimization, making the compilation time increase with probably small improvement in executable size.
min-crossjump-insnsThe minimum number of instructions that must be matched at the end of two blocks before cross-jumping is performed on them. This value is ignored in the case where all instructions in the block being cross-jumped from are matched. The default value is 5.
max-grow-copy-bb-insnsThe maximum code size expansion factor when copying basic blocks instead of jumping. The expansion is relative to a jump instruction. The default value is 8.
max-goto-duplication-insnsThe maximum number of instructions to duplicate to a block that jumps to a computed goto. To avoid O(N^2) behavior in a number of passes, GCC factors computed gotos early in the compilation process, and unfactors them as late as possible. Only computed jumps at the end of a basic blocks with no more than max-goto-duplication-insns are unfactored. The default value is 8.
max-delay-slot-insn-searchThe maximum number of instructions to consider when looking for an instruction to fill a delay slot. If more than this arbitrary number of instructions are searched, the time savings from filling the delay slot are minimal, so stop searching. Increasing values mean more aggressive optimization, making the compilation time increase with probably small improvement in execution time.
max-delay-slot-live-searchWhen trying to fill delay slots, the maximum number of instructions to consider when searching for a block with valid live register information. Increasing this arbitrarily chosen value means more aggressive optimization, increasing the compilation time. This parameter should be removed when the delay slot code is rewritten to maintain the control-flow graph.
max-gcse-memoryThe approximate maximum amount of memory that can be allocated in order to perform the global common subexpression elimination optimization. If more memory than specified is required, the optimization is not done.
max-gcse-insertion-ratioIf the ratio of expression insertions to deletions is larger than this value for any expression, then RTL PRE inserts or removes the expression and thus leaves partially redundant computations in the instruction stream. The default value is 20.
max-pending-list-lengthThe maximum number of pending dependencies scheduling allows before flushing the current state and starting over. Large functions with few branches or calls can create excessively large lists which needlessly consume memory and resources.
max-modulo-backtrack-attemptsThe maximum number of backtrack attempts the scheduler should make when modulo scheduling a loop. Larger values can exponentially increase compilation time.
max-inline-insns-singleSeveral parameters control the tree inliner used in GCC. This number sets the maximum number of instructions (counted in GCC’s internal representation) in a single function that the tree inliner considers for inlining. This only affects functions declared inline and methods implemented in a class declaration (C++). The default value is 400.
max-inline-insns-autoWhen you use -finline-functions (included in -O3), a lot of functions that would otherwise not be considered for inlining by the compiler are investigated. To those functions, a different (more restrictive) limit compared to functions declared inline can be applied. The default value is 40.
inline-min-speedupWhen estimated performance improvement of caller + callee runtime exceeds this threshold (in precent), the function can be inlined regardless the limit on --param max-inline-insns-single and --param max-inline-insns-auto.
large-function-insnsThe limit specifying really large functions. For functions larger than this limit after inlining, inlining is constrained by --param large-function-growth. This parameter is useful primarily to avoid extreme compilation time caused by non-linear algorithms used by the back end. The default value is 2700.
large-function-growthSpecifies maximal growth of large function caused by inlining in percents. The default value is 100 which limits large function growth to 2.0 times the original size.
large-unit-insnsThe limit specifying large translation unit. Growth caused by inlining of units larger than this limit is limited by --param inline-unit-growth. For small units this might be too tight. For example, consider a unit consisting of function A that is inline and B that just calls A three times. If B is small relative to A, the growth of unit is 300\% and yet such inlining is very sane. For very large units consisting of small inlineable functions, however, the overall unit growth limit is needed to avoid exponential explosion of code size. Thus for smaller units, the size is increased to --param large-unit-insns before applying --param inline-unit-growth. The default is 10000.
inline-unit-growthSpecifies maximal overall growth of the compilation unit caused by inlining. The default value is 20 which limits unit growth to 1.2 times the original size. Cold functions (either marked cold via an attribute or by profile feedback) are not accounted into the unit size.
ipcp-unit-growthSpecifies maximal overall growth of the compilation unit caused by interprocedural constant propagation. The default value is 10 which limits unit growth to 1.1 times the original size.
large-stack-frameThe limit specifying large stack frames. While inlining the algorithm is trying to not grow past this limit too much. The default value is 256 bytes.
large-stack-frame-growthSpecifies maximal growth of large stack frames caused by inlining in percents. The default value is 1000 which limits large stack frame growth to 11 times the original size.
max-inline-insns-recursivemax-inline-insns-recursive-autoSpecifies the maximum number of instructions an out-of-line copy of a self-recursive inline function can grow into by performing recursive inlining.
--param max-inline-insns-recursive applies to functions declared inline. For functions not declared inline, recursive inlining happens only when -finline-functions (included in -O3) is enabled; --param max-inline-insns-recursive-auto applies instead. The default value is 450.
max-inline-recursive-depthmax-inline-recursive-depth-autoSpecifies the maximum recursion depth used for recursive inlining.
--param max-inline-recursive-depth applies to functions declared inline. For functions not declared inline, recursive inlining happens only when -finline-functions (included in -O3) is enabled; --param max-inline-recursive-depth-auto applies instead. The default value is 8.
min-inline-recursive-probabilityRecursive inlining is profitable only for function having deep recursion in average and can hurt for function having little recursion depth by increasing the prologue size or complexity of function body to other optimizers.
When profile feedback is available (see -fprofile-generate) the actual recursion depth can be guessed from probability that function recurses via a given call expression. This parameter limits inlining only to call expressions whose probability exceeds the given threshold (in percents). The default value is 10.
early-inlining-insnsSpecify growth that the early inliner can make. In effect it increases the amount of inlining for code having a large abstraction penalty. The default value is 14.
max-early-inliner-iterationsLimit of iterations of the early inliner. This basically bounds the number of nested indirect calls the early inliner can resolve. Deeper chains are still handled by late inlining.
comdat-sharing-probabilityProbability (in percent) that C++ inline function with comdat visibility are shared across multiple compilation units. The default value is 20.
profile-func-internal-idA parameter to control whether to use function internal id in profile database lookup. If the value is 0, the compiler uses an id that is based on function assembler name and filename, which makes old profile data more tolerant to source changes such as function reordering etc. The default value is 0.
min-vect-loop-boundThe minimum number of iterations under which loops are not vectorized when -ftree-vectorize is used. The number of iterations after vectorization needs to be greater than the value specified by this option to allow vectorization. The default value is 0.
gcse-cost-distance-ratioScaling factor in calculation of maximum distance an expression can be moved by GCSE optimizations. This is currently supported only in the code hoisting pass. The bigger the ratio, the more aggressive code hoisting is with simple expressions, i.e., the expressions that have cost less than gcse-unrestricted-cost. Specifying 0 disables hoisting of simple expressions. The default value is 10.
gcse-unrestricted-costCost, roughly measured as the cost of a single typical machine instruction, at which GCSE optimizations do not constrain the distance an expression can travel. This is currently supported only in the code hoisting pass. The lesser the cost, the more aggressive code hoisting is. Specifying 0 allows all expressions to travel unrestricted distances. The default value is 3.
max-hoist-depthThe depth of search in the dominator tree for expressions to hoist. This is used to avoid quadratic behavior in hoisting algorithm. The value of 0 does not limit on the search, but may slow down compilation of huge functions. The default value is 30.
max-tail-merge-comparisonsThe maximum amount of similar bbs to compare a bb with. This is used to avoid quadratic behavior in tree tail merging. The default value is 10.
max-tail-merge-iterationsThe maximum amount of iterations of the pass over the function. This is used to limit compilation time in tree tail merging. The default value is 2.
max-unrolled-insnsThe maximum number of instructions that a loop may have to be unrolled. If a loop is unrolled, this parameter also determines how many times the loop code is unrolled.
max-average-unrolled-insnsThe maximum number of instructions biased by probabilities of their execution that a loop may have to be unrolled. If a loop is unrolled, this parameter also determines how many times the loop code is unrolled.
max-unroll-timesThe maximum number of unrollings of a single loop.
max-peeled-insnsThe maximum number of instructions that a loop may have to be peeled. If a loop is peeled, this parameter also determines how many times the loop code is peeled.
max-peel-timesThe maximum number of peelings of a single loop.
max-peel-branchesThe maximum number of branches on the hot path through the peeled sequence.
max-completely-peeled-insnsThe maximum number of insns of a completely peeled loop.
max-completely-peel-timesThe maximum number of iterations of a loop to be suitable for complete peeling.
max-completely-peel-loop-nest-depthThe maximum depth of a loop nest suitable for complete peeling.
max-unswitch-insnsThe maximum number of insns of an unswitched loop.
max-unswitch-levelThe maximum number of branches unswitched in a single loop.
lim-expensiveThe minimum cost of an expensive expression in the loop invariant motion.
iv-consider-all-candidates-boundBound on number of candidates for induction variables, below which all candidates are considered for each use in induction variable optimizations. If there are more candidates than this, only the most relevant ones are considered to avoid quadratic time complexity.
iv-max-considered-usesThe induction variable optimizations give up on loops that contain more induction variable uses.
iv-always-prune-cand-set-boundIf the number of candidates in the set is smaller than this value, always try to remove unnecessary ivs from the set when adding a new one.
scev-max-expr-sizeBound on size of expressions used in the scalar evolutions analyzer. Large expressions slow the analyzer.
scev-max-expr-complexityBound on the complexity of the expressions in the scalar evolutions analyzer. Complex expressions slow the analyzer.
vect-max-version-for-alignment-checksThe maximum number of run-time checks that can be performed when doing loop versioning for alignment in the vectorizer.
vect-max-version-for-alias-checksThe maximum number of run-time checks that can be performed when doing loop versioning for alias in the vectorizer.
vect-max-peeling-for-alignmentThe maximum number of loop peels to enhance access alignment for vectorizer. Value -1 means no limit.
max-iterations-to-trackThe maximum number of iterations of a loop the brute-force algorithm for analysis of the number of iterations of the loop tries to evaluate.
hot-bb-count-ws-permilleA basic block profile count is considered hot if it contributes to the given permillage (i.e. 0...1000) of the entire profiled execution.
hot-bb-frequency-fractionSelect fraction of the entry block frequency of executions of basic block in function given basic block needs to have to be considered hot.
max-predicted-iterationsThe maximum number of loop iterations we predict statically. This is useful in cases where a function contains a single loop with known bound and another loop with unknown bound. The known number of iterations is predicted correctly, while the unknown number of iterations average to roughly 10. This means that the loop without bounds appears artificially cold relative to the other one.
builtin-expect-probabilityControl the probability of the expression having the specified value. This parameter takes a percentage (i.e. 0 ... 100) as input. The default probability of 90 is obtained empirically.
align-thresholdSelect fraction of the maximal frequency of executions of a basic block in a function to align the basic block.
align-loop-iterationsA loop expected to iterate at least the selected number of iterations is aligned.
tracer-dynamic-coveragetracer-dynamic-coverage-feedbackThis value is used to limit superblock formation once the given percentage of executed instructions is covered. This limits unnecessary code size expansion.
The tracer-dynamic-coverage-feedback parameter is used only when profile feedback is available. The real profiles (as opposed to statically estimated ones) are much less balanced allowing the threshold to be larger value.
tracer-max-code-growthStop tail duplication once code growth has reached given percentage. This is a rather artificial limit, as most of the duplicates are eliminated later in cross jumping, so it may be set to much higher values than is the desired code growth.
tracer-min-branch-ratioStop reverse growth when the reverse probability of best edge is less than this threshold (in percent).
tracer-min-branch-probabilitytracer-min-branch-probability-feedbackStop forward growth if the best edge has probability lower than this threshold.
Similarly to tracer-dynamic-coverage two parameters are provided. tracer-min-branch-probability-feedback is used for compilation with profile feedback and tracer-min-branch-probability compilation without. The value for compilation with profile feedback needs to be more conservative (higher) in order to make tracer effective.
max-cse-path-lengthThe maximum number of basic blocks on path that CSE considers. The default is 10.
max-cse-insnsThe maximum number of instructions CSE processes before flushing. The default is 1000.
ggc-min-expandGCC uses a garbage collector to manage its own memory allocation. This parameter specifies the minimum percentage by which the garbage collector’s heap should be allowed to expand between collections. Tuning this may improve compilation speed; it has no effect on code generation.
The default is 30% + 70% * (RAM/1GB) with an upper bound of 100% when
RAM >= 1GB. If getrlimit is available, the notion of “RAM” is
the smallest of actual RAM and RLIMIT_DATA or RLIMIT_AS. If
GCC is not able to calculate RAM on a particular platform, the lower
bound of 30% is used. Setting this parameter and
ggc-min-heapsize to zero causes a full collection to occur at
every opportunity. This is extremely slow, but can be useful for
debugging.
ggc-min-heapsizeMinimum size of the garbage collector’s heap before it begins bothering to collect garbage. The first collection occurs after the heap expands by ggc-min-expand% beyond ggc-min-heapsize. Again, tuning this may improve compilation speed, and has no effect on code generation.
The default is the smaller of RAM/8, RLIMIT_RSS, or a limit that tries to ensure that RLIMIT_DATA or RLIMIT_AS are not exceeded, but with a lower bound of 4096 (four megabytes) and an upper bound of 131072 (128 megabytes). If GCC is not able to calculate RAM on a particular platform, the lower bound is used. Setting this parameter very large effectively disables garbage collection. Setting this parameter and ggc-min-expand to zero causes a full collection to occur at every opportunity.
max-reload-search-insnsThe maximum number of instruction reload should look backward for equivalent register. Increasing values mean more aggressive optimization, making the compilation time increase with probably slightly better performance. The default value is 100.
max-cselib-memory-locationsThe maximum number of memory locations cselib should take into account. Increasing values mean more aggressive optimization, making the compilation time increase with probably slightly better performance. The default value is 500.
max-sched-ready-insnsThe maximum number of instructions ready to be issued the scheduler should consider at any given time during the first scheduling pass. Increasing values mean more thorough searches, making the compilation time increase with probably little benefit. The default value is 100.
max-sched-region-blocksThe maximum number of blocks in a region to be considered for interblock scheduling. The default value is 10.
max-pipeline-region-blocksThe maximum number of blocks in a region to be considered for pipelining in the selective scheduler. The default value is 15.
max-sched-region-insnsThe maximum number of insns in a region to be considered for interblock scheduling. The default value is 100.
max-pipeline-region-insnsThe maximum number of insns in a region to be considered for pipelining in the selective scheduler. The default value is 200.
min-spec-probThe minimum probability (in percents) of reaching a source block for interblock speculative scheduling. The default value is 40.
max-sched-extend-regions-itersThe maximum number of iterations through CFG to extend regions. A value of 0 (the default) disables region extensions.
max-sched-insn-conflict-delayThe maximum conflict delay for an insn to be considered for speculative motion. The default value is 3.
sched-spec-prob-cutoffThe minimal probability of speculation success (in percents), so that speculative insns are scheduled. The default value is 40.
sched-state-edge-prob-cutoffThe minimum probability an edge must have for the scheduler to save its state across it. The default value is 10.
sched-mem-true-dep-costMinimal distance (in CPU cycles) between store and load targeting same memory locations. The default value is 1.
selsched-max-lookaheadThe maximum size of the lookahead window of selective scheduling. It is a depth of search for available instructions. The default value is 50.
selsched-max-sched-timesThe maximum number of times that an instruction is scheduled during selective scheduling. This is the limit on the number of iterations through which the instruction may be pipelined. The default value is 2.
selsched-insns-to-renameThe maximum number of best instructions in the ready list that are considered for renaming in the selective scheduler. The default value is 2.
sms-min-scThe minimum value of stage count that swing modulo scheduler generates. The default value is 2.
max-last-value-rtlThe maximum size measured as number of RTLs that can be recorded in an expression in combiner for a pseudo register as last known value of that register. The default is 10000.
max-combine-insnsThe maximum number of instructions the RTL combiner tries to combine. The default value is 2 at -Og and 4 otherwise.
integer-share-limitSmall integer constants can use a shared data structure, reducing the compiler’s memory usage and increasing its speed. This sets the maximum value of a shared integer constant. The default value is 256.
ssp-buffer-sizeThe minimum size of buffers (i.e. arrays) that receive stack smashing protection when -fstack-protection is used.
min-size-for-stack-sharingThe minimum size of variables taking part in stack slot sharing when not optimizing. The default value is 32.
max-jump-thread-duplication-stmtsMaximum number of statements allowed in a block that needs to be duplicated when threading jumps.
max-fields-for-field-sensitiveMaximum number of fields in a structure treated in a field sensitive manner during pointer analysis. The default is zero for -O0 and -O1, and 100 for -Os, -O2, and -O3.
prefetch-latencyEstimate on average number of instructions that are executed before prefetch finishes. The distance prefetched ahead is proportional to this constant. Increasing this number may also lead to less streams being prefetched (see simultaneous-prefetches).
simultaneous-prefetchesMaximum number of prefetches that can run at the same time.
l1-cache-line-sizeThe size of cache line in L1 cache, in bytes.
l1-cache-sizeThe size of L1 cache, in kilobytes.
l2-cache-sizeThe size of L2 cache, in kilobytes.
min-insn-to-prefetch-ratioThe minimum ratio between the number of instructions and the number of prefetches to enable prefetching in a loop.
prefetch-min-insn-to-mem-ratioThe minimum ratio between the number of instructions and the number of memory references to enable prefetching in a loop.
use-canonical-typesWhether the compiler should use the “canonical” type system. By default, this should always be 1, which uses a more efficient internal mechanism for comparing types in C++ and Objective-C++. However, if bugs in the canonical type system are causing compilation failures, set this value to 0 to disable canonical types.
switch-conversion-max-branch-ratioSwitch initialization conversion refuses to create arrays that are bigger than switch-conversion-max-branch-ratio times the number of branches in the switch.
max-partial-antic-lengthMaximum length of the partial antic set computed during the tree partial redundancy elimination optimization (-ftree-pre) when optimizing at -O3 and above. For some sorts of source code the enhanced partial redundancy elimination optimization can run away, consuming all of the memory available on the host machine. This parameter sets a limit on the length of the sets that are computed, which prevents the runaway behavior. Setting a value of 0 for this parameter allows an unlimited set length.
sccvn-max-scc-sizeMaximum size of a strongly connected component (SCC) during SCCVN processing. If this limit is hit, SCCVN processing for the whole function is not done and optimizations depending on it are disabled. The default maximum SCC size is 10000.
sccvn-max-alias-queries-per-accessMaximum number of alias-oracle queries we perform when looking for redundancies for loads and stores. If this limit is hit the search is aborted and the load or store is not considered redundant. The number of queries is algorithmically limited to the number of stores on all paths from the load to the function entry. The default maximum number of queries is 1000.
ira-max-loops-numIRA uses regional register allocation by default. If a function contains more loops than the number given by this parameter, only at most the given number of the most frequently-executed loops form regions for regional register allocation. The default value of the parameter is 100.
ira-max-conflict-table-sizeAlthough IRA uses a sophisticated algorithm to compress the conflict table, the table can still require excessive amounts of memory for huge functions. If the conflict table for a function could be more than the size in MB given by this parameter, the register allocator instead uses a faster, simpler, and lower-quality algorithm that does not require building a pseudo-register conflict table. The default value of the parameter is 2000.
ira-loop-reserved-regsIRA can be used to evaluate more accurate register pressure in loops for decisions to move loop invariants (see -O3). The number of available registers reserved for some other purposes is given by this parameter. The default value of the parameter is 2, which is the minimal number of registers needed by typical instructions. This value is the best found from numerous experiments.
lra-inheritance-ebb-probability-cutoffLRA tries to reuse values reloaded in registers in subsequent insns. This optimization is called inheritance. EBB is used as a region to do this optimization. The parameter defines a minimal fall-through edge probability in percentage used to add BB to inheritance EBB in LRA. The default value of the parameter is 40. The value was chosen from numerous runs of SPEC2000 on x86-64.
loop-invariant-max-bbs-in-loopLoop invariant motion can be very expensive, both in compilation time and in amount of needed compile-time memory, with very large loops. Loops with more basic blocks than this parameter won’t have loop invariant motion optimization performed on them. The default value of the parameter is 1000 for -O1 and 10000 for -O2 and above.
loop-max-datarefs-for-datadepsBuilding data dependencies is expensive for very large loops. This parameter limits the number of data references in loops that are considered for data dependence analysis. These large loops are no handled by the optimizations using loop data dependencies. The default value is 1000.
max-vartrack-sizeSets a maximum number of hash table slots to use during variable tracking dataflow analysis of any function. If this limit is exceeded with variable tracking at assignments enabled, analysis for that function is retried without it, after removing all debug insns from the function. If the limit is exceeded even without debug insns, var tracking analysis is completely disabled for the function. Setting the parameter to zero makes it unlimited.
max-vartrack-expr-depthSets a maximum number of recursion levels when attempting to map variable names or debug temporaries to value expressions. This trades compilation time for more complete debug information. If this is set too low, value expressions that are available and could be represented in debug information may end up not being used; setting this higher may enable the compiler to find more complex debug expressions, but compile time and memory use may grow. The default is 12.
min-nondebug-insn-uidUse uids starting at this parameter for nondebug insns. The range below the parameter is reserved exclusively for debug insns created by -fvar-tracking-assignments, but debug insns may get (non-overlapping) uids above it if the reserved range is exhausted.
ipa-sra-ptr-growth-factorIPA-SRA replaces a pointer to an aggregate with one or more new parameters only when their cumulative size is less or equal to ipa-sra-ptr-growth-factor times the size of the original pointer parameter.
sra-max-scalarization-size-Ospeedsra-max-scalarization-size-OsizeThe two Scalar Reduction of Aggregates passes (SRA and IPA-SRA) aim to replace scalar parts of aggregates with uses of independent scalar variables. These parameters control the maximum size, in storage units, of aggregate which is considered for replacement when compiling for speed (sra-max-scalarization-size-Ospeed) or size (sra-max-scalarization-size-Osize) respectively.
tm-max-aggregate-sizeWhen making copies of thread-local variables in a transaction, this parameter specifies the size in bytes after which variables are saved with the logging functions as opposed to save/restore code sequence pairs. This option only applies when using -fgnu-tm.
graphite-max-nb-scop-paramsTo avoid exponential effects in the Graphite loop transforms, the number of parameters in a Static Control Part (SCoP) is bounded. The default value is 10 parameters. A variable whose value is unknown at compilation time and defined outside a SCoP is a parameter of the SCoP.
graphite-max-bbs-per-functionTo avoid exponential effects in the detection of SCoPs, the size of the functions analyzed by Graphite is bounded. The default value is 100 basic blocks.
loop-block-tile-sizeLoop blocking or strip mining transforms, enabled with -floop-block or -floop-strip-mine, strip mine each loop in the loop nest by a given number of iterations. The strip length can be changed using the loop-block-tile-size parameter. The default value is 51 iterations.
loop-unroll-jam-sizeSpecify the unroll factor for the -floop-unroll-and-jam option. The default value is 4.
loop-unroll-jam-depthSpecify the dimension to be unrolled (counting from the most inner loop) for the -floop-unroll-and-jam. The default value is 2.
ipa-cp-value-list-sizeIPA-CP attempts to track all possible values and types passed to a function’s parameter in order to propagate them and perform devirtualization. ipa-cp-value-list-size is the maximum number of values and types it stores per one formal parameter of a function.
ipa-cp-eval-thresholdIPA-CP calculates its own score of cloning profitability heuristics and performs those cloning opportunities with scores that exceed ipa-cp-eval-threshold.
ipa-cp-recursion-penaltyPercentage penalty the recursive functions will receive when they are evaluated for cloning.
ipa-cp-single-call-penaltyPercentage penalty functions containg a single call to another function will receive when they are evaluated for cloning.
ipa-max-agg-itemsIPA-CP is also capable to propagate a number of scalar values passed in an aggregate. ipa-max-agg-items controls the maximum number of such values per one parameter.
ipa-cp-loop-hint-bonusWhen IPA-CP determines that a cloning candidate would make the number of iterations of a loop known, it adds a bonus of ipa-cp-loop-hint-bonus to the profitability score of the candidate.
ipa-cp-array-index-hint-bonusWhen IPA-CP determines that a cloning candidate would make the index of an array access known, it adds a bonus of ipa-cp-array-index-hint-bonus to the profitability score of the candidate.
ipa-max-aa-stepsDuring its analysis of function bodies, IPA-CP employs alias analysis in order to track values pointed to by function parameters. In order not spend too much time analyzing huge functions, it gives up and consider all memory clobbered after examining ipa-max-aa-steps statements modifying memory.
lto-partitionsSpecify desired number of partitions produced during WHOPR compilation. The number of partitions should exceed the number of CPUs used for compilation. The default value is 32.
lto-min-partitionSize of minimal partition for WHOPR (in estimated instructions). This prevents expenses of splitting very small programs into too many partitions.
lto-max-partitionSize of max partition for WHOPR (in estimated instructions). to provide an upper bound for individual size of partition. Meant to be used only with balanced partitioning.
cxx-max-namespaces-for-diagnostic-helpThe maximum number of namespaces to consult for suggestions when C++ name lookup fails for an identifier. The default is 1000.
sink-frequency-thresholdThe maximum relative execution frequency (in percents) of the target block relative to a statement’s original block to allow statement sinking of a statement. Larger numbers result in more aggressive statement sinking. The default value is 75. A small positive adjustment is applied for statements with memory operands as those are even more profitable so sink.
max-stores-to-sinkThe maximum number of conditional store pairs that can be sunk. Set to 0 if either vectorization (-ftree-vectorize) or if-conversion (-ftree-loop-if-convert) is disabled. The default is 2.
allow-store-data-racesAllow optimizers to introduce new data races on stores. Set to 1 to allow, otherwise to 0. This option is enabled by default at optimization level -Ofast.
case-values-thresholdThe smallest number of different values for which it is best to use a jump-table instead of a tree of conditional branches. If the value is 0, use the default for the machine. The default is 0.
tree-reassoc-widthSet the maximum number of instructions executed in parallel in reassociated tree. This parameter overrides target dependent heuristics used by default if has non zero value.
sched-pressure-algorithmChoose between the two available implementations of -fsched-pressure. Algorithm 1 is the original implementation and is the more likely to prevent instructions from being reordered. Algorithm 2 was designed to be a compromise between the relatively conservative approach taken by algorithm 1 and the rather aggressive approach taken by the default scheduler. It relies more heavily on having a regular register file and accurate register pressure classes. See haifa-sched.c in the GCC sources for more details.
The default choice depends on the target.
max-slsr-cand-scanSet the maximum number of existing candidates that are considered when seeking a basis for a new straight-line strength reduction candidate.
asan-globalsEnable buffer overflow detection for global objects. This kind of protection is enabled by default if you are using -fsanitize=address option. To disable global objects protection use --param asan-globals=0.
asan-stackEnable buffer overflow detection for stack objects. This kind of protection is enabled by default when using -fsanitize=address. To disable stack protection use --param asan-stack=0 option.
asan-instrument-readsEnable buffer overflow detection for memory reads. This kind of protection is enabled by default when using -fsanitize=address. To disable memory reads protection use --param asan-instrument-reads=0.
asan-instrument-writesEnable buffer overflow detection for memory writes. This kind of protection is enabled by default when using -fsanitize=address. To disable memory writes protection use --param asan-instrument-writes=0 option.
asan-memintrinEnable detection for built-in functions. This kind of protection is enabled by default when using -fsanitize=address. To disable built-in functions protection use --param asan-memintrin=0.
asan-use-after-returnEnable detection of use-after-return. This kind of protection is enabled by default when using -fsanitize=address option. To disable use-after-return detection use --param asan-use-after-return=0.
asan-instrumentation-with-call-thresholdIf number of memory accesses in function being instrumented is greater or equal to this number, use callbacks instead of inline checks. E.g. to disable inline code use --param asan-instrumentation-with-call-threshold=0.
chkp-max-ctor-sizeStatic constructors generated by Pointer Bounds Checker may become very large and significantly increase compile time at optimization level -O1 and higher. This parameter is a maximum nubmer of statements in a single generated constructor. Default value is 5000.
max-fsm-thread-path-insnsMaximum number of instructions to copy when duplicating blocks on a finite state automaton jump thread path. The default is 100.
max-fsm-thread-lengthMaximum number of basic blocks on a finite state automaton jump thread path. The default is 10.
max-fsm-thread-pathsMaximum number of new jump thread paths to create for a finite state automaton. The default is 50.
parloops-chunk-sizeChunk size of omp schedule for loops parallelized by parloops. The default is 0.
parloops-scheduleSchedule type of omp schedule for loops parallelized by parloops (static, dynamic, guided, auto, runtime). The default is static.
max-ssa-name-query-depthMaximum depth of recursion when querying properties of SSA names in things like fold routines. One level of recursion corresponds to following a use-def chain.
hsa-gen-debug-storesEnable emission of special debug stores within HSA kernels which are then read and reported by libgomp plugin. Generation of these stores is disabled by default, use --param hsa-gen-debug-stores=1 to enable it.
max-speculative-devirt-maydefsThe maximum number of may-defs we analyze when looking for a must-def specifying the dynamic type of an object that invokes a virtual call we may be able to devirtualize speculatively.
Next: Instrumentation Options, Previous: Debugging Options, Up: Invoking GCC [Contents][Index]