Selecting one of the following will take you directly to that section:
Somewhere between -O0 and -O2.
If multiple "O" options are used, with or without level numbers, the last such option is the one that is effective.
Moderate level of optimization which enables most optimizations. This is the default when no "-O" option is specified, or if no value is specified (i.e. "-O").
If multiple "O" options are used, with or without level numbers, the last such option is the one that is effective.
Like -O2, except that it enables optimizations that take longer to perform or that may generate larger code (in an attempt to make the program run faster).
If multiple "O" options are used, with or without level numbers, the last such option is the one that is effective.
Enables all the optimizations from -O3 along with other aggressive optimizations that may violate strict compliance with language standards. Refer to the AOCC options document for the language you're using for more detailed documentation of optimizations enabled under -Ofast.
The option zopt enables architecture specific optimizations tuned for Zen version 4 processors. These safe but aggressive optimizations include various scalar, vector and loop transformations including loop-fusion, loop-interchange, loop-unswitch, loop-versioning and more.
Specify that Clang should generate code for a specific processor family member and later. For example, if you specify -march=znver1, the compiler is allowed to generate instructions that are valid on AMD Zen processors, but which may not exist on earlier products. -march=znver4 enables AVX 512 ISA for Genoa (znver4) processors.
Generate output files in LLVM formats suitable for link time optimization. When used with -S this generates LLVM intermediate language assembly files, otherwise this generates LLVM bitcode format object files (which may be passed to the linker depending on the stage selection options).
Generate code for a 32-bit environment. The 32-bit environment sets int, long and pointer to 32 bits and generates code that runs on any i386 system. The compiler generates x86 or IA32 32-bit ABI. The default on a 32-bit host is 32-bit ABI. The default on a 64-bit host is 64-bit ABI if the target platform specified is 64-bit, otherwise the default is 32-bit.
Generate code for a 64-bit environment. The 64-bit environment sets int to 32 bits and long and pointer to 64 bits and generates code for AMD's x86-64 architecture. The compiler generates AMD64, INTEL64, x86-64 64-bit ABI. The default on a 32-bit host is 32-bit ABI. The default on a 64-bit host is 64-bit ABI if the target platform specified is 64-bit, otherwise the default is 32-bit.
Enables a range of optimizations that provide faster, though sometimes less precise, mathematical operations that may not conform to the IEEE-754 specifications. When this option is specified, the __STDC_IEC_559__ macro is ignored even if set by the system headers.
Enable handling of OpenMP directives and generate parallel code. The openmp library to be linked can be specified through -fopenmp=library option.
Instructs the compiler to conform to the IEEE-754 specifications. The compiler will Perform floating-point operations in strict conformance with the IEEE 754 standard. Some optimizations are disabled when this option is specified
Instructs the compiler to link with system math libraries.
Instructs the compiler to link with AMD-supported optimized math library.
Instructs the linker to use the first definition encountered for a symbol, and ignore all others.
The amdalloc library is a variant of jemalloc library. Use the jemalloc library, which is a general purpose malloc(3) implementation that emphasizes fragmentation avoidance and scalable concurrency support.
The amdalloc library is a variant of jemalloc library. Use the jemalloc library, which is a general purpose malloc(3) implementation that emphasizes fragmentation avoidance and scalable concurrency support. The amdalloc-ext further tries to optimize the library based on system page size.
Instructs the compiler to link with the OpenMP runtime libraries.
Instructs the compiler to link with flang Fortran runtime libraries.
Enable various transformations to generate vector instructions. It would be enabled under -zopt as well.
Enable various transformation involving code versioning, code movements and conversion optimzation.It would be enabled under -zopt as well.
Enable various loop transformations such as loop fusion and loop interchange.
Enable various loop transforms that can potentially increase code size such as loop unswitch and loop versioning.
This option enables the GVN hoist pass, which is used to hoist computations from branches.
Run cleanup optimization passes after vectorization.
Enables splitting of long live ranges of loop induction variables which span loop boundaries. This helps reduce register pressure and can help avoid needless spills to memory and reloads from memory.
This optimization does partial unswitching of loops where some part of the unswitched control flow remains in the loop.
This option enables aggressive loop unswitching heuristic (including -enable-partial-unswitch) based on the usage of the branch conditional values. Loop unswitching leads to code-bloat. Code-bloat can be minimized if the hoisted condition is executed more often. This heuristic prioritizes the conditions based on the number of times they are used within the loop. The heuristic can be controlled with the following options:
Enables unswitching of a loop with respect to a branch conditional value (B), where B appears in at least <n> compares in the loop. This option is enabled with -aggressive-loop-unswitch. Default value is 3.
Usage: -mllvm -aggressive-loop-unswitch -mllvm -unswitch-identical-branches-min-count=<n> where n is a positive integer and lower value of <n> facilitates more unswitching
Enables unswitching of a loop with respect to a branch conditional value (B), where B appears in at most <n> compares in the loop. This option is enabled with -aggressive-loop-unswitch. Default value is 6.
Usage: -mllvm -aggressive-loop-unswitch -mllvm -unswitch-identical-branches-max-count=<n> where n is a positive integer and higher value of <n> facilitates more unswitching
Note: These options may facilitate more unswitching in some of the workloads. Since loop-unswitching inherently leads to code bloat, facilitating more unswitching may significantly increase the code size and hence may also lead to longer compilation times.
This optimization enables generation of prefetch instructions for tightly coupled loops
This option enables an optimization that generates and calls specialized function versions when the loops inside function are vectorizable and the arguments are not aliased with each other. This optimization helps in function inlining and vectorization.
This option enables an optimization that does the slp vectorization across basic blocks. The SLP vectorizer vectorizes instructions within basic blocks. The global slp vectorizer analyzes instructions across basic blocks and vectorizes them.
Attempts epilog vectorization on loops that required induction variables to be widened.
Converts the call to floating point exponent version of pow to its integer exponent version if the floating-point exponent can be converted to integer. This option is set to true by default.
Enables estimation of the virtual register pressure before performing loop invariant code motion. This estimation is used to decide the invariants that will be hoisted during loop invariant code motion.
Enables fusion of adjacent tiled loops as a part of loop tiling transformation. This option is set to false by default.
Optimizes the cost model for strided access computation.
This option enables an optimization that transforms the data layout of a single dimensional array to provide better cache locality by analysing the access patterns.
Enables dead virtual function elimination optimization. Requires -flto=full.
This flag enables vectorization of loops with complex control flow that can not be vectorized by loop and slp vectorizers.
This option eliminates the array computations based on their usage. The computations on unused array elements and computations on zero valued array elements are eliminated with this optimization. -flto as whole program analysis is required to perform this optimization.
Possible values:
Block Reordering
Possible values:
Analyzes the whole program to determine if the structures in the code can be peeled, if dead fields can be deleted, and if the pointer or integer fields in the structure can be compressed. If feasible, this optimization transforms the code to enable these improvements. This transformation is likely to improve cache utilization and memory bandwidth. It is expected to improve the scalability of programs executed on multiple cores. This is effective only under flto as the whole program analysis is required to perform this optimization. You can choose different levels of aggressiveness with which this optimization can be applied to your application; with 1 being the least aggressive and 7 being the most aggressive level.
Possible values:
fstruct-layout=4 and fstruct-layout=5 are derived from fstruct-layout=2 and fstruct-layout=3 respectively with the added feature of safe compression of 64-bit integer fields to 32-bit integer fields in structures. Going from fstruct-layout=4 to fstruct-layout=5 may result in higher performance if the pointer values are such that the pointers can be compressed to 16-bits.
fstruct-layout=6 and fstruct-layout=7 are derived from fstruct-layout=2 and fstructlayout=3 respectively, with the added feature of safe compression of 64 bit integer fields to 16 bit integer in structures. Going from fstruct-layout=6 to fstruct-layout=7 may result in higher performance if the pointer values are such that the pointers can be compressed to 16-bits.
Turns on LLVM's instrumenation based profiling.
Disable the loop reroll.
This option enables generation of gather instructions for the cases where it is profitable.
Enables loop strip mining optimization. This optimization fragments a large loop into smaller segments or strips to improve temporal and spatial locality in the data cache.
Uses the profiling files generated from a program compiled with -fprofile-instr-generate to guide optimization decisions.
In the 502/602.gcc benchmark description, "multiple definitions of symbols" is listed under the "Known Portability Issues" section, and this option is one of the suggested workarounds. This option causes Clang to revert to the same inlining behavior that GCC does when in pre-C99 mode.
Sets the compiler's inlining heuristics to an aggressive level by increasing the inline thresholds.
Sets the compiler's inlining threshold level to the value passed as the argument. The inline threshold is used in the inliner heuristics to decide which functions should be inlined.
Enables inlining for recursive functions based on heuristics, with level 4 being most aggressive. Higher levels may lead to code bloat due to expansion of recursive functions at call sites.
Levels:
This option enables aggressive loop unswitching heuristic based on usage of branch conditions.
Enables loop strength reduction for nested loop structures. By default, the compiler will do loop strength reduction only for the innermost loop.
Enables loop distribution with Scalar expansion for better vectorization. This should be used along with -mllvm -enable-loop-distribute-adv flag.
Allocate local variables on the stack, thus allowing recursion. SAVEd, data-initialized, or namelist members are always allocated statically, regardless of the setting of this switch.
Disable generation of fma instructions when there is a chain of fma instructions and output of one fma instruction is used as input to other fma instruction.
Sets the limit at which loops will be unrolled. For example, if unroll threshold is set to 100 then only loops with 100 or fewer instructions will be unrolled.
Sets the limit at which loops will be unswitched. For example, if unswitch threshold is set to 100 then only loops with 100 or fewer instructions will be unswtched.
Enables aggressive heuristics to get loop unrolling.
This option instructs the compiler to unroll loops wherever possible.
Use the given vector functions library.
This option controls whether AOCC emits (true) or does not emit (false) a vzeroupper instruction before a transfer of control flow. Not emitting the vzeroupper instruction can help minimize the AVX to SSE transition penalty.
Force the alignment of all blocks that have no fall-through predecessors (i.e. don't add nops that are executed). In log2 format (e.g 4 means align on 16B boundaries).
Do not generate an error when linking multiple symbols of the same name.
Set the default symbol visibility for all global declarations.
The binary datasets for some of the Fortran benchmarks in the SPEC CPU suites are stored in big-endian format. This option is necessary for those datasets to be read in correctly.
Specifies size of off_t data type.
This option instructs the compiler to treat char type as unsigned.
clang is a C compiler which encompasses preprocessing, parsing, optimization, code generation, assembly, and linking. Depending on which high-level mode setting is passed, Clang will stop before doing a full link.
clang++ C++ compiler which encompasses preprocessing, parsing, optimization, code generation, assembly, and linking. Depending on which high-level mode setting is passed, Clang will stop before doing a full link.
flang is a Fortran compiler which encompasses parsing, optimization, code generation, assembly, and linking. Depending on which high-level mode setting is passed, Flang will stop before doing a full link.
Do not warn about functions defined with a return type that defaults to "int" or which return something other than what they were declared to.
Do not warn about unused command line arguments.
Specifies a directory to search for libraries. Use -L to add directories to the search path for library files. Multiple -L options are valid. However, the position of multiple -L options is important relative to -l options supplied.
Specifies a directory to search for include files. Use -I to add directories to the search path for include files. Multiple -I options are valid.
Using numactl to bind processes and memory to cores
For multi-copy runs or single copy runs on systems with multiple sockets, it is advantageous to bind a process to a
particular core. Otherwise, the OS may arbitrarily move your process from one core to another. This can affect
performance. To help, SPEC allows the use of a "submit" command where users can specify a utility to use to bind
processes. We have found the utility 'numactl' to be the best choice.
numactl runs processes with a specific NUMA scheduling or memory placement policy. The policy is set for a
command and inherited by all of its children. The numactl flag "--physcpubind" specifies
which core(s) to bind the process. "-l" instructs numactl to keep a process's memory on the
local node while "-m" specifies which node(s) to place a process's memory. For full details on using
numactl, please refer to your Linux documentation, 'man numactl'
Note that some older versions of numactl incorrectly interpret application arguments as its own. For
example, with the command "numactl --physcpubind=0 -l a.out -m a", numactl will interpret
a.out's "-m" option as its own "-m" option. To work around this problem, we put
the command to be run in a shell script and then run the shell script using numactl. For example:
"echo 'a.out -m a' > run.sh ; numactl --physcpubind=0 bash run.sh"
numactl --interleave=all runcpu
numactl --interleave=all runcpu executes the SPEC CPU command runcpu so that memory is consumed across NUMA nodes rather than consumed from a single node. This helps prevent local node out-of-memory conditions which can occur when runcpu is executed without interleaving.
For full details on using numactl, please refer to your Linux documentation, 'man numactl'
Transparent Huge Pages (THP)
THP is an abstraction layer that automates most aspects of creating, managing, and using huge pages. It is designed to hide much of the complexity in using huge pages from system administrators and developers. Huge pages increase the memory page size from 4 kilobytes to 2 megabytes. This provides significant performance advantages on systems with highly contended resources and large memory workloads. If memory utilization is too high or memory is badly fragmented which prevents huge pages being allocated, the kernel will assign smaller 4k pages instead. Most recent Linux OS releases have THP enabled by default.
THP usage is controlled by the sysfs setting /sys/kernel/mm/transparent_hugepage/enabled.
Possible values:
The SPEC CPU benchmark codes themselves never explicitly request huge pages, as the mechanism to do that is OS-specific and can change over time. Libraries such as amdalloc which are used by the benchmarks may explicitly request huge pages, and use of such libraries can make the "madvise" setting relevant and useful.
When no huge pages are immediately available and one is requested, how the system handles the request for THP creation is
controlled by the sysfs setting /sys/kernel/mm/transparent_hugepage/defrag.
Possible values:
An application that "always" requests THP often can benefit from waiting for an allocation until those huge pages can be assembled.
For more information see the Linux transparent hugepage documentation.
ulimit -s <n>
Sets the stack size to n kbytes, or unlimited to allow the stack size to grow without limit.
ulimit -l <n>
Sets the maximum size of memory that may be locked into physical memory.
powersave -f (on SuSE)
Makes the powersave daemon set the CPUs to the highest supported frequency.
/etc/init.d/cpuspeed stop (on Red Hat)
Disables the cpu frequency scaling program in order to set the CPUs to the highest supported frequency.
LD_LIBRARY_PATH
An environment variable that indicates the location in the filesystem of bundled libraries to use when running the benchmark binaries.
sysctl -w vm.dirty_ratio=8
Limits dirty cache to 8% of memory.
sysctl -w vm.swappiness=1
Limits swap usage to minimum necessary.
sysctl -w vm.zone_reclaim_mode=1
Frees local node memory first to avoid remote memory usage.
kernel/numa_balancing
This OS setting controls automatic NUMA balancing on memory mapping and process placement. NUMA balancing incurs overhead for no benefit on workloads that are already bound to NUMA nodes.
Possible settings:
For more information see the numa_balancing entry in the
Linux sysctl documentation.
kernel/randomize_va_space (ASLR)
This setting can be used to select the type of process address space randomization. Defaults differ based on whether the architecture supports ASLR, whether the kernel was built with the CONFIG_COMPAT_BRK option or not, or the kernel boot options used.
Possible settings:
norandmaps" parameter.CONFIG_COMPAT_BRK option is enabled at kernel build time.CONFIG_COMPAT_BRK is
disabled.
Disabling ASLR can make process execution more deterministic and runtimes more consistent.
For more information see the randomize_va_space entry in the
Linux sysctl documentation.
vm/drop_caches
The two commands are equivalent: echo 3> /proc/sys/vm/drop_caches and sysctl -w vm.drop_caches=3 Both must be run as root. The commands are used to free up the filesystem page cache, dentries, and inodes.
Possible settings:
MALLOC_CONF
The amdalloc library is a variant of jemalloc library. The amdalloc
library has tunable parameters, many of which may be changed at run-time via several mechanisms, one of which
is the MALLOC_CONF environment variable. Other methods, as well as the order in which they're referenced,
are detailed in the jemalloc documentation's TUNING section.
The options that can be tuned at run-time are everything in the jemalloc documentation's
MALLCTL NAMESPACE section that begins with
"opt.".
The options that may be encountered in SPEC CPU 2017 results are detailed here:
retain:true - Causes unused virtual memory to
be retained for later reuse rather than discarding it. This is the default for 64-bit Linux.thp:never - Attempts to never utilize huge pages
by using MADV_NOHUGEPAGE on all mappings. This option has no effect except when THP is set to
"madvise".PGHPF_ZMEM
An environment variable used to initialize the allocated memory. Setting PGHPF_ZMEM to "Yes" has the effect of initializing all allocated memory to zero.
GOMP_CPU_AFFINITY
This environment variable is used to set the thread affinity for threads spawned by OpenMP.
OMP_DYNAMIC
This environment variable is defined as part of the OpenMP standard. Setting it to "false" prevents the OpenMP runtime from dynamically adjusting the number of threads to use for parallel execution.
For more information, see chapter 4 ("Environment Variables") in the OpenMP 4.5 Specification.
OMP_SCHEDULE
This environment variable is defined as part of the OpenMP standard. Setting it to "static" causes loop iterations to be assigned to threads in round-robin fashion in the order of the thread number.
For more information, see chapter 4 ("Environment Variables") in the OpenMP 4.5 Specification.
OMP_STACKSIZE
This environment variable is defined as part of the OpenMP standard and controls the size of the stack for threads created by OpenMP.
For more information, see chapter 4 ("Environment Variables") in the OpenMP 4.5 Specification.
OMP_THREAD_LIMIT
This environment variable is defined as part of the OpenMP standard and limits the maximum number of OpenMP threads that can be created.
For more information, see chapter 4 ("Environment Variables") in the OpenMP 4.5 Specification.