Details:
- Added a new script, monolithify-header.sh, to the 'build' directory.
This script recursively replaces all #include directives in a selected
file with the contents of the header files referenced by each directive.
The idea is to "flatten" a tree of .h files into a single file, with
the script acting as a C preprocessor that only processes #include
directives.
Details:
- Removed an errant #include "blis.h directive from bli_cntx_ind_stage.h.
The generaly policy is that no header file in BLIS should include
blis.h. This will be important in the near future when using a tool to
recursively create a monolithic blis.h file from its consitutent
headers.
Details:
- Updated bli_cpuid_query_id() so that BLIS_ARCH_GENERIC is always returned
if the hardware fails to test positive for any supported sub-configuration.
- Defined bli_gks_init_ref_cntx(), which will call the context initialization
function bli_cntx_init_configname() for the sub-configuration 'configname'
associated with the arch_t id returned by bli_arch_query_id(). This makes
initializing a reference context easy for experts who wish to construct
those contexts.
Details:
- Fixed incorrect calling sequence in bli_cntx_init_knl.c--an instance of
bli_blksz_init_easy() that should have been bli_blksz_init().
- Fixed a bug in code that is supposed to output the list of sub-directories
in the 'config' directory when configure script is run with no arguments.
- Expanded the output of "make showconfig" to include more info from config.mk.
- Minor changes to build/auto-detect/cpuid_x86.c, mostly in preparation for
someone to add excavator and zen support.
- Added a link to the ConfigurationHowTo wiki to config_registry.
- Other minor tweaks to configure.
Details:
- Added runtime support for selecting an appropriate arch_t value based
on the results of the cpuid instruction (for x86_64). This allows
deferral of choosing a context (kernels, blocksizes, etc.) until
runtime, which allows BLIS to be built with support for multiple
microarchitectures. Currently, only amd64 and intel64 configurations
are registered in the config_registry; however, one could create
custom configuration families to support arbitrary sets of x86_64
microarchitectures.
- Current Intel microarchitectures supported via cpuid are knl, haswell,
sandybridge, and penryn.
- Current AMD microarchitectures supported via cpuid are: zen, excavator,
steamroller, piledriver, and bulldozer.
Details:
- Removed the default-overriding #define of BLIS_SIMD_ALIGN_SIZE set in
bli_family_bulldozer.h. Not sure where this value came from, but it
would seem to allow for insufficient starting address alignment for
any matrices created via bli_malloc_user(), such as via
bli_obj_create(). Thanks to Rene Sitt for reporting the behavior that
led us to this bug.
Details:
- Discovered that sed commands where the replacement string contains '\n'
are problematic with the version of sed present in OS X. For these cases
cases in the configure script, we instead use 'perl -pe' for
search-and-replace functionality.
- Various other minor comment/whitespace tweaks to configure.
- Removed remaining lines of code related to setting/checking variables to
track "unregistered" configurations.
Details:
- Typecast l1mkr_t enum value in bli_cntx.h to guint_t before testing for
out-of-range value. This is an attempt to pacify a strange warning from
clang on OS X that is seemingly the result of the following compiler
warning flag:
-Wtautological-constant-out-of-range-compare
Details:
- Implemented a replacement for associative arrays in the configure script
that does not utilize arrays, and therefore works in pre-4.0 versions of
bash. (It appears that Mac OS X will be stuck with version 3.2 indefinitely
due to bash switching to the GPL 3.0 license starting with version 4.0.)
Details:
- Added a "generic" configuration that leaves the default blocksizes and
kernels unchanged. This replaces the older "reference" configuration.
Updated auto-detect script and code accordingly.
- Added support for generic configuration to arch_t (bli_type_defs.h),
bli_gks_init() (bli_gks.c), and bli_arch_config.h
- Moved bli_arch_query_id() to bli_arch.c (and prototype to bli_arch.h).
- Whitespace changes to configurations' make_defs.mk files.
Details:
- Renamed the various configurations' "bli_arch_<configname>.h" header files
(replacing "arch" with "family") to free up the 'bli_arch' namespace for a
different purpose (hardware detection).
- Renamed "bli_arch.h" and "bli_arch_pre_macro_defs.h" in frame/include to
"bli_arch_config.h" and "bli_arch_config_pre.h", respectively.
Details:
- Updated the top-level Makefile's build rule for testsuite object files to
properly obtain CFLAGS via get-frame-cflags-for() function instead of
simply using the $(CFLAGS) variable (which is empty). This means that
'make test' should now work as expected.
Details:
- Fixed semi-broken testsuite Makefile and very-broken test driver Makefiles,
as well as those for test/3m4m, test/thread_ranges, and test/exec_sizes
sub-directories.
- Factored out much of the top-level Makefile into common.mk. A Makefile
needs only set DIST_PATH to the relative path to the top level of the
BLIS source distribution before including common.mk in order to acquire
all of the definitions typically needed in a Makefile that tests BLIS.
Details:
- Removed a duplicate header file that was causing problems during
installation for the 'knl' configuration. Thanks to Victor Eijkhout
for reporting this issue.
Details:
- Reworked the build system around a configuration registry file, named
config_registry', that identifies valid configuration targets, their
constituent sub-configurations, and the kernel sets that are needed by
those sub-configurations. The build system now facilitates the building
of a single library that can contains kernels and cache/register
blocksizes for multiple configurations (microarchitectures). Reference
kernels are also built on a per-configuration basis.
- Updated the Makefile to use new variables set by configure via the
config.mk.in template, such as CONFIG_LIST, KERNEL_LIST, and KCONFIG_MAP,
in determining which sub-configurations (CONFIG_LIST) and kernel sets
(KERNEL_LIST) are included in the library, and which make_defs.mk files'
CFLAGS (KCONFIG_MAP) are used when compiling kernels.
- Reorganized 'kernels' directory into a "flat" structure. Renamed kernel
functions into a standard format that includes the kernel set name
(e.g. 'haswell'). Created a "bli_kernels_<kernelset>.h" file in each
kernels sub-directory. These files exist to provide prototypes for the
kernels present in those directories.
- Reorganized reference kernels into a top-level 'ref_kernels' directory.
This directory includes a new source file, bli_cntx_ref.c (compiled on
a per-configuration basis), that defines the code needed to initialize
a reference context and a context for induced methods for the
microarchitecture in question.
- Rewrote make_defs.mk files in each configuration so that the compiler
variables (e.g. CFLAGS) are "stored" (renamed) on a per-configuration
basis.
- Modified bli_config.h.in template so that bli_config.h is generated with
#defines for the config (family) name, the sub-configurations that are
associated with the family, and the kernel sets needed by those
sub-configurations.
- Deprecated all kernel-related information in bli_kernel.h and transferred
what remains to new header files named "bli_arch_<configname>.h", which
are conditionally #included from a new header bli_arch.h. These files
are still needed to set library-wide parameters such as custom
malloc()/free() functions or SIMD alignment values.
- Added bli_cntx_init_<configname>.c files to each configuration directory.
The files contain a function, named the same as the file, that initializes
a "native" context for a particular configuration (microarchitecture). The
idea is that optimized kernels, if available, will be initialized into
these contexts. Other fields will retain pointers to reference functions,
which will be compiled on a per-configuration basis. These bli_cntx_init_*()
functions will be called during the initialization of the global kernel
structure. They are thought of as initializing for "native" execution, but
they also form the basis for contexts that use induced methods. These
functions are prototyped, along with their _ref() and _ind() brethren, by
prototype-generating macros in bli_arch.h.
- Added a new typedef enum in bli_type_defs.h to define an arch_t, which
identifies the various sub-configurations.
- Redesigned the global kernel structure (gks) around a 2D array of cntx_t
structures (pointers to cntx_t, actually). The first dimension is indexed
over arch_t and the inner dimension is the ind_t (induced method) for
each microarchitecture. When a microarchitecture (configuration) is
"registered" at init-time, the inner array for that configuration in the
2D array is initialized (and allocated, if it hasn't been already). The
cntx_t slot for BLIS_NAT is initialized immediately and those for other
induced method types are initialized and cached on-demand, as needed. At
cntx_t registration, we also store function pointers to cntx_init functions
that will initialize (a) "reference" contexts and (b) contexts for use with
induced methods. We don't cache the full contexts for reference contexts
since they are rarely needed. The functions that initialize these two kinds
of contexts are generated automatically for each targeted sub-configuration
from cpp-templatized code at compile-time. Induced method contexts that
need "stage" adjustments can still obtain them via functions in
bli_cntx_ind_stage.c.
- Added new functions and functionality to bli_cntx.c, such as for setting
the level-1f, level-1v, and packm kernels, and for converting a native
context into one for executing an induced method.
- Moved the checking of register/cache blocksize consistency from being cpp
macros in bli_kernel_macro_defs.h to being runtime checks defined in
bli_check.c and called from bli_gks_register_cntx() at the time that the
global kernel structure's internal context is initialized for a given
microarchitecture/configuration.
- Deprecated all of the old per-operation bli_*_cntx.c files and removed
the previous operation-level cntx_t_init()/_finalize() invocations.
Instead, we now query the gks for a suitable context, usually via
bli_gks_query_cntx().
- Deprecated support for the 3m2 and 3m3 induced methods. (They required
hackery that I was no longer willing to support.)
- Consolidated the 1e and 1r packm kernels for any given register blocksize
into a single kernel that will branch on the schema and support packing
to both formats.
- Added the cntx_t* argument to all packm kernel signatures.
- Deprecated the local function pointer array in all bli_packm_cxk*.c files
and instead obtain the packm kernel from the cntx_t.
- Added bli_calloc_intl(), which serves as the calloc-equivalent to to
bli_malloc_intl(). Useful when we wish to allocate and initialize to
zero/NULL.
- Converted existing cpp macro functions defined in bli_blksz.h, bli_func.h,
bli_cntx.h into static functions.
Details:
- Fixed a bug in gemmtrsm test module that was due to improper partitioning
into a k x k triangular matrix for the purposes of obtaining an mr x k
micropanel of A with which to test.
- Fixed a bug in gemm and gemmtrsm test modules that would only manifest for
very large k (depending on the product of mr x kc on that architecture).
The bug arose from the fact that the test module was triggering the
allocation of blocks from the internal memory pools, which are limited in
size. This allocation imposes an implicit assumption that the micro-
panel being tested with will fit inside, and this assumption is violated
for large values of k. Arbitrarily large k may now be tested for both
operation tests.
- Added OpenMP/pthread critical sections around the setting or getting of
statuses from the induced method operation lookup table in bli_l3_ind.c.
- Added the 'static' keyword to all pthread_mutex_t global variables in BLIS.
- Thanks to Nisanth Padinharepatt of AMD for reporting the first and third
issues.
Details:
- Removed trailing commas from enums in bli_type_defs.h. Thanks to
Erling Andersen for pointing out this inconsistency and suggesting
the change.
Details:
- Added explicit handling of situations where i == dim to
bli_determine_blocksize_b_sub(). This isn't actually needed by any
current use case within BLIS, but handling the situation is nonetheless
prudent. Thanks to Minh Quan for reporting this issue and requesting
the fix.
Details:
- Fixed a bug in bli_l3_packm() that caused cntl_t-cached packed mem_t
entries to be released and then re-acquired unnecessarily. (In essence,
the "<" operands in the conditional that guards the
release-and-reacquire code block simply needed to be swapped.) The bug
should have only affected performance (rather than the computed result).
Thanks to Minh Quan for identifying and reporting the bug.
Details:
- Removed the family field inside the cntx_t struct and re-added it to the
cntl_t struct. Updated all accessor functions/macros accordingly, as well
as all consumers and intermediaries of the family parameter (such as
bli_l3_thread_decorator(), bli_l3_direct(), and bli_l3_prune_*()). This
change was motivated by the desire to keep the context limited, as much
as possible, to information about the computing environment. (The family
field, by contrast, is a descriptor about the operation being executed.)
- Added additional functions to bli_blksz_*() API.
- Added additional functions to bli_cntx_*() API.
- Minor updates to bli_func.c, bli_mbool.c.
- Removed 'obj' from bli_blksz_*() API names.
- Removed 'obj' from bli_cntx_*() API names.
- Removed 'obj' from bli_cntl_*(), bli_*_cntl_*() API names. Renamed routines
that operate only on a single struct to contain the "_node" suffix to
differentiate with those routines that operate on the entire tree.
- Added enums for packm and unpackm kernels to bli_type_defs.h.
- Removed BLIS_1F and BLIS_VF from bszid_t definition in bli_type_defs.h.
They weren't being used and probably never will be.
Details:
- Dropped 'u' from the list of modifiers passed into the library archiver
ar. Previously, "cru" was used, while now we employ only "cr". This
change was prompted by a warning observed on Ubuntu 16.04:
ar: `u' modifier ignored since `D' is the default (see `U')
This caused me to realize that the default mode causes timestamps to be
zero, and thus the 'u' option, which causes only changed object files to
be inserted, is not applicable.
Details:
- Added an option to configure that allows the user to force an arbitrary
version string at configure-time. The help text also now describes the
usage information.
- Changed the way the version string is communicated to the Makefile.
Previously, it was read into the VERSION variable from the 'version' file
via $(shell cat ...). Now, the VERSION variable is instead set in
config.mk (via a configure-substituted anchor from config.mk.in).
Details:
- Updated the non-tree openmp and pthreads barriers defined in
bli_thrcomm_openmp.c and bli_thrcomm_pthreads.c to instead call a common
implementation in bli_thrcomm.c, bli_thrcomm_barrier_atomic(). This new
implementation goes through the same motions as the previous codes, but
protects its loads and increments with GNU atomic built-ins. These atomic
statements take memory ordering parameters that allow us to specify just
enough constraints for the barrier to work as intended on weakly-ordered
hardware. The prior implementation was only guaranteed to work on systems
with strongly- ordered memory. (Thanks to Devin Matthews for suggesting
this change and his crash-course in atomics and memory ordering.)
- Removed 'volatile' from structs' barrier field declarations in
bli_thrcomm_*.h.
- Updated bli_thrcomm_pthread.? files to use renamed struct barrier fields
consistent with that of the _openmp.? files.
- Updated other bli_thrcomm_* files to rename "communicator" variables to
simply "comm".
Details:
- Renamed bli_env_get_nway() -> bli_thread_get_env().
- Added bli_thread_set_env() to allow setting environment variables
pertaining to multithreading, such as BLIS_JC_NT or BLIS_NUM_THREADS.
- Added the following convenience wrapper routines:
bli_thread_get_jc_nt()
bli_thread_get_ic_nt()
bli_thread_get_jr_nt()
bli_thread_get_ir_nt()
bli_thread_get_num_threads()
bli_thread_set_jc_nt()
bli_thread_set_ic_nt()
bli_thread_set_jr_nt()
bli_thread_set_ir_nt()
bli_thread_set_num_threads()
- Added #include "errno.h" to bli_system.h.
- This commit addresses issue #140.
- Thanks to Chris Goodyer for inspiring these updates.
