Details:
- Consolidated most of the code relating to induced complex methods
(e.g. 4mh, 4m1, 3mh, 3m1, etc.) into frame/ind. Induced methods
are now enabled on a per-operation basis. The current "available"
(enabled and implemented) implementation can then be queried on
an operation basis. Micro-kernel func_t objects as well as blksz_t
objects can also be queried in a similar maner.
- Redefined several micro-kernel and operation-related functions in
bli_info_*() API, in accordance with above changes.
- Added mr and nr fields to blksz_t object, which point to the mr
and nr blksz_t objects for each cache blocksize (and are NULL for
register blocksizes). Renamed the sub-blocksize field "sub" to
"mult" since it is really expressing a blocksize multiple.
- Updated bli_*_determine_kc_[fb]() for gemm/hemm/symm, trmm, and
trsm to correctly query mr and nr (for purposes of nudging kc).
- Introduced an enumerated opid_t in bli_type_defs.h that uniquely
identifies an operation. For now, only level-3 id values are defined,
along with a generic, catch-all BLIS_NOID value.
- Reworked testsuite so that all induced methods that are enabled
are tested (one at a time) rather than only testing the first
available method.
- Reformated summary at the beginning of testsuite output so that
blocksize and micro-kernel info is shown for each induced method
that was requested (as well as native execution).
- Reduced the number of columns needed to display non-matlab
testsuite output (from approx. 90 to 80).
Details:
- Fixed a bug that affects all level-2 and level-3 blocked variants. The
bug only manifested, however, if the packing of operands (A and B in
gemm, for example) spanned multiple nodes in the control tree. Until
recently, the main consumers of packm were level-3 operations, all of
which packed both input operands from blocked variant 1 (B outside of
the loop, and A within the loop). This particular usage masked a flaw
in the code whereby bli_obj_release_pack() would always release the
underlying mem_t buffer (provided it was allocated), even if the buffer
was not allocated in the current variant. This has been fixed by
replacing all calls to bli_obj_release_pack() with calls to a new
function, bli_packm_release(), which takes the same control tree node
argument passed into the object's corresponding call to packm_init()
or packv_init(). bli_packm_release() then proceeds to invoke
bli_obj_release_pack() only if the control tree node indicates that
packing was requested. Thanks to Devangi Parikh for identifying this
bug.
Details:
- Renamed all remaining 3m/4m packing files and symbols to 3mi/4mi
('i' for "interleaved"). Similar changes to 3M/4M macros.
- Renamed all 3m/4m files and functions to 3m1/4m1.
- Whitespace changes.
Details:
- Relocated bli_4mh.c, bli_4mb.c, bli_4m.c, bli_3mh.c, bli_3m.c (and
associated headers) from frame/base to frame/base/induced.
- Added bli_xm.? to frame/base/induced, which implements
bli_xm_is_enabled(), which detects whether ANY induced complex method
is currently enabled.
- The new function bli_xm_is_enabled() is now used in bli_info.c to
detect when an induced complex method is used, so we know when to
return blocksizes from one of the induced methods' blocksize objects.
Details:
- Changed the static memory definitions to compute the maximum register
blocksize for each datatype and add it to kc when computing the size
of blocks of A and B. This formally accounts for the nudging of kc
up to a multiple of mr or nr at runtime for triangular operations
(e.g. trmm).
Details:
- Renamed several variables and macros from 3m/4m to 3mi/4mi. This is
because those packing schemas were always implicitly "interleaved".
This new naming scheme will make way for new schemas that separate
instead of interleve the real and imaginary (and summed) parts.
- Expanded the pack format sub-field of the pack schema field of the
info_t to 4 bits (from 3). This will allow for more schema types
going forward.
- Removed old _cntl.c files for herk3m, herk4m, trmm3m, trmm4m.
Details:
- Fixed a bug that only affected trmm3 when performed via 3mh or 4mh,
whereby micro-panels of the triangular matrix were packed with "dead
space" between them due to failing to adjust for the fact that pointer
arithmetic was occurring in units of complex elements while the data
being packed consisted of real elements. It turns out that the macro-
kernel suffered from the same bug, meaning the panels were actually
being packed and read consistently. The only way I was able to
discover the bug in the first place was because the packed block of A
was overflowing into the beginning of the packed row panel of B using
the sandybridge configuration.
Details:
- Modified bli_packm_blk_var1.c and _var2.c to increase the triangular
case's panel increment by 1 if it would otherwise be odd. This is
particularly necessary in _var2.c when handling the interleaved 3m
or ro/io/rpi pack schemas, since division of an odd number by 2 can
happen if both the panel length and the panel packing dimension
(register packing blocksize) are odd, thus making their product odd.
- Modified bli_packm_init.c so that panel strides are increased by 1
if they would otherwise be odd, even for non-3m related packing.
- Modified the trmm and trsm macro-kernels so that triangular packed
micro-panels are traversed with this new "increment by 1 if odd"
policy.
- Added sanity checks in trmm and trsm macro-kernels that would result
in an abort() if the conditions that would lead to a "divide odd
integer by 2" scenario ever manifest.
- Defined bli_is_odd(), _is_even() macros in bli_scalar_macro_defs.h.
Details:
- Added an imaginary stride field ("is") to obj_t.
- Renamed bli_obj_set_incs() macro to bli_obj_set_strides().
- Defined bli_obj_imag_stride() and bli_obj_set_imag_stride() and
added invocations in key locations.
- Added some basic error-checking related to imaginary stride.
- For now, imaginary stride will not be exposed into the most-used
BLIS APIs such as bli_obj_create(), and certainly not the
computational APIs such as bli_dgemm().
Details:
- Added a new function, bli_gemm_cntl_ukrs(), that returns the func_t*
for the gemm micro-kernels from the leaf node of the control tree.
This allows all the func_t* fields from higher-level nodes in the tree
to be NULL, which makes the function that builds the control trees
slightly easier to read.
- Call bli_gemm_cntl_ukrs() instead of the cntl_gemm_ukrs() macro in
all bli_*_front() functions (which is needed to apply the row/column
preference optimization).
- In all level-3 bli_*_cntl_init() functions, changed the _obj_create()
function arguments corresponding to the gemm_ukrs fields in higher-
level cntl tree nodes to NULL.
- Removed some old her2k macro-kernels.
Details:
- Added optional flop counting to all level-3 front-ends, which is
enabled via BLIS_ENABLE_FLOP_COUNT. The flop count can be
reset at any time via bli_flop_count_reset() and queried via
bli_flop_count(). Caveats:
- flop counts are approximate for her[2]k, syr[2]k, trmm, and
trsm operations;
- flop counts ignore extra flops due to non-unit alpha;
- flop counts do not account for situations where beta is zero.
Details:
- Implemented a new function, bli_obj_imag_equals(), which compares the
imaginary part of the first argument to the second argument, which may
be a BLIS_CONSTANT or of a regular real datatype.
Details:
- Changed gemm_kc blocksizes to be reduced by two-thirds instead of
half.
- Changed 3m4m/test_gemm.c driver to divide by 3 instead of 2 when
computing the fixed k dimension.
- Fixed runme.sh so that it would use multiple threads for s/dgemm
cases.
Details:
- Added yet another 4m-based implementation for complex domain level-3
operations. This method, which the 3m/4m paper identifies as Algorithm
"4m_1b" fissures the first loop around the micro-kernel so that the
real sub-panel of the current micro-panel of B is multiplied against
(both sub-panels of) all micro-panels of A, before doing the same for
the imaginary sub-panel of the micro-panel of B. For now, only gemm is
supported, and 4m_1b (labeled "4mb" within the framework) is not yet
integrated into the test suite.
Details:
- Fixed a bug in certain structured level-2 operations that manifested
only when the structured matrix was provided to BLIS as matrix stored
with general stride. The bug was introduced in c472993b when the
densify field was removed from the packm control tree node and
associated APIs. Since then, the packed object was unconditionally
marked with an uplo field of BLIS_DENSE. This is fine for level-3
operations where micro-panels are always densified, but in level-2
contexts, the underlying unblocked variant (fused or unfused) of
structured operations (e.g. trmv) still needs to know whether to
execute its "lower" or "upper" branches of code. Since this field
was unconditionally being set to BLIS_DENSE, the unblocked variants
were always executed the "else" branch, which happened to be the
"lower" case code. Thus, running an upper case produced the wrong
answer. This most obviously manifested in the form of failures for
trmm, trmm3, and trsm in the test suite.
The bug was fixed by setting the packed object's uplo field to
BLIS_DENSE only if the schema indicated that micro-panels were to be
packed. Otherwise, we can assume we are packing to regular row or
column storage, as is the case with level-2 packing. Thanks to
Francisco Igual for reporting the testsuite failures and ultimately
leading us to this bug.
Lead thread of the inner thread communicator could exit subproblem, move on the next iteration of the loop and modify a1_pack, b1_pack, or c1_pack while other threads were still using those.
Barriers were inserted to fix this.
Details:
- No longer need to define BLIS_ENABLE_MULTITHREADING manually in
bli_config.h; it now gets defined when BLIS_ENABLE_OPENMP or
BLIS_ENABLE_PTHREADS is defined.
- Added sanity check to prevent both BLIS__ENABLE_OPENMP and
BLIS_ENABLE_PTHREADS from being enabled simultaneously.
- Reorganization of bli_threading*.h header files, which led to
simplification of threading-related part of blis.h.
- added "-fopenmp -lpthread" to LDFLAGS of sandybridge make_defs.mk
file.
Details:
- Since 4674ca8c, the constraint that KC be a multiple of both MR and
NR have been relaxed, and thus it was time to remove the comments
from the top of the bli_kernel.h files of all configurations.
Details:
- Added new micro-kernels for the AMD piledriver architecture (one
for each datatype).
- Updates and tweaks to piledriver configuration.
- Added 3xk packm micro-kernel support.
- Explicitly unrolled some of the smaller packm micro-kernels.
- Added notes to avx/sandybridge and piledriver micro-kernel files
acknowledging the influence of the corresponding kernel code in
OpenBLAS.
Details:
- Changed the recently-added micro-panel alignment macros so that we now
have two sets--one for micro-panels of matrix A and one for micro-
panels of matrix B: BLIS_UPANEL_[AB]_ALIGN_SIZE_?.
- Store each set of alignment values into a separate blksz_t object in
bli_gemm_cntl_init().
- Adjusted packm_init() to use the separate alignment values.
- Added query routines for the new alignment values to bli_info.c.
- Modified test suite output accordingly.
Details:
- Split bli_determine_blocksize_[fb]() into two functions each, the
newer ones ending with the _sub suffix. These new sub-functions are
now called from bli_[gemm|trmm|trsm]_determine_kc_[fb](), which
eliminates redundant code and will allow any future tweaks to the
core sub-functions to automatically be inherited by the operation-
specific versions.
Details:
- These changes were intended for the previous commit.
- Defined bli_gemm_determine_kc_[fb]() and bli_gemm_determine_kc_[fb](),
which determine blocksizes for gemm-based operations, taking special
care to "nudge" the kc dimension up to a multiple of MR or NR for
hemm and symm operations, as needed.
- Changed bli_gemm_blk_var3f.c to call bli_gemm_determine_kc_f().
instead of bli_determine_blocksize_f().
- Comment updates to bli_trmm_blocksize.c, bli_trsm_blocksize.c.
Details:
- Relaxed a long-held requirement in register blocksizes that required
the kernel programmer to choose a KC that was divisible by both MR
and NR. This was very constraining on some architectures that did not
use register blocksizes that were powers of two. The constraint is
now enforced only for trmm and trsm, where it is needed, and it is
now handled by "nudging" kc upward at runtime, if necessary, to be a
multiple of MR or NR, as needed.
- Defined bli_trmm_determine_kc_[fb]() and bli_trsm_determine_kc_[fb](),
which determine blocksizes for trmm and trsm, taking special care to
"nudge" the kc dimension up to a multiple of MR or NR, as needed.
- Changed bli_trmm_blk_var3[fb].c to call bli_trmm_determine_kc_[fb]()
instead of bli_determine_blocksize_[fb]().
- Added safeguard to bli_align_dim_to_mult() that returns the dimension
unmodified if the dimension multiple is zero (to avoid division by
zero).
- Removed cpp guard/check for KC % MR == 0 and KC % NR == 0 from
bli_kernel_macro_defs.h.
- Whitespace, variable name changes to bli_blocksize.c.
- Removed old commented code from bli_gemm_cntl.c.
Details:
- This commit re-implements a feature that was removed in commit
c2b2ab62. It was removed because, at the time, I wasn't sure how the
micro-panel alignment feature would interact with the 4m method (when
applied at the micro-kernrel level), and so it seemed safer to disable
the feature entirely rather than allow possible breakage. This commit
revisits the issue and safely re-implements the feature in a way that
is compatible with 4m, 3m, 4mh, and 3mh (and native execution).
- Modified the static memory pool to account for micro-panel alignment
space.
- Modified packm_init and blocked variants to align whole micro-panels
by a datatype-specific alignment value that may be set by the
configuration. (If it is not set by the configuration, it will default
to BLIS_SIZEOF_?.)
- Modified macro-kernels so that:
- storage stride is handled properly given the new micro-panel
alignment behavior;
- indexing through 3m/4m/rih-type sub-panels, as is done by trmm and
trsm, is more robust (e.g. will work if the applicable packing
register blocksize is odd);
- imaginary strides are computed and stored within auxinfo_t structs,
which allows the virtual micro-kernels to more easily determine how
to index into the micro-panel operands.
- Modified virtual 3m and 4m micro-kernels to use the imaginary strides
within the auxinfo_t structs instead of panel strides.
- Deprecated the panel stride fields from the auxinfo_t structs.
- Updated test suite to print out the micro-panel alignment values.
Details:
- Added a modified test driver for [cz]gemm that will test all 3m/4m
as well as assembly-based and OpenBLAS implementations of gemm
in single and multithreaded modes.
Details:
- Redefined bli_is_last_iter() to take thread_id and num_thread
arguments, which allows the macro to correctly compute whether a
given iteration is the last that the thread will compute in that
particular loop. The new definition, however, remains disabled
(commented out) until someone can look at this more closely, as
the new definition seems to actually hurt performance slightly.
- Whitespace and related updates to level-3 macro-kernels.
- Updated test suite so that performance results in the hundreds of
gigaflops does not disrupt the column alignment of the output.
