Details:
- Expanded the BLIS_PACK_SCHEMA_BITS value in bli_type_defs.h to
include all six bits presently used in the pack schema bitfield of
the info field of obj_t structs. Prior to this commit, the macro
constant only included the lowest five bits, which excluded the
"is or is not packed" bit. This manifested as a strange bug in
probably many level-2 codes that invoked packing, though we only
observed it in ger before fixing. Thanks to Devin Matthews for
finding and reporting this bug.
Details:
- Updated bli_kernel_*_macro_defs.h headers to include default
definitions for 30xk packm kernels.
- Extended function pointer arrays in bli_packm_cxk_*() out to 31 and
included 30xk kernels.
- Addex 30xk kernels to frame/1m/packm/ukernels/bli_packm_ref_cxk_*.c.
Details:
- Applied a patch that allows the top-level Makefile to work on certain
systems. The patch simply separates out the source-to-object code
generation rules for .c and .S files into two separate rules. Thanks
to Devin Matthews for submitting this patch.
Details:
- We actually need to check alignment of lda*sizeof(double) and NOT
a+lda because in the latter case, alignment could cancel out and
still allow the optimized code to run when it shouldn't. Thanks
to Devin for pointing this out.
Details:
- The bugfix in a68b316c was inadvertantly checkin alignment of the
leading dimension itself, rather than the byte size of the leading
dimension. Now, we simply check alignment of a+lda.
Details:
- Added _string() suffix to bli_info_get_*_ukr_type() function names.
This makes them consistent with the bli_info_get_*_impl_string()
functions.
Details:
- Fixed bugs whereby the level-1f dotxf, axpyxf, and dotxaxpyf kernels
were attempting to compute problems with unaligned leading dimensions
with optimized code, rather than (correctly) using the reference
implementations. Thanks to Devin Matthews for reporting this bug.
Details:
- Added "4mh" and "3mh" APIs, which implement the 4m and 3m methods at
high levels, respectively. APIs for trmm and trsm were NOT added due
to the fact that these approaches are inherently incompatible with
implementing 4m or 3m at high levels (because the input right-hand
side matrix is overwritten).
- Added 4mh, 3mh virtual micro-kernels, and updated the existing 4m and
3m so that all are stylistically consistent.
- Added new "rih" packing kernels (both low-level and structure-aware)
to support both 4mh and 3mh.
- Defined new pack_t schemas to support real-only, imaginary-only, and
real+imaginary packing formats.
- Added various level0 scalar macros to support the rih packm kernels.
- Minor tweaks to trmm macro-kernels to facilitate 4mh and 3mh.
- Added the ability to enable/disable 4mh, 3m, and 3mh, and adjusted
level-3 front-ends to check enabledness of 3mh, 3m, 4mh, and 4m (in
that order) and execute the first one that is enabled, or the native
implementation if none are enabled.
- Added implementation query functions for each level-3 operation so
that the user can query a string that describes the implementation
that is currently enabled.
- Updated test suite to output implementation types for reach level-3
operation, as well as micro-kernel types for each of the five micro-
kernels.
- Renamed BLIS_ENABLE_?COMPLEX_VIA_4M macros to _ENABLE_VIRTUAL_?COMPLEX.
- Fixed an obscure bug when packing Hermitian matrices (regular packing
type) whereby the diagonal elements of the packed micro-panels could
get tainted if the source matrix's imaginary diagonal part contained
garbage.
Details:
- Modified macro-kernels to pass the pack_t schema values for matrices
A and B into the datatype-specific functions, where they are now
inserted into a newly-expanded auxinfo_t struct. This gives gives the
micro-kernels access to the pack_t schema values embedded in the
control trees, which determine the precise format into which the
matrix elements are packed.
- Updated a call to bli_packm_init_pack() in src/test_libblis.c to
remove densify argument. Meant to include this in commit c472993b.
Details:
- Removed the "densify" bool_t argument to bli_packm_cntl_obj_create().
This argument was inserted very early in BLIS's development, when it
was anticipated that the developer may sometimes wish to pack a
Hermitian, symmetric, or triangular matrix without making it dense.
But as it turns out, if we are packing a matrix, we always want to
make it dense in some way or another due to the fact that the micro-
kernel only multiplies dense micro-panels. Thus, unless/until there
is a real need for the feature, it seems reasonable to remove it from
the packm_cntl API.
Details:
- Replaced all trmm_t control tree instances and usage with that of
gemm_t. This change is similar to the recent retirement of the herk_t
control tree.
- Tweaked packm blocked variants so that the triangular code does NOT
assume that k is a multiple of MR (when A is triangular) or NR (when
B is triangular). This means that bottom-right micro-panels packed for
trmm will have different zero-padding when k is not already a multiple
of the relevant register blocksize. While this creates a seemingly
arbitrary and unnecessary distinction between trmm and trsm packing,
it actually allows trmm to be handled with one control tree, instead
of one for left and one for right side cases. Furthermore, since only
one tree is required, it can now be handled by the gemm tree, and thus
the trmm control tree definitions can be disposed of entirely.
- Tweaked trmm macro-kernels so that they do NOT inflate k up to a
multiple of MR (when A is triangular) or NR (when B is triangular).
- Misc. tweaks and cleanups to bli_packm_struc_cxk_4m.c and _3m.c, some
of which are to facilitate above-mentioned changes whereby k is no
longer required to be a multiple of register blocksize when packing
triangular micro-panels.
- Adjusted trmm3 according to above changes.
- Retired trmm_t control tree creation/initialization functions.
Details:
- Replaced all herk_t control tree instances and usage with that of
gemm_t, since the two types presently have the same fields. This means
that herk, her2k, syrk, and syr2k can simply use the gemm control tree
as-is, just as hemm and symm have been doing for some time now.
- Retired herk_t control tree creation/initialization functions.
- Retired many _target.c and .h files into 'old' directories.
Details:
- Realized that we don't need to track rs_p11 and cs_p11 for
Hermitian/symmetric case of bli_packm_struc_cxk*(). They are always
equal to rs_p and cs_p.
Details:
- Changed m and n dimension parameter names to panel_dim and panel_len,
respectively, in packm_cxk, packm_cxk_3m, packm_cxk_4m kernel wrapper
functions. This makes the code a little easier to read since "m" and
"n" have connotations that are not applicable here.
- Comment updates.
Details:
- Removed sections of bli_kernel_[4m|3m]_macro_defs.h that defined
4m/3m-specific blocksizes after realizing that this can be done in
bli_gemm[4m|3m]_cntl.c, since that is (mostly) the only place they
are used.
- The maximum cache values for 4m/3m are stll needed when computing mem
pool dimensions in bli_mem_pool_macro_defs.h. As a workaround, "local"
definitions in terms of the regular cache blocksizes are now in place.
- Similarly, the register blocksizes for 4m/3m are still needed in
bli_kernel_post_macro_defs.h. As a workaround, "local" definitions in
terms of the regular register blocksizes are now in place.
Details:
- Changed semantics of cache and register blocksize extensions so that
the extended values are tracked, rather than just the marginal
extensions.
- BLIS_EXTEND_[MKN]C_? has been renamed BLIS_MAXIMUM_[MKN]C_?.
- BLIS_EXTEND_[MKN]R_? has been renamed BLIS_PACKDIM_[MKN]R_?.
- bli_blksz_ext_*() APIs have been renamed to bli_blksz_max_*(). Note
that these "max" query routines grab the maximum value for cache
blocksizes and the packdim value for register blocksizes.
- bli_info_*() API has been updated accordingly.
- All configurations have been updated accordingly.
Details:
- Changed the interface to the packm_struc_cxk*() kernels to include
the pack_t schema. This allows the implementation to more easily
determine how the micro-panel is stored (row-stored column panel
or column-stored row panel).
- Updated packm blocked variants to pass in the schema.
- Updated packm_ker_t function pointer definition accordingly.
Details:
- Reorganized packm variants and structure-aware kernels so that all
routines for a given pack format (4m, 3m, regular) reside in a single
file.
- Renamed _blk_var4 to _blk_var2 and generalized so that it will work
for
both 4m and 3m, and adjusted 4m/3m _cntl_init() functions accordingly.
- Added a new packm_ker_t function pointer type to
bli_kernel_type_defs.h
to facilitate function pointer typecasting in the datatype-specific
packm_blk_var2() functions.
- Deprecated _blk_var3.
- Fixed a bug in the triangular micro-panel packing facility that
affected trmm and trmm3 with unit diagonals.
Details:
- Reordered the #include statements in bli_scalar_macro_defs.h so that
conventional, ri-, and ri3-based macros are grouped together.
- Renamed bli_eqri.h (and macros within) to end with 'ris' suffix.
Details:
- Added 'const' qualifier to inlined packing code that handles
micro-panel packing that is too large for an existing packm ukernel.
- Comment updates.
Details:
- Rather than passing the packm blocked routines a boolean value that
represents whether the matrix is being packed to row or column storage,
we now pass in the pack schema itself.
Details:
- The packm implementation (i.e. sources files in frame/1m/packm and
frame/1m/packm/ukernels), interchangeably used the names "beta" and
"kappa" to refer to the optional scalar to be applied during packing.
This commit renames all uses of "beta" to be "kappa", since "beta"
sometimes evokes the scalar specifically on the output matrix of a
level-2 or level-3 operation.
Details:
- Reorganized unpackm ukernels into a single file,
bli_unpackm_ref_cxk.c, in a manner similar to what was done for packm
ukernels in commit 4cc2b46.
Details:
- Added extra spaces to align arguments of
bli_obj_scalar_init_detached_copy_of(). This misalignment was due to
the fact that the function was previously named
bli_obj_init_scalar_copy_of() and the name change, performed in
b444489f, was done via recursive sed commands which left subsequent
lines untouched.
Details:
- Combined the 4m/3m bits into an expanded bitfield, which will encode
the packing "format" of the micro-panels. This will allow for more
easily and compactly encoding additional formats.
- Other minor comment/whitespace updates to bli_type_defs.h.
- Updated bli_obj_macro_defs.h and bli_param_macro_defs.h to use the new
format bitfield.
- Comment update to bli_kernel_post_macro_defs.h.
- Whitespace changes to bli_kernel_3m_macro_defs.h, _4m_macro_defs.h.
Details:
- Renamed packm ukernels, _cxk dispatcher, and structure-aware _cxk
helper functions to use _4m and _3m instead of _ri and _ri3 suffixes.
- Updated names of cpp macros that correspond to packm ukernels.
Details:
- Added an extra layer to level-3 front-ends (examples: bli_gemm_entry()
and bli_gemm4m_entry()) to hide the control trees from the code that
decides whether to execute native or 4m-based implementations. The
layering was also applied to 3m.
- Branch to 4m code based on the return value of bli_4m_is_enabled(),
rather than the cpp macros BLIS_ENABLE_?COMPLEX_VIA_4M. This lays
the groundwork for users to be able to change at runtime which
implementation is called by the main front-ends (e.g. bli_gemm()).
- Retired some experimental gemm code that hadn't been touched in
months.
Details:
- Added bli_4m.c (and header), which defines a simple API that can be
used to query, enable, and disable 4m-based complex support in BLIS.
The macros BLIS_ENABLE_?COMPLEX_VIA_4M are now used to initialize
the variable that determines the state (enabled or disabled).
- Changed bli_info*() API so that all cache and register blocksize-
related query routines return the blksz_t objects' values as they
exist at runtime, rather than return the values as determined by the
configuration system (e.g. bli_kernel.h, or defaults for those values
not specified). This sets the foundation for being able to change
those blocksizes at runtime.
Details:
- Changed sandybridge MC and KC for single-precision real to 128 and 384,
respectively.
- Updated comments in template configuration's gemm micro-kernel file
to document the new "contiguous row preference" macro.
Details:
- Added the ability for the kernel developer to indicate the gemm micro-
kernel as having a preference for accessing the micro-tile of C via
contiguous rows (as opposed to contiguous columns). This property may
be encoded in bli_kernel.h as BLIS_?GEMM_UKERNEL_PREFERS_CONTIG_ROWS,
which may be defined or left undefined. Leaving it undefined leads to
the default assumption of column preference.
- Changed conditionals in frame/3/*/*_front.c that induce transposition
of the operation so that the transposition is induced only if there
is disagreement between the storage of C and the preference of the
micro-kernel. Previously, the only conditional that needed to be met
was that C was row-stored, which is to say that we assumed the micro-
kernel preferred column-contiguous access on C.
- Added a "prefers_contig_rows" property to func_t objects, and updated
calls to bli_func_obj_create() in _cntl.c files in order to support
the above changes.
- Removed the row-storage optimization from bli_trsm_front.c because
it is actually ineffective. This is because the right-side case of
trsm flips the A and B micro-panel operands (since BLIS only requires
left-side gemmtrsm/trsm kernels), meaning any transposition done
at the high level is then undone at the low level.
- Tweaked trmm, trmm3 _front.c files to eliminate a possible redundant
invocation of the bli_obj_swap() macro.
Details:
- Previously, packm micro-kernels were organized by the implied register
blocksize (panel dimension) assumed by the kernel, meaning conventional,
ri, and ri3 variations of some micro-kernel size were housed in the same
file. This commit reorganizes the micro-kernels so that all sizes reside
in the same file for each format type (conventional, ri, and ri3).
