mirror of
https://github.com/amd/blis.git
synced 2026-04-20 15:48:50 +00:00
5fec95b99f
Details: - Implemented support for gemm where A, B, and C may have different storage datatypes, as well as a computational precision (and implied computation domain) that may be different from the storage precision of either A or B. This results in 128 different combinations, all which are implemented within this commit. (For now, the mixed-datatype functionality is only supported via the object API.) If desired, the mixed-datatype support may be disabled at configure-time. - Added a memory-intensive optimization to certain mixed-datatype cases that requires a single m-by-n matrix be allocated (temporarily) per call to gemm. This optimization aims to avoid the overhead involved in repeatedly updating C with general stride, or updating C after a typecast from the computation precision. This memory optimization may be disabled at configure-time (provided that the mixed-datatype support is enabled in the first place). - Added support for testing mixed-datatype combinations to testsuite. The user may test gemm with mixed domains, precisions, both, or neither. - Added a standalone test driver directory for building and running mixed-datatype performance experiments. - Defined a new variation of castm, castnzm, which operates like castm except that imaginary values are not touched when casting a real operand to a complex operand. (By contrast, in these situations castm sets the imaginary components of the destination matrix to zero.) - Defined bli_obj_imag_is_zero() and substituted calls in lieu of all usages of bli_obj_imag_equals() that tested against BLIS_ZERO, and also simplified the implementation of bli_obj_imag_equals(). - Fixed bad behavior from bli_obj_is_real() and bli_obj_is_complex() when given BLIS_CONSTANT objects. - Disabled dt_on_output field in auxinfo_t structure as well as all accessor functions. Also commented out all usage of accessor functions within macrokernels. (Typecasting in the microkernel is still feasible, though probably unrealistic for now given the additional complexity required.) - Use void function pointer type (instead of void*) for storing function pointers in bli_l0_fpa.c. - Added documentation for using gemm with mixed datatypes in docs/MixedDatatypes.md and example code in examples/oapi/11gemm_md.c. - Defined level-1d operation xpbyd and level-1m operation xpbym. - Added xpbym test module to testsuite. - Updated frame/include/bli_x86_asm_macros.h with additional macros (courtsey of Devin Matthews).
50 lines
2.3 KiB
Plaintext
50 lines
2.3 KiB
Plaintext
# ----------------------------------------------------------------------
|
|
#
|
|
# input.general.fast
|
|
# BLIS test suite
|
|
#
|
|
# This file contains input values that control how BLIS operations are
|
|
# tested. Comments explain the purpose of each parameter as well as
|
|
# accepted values.
|
|
#
|
|
|
|
1 # Number of repeats per experiment (best result is reported)
|
|
rc # Matrix storage scheme(s) to test:
|
|
# 'c' = col-major storage; 'g' = general stride storage;
|
|
# 'r' = row-major storage
|
|
cj # Vector storage scheme(s) to test:
|
|
# 'c' = colvec / unit stride; 'j' = colvec / non-unit stride;
|
|
# 'r' = rowvec / unit stride; 'i' = rowvec / non-unit stride
|
|
0 # Test all combinations of storage schemes?
|
|
1 # Perform all tests with alignment?
|
|
# '0' = do NOT align buffers/ldims; '1' = align buffers/ldims
|
|
0 # Randomize vectors and matrices using:
|
|
# '0' = real values on [-1,1];
|
|
# '1' = powers of 2 in narrow precision range
|
|
32 # General stride spacing (for cases when testing general stride)
|
|
sdcz # Datatype(s) to test:
|
|
# 's' = single real; 'c' = single complex;
|
|
# 'd' = double real; 'z' = double complex
|
|
0 # Test gemm with mixed-domain operands?
|
|
0 # Test gemm with mixed-precision operands?
|
|
100 # Problem size: first to test
|
|
100 # Problem size: maximum to test
|
|
100 # Problem size: increment between experiments
|
|
# Complex level-3 implementations to test:
|
|
0 # 3mh ('1' = enable; '0' = disable)
|
|
0 # 3m1 ('1' = enable; '0' = disable)
|
|
0 # 4mh ('1' = enable; '0' = disable)
|
|
0 # 4m1b ('1' = enable; '0' = disable)
|
|
0 # 4m1a ('1' = enable; '0' = disable)
|
|
0 # 1m ('1' = enable; '0' = disable)
|
|
1 # native ('1' = enable; '0' = disable)
|
|
1 # Simulate application-level threading:
|
|
# '1' = disable / use one testsuite thread;
|
|
# 'n' = enable and use n testsuite threads
|
|
1 # Error-checking level:
|
|
# '0' = disable error checking; '1' = full error checking
|
|
i # Reaction to test failure:
|
|
# 'i' = ignore; 's' = sleep() and continue; 'a' = abort
|
|
0 # Output results in matlab/octave format? ('1' = yes; '0' = no)
|
|
0 # Output results to stdout AND files? ('1' = yes; '0' = no)
|