- Currently the pointer received as function argument is
used for packing which causes only a partial copy of
input buffer to output buffer due to strange optimizations
by compiler.
- To fix this, instead of using a normal pointer for output
buffer, we define a "restrict" local pointer variable.
- "restrict" keyword tells the compiler that the pointer is
the only way to access the object pointed by the pointer.
- By defining "restrict" local pointer pointing to output
buffer, the mysterious problem of incomplete copy has
been solved.
Change-Id: Ie2355beb1d43ff4b60b940dd88c4e2bf6f361646
- In zen4 arch TRSM and GEMM have different blocksizes.
TRSM call will update blockize in global cntx object
which is incorrect for GEMM, when GEMM and TRSM are
called in parallel.
- Hence using a local copy of cntx which holds blocksizes
would help.
AMD-Internal: [CPUPL-3019]
Change-Id: I5f0f5675b3917d2a11d582ac626ca5d8f4752c53
Description:
1. Developed row variant intrinsic Kernels for float32/sgemm
which are called from lpgemm api aocl_gemm_f32f32f32of32()
2. 6x64m, 6x48m, 6x32m kernels and respective fringe kernels are
developed using avx512.
3. 6x16m main kernel and respective n fringe and mn fringe are
are developed based on avx2 and avx
4. Modularizing, K loop unroll, perf tuning, post-ops and dynamic
dispatch are planned next
5. When leading dims are greater than dims bench_lpgemm need
to be updated to test it and this is planned next.
Change-Id: I54c78fef639ea109d6ef2c2b05c07ce396c81370
- Added kernels for all rv and rd variants.
- Main kernel is of size 6x64, and the associated fringe kernels
added are
- 4x64, 2x64, 1x64
- 6x32, 4x32, 2x32, 1x32
- 6x16, 4x16, 2x16, 1x16
- Updated the zen4 config to enable these kernels in zen4 path.
- Added C-prefetching to 6x? row-stored main kernels.
- C-prefetching for column storage yet to be added.
- K-loop unrolling for fringe kernels yet to be added.
AMD-Internal: [CPUPL-3002]
Change-Id: Ide18412cc6178b43a12a3bc7a608ce9d298fb2e4
- Added ZCOPYV kernel that uses AVX2 and SSE instructions for
vectorization.
- The routine returns early when the vector dimension is zero.
- The kernel takes one among the two available paths based on
conjugation requirement of X vector.
- VZEROUPPER is added before transitioning from AVX2 to SEE.
- Added function pointer to ZEN, ZEN 2, ZEN 3 and ZEN 4 contexts.
AMD-Internal: [CPUPL-2773]
Change-Id: Ibd8a2de42060716395ef698d753c8462654cc0f0
- 8x8 kernels are used for DTRSM SMALL
- Matrix A(a10) is packed for GEMM operations.
- Packed martix A will be re-used in all the col-block
along N-dimension.
- Diagonal elements of A matrix are packed(a11) for
TRSM operations.
- Implemented fringe cases with following block sizes
8x8, 8x4, 8x3, 8x2, 8x1
4x8, 4x4, 4x3, 4x2, 4x1
3x8, 3x4, 3x3, 3x2, 3x1
2x8, 2x4, 2x3, 2x2, 2x1
1x8, 1x4, 1x3, 1x2, 1x1
AMD-Internal: [CPUPL-2745]
Change-Id: I6a174e7f88a4c2c5778052525879552a1e82f6ad
-As of now, memcpy is used in u8s8s32 micro-kernel for copying in k
fringe loop (( k % 4 )!= 0) and NR' < 16 fringe kernels. However for
small k/n dimensions, memcpy invocation has high overhead.
-This issue is fixed by replacing memcpy with a MACRO based
implementation of copy routine, specifically optimized for the sizes
that will be encountered in fringe cases (k < 4, NR' < 16).
AMD-Internal: [CPUPL-3008]
Change-Id: I376bab0aac325832e42e370b291614e5fd5272dc
Details:
1. Reference kernel File names are to be mirrored with unique names
for each architecture configuration while generating fat binary.
2. Python script "blis_ref_kernel_mirror.py" is updated to append
the filenames with each configuration (zen, zen2, zen3, zen4
and generic) that is built for windows.
AMD-Internal: [CPUPL-3009]
Change-Id: Ib02206382199cf2aebe14ff9c869b6089228e1c2
- When alpha is equal 1 in scal2v the computation can be reduced to
copyv. The condition check in else if condition has to be a check
for alpha equal to one and not zero.
- BLIS_NO_CONJ has been passed while invoking copy. In case of single
and double complex, the appropriate conjx needs to be passed with
copyv so that the vector can be conjugated before copying to the
destination.
AMD-Internal: [CPUPL-2985]
Change-Id: Ieb1ad111dff68098a08d870da7215bbd4ff5c9d0
- Added ZSCALV that uses AVX2 and SSE instructions for vectorization.
- Return early when the vector dimension is zero. When alpha is 1 there
is no need to perform computation hence return early.
- When alpha is zero expert interface of ZSETV is invoked. In this case,
all the elements of the input vector are set 0.
- Invocation of expert interface means that NULL pointer can be passed
to the function in place of context. Expert interface of ZSETV will
query the context and get the approriate function pointer.
- Added BLAS interface for ZSCALV. The architecture ID is used to decide
the function that is to be invoked.
- Created a new macro INSERT_GENTFUNCSCAL_BLAS_C to instantiate SCALV
BLAS macro interface only for single complex type and single complex,
float mixed type
AMD-Internal: [CPUPL-2773]
Change-Id: I0d6995bce883c0ebdc5da0046608fc59d03f6050
- Vectorized alpha scaling of X vector using SSE instructions. This
can be done irrespective of incx.
- Added code to prefetch A matrix and Y vector to L1 cache
- Vectorized fringe case computation and non-unit stride computation
with SSE instructions.
- Increased unroll in unit stride cases for better register
utilization.
AMD-Internal: [CPUPL-2773]
Change-Id: I217e6ce9e3f5753ebe271c684abd9a2274fd2715
-Inefficient assembly is generated for s16 gemm micro-kernel(intrinsics
code) when compiled using gcc. The presence of -fschedule-insns +
-fschedule-insns2 + -ftree-pre in O2 compiler optimization flags
results in the code being optimized to reduce data stalls, and results
in the usage of stack to store intermediate C register output. Disabling
-ftree-pre in gcc fixes the issue, even in the presence of the other
two flags.
AMD-Internal: [CPUPL-2971]
Change-Id: Ibf0dcde20b5a18708a05faad34e684eb0a9a5463
1. Added Tanh approximation based GeLU Post-Op for S16, S32 and BF16
2. Changes are done at frame and micro-kernel level to
implement this post-op.
3. Efficient AVX-512 and AVX-2 vector versions of TANHF and EXPF
functions are implemented for the GeLU post-operation.
4. TANH and EXPF math functions are efficiently implemented in
macro-based fashion to exploit register level fusion of GeLU
with GEMM operations for improved performance
5. LPGEMM bench is changed to pass GeLU post-op as input and
support accuracy check to verify functional correctness
AMD-Internal: [CPUPL-2978]
Change-Id: I472ac35c00a4ea1ab983cc5f6ff6a123c8035f28
Details:
- DAXPYV is parallelized based on the number of threads
set by Application.
- Using the empirical data collected on genoa, optimal
number of threads is set.
- If intended number of threads to be spawned is not
equal to actual number of threads, AXPY routine will
execute sequentially.
Change-Id: I48015a712ada6428056af6320aa5570596a02b1f
AMD-Internal: [CPUPL-2747]
Details:
- Now AOCL BLIS uses AX512 - 32x6 DGEMM kernel for native code path.
Thanks to Moore, Branden <Branden.Moore@amd.com> for suggesting and
implementing these optimizations.
- In the initial version of 32x6 DGEMM kernel, to broadcast elements of B packed
we perform load into xmm (2 elements), broadcast into zmm from xmmm and then to get the
next element, we do vpermilpd(xmm). This logic is replaced with direct broadcast from
memory, since the elements of Bpack are stored contiguously, the first broadcast fetches
the cacheline and then subsequent broadcasts happen faster. We use two registers for broadcast
and interleave broadcast operation with FMAs to hide any memory latencies.
- Native dTRSM uses 16x14 dgemm - therefore we need to override the default blkszs (MR,NR,..)
when executing trsm. we call bli_zen4_override_trsm_blkszs(cntx_local) on a local cntx_t object
for double data-type as well in the function bli_trsm_front(), bli_trsm_xx_ker_var2, xx = {ll,lu,rl,ru}.
Renamed "BLIS_GEMM_AVX2_UKR" to "BLIS_GEMM_FOR_TRSM_UKR" and in the bli_cntx_init_zen4() we replaced
dgemm kernel for TRSM with 16x14 dgemm kernel.
- New packm kernels - 16xk, 24xk and 32xk are added.
- New 32xk packm reference kernel is added in bli_packm_cxk_ref.c and it is
enabled for zen4 config (bli_dpackm_32xk_zen4_ref() )
- Copyright year updated for modified files.
- cleaned up code for "zen" config - removed unused packm kernels declaration in kernels/zen/bli_kernels.h
- [SWLCSG-1374], [CPUPL-2918]
Change-Id: I576282382504b72072a6db068eabd164c8943627
-The f32 gemm framework is modified to swap input column major matrices
and compute gemm for the transposed matrices (now row major) using the
existing row-major kernels. The output is written to C matrix assuming
it is transposed.
-Framework changes to support leading dimensions that are greater than
matrix widths.
AMD-Internal: [CPUPL-2919]
Change-Id: I805f1cb9ff934bb3106e01eb74e528915ffb90a3
-The post operations attributes are moved to a new struct
lpgemm_post_op_attr, and an object of this struct is passed to the
low precision gemm kernels in place of the multiple parameters.
-The u8s8s32s8 api (downscale api) performance is low when the k
value is less (k < KC). Two scenarios are observed here:
a. beta = 0: Currently, for downscale api, a temporary buffer is
used to accumulate intermediate s32 output, so that it can be used
in later iterations of pc loop (k dim). The usage of this buffer
(store) can be avoided if k < KC. Here intermediate accumulation
is not required, since the after the first iteration of the pc loop,
the output can be downscaled and stored.
b. beta != 0: In this case the existing values of the original s8 C
output matrix needs to be converted to s32 and beta scaled. Currently
the s8 values are converted to s32 and stored in temporary buffer in
pc loop (5 loop algorithm) in blocks of mxNC. This temporary buffer
is passed to the micro kernel and beta scaling is applied on this.
However the mxNC block copy is costly and can be avoided if a new
condition is introduced for beta scaling in the micro kernel, whereby
the original s8 data is loaded instead of from the temporary buffer
to a register, converted to s32 and beta scaling applied on it.
AMD-Internal: [CPUPL-2884]
Change-Id: Id9b4650d500e1b553e48c4f1e4c902b3f553211c
Details: k0 is always positive in bli_dgemm_haswell_asm_6x8(), the operation involved with
k0 is typecasted to uint64_t to enable AOCC generate optimized code.
Thanks for Jini Susan (jinisusan.george@amd.com) from compiler team for suggesting
this change. Similar change was applied to sgemm, cgemm and zgemm kernels.
Change-Id: I423c949e0c1835652142a6931dadf4a7d190aeb9
- In zen4 TRSM and GEMM have different blocksizes,
when trsm is called, blocksizes are changed in
global cntx object. If GEMM and TRSM are called
in parallel, blocksizes in global cntx will not
be correct for GEMM which will cause a seg fault.
- To fix this, a local copy of cntx is created
and blocksizes are changed only in the local copy.
AMD-Internal:[CPUPL-2896]
Change-Id: I0e724520a92fc3b2ed0becf385ec41ab5d1b4490
Corrections for some occurances of:
- Compiler warnings about initialization of float from double
- Spelling mistakes in comments
- Incorrect indentation of code and comments
AMD-Internal: [CPUPL-2870]
Change-Id: Icb68c789687bd0684844331d43071bfffecac9fc
Details:
- In case of dzgemm, if the microkernel prefers column output,
We will induce a transposition and perform C += A*B
where A (formerly B) becomes complex and B(formerly A)
becomes real. Hence attach complex alpha object to A instead of B.
- This commit reverts all the changes made by
d62f12a18a and
0b81f53074 as they are causing
failures in make checkblis-md.
AMD-Internal: [CPUPL-2893]
Change-Id: I56b94ac136fb96003302c568ae2587142c836620
-Implemented (r)ow preferential (d)ot product milli-kernels
(m and n variants) for dcomplex datatype along SUP path.
-These computational kernels extend the support for handling RRC and
CRC storage schemes along the SUP path. In case of BLAS api call,
it corresponds to the input cases with transa equal to T and
transb equal to N.
-In case of the B matrix being packed(conditionally), the inputs are
redirected to the existing (r)ow preferential (v)ector load optimized
kernels due to better performance.
-Added macro for vhsubpd assembly instruction, to support the arithmetic
for complex datatype in its interleaved storage.
AMD-Internal: [CPUPL-2593]
Change-Id: If90834e55e9e31aa87d3d5b711efad9ef2458da8
- Only for RRR case, var2m kernels are added
- Main kernel is of 12x32 (AVX512), associated fringe kernels of
- 8x32, 4x32, 2x32, 1x32 (AVX512)
- 12x16, 8x16, 4x16, 2x16, 1x16 (AVX512)
- 12x8, 8x8 (AVX2)
- 12x4, 8x4 (SSE4)
- 12x2, 8x2 (SSE4)
- existing AVX2/SSE4 kernels are used for other fringe
cases
- Currently, these kernels are not invoked in zen4 path
- Once all AVX512 kernels (n and rd) are done, invoke all of them
together in zen4 config
AMD-Internal: [CPUPL-2801]
Change-Id: I7a206fee9151e92319d83dcc5f3eed61d3bf1196
- For the cases where AVX2 is available, an optimized function is called,
based on Blue's algorithm. The fallback method based on sumsqv is used
otherwise.
- Scaling is used to avoid overflow and underflow.
- Works correctly for negative increments.
AMD-Internal: [SWLCSG-1080]
Change-Id: I6bf2f42652ba6b8a8631a0a9e6f6297d5b3ea5d9
Bugfix for parallel BLAS1 and BLAS2 routines. Threading information
was not being set correctly when initializing local rntm from global.
Also ensure th_rntm is initialized along with global_rntm by updating
it in bli_thread_init(), called by bli_init_once()
AMD-Internal: [CPUPL-2433]
Change-Id: Iba658f87ae13fe16a57ca1fc279e149b7fa294cf
- GoogleTest headers removed. GoogleTest gets fetched at
configuration time.
- BLIS headers removed. A BLIS installation path is required at
configuration time.
- Windows has been temporarily disabled.
AMD-Internal: [CPUPL-2732]
Change-Id: I9e55c8e43b2733f96cd8b6e5449d79623decad5c
Improvements to recent parallelism changes:
1. BLIS specific threading options: In bli_thread_update_rntm_from_env()
set threading variables in tl_rntm to serial values when OpenMP level
for parallelism within BLIS will not be active. User supplied BLIS
threading values remain unchanged in global_rntm.
2. Simplify code structure in bli_thread_update_rntm_from_env().
3. Change variable declarations in bli_thread_init_rntm_from_env()
and bli_thread_update_rntm_from_env() to avoid unused variable
warnings in non-OpenMP builds.
AMD-Internal: [CPUPL-2433]
Change-Id: I5505657e3d2722e69bc4a1c1bb9fd8df55407fdd
HPL script was using BLIS manual way to set threading, i.e. setting
BLIS_IC_NT explicitly. This causes bli_rntm_num_threads() to return
-1, which wasn't trapped in parallelised BLAS1 and BLAS2 routines.
Fix: if this occurs, set local number of threads based on product of
BLIS_JC_NT * BLIS_PC_NT * BLIS_IC_NT * BLIS_JR_NT * BLIS_IR_NT values.
Note: BLIS_PC_NT should always be 1, but this environment variable
is currently being read (contrary to documentation), so include it
for now.
Other changes:
* implement _Pragma convention in all code used on AMD
* frame/2/gemv/bli_gemv_unf_var1_amd.c: Remove is_omp_mt_enabled flag
AMD-Internal: [CPUPL-2803]
Change-Id: I37e8b038e5640d6693a87be0609888186322b465
- In the existing implementation, when the actual number of threads
spawned is different from the indicated number of threads for the
parallel region, partial job is being executed.
- Fix added to identify actual number of threads spawned and
allocate the work load to single thread in case of discrepancy
in the number of threads spawned vs indicated.
AMD-Internal: [CPUPL-2761]
Change-Id: I9eb9baccaef7f2149346be705ee151ae7af0a0e5
- In the existing implementation, when the actual number of threads
spawned is different from the indicated number of threads for the
parallel region, partial job is being executed.
- Fix added to identify actual number of threads spawned and
allocate the work load to single thread in case of discrepancy
in the number of threads spawned vs indicated.
AMD-Internal: [CPUPL-2761]
Change-Id: Ife36e6e4993bdcc5a506349b54b2177173866e32
- In the existing implementation, when the actual number of threads
spawned is different from the indicated number of threads for the
parallel region, partial job is being executed.
- Fix added to identify actual number of threads spawned and
allocate the work load to single thread in case of discrepancy
in the number of threads spawned vs indicated.
AMD-Internal: [SWLCSG-1659]
Change-Id: I882d0af608009e502189a4469eb3483d659c2b3f
- Broke down the KR loop inside the compute kernel into
two pieces
- Added C matrix prefetch between the two decomposed
pieces of KR loop
AMD-Internal: [CPUPL-2693]
Change-Id: Ib73bc2145de4c75bc8153d7d7d20fb057270c94e
- NC value is adjusted based on the n dimension of the B
matrix.
- KC value is adjusted based on the k dimension input
matrices.
- In order to ensure a handshake between the reorder
function and compute function, the same dynamic block
tuning logic is called in both places.
- Reorder followed by compute scenarios mean that tuning
KC value based on MC value and vice versa is not feasible.
AMD-Internal: [CPUPL-2638]
Change-Id: Ie515da7de83e3dfb59946e873c92d67f11559429
-Enabled the sup path for zgemm in case of single thread.
-Fine tuned the thresholds for small path in order to handle
certain spectrum of skinny matrices.
-The inputs are now redirected among small, sup and native
paths in case of single thread.
AMD-Internal: [CPUPL-2632]
Change-Id: I3eca315bcb8bfa8e015349a7b2bc4ebd5448e065
1. Check OpenMP active level against max active levels when setting
number of threads for starting a new parallel region in
./frame/thread/bli_thread.c to ensure the correct number of threads
is used when BLIS is called within nested OpenMP parallelism.
2. In subsequent BLIS calls, threading choices could be incorrectly
set based on values used and stored in global_rntm by a previous
call. This could apply when the OpenMP number of threads differ from
call to call, different nested parallelism is used in different
parts of a user's code, or different threads at the user level
request different numbers of OpenMP threads for BLIS calls.
Keep threading information in both global_rntm and a new Thread
Local Storage copy tl_rntm. Update tl_rntm from OpenMP runtime
environment (as appropriate) during bli_init_auto() calls in each
BLIS routine. The details are:
* global_rntm is initialized on first BLIS call based on OpenMP and
BLIS threading environment variables.
* global_rntm is updated by any BLIS threading function calls.
* In bli_thread_update_tl(), called by bli_init_auto(), sync with
any BLIS values set or updated in global_rntm. Then, if BLIS
threading control is not used, check OpenMP ICVs and set thread
count and auto_factor appropriately.
* Setting BLIS threading locally (using expert interfaces to pass
a user defined rntm data structure) should work as before.
3. bli_thread_get_is_parallel can now only be called outside of
parallelism within BLIS routines. Change calls in trsm to reflect
this.
4. Ensure blis_mt is set to TRUE in bli_thread_init_rntm_from_env()
if any BLIS_*_NT environment variables are set.
5. Set auto_factor = FALSE when the number of threads is 1.
6. bli_rntm_set_num_threads() and bli_rntm_set_ways() set blis_mt=TRUE.
7. Set blis_mt=FALSE in BLIS_RNTM_INITIALIZER and bli_rntm_init().
8. For debugging, internal information on the rntm threading data can
be printed by defining "PRINT_THREADING" at the top of bli_rntm.h
9. bli_rntm_print() now also prints the value of blis_mt.
10. Function prototypes in bli_rntm.h moved to top of file, so that
bli_rntm_print() can be used within inline functions defined in
this header file.
11. Comment out bli_init_auto() and bli_finalize_auto() calls in
Fortran interfaces in frame/compat/blis/thread/b77_thread.c
12. In frame/3/bli_l3_sup_int_amd.c move two calls to set_pack_a and
set_pack_b functions outside of the auto_factor if statements.
13. Misc code tidying.
AMD-Internal: [CPUPL-2433]
Change-Id: I8342c37fb4e280118e5e55164fbd6ea636f858ee
-Disabled the sup path for zgemm with single thread temporarily
until we tune the performance and update the thresholds.
The redirection of inputs in zgemm is now among small and native path
for single thread.
-Fine tuned the input size constraints for entry to small path in zgemm
in case of single thread.
AMD-Internal: [CPUPL-2495]
Change-Id: Ifa19116757ca132d9636c3f2a0661524d2c1d3ec
Clipping is done during the downscaling of the accumulated result
from s32 to s8 for u8s8s32os8 and from s16 to s8 for u8s8s16os8,
to saturate the final output values between [-128,127]
AMD-Internal: [LWPZENDNN-493]
Change-Id: Ica9bba5044e87b815e2b4e35809bf440bb9dd41f
1. Fixed accuracy issues in dtrsm small multi-thread.
2. Fixed out of bound memory accesses in the patch
"Fixes to avoid Out of Bound Memory Access in TRSM small algorithm"
3. Re-enabled DTRSM small MT with above fixes.
AMD-Internal: [CPUPL-2567]
Change-Id: Ibf2949b25fde4007a92cc635526bed0e0d897800
Description:
- When the value of the result in s8 for u8s8s32 and u8s8s16 are
close to 0. Values are getting ceiled to 1.
- Used nearbyintf to round the downscaled values in bench reference.
This fixed the result mismatch issue between the vectorized kernel
implementation and reference implementation in bench accuracy test.
AMD-Internal: [CPUPL-2617]
Change-Id: Ie42d612b1933bf622e6bd80eaf3db4bcb7a3ce82
Zen4 kernel bli_damaxv_zen_int_avx512 is causing incorrect results in
the netlib LAPACK tests, specifically in:
./xlintstd < ../dtest.in > dtest.out
in the TESTING/LIN directory. Given time constraints, i.e. the need to
finalize code for AOCL 4.0 release, disable calls to AVX512 kernel
(i.e. always use the AVX2 kernel) for now, and aim to correct
bli_damaxv_zen_int_avx512 for AOCL 4.1.
AMD-Internal: [CPUPL-2590]
Change-Id: I2603dd97c3931acb9730563e8126b109ec2b2572
- Added DTRSM AVX512 kernels for lower and upper variants in the native path.
- Changes in framework are made to accommodate these kernels.
AMD-Internal: [CPUPL-2588]
Change-Id: I1f74273ef2389018343c0645870290373ce25efe
