Skip to content

Commit 03b3395

Browse files
JaccovGJaccovG
authored
Merge pull request #77 from foss-for-synopsys-dwc-arc-processors/mli_dev
Release version 1.0
2 parents 690708d + dabcb77 commit 03b3395

17 files changed

+1603
-11620
lines changed

‎README.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -14,7 +14,7 @@ documentation and examples. Read the documentation at [embarc.org](https://embar
1414
* Max and average pooling
1515
* LSTM, Basic RNN
1616
* Elementwise (add, sub, mul, min, max)
17-
* Data manipulation (concatanation, permute, 2D padding)
17+
* Data manipulation (concatenation, permute, 2D padding)
1818
* ReLU, Leaky ReLu, ReLu1, ReLu6
1919
* Softmax, Sigmoid, TanH
2020
3. Supported data layout CHW (Channel-Height-Width standard for Caffe)

‎examples/example_cifar10_caffe/Makefile

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -29,6 +29,10 @@ OUT_NAME ?= example_cifar10_caffe
2929
ifeq ($(TOOLCHAIN),mwdt)
3030
# MWDT specific options
3131
CFLAGS = -Hnocopyr -Hpurge -Hheap=8K -Hstack=1K -Hfxapi -e_start -Bgrouplib -Hldopt=-q -Hsdata0 -Xdsp_ctrl=postshift,guard,convergent -Hdense_prologue
32+
33+
# use compact CRT
34+
CFLAGS += -Hcl -Hcrt_argv -Hcrt_fast_memcpy -Hcrt_fast_memset -Hxcheck -Hcrt_initbss
35+
3236
else
3337
PREBUILT_LIB ?= $(EMBARC_MLI_DIR)/examples/prebuilt/libmli.a
3438

‎examples/example_har_smartphone/Makefile

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -29,6 +29,10 @@ OUT_NAME ?= example_har_smartphone
2929
ifeq ($(TOOLCHAIN),mwdt)
3030
# MWDT specific options
3131
CFLAGS = -Hnocopyr -Hpurge -Hheap=8K -Hstack=1K -Hfxapi -e_start -Bgrouplib -Hldopt=-q -Hsdata0 -Xdsp_ctrl=postshift,guard,convergent -Hdense_prologue
32+
33+
# use compact CRT
34+
CFLAGS += -Hcl -Hcrt_argv -Hcrt_fast_memcpy -Hcrt_fast_memset -Hxcheck -Hcrt_initbss
35+
3236
else
3337
PREBUILT_LIB ?= $(EMBARC_MLI_DIR)/examples/prebuilt/libmli.a
3438

‎include/api/mli_krn_avepool_spec_api.h

Lines changed: 48 additions & 66 deletions
Original file line numberDiff line numberDiff line change
@@ -26,84 +26,66 @@ extern "C" {
2626

2727
mli_status mli_krn_avepool_chw_fx16_k2x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
2828
mli_status mli_krn_avepool_chw_fx16_k4x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
29-
mli_status mli_krn_avepool_chw_fx16_k5x5_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
30-
mli_status mli_krn_avepool_chw_fx16_k7x7_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
31-
mli_status mli_krn_avepool_chw_fx16_k9x9_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
32-
mli_status mli_krn_avepool_chw_fx16_k4x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
33-
mli_status mli_krn_avepool_chw_fx16_k4x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
34-
mli_status mli_krn_avepool_chw_fx16_k4x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
35-
mli_status mli_krn_avepool_chw_fx16_k4x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
36-
mli_status mli_krn_avepool_chw_fx16_k6x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
37-
mli_status mli_krn_avepool_chw_fx16_k6x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
38-
mli_status mli_krn_avepool_chw_fx16_k6x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
39-
mli_status mli_krn_avepool_chw_fx16_k6x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
40-
mli_status mli_krn_avepool_chw_fx16_k8x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
41-
mli_status mli_krn_avepool_chw_fx16_k8x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
42-
mli_status mli_krn_avepool_chw_fx16_k8x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
43-
mli_status mli_krn_avepool_chw_fx16_k8x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
44-
mli_status mli_krn_avepool_chw_fx16_k4x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
45-
mli_status mli_krn_avepool_chw_fx16_k6x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
46-
mli_status mli_krn_avepool_chw_fx16_k6x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
47-
mli_status mli_krn_avepool_chw_fx16_k6x6_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
48-
mli_status mli_krn_avepool_chw_fx16_k6x8_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
49-
mli_status mli_krn_avepool_chw_fx16_k8x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
50-
mli_status mli_krn_avepool_chw_fx16_k8x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
51-
mli_status mli_krn_avepool_chw_fx16_k8x6_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
52-
mli_status mli_krn_avepool_chw_fx16_k8x8_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
53-
mli_status mli_krn_avepool_chw_fx16_k3x3_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
54-
mli_status mli_krn_avepool_chw_fx16_k5x5_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
55-
mli_status mli_krn_avepool_chw_fx16_k7x7_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
56-
mli_status mli_krn_avepool_chw_fx16_k9x9_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
57-
mli_status mli_krn_avepool_chw_fx16_k2x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
5829
mli_status mli_krn_avepool_chw_fx16_k3x3_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
59-
mli_status mli_krn_avepool_chw_fx16_k4x4_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6030
mli_status mli_krn_avepool_chw_fx16_k5x5_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
61-
mli_status mli_krn_avepool_chw_fx16_k6x6_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6231
mli_status mli_krn_avepool_chw_fx16_k7x7_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
63-
mli_status mli_krn_avepool_chw_fx16_k8x8_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6432
mli_status mli_krn_avepool_chw_fx16_k9x9_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
65-
mli_status mli_krn_avepool_chw_fx16_k10x10_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
33+
mli_status mli_krn_avepool_chw_fx16_k2x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
34+
mli_status mli_krn_avepool_chw_fx16_k4x4_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
35+
mli_status mli_krn_avepool_chw_fx16_k6x6_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
36+
mli_status mli_krn_avepool_chw_fx16_k8x8_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
37+
mli_status mli_krn_avepool_chw_fx16_k3x3_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
38+
mli_status mli_krn_avepool_chw_fx16_k5x5_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
39+
mli_status mli_krn_avepool_chw_fx16_k7x7_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
40+
mli_status mli_krn_avepool_chw_fx16_k9x9_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
41+
mli_status mli_krn_avepool_chw_fx16_k2x2_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
42+
mli_status mli_krn_avepool_chw_fx16_k4x4_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
43+
mli_status mli_krn_avepool_chw_fx16_k6x6_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
44+
mli_status mli_krn_avepool_chw_fx16_k8x8_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
45+
mli_status mli_krn_avepool_chw_fx16_k1xn_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
46+
mli_status mli_krn_avepool_chw_fx16_k1x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
47+
mli_status mli_krn_avepool_chw_fx16_k1x3_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
48+
mli_status mli_krn_avepool_chw_fx16_knx1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
49+
mli_status mli_krn_avepool_chw_fx16_k2x1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
50+
mli_status mli_krn_avepool_chw_fx16_k3x1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
51+
mli_status mli_krn_avepool_chw_fx16_k1xn_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
52+
mli_status mli_krn_avepool_chw_fx16_k1x2_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
53+
mli_status mli_krn_avepool_chw_fx16_k1x3_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
54+
mli_status mli_krn_avepool_chw_fx16_knx1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
55+
mli_status mli_krn_avepool_chw_fx16_k2x1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
56+
mli_status mli_krn_avepool_chw_fx16_k3x1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6657
mli_status mli_krn_avepool_chw_fx16_generic(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6758

6859
mli_status mli_krn_avepool_chw_fx8_k2x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
6960
mli_status mli_krn_avepool_chw_fx8_k4x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
70-
mli_status mli_krn_avepool_chw_fx8_k5x5_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
71-
mli_status mli_krn_avepool_chw_fx8_k7x7_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
72-
mli_status mli_krn_avepool_chw_fx8_k9x9_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
73-
mli_status mli_krn_avepool_chw_fx8_k4x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
74-
mli_status mli_krn_avepool_chw_fx8_k4x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
75-
mli_status mli_krn_avepool_chw_fx8_k4x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
76-
mli_status mli_krn_avepool_chw_fx8_k4x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
77-
mli_status mli_krn_avepool_chw_fx8_k6x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
78-
mli_status mli_krn_avepool_chw_fx8_k6x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
79-
mli_status mli_krn_avepool_chw_fx8_k6x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
80-
mli_status mli_krn_avepool_chw_fx8_k6x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
81-
mli_status mli_krn_avepool_chw_fx8_k8x2_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
82-
mli_status mli_krn_avepool_chw_fx8_k8x4_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
83-
mli_status mli_krn_avepool_chw_fx8_k8x6_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
84-
mli_status mli_krn_avepool_chw_fx8_k8x8_str1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
85-
mli_status mli_krn_avepool_chw_fx8_k4x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
86-
mli_status mli_krn_avepool_chw_fx8_k6x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
87-
mli_status mli_krn_avepool_chw_fx8_k6x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
88-
mli_status mli_krn_avepool_chw_fx8_k6x6_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
89-
mli_status mli_krn_avepool_chw_fx8_k6x8_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
90-
mli_status mli_krn_avepool_chw_fx8_k8x2_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
91-
mli_status mli_krn_avepool_chw_fx8_k8x4_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
92-
mli_status mli_krn_avepool_chw_fx8_k8x6_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
93-
mli_status mli_krn_avepool_chw_fx8_k8x8_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
94-
mli_status mli_krn_avepool_chw_fx8_k3x3_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
95-
mli_status mli_krn_avepool_chw_fx8_k5x5_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
96-
mli_status mli_krn_avepool_chw_fx8_k7x7_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
97-
mli_status mli_krn_avepool_chw_fx8_k9x9_str1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
98-
mli_status mli_krn_avepool_chw_fx8_k2x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
9961
mli_status mli_krn_avepool_chw_fx8_k3x3_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
100-
mli_status mli_krn_avepool_chw_fx8_k4x4_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
10162
mli_status mli_krn_avepool_chw_fx8_k5x5_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
102-
mli_status mli_krn_avepool_chw_fx8_k6x6_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
10363
mli_status mli_krn_avepool_chw_fx8_k7x7_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
104-
mli_status mli_krn_avepool_chw_fx8_k8x8_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
10564
mli_status mli_krn_avepool_chw_fx8_k9x9_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
106-
mli_status mli_krn_avepool_chw_fx8_k10x10_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
65+
mli_status mli_krn_avepool_chw_fx8_k2x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
66+
mli_status mli_krn_avepool_chw_fx8_k4x4_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
67+
mli_status mli_krn_avepool_chw_fx8_k6x6_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
68+
mli_status mli_krn_avepool_chw_fx8_k8x8_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
69+
mli_status mli_krn_avepool_chw_fx8_k3x3_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
70+
mli_status mli_krn_avepool_chw_fx8_k5x5_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
71+
mli_status mli_krn_avepool_chw_fx8_k7x7_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
72+
mli_status mli_krn_avepool_chw_fx8_k9x9_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
73+
mli_status mli_krn_avepool_chw_fx8_k2x2_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
74+
mli_status mli_krn_avepool_chw_fx8_k4x4_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
75+
mli_status mli_krn_avepool_chw_fx8_k6x6_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
76+
mli_status mli_krn_avepool_chw_fx8_k8x8_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
77+
mli_status mli_krn_avepool_chw_fx8_k1xn_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
78+
mli_status mli_krn_avepool_chw_fx8_k1x2_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
79+
mli_status mli_krn_avepool_chw_fx8_k1x3_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
80+
mli_status mli_krn_avepool_chw_fx8_knx1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
81+
mli_status mli_krn_avepool_chw_fx8_k2x1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
82+
mli_status mli_krn_avepool_chw_fx8_k3x1_krnpad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
83+
mli_status mli_krn_avepool_chw_fx8_k1xn_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
84+
mli_status mli_krn_avepool_chw_fx8_k1x2_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
85+
mli_status mli_krn_avepool_chw_fx8_k1x3_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
86+
mli_status mli_krn_avepool_chw_fx8_knx1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
87+
mli_status mli_krn_avepool_chw_fx8_k2x1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
88+
mli_status mli_krn_avepool_chw_fx8_k3x1_nopad(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
10789
mli_status mli_krn_avepool_chw_fx8_generic(const mli_tensor * in, const mli_pool_cfg * cfg, mli_tensor * out);
10890

10991
#ifdef __cplusplus

0 commit comments

Comments
 (0)