2 * Copyright © 2017 Broadcom
4 * Permission is hereby granted, free of charge, to any person obtaining a
5 * copy of this software and associated documentation files (the "Software"),
6 * to deal in the Software without restriction, including without limitation
7 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8 * and/or sell copies of the Software, and to permit persons to whom the
9 * Software is furnished to do so, subject to the following conditions:
11 * The above copyright notice and this permission notice (including the next
12 * paragraph) shall be included in all copies or substantial portions of the
15 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
18 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
20 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
24 /** @file vc4_tiling_lt.c
26 * Helper functions from vc4_tiling.c that will be compiled for using NEON
31 #include "pipe/p_state.h"
32 #include "vc4_tiling.h"
34 /** Returns the stride in bytes of a 64-byte microtile. */
36 vc4_utile_stride(int cpp
)
46 unreachable("bad cpp");
51 vc4_load_utile(void *dst
, void *src
, uint32_t dst_stride
, uint32_t cpp
)
53 uint32_t src_stride
= vc4_utile_stride(cpp
);
55 for (uint32_t src_offset
= 0; src_offset
< 64; src_offset
+= src_stride
) {
56 memcpy(dst
, src
+ src_offset
, src_stride
);
62 vc4_store_utile(void *dst
, void *src
, uint32_t src_stride
, uint32_t cpp
)
64 uint32_t dst_stride
= vc4_utile_stride(cpp
);
66 for (uint32_t dst_offset
= 0; dst_offset
< 64; dst_offset
+= dst_stride
) {
67 memcpy(dst
+ dst_offset
, src
, dst_stride
);
73 vc4_load_lt_image(void *dst
, uint32_t dst_stride
,
74 void *src
, uint32_t src_stride
,
75 int cpp
, const struct pipe_box
*box
)
77 uint32_t utile_w
= vc4_utile_width(cpp
);
78 uint32_t utile_h
= vc4_utile_height(cpp
);
79 uint32_t xstart
= box
->x
;
80 uint32_t ystart
= box
->y
;
82 for (uint32_t y
= 0; y
< box
->height
; y
+= utile_h
) {
83 for (int x
= 0; x
< box
->width
; x
+= utile_w
) {
84 vc4_load_utile(dst
+ (dst_stride
* y
+
86 src
+ ((ystart
+ y
) * src_stride
+
87 (xstart
+ x
) * 64 / utile_w
),
94 vc4_store_lt_image(void *dst
, uint32_t dst_stride
,
95 void *src
, uint32_t src_stride
,
96 int cpp
, const struct pipe_box
*box
)
98 uint32_t utile_w
= vc4_utile_width(cpp
);
99 uint32_t utile_h
= vc4_utile_height(cpp
);
100 uint32_t xstart
= box
->x
;
101 uint32_t ystart
= box
->y
;
103 for (uint32_t y
= 0; y
< box
->height
; y
+= utile_h
) {
104 for (int x
= 0; x
< box
->width
; x
+= utile_w
) {
105 vc4_store_utile(dst
+ ((ystart
+ y
) * dst_stride
+
106 (xstart
+ x
) * 64 / utile_w
),
107 src
+ (src_stride
* y
+