1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
|
/*
* Copyright 2013 Advanced Micro Devices, Inc.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*
* Authors: Alex Deucher
*/
#include "radeon.h"
#include "radeon_asic.h"
#include "radeon_trace.h"
#include "si.h"
#include "sid.h"
/**
* si_dma_is_lockup - Check if the DMA engine is locked up
*
* @rdev: radeon_device pointer
* @ring: radeon_ring structure holding ring information
*
* Check if the async DMA engine is locked up.
* Returns true if the engine appears to be locked up, false if not.
*/
bool si_dma_is_lockup(struct radeon_device *rdev, struct radeon_ring *ring)
{
u32 reset_mask = si_gpu_check_soft_reset(rdev);
u32 mask;
if (ring->idx == R600_RING_TYPE_DMA_INDEX)
mask = RADEON_RESET_DMA;
else
mask = RADEON_RESET_DMA1;
if (!(reset_mask & mask)) {
radeon_ring_lockup_update(rdev, ring);
return false;
}
return radeon_ring_test_lockup(rdev, ring);
}
/**
* si_dma_vm_copy_pages - update PTEs by copying them from the GART
*
* @rdev: radeon_device pointer
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
* @src: src addr where to copy from
* @count: number of page entries to update
*
* Update PTEs by copying them from the GART using the DMA (SI).
*/
void si_dma_vm_copy_pages(struct radeon_device *rdev,
struct radeon_ib *ib,
uint64_t pe, uint64_t src,
unsigned count)
{
while (count) {
unsigned bytes = count * 8;
if (bytes > 0xFFFF8)
bytes = 0xFFFF8;
ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY,
1, 0, 0, bytes);
ib->ptr[ib->length_dw++] = lower_32_bits(pe);
ib->ptr[ib->length_dw++] = lower_32_bits(src);
ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
ib->ptr[ib->length_dw++] = upper_32_bits(src) & 0xff;
pe += bytes;
src += bytes;
count -= bytes / 8;
}
}
/**
* si_dma_vm_write_pages - update PTEs by writing them manually
*
* @rdev: radeon_device pointer
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
* @addr: dst addr to write into pe
* @count: number of page entries to update
* @incr: increase next addr by incr bytes
* @flags: access flags
*
* Update PTEs by writing them manually using the DMA (SI).
*/
void si_dma_vm_write_pages(struct radeon_device *rdev,
struct radeon_ib *ib,
uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags)
{
uint64_t value;
unsigned ndw;
while (count) {
ndw = count * 2;
if (ndw > 0xFFFFE)
ndw = 0xFFFFE;
/* for non-physically contiguous pages (system) */
ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, ndw);
ib->ptr[ib->length_dw++] = pe;
ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
for (; ndw > 0; ndw -= 2, --count, pe += 8) {
if (flags & R600_PTE_SYSTEM) {
value = radeon_vm_map_gart(rdev, addr);
} else if (flags & R600_PTE_VALID) {
value = addr;
} else {
value = 0;
}
addr += incr;
value |= flags;
ib->ptr[ib->length_dw++] = value;
ib->ptr[ib->length_dw++] = upper_32_bits(value);
}
}
}
/**
* si_dma_vm_set_pages - update the page tables using the DMA
*
* @rdev: radeon_device pointer
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
* @addr: dst addr to write into pe
* @count: number of page entries to update
* @incr: increase next addr by incr bytes
* @flags: access flags
*
* Update the page tables using the DMA (SI).
*/
void si_dma_vm_set_pages(struct radeon_device *rdev,
struct radeon_ib *ib,
uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags)
{
uint64_t value;
unsigned ndw;
while (count) {
ndw = count * 2;
if (ndw > 0xFFFFE)
ndw = 0xFFFFE;
if (flags & R600_PTE_VALID)
value = addr;
else
value = 0;
/* for physically contiguous pages (vram) */
ib->ptr[ib->length_dw++] = DMA_PTE_PDE_PACKET(ndw);
ib->ptr[ib->length_dw++] = pe; /* dst addr */
ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
ib->ptr[ib->length_dw++] = flags; /* mask */
ib->ptr[ib->length_dw++] = 0;
ib->ptr[ib->length_dw++] = value; /* value */
ib->ptr[ib->length_dw++] = upper_32_bits(value);
ib->ptr[ib->length_dw++] = incr; /* increment size */
ib->ptr[ib->length_dw++] = 0;
pe += ndw * 4;
addr += (ndw / 2) * incr;
count -= ndw / 2;
}
}
void si_dma_vm_flush(struct radeon_device *rdev, struct radeon_ring *ring,
unsigned vm_id, uint64_t pd_addr)
{
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
if (vm_id < 8) {
radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + (vm_id << 2)) >> 2));
} else {
radeon_ring_write(ring, (0xf << 16) | ((VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + ((vm_id - 8) << 2)) >> 2));
}
radeon_ring_write(ring, pd_addr >> 12);
/* flush hdp cache */
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
radeon_ring_write(ring, (0xf << 16) | (HDP_MEM_COHERENCY_FLUSH_CNTL >> 2));
radeon_ring_write(ring, 1);
/* bits 0-7 are the VM contexts0-7 */
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
radeon_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST >> 2));
radeon_ring_write(ring, 1 << vm_id);
/* wait for invalidate to complete */
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0));
radeon_ring_write(ring, VM_INVALIDATE_REQUEST);
radeon_ring_write(ring, 0xff << 16); /* retry */
radeon_ring_write(ring, 1 << vm_id); /* mask */
radeon_ring_write(ring, 0); /* value */
radeon_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */
}
/**
* si_copy_dma - copy pages using the DMA engine
*
* @rdev: radeon_device pointer
* @src_offset: src GPU address
* @dst_offset: dst GPU address
* @num_gpu_pages: number of GPU pages to xfer
* @resv: reservation object to sync to
*
* Copy GPU paging using the DMA engine (SI).
* Used by the radeon ttm implementation to move pages if
* registered as the asic copy callback.
*/
struct radeon_fence *si_copy_dma(struct radeon_device *rdev,
uint64_t src_offset, uint64_t dst_offset,
unsigned num_gpu_pages,
struct dma_resv *resv)
{
struct radeon_fence *fence;
struct radeon_sync sync;
int ring_index = rdev->asic->copy.dma_ring_index;
struct radeon_ring *ring = &rdev->ring[ring_index];
u32 size_in_bytes, cur_size_in_bytes;
int i, num_loops;
int r = 0;
radeon_sync_create(&sync);
size_in_bytes = (num_gpu_pages << RADEON_GPU_PAGE_SHIFT);
num_loops = DIV_ROUND_UP(size_in_bytes, 0xfffff);
r = radeon_ring_lock(rdev, ring, num_loops * 5 + 11);
if (r) {
DRM_ERROR("radeon: moving bo (%d).\n", r);
radeon_sync_free(rdev, &sync, NULL);
return ERR_PTR(r);
}
radeon_sync_resv(rdev, &sync, resv, false);
radeon_sync_rings(rdev, &sync, ring->idx);
for (i = 0; i < num_loops; i++) {
cur_size_in_bytes = size_in_bytes;
if (cur_size_in_bytes > 0xFFFFF)
cur_size_in_bytes = 0xFFFFF;
size_in_bytes -= cur_size_in_bytes;
radeon_ring_write(ring, DMA_PACKET(DMA_PACKET_COPY, 1, 0, 0, cur_size_in_bytes));
radeon_ring_write(ring, lower_32_bits(dst_offset));
radeon_ring_write(ring, lower_32_bits(src_offset));
radeon_ring_write(ring, upper_32_bits(dst_offset) & 0xff);
radeon_ring_write(ring, upper_32_bits(src_offset) & 0xff);
src_offset += cur_size_in_bytes;
dst_offset += cur_size_in_bytes;
}
r = radeon_fence_emit(rdev, &fence, ring->idx);
if (r) {
radeon_ring_unlock_undo(rdev, ring);
radeon_sync_free(rdev, &sync, NULL);
return ERR_PTR(r);
}
radeon_ring_unlock_commit(rdev, ring, false);
radeon_sync_free(rdev, &sync, fence);
return fence;
}
|
