1 /*
2 * Copyright 2019 Advanced Micro Devices, Inc.
3 * All Rights Reserved.
4 *
5 * Permission is hereby granted, free of charge, to any person obtaining a
6 * copy of this software and associated documentation files (the
7 * "Software"), to deal in the Software without restriction, including
8 * without limitation the rights to use, copy, modify, merge, publish,
9 * distribute, sub license, and/or sell copies of the Software, and to
10 * permit persons to whom the Software is furnished to do so, subject to
11 * the following conditions:
12 *
13 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
14 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
15 * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
16 * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
17 * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
18 * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
19 * USE OR OTHER DEALINGS IN THE SOFTWARE.
20 *
21 * The above copyright notice and this permission notice (including the
22 * next paragraph) shall be included in all copies or substantial portions
23 * of the Software.
24 *
25 */
26
27 #include "amdgpu.h"
28 #include "amdgpu_jpeg.h"
29 #include "amdgpu_pm.h"
30 #include "soc15d.h"
31 #include "soc15_common.h"
32
33 #define JPEG_IDLE_TIMEOUT msecs_to_jiffies(1000)
34
35 static void amdgpu_jpeg_idle_work_handler(struct work_struct *work);
36
amdgpu_jpeg_sw_init(struct amdgpu_device * adev)37 int amdgpu_jpeg_sw_init(struct amdgpu_device *adev)
38 {
39 INIT_DELAYED_WORK(&adev->jpeg.idle_work, amdgpu_jpeg_idle_work_handler);
40 mutex_init(&adev->jpeg.jpeg_pg_lock);
41 atomic_set(&adev->jpeg.total_submission_cnt, 0);
42
43 return 0;
44 }
45
amdgpu_jpeg_sw_fini(struct amdgpu_device * adev)46 int amdgpu_jpeg_sw_fini(struct amdgpu_device *adev)
47 {
48 int i;
49
50 for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
51 if (adev->jpeg.harvest_config & (1 << i))
52 continue;
53
54 amdgpu_ring_fini(&adev->jpeg.inst[i].ring_dec);
55 }
56
57 mutex_destroy(&adev->jpeg.jpeg_pg_lock);
58
59 return 0;
60 }
61
amdgpu_jpeg_suspend(struct amdgpu_device * adev)62 int amdgpu_jpeg_suspend(struct amdgpu_device *adev)
63 {
64 cancel_delayed_work_sync(&adev->jpeg.idle_work);
65
66 return 0;
67 }
68
amdgpu_jpeg_resume(struct amdgpu_device * adev)69 int amdgpu_jpeg_resume(struct amdgpu_device *adev)
70 {
71 return 0;
72 }
73
amdgpu_jpeg_idle_work_handler(struct work_struct * work)74 static void amdgpu_jpeg_idle_work_handler(struct work_struct *work)
75 {
76 struct amdgpu_device *adev =
77 container_of(work, struct amdgpu_device, jpeg.idle_work.work);
78 unsigned int fences = 0;
79 unsigned int i;
80
81 for (i = 0; i < adev->jpeg.num_jpeg_inst; ++i) {
82 if (adev->jpeg.harvest_config & (1 << i))
83 continue;
84
85 fences += amdgpu_fence_count_emitted(&adev->jpeg.inst[i].ring_dec);
86 }
87
88 if (!fences && !atomic_read(&adev->jpeg.total_submission_cnt))
89 amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_JPEG,
90 AMD_PG_STATE_GATE);
91 else
92 schedule_delayed_work(&adev->jpeg.idle_work, JPEG_IDLE_TIMEOUT);
93 }
94
amdgpu_jpeg_ring_begin_use(struct amdgpu_ring * ring)95 void amdgpu_jpeg_ring_begin_use(struct amdgpu_ring *ring)
96 {
97 struct amdgpu_device *adev = ring->adev;
98
99 atomic_inc(&adev->jpeg.total_submission_cnt);
100 cancel_delayed_work_sync(&adev->jpeg.idle_work);
101
102 mutex_lock(&adev->jpeg.jpeg_pg_lock);
103 amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_JPEG,
104 AMD_PG_STATE_UNGATE);
105 mutex_unlock(&adev->jpeg.jpeg_pg_lock);
106 }
107
amdgpu_jpeg_ring_end_use(struct amdgpu_ring * ring)108 void amdgpu_jpeg_ring_end_use(struct amdgpu_ring *ring)
109 {
110 atomic_dec(&ring->adev->jpeg.total_submission_cnt);
111 schedule_delayed_work(&ring->adev->jpeg.idle_work, JPEG_IDLE_TIMEOUT);
112 }
113
amdgpu_jpeg_dec_ring_test_ring(struct amdgpu_ring * ring)114 int amdgpu_jpeg_dec_ring_test_ring(struct amdgpu_ring *ring)
115 {
116 struct amdgpu_device *adev = ring->adev;
117 uint32_t tmp = 0;
118 unsigned i;
119 int r;
120
121 WREG32(adev->jpeg.inst[ring->me].external.jpeg_pitch, 0xCAFEDEAD);
122 r = amdgpu_ring_alloc(ring, 3);
123 if (r)
124 return r;
125
126 amdgpu_ring_write(ring, PACKET0(adev->jpeg.internal.jpeg_pitch, 0));
127 amdgpu_ring_write(ring, 0xDEADBEEF);
128 amdgpu_ring_commit(ring);
129
130 for (i = 0; i < adev->usec_timeout; i++) {
131 tmp = RREG32(adev->jpeg.inst[ring->me].external.jpeg_pitch);
132 if (tmp == 0xDEADBEEF)
133 break;
134 udelay(1);
135 }
136
137 if (i >= adev->usec_timeout)
138 r = -ETIMEDOUT;
139
140 return r;
141 }
142
amdgpu_jpeg_dec_set_reg(struct amdgpu_ring * ring,uint32_t handle,struct dma_fence ** fence)143 static int amdgpu_jpeg_dec_set_reg(struct amdgpu_ring *ring, uint32_t handle,
144 struct dma_fence **fence)
145 {
146 struct amdgpu_device *adev = ring->adev;
147 struct amdgpu_job *job;
148 struct amdgpu_ib *ib;
149 struct dma_fence *f = NULL;
150 const unsigned ib_size_dw = 16;
151 int i, r;
152
153 r = amdgpu_job_alloc_with_ib(ring->adev, NULL, NULL, ib_size_dw * 4,
154 AMDGPU_IB_POOL_DIRECT, &job);
155 if (r)
156 return r;
157
158 ib = &job->ibs[0];
159
160 ib->ptr[0] = PACKETJ(adev->jpeg.internal.jpeg_pitch, 0, 0,
161 PACKETJ_TYPE0);
162 ib->ptr[1] = 0xDEADBEEF;
163 for (i = 2; i < 16; i += 2) {
164 ib->ptr[i] = PACKETJ(0, 0, 0, PACKETJ_TYPE6);
165 ib->ptr[i+1] = 0;
166 }
167 ib->length_dw = 16;
168
169 r = amdgpu_job_submit_direct(job, ring, &f);
170 if (r)
171 goto err;
172
173 if (fence)
174 *fence = dma_fence_get(f);
175 dma_fence_put(f);
176
177 return 0;
178
179 err:
180 amdgpu_job_free(job);
181 return r;
182 }
183
amdgpu_jpeg_dec_ring_test_ib(struct amdgpu_ring * ring,long timeout)184 int amdgpu_jpeg_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
185 {
186 struct amdgpu_device *adev = ring->adev;
187 uint32_t tmp = 0;
188 unsigned i;
189 struct dma_fence *fence = NULL;
190 long r = 0;
191
192 r = amdgpu_jpeg_dec_set_reg(ring, 1, &fence);
193 if (r)
194 goto error;
195
196 r = dma_fence_wait_timeout(fence, false, timeout);
197 if (r == 0) {
198 r = -ETIMEDOUT;
199 goto error;
200 } else if (r < 0) {
201 goto error;
202 } else {
203 r = 0;
204 }
205
206 for (i = 0; i < adev->usec_timeout; i++) {
207 tmp = RREG32(adev->jpeg.inst[ring->me].external.jpeg_pitch);
208 if (tmp == 0xDEADBEEF)
209 break;
210 udelay(1);
211 }
212
213 if (i >= adev->usec_timeout)
214 r = -ETIMEDOUT;
215
216 dma_fence_put(fence);
217 error:
218 return r;
219 }
220
amdgpu_jpeg_process_poison_irq(struct amdgpu_device * adev,struct amdgpu_irq_src * source,struct amdgpu_iv_entry * entry)221 int amdgpu_jpeg_process_poison_irq(struct amdgpu_device *adev,
222 struct amdgpu_irq_src *source,
223 struct amdgpu_iv_entry *entry)
224 {
225 struct ras_common_if *ras_if = adev->jpeg.ras_if;
226 struct ras_dispatch_if ih_data = {
227 .entry = entry,
228 };
229
230 if (!ras_if)
231 return 0;
232
233 ih_data.head = *ras_if;
234 amdgpu_ras_interrupt_dispatch(adev, &ih_data);
235
236 return 0;
237 }
238
jpeg_set_ras_funcs(struct amdgpu_device * adev)239 void jpeg_set_ras_funcs(struct amdgpu_device *adev)
240 {
241 if (!adev->jpeg.ras)
242 return;
243
244 amdgpu_ras_register_ras_block(adev, &adev->jpeg.ras->ras_block);
245
246 strcpy(adev->jpeg.ras->ras_block.ras_comm.name, "jpeg");
247 adev->jpeg.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__JPEG;
248 adev->jpeg.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__POISON;
249 adev->jpeg.ras_if = &adev->jpeg.ras->ras_block.ras_comm;
250
251 /* If don't define special ras_late_init function, use default ras_late_init */
252 if (!adev->jpeg.ras->ras_block.ras_late_init)
253 adev->jpeg.ras->ras_block.ras_late_init = amdgpu_ras_block_late_init;
254 }
255