summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c
blob: ce2bf1fb79ed12a2c7a11cb30eb5c8c9f49eebe0 (plain)
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
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
/*
 * Copyright 2014 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.
 *
 */

#include "amdgpu.h"
#include "nbio/nbio_2_3_offset.h"
#include "nbio/nbio_2_3_sh_mask.h"
#include "gc/gc_10_1_0_offset.h"
#include "gc/gc_10_1_0_sh_mask.h"
#include "soc15.h"
#include "navi10_ih.h"
#include "soc15_common.h"
#include "mxgpu_nv.h"

static void xgpu_nv_mailbox_send_ack(struct amdgpu_device *adev)
{
	WREG8(NV_MAIBOX_CONTROL_RCV_OFFSET_BYTE, 2);
}

static void xgpu_nv_mailbox_set_valid(struct amdgpu_device *adev, bool val)
{
	WREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE, val ? 1 : 0);
}

/*
 * this peek_msg could *only* be called in IRQ routine becuase in IRQ routine
 * RCV_MSG_VALID filed of BIF_BX_PF_MAILBOX_CONTROL must already be set to 1
 * by host.
 *
 * if called no in IRQ routine, this peek_msg cannot guaranteed to return the
 * correct value since it doesn't return the RCV_DW0 under the case that
 * RCV_MSG_VALID is set by host.
 */
static enum idh_event xgpu_nv_mailbox_peek_msg(struct amdgpu_device *adev)
{
	return RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW0);
}


static int xgpu_nv_mailbox_rcv_msg(struct amdgpu_device *adev,
				   enum idh_event event)
{
	u32 reg;

	reg = RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW0);
	if (reg != event)
		return -ENOENT;

	xgpu_nv_mailbox_send_ack(adev);

	return 0;
}

static uint8_t xgpu_nv_peek_ack(struct amdgpu_device *adev)
{
	return RREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE) & 2;
}

static int xgpu_nv_poll_ack(struct amdgpu_device *adev)
{
	int timeout  = NV_MAILBOX_POLL_ACK_TIMEDOUT;
	u8 reg;

	do {
		reg = RREG8(NV_MAIBOX_CONTROL_TRN_OFFSET_BYTE);
		if (reg & 2)
			return 0;

		mdelay(5);
		timeout -= 5;
	} while (timeout > 1);

	pr_err("Doesn't get TRN_MSG_ACK from pf in %d msec\n", NV_MAILBOX_POLL_ACK_TIMEDOUT);

	return -ETIME;
}

static int xgpu_nv_poll_msg(struct amdgpu_device *adev, enum idh_event event)
{
	int r, timeout = NV_MAILBOX_POLL_MSG_TIMEDOUT;

	do {
		r = xgpu_nv_mailbox_rcv_msg(adev, event);
		if (!r)
			return 0;

		msleep(10);
		timeout -= 10;
	} while (timeout > 1);


	return -ETIME;
}

static void xgpu_nv_mailbox_trans_msg (struct amdgpu_device *adev,
	      enum idh_request req, u32 data1, u32 data2, u32 data3)
{
	int r;
	uint8_t trn;

	/* IMPORTANT:
	 * clear TRN_MSG_VALID valid to clear host's RCV_MSG_ACK
	 * and with host's RCV_MSG_ACK cleared hw automatically clear host's RCV_MSG_ACK
	 * which lead to VF's TRN_MSG_ACK cleared, otherwise below xgpu_nv_poll_ack()
	 * will return immediatly
	 */
	do {
		xgpu_nv_mailbox_set_valid(adev, false);
		trn = xgpu_nv_peek_ack(adev);
		if (trn) {
			pr_err("trn=%x ACK should not assert! wait again !\n", trn);
			msleep(1);
		}
	} while (trn);

	WREG32_NO_KIQ(mmMAILBOX_MSGBUF_TRN_DW0, req);
	WREG32_NO_KIQ(mmMAILBOX_MSGBUF_TRN_DW1, data1);
	WREG32_NO_KIQ(mmMAILBOX_MSGBUF_TRN_DW2, data2);
	WREG32_NO_KIQ(mmMAILBOX_MSGBUF_TRN_DW3, data3);
	xgpu_nv_mailbox_set_valid(adev, true);

	/* start to poll ack */
	r = xgpu_nv_poll_ack(adev);
	if (r)
		pr_err("Doesn't get ack from pf, continue\n");

	xgpu_nv_mailbox_set_valid(adev, false);
}

static int xgpu_nv_send_access_requests(struct amdgpu_device *adev,
					enum idh_request req)
{
	int r;
	enum idh_event event = -1;

	xgpu_nv_mailbox_trans_msg(adev, req, 0, 0, 0);

	switch (req) {
	case IDH_REQ_GPU_INIT_ACCESS:
	case IDH_REQ_GPU_FINI_ACCESS:
	case IDH_REQ_GPU_RESET_ACCESS:
		event = IDH_READY_TO_ACCESS_GPU;
		break;
	case IDH_REQ_GPU_INIT_DATA:
		event = IDH_REQ_GPU_INIT_DATA_READY;
		break;
	default:
		break;
	}

	if (event != -1) {
		r = xgpu_nv_poll_msg(adev, event);
		if (r) {
			if (req != IDH_REQ_GPU_INIT_DATA) {
				pr_err("Doesn't get msg:%d from pf, error=%d\n", event, r);
				return r;
			}
			else /* host doesn't support REQ_GPU_INIT_DATA handshake */
				adev->virt.req_init_data_ver = 0;
		} else {
			if (req == IDH_REQ_GPU_INIT_DATA)
			{
				adev->virt.req_init_data_ver =
					RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW1);

				/* assume V1 in case host doesn't set version number */
				if (adev->virt.req_init_data_ver < 1)
					adev->virt.req_init_data_ver = 1;
			}
		}

		/* Retrieve checksum from mailbox2 */
		if (req == IDH_REQ_GPU_INIT_ACCESS || req == IDH_REQ_GPU_RESET_ACCESS) {
			adev->virt.fw_reserve.checksum_key =
				RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW2);
		}
	}

	return 0;
}

static int xgpu_nv_request_reset(struct amdgpu_device *adev)
{
	return xgpu_nv_send_access_requests(adev, IDH_REQ_GPU_RESET_ACCESS);
}

static int xgpu_nv_request_full_gpu_access(struct amdgpu_device *adev,
					   bool init)
{
	enum idh_request req;

	req = init ? IDH_REQ_GPU_INIT_ACCESS : IDH_REQ_GPU_FINI_ACCESS;
	return xgpu_nv_send_access_requests(adev, req);
}

static int xgpu_nv_release_full_gpu_access(struct amdgpu_device *adev,
					   bool init)
{
	enum idh_request req;
	int r = 0;

	req = init ? IDH_REL_GPU_INIT_ACCESS : IDH_REL_GPU_FINI_ACCESS;
	r = xgpu_nv_send_access_requests(adev, req);

	return r;
}

static int xgpu_nv_request_init_data(struct amdgpu_device *adev)
{
	return xgpu_nv_send_access_requests(adev, IDH_REQ_GPU_INIT_DATA);
}

static int xgpu_nv_mailbox_ack_irq(struct amdgpu_device *adev,
					struct amdgpu_irq_src *source,
					struct amdgpu_iv_entry *entry)
{
	DRM_DEBUG("get ack intr and do nothing.\n");
	return 0;
}

static int xgpu_nv_set_mailbox_ack_irq(struct amdgpu_device *adev,
					struct amdgpu_irq_src *source,
					unsigned type,
					enum amdgpu_interrupt_state state)
{
	u32 tmp = RREG32_NO_KIQ(mmMAILBOX_INT_CNTL);

	if (state == AMDGPU_IRQ_STATE_ENABLE)
		tmp |= 2;
	else
		tmp &= ~2;

	WREG32_NO_KIQ(mmMAILBOX_INT_CNTL, tmp);

	return 0;
}

static void xgpu_nv_mailbox_flr_work(struct work_struct *work)
{
	struct amdgpu_virt *virt = container_of(work, struct amdgpu_virt, flr_work);
	struct amdgpu_device *adev = container_of(virt, struct amdgpu_device, virt);
	int timeout = NV_MAILBOX_POLL_FLR_TIMEDOUT;
	int locked;

	/* block amdgpu_gpu_recover till msg FLR COMPLETE received,
	 * otherwise the mailbox msg will be ruined/reseted by
	 * the VF FLR.
	 *
	 * we can unlock the lock_reset to allow "amdgpu_job_timedout"
	 * to run gpu_recover() after FLR_NOTIFICATION_CMPL received
	 * which means host side had finished this VF's FLR.
	 */
	locked = mutex_trylock(&adev->lock_reset);
	if (locked)
		adev->in_gpu_reset = true;

	do {
		if (xgpu_nv_mailbox_peek_msg(adev) == IDH_FLR_NOTIFICATION_CMPL)
			goto flr_done;

		msleep(10);
		timeout -= 10;
	} while (timeout > 1);

flr_done:
	if (locked) {
		adev->in_gpu_reset = false;
		mutex_unlock(&adev->lock_reset);
	}

	/* Trigger recovery for world switch failure if no TDR */
	if (amdgpu_device_should_recover_gpu(adev)
		&& (adev->sdma_timeout == MAX_SCHEDULE_TIMEOUT ||
		adev->gfx_timeout == MAX_SCHEDULE_TIMEOUT ||
		adev->compute_timeout == MAX_SCHEDULE_TIMEOUT ||
		adev->video_timeout == MAX_SCHEDULE_TIMEOUT))
		amdgpu_device_gpu_recover(adev, NULL);
}

static int xgpu_nv_set_mailbox_rcv_irq(struct amdgpu_device *adev,
				       struct amdgpu_irq_src *src,
				       unsigned type,
				       enum amdgpu_interrupt_state state)
{
	u32 tmp = RREG32_NO_KIQ(mmMAILBOX_INT_CNTL);

	if (state == AMDGPU_IRQ_STATE_ENABLE)
		tmp |= 1;
	else
		tmp &= ~1;

	WREG32_NO_KIQ(mmMAILBOX_INT_CNTL, tmp);

	return 0;
}

static int xgpu_nv_mailbox_rcv_irq(struct amdgpu_device *adev,
				   struct amdgpu_irq_src *source,
				   struct amdgpu_iv_entry *entry)
{
	enum idh_event event = xgpu_nv_mailbox_peek_msg(adev);

	switch (event) {
	case IDH_FLR_NOTIFICATION:
		if (amdgpu_sriov_runtime(adev))
			schedule_work(&adev->virt.flr_work);
		break;
		/* READY_TO_ACCESS_GPU is fetched by kernel polling, IRQ can ignore
		 * it byfar since that polling thread will handle it,
		 * other msg like flr complete is not handled here.
		 */
	case IDH_CLR_MSG_BUF:
	case IDH_FLR_NOTIFICATION_CMPL:
	case IDH_READY_TO_ACCESS_GPU:
	default:
		break;
	}

	return 0;
}

static const struct amdgpu_irq_src_funcs xgpu_nv_mailbox_ack_irq_funcs = {
	.set = xgpu_nv_set_mailbox_ack_irq,
	.process = xgpu_nv_mailbox_ack_irq,
};

static const struct amdgpu_irq_src_funcs xgpu_nv_mailbox_rcv_irq_funcs = {
	.set = xgpu_nv_set_mailbox_rcv_irq,
	.process = xgpu_nv_mailbox_rcv_irq,
};

void xgpu_nv_mailbox_set_irq_funcs(struct amdgpu_device *adev)
{
	adev->virt.ack_irq.num_types = 1;
	adev->virt.ack_irq.funcs = &xgpu_nv_mailbox_ack_irq_funcs;
	adev->virt.rcv_irq.num_types = 1;
	adev->virt.rcv_irq.funcs = &xgpu_nv_mailbox_rcv_irq_funcs;
}

int xgpu_nv_mailbox_add_irq_id(struct amdgpu_device *adev)
{
	int r;

	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 135, &adev->virt.rcv_irq);
	if (r)
		return r;

	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_BIF, 138, &adev->virt.ack_irq);
	if (r) {
		amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
		return r;
	}

	return 0;
}

int xgpu_nv_mailbox_get_irq(struct amdgpu_device *adev)
{
	int r;

	r = amdgpu_irq_get(adev, &adev->virt.rcv_irq, 0);
	if (r)
		return r;
	r = amdgpu_irq_get(adev, &adev->virt.ack_irq, 0);
	if (r) {
		amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
		return r;
	}

	INIT_WORK(&adev->virt.flr_work, xgpu_nv_mailbox_flr_work);

	return 0;
}

void xgpu_nv_mailbox_put_irq(struct amdgpu_device *adev)
{
	amdgpu_irq_put(adev, &adev->virt.ack_irq, 0);
	amdgpu_irq_put(adev, &adev->virt.rcv_irq, 0);
}

const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
	.req_full_gpu	= xgpu_nv_request_full_gpu_access,
	.rel_full_gpu	= xgpu_nv_release_full_gpu_access,
	.req_init_data  = xgpu_nv_request_init_data,
	.reset_gpu = xgpu_nv_request_reset,
	.wait_reset = NULL,
	.trans_msg = xgpu_nv_mailbox_trans_msg,
};