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
|
/*
* Copyright (c) 2014-2018, NVIDIA CORPORATION. All rights reserved.
*
* 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 AUTHORS OR COPYRIGHT HOLDERS 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 <nvgpu/enabled.h>
#include <nvgpu/bus.h>
#include <nvgpu/vgpu/vgpu_ivc.h>
#include <nvgpu/vgpu/vgpu.h>
#include "gk20a/gk20a.h"
#include "fecs_trace_vgpu.h"
int vgpu_comm_init(struct gk20a *g)
{
size_t queue_sizes[] = { TEGRA_VGPU_QUEUE_SIZES };
return vgpu_ivc_init(g, 3, queue_sizes, TEGRA_VGPU_QUEUE_CMD,
ARRAY_SIZE(queue_sizes));
}
void vgpu_comm_deinit(void)
{
size_t queue_sizes[] = { TEGRA_VGPU_QUEUE_SIZES };
vgpu_ivc_deinit(TEGRA_VGPU_QUEUE_CMD, ARRAY_SIZE(queue_sizes));
}
int vgpu_comm_sendrecv(struct tegra_vgpu_cmd_msg *msg, size_t size_in,
size_t size_out)
{
void *handle;
size_t size = size_in;
void *data = msg;
int err;
err = vgpu_ivc_sendrecv(vgpu_ivc_get_server_vmid(),
TEGRA_VGPU_QUEUE_CMD, &handle, &data, &size);
if (!err) {
WARN_ON(size < size_out);
memcpy(msg, data, size_out);
vgpu_ivc_release(handle);
}
return err;
}
u64 vgpu_connect(void)
{
struct tegra_vgpu_cmd_msg msg;
struct tegra_vgpu_connect_params *p = &msg.params.connect;
int err;
msg.cmd = TEGRA_VGPU_CMD_CONNECT;
p->module = TEGRA_VGPU_MODULE_GPU;
err = vgpu_comm_sendrecv(&msg, sizeof(msg), sizeof(msg));
return (err || msg.ret) ? 0 : p->handle;
}
int vgpu_get_attribute(u64 handle, u32 attrib, u32 *value)
{
struct tegra_vgpu_cmd_msg msg;
struct tegra_vgpu_attrib_params *p = &msg.params.attrib;
int err;
msg.cmd = TEGRA_VGPU_CMD_GET_ATTRIBUTE;
msg.handle = handle;
p->attrib = attrib;
err = vgpu_comm_sendrecv(&msg, sizeof(msg), sizeof(msg));
if (err || msg.ret)
return -1;
*value = p->value;
return 0;
}
static void vgpu_handle_channel_event(struct gk20a *g,
struct tegra_vgpu_channel_event_info *info)
{
struct tsg_gk20a *tsg;
if (!info->is_tsg) {
nvgpu_err(g, "channel event posted");
return;
}
if (info->id >= g->fifo.num_channels ||
info->event_id >= TEGRA_VGPU_CHANNEL_EVENT_ID_MAX) {
nvgpu_err(g, "invalid channel event");
return;
}
tsg = &g->fifo.tsg[info->id];
gk20a_tsg_event_id_post_event(tsg, info->event_id);
}
int vgpu_intr_thread(void *dev_id)
{
struct gk20a *g = dev_id;
struct vgpu_priv_data *priv = vgpu_get_priv_data(g);
while (true) {
struct tegra_vgpu_intr_msg *msg;
u32 sender;
void *handle;
size_t size;
int err;
err = vgpu_ivc_recv(TEGRA_VGPU_QUEUE_INTR, &handle,
(void **)&msg, &size, &sender);
if (err == -ETIME)
continue;
if (WARN_ON(err))
continue;
if (msg->event == TEGRA_VGPU_EVENT_ABORT) {
vgpu_ivc_release(handle);
break;
}
switch (msg->event) {
case TEGRA_VGPU_EVENT_INTR:
if (msg->unit == TEGRA_VGPU_INTR_GR)
vgpu_gr_isr(g, &msg->info.gr_intr);
else if (msg->unit == TEGRA_VGPU_NONSTALL_INTR_GR)
vgpu_gr_nonstall_isr(g,
&msg->info.gr_nonstall_intr);
else if (msg->unit == TEGRA_VGPU_INTR_FIFO)
vgpu_fifo_isr(g, &msg->info.fifo_intr);
else if (msg->unit == TEGRA_VGPU_NONSTALL_INTR_FIFO)
vgpu_fifo_nonstall_isr(g,
&msg->info.fifo_nonstall_intr);
else if (msg->unit == TEGRA_VGPU_NONSTALL_INTR_CE2)
vgpu_ce2_nonstall_isr(g,
&msg->info.ce2_nonstall_intr);
break;
#ifdef CONFIG_GK20A_CTXSW_TRACE
case TEGRA_VGPU_EVENT_FECS_TRACE:
vgpu_fecs_trace_data_update(g);
break;
#endif
case TEGRA_VGPU_EVENT_CHANNEL:
vgpu_handle_channel_event(g, &msg->info.channel_event);
break;
case TEGRA_VGPU_EVENT_SM_ESR:
vgpu_gr_handle_sm_esr_event(g, &msg->info.sm_esr);
break;
default:
nvgpu_err(g, "unknown event %u", msg->event);
break;
}
vgpu_ivc_release(handle);
}
while (!nvgpu_thread_should_stop(&priv->intr_handler))
nvgpu_msleep(10);
return 0;
}
void vgpu_remove_support_common(struct gk20a *g)
{
struct vgpu_priv_data *priv = vgpu_get_priv_data(g);
struct tegra_vgpu_intr_msg msg;
int err;
if (g->dbg_regops_tmp_buf)
nvgpu_kfree(g, g->dbg_regops_tmp_buf);
if (g->pmu.remove_support)
g->pmu.remove_support(&g->pmu);
if (g->gr.remove_support)
g->gr.remove_support(&g->gr);
if (g->fifo.remove_support)
g->fifo.remove_support(&g->fifo);
if (g->mm.remove_support)
g->mm.remove_support(&g->mm);
msg.event = TEGRA_VGPU_EVENT_ABORT;
err = vgpu_ivc_send(vgpu_ivc_get_peer_self(), TEGRA_VGPU_QUEUE_INTR,
&msg, sizeof(msg));
WARN_ON(err);
nvgpu_thread_stop(&priv->intr_handler);
}
void vgpu_detect_chip(struct gk20a *g)
{
struct nvgpu_gpu_params *p = &g->params;
struct vgpu_priv_data *priv = vgpu_get_priv_data(g);
p->gpu_arch = priv->constants.arch;
p->gpu_impl = priv->constants.impl;
p->gpu_rev = priv->constants.rev;
gk20a_dbg_info("arch: %x, impl: %x, rev: %x\n",
p->gpu_arch,
p->gpu_impl,
p->gpu_rev);
}
int vgpu_init_gpu_characteristics(struct gk20a *g)
{
int err;
gk20a_dbg_fn("");
err = gk20a_init_gpu_characteristics(g);
if (err)
return err;
__nvgpu_set_enabled(g, NVGPU_SUPPORT_MAP_BUFFER_BATCH, false);
/* features vgpu does not support */
__nvgpu_set_enabled(g, NVGPU_SUPPORT_RESCHEDULE_RUNLIST, false);
return 0;
}
int vgpu_read_ptimer(struct gk20a *g, u64 *value)
{
struct tegra_vgpu_cmd_msg msg = {0};
struct tegra_vgpu_read_ptimer_params *p = &msg.params.read_ptimer;
int err;
gk20a_dbg_fn("");
msg.cmd = TEGRA_VGPU_CMD_READ_PTIMER;
msg.handle = vgpu_get_handle(g);
err = vgpu_comm_sendrecv(&msg, sizeof(msg), sizeof(msg));
err = err ? err : msg.ret;
if (!err)
*value = p->time;
else
nvgpu_err(g, "vgpu read ptimer failed, err=%d", err);
return err;
}
int vgpu_get_timestamps_zipper(struct gk20a *g,
u32 source_id, u32 count,
struct nvgpu_cpu_time_correlation_sample *samples)
{
struct tegra_vgpu_cmd_msg msg = {0};
struct tegra_vgpu_get_timestamps_zipper_params *p =
&msg.params.get_timestamps_zipper;
int err;
u32 i;
gk20a_dbg_fn("");
if (count > TEGRA_VGPU_GET_TIMESTAMPS_ZIPPER_MAX_COUNT) {
nvgpu_err(g, "count %u overflow", count);
return -EINVAL;
}
msg.cmd = TEGRA_VGPU_CMD_GET_TIMESTAMPS_ZIPPER;
msg.handle = vgpu_get_handle(g);
p->source_id = TEGRA_VGPU_GET_TIMESTAMPS_ZIPPER_SRC_ID_TSC;
p->count = count;
err = vgpu_comm_sendrecv(&msg, sizeof(msg), sizeof(msg));
err = err ? err : msg.ret;
if (err) {
nvgpu_err(g, "vgpu get timestamps zipper failed, err=%d", err);
return err;
}
for (i = 0; i < count; i++) {
samples[i].cpu_timestamp = p->samples[i].cpu_timestamp;
samples[i].gpu_timestamp = p->samples[i].gpu_timestamp;
}
return err;
}
int vgpu_init_hal(struct gk20a *g)
{
u32 ver = g->params.gpu_arch + g->params.gpu_impl;
int err;
switch (ver) {
case NVGPU_GPUID_GP10B:
gk20a_dbg_info("gp10b detected");
err = vgpu_gp10b_init_hal(g);
break;
case NVGPU_GPUID_GV11B:
err = vgpu_gv11b_init_hal(g);
break;
default:
nvgpu_err(g, "no support for %x", ver);
err = -ENODEV;
break;
}
return err;
}
int vgpu_get_constants(struct gk20a *g)
{
struct tegra_vgpu_cmd_msg msg = {};
struct tegra_vgpu_constants_params *p = &msg.params.constants;
struct vgpu_priv_data *priv = vgpu_get_priv_data(g);
int err;
gk20a_dbg_fn("");
msg.cmd = TEGRA_VGPU_CMD_GET_CONSTANTS;
msg.handle = vgpu_get_handle(g);
err = vgpu_comm_sendrecv(&msg, sizeof(msg), sizeof(msg));
err = err ? err : msg.ret;
if (unlikely(err)) {
nvgpu_err(g, "%s failed, err=%d", __func__, err);
return err;
}
if (unlikely(p->gpc_count > TEGRA_VGPU_MAX_GPC_COUNT ||
p->max_tpc_per_gpc_count > TEGRA_VGPU_MAX_TPC_COUNT_PER_GPC)) {
nvgpu_err(g, "gpc_count %d max_tpc_per_gpc %d overflow",
(int)p->gpc_count, (int)p->max_tpc_per_gpc_count);
return -EINVAL;
}
priv->constants = *p;
return 0;
}
|