nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-24 10:34:43 +03:00
Code Issues Packages Projects Releases Wiki Activity

gpu: nvgpu: hal for timestamps correlation

Browse Source 
In order to perform timestamps correlation for FECS
traces, we need to collect GPU / GPU timestamps
samples. In virtualization case, it is possible for
a guest to get GPU timestamps by using read_ptimer.
However, if the CPU timestamp is read on guest side,
and the GPU timestamp is read on vm-server side,
then it introduces some latency that will create an
artificial offset for GPU timestamps (~2 us in
average). For better CPU / GPU timestamps correlation,
Added a command to collect all timestamps on vm-server
side.

Bug 1900475

Change-Id: Idfdc6ae4c16c501dc5e00053a5b75932c55148d6
Signed-off-by: Thomas Fleury <tfleury@nvidia.com>
Reviewed-on: http://git-master/r/1472447
Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Aparna Das <aparnad@nvidia.com>
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
This commit is contained in:
Thomas Fleury
2017-04-27 11:28:27 -07:00
committed by mobile promotions
parent 70f507eec7
commit 56f56b5cd9
9 changed files with 160 additions and 50 deletions
Download Patch File Download Diff File Expand all files Collapse all files

111
drivers/gpu/nvgpu/gk20a/ctrl_gk20a.c
View File

@@ -643,58 +643,17 @@ static inline u64 get_cpu_timestamp_timeofday(void)
return timeval_to_jiffies(&tv);
}
static inline int get_timestamps_zipper(struct gk20a *g,
u64 (*get_cpu_timestamp)(void),
struct nvgpu_gpu_get_cpu_time_correlation_info_args *args)
int gk20a_get_timestamps_zipper(struct gk20a *g,
u32 source_id, u32 count,
struct nvgpu_cpu_time_correlation_sample *samples)
{
int err = 0;
unsigned int i = 0;
u32 gpu_timestamp_hi_new = 0;
u32 gpu_timestamp_hi_old = 0;
if (gk20a_busy(g)) {
gk20a_err(dev_from_gk20a(g), "GPU not powered on\n");
err = -EINVAL;
goto end;
}
/* get zipper reads of gpu and cpu counter values */
gpu_timestamp_hi_old = gk20a_readl(g, timer_time_1_r());
for (i = 0; i < args->count; i++) {
u32 gpu_timestamp_lo = 0;
u32 gpu_timestamp_hi = 0;
gpu_timestamp_lo = gk20a_readl(g, timer_time_0_r());
args->samples[i].cpu_timestamp = get_cpu_timestamp();
rmb(); /* maintain zipper read order */
gpu_timestamp_hi_new = gk20a_readl(g, timer_time_1_r());
/* pick the appropriate gpu counter hi bits */
gpu_timestamp_hi = (gpu_timestamp_lo & (1L << 31)) ?
gpu_timestamp_hi_old : gpu_timestamp_hi_new;
args->samples[i].gpu_timestamp =
((u64)gpu_timestamp_hi << 32) | (u64)gpu_timestamp_lo;
gpu_timestamp_hi_old = gpu_timestamp_hi_new;
}
end:
gk20a_idle(g);
return err;
}
static int nvgpu_gpu_get_cpu_time_correlation_info(
struct gk20a *g,
struct nvgpu_gpu_get_cpu_time_correlation_info_args *args)
{
int err = 0;
u64 (*get_cpu_timestamp)(void) = NULL;
if (args->count > NVGPU_GPU_GET_CPU_TIME_CORRELATION_INFO_MAX_COUNT)
return -EINVAL;
switch (args->source_id) {
switch (source_id) {
case NVGPU_GPU_GET_CPU_TIME_CORRELATION_INFO_SRC_ID_TSC:
get_cpu_timestamp = get_cpu_timestamp_tsc;
break;
@@ -709,7 +668,67 @@ static int nvgpu_gpu_get_cpu_time_correlation_info(
return -EINVAL;
}
err = get_timestamps_zipper(g, get_cpu_timestamp, args);
if (gk20a_busy(g)) {
gk20a_err(dev_from_gk20a(g), "GPU not powered on\n");
err = -EINVAL;
goto end;
}
/* get zipper reads of gpu and cpu counter values */
gpu_timestamp_hi_old = gk20a_readl(g, timer_time_1_r());
for (i = 0; i < count; i++) {
u32 gpu_timestamp_lo = 0;
u32 gpu_timestamp_hi = 0;
gpu_timestamp_lo = gk20a_readl(g, timer_time_0_r());
samples[i].cpu_timestamp = get_cpu_timestamp();
rmb(); /* maintain zipper read order */
gpu_timestamp_hi_new = gk20a_readl(g, timer_time_1_r());
/* pick the appropriate gpu counter hi bits */
gpu_timestamp_hi = (gpu_timestamp_lo & (1L << 31)) ?
gpu_timestamp_hi_old : gpu_timestamp_hi_new;
samples[i].gpu_timestamp =
((u64)gpu_timestamp_hi << 32) | (u64)gpu_timestamp_lo;
gpu_timestamp_hi_old = gpu_timestamp_hi_new;
}
end:
gk20a_idle(g);
return err;
}
static int nvgpu_gpu_get_cpu_time_correlation_info(
struct gk20a *g,
struct nvgpu_gpu_get_cpu_time_correlation_info_args *args)
{
struct nvgpu_cpu_time_correlation_sample *samples;
int err;
u32 i;
if (args->count >= NVGPU_GPU_GET_CPU_TIME_CORRELATION_INFO_MAX_COUNT) {
return -EINVAL;
}
samples = kzalloc(args->count *
sizeof(struct nvgpu_cpu_time_correlation_sample), GFP_KERNEL);
if (!samples) {
return -ENOMEM;
}
err = g->ops.bus.get_timestamps_zipper(g,
args->source_id, args->count, samples);
if (!err) {
for (i = 0; i < args->count; i++) {
args->samples[i].cpu_timestamp = samples[i].cpu_timestamp;
args->samples[i].gpu_timestamp = samples[i].gpu_timestamp;
}
}
kfree(samples);
return err;
}