Hi Lucas
2017-08-08 12:34 GMT+02:00 Lucas Stach <l.stach@xxxxxxxxxxxxxx>:
> Am Samstag, den 22.07.2017, 11:53 +0200 schrieb Christian Gmeiner:
>> In order to support performance counters in a sane way we need to provide
>> a method to sync the GPU with the CPU. The GPU can process multpile command
>> buffers/events per irq. With the help of a 'sync point' we can trigger an event
>> and stop the GPU/FE immediately. When the CPU is done with is processing it
>> simply needs to restart the FE and the GPU will process the command stream.
>>
>> Changes from v1 -> v2:
>> - process sync point with a work item to keep irq as fast as possible
>>
>> Signed-off-by: Christian Gmeiner <christian.gmeiner@xxxxxxxxx>
>> ---
>> drivers/gpu/drm/etnaviv/etnaviv_buffer.c | 36 ++++++++++++++++++++++++++++++++
>> drivers/gpu/drm/etnaviv/etnaviv_drv.h | 1 +
>> drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 25 ++++++++++++++++++++++
>> drivers/gpu/drm/etnaviv/etnaviv_gpu.h | 6 ++++++
>> 4 files changed, 68 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
>> index ed9588f36bc9..9e7098e3207f 100644
>> --- a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
>> +++ b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
>> @@ -250,6 +250,42 @@ void etnaviv_buffer_end(struct etnaviv_gpu *gpu)
>> }
>> }
>>
>> +/* Append a 'sync point' to the ring buffer. */
>> +void etnaviv_sync_point_queue(struct etnaviv_gpu *gpu, unsigned int event)
>> +{
>> + struct etnaviv_cmdbuf *buffer = gpu->buffer;
>> + unsigned int waitlink_offset = buffer->user_size - 16;
>> + u32 dwords, target;
>> +
>> + /*
>> + * We need at most 3 dwords in the return target:
>> + * 1 event + 1 end + 1 wait + 1 link.
>> + */
>> + dwords = 4;
>> + target = etnaviv_buffer_reserve(gpu, buffer, dwords);
>> +
>> + /* Signal sync point event */
>> + CMD_LOAD_STATE(buffer, VIVS_GL_EVENT, VIVS_GL_EVENT_EVENT_ID(event) |
>> + VIVS_GL_EVENT_FROM_PE);
>> +
>> + /* Stop the FE to 'pause' the GPU */
>> + CMD_END(buffer);
>> +
>> + /* Append waitlink */
>> + CMD_WAIT(buffer);
>> + CMD_LINK(buffer, 2, etnaviv_cmdbuf_get_va(buffer) +
>> + buffer->user_size - 4);
>> +
>> + /*
>> + * Kick off the 'sync point' command by replacing the previous
>> + * WAIT with a link to the address in the ring buffer.
>> + */
>> + etnaviv_buffer_replace_wait(buffer, waitlink_offset,
>> + VIV_FE_LINK_HEADER_OP_LINK |
>> + VIV_FE_LINK_HEADER_PREFETCH(dwords),
>> + target);
>> +}
>> +
>> /* Append a command buffer to the ring buffer. */
>> void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, unsigned int event,
>> struct etnaviv_cmdbuf *cmdbuf)
>> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
>> index 058389f93b69..f6cdd694ca51 100644
>> --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h
>> +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
>> @@ -101,6 +101,7 @@ int etnaviv_gem_new_userptr(struct drm_device *dev, struct drm_file *file,
>> u16 etnaviv_buffer_init(struct etnaviv_gpu *gpu);
>> u16 etnaviv_buffer_config_mmuv2(struct etnaviv_gpu *gpu, u32 mtlb_addr, u32 safe_addr);
>> void etnaviv_buffer_end(struct etnaviv_gpu *gpu);
>> +void etnaviv_sync_point_queue(struct etnaviv_gpu *gpu, unsigned int event);
>> void etnaviv_buffer_queue(struct etnaviv_gpu *gpu, unsigned int event,
>> struct etnaviv_cmdbuf *cmdbuf);
>> void etnaviv_validate_init(void);
>> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
>> index ce6c869e214f..bc1b96b4c7e9 100644
>> --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
>> +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
>> @@ -25,6 +25,7 @@
>> #include "etnaviv_gpu.h"
>> #include "etnaviv_gem.h"
>> #include "etnaviv_mmu.h"
>> +#include "etnaviv_perfmon.h"
>> #include "common.xml.h"
>> #include "state.xml.h"
>> #include "state_hi.xml.h"
>> @@ -1353,6 +1354,7 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
>> }
>>
>> gpu->event[event].fence = fence;
>> + gpu->event[event].sync_point = NULL;
>> submit->fence = dma_fence_get(fence);
>> gpu->active_fence = submit->fence->seqno;
>>
>> @@ -1398,6 +1400,23 @@ int etnaviv_gpu_submit(struct etnaviv_gpu *gpu,
>> return ret;
>> }
>>
>> +static void etnaviv_process_sync_point(struct etnaviv_gpu *gpu,
>> + struct etnaviv_event *event)
>> +{
>> + u32 addr = gpu_read(gpu, VIVS_FE_DMA_ADDRESS);
>> +
>> + event->sync_point(gpu, event);
>> + etnaviv_gpu_start_fe(gpu, addr + 2, 2);
>> +}
>> +
>> +static void pmrs_worker(struct work_struct *work)
>> +{
>> + struct etnaviv_gpu *gpu = container_of(work, struct etnaviv_gpu,
>> + pmrs_work);
>> +
>> + etnaviv_process_sync_point(gpu, &gpu->event[gpu->pmrs_event]);
>> +}
>> +
>> /*
>> * Init/Cleanup:
>> */
>> @@ -1444,6 +1463,11 @@ static irqreturn_t irq_handler(int irq, void *data)
>>
>> dev_dbg(gpu->dev, "event %u\n", event);
>>
>> + if (gpu->event[event].sync_point) {
>> + gpu->pmrs_event = event;
>> + etnaviv_queue_work(gpu->drm, &gpu->pmrs_work);
>
> If the handler is delayed we might handle multiple events per
> invocation, in which case the events might not be in order. E.g. the FE
> stop event might be event 30, while the FE start event might be event 0.
> In that case you would execute the FE start before the FE stop has been
> queued -> not good. You need to make sure that your PMRS events are
> processed in the correct order.
>
I thought about this problem for some time and I do not fully get your
point - sorry.
First there is no FE start event. I am using 'sync' points for pre and
post pmrs points.
A 'sync' point is represented like this in the cmd stream:
| EVENT | FE STOP | WAIT | LINK back to WAIT | - see
etnaviv_sync_point_queue(..) in etnaviv_buffer.c
The restarting of the FE happens in pmrs_worker(..) triggered by EVENT
and the FE starts processing the
cmd stream at "| WAIT | LINK back to WAIT |".
Maybe you can point me into the right direction.
>> + }
>> +
>> fence = gpu->event[event].fence;
>> gpu->event[event].fence = NULL;
>> dma_fence_signal(fence);
>> @@ -1647,6 +1671,7 @@ static int etnaviv_gpu_bind(struct device *dev, struct device *master,
>>
>> INIT_LIST_HEAD(&gpu->active_cmd_list);
>> INIT_WORK(&gpu->retire_work, retire_worker);
>> + INIT_WORK(&gpu->pmrs_work, pmrs_worker);
>> INIT_WORK(&gpu->recover_work, recover_worker);
>> init_waitqueue_head(&gpu->fence_event);
>>
>> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
>> index 70e6590aacdf..7d5f785b0f08 100644
>> --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
>> +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
>> @@ -89,6 +89,8 @@ struct etnaviv_chip_identity {
>>
>> struct etnaviv_event {
>> struct dma_fence *fence;
>> +
>> + void (*sync_point)(struct etnaviv_gpu *gpu, struct etnaviv_event *event);
>> };
>>
>> struct etnaviv_cmdbuf_suballoc;
>> @@ -135,6 +137,10 @@ struct etnaviv_gpu {
>> /* worker for handling active-list retiring: */
>> struct work_struct retire_work;
>>
>> + /* worker for handling performance monitor requests: */
>> + struct work_struct pmrs_work;
>> + int pmrs_event;
>> +
>> void __iomem *mmio;
>> int irq;
>>
>
>
greets
--
Christian Gmeiner, MSc
https://christian-gmeiner.info
_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel
