[Mesa-dev] [PATCH] radv: Actually check for vm faults.

[Samuel Pitoiset]

I have something similar locally. :)

Reviewed-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>

On 09/07/2017 10:23 PM, Bas Nieuwenhuizen wrote:
> The code can check for vm faults having happened. If we only do it
> on a hang we don't know when the faults happened. This changes the
> behavior to when the first VM faults is found, even without a hang.
> ---
>   src/amd/vulkan/radv_debug.c  | 10 +++++++---
>   src/amd/vulkan/radv_debug.h  |  1 +
>   src/amd/vulkan/radv_device.c |  1 +
>   3 files changed, 9 insertions(+), 3 deletions(-)
> 
> diff --git a/src/amd/vulkan/radv_debug.c b/src/amd/vulkan/radv_debug.c
> index 949eeea2f36..4b70c8a141c 100644
> --- a/src/amd/vulkan/radv_debug.c
> +++ b/src/amd/vulkan/radv_debug.c
> @@ -87,11 +87,15 @@ radv_check_gpu_hangs(struct radv_queue *queue, struct radeon_winsys_cs *cs)
>   	struct radv_device *device = queue->device;
>   	uint64_t addr;
>   
> -	if (!radv_gpu_hang_occured(queue))
> +	bool hang_occurred = radv_gpu_hang_occured(queue);
> +	bool vm_fault_occurred = false;
> +	if (queue->device->instance->debug_flags & RADV_DEBUG_VM_FAULTS)
> +		vm_fault_occurred = ac_vm_fault_occured(device->physical_device->rad_info.chip_class,
> +		                                        &device->dmesg_timestamp, &addr);
> +	if (!hang_occurred && !vm_fault_occurred)
>   		return;
>   
> -	if (ac_vm_fault_occured(device->physical_device->rad_info.chip_class,
> -				&device->dmesg_timestamp, &addr)) {
> +	if (vm_fault_occurred) {
>   		fprintf(stderr, "VM fault report.\n\n");
>   		fprintf(stderr, "Failing VM page: 0x%08"PRIx64"\n\n", addr);
>   	}
> diff --git a/src/amd/vulkan/radv_debug.h b/src/amd/vulkan/radv_debug.h
> index 240ce162047..18ef7e7e0a8 100644
> --- a/src/amd/vulkan/radv_debug.h
> +++ b/src/amd/vulkan/radv_debug.h
> @@ -38,6 +38,7 @@ enum {
>   	RADV_DEBUG_ALL_BOS           = 0x100,
>   	RADV_DEBUG_NO_IBS            = 0x200,
>   	RADV_DEBUG_DUMP_SPIRV        = 0x400,
> +	RADV_DEBUG_VM_FAULTS         = 0x800,
>   };
>   
>   enum {
> diff --git a/src/amd/vulkan/radv_device.c b/src/amd/vulkan/radv_device.c
> index bf7552e63da..4f4fa20209b 100644
> --- a/src/amd/vulkan/radv_device.c
> +++ b/src/amd/vulkan/radv_device.c
> @@ -410,6 +410,7 @@ static const struct debug_control radv_debug_options[] = {
>   	{"allbos", RADV_DEBUG_ALL_BOS},
>   	{"noibs", RADV_DEBUG_NO_IBS},
>   	{"spirv", RADV_DEBUG_DUMP_SPIRV},
> +	{"vmfaults", RADV_DEBUG_VM_FAULTS},
>   	{NULL, 0}
>   };
>   
>