[PATCH] drm/amdgpu: limit the number of dst address in trace
Christian König
christian.koenig at amd.com
Mon Jan 17 07:24:45 UTC 2022
Am 17.01.22 um 07:32 schrieb Somalapuram Amaranath:
> trace_amdgpu_vm_update_ptes trace unable to log when nptes too large
>
> Signed-off-by: Somalapuram Amaranath <Amaranath.Somalapuram at amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index 0e7dc23f78e7..f367fcc5c535 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -1634,7 +1634,7 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
> nptes = max(nptes, 1u);
>
> trace_amdgpu_vm_update_ptes(params, frag_start, upd_end,
> - nptes, dst, incr, upd_flags,
> + min(nptes, 32u), dst, incr, upd_flags,
You still need to remove the same workaround in amdgpu_trace.h.
E.g. something like this:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
index d855cb53c7e0..a48b34d4ce31 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -358,11 +358,10 @@ TRACE_EVENT(amdgpu_vm_update_ptes,
}
),
TP_printk("pid:%u vm_ctx:0x%llx start:0x%010llx end:0x%010llx,"
- " flags:0x%llx, incr:%llu, dst:\n%s%s", __entry->pid,
+ " flags:0x%llx, incr:%llu, dst:\n%s", __entry->pid,
__entry->vm_ctx, __entry->start, __entry->end,
__entry->flags, __entry->incr, __print_array(
- __get_dynamic_array(dst), min(__entry->nptes, 32u), 8),
- __entry->nptes > 32 ? "..." : "")
+ __get_dynamic_array(dst), __entry->nptes, 8))
);
TRACE_EVENT(amdgpu_vm_set_ptes,
Regards,
Christian.
> vm->task_info.pid,
> vm->immediate.fence_context);
> amdgpu_vm_update_flags(params, to_amdgpu_bo_vm(pt),
More information about the amd-gfx
mailing list