You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey, I'm debugging the performance of cuMemGetHandleForAddressRange, sometimes it takes longer time than expected when registering dma buff. Upon on further investigation, I found the problem within nv_dma_buf_export.
But inside the kernel-open/nvidia/nv-dmabuf.c, only a few functions are available for tracing, I was wondering how to enable the traces for all functions inside this file? like all the function called in nv_dma_buf_create. and even to trace the potential lock contention problem.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hey, I'm debugging the performance of cuMemGetHandleForAddressRange, sometimes it takes longer time than expected when registering dma buff. Upon on further investigation, I found the problem within
nv_dma_buf_export
.But inside the
kernel-open/nvidia/nv-dmabuf.c
, only a few functions are available for tracing, I was wondering how to enable the traces for all functions inside this file? like all the function called innv_dma_buf_create
. and even to trace the potential lock contention problem.Thanks!
cc: @mtijanic could you please take a look? could we do similar thing in #675?
Beta Was this translation helpful? Give feedback.
All reactions