We have a complicated pipeline which in most time can run smoothly.
But somehow we observed following error message and pipeline stopped.
2022-02-16 10:09:14.566683 13437-19700 nvvideoconvert ERROR: buffer transform failed 2022-02-16 10:09:14.566694 13437-19709 nvinfer WARN : error: NvBufSurfTransform failed with error -2 while converting buffer 2022-02-16 10:09:14.567077 13437-19709 GST_ERROR_SYSTEM INFO : posting message: NvBufSurfTransform failed with error -2 while converting buffer 2022-02-16 10:09:14.567397 13437-19709 GST_ERROR_SYSTEM INFO : posted error message: NvBufSurfTransform failed with error -2 while converting bufferPlease provide complete information as applicable to your setup.
It is not easy to reproduce.
But I’d like to know what is the possible reason causing the error?
Is there some easy way to help debug this issue if we meet it in future?
Thanks a lot!
• Hardware Platform (Jetson / GPU) Jetson Xavier Nx • DeepStream Version 6.0 • JetPack Version (valid for Jetson only) 4.6 • TensorRT Version • NVIDIA GPU Driver Version (valid for GPU only) • Issue Type( questions, new requirements, bugs) questions • How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing) • Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)
Hi,
This should not happen and may be triggered by memory corruption. Seems like the data of the buffers is not correctly read. For further investigation we would need to reproduce it and check further. Does it happen in specific time(such as beginnig or termination) or randomly? How long would it take to see the issue(such as see once in 30-minute run)? Does it happen in single input source or multiple sources?
BTW, we have another issue which I am not sure if it has connection with this one.
In our APP, we repeatedly run our pipeline and then stop it after a while.
After several rounds, we faced a NvBufferDestroy abort issue in mymediacompositor plugin when stopping the pipeline.
Plugin mymediacompositor is derived from nvmediacompositor plugin.
From callstack and disassembly code, we can see a valid dmabuf_fd was passed to NvBufferDestroy in function gst_nv_compositor_memory_allocator_free().
But somehow NvBufferDestroy called free() with NULL pointer which led to abort() was called by libc.
Hi,
Do you mean nvcompositor? We shall not have nvmediacompositor plugin. And the nvcompositor does not works in DeepStream SDK. You would need to use nvmultistreamtiler.
Hi,
The nvmultistreamtiler plugin is not open source. There are multiple samples demonstrating how to link to the plugin and set the properties. Please take a look.