I find a similar question.
https://forums.developer.nvidia.com/t/kernel-call-trace-observed-when-calling-cudafreehost-cudahostalloc-for-buffers-on-amd-cpu-with-nvi/72930
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Kernel call trace observed when calling cudaFreeHost/cudaHostAlloc for buffers on AMD CPU with NVI... | 5 | 2108 | April 19, 2019 | |
| Pinned memory limit | 16 | 14075 | May 1, 2016 | |
| Registering Mapped Linux Character Device Memory with cudaHostRegister Results in Invalid Argument | 2 | 2976 | June 27, 2017 | |
| cudaHostRegister crash or poor performance unknow error (30) in kernel for 64bit host operating syst | 23 | 6036 | May 8, 2012 | |
| pci_map_sg sometimes leads to kernel-oops | 8 | 2139 | December 21, 2016 | |
| cudaFreeHost very slow for buffers > 2GB [newbee here!] | 30 | 2847 | October 3, 2019 | |
| iommu unhandled context fault on PCI device DMA | 20 | 9703 | September 11, 2018 | |
| cudaHostAlloc --> invalid argument it works with fermi, not with 9800gtx | 5 | 6722 | November 11, 2011 | |
| CUDA 4.0: linux malloc for page-aligned memory and cudaHostRegister How to malloc page-aligned memor | 9 | 19499 | March 11, 2011 | |
| cudaHostAlloc can only allocate about 3.5GB of memory out of 128GB | 7 | 624 | June 2, 2023 |