On Tue, Aug 30, 2022 at 01:59:16PM +0300, Dmitry Kozlyuk wrote: > Thank you for the most detailed info! > > 1. If you run the poisoner program the second time, > does it also see dirty memory immediately after mmap()?
No, running the poisoner program again doesn't show dirty memory immediately after mmap. I assume that there are some differences on how my poisoner program using hugepage and how testpmd using hugepage. I notice that testpmd leaves some files under /dev/hugepages/rtemap_xxx even after testpmd process exits. But my program didn't create any file under /dev/hugepages/. > > 2. Kernel 4.19.90-2102 patchlevel 2102 is very high, > can there be any unusual patches applied? > Your host has "compute" in its name, > can it have patches that trade security for performance? I may need to talk to the kernel team. Indeed, I failed to reproduce the issue on a Virtual Machine of kernel 4.18.0-305.12.1.el8_4.x86_64.(2M page size insetad of 1G)
