On Tue, Jul 30, 2024 at 03:19:30PM +0200, Borislav Petkov wrote:

On Mon, Jul 29, 2024 at 11:58:50AM -0300, Jason Gunthorpe wrote:

quoted

Like continuous memory scrubbing and EDAC is not really fwctl since it
is part of the main mission of a memory device. However evaluating the
memory to measure current ECC error rate for data collection and
debugging would be appropriate for fwctl.

fwctl?

What's that? Something to do with "firmware"? If so, how does this thing have
anything to do with RAS stuff?

https://lore.kernel.org/linux-rdma/6-v2-940e479ceba9+3821-fwctl_jgg@nvidia.com/ (local)
https://lore.kernel.org/r/20240718213446.1750135-1-dave.jiang@intel.com (local)

Jonathan is concerned that Dave exposed some of the scrubber FW RPCs
through fwctl.

We have rasdaemon for that where all the RAS controlling and evaluation goes
into.

Sure

Jaso