On Wed, Apr 3, 2024 at 11:54 AM Sunil V L <sunilvl@xxxxxxxxxxxxxxxx> wrote: > > On Mon, Apr 01, 2024 at 04:57:30PM +0000, Luck, Tony wrote: > > >> If the firmware didn't provide the SRAT/SLIT on ACPI based systems, then > > >> there will be a message "Failed to initialise from firmware" from > > >> arch_acpi_numa_init(). This is not specific to RISC-V. But I am > > >> wondering why should it be pr_info instead of pr_debug. > > >> > > > > > > My understanding is maybe it just wants to expose explicit logs to > > > avoid any potential bugs from FW or Kernel. > > > > There are lots of ACPI enabled systems that aren't NUMA (single > > socket servers, desktops, laptops). Making this "pr_info()" would just > > add noise to the boot on all of those. > > > Exactly. But this is an existing pr_info message across architectures. > My suggestion is to add one more patch in this series to convert > this to pr_debug unless someone has strong reason to keep it pr_info. > Sure. I will add a patch to fix it in v3. Thanks, Haibo > Thanks, > Sunil
