Re: [PATCH v2 3/4] efi/arm: map UEFI memory map earlier on boot

From: Will Deacon
Date: Fri Jul 06 2018 - 09:37:28 EST


Hi Akashi,

On Fri, Jul 06, 2018 at 10:33:13AM +0900, AKASHI Takahiro wrote:
> On Fri, Jul 06, 2018 at 09:42:28AM +0900, AKASHI Takahiro wrote:
> > I almost believed that my patch#2 was just a preparatory one for patch#3
> > where arm_enable_runtime_services() is moved aggressively forward.
> > But acpi_os_ioremap() is not a __init function and I can now agree to
> > keeping patch#2.
> >
> > Meanwhile, the consequent code with Ard's patch would look like:
> > ---8<---
> > static int __init arm_enable_runtime_services(void)
> > {
> > ...
> > efi_memmap_unmap();
> >
> > mapsize = efi.memmap.desc_size * efi.memmap.nr_map;
> >
> > if (efi_memmap_init_late(efi.memmap.phys_map, mapsize)) {
> > pr_err("Failed to remap EFI memory map\n");
> > return 0;
> > }
> > ...
> > }
> > --->8---
> > It seems to me that it makes no sense.
>
> Oops, it does. Comments at efi_memmap_init_late() say:
> ---8<---
> * The reason there are two EFI memmap initialisation
> * (efi_memmap_init_early() and this late version) is because the
> * early EFI memmap should be explicitly unmapped once EFI
> * initialisation is complete as the fixmap space used to map the EFI
> * memmap (via early_memremap()) is a scarce resource.
> --->8---
>
> > Is it okay to take them out?
>
> Never mind.

I'm struggling with your monologue...

Please can you send a v3 of the series, containing the patches that you
think are necessary, along with the Acks you've collected?

Thanks,

Will