Re: [GIT PULL] MM updates for 6.12-rc1

[Geert Uytterhoeven <geert@linux-m68k.org>]

Hi Linus,

On Sun, Sep 22, 2024 at 6:35 PM Linus Torvalds
<torvalds@linux-foundation.org> wrote:
> On Sun, 22 Sept 2024 at 02:32, Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> >     kernel/resource.c: In function ‘gfr_start’:
> >     ./include/linux/minmax.h:93:30: error: conversion from ‘long long
> > unsigned int’ to ‘resource_size_t’ {aka ‘unsigned int’} changes value
> > from ‘18446744073709551615’ to ‘4294967295’ [-Werror=overflow]
> >
> > Due to
> >
> >     #define PHYSMEM_END  (-1ULL)
> >
> > not being correct on 32-bit without LPAE.
>
> Hmm. Can you check if making it be
>
>       #define PHYSMEM_END  ((phys_addr_t)-1)
>
> fixes things for you?

That fixes the warning/error, as expected.

> That said, it would probably be even better if we got rid of these
> games entirely, and m68k just defined MAX_PHYSMEM_BITS instead. Maybe
> as a config option, since I assume it's going to be either 24 or 32
> depending on CPU (or are there other choices? I used to know the old
> m68k, but...)

Indeed, on 68000 and 68008 it could be 24 resp. 20 ;-)
But all systems with a MMU (and even MC68328 without MMU) do support
a 32-bit external address space, so 32 seems fine.

BTW, other 32-bit architectures are suffering from the same issue
(kisskb shows mips, xtensa, parisc, powerpc failures).

BTW2, the following may not work with the default PHYSMEM_END due
to integer overflow, on both 32-bit and 64-bit:

mm/sparse.c:    unsigned long max_sparsemem_pfn = (PHYSMEM_END + 1) >>
PAGE_SHIFT;

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds