[PATCH] x86/coreboot: Exclude memory regions starting above 4GB

[Tom Rini]

On Fri, Feb 13, 2026 at 04:06:58PM -0700, Jeremy Compostella wrote:

> Hi,
> 
> Is there any update on this commit?

I'll be putting it in master for v2026.04 sometime next week.

> 
> Jeremy Compostella <jeremy.compostella at intel.com> writes:
> 
> > Tom Rini <trini at konsulko.com> writes:
> >
> >> On Thu, Feb 05, 2026 at 10:50:24AM -0700, Jeremy Compostella wrote:
> >>
> >>> Is there anything more I should be doing? This is a bug fix I would
> >>> really appreciate seeing merged.
> >>
> >> It would be nice if someone else reviewed it. It's also been a day since
> >> posting. I'll put it in my list to make sure doesn't get lost for the
> >> next release, thanks.
> >
> > Thank you for looking into this. It would indeed help if someone else could review it.
> >
> >>> Jeremy Compostella <jeremy.compostella at intel.com> writes:
> >>> 
> >>> > This commit updates the RAM region filtering logic in
> >>> > board_get_usable_ram_top() to skip any memory regions whose start address
> >>> > is above 4GB. Previously, only the end address was capped at 4GB, but
> >>> > regions entirely above this threshold were still considered.
> >>> >
> >>> > Typically, the following memory map entries would cause
> >>> > board_get_usable_ram_top() to return 0x100000000, which is incorrect.
> >>> >
> >>> >     start=00000000, end=00001000, type=16
> >>> >     start=00001000, end=000a0000, type=1
> >>> >     start=000a0000, end=000f6000, type=2
> >>> >     start=000f6000, end=000f7000, type=16
> >>> >     start=000f7000, end=00100000, type=2
> >>> >     start=00100000, end=6f170000, type=1
> >>> >     start=6f170000, end=70000000, type=16
> >>> >     start=70000000, end=80800000, type=2
> >>> >     start=e0000000, end=f8000000, type=2
> >>> >     start=fa000000, end=fc000000, type=2
> >>> >     start=fc800000, end=fc880000, type=2
> >>> >     start=fd800000, end=fe800000, type=2
> >>> >     start=feb00000, end=feb80000, type=2
> >>> >     start=fec00000, end=fed00000, type=2
> >>> >     start=fed20000, end=fed80000, type=2
> >>> >     start=feda1000, end=feda2000, type=2
> >>> >     start=fedc0000, end=fede0000, type=2
> >>> >     start=100000000, end=102400000, type=2
> >>> >     start=102400000, end=47f800000, type=1
> >>> >     start=4000000000, end=4020000000, type=2
> >>> >
> >>> > By adding a check to continue the loop if the region's start address
> >>> > exceeds 0xffffffffULL, the function now properly ignores regions that are
> >>> > not usable in 32-bit address space.
> >>> >
> >>> > Change-Id: I8cbe0281aeea67a2f5bb9f6669456f5c7df9b409
> >>> > Signed-off-by: Jeremy Compostella <jeremy.compostella at intel.com>
> >>> > ---
> >>> >  arch/x86/cpu/coreboot/sdram.c | 2 ++
> >>> >  1 file changed, 2 insertions(+)
> >>> >
> >>> > diff --git a/arch/x86/cpu/coreboot/sdram.c b/arch/x86/cpu/coreboot/sdram.c
> >>> > index 013225f129a..cc1edd7badd 100644
> >>> > --- a/arch/x86/cpu/coreboot/sdram.c
> >>> > +++ b/arch/x86/cpu/coreboot/sdram.c
> >>> > @@ -42,6 +42,8 @@ phys_addr_t board_get_usable_ram_top(phys_size_t total_size)
> >>> >  			continue;
> >>> >  
> >>> >  		/* Filter memory over 4GB. */
> >>> > +		if (start > 0xffffffffULL)
> >>> > +			continue;
> >>> >  		if (end > 0xffffffffULL)
> >>> >  			end = 0x100000000ULL;
> >>> >  		/* Skip this region if it's too small. */
> >>> 
> >>> -- 
> >>> Jeremy
> >>> One Emacs to rule them all
> 
> -- 
> Jeremy
> One Emacs to rule them all


-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 228 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20260213/5f78d6b9/attachment.sig>