Re: [edk2-devel] [PATCH edk2-platforms 0/2] Switch all platforms to [Peiless]Sec

["Ard Biesheuvel" <ardb@kernel.org>]

On Thu, 1 Aug 2024 at 17:19, Jeremy Linton <jeremy.linton@arm.com> wrote:
>
> Hi,
>
>
> On 8/1/24 09:44, Ard Biesheuvel wrote:
> > On Thu, 1 Aug 2024 at 16:11, Ard Biesheuvel <ardb@kernel.org> wrote:
> >>
> >> On Thu, 1 Aug 2024 at 15:49, Jeremy Linton <jeremy.linton@arm.com> wrote:
> >>>
> >>> Hi,
> >>>
> >>> On 7/31/24 11:33, Ard Biesheuvel wrote:
> >>>> Switch all ARM platforms that use the SEC drivers in edk2/ArmPlatformPkg
> >>>> to the new versions called Sec or PeilessSec - these have been cleaned
> >>>> up and stripped of obsolete functionality related to multicore boot,
> >>>> which is not something UEFI should concern itself with.
> >>>>
> >>>> This series can be merged once Tianocore/edk2 PR #5997 is merged first.
> >>>> After that, ArmPlatformStackLib and the old PrePi / PrePeiCore drivers
> >>>> can be retired.
> >>>
> >>>
> >>> Hi,
> >>>
> >>> Thanks for cleaning this up, but the rpi4 fails with:
> >>>
> >>> Decompress GetInfo Failed - Invalid Parameter
> >>>
> >>> ASSERT_EFI_ERROR (Status = Not Found)
> >>> ASSERT [PeilessSec]
> >>> /home/jlinton/rpi2/edk2/ArmPlatformPkg/PeilessSec/PeilessSec.c(158):
> >>> !(((INTN)(RETURN_STATUS)(Status)) < 0)
> >>>
> >>
> >> Weird. I actually tried RPi4 myself. Maybe I should have tried a clean rebuild.
> >>
> >> I'll look into it.
> >
> > The below should fix it - I'll update all DSCs with this if it works
> > for you as well.
>
> That fixes the problem above. Thanks!
>
>
>
> I just spent an hour hunting a sd stall heisenbug I've never seen before
> too, which initially I thought might be related to the set from earlier
> this week, but it seems doubtful since the controller stalled far to
> late in the firmware boot for it to be related. Looks like it just
> glitched out and stopped responding to commands, and I either got
> impatient and power cycled it to soon, or the reset/recovery logic isn't
> right, but I don't know since I killed it to quickly and now can't
> reproduce it.
>

Haven't noticed this one myself. The only issue I hit once in a while
(but only with DEBUG builds as far as I am aware) is an ASSERT() on
some XhciDxe control register read (or perhaps a timeout?)


-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#120188): https://edk2.groups.io/g/devel/message/120188
Mute This Topic: https://groups.io/mt/107649429/7686176
Group Owner: devel+owner@edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [rebecca@openfw.io]
-=-=-=-=-=-=-=-=-=-=-=-