Re: Shim update breaks booting
On Sat, Aug 26, 2017 at 8:24 PM, Björn 'besser82' Esser
<besser82(a)fedoraproject.org> wrote:
Am 26.08.2017 um 21:16 schrieb Peter Robinson:
>
> On Sat, Aug 26, 2017 at 5:57 PM, Lars Seipel <lars.seipel(a)gmail.com>
> wrote:
>>
>> On Fri, Aug 25, 2017 at 08:24:59PM +0000, Fedora Branched Report wrote:
>>>
>>> Package: shim-signed-13-0.2
>>> Old package: shim-signed-0.8-10
>>> Summary: First-stage UEFI bootloader
>>> RPMs: shim-aa64 shim-ia32 shim-x64
>>> Added RPMs: shim-aa64 shim-ia32 shim-x64
>>> Dropped RPMs: shim
>>> Size: 2238688 bytes
>>> Size change: 1219336 bytes
>>> Changelog:
>>> * Thu Mar 23 2017 Petr Šabata <contyk(a)redhat.com> - 0.8-9
>>> - Re-enable dist tag for module builds
>>>
>>> * Tue Aug 22 2017 Peter Jones <pjones(a)redhat.com> - 13-0.1
>>> - Initial (partially unsigned) build for multi-arch support on
>>> x64/ia32.
>>>
>>> * Thu Aug 24 2017 Peter Jones <pjones(a)redhat.com> - 13-0.2}
>>> - Obsolete old shim builds.
>>
>> The package no longer ships /boot/efi/EFI/fedora/shim.efi which is what
>> people's boot entries refer to. It's shimx64.efi now. This causes
>> failure to boot on EFI systems.
>>
>> Which component, if any, is tasked with adjusting EFI boot entries in
>
> Probably efibootmgr or grubby, the former is more ljkely but I'd just
> file a bug against shim-signed as they're all maintained by the same
> group/person.
>
> Peter
Can we be sure this is a bug in one of those packages? There are other
factors that have an influence on automatically updating the boot entry in
UEFI, like users write protecting those entries on purpose on their systems…
No, but the maintainer of these packages will know, he deals with the
entire stack and knows all the details, he will re-assign the bug as
appropriate.