A package containing your fix is currently in experimental. My expectation is that the grub2 will be synced during the Mantic cycle (or merged leaving a minimal delta). This will bring your fix to Ubuntu. Therefore I am not sure a separate upload to Ubuntu is really worth it, also considering that the Importance of LP: #1475620 is Low. What do you think?
Speaking about your MP: you modified an already released d/changelog entry, while you should have added a new one (2.06-2ubuntu19). However, again, I am not convinced an upload is justified.
> Hello,
>
> If I understand it right this change already landed upstream (Debian) as part
> of this MR:
>
> https://salsa.debian.org/grub-team/grub/-/merge_requests/33/commits
>
> A package containing your fix is currently in experimental. My expectation is
> that the grub2 will be synced during the Mantic cycle (or merged leaving a
> minimal delta). This will bring your fix to Ubuntu. Therefore I am not sure a
> separate upload to Ubuntu is really worth it, also considering that the
> Importance of LP: #1475620 is Low. What do you think?
>
> Speaking about your MP: you modified an already released d/changelog entry,
> while you should have added a new one (2.06-2ubuntu19). However, again, I am
> not convinced an upload is justified.
Thanks for the review - I wasn't expecting my change to be merged in that MR, that was why I opened this proposal "upstream" (relative to that MR). I think this proposal can be closed/rejected.
grub-common.service: Fix service run before actual system sleep
The recordfail functionality is intended to detect if resuming after
hibernation failed:
* On boot, grub sets recordfail
* Kernel attempts to resume
* If resuming was successful, grub-common.service clears recordfail
* If resuming was unsuccessful, on next boot grub finds that recordfail
is already set and adjusts the timeout value (to
GRUB_RECORDFAIL_TIMEOUT or 30 seconds if GRUB_RECORDFAIL_TIMEOUT is
not set)
However, grub-common.service is configured to run after sleep.target.
sleep.target is reached before actual system sleep occurs, and so
grub-common.service will very likely be run before the system shuts
down, i.e. before grub sets recordfail for the attempted resume.
This replaces sleep.target with hibernate.target, hybrid-sleep.target,
and suspend-then-hibernate.target; these targets are reached after the
system has resumed. (suspend.target is omitted as the system is not
fully shut down during suspend and so grub is not involved.)
Hello,
If I understand it right this change already landed upstream (Debian) as part of this MR:
https:/ /salsa. debian. org/grub- team/grub/ -/merge_ requests/ 33/commits
A package containing your fix is currently in experimental. My expectation is that the grub2 will be synced during the Mantic cycle (or merged leaving a minimal delta). This will bring your fix to Ubuntu. Therefore I am not sure a separate upload to Ubuntu is really worth it, also considering that the Importance of LP: #1475620 is Low. What do you think?
Speaking about your MP: you modified an already released d/changelog entry, while you should have added a new one (2.06-2ubuntu19). However, again, I am not convinced an upload is justified.