Repository navigation
external: pull the Raspberry Pi archive suite that matches the release - #491
Merged
Merged
Conversation
Every Debian-side rpi-* source pulled KEY="bookworm main", including the ones serving trixie, forky and sid -- two of them are even named for a suite they did not pull. archive.raspberrypi.com now publishes bookworm (oldstable), trixie (stable) and forky (testing), so trixie and later images have been getting oldstable packages: raspberrypi-sys-mods bookworm 20250930~bookworm trixie 1:20260914 raspi-utils bookworm 20251002-1~bookworm trixie 20260626-1 python3-libcamera bookworm 0.5.2+rpt20250903 trixie 0.7.2+rpt20260817 firmware-brcm80211 bookworm 1:20240709-2 trixie 1:20260519-1 bluez bookworm 5.66-1 trixie 5.82-1.1 This is also the root cause behind armbian/build#10827: the sudoers rule /etc/sudoers.d/010_pi-nopasswd is shipped by the bookworm build of raspberrypi-sys-mods and is already gone from the trixie build. The images carry it only because they mirror the wrong suite. Split the three affected sources per release rather than widening one: the existing file keeps serving bookworm, a new -trixie sibling serves trixie:forky:sid. forky and sid go to trixie because upstream's forky suite holds three packages (binutils-source, gcc-16-source, gdb-source) and none of ours. raspi-gpio is left out of the trixie glob: upstream retired it in favour of the raspi-utils tools and it does not exist in that suite. It stays available for bookworm. rpi-libraspberrypi-bin-{trixie,forky} are deliberately NOT changed -- libraspberrypi-bin and libraspberrypi0 exist only in bookworm, so pointing them at trixie would mirror nothing. Both now say so in a comment, since the next person to read this will otherwise "fix" them. Verified every rpi-* source against the live indexes it now points at, both architectures: each glob resolves to real packages, nothing resolves empty. Signed-off-by: Igor Pecovnik <igor@armbian.com>
1 of 2 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Every Debian-side
rpi-*source pulledKEY="bookworm main"— including the ones serving trixie, forky and sid. Two of them (rpi-libraspberrypi-bin-trixie,-forky) are even named for a suite they did not pull.archive.raspberrypi.compublishes three Debian suites now:oldstablestabletestingSo trixie and later images have been getting oldstable packages:
raspberrypi-sys-mods20250930~bookworm1:20260914raspi-utils20251002-1~bookworm20260626-1python3-libcamera0.5.2+rpt202509030.7.2+rpt20260817firmware-brcm802111:20240709-21:20260519-1bluez5.66-15.82-1.1This is the root cause of armbian/build#10827
That PR removes
/etc/sudoers.d/010_pi-nopasswd(pi ALL=(ALL) NOPASSWD: ALL) from Debian bcm2711 images with apre_umount_final_imagehook. I checked both packages:Upstream already removed it. The images carry that rule only because they mirror the wrong suite. After this change, #10827 is a bookworm-scoped mitigation rather than the fix — still worth having there, since bookworm is
oldstableupstream and won't get the newer package.What changed
Split per release rather than widening one source, following the pattern already used by
rpi-libraspberrypi-bin-{bookworm,trixie,forky}:rpi-debian-firmware-raspi.confrpi-debian-firmware-raspi-trixie.conf(new)rpi-libcamera-debian.confrpi-libcamera-debian-trixie.conf(new)rpi-pi-bluetooth-debian.confrpi-pi-bluetooth-debian-trixie.conf(new)forky and sid point at trixie, not forky: upstream's forky suite contains three packages (
binutils-source,gcc-16-source,gdb-source) and none of ours.raspi-gpiois left out of the trixie glob — upstream retired it in favour of theraspi-utilstools and it does not exist in that suite. It stays available for bookworm.Deliberately not changed
rpi-libraspberrypi-bin-trixie.confand-forky.confkeep pulling bookworm.libraspberrypi-binandlibraspberrypi0exist only in the bookworm suite — pointing them at trixie would mirror nothing. Both files now carry a comment saying so, because otherwise the next person reading this will "fix" them.Verification
Every
rpi-*source checked against the live index of the suite it now points at, botharm64andarmhf— each glob resolves to real packages, nothing resolves empty:Worth knowing before merge
This bumps several packages by a year or more in trixie+ images —
bluez5.66 → 5.82 andfirmware-brcm802112024 → 2026 are the notable ones. That is the point, but it is a real change in what ships, so the first nightly after this deserves a look rather than a rubber stamp.