[V2]Update GStreamer packages to 1.28.7 + fix ARM32

[commodo]

📦 Package Details

Maintainer: @thess @flyn-org @dangowrt

Description:

V2 of PR #29616
But trying to fix the ARM32 stuff here.

If this works, we will try to find an upstreamable fix for ARM32

---

🧪 Run Testing Details

• OpenWrt Version:
• OpenWrt Target/Subtarget:
• OpenWrt Device:

---

✅ Formalities

• I have reviewed the CONTRIBUTING.md file for detailed contributing guidelines.

[openwrt-ai]

Reviewed 7 commits (fresh review, full PR diff). Version/hash bumps are consistent across the six packages, PKG_RELEASE is correctly reset to 1 on gstreamer1 and gst1-plugins-base (both were at 2), no stale 1.26 references are left in the feed, and each commit subject matches the files it touches — no commit-message problems found. The new NEON patch is clean (18/18 add.w → add, no CRLF or trailing whitespace, diffstat matches).

Nothing here blocks a merge from my side. The three inline notes are: an undocumented user-visible package removal (gst1-mod-y4mdec) plus a question about whether the y4m functionality should move to gst-plugins-good, an undocumented dependency-graph change (libgst1video → libgst1allocators → libdrm), and a request for an upstream reference on the new patch.

Note that I could not reach the upstream GStreamer sources from this environment, so anything about what 1.28 changed upstream is phrased as a question rather than an assertion.

---

Generated by Claude Code

[openwrt-ai]

nit: this drops the user-selectable gst1-mod-y4mdec package (together with the -Dy4m= meson arg on line 159), which is a user-visible removal that the commit message ("update to 1.28.6" + release-notes link) does not mention. Worth a sentence in the commit body saying the y4m plugin is gone from gst-plugins-bad in 1.28 so anyone bisecting a missing package finds it.

Related question: gst-plugins-good already carries the plumbing for a y4m plugin — $(call GST_COND_SELECT,y4m) at gst1-plugins-good/Makefile:139 with the package itself commented out at line 265. If upstream folded y4mdec into the -good y4m plugin, enabling that entry would keep the functionality available instead of losing it in this bump. I could not verify the upstream source from this environment, so this is a question rather than a claim — is the drop intended to be a straight removal?

---

Generated by Claude Code

[openwrt-ai]

fixed, thanks — documented in the commit body and the plugin is kept available via the new gst1-mod-y4m in gst-plugins-good.

---

Generated by Claude Code

[openwrt-ai]

Adding allocators here makes libgst1video depend on libgst1allocators, which in turn carries +libdrm (GstBuildLibrary,allocators at line 258). Since practically every gst consumer in the feed pulls libgst1video (gst1-libav, the -good/-bad/-ugly plugins), this quietly adds libdrm to a lot of installs. If it reflects a real link-time dependency in 1.28 that is fine, but it is a dependency-graph change rather than a version bump, and neither it nor the audio addition on line 318 is mentioned in the "update to 1.28.6" commit message — please note both in the commit body so the reason survives.

---

Generated by Claude Code

[openwrt-ai]

nit: "Reported upstream." without a pointer makes this hard to retire later. This patch is a real code change that is expected to go upstream (the PR body says so), so please add the GStreamer issue/MR URL — or an Upstream-Status: line — so a future bump can tell at a glance whether the patch can be dropped.

For context on validation: on this head the Test arm_cortex-a15_neon-vfpv4 job was cancelled at the 6h runner limit rather than completing, so the NEON target this patch exists for is not actually covered by a green run yet (aarch64_generic was cancelled the same way). Not something to fix in the diff, just worth knowing before merge.

---

Generated by Claude Code

[openwrt-ai]

fixed, thanks

---

Generated by Claude Code

[openwrt-ai]

Re-reviewed at 5371599 (previous review was at 6bcca65).

The push since my last review is a pure rebase onto a newer master: git range-diff 6bcca65~7..6bcca65 5371599~7..5371599 reports all seven commits as identical (=), and nothing landed on master in the rebase window that touches multimedia/gst*. No new commits and no content changes, so no new issues found.

The three inline notes from the previous review are still open and still apply to the current head:

• the undocumented removal of gst1-mod-y4mdec (and the related question about the y4m plugin in gst-plugins-good),
• the undocumented libgst1video → libgst1allocators (→ libdrm) dependency-graph change, plus the audio addition to pbtypes,
• the missing upstream reference / Upstream-Status: on 100-audio-resampler-neon-a32-syntax.patch.

One piece of context worth repeating for this head: Test arm_cortex-a15_neon-vfpv4 and Test aarch64_generic were again cancelled at the ~6h runner limit rather than completing, so the NEON target this patch exists for still has no completed build behind it. That is a harness limitation, not a defect in the diff — nothing here blocks a merge from my side.

---

Generated by Claude Code

[flyn-org]

I see two build targets timed out. It looks like the build succeeded in both cases. Can anyone else comment on this?Please add a reference to where you submitted/reported 100-audio-resampler-neon-a32-syntax.patch upstream. Thank you—I am very glad to see this progress.

[commodo]

I see two build targets timed out. It looks like the build succeeded in both cases. Can anyone else comment on this?Please add a reference to where you submitted/reported 100-audio-resampler-neon-a32-syntax.patch upstream. Thank you—I am very glad to see this progress.

Right.
I wanted to start upstreaming this but got derailed by other stuff + vacation.
Will get back

[openwrt-ai]

Re-reviewed at 26a541f (previous review was at 5371599).

The only change to this feed since then is a rebase plus a rewrite of the commit body and the patch header on the last commit. git diff 5371599..26a541f -- multimedia/ touches nothing but the header of 100-audio-resampler-neon-a32-syntax.patch; all 18 add.w → add hunks are byte-identical, and git range-diff reports the six version/hash-bump commits as =.

Two things worth recording for this head:

• Test arm_cortex-a15_neon-vfpv4 completed successfully instead of being cancelled at the runner limit, so the NEON A32 target this patch exists for finally has a green build behind it. That was the main open unknown in my last two reviews.
• I checked the "OpenWrt builds ARM in A32 mode" premise against openwrt/openwrt — there is no -mthumb or USE_THUMB anywhere in the userspace build flags, only CONFIG_ARM_THUMB in kernel configs — so the reasoning in the patch header holds. The encoding claims check out too (e0888001 is add r8, r8, rN in A32; the 16-bit Thumb ADD (register) form supports high registers and leaves flags alone).

Commit checks

• 26a541f "gst1-plugins-base: fix NEON resampler build in ARM A32 mode" — the reworded body now says "Backport a patch dropping the .w suffix". "Backport" asserts the fix already exists upstream, but as of your 2026-08-24 reply upstreaming had not started, and the sentence that previously hedged this ("Reported upstream.") was dropped rather than replaced with a link. If the fix has landed upstream since, please cite the commit/MR in the body; if not, "Add a patch" is the accurate verb. I could not reach the GStreamer repo from this environment, so this is a question rather than an assertion. Same point inline on the patch header.

Still open from the first review and unchanged on this head: the undocumented removal of gst1-mod-y4mdec in gst1-plugins-bad, and the libgst1video → libgst1allocators (→ libdrm) dependency change plus the audio addition to pbtypes.

One CI note so it is not mistaken for a regression: Test x86_64 failed at the runtime-test install step with libmesa (no such package): required by: libgst1gl-1.28.6-r1[libmesa]. That +libmesa dependency is pre-existing on master (GstBuildLibrary,gl at gst1-plugins-base/Makefile:262, untouched by this PR), so it reads as a feed/index gap in the test harness rather than something this diff introduced. aarch64_generic was again cancelled at the 6h limit.

Nothing here blocks a merge from my side.

---

Generated by Claude Code

[openwrt-ai]

The upstream reference is still missing, and this reword moved it backwards rather than forwards: the previous header ended with "Reported upstream.", and that sentence was deleted without a link taking its place — while the commit body now calls the change a "Backport". flyn-org asked for the reference explicitly and your reply said upstreaming had not started yet, so as the patch stands there is nothing a future maintainer can check to decide whether it can be dropped at the next version bump.

The minimum that keeps the patch retirable — swap Pending for the MR URL once it is filed:

Suggested change

	Signed-off-by: Alexandru Ardelean <alex@shruggie.ro>
	Signed-off-by: Alexandru Ardelean <alex@shruggie.ro>
	Upstream-Status: Pending

Related, on line 1: From 3664d240c434969e6e460faa323704a4c579261e Mon Sep 17 00:00:00 2001 is the id of a commit in your own tree, not an upstream one (unless the fix has landed upstream since — I could not reach the GStreamer repo from this environment to check). Combined with "Backport a patch..." in the commit body, that header reads like an upstream commit someone is meant to be able to look up. Either drop the From <sha> line or make the commit body say "Add a patch", so the provenance is unambiguous.

---

Generated by Claude Code

[openwrt-ai]

fixed, thanks — the patch is now a real upstream backport with Upstream-Status and the MR link, and the From line is an upstream SHA.

---

Generated by Claude Code

[commodo]

@flyn-org

Sync-ed with upstream.
Resolved ARM32 Thumb-2/1/A32 issues on this PR https://gitlab.freedesktop.org/gstreamer/gstreamer/-/merge_requests/12490

[openwrt-ai]

Reviewed 8 commits.

---

Generated by Claude Code

[openwrt-ai]

Applying patch 001-gst-libav-Fix-build-failure-with-FFmpeg-9.0.patch
Hunk #1 FAILED at 2760.
1 out of 1 hunk FAILED -- rejects in file ext/libav/gstavviddec.c
Patch 001-gst-libav-Fix-build-failure-with-FFmpeg-9.0.patch can be reverse-applied
ERROR: package/feeds/packages_ci/gst1-libav failed to build.

That patch, added on master in 66c7c55e for 1.26.4, is MR !12217, which 1.28.7 already contains — so quilt reverse-applies it and all ten build jobs die at package/gst1-libav/refresh before anything compiles. Delete multimedia/gst1-libav/patches/001-gst-libav-Fix-build-failure-with-FFmpeg-9.0.patch in this commit.

---

Generated by Claude Code

[openwrt-ai]

fixed, thanks

---

Generated by Claude Code

[commodo]

To the best of my knowledge, all matters of gstreamer have been solved; and the update should be good to go.

So, if no objections, I will merge this in a couple of days.

[flyn-org]

Thank you for your work on this! I think we should merge, assuming the build failures are due to the CI system. The failures look to be during the test phase after the build.

[commodo]

Thank you for your work on this! I think we should merge, assuming the build failures are due to the CI system. The failures look to be during the test phase after the build.

They're CI related
@GeorgeSapkin @aparcar @hnyman @BKPepe
Any idea how to unblock CI?

[openwrt-ai]

Reviewed 1 new commit; no new issues found.

---

Generated by Claude Code

[BKPepe]

They're CI related @GeorgeSapkin @aparcar @hnyman @BKPepe Any idea how to unblock CI?

I think it will be fixed tomorrow as there are scheduled runs for new Docker images.
This is related: openwrt/docker#213

[flyn-org]

Thank you, @commodo.