STR:

• cross compile Firefox for aarch64-unknown-linux-gnu (using the sysroot from --enable-bootstrap).
• push firefox to a pinephone
• launch weston with the kiosk-shell (weston --shell=kiosk-shell.so)

Expected:

• firefox opens in full screen.

Observed:

• the screen turns full black (it's gray when no app runs).

Here's the log: https://pastebin.com/SadcqrQe

There's some crash in glxtest that might be related to bug 1676883, see No GPUs detected via PCI on ARM in the logs.

For the record: this appears to work fine on desktops/laptops as well as the desktop mode of weston on the pinephone.

From https://gitlab.freedesktop.org/wayland/weston/-/issues/473#note_822381:

My rough understanding of the situation (thanks @marius.vlad0 for the detailed investigation!) is this:

We send a keyboard.enter event to an unmapped surface.
This somehow leads Firefox to commit that surface without a buffer but with a frame callback.
Since that surface is not visible, the compositor doesn't ever send back a wl_callback.done().

The first issue is that kiosk-shell send a keyboard.enter to an unmapped surface. This certainly isn't useful in general, although I don't think it's a protocol correctness violation. In any case, !564 fixes this.
This stray enter event uncovers the second issue, which is that Firefox sends a frame callback when committing without a buffer and effectively blocks on that callback signalling. If the stray enter, which seems to trigger this, is in fact not a protocol correctness violation, then Firefox's behavior is also independently problematic and should also be fixed.

So we should fix that behaviour indeed.

This is something that Martin or I should probably look into as it's quite Wayland specific.

Sounds good - thanks.

That's Weston specific issue, right? I wonder if the commit is for SW renderer or VSync. Robert, do we have the VSync with empty commits enabled? I think we also do the wl_surface commit somewhere with actual wl_buffer attach.

Following up on https://gitlab.freedesktop.org/wayland/weston/-/issues/473#note_823635:

I do think that sending keyboard-enter to an unmapped surface is very close to protocol violation because I believe that unmapped surfaces should not be able to gain or have input focus.

Given that:

• this is a very close call on whether it's valid behaviour of the compositor or not
• it's now fixed in weston
• we haven't seen that from any other compositors yet

I'd say let's close this as fixed by https://gitlab.freedesktop.org/wayland/weston/-/merge_requests/564