Local display function fails with HDMI in moOde 7.0.1

[TheOldPresbyope]

Sigh.

Even though the title of this thread clearly says "with HDMI" it seems to have mutated into an issue with DSI displays.

My official RPi 7" DSI touchscreen is encased with an RPi that drives an external USB DAC. With no release of moOde has this system experienced the reported local display failure mode. I've seen it occur only with HDMI displays.

What I find interesting about the me-too DSI posts to this thread is that Alaini93 is using a HiFiBerry DAC+ Pro while pgzh1, ciro73, and new_bloke are using a HiFiBerry AMP2 DAC. Is it just a coincidence that these few DSI failure reports involve I2S and the same driver or are we seeing real evidence here? [I'm beginning to feel like an epidemiologist here.]

I'm not going to pull my encased system apart to test with an I2S DAC.

Perhaps one of you four who report experiencing the failure using a DSI display could try switching to a different audio output---onboard audio jack or USB DAC---and seeing if the local display function continues to result in the chromium-browser eating up system memory.

Regards,
Kent

I'm sorry that my post has irritated you. I'd be happy to delete it and start a new thread if you'd prefer. I test software for a living, and it's often the case that what looks like a specific scenario may have a root cause that manifests via other routes as well. I'd be more than happy to help investigate if it can be of use to the community of moode users. I have purchased a separate USB thumb drive for this purpose so I can switch between versions without a reinstall. Please let me know what information would be most useful in terms of output (other settings or logs).

Re "irritation": Sorry, I was feeling more bemused than irritated. There may well be a root cause but posts should not get buried in pre-existing threads before it's known if there is a connection.

IMO its better to have multiple threads with specific titles that later get merged when more details are discovered. This makes it much easier for readers and testers to find relevant conversations. If you've tried the forum search feature you'll know what I mean. 

Still, I'm a curmudgeon channeling my college librarian on this issue (her motto: a place for everything and everything in its place). This forum is as messily organized as most other forums are. It certainly beats the old days when all things "moOde" were posted to a single thread on the diyaudio.com forum.

Re: more info

1) I've seen no similar posts from users of DSI displays and DACs other than HiFiBerry Hats. If there are such folks having similar problems, we need to hear from them.
2) I've already suggested a test which could show whether the coincidence of HiFiBerry Hats/drivers is a red herring.

Other than that, for either display subsystem we need some way of understand what's driving the chromium-browser's render process to binge feed on system memory. The usual Linux logs and tools suffice to tell us which process it is but that's about it. So far, we haven't found a way to characterize this with enough supporting detail to make an intelligent case to the chromium devs and/or the Raspberry Pi OS devs (or even which group to approach).

All suggestions welcome.

Regards,
Kent