Add suspend and shutdown options to screensaver

[wbrawner]

This adds SUSPEND and SHUTDOWN as options to the screensaver. I put the options under screensaver to avoid confusion about the timeouts (e.g. if you have the device set to shutdown after 2m but screensaver after 5m, should the timeout for shutdown occur before the screensaver or does the shutdown timeout wait until the screensaver has started?). I've attached screenshots for reference as well, taken from my RGB30.

Of note is that the behavior only affects emulationstation, so the device still won't suspend or sleep while a game is open. I have a WIP branch¹ where I've added swayidle as a package to attempt to get the device to suspend mid-game after some inactivity, but it seems that XWayland and/or the emulators don't properly support idle-inhibit², so in my testing my device would just suspend or shutdown in the middle of gameplay despite actively playing.

Footnotes

1. https://github.com/wbrawner/distribution/tree/idle-suspend ↩

2. https://github.com/swaywm/swayidle/issues/5 ↩

[aenertia]

Do you know if you can use hyrbid suspend (i.e suspend to disk after X time/dischage) along with a resume partition to achieve this. IIRC this would require an additional kernel module and a dedicated swap partition. https://linux.die.net/man/8/pm-suspend-hybrid

[wbrawner]

I think we would need the swap partition, not sure about the kernel module though. I think systemd would handle this now though, with systemctl suspend-then-hibernate

[aenertia]

You can actually use an offset location and a swap-file rather than dedicated partition, although dedicated partitions tend to be easier to deal with as you don't need to get the block offset added to the kernel cmd

[wbrawner]

Yeah I saw that too, wasn't even going to suggest it though since I figured the partition would just be easier

[porschemad911]

Yes, would just need to size the partition correctly based on device RAM.

[aenertia]

I mean if it's a toggle/script exposed as a UX, making a swapfile and calculating an offset and adding it to the kernel cmd is possibly less work than a separate partition layout/logic tbh.

…

On Wed, 4 Sept 2024 at 12:23, John Williams ***@***.***> wrote: Yes, would just need to size it correctly based on device RAM. — Reply to this email directly, view it on GitHub <#29 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACF5L5O7IGLXZJK2NPXPGTZUZHI5AVCNFSM6AAAAABNSVBFFKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMRXGY3TOMBVHE> . You are receiving this because you commented.Message ID: ***@***.***>

[wbrawner]

Fair enough! Sorry for the delay, I was out of my home country for a couple of weeks so I didn't have access to my home computer. I'll take a look at getting a swapfile created and test hibernation out on my device

[aenertia]

I'm interested to see what the clock time is rebooting to a resume vs normal boot. My experience is hibernate boost tend to be slower than reboots

…

On Sat, 5 Oct 2024 at 10:06, William Brawner ***@***.***> wrote: Fair enough! Sorry for the delay, I was out of my home country for a couple of weeks so I didn't have access to my home computer. I'll take a look at getting a swapfile created and test hibernation out on my device — Reply to this email directly, view it on GitHub <#29 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACF5L66KRYCMXIIVEFVRYTZZ5CTBAVCNFSM6AAAAABNSVBFFKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOJUHA3DEMZVGM> . You are receiving this because you commented.Message ID: ***@***.***>