Linux Adding Wake-On-Connect/Disconnect For USB4 Ports

[Quackdoc]

this is nice, plugging into a dock even while the screen is closed should wake up the device. this is the same whether it be a laptop or a phone IMO.

[johanb]

Windows has what they call hybrid sleep mode -- it writes the ram to disk but keeps the power on to resume from standby. It's somewhere in-between hibernate and sleep mode and is supposed to use less power than sleep. I wouldn't know. I don't sleep or hibernate anything I own since shit turns on and off so fast these days.

Hybrid mode , like traditional standby, can wake stuff up, too. My first assumption is that's what they're actually using and not realizing it since, IIRC, that's the default option. My 2nd assumption is their UEFI has wake up abilities. My 3rd assumption, very closely related to the 2nd one, is they're on a network and receiving wake up packets. Based on the battery comment, my 4th assumption is that one of the above kicks in when the laptop is plugged into AC power and not battery only.

That is not what hybrid sleep is... Hybrid sleep has been available on both Windows and Linux for decades, and that feature has nothing to do with getting woken up by UEFI or whatever you're talking about. Maybe that exists, I don't know, but don't confuse people with the name "hybrid sleep" feature which has existed for many many years.

REAL hybrid sleep is just a mix of suspend and hibernation, when turned on your computer will both suspend to RAM and to disk at the same time. What's the point of this? The primary use-case of this in case you have a desktop with unreliable power, that way you can quickly wake up from a suspend if power has been available all the time and if the power went out you still didn't lose your stuff. Laptops usually don't have this enabled, because if configured correctly they can wake up at a low battery level and automatically suspend (not sure if any linux distro has that feature though).

Googling/DDG "hybrid sleep" and "hybrid sleep windows" gives me no results about whatever you are talking about, only what I just said.

[linuxgeex]

RE: "running in hibernate mode"

How that is made? I'd like to run all my servers e.g. in AWS in that kind of mode and save money a lot.

Don't take it literally. Every system I've owned that can hibernate could also wake itself via a timer, so it could periodically test the environment to see whether it should stay awake, or run some small maintenance task and then hibernate again. Useful if you only want it to run 1 minute per day, and if your power supply is constrained but doesn't self-deplete at a high rate.

My partner's Dell XPS-13 self-depletes in about 60 hours even when hibernated, because it keeps the USB-PD port hot, making S2D last identical to S2R. No there's no way to disable the hot PD port.

[Frenzie]

My partner's Dell XPS-13 self-depletes in about 60 hours even when hibernated, because it keeps the USB-PD port hot, making S2D last identical to S2R. No there's no way to disable the hot PD port.

Is this by any chance from the same series as my 9350? I'd made the mistake of upgrading the BIOS one time too many and I had to downgrade it a little bit to get a non-broken suspend behavior back (although I'm talking about suspend, not hibernate). Anyway, that's a long-winded way of saying a BIOS upgrade or downgrade could be worth investigating even if the changelog doesn't mention anything about it.

[linuxgeex]

Is this by any chance from the same series as my 9350? I'd made the mistake of upgrading the BIOS one time too many and I had to downgrade it a little bit to get a non-broken suspend behavior back (although I'm talking about suspend, not hibernate). Anyway, that's a long-winded way of saying a BIOS upgrade or downgrade could be worth investigating even if the changelog doesn't mention anything about it.

9360, the BIOS updates have all been security-related, she uses it for business and PCI-DSS compliance prevents unpatched operation. ie when stating the business-critical reason for not applying the patch, "it's inconvenient" doesn't cut it. If there was a breach that's the sort of thing that makes you liable and uninsured.

[Frenzie]

9360, the BIOS updates have all been security-related, she uses it for business and PCI-DSS compliance prevents unpatched operation. ie when stating the business-critical reason for not applying the patch, "it's inconvenient" doesn't cut it. If there was a breach that's the sort of thing that makes you liable and uninsured.

That's too bad. Anyway, I checked and I had to downgrade to 1.12.2. The next and current version, 1.13.0, is the one that's borked.