So I tried to revitalize a very old Toshiba NB505 by installing Linux on it (Debian stable/Stretch). For the most part, it works perfectly, except for one annoying aspect: if I suspend it by closing the lid, it won't resume successfully afterwards.

It goes to sleep ok. Nothing seems amiss in the logs. However, when I wake it up, the computer spins up, but the screen is blank, and there's no reaction to input (including ctrl-alt-del). Logs from after the resume don't exist. Waiting for a very long time doesn't make a difference.

This is on a 4.16.0 kernel, although I had the same issue on a 4.9 kernel just before that. Suspend/resume did work correctly under Windows before I blew it away.
edit: I should also mention that suspend/resume works fine on my other Toshiba (Satellite) Debian laptop, which is set up similarly.

Any ideas?

Try hibernation instead. In the Linux world they try for one or the other. Frequently they don't both work.

Yes, hibernate works.

Who's "they"?

update: I just tried upgrading to the most recent BIOS version (which by the way specifically mentions a Linux-related fix), but the problem persists.

update: I found out that suspend works if I use s2ram instead of pm-suspend (or closing the lid). So that at least shifts it from being a hardware issue to being a configuration issue.

I tried including a file with SLEEP_MODULE="uswsusp" in /etc/pm/config.d/ as indicated by man pm-action, but no luck. I'm a bit tired to play with it more this evening.

What initially gave me hope was the fact that testing suspend by writing core to /sys/power/pm_test worked fine.

I fixed it by editing /etc/acpi/lid.sh to use s2ram instead of pm-suspend, but if anyone can figure out the mystery as to why pm-suspend doesn't work, I'd be interested to hear about it.

1 members found this post helpful.