On Wed, Oct 3, 2018 at 5:39 PM Zbigniew Jędrzejewski-Szmek <zbyszek@xxxxxxxxx> wrote:
We have had a long discussion about hibernate (suspend to disk)
being unreliable. But there seems to be no hard data. Let's gather
some!
If you perform hibernation (systemctl hibernate, or the equivalent
through the GUI), does _your_ system suspend and resume correctly?
Note: I'm not talking about the user-space configuration issues
(resume= not set on the kernel command line, no swap, swap encrypted
with temporary keys, whatever), but only about any potential kernel
driver issues.
On all my Thinkpads (R61, T500, X220, T450s, T480s) hibernation worked correctly once I've dealt with the user space issues (those are usually the most painful part of the whole process). I've had some rare issues on T500 where resume from hibernation would cause some cpu lockup, but it only happened when I had a file-based swap, so perhaps it can be attributed to that.
On my Intel-based desktop (with Gigabyte GA-Z87-D3HP motherboard) the resume from hibernation failed randomly (probably in about 20% of cases or so) with errors about e820 memory mapping changed. There were months when it happened and months when it didn't, so perhaps it was related to the currently used kernel. After I configured all my USB ports to show up as XHCI in UEFI setup, the problems seem to have went away. So either this is a firmware problem on my motherboard, or the kernels have been working fine since then.
Quite interestingly, regular suspend (to RAM) has been more problematic for me than hibernation. During certain periods of time (i.e. using certain kernels), I've seen my laptops immediately resume from suspend every time I tried it (but hibernation worked fine), or seeing GNOME freeze/crash after resume (but that can easily be a userspace issue). On my desktop, I see frequent memory corruption after resume from suspend and I wasn't able to figure out why (hardware is not faulty), so I had to resort to using hibernation only.
On the other hand, on average I use regular suspend much more often than hibernation (also due to GNOME being quite unfriendly to any user choice in this matter), so maybe that's why I see suspend issues more often than hibernation issues.
_______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@xxxxxxxxxxxxxxxxxxxxxxx
