[Bug 215880] Resume process hangs for 5-6 seconds starting sometime in 5.16

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=215880

--- Comment #57 from Paul Ausbeck (paula@alumni.cse.ucsc.edu) ---
The m.2/SATA to 7pin SATA adapter that I mentioned previously as on order,
arrived. Using this adapter I connected a spare Barracuda 4TB drive to one of
the m.2 slots on my Chuwi Gemibook Pro laptop. With this setup, the lack of
mouse pointer interactivity on resume exhibits exactly the same as on my Intel
dh77kc desktop.

The mouse pointer is not interactive until after "PM: suspend exit" is written
to the kernel log. Additionally, an active GUI terminal is not interactive
until the same time. This is noticeable by typing "sudo pm-suspend" in a
terminal. The resulting CR/LF and next command prompt do not appear following
resume until the same time that the mouse cursor comes alive. The Barracuda
drive connected to the Gemibook isn't as slow to spin up as the relatively old
WD 2TB in my desktop so the delay to interactivity is only about 5 seconds
instead of 10+.

The interactivity delay is present on both machines when using kernel 6.1x and
not present on both machines when using kernel 5.10x. When using kernel 5.10x,
the dmesg log on both machines shows "PM suspend exit" happens before the
slower ATA drives are "configured for UDMA/133". When using kernel 6.1x, the
dmesg log shows "PM suspend exit" occurring after the slower ATA dries are
"configured for UDMA/133".

At this point, I am convinced that this is a regression related to changes in
libata. I will take this offline now to see what I can do to help recover the
previous resume interactivity behavior.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.