On Thu, 4 May 2017, Paul Menzel wrote:
> Dear Linux USB folks,
>
>
> With commit 85724edec (Merge tag 'leds_for_4.12' of
> git://git.kernel.org/pub/scm/linux/kernel/git/j.anaszewski/linux-leds),
> testing the resume time on a Lenovo X60t connected to the docking
> station with the script `analyze_suspend.py` [1][2], one USB
> controller(?) needs the most time with 376 ms.
Well, it's always going to be true that _one_ of the USB controllers
needs more time than the others. :-)
Where do you get than 376 ms number from? The attached graph and
listing show that 0000:00:1d.7 and usb5 (the PCI and USB interfaces of
the EHCI controller) take 18.971 ms in resume_noirq and 119.183 ms in
resume, for a total of 138.154 ms.
> > usb @ 17ef:1000 [5-6] {usb} async_device (Total Suspend: 88.202 ms Total Resume: 376.385 ms)
5-6 is not the EHCI controller. It is a USB device plugged into that
controller.
> Around 270 ms are spent in `generic_resume [usbcore]`.
generic_resume doesn't do anything; it just calls usb_port_resume to do
the real work.
> Please find the compressed HTML result page attached, which was created
> with `sudo ./analyze_suspend.py -config config/suspend-callgraph.cfg`.
>
> Here is some more information about the devices.
You didn't include any information about the 5-6 device.
> Can there something be done about that?
Hard to say without knowing anything about the device. Or what driver
it uses.
Alan Stern
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
