> Interestingly, "wakeup_count" does not increase for both? Version, thanks for reporting I had no idea to check for this. Seems to work as you describe, "active_count" and "event_count" are Wakeup source was WoL :-) I added this to sh_eth and now the bookkeeping > So something seems to be missing from a bookkeeping point of view.Ĭool, now I know why some net drivers call pm_wakeup_event() if the > they do for the GPIO keyboard when using the keyboard for wake up. > "event_count" for the Ethernet device do not increase when using WoL, while > However, if I look at /sys/kernel/debug/wakeup_sources, "active_count" and > I've verified WoL is working on r8a7791/koelsch and r8a7740/armadillo. > count it won't reach 0 and be switched off. Then when Runtime PM decreases the clock usage > suspend callback need to call clk_enable() directly to increase the > One quirk needed for WoL is that the module clock needs to be prevented > close and open the device in the resume callback since the device needs > in a normal suspend/resume scenario without WoL enabled, but it both > hardware the driver close and reopens the device just like it would do > that sh_eth was not suspended to reduce resume time. > the driver from simply restoring the registers and to take advantage of > driver needs to reset the hardware to rearm the WoL logic, this prevents > and disabling all interrupts expect MagicPacket. > WoL is enabled in the suspend callback by setting MagicPacket detection > functionality for WoL, no specific sh_eth version are marked to support > are supported by at least a few versions of sh_eth. > Add generic functionality to support Wake-on-Lan using MagicPacket which Linux-Renesas Re: sh_eth: add generic wake-on-lan support via magic packetĭate: Tue, 17:27:17 +0100 Re: sh_eth: add generic wake-on-lan support via magic packet - Niklas Söderlund Netdev Archive on help / color / mirror / Atom feed From: "Niklas Söderlund"
0 Comments
Leave a Reply. |