Messages in this thread | | | Date | Wed, 26 Jan 2022 13:27:34 +0000 | From | "Russell King (Oracle)" <> | Subject | Re: [PATCH] net: stmmac: don't stop RXC during LPI |
| |
On Wed, Jan 26, 2022 at 08:55:22PM +0800, Jisheng Zhang wrote: > On Sun, Jan 23, 2022 at 06:39:26PM +0100, Andrew Lunn wrote: > > > I think this is a common issue because the MAC needs phy's RXC for RX > > > logic. But it's better to let other stmmac users verify. The issue > > > can easily be reproduced on platforms with PHY_POLL external phy. > > > > What is the relevance of PHY polling here? Are you saying if the PHY > > is using interrupts you do not see this issue? > > I tried these two days, if the PHY is using interrupts, I can't > reproduce the issue. It looks a bit more complex. Any suggestions?
I suppose it could be that there is a delay between the PHY reporting the link loss, raising an interrupt, which is then processed to disable the receive side, and the PHY going into LPI. The problem with polling is, well, it's polling, and at a one second rate - which probably is too long between the PHY noticing the loss of link and going into LPI.
What this also probably means is that if interrupt latency is high enough, the same problem will occur.
So maybe the EEE support to be a little more clever - so we only enable clock stop after the MAC driver has disabled the receive side.
-- RMK's Patch system: https://www.armlinux.org.uk/developer/patches/ FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
| |