Re: kern/42314: IC Plus IP100x PHY support

[SAITOH Masanobu <msaitoh%execsw.org@localhost>]

The following reply was made to PR kern/42314; it has been noted by GNATS.

From: SAITOH Masanobu <msaitoh%execsw.org@localhost>
To: Andrius V <vezhlys%gmail.com@localhost>
Cc: msaitoh%execsw.org@localhost, gnats-bugs%netbsd.org@localhost, msaitoh%netbsd.org@localhost,
 gnats-admin%netbsd.org@localhost, netbsd-bugs%netbsd.org@localhost, tharada%oucrc.org@localhost
Subject: Re: kern/42314: IC Plus IP100x PHY support
Date: Thu, 21 Nov 2019 00:14:35 +0900

 On 2019/11/20 15:25, Andrius V wrote:
 > Hi,
 > 
 > With the latest code ifconfig selects media correctly and I don't see
 > watchdog timeout messages but dhcpcd still fails to get IP address
 > properly though (assigns 169.254.161.13/16) and network is not working
 > because of that. Same happens with 1000baseT-FDX and 1000baseT now
 > (for 1000baseT it is a regression since previous commit). Reselecting
 > auto or lower media types (100baseT, 10baseT) works correctly and dhcp
 > reassigns proper the IP address.
 
 Could you test the following diff?
 - Set duplex correctly when user setting is not IFM_AUTO.
 - When the link is up, set VGE_DIAGCTL not from user media setting but from
   the current active link status.
 
 Index: if_vge.c
 ===================================================================
 RCS file: /cvsroot/src/sys/dev/pci/if_vge.c,v
 retrieving revision 1.76
 diff -u -p -r1.76 if_vge.c
 --- if_vge.c	19 Nov 2019 09:54:07 -0000	1.76
 +++ if_vge.c	20 Nov 2019 15:08:28 -0000
 @@ -1928,33 +1928,34 @@ vge_miibus_statchg(struct ifnet *ifp)
  	 * always implied, so we turn on the forced mode bit but leave
  	 * the FDX bit cleared.
  	 */
 -
  	dctl = CSR_READ_1(sc, VGE_DIAGCTL);
 
 -	switch (IFM_SUBTYPE(ife->ifm_media)) {
 -	case IFM_AUTO:
 +	if (IFM_SUBTYPE(ife->ifm_media) == IFM_AUTO) {
  		dctl &= ~VGE_DIAGCTL_MACFORCE;
  		dctl &= ~VGE_DIAGCTL_FDXFORCE;
 -		break;
 -	case IFM_1000_T:
 -		dctl |= VGE_DIAGCTL_MACFORCE;
 -		dctl &= ~VGE_DIAGCTL_FDXFORCE;
 -		dctl |= VGE_DIAGCTL_GMII;
 -		break;
 -	case IFM_100_TX:
 -	case IFM_10_T:
 +	} else {
 +		u_int ifmword;
 +
 +		/* If the link is up, use the current active media. */
 +		if ((mii->mii_media_status & IFM_ACTIVE) != 0)
 +			ifmword = mii->mii_media_active;
 +		else
 +			ifmword = ife->ifm_media;
 +
  		dctl |= VGE_DIAGCTL_MACFORCE;
 -		dctl &= ~VGE_DIAGCTL_GMII;
 -		if ((ife->ifm_media & IFM_FDX) != 0)
 +		if ((ifmword & IFM_FDX) != 0)
  			dctl |= VGE_DIAGCTL_FDXFORCE;
  		else
  			dctl &= ~VGE_DIAGCTL_FDXFORCE;
 -		break;
 -	default:
 -		printf("%s: unknown media type: %x\n",
 -		    device_xname(sc->sc_dev),
 -		    IFM_SUBTYPE(ife->ifm_media));
 -		break;
 +
 +		if (IFM_SUBTYPE(ifmword) == IFM_1000_T) {
 +			/*
 +			 * It means the user setting is not auto and it's
 +			 * 1000baseT-FDX or 1000baseT.
 +			 */
 +			dctl |= VGE_DIAGCTL_GMII;
 +		} else
 +			dctl &= ~VGE_DIAGCTL_GMII;
  	}
 
  	CSR_WRITE_1(sc, VGE_DIAGCTL, dctl);
 -----------
 
 
 
 The same diff is at:
 
 	http://www.netbsd.org/~msaitoh/vge-20191120-0.dif
 
 Thanks in advance.
 
 -- 
 -----------------------------------------------
                 SAITOH Masanobu (msaitoh%execsw.org@localhost
                                  msaitoh%netbsd.org@localhost)