Am 19.11.2014 um 21:19 schrieb Felix Fietkau: > On 2014-10-31 21:38, Heiner Kallweit wrote: >> Kernel 3.14 introduced a switch reset in phy_init_hw in drivers/net/phy >> causing BMCR_ANENABLE to get cleared. >> >> Due to the fact that ar8xxx_phy_config_aneg does nothing for >> PHY 0 autonegatiation support remains disabled. >> This can cause ports to operate at 10MBit/half-duplex only. >> >> Fix this by calling genphy_config_aneg for PHY 0 too as >> genphy_config_aneg sets BMCR_ANENABLE if it's not yet set. >> Fixes: ticket 17800 >> >> Signed-off-by: Heiner Kallweit <hkallwe...@gmail.com> > Applied the series with a few minor modifications. > > Thanks! > > - Felix > Great! Not sure whether you've seen the follow-up on some patches of the series.
Patch 4: Nov 1st I submitted a version 2 which simplifies ar8xxx_phy_init by using ar8xxx_has_gige to determine chip GBit capability instead of passing this information as a parameter to the function. Patch 3/5: I use my routers w/o VLAN functionality and the patches work fine. However on Nov 8th Dirk Neukirchen reported a problem that with this patch series his WAN port stopped working. I have no idea how this patch series can impact VLAN functionality. However I provided him with updated patches which fixed the issue. (included in a mail sent Nov 9th). Eventually I moved the ANEG fixup code from ar8xxx_phy_config_aneg to ar8xxx_phy_config_init what seems to be cleaner to me. That's what we discussed about on Nov 13th/14th. Having said that I'm totally fine with patches 1 and 2 being applied. For patches 3 and 4 I'll provide the final (IMHO) versions to be applied. Regarding patch 5 I'd be interested in other's opinion whether this ANEG fixup code better fits into config_init or config_aneg to eventually decide with which version of the patch to go. Heiner _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
