* [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP @ 2011-04-21 14:03 Matteo Fortini 2011-04-21 18:10 ` Baruch Siach 0 siblings, 1 reply; 4+ messages in thread From: Matteo Fortini @ 2011-04-21 14:03 UTC (permalink / raw) To: linux-arm-kernel We are experiencing unrecoverable timeouts if we disconnect a cable from the FEC. This patch solves the issue by keeping the Ethernet enabled even in STOP. The RM doesn't state it, but i seems that if disabled, the Ethernet is not issuing interrupts to the core. (See http://forums.freescale.com/t5/i-MX-Microprocessors/iMX28-Network-MDIO-timeout-recovery-and-lost-IRQs/td-p/73309 ) --- ?drivers/net/fec.c | ? ?7 ++++++- ?1 files changed, 6 insertions(+), 1 deletions(-) diff --git a/drivers/net/fec.c b/drivers/net/fec.c index d0e2e69..26ea72d 100644 --- a/drivers/net/fec.c +++ b/drivers/net/fec.c @@ -121,8 +121,10 @@ ?#if defined(CONFIG_FEC_1588) && defined(CONFIG_ARCH_MX28) ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII | \ ? ? ? ? ? ? ? ? FEC_ENET_TS_AVAIL | FEC_ENET_TS_TIMER) +#define FEC_STOP_IMASK (FEC_ENET_MII) ?#else ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII) +#define FEC_STOP_IMASK (FEC_ENET_MII) ?#endif ?/* The FEC stores dest/src/type, data, and checksum for receive packets. @@ -1409,6 +1411,9 @@ fec_stop(struct net_device *dev) ? ? writel(1, fep->hwp + FEC_ECNTRL); ? ? udelay(10); + ? ?/* Reactivate the controller to get the IRQs */ + ? ?writel(0x00000002, fep->hwp + FEC_ECNTRL); + ?#ifdef CONFIG_ARCH_MXS ? ? /* Check MII or RMII */ ? ? if (fep->phy_interface == PHY_INTERFACE_MODE_RMII) @@ -1423,7 +1428,7 @@ fec_stop(struct net_device *dev) ? ? writel(fep->phy_speed, fep->hwp + FEC_MII_SPEED); ? ? if (fep->ptimer_present) ? ? ? ? fec_ptp_stop(fep->ptp_priv); - ? ?writel(FEC_DEFAULT_IMASK, fep->hwp + FEC_IMASK); + ? ?writel(FEC_STOP_IMASK, fep->hwp + FEC_IMASK); ? ? netif_stop_queue(dev); ? ? fep->link = 0; -- 1.7.4.2 ^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP 2011-04-21 14:03 [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP Matteo Fortini @ 2011-04-21 18:10 ` Baruch Siach 2011-04-22 6:36 ` Matteo Fortini 0 siblings, 1 reply; 4+ messages in thread From: Baruch Siach @ 2011-04-21 18:10 UTC (permalink / raw) To: linux-arm-kernel Hi Matteo, On Thu, Apr 21, 2011 at 04:03:49PM +0200, Matteo Fortini wrote: > We are experiencing unrecoverable timeouts if we disconnect a cable from the > FEC. This patch solves the issue by keeping the Ethernet enabled even in > STOP. The RM doesn't state it, but i seems that if disabled, the Ethernet is > not issuing interrupts to the core. > (See > http://forums.freescale.com/t5/i-MX-Microprocessors/iMX28-Network-MDIO-timeout-recovery-and-lost-IRQs/td-p/73309 > ) Please send network driver patches to the network drivers mailing list at netdev at vger.kernel.org. This patch does not apply to the current mainline kernel. Which kernel tree are you using? baruch > --- > ?drivers/net/fec.c | ? ?7 ++++++- > ?1 files changed, 6 insertions(+), 1 deletions(-) > > diff --git a/drivers/net/fec.c b/drivers/net/fec.c > index d0e2e69..26ea72d 100644 > --- a/drivers/net/fec.c > +++ b/drivers/net/fec.c > @@ -121,8 +121,10 @@ > ?#if defined(CONFIG_FEC_1588) && defined(CONFIG_ARCH_MX28) > ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII | \ > ? ? ? ? ? ? ? ? FEC_ENET_TS_AVAIL | FEC_ENET_TS_TIMER) > +#define FEC_STOP_IMASK (FEC_ENET_MII) > ?#else > ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII) > +#define FEC_STOP_IMASK (FEC_ENET_MII) > ?#endif > > ?/* The FEC stores dest/src/type, data, and checksum for receive packets. > @@ -1409,6 +1411,9 @@ fec_stop(struct net_device *dev) > ? ? writel(1, fep->hwp + FEC_ECNTRL); > ? ? udelay(10); > > + ? ?/* Reactivate the controller to get the IRQs */ > + ? ?writel(0x00000002, fep->hwp + FEC_ECNTRL); > + > ?#ifdef CONFIG_ARCH_MXS > ? ? /* Check MII or RMII */ > ? ? if (fep->phy_interface == PHY_INTERFACE_MODE_RMII) > @@ -1423,7 +1428,7 @@ fec_stop(struct net_device *dev) > ? ? writel(fep->phy_speed, fep->hwp + FEC_MII_SPEED); > ? ? if (fep->ptimer_present) > ? ? ? ? fec_ptp_stop(fep->ptp_priv); > - ? ?writel(FEC_DEFAULT_IMASK, fep->hwp + FEC_IMASK); > + ? ?writel(FEC_STOP_IMASK, fep->hwp + FEC_IMASK); > > ? ? netif_stop_queue(dev); > ? ? fep->link = 0; -- ~. .~ Tk Open Systems =}------------------------------------------------ooO--U--Ooo------------{= - baruch at tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il - ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP 2011-04-21 18:10 ` Baruch Siach @ 2011-04-22 6:36 ` Matteo Fortini 2011-04-22 7:25 ` Baruch Siach 0 siblings, 1 reply; 4+ messages in thread From: Matteo Fortini @ 2011-04-22 6:36 UTC (permalink / raw) To: linux-arm-kernel Ok I sent it to netdev, let's see if they accept it. The interface is not working without it. I patched the opensource.freescale.com git tree, since it is the most common start for iMX. Regards On Thu, Apr 21, 2011 at 8:10 PM, Baruch Siach <baruch@tkos.co.il> wrote: > Hi Matteo, > > On Thu, Apr 21, 2011 at 04:03:49PM +0200, Matteo Fortini wrote: >> We are experiencing unrecoverable timeouts if we disconnect a cable from the >> FEC. This patch solves the issue by keeping the Ethernet enabled even in >> STOP. The RM doesn't state it, but i seems that if disabled, the Ethernet is >> not issuing interrupts to the core. >> (See >> http://forums.freescale.com/t5/i-MX-Microprocessors/iMX28-Network-MDIO-timeout-recovery-and-lost-IRQs/td-p/73309 >> ) > > Please send network driver patches to the network drivers mailing list at > netdev at vger.kernel.org. > > This patch does not apply to the current mainline kernel. Which kernel tree > are you using? > > baruch > >> --- >> ?drivers/net/fec.c | ? ?7 ++++++- >> ?1 files changed, 6 insertions(+), 1 deletions(-) >> >> diff --git a/drivers/net/fec.c b/drivers/net/fec.c >> index d0e2e69..26ea72d 100644 >> --- a/drivers/net/fec.c >> +++ b/drivers/net/fec.c >> @@ -121,8 +121,10 @@ >> ?#if defined(CONFIG_FEC_1588) && defined(CONFIG_ARCH_MX28) >> ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII | \ >> ? ? ? ? ? ? ? ? FEC_ENET_TS_AVAIL | FEC_ENET_TS_TIMER) >> +#define FEC_STOP_IMASK (FEC_ENET_MII) >> ?#else >> ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII) >> +#define FEC_STOP_IMASK (FEC_ENET_MII) >> ?#endif >> >> ?/* The FEC stores dest/src/type, data, and checksum for receive packets. >> @@ -1409,6 +1411,9 @@ fec_stop(struct net_device *dev) >> ? ? writel(1, fep->hwp + FEC_ECNTRL); >> ? ? udelay(10); >> >> + ? ?/* Reactivate the controller to get the IRQs */ >> + ? ?writel(0x00000002, fep->hwp + FEC_ECNTRL); >> + >> ?#ifdef CONFIG_ARCH_MXS >> ? ? /* Check MII or RMII */ >> ? ? if (fep->phy_interface == PHY_INTERFACE_MODE_RMII) >> @@ -1423,7 +1428,7 @@ fec_stop(struct net_device *dev) >> ? ? writel(fep->phy_speed, fep->hwp + FEC_MII_SPEED); >> ? ? if (fep->ptimer_present) >> ? ? ? ? fec_ptp_stop(fep->ptp_priv); >> - ? ?writel(FEC_DEFAULT_IMASK, fep->hwp + FEC_IMASK); >> + ? ?writel(FEC_STOP_IMASK, fep->hwp + FEC_IMASK); >> >> ? ? netif_stop_queue(dev); >> ? ? fep->link = 0; > > -- > ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ~. .~ ? Tk Open Systems > =}------------------------------------------------ooO--U--Ooo------------{= > ? - baruch at tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il - > ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP 2011-04-22 6:36 ` Matteo Fortini @ 2011-04-22 7:25 ` Baruch Siach 0 siblings, 0 replies; 4+ messages in thread From: Baruch Siach @ 2011-04-22 7:25 UTC (permalink / raw) To: linux-arm-kernel Hi Metteo, On Fri, Apr 22, 2011 at 08:36:17AM +0200, Matteo Fortini wrote: > Ok I sent it to netdev, let's see if they accept it. Keeping the linux-arm-kernel list Cc on the patch is a good idea. > The interface is not working without it. > I patched the opensource.freescale.com git tree, since it is the most > common start for iMX. The networking drivers maintainer (David Miller) can not possibly accept this patch because it does not apply to the current mainline kernel. For patches against the Freescale git tree you should contact Freescale. Which i.MX variant is this? I suggest you to try the latest mainline kernel, or the i.MX kernel development tree at http://git.pengutronix.de/?p=imx/linux-2.6.git;a=summary. baruch > On Thu, Apr 21, 2011 at 8:10 PM, Baruch Siach <baruch@tkos.co.il> wrote: > > Hi Matteo, > > > > On Thu, Apr 21, 2011 at 04:03:49PM +0200, Matteo Fortini wrote: > >> We are experiencing unrecoverable timeouts if we disconnect a cable from the > >> FEC. This patch solves the issue by keeping the Ethernet enabled even in > >> STOP. The RM doesn't state it, but i seems that if disabled, the Ethernet is > >> not issuing interrupts to the core. > >> (See > >> http://forums.freescale.com/t5/i-MX-Microprocessors/iMX28-Network-MDIO-timeout-recovery-and-lost-IRQs/td-p/73309 > >> ) > > > > Please send network driver patches to the network drivers mailing list at > > netdev at vger.kernel.org. > > > > This patch does not apply to the current mainline kernel. Which kernel tree > > are you using? > > > > baruch > > > >> --- > >> ?drivers/net/fec.c | ? ?7 ++++++- > >> ?1 files changed, 6 insertions(+), 1 deletions(-) > >> > >> diff --git a/drivers/net/fec.c b/drivers/net/fec.c > >> index d0e2e69..26ea72d 100644 > >> --- a/drivers/net/fec.c > >> +++ b/drivers/net/fec.c > >> @@ -121,8 +121,10 @@ > >> ?#if defined(CONFIG_FEC_1588) && defined(CONFIG_ARCH_MX28) > >> ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII | \ > >> ? ? ? ? ? ? ? ? FEC_ENET_TS_AVAIL | FEC_ENET_TS_TIMER) > >> +#define FEC_STOP_IMASK (FEC_ENET_MII) > >> ?#else > >> ?#define FEC_DEFAULT_IMASK (FEC_ENET_TXF | FEC_ENET_RXF | FEC_ENET_MII) > >> +#define FEC_STOP_IMASK (FEC_ENET_MII) > >> ?#endif > >> > >> ?/* The FEC stores dest/src/type, data, and checksum for receive packets. > >> @@ -1409,6 +1411,9 @@ fec_stop(struct net_device *dev) > >> ? ? writel(1, fep->hwp + FEC_ECNTRL); > >> ? ? udelay(10); > >> > >> + ? ?/* Reactivate the controller to get the IRQs */ > >> + ? ?writel(0x00000002, fep->hwp + FEC_ECNTRL); > >> + > >> ?#ifdef CONFIG_ARCH_MXS > >> ? ? /* Check MII or RMII */ > >> ? ? if (fep->phy_interface == PHY_INTERFACE_MODE_RMII) > >> @@ -1423,7 +1428,7 @@ fec_stop(struct net_device *dev) > >> ? ? writel(fep->phy_speed, fep->hwp + FEC_MII_SPEED); > >> ? ? if (fep->ptimer_present) > >> ? ? ? ? fec_ptp_stop(fep->ptp_priv); > >> - ? ?writel(FEC_DEFAULT_IMASK, fep->hwp + FEC_IMASK); > >> + ? ?writel(FEC_STOP_IMASK, fep->hwp + FEC_IMASK); > >> > >> ? ? netif_stop_queue(dev); > >> ? ? fep->link = 0; -- ~. .~ Tk Open Systems =}------------------------------------------------ooO--U--Ooo------------{= - baruch at tkos.co.il - tel: +972.2.679.5364, http://www.tkos.co.il - ^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2011-04-22 7:25 UTC | newest] Thread overview: 4+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2011-04-21 14:03 [PATCH] Fix for missed MII interrupts and MDIO timeouts when FEC is in STOP Matteo Fortini 2011-04-21 18:10 ` Baruch Siach 2011-04-22 6:36 ` Matteo Fortini 2011-04-22 7:25 ` Baruch Siach
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).