Patch: link state detection for 8139too against 2.5.41

Patch: link state detection for 8139too against 2.5.41

[Stefan Rompf]

[-- Attachment #1: Type: text/plain, Size: 278 bytes --]

Hi Jeff,

attached you find a patch that enables link state detection for the
8139too driver. It applies cleanly against 2.5.41 and with some line
offset against 2.4.19.

I've written this stuff on a friends' computer using 2.4.18, so it is
only slightly tested.

Cheers, Stefan

[-- Attachment #2: patch-linkstate-8139too-2.5.41 --]
[-- Type: text/plain, Size: 709 bytes --]

--- linux/drivers/net/8139too.c.orig	Tue Oct  1 09:05:46 2002
+++ linux/drivers/net/8139too.c	Thu Oct 10 00:37:37 2002
@@ -1534,9 +1534,16 @@
 				 struct rtl8139_private *tp,
 				 void *ioaddr)
 {
-	int mii_lpa;
+	int mii_lpa, mii_bmsr;
 
 	mii_lpa = mdio_read (dev, tp->phys[0], MII_LPA);
+	mii_bmsr = mdio_read(dev, tp->phys[0], MII_BMSR);
+
+	if (mii_bmsr & BMSR_LSTATUS && !netif_carrier_ok(dev)) {
+		netif_carrier_on(dev);
+	} else if (netif_carrier_ok(dev)) {
+		netif_carrier_off(dev);
+	}
 
 	if (!tp->mii.force_media && mii_lpa != 0xffff) {
 		int duplex = (mii_lpa & LPA_100FULL)
@@ -1563,7 +1570,7 @@
 		}
 	}
 
-	next_tick = HZ * 60;
+	next_tick = HZ * 3;
 
 	rtl8139_tune_twister (dev, tp);
 

Re: Patch: link state detection for 8139too against 2.5.41

[Donald Becker]

On Sat, 12 Oct 2002, Stefan Rompf wrote:

> Date: Sat, 12 Oct 2002 01:55:04 +0200
> From: Stefan Rompf <srompf@isg.de>
> To: jgarzik@mandrakesoft.com
> Cc: netdev@oss.sgi.com
> Subject: Patch: link state detection for 8139too against 2.5.41
> 
> Hi Jeff,
> 
> attached you find a patch that enables link state detection for the
> 8139too driver. It applies cleanly against 2.5.41 and with some line
> offset against 2.4.19.
> 
> I've written this stuff on a friends' computer using 2.4.18, so it is
> only slightly tested.

Uhmmm, you don't need to poll with the rtl8139.  There is a link change
interrupt.  Here is the recently added from rtl8139.c

	if (status & RxUnderrun){
		/* This might actually be a link change event. */
		if ((tp->drv_flags & HAS_LNK_CHNG)  &&  link_changed) {
			/* Really link-change on new chips. */
			int lpar = inw(ioaddr + NWayLPAR);
			int duplex = (lpar&0x0100) || (lpar & 0x01C0) == 0x0040
				|| tp->duplex_lock;
			/* Do not use MII_BMSR as that clears sticky bit. */
			if (inw(ioaddr + GPPinData) & 0x0004) {
				netif_link_down(dev);
			} else
				netif_link_up(dev);
			if (tp->msg_level & NETIF_MSG_LINK)
				printk(KERN_DEBUG "%s: Link changed, link partner "
					   "%4.4x new duplex %d.\n",
					   dev->name, lpar, duplex);
			tp->full_duplex = duplex;
			/* Only count as errors with no link change. */
			status &= ~RxUnderrun;
		} else {


-- 
Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993

Re: Patch: link state detection for 8139too against 2.5.41

[Stefan Rompf]

Hi,

> > I've written this stuff on a friends' computer using 2.4.18, so it is
> > only slightly tested.
> 
> Uhmmm, you don't need to poll with the rtl8139.  There is a link change
> interrupt.  Here is the recently added from rtl8139.c

ok, I'll try this next time I have access to the card.

Stefan

Re: Patch: link state detection for 8139too against 2.5.41

[Jeff Garzik]

Take a look at how 8139cp driver does link state...  it should be an 
also exact copy of 8139cp's code.

(8139cp will eventually get GMII/TBI support, but for now this statement 
is correct)

Re: Patch: link state detection for 8139too against 2.4.20rc2 / 2.5

[Stefan Rompf]

[-- Attachment #1: Type: text/plain, Size: 426 bytes --]

Hi,

> > Uhmmm, you don't need to poll with the rtl8139.  There is a link change
> > interrupt.  Here is the recently added from rtl8139.c
> 
> ok, I'll try this next time I have access to the card.

...that was some days ago. I've seen that Jeff is enhancing the MII
interface library, so I used the functions provided there. Patch is
against 2.4.20rc2 and 2.5 with some offset. Please apply if it looks
good.

Cheers, Stefan

[-- Attachment #2: patch-ls-8139too --]
[-- Type: text/plain, Size: 1326 bytes --]

--- linux/drivers/net/8139too.c.old	2002-11-19 00:32:04.000000000 +0100
+++ linux/drivers/net/8139too.c	2002-11-21 22:32:39.000000000 +0100
@@ -1335,18 +1335,7 @@
 	struct rtl8139_private *tp = dev->priv;
 
 	if (tp->phys[0] >= 0) {
-		u16 mii_lpa = mdio_read(dev, tp->phys[0], MII_LPA);
-		if (mii_lpa == 0xffff)
-			;					/* Not there */
-		else if ((mii_lpa & LPA_100FULL) == LPA_100FULL
-				 || (mii_lpa & 0x00C0) == LPA_10FULL)
-			tp->mii.full_duplex = 1;
-
-		printk (KERN_INFO"%s: Setting %s%s-duplex based on"
-				" auto-negotiated partner ability %4.4x.\n",
-		        dev->name, mii_lpa == 0 ? "" :
-				(mii_lpa & 0x0180) ? "100mbps " : "10mbps ",
-			tp->mii.full_duplex ? "full" : "half", mii_lpa);
+		mii_check_media(&tp->mii, 1, 1);
 	}
 }
 
@@ -1994,18 +1983,7 @@
 
 	if ((status & RxUnderrun) && link_changed &&
 	    (tp->drv_flags & HAS_LNK_CHNG)) {
-		/* Really link-change on new chips. */
-		int lpar = RTL_R16 (NWayLPAR);
-		int duplex = (lpar & LPA_100FULL) || (lpar & 0x01C0) == 0x0040
-				|| tp->mii.force_media;
-		if (tp->mii.full_duplex != duplex) {
-			tp->mii.full_duplex = duplex;
-#if 0
-			RTL_W8 (Cfg9346, Cfg9346_Unlock);
-			RTL_W8 (Config1, tp->mii.full_duplex ? 0x60 : 0x20);
-			RTL_W8 (Cfg9346, Cfg9346_Lock);
-#endif
-		}
+		rtl_check_media(dev);
 		status &= ~RxUnderrun;
 	}
 

Re: Patch: link state detection for 8139too against 2.4.20rc2 / 2.5

[Stefan Rompf]

Hi,

Jeff Garzik wrote:

> > ...that was some days ago. I've seen that Jeff is enhancing the MII
> > interface library, so I used the functions provided there. Patch is
> > against 2.4.20rc2 and 2.5 with some offset. Please apply if it looks
> > good.

[...]

> close -- you don't want to unconditionally "initialize" the media 
> (mii_check_media second arg).

I've just seen the reason: No need to read the MII_ADVERTISE register
once the structure has been initialized. Expect another patch the next
days.

> the phy code is gonna be pretty darned similar.  Eventually the phy code 
> needs to move to a common 8139lib.c, because both old-8139 line and 

May be. However, I cannot help there. I just have one card of the 8139
family, and all I could produce is some compiles for me stuff -
something that I would not consider helpful.

Stefan

Re: Patch: link state detection for 8139too against 2.4.20rc2 / 2.5

[Jeff Garzik]

On Mon, Dec 09, 2002 at 11:36:11PM +0100, Stefan Rompf wrote:
> Hi,
> 
> Jeff Garzik wrote:
> 
> > > ...that was some days ago. I've seen that Jeff is enhancing the MII
> > > interface library, so I used the functions provided there. Patch is
> > > against 2.4.20rc2 and 2.5 with some offset. Please apply if it looks
> > > good.
> 
> [...]
> 
> > close -- you don't want to unconditionally "initialize" the media 
> > (mii_check_media second arg).
> 
> I've just seen the reason: No need to read the MII_ADVERTISE register
> once the structure has been initialized. Expect another patch the next
> days.

Thanks.


> > the phy code is gonna be pretty darned similar.  Eventually the phy code 
> > needs to move to a common 8139lib.c, because both old-8139 line and 
> 
> May be. However, I cannot help there. I just have one card of the 8139
> family, and all I could produce is some compiles for me stuff -
> something that I would not consider helpful.

<shrug>  I would consider it helpful.  If you are willing to do the
coding and worksforme testing, I can review and do further testing.
Part of the power of open source :)

	Jeff

Re: Patch: link state detection for 8139too against 2.4.20rc2 / 2.5

[Jeff Garzik]

Stefan Rompf wrote:

> Hi,
>
>
> >>Uhmmm, you don't need to poll with the rtl8139.  There is a link change
> >>interrupt.  Here is the recently added from rtl8139.c
> >
> >ok, I'll try this next time I have access to the card.
>
>
> ...that was some days ago. I've seen that Jeff is enhancing the MII
> interface library, so I used the functions provided there. Patch is
> against 2.4.20rc2 and 2.5 with some offset. Please apply if it looks
> good.
>
> Cheers, Stefan
>
>
> ------------------------------------------------------------------------
>
> --- linux/drivers/net/8139too.c.old	2002-11-19 00:32:04.000000000 +0100
> +++ linux/drivers/net/8139too.c	2002-11-21 22:32:39.000000000 +0100
> @@ -1335,18 +1335,7 @@
>  	struct rtl8139_private *tp = dev->priv;
>
>  	if (tp->phys[0] >= 0) {
> -		u16 mii_lpa = mdio_read(dev, tp->phys[0], MII_LPA);
> -		if (mii_lpa == 0xffff)
> -			;					/* Not there */
> -		else if ((mii_lpa & LPA_100FULL) == LPA_100FULL
> -				 || (mii_lpa & 0x00C0) == LPA_10FULL)
> -			tp->mii.full_duplex = 1;
> -
> -		printk (KERN_INFO"%s: Setting %s%s-duplex based on"
> -				" auto-negotiated partner ability %4.4x.\n",
> -		        dev->name, mii_lpa == 0 ? "" :
> -				(mii_lpa & 0x0180) ? "100mbps " : "10mbps ",
> -			tp->mii.full_duplex ? "full" : "half", mii_lpa);
> +		mii_check_media(&tp->mii, 1, 1);



close -- you don't want to unconditionally "initialize" the media 
(mii_check_media second arg).

You can also kibbitz from 8139cp.c in 2.4.20-rc4 / 2.5.<recent> because 
the phy code is gonna be pretty darned similar.  Eventually the phy code 
needs to move to a common 8139lib.c, because both old-8139 line and 
8139C+ support external MII phys as well as the normal on-chip phy found 
in most 8139s.

	Jeff