Merge with 2.4.0-test1-ac21 + pile of MIPS cleanups to make merging

possible.  Chainsawed RM200 kernel to compile again.  Jazz machine
status unknown.

6 files changed, 85 insertions, 29 deletions

diff --git a/Documentation/networking/8139too.txt b/Documentation/networking/8139too.txt
index 879bf6d2f..f27811a2a 100644
--- a/Documentation/networking/8139too.txt
+++ b/Documentation/networking/8139too.txt
@@ -181,11 +181,25 @@ User-mode (or maybe optional /proc) diagnostics program.
 
 11) RTL8139C support untested.
 
+12) 10base-T support flaky or slow
+
 
 
 
 Change History
 --------------
+Version 0.9.7 - June 11, 2000
+
+* Fix support for older chips (RTL8139 early chips should now work again)
+
+
+Version 0.9.6 - May 30, 2000
+
+* Fix 4-extra-bytes bug
+  (thanks to Markus Westergren, via Santiago Garcia Mantinan)
+* Yet more improved chip recognition
+
+
 Version 0.9.5 - May 17, 2000
 
 * Improved chip version recognition
diff --git a/Documentation/networking/shaper.txt b/Documentation/networking/shaper.txt
index 53ff5aeff..1be0db8c0 100644
--- a/Documentation/networking/shaper.txt
+++ b/Documentation/networking/shaper.txt
@@ -1,6 +1,6 @@
 Traffic Shaper For Linux
 
-This is the current ALPHA release of the traffic shaper for Linux. It works
+This is the current BETA release of the traffic shaper for Linux. It works
 within the following limits:
 
 o	Minimum shaping speed is currently about 9600 baud (it can only
@@ -37,13 +37,12 @@ and will treat it as such unless patched. Note that for mrouted you can run
 mrouted tunnels via a traffic shaper to control bandwidth usage.
 
 	The shaper is device/route based. This makes it very easy to use
-with any setup BUT less flexible. You may well want to combine this patch
-with Mike McLagan <mmclagan@linux.org>'s patch to allow routes to be
-specified by source/destination pairs.
+with any setup BUT less flexible. You may need to use iproute2 to set up
+multiple route tables to get the flexibility.
 
 	There is no "borrowing" or "sharing" scheme. This is a simple
-traffic limiter. I'd like to implement Van Jacobson and Sally Floyd's CBQ
-architecture into Linux one day (maybe in 2.1 sometime) and do this with
-style.
+traffic limiter. We implement Van Jacobson and Sally Floyd's CBQ
+architecture into Linux 2.2. THis is the preferred solution. Shaper is
+for simple or back compatible setups.
 
 Alan
diff --git a/Documentation/networking/sk98lin.txt b/Documentation/networking/sk98lin.txt
index 863090170..510d2c4ef 100644
--- a/Documentation/networking/sk98lin.txt
+++ b/Documentation/networking/sk98lin.txt
@@ -3,7 +3,7 @@
 
 sk98lin.txt created 11-Nov-1999
 
-Readme File for sk98lin.o v3.02
+Readme File for sk98lin.o v3.04
 SK-NET Gigabit Ethernet Adapter SK-98xx Driver for Linux
 
 This file contains
@@ -24,8 +24,8 @@ This file contains
 ============
 
 The sk98lin driver supports the SysKonnect SK-NET Gigabit Ethernet
-Adapter SK-98xx family on Linux 2.2.x.
-It has been tested with Linux on Intel/x86 and ALPHA machines.
+Adapter SK-98xx family on Linux 2.2.x and above.
+It has been tested with Linux on Intel/x86, ALPHA and UltraSPARC machines.
 From v3.02 on, the driver is integrated in the linux kernel source.
 ***
 
@@ -132,7 +132,8 @@ Parameters can be set at the command line while loading the
 module with 'insmod'. The configuration tools of some distributions
 can also give parameters to the driver module.
 If you use the kernel module loader, you can set driver parameters
-in the file /etc/conf.modules. Insert a line of the form:
+in the file /etc/modules.conf (or old name: /etc/conf.modules).
+Insert a line of the form:
 
 options sk98lin ...
 
@@ -281,14 +282,12 @@ The setting must be done on all adapters that can be reached by
 the large frames. If one adapter is not set to receive large frames,
 it will simply drop them.
 
-NOTE: If you look at the statistics (with netstat) in large frame
-      mode while there is traffic on the net, you will see the
-      RX error counter go up. This is because the adapter hardware
-      counts received large frames as errors, although they are
-      received correctly. So ignore this counter in that case.
-
 You can switch back to the standard ethernet frame size with:
   ifconfig eth0 mtu 1500
+
+To make this setting persitent, add a script with the 'ifconfig'
+line to the system startup sequence (named something like "S99sk98lin"
+in /etc/rc.d/rc2.d).
 ***
 
 
@@ -374,15 +373,27 @@ following information is available:
 (8) HISTORY
 ===========
 
-VERSION 3.02
+VERSION 3.04 (In-Kernel version)
+Problems fixed:
+- Driver start failed on UltraSPARC
+- Rx checksum calculation for big endian machines did not work
+- Jumbo frames were counted as input-errors in netstat
+
+VERSION 3.03 (Standalone version)
+Problems fixed:
+- Compilation did not find script "printver.sh" if "." not in PATH
+Known limitations:
+- None
+
+VERSION 3.02 (In-Kernel version)
 Problems fixed:
 - None
 New Features:
-- Integration in linux kernel source.
+- Integration in Linux kernel source (2.2.14 and 2.3.29) 
 Known limitations:
 - None
 
-VERSION 3.02
+VERSION 3.01
 Problems fixed:
 - None
 New Features:
diff --git a/Documentation/networking/tlan.txt b/Documentation/networking/tlan.txt
index 1485930fb..901eb44bc 100644
--- a/Documentation/networking/tlan.txt
+++ b/Documentation/networking/tlan.txt
@@ -2,7 +2,10 @@
 (C) 1998 James Banks
 (C) 1999-2000 Torben Mathiasen <tmm@image.dk, torben.mathiasen@compaq.com>
 
-TLAN driver for Linux, version 1.5
+For driver information/updates visit http://tlan.kernel.dk
+
+
+TLAN driver for Linux, version 1.8a
 README
 
 
@@ -65,12 +68,15 @@ II.   Driver Options
 	   if a card which only supports 10Mbs is forced into 100Mbs
 	   mode.)
 
-	5. If the driver is built into the kernel, you can use the 3rd
+	5. You have to use speed=X duplex=Y together now. If you just
+	   do "insmod tlan.o speed=100" the driver will do Auto-Neg.
+	   To force a 10Mbps Half-Duplex link do "insmod tlan.o speed=10 
+	   duplex=1".
+
+	6. If the driver is built into the kernel, you can use the 3rd
 	   and 4th parameters to set aui and debug respectively.  For
 	   example:
 
-/* kernel-parameters are currently not supported. I will fix this asap. */
-
 	   ether=0,0,0x1,0x7,eth0
 
 	   This sets aui to 0x1 and debug to 0x7, assuming eth0 is a
@@ -79,11 +85,14 @@ II.   Driver Options
 	   The bits in the third byte are assigned as follows:
 
 		0x01 = aui
-		0x04 = use half duplex
-		0x08 = use full duplex
-		0x10 = use 10BaseT
-		0x20 = use 100BaseTx
+		0x02 = use half duplex
+		0x04 = use full duplex
+		0x08 = use 10BaseT
+		0x10 = use 100BaseTx
 
+	   You also need to set both speed and duplex settings when forcing
+	   speeds with kernel-parameters. 
+	   ether=0,0,0x12,0,eth0 will force link to 100Mbps Half-Duplex.
 
 III.  Things to try if you have problems.
 	1. Make sure your card's PCI id is among those listed in
@@ -94,5 +103,5 @@ III.  Things to try if you have problems.
 
 There is also a tlan mailing list which you can join by sending "subscribe tlan"
 in the body of an email to majordomo@vuser.vu.union.edu.
-
+There is also a tlan website at http://tlan.kernel.dk
 
diff --git a/Documentation/networking/tulip.txt b/Documentation/networking/tulip.txt
index be41c4031..b936846af 100644
--- a/Documentation/networking/tulip.txt
+++ b/Documentation/networking/tulip.txt
@@ -142,6 +142,24 @@ tulip_core.c	- Driver core (a.k.a. where "everything else" goes)
 
 Version history
 ===============
+0.9.6 (May 31, 2000):
+* Revert 21143-related support flag patch
+* Add HPPA/media-table debugging printk
+
+0.9.5 (May 30, 2000):
+* HPPA support (willy@puffingroup)
+* CSR6 bits and tulip.h cleanup (Chris Smith)
+* Improve debugging messages a bit
+* Add delay after CSR13 write in t21142_start_nway
+* Remove unused ETHER_STATS code
+* Convert 'extern inline' to 'static inline' in tulip.h (Chris Smith)
+* Update DS21143 support flags in tulip_chip_info[]
+* Use spin_lock_irq, not _irqsave/restore, in tulip_start_xmit()
+* Add locking to set_rx_mode()
+* Fix race with chip setting DescOwned bit (Hal Murray)
+* Request 100% of PIO and MMIO resource space assigned to card
+* Remove error message from pci_enable_device failure
+
 0.9.4.3 (April 14, 2000):
 * mod_timer fix (Hal Murray)
 * PNIC2 resusitation (Chris Smith)
diff --git a/Documentation/networking/vortex.txt b/Documentation/networking/vortex.txt
index 850ce4838..19d30079d 100644
--- a/Documentation/networking/vortex.txt
+++ b/Documentation/networking/vortex.txt
@@ -156,7 +156,12 @@ compaq_device_id=N
 
   "Variables to work-around the Compaq PCI BIOS32 problem"....
 
+watchdog=N
 
+  Sets the time duration (in milliseconds) after which the kernel
+  decides that the transmitter has become stuck and needs to be reset. 
+  This is mainly for debugging purposes.  The default value is 400 (0.4
+  seconds).
 
 Additional resources
 --------------------