pointers in the sk_buff structure.

pointers in the sk_buff structure.

[sudeep list]

hello,

I figured that the pointer sk_buff->tail points to the end of data in
the buffer attached to the sk_buff. What does the pointer sk_buff->end
point to ?

Is there any data that lies between the two pointers, or they are two
pointers to the same address (end of data in the buffer area) ?

Sudeep

[PATCH 4 2.4.27] pcnet32: Add HomePNA parameter for 79C978.

[Richard Hobbs]

I am trying to set up a hybrid network, part wired, part wireless and part
homePNA (the building has some very thick walls and remote computers that can be
reached by existing telephone wires). I am using Fedora core 3 with the 2.6.10
kernel.  I have  checked the pcnet32.c code and the patch 4 2.4.27 appears to be
there. 

I load the pcnet32 module using:
modprobe -v pcnet32 homepna=1

I get reply:
insmod /lib/modules/2.6.10-1.741_FC3/kernel/drivers/net/pcnet32.ko homepna=1

Looks good.

I then try to configure the network through the systems settings interface:
eth0 is a e100 ethernet to my router/modem that is OK.
Disconnecting eth0 to avoid any conflicts.

eth1 should be the AMD79c978 card. Connecting with the telephone wire

I set eth1 as an ethernet (I don't see an option for phoneline network)
I have tried DHCP get the response on activating the network:
Determining IP information for eth1 failed; no link present. Check cable.

I have tried a static IP address, card activates but no device is detected at
the modem/router.

If I unload the module and reload without the homepna option, the card works
fine with a standard ethernet cable.

Any suggestions please.

Thanks
Richard Hobbs

Re: [PATCH 4 2.4.27] pcnet32: Add HomePNA parameter for 79C978.

[Don Fry]

Richard,

2.6.10 has the driver to support the 79c978 card in homepna mode.  I
have only used the card in a point-to-point configuration when in
homepna mode, because all I have is two cards (no modem/router).

When you bring up the card with homepna=1, if you look in
/var/log/messages it should indicate the speed at which the card is
operating.  In my system I see:

pcnet32: media set to 1Mbit mode.
pcnet32: PCnet/Home 79C978 at 0x252c00, 00 90 00 05 dd 27 assigned IRQ 201.
eth8: registered as PCnet/Home 79C978
pcnet32: 8 cards_found.

With the cards cabled together I see link detected using ethtool.

# ethtool eth8
Settings for eth8:
        Current message level: 0x00000007 (7)
        Link detected: yes

On Fri, Feb 04, 2005 at 12:42:07AM +0000, Richard Hobbs wrote:
> 
> I am trying to set up a hybrid network, part wired, part wireless and part
> homePNA (the building has some very thick walls and remote computers that can be
> reached by existing telephone wires). I am using Fedora core 3 with the 2.6.10
> kernel.  I have  checked the pcnet32.c code and the patch 4 2.4.27 appears to be
> there. 
> 
> I load the pcnet32 module using:
> modprobe -v pcnet32 homepna=1
> 
> I get reply:
> insmod /lib/modules/2.6.10-1.741_FC3/kernel/drivers/net/pcnet32.ko homepna=1
> 
> Looks good.
> 
> I then try to configure the network through the systems settings interface:
> eth0 is a e100 ethernet to my router/modem that is OK.
> Disconnecting eth0 to avoid any conflicts.
> 
> eth1 should be the AMD79c978 card. Connecting with the telephone wire
> 
> I set eth1 as an ethernet (I don't see an option for phoneline network)
> I have tried DHCP get the response on activating the network:
> Determining IP information for eth1 failed; no link present. Check cable.
> 
> I have tried a static IP address, card activates but no device is detected at
> the modem/router.
> 
> If I unload the module and reload without the homepna option, the card works
> fine with a standard ethernet cable.
> 
> Any suggestions please.
> 
> Thanks
> Richard Hobbs

-- 
Don Fry
brazilnut@us.ibm.com

Re: pointers in the sk_buff structure.

[Lennert Buytenhek]

On Thu, Feb 03, 2005 at 03:45:15PM -0700, sudeep list wrote:

> hello,

Hi,


> I figured that the pointer sk_buff->tail points to the end of data in
> the buffer attached to the sk_buff. What does the pointer sk_buff->end
> point to ?
> 
> Is there any data that lies between the two pointers, or they are two
> pointers to the same address (end of data in the buffer area) ?

This is a snippet from some docs on the subject I wrote a while ago,
hope it helps.  I should polish it up and submit it for inclusion.


2. The sk_buff structure.

The sk_buff structure ('skb') is actually only used for storing the
metadata corresponding to a packet.  The packet's data is not stored
inside the sk_buff structure itself, but in a separate buffer that is
pointed to by skb->head.  The skb->end member points one byte past the
end of this data buffer.

An important design requirement for sk_buffs is being able to add data
at the end as well as at the front of the packet.  As a packet travels
downwards through the network stack, each layer will usually want to
add its own header in front of the packet, and it would be nice if we
could avoid reallocating and/or copying the entire data portion of the
packet around to make more space at the front of the buffer every time
we want to do this.

To achieve this goal, the packet data is not necessarily stored at the
front of the data buffer, but some space between the front of the buffer
and the front of the packet is left unused.  skb->data and skb->tail
are two extra pointers that point to the beginning and one byte past
the end of the currently used portion of the data buffer, respectively.
Both are guaranteed to point somewhere within the data buffer.
('skb->head <= skb->data <= skb->tail <= skb->end')

        +----------+-------------------------------------+--------------+
        | headroom | packet data                         | tailroom     |
        +----------+-------------------------------------+--------------+
        ^          ^ skb->data                           ^     skb->end ^
        |                                                |
        + skb->head                                      + skb->tail

The function skb_headroom(skb) calculates 'skb->data - skb->head', and
indicates how many bytes we can add to the front of the packet without
having to reallocate the buffer.  Similarly, skb_tailroom(skb) calculates
'skb->end - skb->tail' and indicates how many bytes we can add to the
end of the packet before having to reallocate.

Adding data to and removing data from the front of the buffer is done
with skb_push and skb_pull, respectively.  These wrappers do some sanity
checks to make sure the relevant constraints on the four pointers are
maintained.

When an sk_buff is allocated by alloc_skb, skb->{head,data,tail} are all
initialised to point to the start of the data buffer.  Depending on what
the skb will be used for, the caller will usually want to reserve some
headroom in anticipation of expansion of the data buffer towards the
front.  This is done by calling skb_reserve().

[PATCH 4 2.4.27] pcnet32: Add HomePNA parameter for 79C978.

[Don Fry]

This patch adds a module parameter to select HomePNA mode of operation
for the 79C978 version of the pcnet32.  Tested ia32.

signed-off-by: Patrick Simmons <psimmons@flash.net>
signed-off-by: Don Fry <brazilnut@us.ibm.com>

--- linux-2.4.27/drivers/net/ints.pcnet32.c	Mon Aug  9 15:41:52 2004
+++ linux-2.4.27/drivers/net/pcnet32.c	Mon Aug  9 15:42:04 2004
@@ -22,8 +22,8 @@
  *************************************************************************/
 
 #define DRV_NAME	"pcnet32"
-#define DRV_VERSION	"1.30f"
-#define DRV_RELDATE	"06.16.2004"
+#define DRV_VERSION	"1.30g"
+#define DRV_RELDATE	"06.22.2004"
 #define PFX		DRV_NAME ": "
 
 static const char *version =
@@ -137,6 +137,7 @@ static const char pcnet32_gstrings_test[
 #define MAX_UNITS 8	/* More are supported, limit only on options */
 static int options[MAX_UNITS];
 static int full_duplex[MAX_UNITS];
+static int homepna[MAX_UNITS];
 
 /*
  *				Theory of Operation
@@ -250,6 +251,8 @@ static int full_duplex[MAX_UNITS];
  * v1.30f  16 Jun 2004 Don Fry cleanup IRQ to allow 0 and 1 for PCI,
  * 	   expanding on suggestions from Ralf Baechle <ralf@linux-mips.org>,
  * 	   and Brian Murphy <brian@murphy.dk>.
+ * v1.30g  22 Jun 2004 Patrick Simmons <psimmons@flash.net> added option
+ *	   homepna for selecting HomePNA mode for PCNet/Home 79C978.
  */
 
 
@@ -1083,15 +1086,17 @@ pcnet32_probe1(unsigned long ioaddr, int
 	fdx = 1;
 	/*
 	 * This is based on specs published at www.amd.com.  This section
-	 * assumes that a card with a 79C978 wants to go into 1Mb HomePNA
-	 * mode.  The 79C978 can also go into standard ethernet, and there
-	 * probably should be some sort of module option to select the
-	 * mode by which the card should operate
+	 * assumes that a card with a 79C978 wants to go into standard
+	 * ethernet mode.  The 79C978 can also go into 1Mb HomePNA mode,
+	 * and the module option homepna=1 can select this instead.
 	 */
-	/* switch to home wiring mode */
 	media = a->read_bcr(ioaddr, 49);
+	media &= ~3;		/* default to 10Mb ethernet */
+	if (cards_found < MAX_UNITS && homepna[cards_found])
+	    media |= 1; 	/* switch to home wiring mode */
 	if (pcnet32_debug & NETIF_MSG_PROBE)
-	    printk(KERN_DEBUG PFX "media reset to %#x.\n",  media);
+	    printk(KERN_DEBUG PFX "media set to %dMbit mode.\n", 
+		    (media & 1) ? 1 : 10);
 	a->write_bcr(ioaddr, 49, media);
 	break;
     case 0x2627:
@@ -2255,6 +2260,9 @@ MODULE_PARM(options, "1-" __MODULE_STRIN
 MODULE_PARM_DESC(options, DRV_NAME " initial option setting(s) (0-15)");
 MODULE_PARM(full_duplex, "1-" __MODULE_STRING(MAX_UNITS) "i");
 MODULE_PARM_DESC(full_duplex, DRV_NAME " full duplex setting(s) (1)");
+/* Module Parameter for HomePNA cards added by Patrick Simmons, 2004 */
+MODULE_PARM(homepna,"1-" __MODULE_STRING(MAX_UNITS) "i");
+MODULE_PARM_DESC(homepna, DRV_NAME " mode for 79C978 cards (1 for HomePNA, 0 for Ethernet, default Ethernet");
 
 MODULE_AUTHOR("Thomas Bogendoerfer");
 MODULE_DESCRIPTION("Driver for PCnet32 and PCnetPCI based ethercards");

-- 
Don Fry
brazilnut@us.ibm.com