Re: skb_padto and small fragmented transmits

On Thu, 2003-02-06 at 10:44, David S. Miller wrote:

   From: Chris Leech [off-list ref]
   Date: 06 Feb 2003 11:22:51 -0800
   
   I fail to see how the statement "skb->len + skb->data_len" has any
   usable meaning, or how it can be anything other than a bug.

This equation is the standard way to find the full length
on any skb.  For linear skbs, data_len is always zero.

I asked Alan to use this formula so that greps on the source
tree would always show data_len being taken into account, and
thus usage would be consistent.

OK, now I'm really getting confused.  Every other example I can find in
the networking code, and every scatter-gather capable driver, uses
skb->len as the full length and skb->len - skb->data_len as the length
of the first or linear portion.

   From: Chris Leech [off-list ref]
   Date: 06 Feb 2003 11:22:08 -0800
   
   OK, now I'm really getting confused.  Every other example I can find in
   the networking code, and every scatter-gather capable driver, uses
   skb->len as the full length and skb->len - skb->data_len as the length
   of the first or linear portion.
   
Indeed, Alan you need to fix the skb_padto stuff to use
skb->len, ignore the skb->data_len as skb->len is the
full length.

Sorry for telling you to do the wrong thing Alan, my bad :)

On Thu, 2003-02-06 at 22:43, David S. Miller wrote:

   From: Chris Leech [off-list ref]
   Date: 06 Feb 2003 11:22:08 -0800
   
   OK, now I'm really getting confused.  Every other example I can find in
   the networking code, and every scatter-gather capable driver, uses
   skb->len as the full length and skb->len - skb->data_len as the length
   of the first or linear portion.
   
Indeed, Alan you need to fix the skb_padto stuff to use
skb->len, ignore the skb->data_len as skb->len is the
full length.

Dave just fix it next time you touch the code and push it to Marcelo. It
doesnt affect the 2.2 backport so that will be ok

   From: Alan Cox [off-list ref]
   Date: 07 Feb 2003 13:33:41 +0000

   On Thu, 2003-02-06 at 22:43, David S. Miller wrote:
   > Indeed, Alan you need to fix the skb_padto stuff to use
   > skb->len, ignore the skb->data_len as skb->len is the
   > full length.
   
   Dave just fix it next time you touch the code and push it to Marcelo. It
   doesnt affect the 2.2 backport so that will be ok
   
Ok, I will take care of this.