Re: PATCH: VLAN support for 3c59x/3c90x

[Willy Tarreau]

Hi Herbert,

On Sat, Jul 31, 2004 at 01:57:04PM +1000, Herbert Xu wrote:
> Willy Tarreau <willy@xxxxxxxxx> wrote:
> > no, because the driver has no change_mtu() function, so it uses the generic
> > one, eth_change_mtu(), which is bound to 1500.
> 
> What is preventing you from implementing a change_mtu() function?

well, now I see where you want to bring me :-)

So several reasons :
  - the change_mtu() function might be called at any time after driver
    initialization. I don't know at all if there are things to do to
    "lock" the hardware during such changes, as well as I don't know
    what parts of the code I will need to extract to change the hard
    MTU. The initial MTU is really different since it's used to
    initialize hardware registers. The generic change_mtu() function
    only plays with dev->mtu and not hardware since it never goes
    above standard size. I could try, but if it works I would offer
    no warranties for other hardware.

  - I really, really, really... lack time. I would do this during
    my few hours nighty sleep and I wouldn't want to use the resulting
    code :-)

  - many (all ?) other drivers already have an MTU parameter, and many
    of them don't have a problem with using generic change_mtu(). So why
    would this one in particular need such a change ? (and please don't
    tell me that *I* will have to do this for all others :-))

As previously said, I can take a few minutes to add the 'MODULE_PARM'
line, it's not much more than replying to this mail. At least it will
be a good start.

Cheers,
Willy

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/