Re: [net-next 4/6] e1000: configure and read MDI settings

[Jeff Kirsher <jeffrey.t.kirsher@intel.com>]

[-- Attachment #1: Type: text/plain, Size: 1448 bytes --]

On Sat, 2012-07-21 at 16:37 +0100, Ben Hutchings wrote:
> 
> On Fri, 2012-07-20 at 18:17 -0700, Brandeburg, Jesse wrote:
> > 
> > On Fri, 20 Jul 2012, Ben Hutchings wrote:
> > > Why don't you set ecmd->eth_tp_mdix_ctrl here?
> > > 
> > > If you also leave it as 0, it's impossible for userland to tell
> whether
> > > the current mode was forced or automatically selected.
> > 
> > Thanks for the review, right now the get interface (and ethtool
> display) 
> > doesn't support any way to report if the setting was forced or not.
> I 
> > didn't think about changing the get because I didn't want to modify
> the 
> > userland reporting (I also figured it was a simple interface right
> now, 
> > and didn't need changing, and was focused on the _set_ which is the
> part 
> > fixing the users' reported bugs.)
> 
> Everything else you can change with ETHTOOL_SSET is also reported by
> ETHTOOL_GSET; why would this be any different?
> 
> > I think the patches as they currently stand are okay, do you agree?
> I 
> > would be glad to submit a followon to implement the new "get"
> interface if 
> > we can hash out the interface changes, but I see no reason to gate
> these 
> > patches.
> 
> You left this for 20 months, what's the hurry now? 

Since there some changes that are needed in this patch set, I will drop
this series from my tree so that I can continue pushing additional
ixgbe/ixgbevf patches.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 836 bytes --]