Re: [PATCH v7 ethtool 2/2] ethtool: Support for configurable RSS hash key

[Ben Hutchings <ben@decadent.org.uk>]

[-- Attachment #1: Type: text/plain, Size: 1586 bytes --]

On Sun, 2014-09-21 at 23:15 +0100, Ben Hutchings wrote:
> On Sun, 2014-09-21 at 21:52 +0100, Ben Hutchings wrote:
> > On Tue, 2014-07-22 at 17:51 +0530, Venkat Duvvuru wrote:
> > > This ethtool patch will primarily implement the parser for the options provided
> > > by the user for get and set rxfh before invoking the ioctl.
> > > This patch also has
> > > 1. Ethtool man page changes which describes the Usage of
> > >    get and set rxfh options.
> > > 2. Test cases for get and set rxfh in test-cmdline.c
> > > 
> > > Signed-off-by: Venkat Duvvuru <VenkatKumar.Duvvuru@Emulex.com>
> > [...]
> > 
> > FAIL: test-cmdline
> > ==================
> > 
> > E: ethtool --rxfh devname hkey foo returns 0
> > E: ethtool -X devname hkey foo returns 0
> > 
> > The problem with these test cases is we can't parse the hash key (and
> > reject "foo") until after we start sending ioctls.  test-cmdline doesn't
> > include any mocking of ioctls, but instead makes each test successful if
> > it calls an ioctl.
> 
> It also doesn't work for me with an older kernel version (I tested on
> Linux 3.14).  I don't know why that is, as your fallback code does look
> reasonable.

Oops, I tested the previous version of this on 3.14.  The latest version
does work correctly.

I'll apply your changes and then disable the two failing test cases for
now.  Thanks for persisting with this.

Ben.

-- 
Ben Hutchings
We get into the habit of living before acquiring the habit of thinking.
                                                              - Albert Camus

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 811 bytes --]