[dpdk-dev] [PATCH v2 01/16] fm10k: add new vPMD file

Chen, Jing D jing.d.chen at intel.com
Tue Oct 27 06:26:03 CET 2015


Hi, Bruce,

Best Regards,
Mark


> -----Original Message-----
> From: Richardson, Bruce
> Sent: Friday, October 23, 2015 6:01 PM
> To: Chen, Jing D
> Cc: Stephen Hemminger; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v2 01/16] fm10k: add new vPMD file
> 
> On Fri, Oct 23, 2015 at 08:39:56AM +0000, Chen, Jing D wrote:
> > Hi, Stephen,
> >
> > Best Regards,
> > Mark
> >
> >
> > > -----Original Message-----
> > > From: Stephen Hemminger [mailto:stephen at networkplumber.org]
> > > Sent: Thursday, October 22, 2015 11:59 PM
> > > To: Chen, Jing D
> > > Cc: dev at dpdk.org
> > > Subject: Re: [dpdk-dev] [PATCH v2 01/16] fm10k: add new vPMD file
> > >
> > > On Thu, 22 Oct 2015 17:44:49 +0800
> > > "Chen Jing D(Mark)" <jing.d.chen at intel.com> wrote:
> > >
> > > > +#ifndef __INTEL_COMPILER
> > > > +#pragma GCC diagnostic ignored "-Wcast-qual"
> > > > +#endif
> > >
> > > Since this is new code, can't you make it work correctly
> > > with Gcc. Rather than turning off a useful diagnostic.
> >
> > This macro is necessary for later SSE functions or I'll have to add some
> > Un-necessary cast to avoid compile failure.
> >
> > I can add it in later patch. But it will have to show up anyway, right?
> 
> Actually, casting won't make the warnings go away either. You'll always get a
> warning about removing the volatile from the structure members when you
> pass
> them to the intrinsic functions. Only option is to temporarily disable the
> warning
> [or else write your own versions of the intrinsics that do support volatiles :-)]
> 
> /Bruce

Thanks for the clarifications. :)



More information about the dev mailing list