Re: [dpdk-dev] [PATCH v3 2/4] drivers/net: add drivers for Cavium NICs to meson build

[Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>]

On Fri, Jan 19, 2018 at 05:16:30PM +0000, Bruce Richardson wrote:
> On Fri, Jan 19, 2018 at 06:42:02PM +0530, Pavan Nikhilesh wrote:
> > Add Cavium octeontx and thunder nicvf to meson build infrastructure.
> >
> > Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
> > ---
> >  drivers/net/meson.build                             |  3 ++-
> >  drivers/net/octeontx/base/meson.build               | 21 +++++++++++++++++++++
> >  drivers/net/octeontx/meson.build                    | 13 +++++++++++++
> >  drivers/net/thunderx/Makefile                       |  2 +-
> >  drivers/net/thunderx/base/meson.build               | 14 ++++++++++++++
> >  drivers/net/thunderx/meson.build                    | 14 ++++++++++++++
> >  ...cvf_version.map => rte_pmd_thunderx_version.map} |  0
> >  7 files changed, 65 insertions(+), 2 deletions(-)
> >  create mode 100644 drivers/net/octeontx/base/meson.build
> >  create mode 100644 drivers/net/octeontx/meson.build
> >  create mode 100644 drivers/net/thunderx/base/meson.build
> >  create mode 100644 drivers/net/thunderx/meson.build
> >  rename drivers/net/thunderx/{rte_pmd_thunderx_nicvf_version.map => rte_pmd_thunderx_version.map} (100%)
> >
> <snip>
> > diff --git a/drivers/net/thunderx/meson.build b/drivers/net/thunderx/meson.build
> > new file mode 100644
> > index 000000000..7333738de
> > --- /dev/null
> > +++ b/drivers/net/thunderx/meson.build
> > @@ -0,0 +1,14 @@
> > +# SPDX-License-Identifier: BSD-3-Clause
> > +# Copyright(c) 2017 Cavium, Inc
> > +
> > +subdir('base')
> > +objs = [base_objs]
> > +
> > +sources = files('nicvf_rxtx.c',
> > +		'nicvf_ethdev.c',
> > +		'nicvf_svf.c'
> > +	       )
> > +
> > +cflags += '-fno-prefetch-loop-arrays'
> > +
>
> This flag doesn't work with clang (at least with 5.0.0), so needs to
> be:
>
> 	if cc.has_argument('-fno-prefetch-loop-arrays')
> 	       cflags += '-fno-prefetch-loop-arrays'
> 	endif
>
> I'm also getting errors with gcc builds due to maybe-uninitialized
> variables, so I also had to add:
>
> 	if cc.has_argument('-Wno-maybe-uninitialized')
> 	       cflags += '-Wno-maybe-uninitialized'
> 	endif

Will send out v4 adressing these.

>
> Regards,
> /Bruce

Thanks,
Pavan