[dpdk-dev] [PATCH v5 2/2] build: add support for detecting march on ARM

[Bruce Richardson]

On Mon, Jan 22, 2018 at 06:07:18PM +0530, Pavan Nikhilesh wrote:
> On Mon, Jan 22, 2018 at 12:30:53PM +0000, Bruce Richardson wrote:
> > On Mon, Jan 22, 2018 at 05:16:49PM +0530, Pavan Nikhilesh wrote:
> > > Added support for detecting march and mcpu by reading midr_el1
> > > register.  The implementer, primary part number values read can be
> > > used to figure out the underlying arm cpu.
> > >
> > > Signed-off-by: Pavan Nikhilesh <pbhagavatula at caviumnetworks.com>
> > > --- app/test-pmd/meson.build    |  2 +-
> > > config/arm/armv8_machine.py | 18 +++++++++ config/arm/meson.build
> > > | 95 +++++++++++++++++++++++++++++++++++++++++----
> > > config/meson.build          | 19 ++++----- drivers/meson.build
> > > |  2 +- examples/meson.build        |  2 +- lib/meson.build
> > > |  2 +- meson.build                 |  2 +- test/test/meson.build
> > > |  2 +- 9 files changed, 121 insertions(+), 23 deletions(-) create
> > > mode 100755 config/arm/armv8_machine.py
> > >
> <snip>
> > > + +	if machine == 'generic' +		machine =
> > > impl_generic +		cmd_output = cmd_generic +	endif +
> > > +	message('Implementer : ' + machine[0])
> >
> > When cross-compiling for arm I get an error at this line:
> >
> >   Meson encountered an error in file config/arm/meson.build, line
> >   69, column 1: Index 0 out of bounds of array of size 0.
> >
> Will be sending v2 of the cross patch soon(once directory structure
> and file name is fininlized) that would resolve this issue.
> 
So is there an expected patch ordering here? I believe the set for
adding Octeon drivers to build is ok to merge, though there is a small
change for renaming machine_arg to machine_args in it if these patches
are not applied first. Will I take that set, if my recent rebase on top
of mainline has not broken it, and you can rebase these to take account
of that extra driver?

/Bruce