[Intel-gfx] [PATCH i-g-t v3] lib/igt_core.c: Expand --run-subtest functionality.

[Dave Gordon]

On 15/02/16 16:55, Daniel Vetter wrote:
> On Thu, Feb 04, 2016 at 12:06:57PM +0000, Derek Morton wrote:
>> Added extended wildcard support when specifying --run-subtest.
>>
>> Wildcard format is as specified in rfc3977 and the uwildmat() implementation
>> is taken from libinn.
>> See https://tools.ietf.org/html/rfc3977#section-4 for a description of
>> allowed wildcard expressions.
>>
>> v2: Use comma as list separator (Ville Syrjala)
>> support both ^ and ! as not operators (Dave Gordon)
>>
>> v3: Updated to use uwildmat() (Dave Gordon)
>>
>> Signed-off-by: Derek Morton <derek.j.morton at intel.com>
>> ---
>>   COPYING                 |  21 +++
>>   lib/Makefile.sources    |   2 +
>>   lib/igt_core.c          |  17 +-
>>   lib/uwildmat/uwildmat.c | 474 ++++++++++++++++++++++++++++++++++++++++++++++++
>>   lib/uwildmat/uwildmat.h |  24 +++
>
> Not really a fan of copying other sources into ours. Don't we have
> something ready-made that's generally available, or can we at least pull
> it in as a build-dep?
>
> Thanks, Daniel

It's a standard, RFC3977. The source is readily available for download, 
for example from the INN master site

https://inn.eyrie.org/trac/browser/trunk/lib/uwildmat.c

It seems to be very stable (last change was 2014: "Change Russ' email 
address").

It *is* already available (in Ubuntu at least), but it's in a fairly 
obscure package that not many people will have installed. On Ubuntu, 
/usr/lib/news/libinn.a is part of package 'inn2-dev'; I understand that 
on some other distros its in a different package (e.g. 'inn-devel' on 
CentOS).

So this is probably why Derek chose to include the source from INN 
rather than working out how to make it a prerequisite for building 
i-g-t. And it almost certainly isn't otherwise available for Android!

.Dave.