On 3 December 2014 at 14:22, Daniel Vetter <dan...@ffwll.ch> wrote: > On Tue, Dec 02, 2014 at 05:15:37PM +0000, Thomas Wood wrote: > > diff --git a/lib/igt_core.h b/lib/igt_core.h > > index a258348..5c5ee25 100644 > > --- a/lib/igt_core.h > > +++ b/lib/igt_core.h > > @@ -512,16 +512,20 @@ bool igt_run_in_simulation(void); > > void igt_skip_on_simulation(void); > > > > /* structured logging */ > > +#ifndef IGT_LOG_DOMAIN > > +#define IGT_LOG_DOMAIN (NULL) > > +#endif > > I guess the idea is that different source files have a > > #define IGT_LOG_DOMAIN "foobar" > > at the top before any includes? >
Yes, or defined with -D in CFLAGS. > > Would be great if you can add some good defaults as a patch on top: > - For everything in tests/*c I think we should add a suitable define in > the Makefile, perhaps "test" > The default domain for a test is the program name. > - Same for tools/ with "tools" > - lib/* should probably have per-file log domains, using the basename of > the source file itself. Dunno whether there's a cool Makefile trick to > make that happen. > The helper library is currently all within the same domain (defined in Makefile.am), but separate domains for different parts of the library would improve the filtering options. > > With that added I think patches 1-4 are a nice addition to the igt > logging. > -Daniel > -- > Daniel Vetter > Software Engineer, Intel Corporation > +41 (0) 79 365 57 48 - http://blog.ffwll.ch >
_______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx
