[Mesa-dev] [PATCH] nir: acknowledge the existence of nir_builder.h

[Matt Turner]

On Mon, Mar 30, 2015 at 10:23 AM, Emil Velikov <emil.l.velikov at gmail.com> wrote:
> On 30 March 2015 at 18:10, Matt Turner <mattst88 at gmail.com> wrote:
>> No, it doesn't make sense. Presumably you added it because it fixed a
>> build error? Maybe if you showed what the error was it would make
>> sense.
>>
> Hmm... ok. Can you elaborate if on the scenario mentioned above "one
> may try to "build" nir.h ..." is feasible or a complete BS ? I was
> under the impression that make will not wait for all the BUILT_SOURCES
> to be generated before starting the actual build. Could be wrong
> though :-)

I don't know what "building nir.h" would mean. nir.h isn't a generated file.

The nir/nir.h: nir/nir_opcodes.h dependency looks entirely bogus. Even
if nir.h was a generated file that included nir_opcodes.h, the
dependence would only matter when compiling something that included
nir.h, and make handles header dependencies.

But yes, things listed in BUILT_SOURCES are built before the rest of
the build begins. If you specify "SUBDIRS = foo bar ." in a
Makefile.am with BUILT_SOURCES, the BUILT_SOURCES are even built
before recursing into foo and bar.

I've confirmed that removing the bogus dependency does not cause any
built problems in a clean out-of-tree build.