[Mesa-dev] Compile error in nir/nir_builder.h in current head
Emil Velikov
emil.l.velikov at gmail.com
Thu Sep 17 06:27:31 PDT 2015
Hi Gottfried,
On 17 September 2015 at 13:42, Gottfried Haider
<gottfried.haider at gmail.com> wrote:
> I am getting this error on 7e286506 - comping mesa used to work fine
> on this system a couple of weeks ago
>
> CXX nir/nir_lower_samplers.lo
> CC nir/nir_lower_system_values.lo
> In file included from nir/nir_lower_samplers.cpp:27:0:
> nir/nir_builder.h: In function 'nir_ssa_def*
> nir_imm_float(nir_builder*, float)':
> nir/nir_builder.h:79:28: error: expected primary-expression before '.' token
> nir/nir_builder.h:79:44: warning: extended initializer lists only
> available with -std=c++0x or -std=gnu++0x [enabled by default]
> nir/nir_builder.h: In function 'nir_ssa_def*
> nir_imm_vec4(nir_builder*, float, float, float, float)':
> nir/nir_builder.h:86:28: error: expected primary-expression before '.' token
> nir/nir_builder.h:86:44: warning: extended initializer lists only
> available with -std=c++0x or -std=gnu++0x [enabled by default]
> nir/nir_builder.h: In function 'nir_ssa_def* nir_imm_int(nir_builder*, int)':
> nir/nir_builder.h:93:28: error: expected primary-expression before '.' token
> nir/nir_builder.h:93:44: warning: extended initializer lists only
> available with -std=c++0x or -std=gnu++0x [enabled by default]
> Makefile:1525: recipe for target 'nir/nir_lower_samplers.lo' failed
>
This is caused with the introduction of
commit ef8eebc6ad5d86e524426f0755c0f7d0b4c0cd3e
Author: Timothy Arceri <t_arceri at yahoo.com.au>
Date: Wed Aug 26 22:18:36 2015 +1000
nir: support indirect indexing samplers in struct arrays
The commit adds nir_builder.h to nir_lower_samplers.cpp, with the
former containing designated initializer(s). As this is not allowed by
the the standard (albeit some compilers allow it) you see the error.
>From a very quick look I cannot see any specific reason why we cannot
rename the file to nir_lower_samplers.c, although someone more
familiar with the code will be the judge :-)
Thanks
Emil
More information about the mesa-dev
mailing list