[Mesa-dev] [PATCH 1/3] meson: Define NDEBUG for non-debug-builds

[Michel Dänzer]

On 2018-07-17 03:07 PM, Rob Clark wrote:
> On Tue, Jul 17, 2018 at 9:01 AM, Rob Clark <robdclark at gmail.com> wrote:
>> On Mon, Apr 16, 2018 at 5:19 PM, Dylan Baker <dylan at pnwbakers.com> wrote:
>>> Quoting Jan Alexander Steffens (heftig) (2018-04-14 10:23:20)
>>>> This is for parity with autotools. We were suddenly getting assertion
>>>> failures after switching to meson, see [1].
>>>>
>>>> [1]: https://bugs.archlinux.org/task/58218
>>>>
>>>> Signed-off-by: Jan Alexander Steffens (heftig) <jan.steffens at gmail.com>
>>>> ---
>>>>  meson.build | 2 ++
>>>>  1 file changed, 2 insertions(+)
>>>>
>>>> diff --git a/meson.build b/meson.build
>>>> index a4dfa62255..931e0389e6 100644
>>>> --- a/meson.build
>>>> +++ b/meson.build
>>>> @@ -705,6 +705,8 @@ endif
>>>>  # Define DEBUG for debug builds only (debugoptimized is not included on this one)
>>>>  if get_option('buildtype') == 'debug'
>>>>    pre_args += '-DDEBUG'
>>>> +else
>>>> +  pre_args += '-DNDEBUG'
>>>>  endif
>>>>
>>>>  if get_option('shader-cache')
>>>> --
>>>> 2.16.2
>>>
>>> NAK.
>>>
>>> meson has -Db_ndebug for controlling NDEBUG, which is done separately from
>>> optimization levels, this is in our documentation as well as the meson
>>> documentation. For builds without asserts, set b_ndebug=true.
>>>
>>
>> Is it possible to make b_ndebug default to true for release builds?
>> That would be far less surprising..
>>
> 
> just for context, debug_assert() (which can be used for expensive
> checks), nir_validate_shader() (quite expensive), etc are all
> controlled by NDEBUG.. so b_ndebug defaulting to false for release
> builds basically looks like a big performance regression compared to
> autotools builds.

FWIW, sounds like those things should really only be enabled when DEBUG
is defined.


-- 
Earthling Michel Dänzer               |               http://www.amd.com
Libre software enthusiast             |             Mesa and X developer