Re: [Mesa-dev] DRI Configurator replacement announcement

2018-01-18 Thread Michel Dänzer
On 2018-01-17 08:39 PM, Jean Hertel wrote:
> De: Gert Wollny <gw.foss...@gmail.com>
> Enviado: terça-feira, 16 de janeiro de 2018 05:33
> Para: Jean Hertel; mesa-dev@lists.freedesktop.org
> Assunto: Re: [Mesa-dev] DRI Configurator replacement announcement
>   
>>> - Applications with empty options (all options are the same as
>>> system-wide config or driver default) will be removed automatically
>> I don't think that is such a good idea, what happens if someone wants
>> to try out options to compare to the defaults? The application
>> shouldn't get removed automatically when the user sets all to default.
> 
> I changed this with ed24a672a68529d6aa8e8d886009605411ae5379.
> Now only system-wide applications that have all empty options get removed 
> (after all, they will still be there on the next run, and it doesn't make 
> sense to write empty tags on the user's .drirc file)
> 
> 
> Also, if anyone else reading this has a PRIME setup (two GPUs from same 
> vendor, usually found in notebooks),

FWIW, the GPUs can be from different vendors.


> I would be very happy to take a look on it so the software can correctly 
> support it. 
> If this is your case, please open a github issue and post the contents of the 
> following commands for me:
> xdriinfo

This only reports the primary GPU of each X11 screen.


-- 
Earthling Michel Dänzer   |   http://www.amd.com
Libre software enthusiast | Mesa and X developer
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev


Re: [Mesa-dev] DRI Configurator replacement announcement

2018-01-18 Thread Jean Hertel
De: Gert Wollny <gw.foss...@gmail.com>
Enviado: terça-feira, 16 de janeiro de 2018 05:33
Para: Jean Hertel; mesa-dev@lists.freedesktop.org
Assunto: Re: [Mesa-dev] DRI Configurator replacement announcement
  
>> - Applications with empty options (all options are the same as
>> system-wide config or driver default) will be removed automatically
>I don't think that is such a good idea, what happens if someone wants
>to try out options to compare to the defaults? The application
>shouldn't get removed automatically when the user sets all to default.

I changed this with ed24a672a68529d6aa8e8d886009605411ae5379.
Now only system-wide applications that have all empty options get removed 
(after all, they will still be there on the next run, and it doesn't make sense 
to write empty tags on the user's .drirc file)


Also, if anyone else reading this has a PRIME setup (two GPUs from same vendor, 
usually found in notebooks), I would be very happy to take a look on it so the 
software can correctly support it. 
If this is your case, please open a github issue and post the contents of the 
following commands for me:
xdriinfo
for each screen/driver also run:
xdriinfo options SCREENNUMBER

Please provide also your local configurations located under $HOME/.drirc

The contributions are really appreciated.

Kind Regards,
Jean Hertel
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev


Re: [Mesa-dev] DRI Configurator replacement announcement

2018-01-16 Thread Gustaw Smolarczyk
2018-01-16 8:33 GMT+01:00 Gert Wollny :

> Hello Jean,
>
> Am Montag, den 15.01.2018, 20:15 + schrieb Jean Hertel:
> > I have written a simply application like DRI Conf tool.
> > It is written using GTKmm and C++.
> Great!
>
> Unfortunately, it didn't link properly, so I send you a pull request to
> correct the Boost_LOCALE dependency.
>
>   https://github.com/jlHertel/adriconf/pull/1
>
> Then it crashed with std::bas_alloc that also printed a Python (?!)
> ValueError (opened an issue for that):
>
> https://github.com/jlHertel/adriconf/issues/2


FYI, the python exception is internal to gdb that probably couldn't use the
pretty printer for the mentioned type.

Gustaw
___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev


Re: [Mesa-dev] DRI Configurator replacement announcement

2018-01-15 Thread Gert Wollny
Hello Jean, 

Am Montag, den 15.01.2018, 20:15 + schrieb Jean Hertel:
> I have written a simply application like DRI Conf tool.
> It is written using GTKmm and C++.
Great! 

Unfortunately, it didn't link properly, so I send you a pull request to
correct the Boost_LOCALE dependency. 
  
  https://github.com/jlHertel/adriconf/pull/1

Then it crashed with std::bas_alloc that also printed a Python (?!) 
ValueError (opened an issue for that):

https://github.com/jlHertel/adriconf/issues/2

Because of this I could not really test the application. I'll see
whether I can find the culprit later.

> Main Features (apart from what is already available in driconf):
> - Automatic removal of invalid options (Options that the driver
> doesn't support at all)
> - Options that have the same value as the system wide options or
> driver default will be ignored
Makes sense. 

> - Applications with empty options (all options are the same as
> system-wide config or driver default) will be removed automatically
I don't think that is such a good idea, what happens if someone wants
to try out options to compare to the defaults? The application
shouldn't get removed automatically when the user sets all to default.

> Current TODOs:
[...]
> - Tests? Implementing testing for the software would be very nice
+1

> - Remove Boost dependency? (Currently boost.locale is used to get the
> ISO-639 language id, to properly parse Mesa3d translations)
If the OS provides a propper packaging system having one dependency
more should not be a problem. 

I'll take a look at the code itself later. 

Best, 
Gert 

___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev