Re: Linux-x86-64 and cairo crash

[Eric Wasylishen]

Hi Riccardo,

On 2011-09-15, at 2:07 AM, Riccardo Mottola wrote:

> Hi
>> Ah, I see. Transparent windows works fine for me with XGCairoSurface now - 
>> perhaps it was a limitation in cairo that was fixed since then.
>> 
>> btw, I added the test to GSTest that I was meaning to write for a while, for 
>> testing -[NSWindow setAlphaValue:] and -[NSWindow setBackgroundColor:] for 
>> colors with an alpha<  1. It's called "Transparency-test".
>> 
> just out of curiosity I tried your patch anyway.
> 
ok, you mean you tried switching to XGCairoSurface?

> On the machine exporting to 15/16bit display it did not help at all, same 
> behaviour as before.

Hm, that suggests problem is elsewhere in the back/x11 code then. Is this the 
first time you tested this configuration, or was it working in the past? It may 
be a bug in back/x11 (XWindowBuffer?) that has been there for a long time.

> I tried it also on another computer with a standard GeForce card, not 
> exported. I don't know if the display is 24 or 32bit. How do I know best?
> I noticed a slight speed increase in certain operation, scrolling and 
> resizing seemed a bit better.
> 
> The transparency test didn't work for me before and continued not to work. It 
> says it needs a special compositing windowmanager? I just run windowmaker.

yeah. I use gnome's metacity which has an option to turn on compositing.

another thing you can do is run xcompmgr along with your window manager. 
However, last time I checked windowmaker was incompatible with xcompmgr for 
some reason. :-(

> I always wonder that the animation test makes about 29 frames and the machine 
> becomes almost completely unresponsive. With the new surface type behaviour 
> remained about the same.
> 

It uses close to 100% cpu for me too.

Eric