Mac VM display broken for smaller bit depths

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Mac VM display broken for smaller bit depths

Bert Freudenberg
This hack avoids entering an unusable state. OTOH it makes debugging it more difficult. What to do?

- Bert -

On Tue, Sep 13, 2016 at 10:01 PM, <[hidden email]> wrote:
Bert Freudenberg uploaded a new version of Graphics to project The Trunk:
http://source.squeak.org/trunk/Graphics-bf.365.mcz

==================== Summary ====================

Name: Graphics-bf.365
Author: bf
Time: 13 September 2016, 10:01:12.119134 pm
UUID: 184228fa-4e38-472a-8e51-1cd86ff433f4
Ancestors: Graphics-tfel.364

Work around Mac VM bug: use 32 bpp

=============== Diff against Graphics-tfel.364 ===============

Item was added:
+ ----- Method: DisplayScreen>>primSupportsDisplayDepth: (in category 'other') -----
+ primSupportsDisplayDepth: pixelDepth
+       "Return true if this pixel depth is supported on the current host platform.
+       Primitive. Optional."
+       <primitive: 91>
+       ^#(1 2 4 8 16 32) includes: pixelDepth!

Item was changed:
  ----- Method: DisplayScreen>>supportsDisplayDepth: (in category 'other') -----
  supportsDisplayDepth: pixelDepth
+       "Return true if this pixel depth is supported on the current host platform."
+       Smalltalk platformName = 'Mac OS' ifTrue: [^pixelDepth abs = 32]. "Work around VM bug"
+       ^self primSupportsDisplayDepth: pixelDepth!
-       "Return true if this pixel depth is supported on the current host platform.
-       Primitive. Optional."
-       <primitive: 91>
-       ^#(1 2 4 8 16 32) includes: pixelDepth!





Reply | Threaded
Open this post in threaded view
|

Re: Mac VM display broken for smaller bit depths

Tobias Pape
Hi,

On 14.09.2016, at 09:57, Bert Freudenberg <[hidden email]> wrote:

> This hack avoids entering an unusable state. OTOH it makes debugging it more difficult. What to do?

Is the unstable state present in cocoa or carbon VMs?
Best regards
        -Tobias

>
> - Bert -
>
> On Tue, Sep 13, 2016 at 10:01 PM, <[hidden email]> wrote:
> Bert Freudenberg uploaded a new version of Graphics to project The Trunk:
> http://source.squeak.org/trunk/Graphics-bf.365.mcz
>
> ==================== Summary ====================
>
> Name: Graphics-bf.365
> Author: bf
> Time: 13 September 2016, 10:01:12.119134 pm
> UUID: 184228fa-4e38-472a-8e51-1cd86ff433f4
> Ancestors: Graphics-tfel.364
>
> Work around Mac VM bug: use 32 bpp
>
> =============== Diff against Graphics-tfel.364 ===============
>
> Item was added:
> + ----- Method: DisplayScreen>>primSupportsDisplayDepth: (in category 'other') -----
> + primSupportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform.
> +       Primitive. Optional."
> +       <primitive: 91>
> +       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
> Item was changed:
>   ----- Method: DisplayScreen>>supportsDisplayDepth: (in category 'other') -----
>   supportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform."
> +       Smalltalk platformName = 'Mac OS' ifTrue: [^pixelDepth abs = 32]. "Work around VM bug"
> +       ^self primSupportsDisplayDepth: pixelDepth!
> -       "Return true if this pixel depth is supported on the current host platform.
> -       Primitive. Optional."
> -       <primitive: 91>
> -       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
>
>
>


Reply | Threaded
Open this post in threaded view
|

Re: Mac VM display broken for smaller bit depths

Bert Freudenberg
On Wed, Sep 14, 2016 at 10:02 AM, Tobias Pape <[hidden email]> wrote:
Hi,

On 14.09.2016, at 09:57, Bert Freudenberg <[hidden email]> wrote:

> This hack avoids entering an unusable state. OTOH it makes debugging it more difficult. What to do?

Is the unstable state present in cocoa or carbon VMs?
Best regards
        -Tobias

Cocoa I would think? A current Spur VM. The display just looks garbled e.g. when choosing 16 bpp. I wouldn't call it "unstable" just "unusable".

- Bert -

 

>
> - Bert -
>
> On Tue, Sep 13, 2016 at 10:01 PM, <[hidden email]> wrote:
> Bert Freudenberg uploaded a new version of Graphics to project The Trunk:
> http://source.squeak.org/trunk/Graphics-bf.365.mcz
>
> ==================== Summary ====================
>
> Name: Graphics-bf.365
> Author: bf
> Time: 13 September 2016, 10:01:12.119134 pm
> UUID: 184228fa-4e38-472a-8e51-1cd86ff433f4
> Ancestors: Graphics-tfel.364
>
> Work around Mac VM bug: use 32 bpp
>
> =============== Diff against Graphics-tfel.364 ===============
>
> Item was added:
> + ----- Method: DisplayScreen>>primSupportsDisplayDepth: (in category 'other') -----
> + primSupportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform.
> +       Primitive. Optional."
> +       <primitive: 91>
> +       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
> Item was changed:
>   ----- Method: DisplayScreen>>supportsDisplayDepth: (in category 'other') -----
>   supportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform."
> +       Smalltalk platformName = 'Mac OS' ifTrue: [^pixelDepth abs = 32]. "Work around VM bug"
> +       ^self primSupportsDisplayDepth: pixelDepth!
> -       "Return true if this pixel depth is supported on the current host platform.
> -       Primitive. Optional."
> -       <primitive: 91>
> -       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
>
>
>





Reply | Threaded
Open this post in threaded view
|

Re: Mac VM display broken for smaller bit depths

Nicolas Cellier
could we bissect from which version this happened?

2016-09-14 10:13 GMT+02:00 Bert Freudenberg <[hidden email]>:
On Wed, Sep 14, 2016 at 10:02 AM, Tobias Pape <[hidden email]> wrote:
Hi,

On 14.09.2016, at 09:57, Bert Freudenberg <[hidden email]> wrote:

> This hack avoids entering an unusable state. OTOH it makes debugging it more difficult. What to do?

Is the unstable state present in cocoa or carbon VMs?
Best regards
        -Tobias

Cocoa I would think? A current Spur VM. The display just looks garbled e.g. when choosing 16 bpp. I wouldn't call it "unstable" just "unusable".

- Bert -

 

>
> - Bert -
>
> On Tue, Sep 13, 2016 at 10:01 PM, <[hidden email]> wrote:
> Bert Freudenberg uploaded a new version of Graphics to project The Trunk:
> http://source.squeak.org/trunk/Graphics-bf.365.mcz
>
> ==================== Summary ====================
>
> Name: Graphics-bf.365
> Author: bf
> Time: 13 September 2016, 10:01:12.119134 pm
> UUID: 184228fa-4e38-472a-8e51-1cd86ff433f4
> Ancestors: Graphics-tfel.364
>
> Work around Mac VM bug: use 32 bpp
>
> =============== Diff against Graphics-tfel.364 ===============
>
> Item was added:
> + ----- Method: DisplayScreen>>primSupportsDisplayDepth: (in category 'other') -----
> + primSupportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform.
> +       Primitive. Optional."
> +       <primitive: 91>
> +       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
> Item was changed:
>   ----- Method: DisplayScreen>>supportsDisplayDepth: (in category 'other') -----
>   supportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform."
> +       Smalltalk platformName = 'Mac OS' ifTrue: [^pixelDepth abs = 32]. "Work around VM bug"
> +       ^self primSupportsDisplayDepth: pixelDepth!
> -       "Return true if this pixel depth is supported on the current host platform.
> -       Primitive. Optional."
> -       <primitive: 91>
> -       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
>
>
>









Reply | Threaded
Open this post in threaded view
|

Re: Mac VM display broken for smaller bit depths

Bert Freudenberg
Has it ever worked on the cocoa VM?

- Bert -

On Wed, Sep 14, 2016 at 1:51 PM, Nicolas Cellier <[hidden email]> wrote:
could we bissect from which version this happened?

2016-09-14 10:13 GMT+02:00 Bert Freudenberg <[hidden email]>:
On Wed, Sep 14, 2016 at 10:02 AM, Tobias Pape <[hidden email]> wrote:
Hi,

On 14.09.2016, at 09:57, Bert Freudenberg <[hidden email]> wrote:

> This hack avoids entering an unusable state. OTOH it makes debugging it more difficult. What to do?

Is the unstable state present in cocoa or carbon VMs?
Best regards
        -Tobias

Cocoa I would think? A current Spur VM. The display just looks garbled e.g. when choosing 16 bpp. I wouldn't call it "unstable" just "unusable".

- Bert -

 

>
> - Bert -
>
> On Tue, Sep 13, 2016 at 10:01 PM, <[hidden email]> wrote:
> Bert Freudenberg uploaded a new version of Graphics to project The Trunk:
> http://source.squeak.org/trunk/Graphics-bf.365.mcz
>
> ==================== Summary ====================
>
> Name: Graphics-bf.365
> Author: bf
> Time: 13 September 2016, 10:01:12.119134 pm
> UUID: 184228fa-4e38-472a-8e51-1cd86ff433f4
> Ancestors: Graphics-tfel.364
>
> Work around Mac VM bug: use 32 bpp
>
> =============== Diff against Graphics-tfel.364 ===============
>
> Item was added:
> + ----- Method: DisplayScreen>>primSupportsDisplayDepth: (in category 'other') -----
> + primSupportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform.
> +       Primitive. Optional."
> +       <primitive: 91>
> +       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
> Item was changed:
>   ----- Method: DisplayScreen>>supportsDisplayDepth: (in category 'other') -----
>   supportsDisplayDepth: pixelDepth
> +       "Return true if this pixel depth is supported on the current host platform."
> +       Smalltalk platformName = 'Mac OS' ifTrue: [^pixelDepth abs = 32]. "Work around VM bug"
> +       ^self primSupportsDisplayDepth: pixelDepth!
> -       "Return true if this pixel depth is supported on the current host platform.
> -       Primitive. Optional."
> -       <primitive: 91>
> -       ^#(1 2 4 8 16 32) includes: pixelDepth!
>
>
>
>