Previous Thread
Next Thread
Print Thread
Page 523 of 526 1 2 521 522 523 524 525 526
Re: SVN builds - new driver flood [Re: Darkstar] #114724
02/20/19 05:27 PM
02/20/19 05:27 PM
Joined: Aug 2011
Posts: 15
D
drewcifer Offline
Member
drewcifer  Offline
Member
D
Joined: Aug 2011
Posts: 15
That's super-great to hear! I'm itching for the chance to run Alias PowerAnimator or the original SoftImage again and show kids these days some of the really good features that many modern programs have never really replicated.

Man, I can still hear that sharp'ish metallic click sound when you do certain X-windows operations (was it closing a terminal? Opening?). Very happy to see any and all progress. Thanks again.

Re: SVN builds - new driver flood [Re: Anna Wu] #114726
02/20/19 10:35 PM
02/20/19 10:35 PM
Joined: May 2009
Posts: 1,794
J
Just Desserts Offline
Very Senior Member
Just Desserts  Offline
Very Senior Member
J
Joined: May 2009
Posts: 1,794
The latest fix at top-of-tree solves the hang that was occurring when double-clicking on things in IRIX. It turns out I was only triggering audio DMA on PBUS DMA channels 1 and 2, when it should have been 0-3. There was also an issue with reading back the current buffer address.

[Linked Image]

Re: SVN builds - new driver flood [Re: Anna Wu] #114728
02/21/19 11:03 PM
02/21/19 11:03 PM
Joined: May 2009
Posts: 1,794
J
Just Desserts Offline
Very Senior Member
Just Desserts  Offline
Very Senior Member
J
Joined: May 2009
Posts: 1,794
Some more Newport fixes. Now 'jot' works:

[Linked Image]

Re: SVN builds - new driver flood [Re: Anna Wu] #114729
02/22/19 12:39 AM
02/22/19 12:39 AM
Joined: Apr 2005
Posts: 556
GERMANY
Darkstar Offline
Senior Member
Darkstar  Offline
Senior Member
Joined: Apr 2005
Posts: 556
GERMANY
Wow, that is totally awesome. Have you tried running MipsPRO yet? I wonder if emulation is stable enough for something a bit more CPU-intensive...

BTW, the colors of the icons seem to be incorrect:
[Linked Image]

Still, very impressive work. Remind me to get you a beer or two sometime ;-)

Last edited by Darkstar; 02/22/19 12:39 AM.
Re: SVN builds - new driver flood [Re: Anna Wu] #114730
02/22/19 01:13 AM
02/22/19 01:13 AM
Joined: Mar 2001
Posts: 16,204
USA
R
R. Belmont Online content
Very Senior Member
R. Belmont  Online Content
Very Senior Member
R
Joined: Mar 2001
Posts: 16,204
USA
At a guess, R and G are swapped. Might be more than that though. Given it didn't run at all 2 weeks ago I'll take it smile

Re: SVN builds - new driver flood [Re: Anna Wu] #114731
02/22/19 09:19 AM
02/22/19 09:19 AM
Joined: May 2009
Posts: 1,794
J
Just Desserts Offline
Very Senior Member
Just Desserts  Offline
Very Senior Member
J
Joined: May 2009
Posts: 1,794
Yes, I'm well aware of the palette issues. Someone was courteous enough to send me screenshots of the relevant screens from an actual Indy with an 8-bit board running IRIX 5.3.

Currently I'm not quite sure why the palette is mangled, though I do know that I don't emulate the optional endianness swaps on Display Control Bus transfers. There's a chance that those palette entries are uploaded with that bit set, in which case the result would be more or less exactly as pictured. I'll be taking a look at it this evening.

Re: SVN builds - new driver flood [Re: Just Desserts] #114732
02/22/19 05:54 PM
02/22/19 05:54 PM
Joined: Aug 2011
Posts: 15
D
drewcifer Offline
Member
drewcifer  Offline
Member
D
Joined: Aug 2011
Posts: 15
How bizarre that a single window would use set such a bit to wiggle its palette around (the rest of your colors feel spot on), but hey, it's SGI, they could do whatever they wanted :-).

I also really find it humorous that the diagonal lines in buttonfly all seem to be offset - as letter-based groups, even. Like, all the diagonals from the capital S are laying down just above and to the right of it (mirrored, even). I wonder if they do rise-run encoding for diagonals or some other such trickery instead of the code you've got working now? Pretty cool stuff nonetheless.

Something else I pondered (and will learn from you I'm sure) - I wonder if their GL rasterizer is actually their 2d rasterizer as well? Eg. maybe you don't need a GL context on an X server in Irix - it's all a GL context or something? That'd be pretty rad.

Re: SVN builds - new driver flood [Re: drewcifer] #114733
02/23/19 07:50 AM
02/23/19 07:50 AM
Joined: Nov 2018
Posts: 7
J
Jim Miller Offline
Member
Jim Miller  Offline
Member
J
Joined: Nov 2018
Posts: 7
I've been lurking here for a while and I have to say this is awesome progress.

Perhaps you know this already based on implementation of the XMAP, but most SGI systems had WID planes that allow per-window bit depth and color map selection. Think of it as a per-pixel MUX for interpreting the contents of the 24-bit color buffer. This is something unique to SGI that dates back to the earliest machines where you could trade color depth for interaction (24-bit single-buffered vs 8 or 12-bit double-buffered). Or it could be used to provide a 4k color index mode.

I suspect the above issues may be with interpretation of the bits in the buffer. For example, borders were often rendered in the native depth for the root window, but the content could be in another mode based on the WID bits. In this case, the inner window is probably being interpreted with the wrong bit depth or the wrong color map entries.

There was a tool called 'mag' let you explore the frame buffer to see what's what. It would be interesting to see what it thinks those windows are.

One other thing... some systems like the Indigo implemented WID planes in an RLE list to save memory. It was only 64kx4, so it was possible to overwhelm it by doing things like allocating round windows. smile

Re: SVN builds - new driver flood [Re: Anna Wu] #114735
02/23/19 03:31 PM
02/23/19 03:31 PM
Joined: May 2009
Posts: 1,794
J
Just Desserts Offline
Very Senior Member
Just Desserts  Offline
Very Senior Member
J
Joined: May 2009
Posts: 1,794
I appreciate the info. That said, I've got it sorted already. In this case, the issue was due to simply not handling the XMAP9 mode table bits which specify a non-CI mode. There's a pair of bits in each 24-bit mode table entry which specify whether that particular mode should be handled as CI or one of three direct RGB mappings.

The VC2 chip used on the XL/Newport boards, too, uses an RLE list to determine the starting X position at which to invoke a given 5-bit DID, which gets transmitted to the XMAP9 to determine which of the 32 mode table entries should be active at a given point. smile

Re: SVN builds - new driver flood [Re: Anna Wu] #114736
02/23/19 03:50 PM
02/23/19 03:50 PM
Joined: May 2009
Posts: 1,794
J
Just Desserts Offline
Very Senior Member
Just Desserts  Offline
Very Senior Member
J
Joined: May 2009
Posts: 1,794
The latest changes in Git fix the login screen colors as well as icon colors in certain windows:

[Linked Image]

[Linked Image]

Page 523 of 526 1 2 521 522 523 524 525 526

Who's Online Now
3 registered members (mixmaster, Gregoire Duval, 1 invisible), 37 guests, and 3 spiders.
Key: Admin, Global Mod, Mod
Shout Box
Forum Statistics
Forums9
Topics8,653
Posts113,409
Members4,846
Most Online324
Dec 20th, 2018
Powered by UBB.threads™ PHP Forum Software 7.6.1.1
(Release build 20180111)
Page Time: 0.153s Queries: 14 (0.132s) Memory: 5.7359 MB (Peak: 5.9587 MB) Zlib enabled. Server Time: 2019-05-23 15:05:35 UTC