Created attachment 47875 [details]
Model rendered too bright

It looks like ETQW suffers from a few other rendering glitches. The characters in the team selection seems to be rendered too bright (but are fine in-game), and trees on the map "Island" looks quite funky.

I'm not sure if this is the same problems found in bug 35434.

Created attachment 47876 [details]
Trees on the "Island" map

Rendering of trees (and water) is fixed on master, but broken on non-evergreen, see bug 38280.

(In reply to comment #1)
> Created an attachment (id=47875) [details]
> Model rendered too bright
> 
> It looks like ETQW suffers from a few other rendering glitches. The characters
> in the team selection seems to be rendered too bright (but are fine in-game),
> and trees on the map "Island" looks quite funky.

I tested with my 4890 and the globe renders OK, but the menu model is too bright.

I've no idea when this started as I normally test in spectate mode or timenetdemo - so don't usually see it.

(In reply to comment #4)
> 
> I tested with my 4890 and the globe renders OK, but the menu model is too
> bright.

I guess the globe is evergreen-only then, thanks for testing and commenting.

> I've no idea when this started as I normally test in spectate mode or
> timenetdemo - so don't usually see it.

I've tried a few different older versions, but so far I haven't found one where the model renders rigth, so I'm not sure if it's a regression or not.

Created attachment 48072 [details] [review]
[PATCH] r600g: fix RSQ to use abs value of operand on evergreen

This patch should fix globe rendering on evergreen.

(In reply to comment #6)
> Created an attachment (id=48072) [details]
> [PATCH] r600g: fix RSQ to use abs value of operand on evergreen
> 
> This patch should fix globe rendering on evergreen.

Confirmed, it's working fine now! :)

(In reply to comment #4)
> I tested with my 4890 and the globe renders OK, but the menu model is too
> bright.
> 
> I've no idea when this started as I normally test in spectate mode or
> timenetdemo - so don't usually see it.

I poked around a bit more, and it turns out that this only happens when "Lightning quality" is set to normal, it seems to be okay when set to high.

Doesn't seem to be a regression, it's present as far back as when S3TC support first was added, and I don't think the game runs without it.

(In reply to comment #8)
> (In reply to comment #4)
> > I tested with my 4890 and the globe renders OK, but the menu model is too
> > bright.
> > 
> > I've no idea when this started as I normally test in spectate mode or
> > timenetdemo - so don't usually see it.
> 
> I poked around a bit more, and it turns out that this only happens when
> "Lightning quality" is set to normal, it seems to be okay when set to high.
> 
> Doesn't seem to be a regression, it's present as far back as when S3TC support
> first was added, and I don't think the game runs without it.

Yea, it's OK with lighting high for me also.

Unfortunately setting high causes a rendering error for me - distant foliage is rendered black.

Created attachment 48117 [details] [review]
[PATCH] r600g: fix LIT to handle src==dst properly

> Unfortunately setting high causes a rendering error for me - distant foliage is
> rendered black.

This patch should fix it.

(In reply to comment #10)
> Created an attachment (id=48117) [details]
> [PATCH] r600g: fix LIT to handle src==dst properly
> 
> > Unfortunately setting high causes a rendering error for me - distant foliage is
> > rendered black.
> 
> This patch should fix it.

It does, thanks.

The patch has been merged to master, so let's close this bug. Thanks for all your work on this Vadim! :)

I'm going to open a new bug about the rendering glitch with lightning set to normal.