lightbridges not rendering correctly on open source AMD driver

[randomweej]

Processor Information:
Vendor: AuthenticAMD
CPU Family: 0x10
CPU Model: 0x4
CPU Stepping: 0x2
CPU Type: 0x0
Speed: 3093 Mhz
4 logical processors
4 physical processors
HyperThreading: Unsupported
FCMOV: Supported
SSE2: Supported
SSE3: Supported
SSSE3: Unsupported
SSE4a: Supported
SSE41: Unsupported
SSE42: Unsupported

Network Information:
Network Speed:

Operating System Version:
Ubuntu Trusty Tahr (development branch) (64 bit)
Kernel Name: Linux
Kernel Version: 3.13.4
X Server Vendor: The X.Org Foundation
X Server Release: 11500000
X Window Manager: GNOME Shell
Steam Runtime Version: steam-runtime-release_2014-02-05

Video Card:
Driver: X.Org Gallium 0.4 on AMD TURKS

Driver Version:  3.0 Mesa 10.1.0-rc1
OpenGL Version: 3.0
Desktop Color Depth: 24 bits per pixel
Monitor Refresh Rate: 59 Hz
VendorID:  0x1002
DeviceID:  0x6758
Number of Monitors:  1
Number of Logical Video Cards:  1
Primary Display Resolution:  1366 x 768
Desktop Resolution: 1366 x 768
Primary Display Size: 13.54" x 7.60"  (15.51" diag)
                                        34.4cm x 19.3cm  (39.4cm diag)
Primary VRAM Not Detected

Sound card:
Audio device: Realtek ALC888

Memory:
RAM: 7856 Mb

Miscellaneous:
UI Language: English
LANG: en_GB.UTF-8
Microphone: Not set
Total Hard Disk Space Available: 442411 Mb
Largest Free Hard Disk Block: 360934 Mb

[ghost]

I think this is related to #101.
I've got the same issue when I use bilinear or trilinear filtering. But with anisotropic filtering it's fine.

Mesa 10.2.0-devel (git-80c1b93)

[stevethepocket]

I'm getting something like that too. Can't switch to the official drivers because they don't support my card anymore (Radeon HD 4850). Looks a bit different on mine though:

Note how the glitchy area seems to be confined to the first mipmap zone. This is from sp_a2_bridge_the_gap. I've also got some weird stuff going on with the fluorescent lights:

It doesn't seem to affect other transparent textures with refraction mapping, but I can't be sure. All I know is, something is rendering at half the proper resolution. Note the top of the image especially.

The thermal discouragement beams are a bit off as well, once again mainly in the first mipmap zone:

All of this is, just like you said, not happening when I turn anisotropic filtering on.

[davidw-valve]

Please report this issue to your driver vendor.

[stevethepocket]

FYI, whatever you did in the last patch affected this. Now the bridges look like this:

So, better I guess. Among other things, I can see that mipmap transitions are clearly not abrupt anymore like they were before (also evident on catwalk floors).

[stevethepocket]

We may as well close this as a duplicate of #101, especially since I went to the trouble of getting it reopened.