[9025695] Demo "crystal skull" fails to compile in Shadertoy, after converting to GLSL hacker format it renders weird waves.

[Stefan3D]

Demo https://www.shadertoy.com/view/MsS3WV "crystal skull" fails to compile in Shadertoy, after converting to GLSL hacker format it renders weird waves.

This is with HD5500, works fine.with competitors' GPUs.

[Bryce__Intel]

Hi Stefan3D,

Please provide the details from the sticky so I can get this replicated and proceed with investigation. OGL team already previewed the post and says it's an unknown bug and agreed to look into it further.

.:Bryce:.

[Stefan3D]

Detailed infos except from Direct3D in DXDIAG.TXT, for some reason it crashed today.

EDIT: new DXDIAG.TXT attached, crash was caused by GeForce driver (nvwgf2um.dll)

EDIT 2: the demo fails now to compile on Shadertoy.com with any GPU, i guess they made some changes on server side.

Category

Questions

Answers (N/A if not applicable)

Description

Provide a detailed description of the issue AND 'does it fail every single time, or only sometimes?' If you can offer a % rate please do.

Demo https://www.shadertoy.com/view/MsS3WV "crystal skull" fails to compile in Shadertoy, after converting to GLSL hacker format it renders weird waves.

rate: 100%

Demo renders correcty with NVIDIA or AMD GPUs

Hardware (HW)

Brand and Model of the system.

Acer E5-571G

Hybrid or switchable graphics system? ie Does it have AMD or NV graphics too?

NVIDIA Optimus

Make and model of any Displays that are used to see the issue (see note2 below).

LFP = Local Flat Panel (Laptop panel)

EFP = External Flat Panel (Monitor you plug in) 

LFP

How much memory [RAM] in the system (see note2 below).

8 GB

Provide any other hardware needed to replicate the issue.

ie: Cables&brand, cable type [vga, hdmi, DP, etc], dock, dongles/adapters, etc

NA

Hardware Stepping (see note1 below).

PCI\VEN_8086&DEV_1616&SUBSYS_09401025&REV_09

PCI\VEN_8086&DEV_1616&SUBSYS_09401025

PCI\VEN_8086&DEV_1616&CC_030000

PCI\VEN_8086&DEV_1616&CC_0300

Software (SW)

...

[Bryce__Intel]

Thanks Stefan3D,

We're continuing investigation. I'll report back any updates.

[Bryce__Intel]

Hi Stefan3D,

The bug has been marked as resolved in our system with an internal build. The driver will go into testing and is targeting the next quarterly driver release.

[Stefan3D]

Looks like the fix didn't make it into driver v4352 you released 2 days ago.

Meanwhile i updated the demo for GeeXLab 0.91 with a different cubemap.

[Bryce__Intel]

Issue's not present on SKL with latest driver. Back in Dev for BDW. I'll update when more info is made available, stay tuned...

[Stefan3D]

At Bryce@Intel

BDW and SKL have different OpenGL drivers.

Maybe you forgot to add the fix to ig8icd*.dll ?

If you look at it, there's also a typo: "Declare used insrisnics"

[Bryce__Intel]

Hey Stefan3D,

The Dev was asking if you could try this. Let me know your thoughts. Thanks

In attachment you can find my proposal of slightly changed fragment shader from the "crystal skull" demo, extracted from Crystal_skull_gl2.xmlfrom communities thread: /message/362076# 362076 https://communities.intel.com/message/362076# 362076

How to repro: 2. download shader converted for GLSL Hacker from https://goo.gl/KxDL4B Crystal_skull_gl2.xml - Google Drive

I marked my changes in glsl code with // AIL

----

Because shader is "unstable" I slightly changed the code and with my modifications it works OK both on Intel and NVIDIA (I have the NVIDIA Quadro NVS 290). I'm wondering if the submitter would like to use my version of this shader instead of the original.

The interesting thing is that the original shader does 8-9 FPS while mine does 14-15 FPS J, tested on NVIDIA Quadro NVS 290.

[Stefan3D]

Bryce@Intel

All good now (with that shader fix)

In case of interest i found more shaders at Shadertoy.com that don't render properly on Intel.