topic Re: Possible Direct3D 11 API implementation bug in HD 4000 driver in Graphics

Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Wed, 25 Oct 2017 23:52:50 GMT

Category

Questions

Answers (N/A if not applicable)

Description

Provide a detailed description of the issue

Possible Direct3D 11 API implementation bug in HD 4000 driver

Please place an X to the right of the option showing how often you see this issue using specific steps. (Ex: 'Every few times a game is started it flickers.' <- This would be "Often")<p>

Always (100%):X

Often (51-99%):

Sporadic (20-50%):

Very Sporadic (<20%):

Hardware (HW)

Brand and Model of the system.

i5-3570K

Hybrid or switchable graphics system?

ie Does it have AMD or NV graphics too?

Hybrid, but does not matter

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)<p style="margin-bo...

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

idata — Thu, 26 Oct 2017 15:59:47 GMT

Hello alexanderi_gpu,

Thank you for joining the Graphics Community. I am sorry to hear you are having issues with this matter.

I would like to gather more information regarding this behavior;

Provide the model of your graphics card.
Provide the model of your computer model.
The app you developed is the only one that shows this behavior, correct?

Thank you.

Regards,

Amy C.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Thu, 26 Oct 2017 16:26:27 GMT

Provide the model of your graphics card.

Intel HD Graphics 4000

Provide the model of your computer model.

Custom built PC using ASUS P8H77-I mainboard with Intel Core i5-3570K proccessor (family 6 model 58 stepping 9)

The app you developed is the only one that shows this behavior, correct?

It is the only one, that I have for now, and it passes all the validation checks, imposed by Microsoft Direct3D 11 API, and works fine on any other Direct3D 11 compatible hardware, including Intel HD Graphics 530.

It would be great to know if the strange behavior of the app is reproduced in your labs. If it helps, I can redirect the issue to the driver development team.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

idata — Thu, 26 Oct 2017 18:52:13 GMT

/thread/118987 alexanderi_gpu, thank you.

By graphics card I mean the physical card not the integrated graphics controller. Do you know the model? Is it an AMD graphics card?

Regards,

Amy C.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Thu, 26 Oct 2017 20:00:34 GMT

AMD Vega Frontier Edition

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

idata — Fri, 27 Oct 2017 18:38:19 GMT

Thank you for the information.

Does this behavior happen while using graphics driver from ASUS?

If you remove the graphics card, does the behavior happen?

Regards,

Amy C.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Mon, 30 Oct 2017 09:33:25 GMT

Does this behavior happen while using graphics driver from ASUS?

If you remove the graphics card, does the behavior happen?

I will check this out a little later. But the very same problem occures in HD Graphics 4000 ( driver ver. 10.18.10.4276 ) on a different machine:

Custom built PC running Microsoft Windows 7 SP1 Ultimate edition with Intel Core i7-3770K processor and 2 descreet graphics cards:

AMD Radeon R7 200 Series

NVIDIA GeForce GT 630

That's why I think it is an Intel driver bug.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Mon, 30 Oct 2017 12:26:25 GMT

The very same issue is in integrated Intel HD Graphics driver on Intel Atom Z3745 processor running on Windows 10 Professional on tablet FC-5000 manufactured by Topcon

It seems like it is reproduced on any Intel integrated graphics of 7th generation with Windows OS.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

idata — Mon, 30 Oct 2017 19:47:34 GMT

What was the model of the 7th system you tested?

Regards,

Amy C.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Mon, 30 Oct 2017 21:25:08 GMT

What was the model of the 7th system you tested?

It is not a system, but the 7th generation of Intel graphics systems.

https://en.wikipedia.org/wiki/List_of_Intel_graphics_processing_units# Seventh_generation List of Intel graphics processing units - Wikipedia

It includes Intel HD Graphics 4000, Intel HD Graphics 2500 and Intel HD Graphics

So far I've tested:

Intel Atom Z3745, with Intel HD Graphics

Intel Core i5-3570K, with Intel HD Graphics 4000

Intel Core i7-3770K , with Intel HD Graphics 4000

All processors have Ivy Bridge architecture.

Works fine on Intel Core i5-6600 with Intel HD Graphics 530

I've added short videos when error is present and when everything works fine, see attachements in the first post.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

idata — Tue, 31 Oct 2017 17:14:34 GMT

/thread/118987 alexanderi_gpu, thank you for that information. I will double check this matter and update the thread as soon as possible.

In the meantime, I would like to ask the community the following: do we have any other instance of this bug manifesting itself on a publicly available application?

Regards,

Amy C.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Tue, 31 Oct 2017 19:45:41 GMT

If the driver development team is concerned by this issue, i'm ready to provide a very detailed issue description (beyond what's written in description.txt) or a source code of the demo app.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

RonaldM_Intel — Wed, 06 Dec 2017 21:38:39 GMT

Hello Alexander,

Can you please provide us with a detailed issue description and specifics on how to reproduce it? (like you mentioned, a source code of the demo app).

The reason I'm asking is because 3rd generation graphics are quite old now, and so far this appears to be an obscure bug (i.e. not affecting too many users or even popular applications), thus I might need to provide compelling justification to our development team so that they pick this one from the queue.

I truly appreciate you taking the time to report this to us.

Best Regards,

Ronald M.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Wed, 13 Dec 2017 20:00:48 GMT

Hi, Sorry, I was busy lately. Concerning the source code, I have a project with a demo app, which can reproduce the bug. It can be assembled via CMake and built by MS Visual studio and does not require any dependencies.

Besides the issue text description, I was wondering, in what form You'd like to have a source code. I can send a whole project with a demo app, or I can try to prepare a small demo, which reproduces the bug, but it will take time.

Another annoying bug I found, is present in OpenCL driver implementation, which causes HD4000 driver to stop responding. It also is present on Ivy Bridge Integrated Graphics only. Should I create a separate thread or I can attach it here?

For now You can try to run a demo app in demo1.zip attachment on any Ivy Bridge desktop running MS Windows 7+ to see if the animation is working or not on your side. If the app is running and animation does not happen, - you are dealing with a driver error.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

RonaldM_Intel — Fri, 29 Dec 2017 18:16:25 GMT

Hi again Alexander,

Thanks for the details. I'll move this internally to determine if we're moving on with further debugging. I'll like to point out that there is a chance that this won't be picked by our devs if the bug is obscure and cannot be replicated on popular applications.

I'll reply here as soon as I have further details.

Best Regards,

Ronald M.

Re: Possible Direct3D 11 API implementation bug in HD 4000 driver

ABich1 — Fri, 29 Dec 2017 18:44:00 GMT

Hi, I've decided it's more beneficial to just blacklist certain hardware with drivers issues and move on, than to spend time on fixing it. It's really getting too old. Since issue is not the case on more recent hardware, You can close the thread and the issue.