Home Original page

[OpenJDK 2D-Dev] [9] Review Request: 8146042 Offscreen rendering is different from onscreen one

Philip Race philip.race at oracle.com
Fri Sep 9 15:16:09 UTC 2016 
I think this bug should be downgraded to P4 and deferred.
Put a link in the bug report to the fix so we can revive it later.

-phil.

On 9/9/16, 6:55 AM, Semyon Sadetsky wrote:
>
>
> On 9/9/2016 4:08 PM, Vadim Pakhnushev wrote:
>> On 09.09.2016 15:58, Semyon Sadetsky wrote:
>>> On 09.09.2016 14:45, Vadim Pakhnushev wrote:
>>>
>>>>
>>>>
>>>>>> My cards are HD Graphics 3000 (0x8086/0x0126) with 9.17.10.4101 and
>>>>>> ATI Radeon HD 5700 (0x1002/0x68b8) with driver 8.17.10.1333
>>>>> ATI is unrelated to this fix. Could you upgrade to the latest driver
>>>>> from https://downloadcenter.intel.com/search?keyword=hd+graphics+3000
>>>>> and try to reproduce 803944 again?
>>>>
>>>> Alas, the issue is still there even with the latest driver:
>>>> https://i.imgur.com/1kXVbaI.png
>>> I definitely cannot reproduce this picture on my system. Maybe it 
>>> take place for old Intel cards only. HD 3000 is  5+ years old and 
>>> probably it does not fully support Windows 7 d3d. HD graphics 
>>> declares support for dx11 starting from HD4000.
>>> This bug and all related were resolved by disabling d3d for all 
>>> Intel video cards. We could at least to defer those bugs and disable 
>>> d3d in a separate bug as a temp solution. We are forgetting about 
>>> Intel d3d forever?
>>> Its a pity, because Intel APUs as main graphics card are rather 
>>> popular. Ultrabooks rarely have discrete video.
>>
>> We had a lot of reports on HD4000 as well, listed as duplicates of 
>> 803944.
>> DX11 support is not relevant here at all since we are only use DX9.
>>
>> Since your fix explicitly checks for Intel hardware, it doesn't make 
>> any sense to use it without unblocking them.
>> Moreover, I can't see how this bug (8146042) can be reproduced 
>> without modified build which unblocks Intel hardware.
> This may be explained. D3D on Intel was disabled in April while this 
> bug fix was published in January. Before it was rather annoying. But 
> it seems 803944 fix has resolved all all such issues for all Intel 
> cards at once. And I have lost possibility to reproduce d3d related 
> issues because I have no discrete card and even an option to force d3d 
> is not provided.
>
> --Semyon
>>
>> Vadim
>
More information about the 2d-dev mailing list