I try to use PowerVR OpenGL ES 2.0 Emulator from the last PowerVR SDK 3.5
When i create the OpenGL ES context and try to call eglMakeCurrent, then the app is hangs!
I suspect the issue may be that the Windows API is initialised on a separate thread from the GLES initialisation.
If you use only one thread for these, and use additional threads only for loading resources for GLES asynchronously, the application should behave well.
>I suspect the issue may be that the Windows API is initialised on a separate thread from the GLES initialisation.
Do you have plans for fixing of this issue in new version of PowerVR SDK?
>If you use only one thread for these, and use additional threads only for loading resources for GLES asynchronously, >the application should behave well.
No, I can’t load resources because the application is hangs durinng calling eglMakeCurrent…
Adreno/Mali/AMD OpenGL ES 2.0 Emulators are working very good in the other thread! but PoverVR emulator is better!, i hope that it will be fixed in the new version PoverVR SDK.
Unfortunately, the PVRVFrame lead didn’t have a chance to investigate and resolve the issue before our 4.0 SDK went live. We will aim to resolve the problem in a future release.
I had the same problem. All the time when I create the OpenGL ES context and try to call eglMakeCurrent, the app is hangs. It’s really pity. But now I found information what I needed. Thank you.
I tried the application on my system and it runs fine ( with a one caveat ). There was NO OpenGL ES emulator binary provided in the attached archive (zip/rar) file. I replaced all the binaries and static libraries included with those from the latest SDK and ran with no problem( at least thats what I think ). I tried to capture a screen shot just using Alt-Print Screen, but nothing seems to be getting captured. In either case, I saw a window with a large blue triangle.
I’ve discussed this issue with the PVRVFrame lead. It’s not mentioned in our release notes (I’ll see if we can get this updated), but a fix for the eglMakeCurrent() hang made it into our 2016r1 release. With the new libraries, the PVRVFrame lead was unable to reproduce when running your example application on a Radeon HD 7700 Windows machine.
Can you try using the latest PVRVFrame libraries and let us know if the issue still occurs for you?