
I’m currently running macOS 10.12.5 + gfxCardStatus v2.4.3i It’s usually used to load up older games that get confused by the Mac’s dynamic graphics switching, but it worked well for this case too. GitHub project page: GitHub - steveschow/gfxCardStatus: This is a fork of gfxCardStatus, this fork will handle integrated-only mode a little better Note: The link below is to the steveschow project fork, as it contains patches for post-2010 MacBooks. Hello, I’m relatively new to Unreal and am hoping for a solution to the constant crashes I’m currently experiencing. I upgraded to 4.17.0 today and have not been able to load a project since. I’m able to choose basic project settings such as console, quality, and starter content but after that the project editor loads to 96% then crashes. I’ve attempted two clean installs including deleting all of the files stored in ‘Users/Shared/Epic Games’ and ‘Library/Application Support/Epic’.
#Gfxcardstatus sierra pro#
MacBook Pro running Sierra 10.12 with XCode 8.3.3. My exact error message is: SEGV_MAPERR at 0x3įGenericPlatformMisc::RaiseException(unsigned int) Address = 0x1069d9c9b (filename not found) įOutputDeviceMacError::Serialize(wchar_t const*, ELogVerbosity::Type, FName const&) Address = 0x106bf125f (filename not found) įOutputDevice::Logf(wchar_t const*. Still, switches to the discrete GPU does occur.) Address = 0x106c2a81e (filename not found) įMetalRenderPipelineDesc::CreatePipelineStateForBoundShaderState(FMetalBoundShaderState*, FMetalHashedVertexDescriptor const&) const Address = 0x13c53651a (filename not found) įMetalBoundShaderState::PrepareToDraw(FMetalHashedVertexDescriptor const&, FMetalRenderPipelineDesc const&) Address = 0x13c501c92 (filename not found) įMetalContext::PrepareToDraw(unsigned int, EMetalIndexType) Address = 0x13c4fba28 (filename not found) įMetalRHICommandContext::RHIDrawIndexedPrimitive(FRHIIndexBuffer*, unsigned int, int, unsigned int, unsigned int, unsigned int, unsigned int, unsigned int) Address = 0x13c4ddb72 (filename not found) ĭrawRectangle(FRHICommandList&, float, float, float, float, float, float, float, float, FIntPoint, FIntPoint, FShader*, EDrawRectangleFlags, unsigned int) Address = 0x1183d6b48 (filename not found) įDeferredShadingSceneRenderer::RenderLight(FRHICommandList&, FLightSceneInfo const*, IPooledRenderTarget*, bool, bool) Address = 0x1180ef04f (filename not found) įDeferredShadingSceneRenderer::RenderLights(FRHICommandListImmediate&) Address = 0x1180e95b6 (filename not found) įDeferredShadingSceneRenderer::Render(FRHICommandListImmediate&) Address = 0x117f97a11 (filename not found) įRendererModule::BeginRenderingViewFamily(FCanvas*, FSceneViewFamily*)::EURCMacro_FDrawSceneCommand::DoTask(ENamedThreads::Type, TRefCountPtr const&) Address = 0x1186223d0 (filename not found) ) Address = 0x106cd794d (filename not found) įDebug::AssertFailed(char const*, char const*, int, wchar_t const*. I applied the fix from GitHub that apparently limits use of highest power graphics state. Still, switches to the discrete GPU does occur.Ī wrench in the mix? : I have the MacBookPro that is prone to kernel panics when graphics switch occurs. And I don't really know what the bug is.Ī wrench in the mix? : I have the MacBookPro that is prone to kernel panics when graphics switch occurs. I wouldn't know where to begin to file a bug. Isn't it desirable to have dynamic switching? Is Chrome 80 designed to switch dynamically? My Chrome 80 doesn't. So, Firefox doesn't behave the same as Chrome - at least in my MacbookPro setup. This is the behavior I had seen before 74. In Firefox 74, when I set fault-low-power to False, Firefox switches between the discrete and the integrated GPU dynamically. "Use hardware acceleration when available" has always been enabled.
With Firefox 74, switch to discrete GPU never happens.

Switch back to integrated GPU never happens. In Chrome, when I turn on "Use hardware acceleration when available", graphics switches immediately to the discrete GPU.
