PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 @ichelash I have tested and HEVC works with a simple configuration and igfxagdc=0 in 13.5 beta 😆 2 Quote Link to comment Share on other sites More sharing options...
ichelash Posted May 30, 2023 Share Posted May 30, 2023 Just now, PoMpIs said: @ichelash I have tested and HEVC works with a simple configuration and igfxagdc=0 in 13.5 beta 😆 Why did you choose Iris 640? 2 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 8 hours ago, ichelash said: Why did you choose Iris 640? It is that in my case I venture 13.5 beta or the VideoProc fails, or the HDMI image goes out when the desktop loads if I use device-id 1259, 1659...... I'm still using my full patch that I've been testing thoroughly and I don't see any failure. So I keep staying like this: 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 30, 2023 Share Posted May 30, 2023 (edited) @PoMpIs I had indicated here that I saw the use of "Pipes=0x12" in another forum. Have you confirmed that setting con0 and con1 pipe = 0x12 is necessary? Also, now that you have disabled con2 (con2-index = 0xffffffff) and set port count to 2, do you still need to set fbmem, stolenmem and unifiedmem? When I was hacking a real Kabylake rig (not spoofed) and I reduced port count to 2, I found that setting stolenmem and unifiedmem actually hurt my graphics performance (graphics performed better without setting fbmem, stolenmem and unifiedmem). My theory was that with the reduced port count, I no longer needed to restrict memory available for the remaining 2 ports. EDIT: What I mean by "hurt my graphics performance" is that with fbmem, stolenmem, and unifiedmem settings, I had graphics problems when I tried to open too many browser windows (don't remember the exact test results). Without setting fbmem, stolenmem and unifiedmem, I did not have the browser window limitation. Edited May 30, 2023 by deeveedee 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 (edited) 1 hour ago, deeveedee said: @PoMpIs I had indicated here that I saw the use of "Pipes=0x12" in another forum. Have you confirmed that setting con0 and con1 pipe = 0x12 is necessary? Also, now that you have disabled con2 (con2-index = 0xffffffff) and set port count to 2, do you still need to set fbmem, stolenmem and unifiedmem? When I was hacking a real Kabylake rig (not spoofed) and I reduced port count to 2, I found that setting stolenmem and unifiedmem actually hurt my graphics performance (graphics performed better without setting fbmem, stolenmem and unifiedmem). My theory was that with the reduced port count, I no longer needed to restrict memory available for the remaining 2 ports. EDIT: What I mean by "hurt my graphics performance" is that with fbmem, stolenmem, and unifiedmem settings, I had graphics problems when I tried to open too many browser windows (don't remember the exact test results). Without setting fbmem, stolenmem and unifiedmem, I did not have the browser window limitation. My board has only two outputs, that's why I did it like this... Stolenmem and unifiedmem, well, I haven't noticed anything strange, and I've opened many browser tabs and everything is fine, but I'm going to try removing those options, which will probably work fine too... It's just that when I finally saw that it was working 100% I already left it like that 😆.. I'm going to try without those options and see what happens 😁.. I'll tell you in a moment how the test went... UPDATE: So far it still works fine without those options 👌 And the performance is the same: https://browser.geekbench.com/v6/compute/499390 Edited May 30, 2023 by PoMpIs 2 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 30, 2023 Share Posted May 30, 2023 (edited) @PoMpIs I totally agree with you - if it works well, leave it alone. I was just curious. Thanks for being patient and testing. I'd also be curious to know if you still need con0-pipe=0x12 and con1-pipe=0x12, but definitely not important and not urgent - just curious. Great job on your hack. EDIT: One other note: hda-gfx=onboard-1 should be inserted automatically by AppleALC.kext, so if you use that kext, you shouldn't need hda-gfx in your DeviceProperties (but it doesn't hurt anything). This document is good reading (it is where I learned that hda-gfx is handled by AppleALC.kext). Edited May 30, 2023 by deeveedee 1 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 31 minutes ago, deeveedee said: @PoMpIs I totally agree with you - if it works well, leave it alone. I was just curious. Thanks for being patient and testing. I'd also be curious to know if you still need con0-pipe=0x12 and con1-pipe=0x12, but definitely not important and not urgent - just curious. Great job on your hack. EDIT: One other note: hda-gfx=onboard-1 should be inserted automatically by AppleALC.kext, so if you use that kext, you shouldn't need hda-gfx in your DeviceProperties (but it doesn't hurt anything). This document is good reading (it is where I learned that hda-gfx is handled by AppleALC.kext). Thank you so much... I am already testing, it has been working for a while and everything is still perfect 👌 2 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 30, 2023 Share Posted May 30, 2023 (edited) @PoMpIs Nice! When you have eliminated patches for a connector X, you no longer need framebuffer-conX-enable. For example, now that you don't have any patches for con0, you can eliminate framebuffer-con0-enable. framebuffer-conX-enable enables WhateverGreen patching of that connector and is not needed if you are not patching the connector. EDIT: One last thing and then I'll shut up - I found that I couldn't be certain that my framebuffer patch changes were implemented until I performed an NVRAM Reset after making the changes. I don't know why that was the case, but there were times that I would make DeviceProperties changes and I thought they were working, only to find that after an NVRAM Reset, my framebuffer patches were no longer working. My lesson learned was that, in order to be certain that framebuffer patch changes were working, I needed to Reset NVRAM after making the changes (before booting macOS with the framebuffer changes). Again, nice job on your hack. Edited May 30, 2023 by deeveedee 1 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 41 minutes ago, deeveedee said: @PoMpIs Nice! When you have eliminated patches for a connector X, you no longer need framebuffer-conX-enable. For example, now that you don't have any patches for con0, you can eliminate framebuffer-con0-enable. framebuffer-conX-enable enables WhateverGreen patching of that connector and is not needed if you are not patching the connector. EDIT: One last thing and then I'll shut up - I found that I couldn't be certain that my framebuffer patch changes were implemented until I performed an NVRAM Reset after making the changes. I don't know why that was the case, but there were times that I would make DeviceProperties changes and I thought they were working, only to find that after an NVRAM Reset, my framebuffer patches were no longer working. My lesson learned was that, in order to be certain that framebuffer patch changes were working, I needed to Reset NVRAM after making the changes (before booting macOS with the framebuffer changes). Again, nice job on your hack. I have a problem!!! 😆 it broke lol if i remove framebuffer-conX-enable it gives a black screen when loading the desktop I'll fix it tomorrow and I'm going to sleep Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 30, 2023 Share Posted May 30, 2023 @PoMpIs You should be removing only framebuffer-con0-enable. You still need framebuffer-con1-enable and framebuffer-con2-enable. If your screen goes black when you remove only framebuffer-con0-enable, I'll have to think about why that would be. 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 30, 2023 Share Posted May 30, 2023 (edited) 15 minutes ago, deeveedee said: @PoMpIs You should be removing only framebuffer-con0-enable. You still need framebuffer-con1-enable and framebuffer-con2-enable. If your screen goes black when you remove only framebuffer-con0-enable, I'll have to think about why that would be. It's exactly what I did, just delete con-0 the rest is the same, and never reaches the desktop 😄 Edited May 30, 2023 by PoMpIs 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 30, 2023 Share Posted May 30, 2023 (edited) @PoMpIs Very interesting. Does anyone know what framebuffer-conX-enable does when there are no conX patches? Is there a type patch that WEG does automatically? @PoMpIs Just to make sure I'm not missing anything, could you post your latest config.plist with framebuffer-con0-enable deleted (the one that doesn't work)? Thanks! EDIT: @PoMpIs Have you mapped your logical ports con0 and con1 to physical ports? If so, what is the physical connector type associated with con0 and the physical connector type associated with con1. Are you using any video adapters (e.g., DP->HDMI)? Edited May 30, 2023 by deeveedee Quote Link to comment Share on other sites More sharing options...
ichelash Posted May 31, 2023 Share Posted May 31, 2023 14 hours ago, PoMpIs said: It is that in my case I venture 13.5 beta or the VideoProc fails, or the HDMI image goes out when the desktop loads if I use device-id 1259, 1659...... I'm still using my full patch that I've been testing thoroughly and I don't see any failure. So I keep staying like this: Mine still trying to make DVI-D work..if anyone can i will appreciate @deeveedee 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 31, 2023 Share Posted May 31, 2023 (edited) On 5/25/2023 at 7:12 PM, Rocky12 said: @Rocky12 I was reviewing your framebuffer patches and I am starting to think that my understanding of frame buffer index numbers is incorrect. The framebuffer that you have chosen (0x59120000) is a 3-connector frame buffer: Framebuffer 0x59120000 Spoiler ID: 59120000, STOLEN: 38 MB, FBMEM: 0 bytes, VRAM: 1536 MB, Flags: 0x0000110B TOTAL STOLEN: 39 MB, TOTAL CURSOR: 1 MB (1572864 bytes), MAX STOLEN: 115 MB, MAX OVERALL: 116 MB (122171392 bytes) Model name: Intel HD Graphics KBL CRB Camellia: CamelliaDisabled (0), Freq: 1388 Hz, FreqMax: 1388 Hz Mobile: 0, PipeCount: 3, PortCount: 3, FBMemoryCount: 3 [1] busId: 0x05, pipe: 9, type: 0x00000400, flags: 0x00000187 - ConnectorDP [2] busId: 0x04, pipe: 10, type: 0x00000400, flags: 0x00000187 - ConnectorDP [3] busId: 0x06, pipe: 10, type: 0x00000400, flags: 0x00000187 - ConnectorDP 01050900 00040000 87010000 02040A00 00040000 87010000 03060A00 00040000 87010000 Your DeviceProperties patch 4 logical connectors: con1, con2 and con3 are enabled and you "disable" con0 by setting con0-index = 0xff. Have you tested all three displays (the HDMI and both DP ports) to confirm that your logical connectors are numbered con1, con2 and con3? Do all three of your Displays (HDMI, DP, DP) work? EDIT: @Rocky12 The questions I have about your patching are as follows: Why did you choose to patch 4 connectors con0, con1, con2, con3 for a 3-connector framebuffer? If you did not use the all-data method and you simplified your patches as follows (removing your existing patches for con0, con1, con2 and con3, remove port count and remove pipe count), would this have worked? framebuffer-con0-enable <01000000> framebuffer-con0-type <00800000> What I'm trying to understand is if the logical connectors in the framebuffer 0x59120000 default to con0, con1 and con2 or do they default to con1, con2, con3? Edited May 31, 2023 by deeveedee 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 31, 2023 Share Posted May 31, 2023 (edited) 13 hours ago, ichelash said: Mine still trying to make DVI-D work..if anyone can i will appreciate @deeveedee I discovered with my HackMini8,1 that starting with Monterey, DVI-D requires port type HDMI. My hack is CoffeeLake. I suspect that HDMI port type is also required to enable DVI-D with Skylake/Kabylake. EDIT: @ichelash Is there still a problem that you are trying to solve? When you say you are trying to get DVI-D to work, do you have a physical DVI-D port on your PC or are you using an adapter (DP->DVI or HDMI->DVI)? Edited May 31, 2023 by deeveedee 2 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 31, 2023 Share Posted May 31, 2023 6 minutes ago, deeveedee said: I discovered with my HackMini8,1 that starting with Monterey, DVI-D requires port type HDMI. My hack is CoffeeLake. I suspect that HDMI port type is also required to enable DVI-D with Skylake/Kabylake. Hello @ ... @deeveedee ☺️ Yes, that's correct, for DVI to work I put it as HDMI output.. Now going back to where we left off yesterday... I had gotten confused with the port mapping, the one I should disable is 0, not 2... (although it worked perfectly as I had it) I did that and everything works fine, but I can't remove the conX but I get a nice black screen when I should get to the desktop.. If I delete everything and leave only the Device-id and platform-id it works too, but I get the 3 active ports in iroeg, nothing happens, but the motherboard doesn't really have them.. I leave you the config.plist with which it works perfectly... if I remove something from the GPU config I no longer have video... only if I remove them all con-x do I have video 🤷♂️ (and the 3 active outputs) config.plist 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 31, 2023 Share Posted May 31, 2023 (edited) @deeveedee More tests... This is the maximum that can be taken in the configuration: If I remove more patches, it gives a black screen And another option that also works: But the system says that I have 3 outputs.. P.S: Remember, this is on Ventura 13.5 beta... which already doesn't seem to work the same way as 13.4 and below on the HD530 Edited May 31, 2023 by PoMpIs 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 31, 2023 Share Posted May 31, 2023 (edited) EDIT: @PoMpIs We were typing at the same time You just confirmed what I asked below. It looks to me as though your framebuffer patching is working as I would expect it to. Nice work and thanks for your patience. ============================================== @PoMpIs Good to know that you need to disable con0. This means that your physical video ports map to logical connectors con1 and con2: Framebuffer 0x59260000 Spoiler ID: 59260000, STOLEN: 38 MB, FBMEM: 0 bytes, VRAM: 1536 MB, Flags: 0x00030B8B TOTAL STOLEN: 39 MB, TOTAL CURSOR: 1 MB (1572864 bytes), MAX STOLEN: 115 MB, MAX OVERALL: 116 MB (122171392 bytes) Model name: Intel HD Graphics KBL CRB Camellia: CamelliaDisabled (0), Freq: 1388 Hz, FreqMax: 1388 Hz Mobile: 0, PipeCount: 3, PortCount: 3, FBMemoryCount: 3 [0] busId: 0x00, pipe: 8, type: 0x00000002, flags: 0x00000098 - ConnectorLVDS [1] busId: 0x05, pipe: 9, type: 0x00000400, flags: 0x00000187 - ConnectorDP [2] busId: 0x04, pipe: 10, type: 0x00000400, flags: 0x00000187 - ConnectorDP 00000800 02000000 98000000 01050900 00040000 87010000 02040A00 00040000 87010000 You need framebuffer-con0-enable = 1 to enable framebuffer-con0-index = 0xffffffff Can you still delete framebuffer-conX-pipe and hda-gfx from your DeviceProperties from your latest config.plist without breaking graphics? Edited May 31, 2023 by deeveedee 2 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 31, 2023 Share Posted May 31, 2023 (edited) @deeveedee We respond very quickly lol 😆 If I can remove with X-pipe and hda-gfx I leave you the last config 👌 config.plist Edited May 31, 2023 by PoMpIs 1 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted May 31, 2023 Share Posted May 31, 2023 (edited) 13 minutes ago, PoMpIs said: P.S: Remember, this is on Ventura 13.5 beta... which already doesn't seem to work the same way as 13.4 and below on the HD530 Based on what you have learned, are you able to summarize the graphics patching differences between 13.5 and 13.4? Also, I assume you can remove hda-gfx from DeviceProperties since you are using AppleALC.kext. Edited May 31, 2023 by deeveedee 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 31, 2023 Share Posted May 31, 2023 (edited) 1 hour ago, deeveedee said: Based on what you have learned, are you able to summarize the graphics patching differences between 13.5 and 13.4? Also, I assume you can remove hda-gfx from DeviceProperties since you are using AppleALC.kext. Yes yes I can, I forgot to remove it, but it works just fine without hda-gfx 👌 P.S: The settings I used in Ventura 13.2.1 lost HEVC in VideoProc when upgrading to 13.5 beta 🤷♂️ Edited May 31, 2023 by PoMpIs 1 Quote Link to comment Share on other sites More sharing options...
PoMpIs Posted May 31, 2023 Share Posted May 31, 2023 @deeveedee @ichelash You are not going to believe it, but I have been trying to install 13.5 beta 2 for two hours and it has never passed this: It occurred to me that the problem could come from using the Device-id of the iris 640 (2659) .... And that was it... I changed the device-id to 1659 and it has been updated at the moment... and the HD530 is still working 👌 3 Quote Link to comment Share on other sites More sharing options...
ichelash Posted June 1, 2023 Share Posted June 1, 2023 7 hours ago, PoMpIs said: @deeveedee @ichelash You are not going to believe it, but I have been trying to install 13.5 beta 2 for two hours and it has never passed this: It occurred to me that the problem could come from using the Device-id of the iris 640 (2659) .... And that was it... I changed the device-id to 1659 and it has been updated at the moment... and the HD530 is still working 👌 That's why i asked before, why do you prefer Iris 640...😁 Glad that the issue is now fixed. 3 Quote Link to comment Share on other sites More sharing options...
deeveedee Posted June 1, 2023 Share Posted June 1, 2023 @ichelash did you get DVI-D working on your hack? Quote Link to comment Share on other sites More sharing options...
ichelash Posted June 1, 2023 Share Posted June 1, 2023 (edited) 2 hours ago, deeveedee said: @ichelash did you get DVI-D working on your hack? Not yet....even i tried to update to OC 0.9.2, everything died 🤣 ..Now even USB 3.0 ports are dead..can you take a look at this..and tell me what i am doing wrong..Kindly! EFI.zip Edited June 1, 2023 by ichelash Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.