iCanaro Posted November 30, 2020 Share Posted November 30, 2020 with hacks z97 and devirtualisemmio enabled oc 062 and clover 5127 have the same behavior kernel panic at boot 3 Link to comment Share on other sites More sharing options...
Jief_Machak Posted November 30, 2020 Share Posted November 30, 2020 (edited) 2 hours ago, iCanaro said: on the X570 I don't have the same behavior So that's 2 cases where behavior is different, yours and mine. I guess I'll open a bug on github, but since nobody is annoyed with that (because disabled is OK), we'll keep the resolution for later. On interesting thing. If a put this : <key>MmioWhitelist</key> <array> <dict> <key>Address</key> <integer>4278190080</integer> <key>Comment</key> <string>Generic: PCI root is a 0x1000 page memory region used by some firmwares</string> <key>Enabled</key> <true/> </dict> </array> it boots fine with devirtualisemmio enabled, Clover as well as OC. I'm confused with this bug... Edited November 30, 2020 by Jief_Machak 1 2 Link to comment Share on other sites More sharing options...
JorgeMax Posted November 30, 2020 Share Posted November 30, 2020 About the chipset: Big Sur: Spoiler Catalina: Spoiler In the normal Catalina the chipset is recognized, but in Big Sur it shows "Generic" Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 19 minutes ago, Jief_Machak said: it boots fine with devirtualisemmio enabled, Clover as well as OC. I'm confused with this bug... now I verify, address 4278190080 is among the 12 of my list on the X570 that I got with DEBUG OC and converted values for the MMIO white list Link to comment Share on other sites More sharing options...
Guest Posted November 30, 2020 Share Posted November 30, 2020 17 minutes ago, Jief_Machak said: So that's 2 cases where behavior is different, yours and mine. I guess I'll open a bug on github, but since nobody is annoyed with that (because disabled is OK), we'll keep the resolution for later. On interesting thing. If a put this : <key>MmioWhitelist</key> <array> <dict> <key>Address</key> <integer>4278190080</integer> <key>Comment</key> <string>Generic: PCI root is a 0x1000 page memory region used by some firmwares</string> <key>Enabled</key> <true/> </dict> </array> it boots fine with devirtualisemmio enabled, Clover as well as OC. I'm confused with this bug... Have you verified via your debug log if you have that location in your bios... It seems many of you are testing this stuff without understanding well how they work and if they are useful or not for your rig to use it But, said this..first thing is to verify if that MMIO location is present or not in your firmware Link to comment Share on other sites More sharing options...
Jief_Machak Posted November 30, 2020 Share Posted November 30, 2020 2 minutes ago, fabiosun said: first thing is to verify if that MMIO location is present or not in your firmware I'm looking at behavior difference between OC and Clover. Difference that should exist. I'm NOT looking at mmio addresses accuracy. Even if this address does NOT exist, if it generates a behavior difference, it should be looked at. 1 Link to comment Share on other sites More sharing options...
Slice Posted November 30, 2020 Share Posted November 30, 2020 13 minutes ago, JorgeMax said: About the chipset: Big Sur: Hide contents Spoiler Catalina: Hide contents Spoiler In the normal Catalina the chipset is recognized, but in Big Sur it shows "Generic" Same for me. I don't know why and propose it is BigSur beta bug. 1 Link to comment Share on other sites More sharing options...
Guest Posted November 30, 2020 Share Posted November 30, 2020 (edited) 6 minutes ago, Jief_Machak said: I'm looking at behavior difference between OC and Clover. Difference that should exist. I'm NOT looking at mmio addresses accuracy. Even if this address does NOT exist, if it generates a behavior difference, it should be looked at. if it does not exist and if your boot loader works well as OC does..that MMIO in MMIO debug list must be to skip to 0 otherwise..you are devirtualizing some areas randomly...as many of you are doing if I read well here Accuracy is a word I would not associate at this kind of testing... Edited November 30, 2020 by Guest really instead read Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 37 minutes ago, Jief_Machak said: So that's 2 cases where behavior is different, yours and mine. I guess I'll open a bug on github, but since nobody is annoyed with that (because disabled is OK), we'll keep the resolution for later. On interesting thing. If a put this : <key>MmioWhitelist</key> <array> <dict> <key>Address</key> <integer>4278190080</integer> <key>Comment</key> <string>Generic: PCI root is a 0x1000 page memory region used by some firmwares</string> <key>Enabled</key> <true/> </dict> </array> it boots fine with devirtualisemmio enabled, Clover as well as OC. I'm confused with this bug... with this MMIO enabled and all others disabled, on the X570 I get: Clover --> Kernel Panic OC --> black screen 2 Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 @Jief_Machak I ran new test on X570 with DevirtualiseMmio enabled but without any entry in the white list, that is, all active MMIO, Clover and OC have the same behavior, I arrive at desktop and there does not seem to be any anomalies 2 Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 I recommend to those who want to play with MMIO, first understand which ones it has through OC debug, so as to have the complete list, then convert the HEX value to decimal and add the entries in the MmioWhiteList. Then with a lot of patience and time to check which ones you need to have active and which do not. Remember that if you change the BIOS the list of MMIO may change 2 1 Link to comment Share on other sites More sharing options...
Matgen84 Posted November 30, 2020 Share Posted November 30, 2020 17 minutes ago, iCanaro said: @Jief_Machak I ran new test on X570 with DevirtualiseMmio enabled but without any entry in the white list, that is, all active MMIO, Clover and OC have the same behavior, I arrive at desktop and there does not seem to be any anomalies @Jief_Machak I do the same test on my Z390: DevirtualiseMmio enabled, without any entry in the MMIO white list. Clover and OC works fine, I reach the Big Sur Desktop successfully. Thanks for your effort. 1 Link to comment Share on other sites More sharing options...
naiclub Posted November 30, 2020 Share Posted November 30, 2020 26 minutes ago, Jief_Machak said: Spoiler Clover this floor And collect commits (Powered by Clover revision: 5126 (master, commit d042fa25a)) Definitely the best for me Because it can boot from 10.8-11 Which other versions do not work at all It works like 10.8- 10.10- 10.11- 10.12- 10.13- 10.14- 10.15 11 not 10.9. I wonder if the reason that works all of them is because of you. Thank you very much Spoiler Now I have another story I want to update everyone to know with this. And there is no need to add kext .. For SATA as well It works with all versions 2 Link to comment Share on other sites More sharing options...
Jief_Machak Posted November 30, 2020 Share Posted November 30, 2020 36 minutes ago, fabiosun said: So we have to apply MMIO whitelisting..without knowing as it seems you are NOT knowing their meaning..doing damage in our system..applying it without any check as in the past it happened with kernel patches? Maybe is my English..I have said above how to..and for sure not all of people testing MMIO need to use it (overall DevirtualizeMMIO quirk on I mean) then Please do not interfere really? in a civil way and in a public forum? funny The test I asked is devirtualisemmio without any mmiowhitelist. Then I made that remark that I could get it boot by adding an address in the whitelist. Please have the elegance to see that your comment doesn't apply here, as I wasn't trying to make it work by trying random addresses. I know I'm overreacting when people seems to say "you are all wrong" instead of proposing their help. I have my result : there is sometime a behavior difference so there may be a bug. Let's say that I misread your post. I apologize. Let's move on, shall we ? Thanks for your understanding. 3 Link to comment Share on other sites More sharing options...
Guest Posted November 30, 2020 Share Posted November 30, 2020 40 minutes ago, Jief_Machak said: The test I asked is devirtualisemmio without any mmiowhitelist. Then I made that remark that I could get it boot by adding an address in the whitelist. Please have the elegance to see that your comment doesn't apply here, as I wasn't trying to make it work by trying random addresses. I know I'm overreacting when people seems to say "you are all wrong" instead of proposing their help. I have my result : there is sometime a behavior difference so there may be a bug. Let's say that I misread your post. I apologize. Let's move on, shall we ? Thanks for your understanding. Jieef, I appreciate your work even though I don't always say how I often read here it works I do not appreciate the methods and the type of beta testing you have imposed here and I personally preferred to participate by opening issue on github I have had no answer by the way! Returning to the topic and then if we do not understand each other I will not talk about it anymore When you activate DEVIRTUALIZEMMIO do you understand what happens? Have you also tried to check in the debug, both of clover and of OpenCore from which you were inspired if the values that the debug prints are equal and set to 0? Do you know that I personally fought a lot with this quirk and above all with its meaning on platforms .. so to speak complicated? Link to comment Share on other sites More sharing options...
maclinuxG4 Posted November 30, 2020 Share Posted November 30, 2020 26 minutes ago, fabiosun said: Jieef, I appreciate your work even though I don't always say how I often read here it works I do not appreciate the methods and the type of beta testing you have imposed here and I personally preferred to participate by opening issue on github I have had no answer by the way! we need to investigate if it is a present trouble or not. so first, i agree with @jief to check. but if it is a bug, an issue ould be raise, of course. i aggre with fabiosun, plz integration pahse need a good asssement, and tests as possible 1 Link to comment Share on other sites More sharing options...
naiclub Posted November 30, 2020 Share Posted November 30, 2020 Just now, fabiosun said: Jieef, I appreciate your work even though I don't always say how I often read here it works I do not appreciate the methods and the type of beta testing you have imposed here and I personally preferred to participate by opening issue on github I have had no answer by the way! Returning to the topic and then if we do not understand each other I will not talk about it anymore When you activate DEVIRTUALIZEMMIO do you understand what happens? Have you also tried to check in the debug, both of clover and of OpenCore from which you were inspired if the values that the debug prints are equal and set to 0? Do you know that I personally fought a lot with this quirk and above all with its meaning on platforms .. so to speak complicated? I think talking with each other is understandable. Will bring good things to everyone It's not good that you and Jieef are angry. He would want everyone to have the best use. I hope there will be no resentment towards each other. Link to comment Share on other sites More sharing options...
naiclub Posted November 30, 2020 Share Posted November 30, 2020 Just now, fabiosun said: I personally have no resentment for anyone here I'm just asking, reading some things .. do we know what MMIOs are? Or the DevirtualizeMMIO quirk do we know what it does if it's activated? Why do so many here activate it even though they don't need it? Do you know what happens when it is active? and why for some you need to activate it see the trx40 platforms? And why 'then, the aforementioned platforms are afraid to reassign MMIOs to OSX to make them non-virtualized? If you answer this you can try to understand where the problem lies ... Otherwise..we say it works sometimes and we are all happy As far as I understand it or maybe it is wrong. MMIO It is a virtual memory emulator for use with a discrete graphics card. Is it only available with newer motherboards? I'm not sure this one, but it definitely works well with discrete graphics. For new models 1 Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 50 minutes ago, fabiosun said: I personally have no resentment for anyone here I'm just asking, reading some things .. do we know what MMIOs are? Or the DevirtualizeMMIO quirk do we know what it does if it's activated? Why do so many here activate it even though they don't need it? Do you know what happens when it is active? and why for some you need to activate it see the trx40 platforms? And why 'then, the aforementioned platforms are afraid to reassign MMIOs to OSX to make them non-virtualized? If you answer this you can try to understand where the problem lies ... all true, those who have TRX40 had to deal with it, in fact it is necessary to identify them and activate part of these MMIO otherwise no party with hacks. All other platforms can nicely rip it off and keep DevirtualiseMMIO off. In case you need to locate them and work on them, it's not like I've seen a lot of guides around, so the average hack user, he has no idea what they are Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 2 hours ago, iCanaro said: I recommend to those who want to play with MMIO, first understand which ones it has through OC debug, so as to have the complete list, then convert the HEX value to decimal and add the entries in the MmioWhiteList. Then with a lot of patience and time to check which ones you need to have active and which do not. Remember that if you change the BIOS the list of MMIO may change DevirtualiseMmio must be enabled, no matter if the boot then goes into kernel panic or freezes, important to print the MMIO list Link to comment Share on other sites More sharing options...
naiclub Posted November 30, 2020 Share Posted November 30, 2020 IO/MMIO means Input/Output or Memory - Mapped Input/Output 1 Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 18 minutes ago, fabiosun said: and above all, because here they are also used in systems that do not need them to function I had prepared the answer, but I deleted everything after screenshots, I am not interested in continuing discussions that do not interest me and feed flame. I only remember a request for a check if activating DevirtualiseMmio then Clover and OC had the same behavior, end of story 2 1 Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 3 hours ago, naiclub said: As far as I understand it or maybe it is wrong. MMIO It is a virtual memory emulator for use with a discrete graphics card. Is it only available with newer motherboards? I'm not sure this one, but it definitely works well with discrete graphics. For new models let's say my Z68 has 6 nice MMIO 1 Link to comment Share on other sites More sharing options...
Jief_Machak Posted November 30, 2020 Share Posted November 30, 2020 5 hours ago, fabiosun said: I do not appreciate the methods and the type of beta testing you have imposed here I don't understand that. I don't know what I imposed. I don't know even how I could impose anything on anyone. I never trade my help or my work. I never promise a Clover improvement in exchange of something. 2 hours ago, iCanaro said: a request for a check if activating DevirtualiseMmio then Clover and OC had the same behavior @fabiosun Even if it's screwing up the boot of macOs, it doesn't matter as long as it's screwing up the same way in OC and Clover. For that test, we don't need to know what's it's doing or how to use it. I told that few times. I don't understand why this conversation went to "you all should not use it if you don't know exactly how that works". That is not the point here. 4 hours ago, fabiosun said: Otherwise..we say it works sometimes and we are all happy Make it work is not the point here. 3 hours ago, iCanaro said: I am not interested in continuing discussions iCanaro, you're right. I imagine fabiosun will answer to that because he's not able to drop it, although what he said is off topic. Fabiosun, you can have the last word on this, I won't answer anymore. He accused me to impose methods and beta-testing. iCanaro and others, if you feel I'm imposing something, let me know. Link to comment Share on other sites More sharing options...
iCanaro Posted November 30, 2020 Share Posted November 30, 2020 27 minutes ago, Jief_Machak said: He accused me to impose methods and beta-testing. iCanaro and others, if you feel I'm imposing something, let me know. I have no complaints, the day I have it you can rest assured that I will let you know. @Jief_Machak After entering all the MMIO of the Z68 into the config, I located the one that caused the boot to be blocked and kernel panic, once disabled, both OC and Clover start and operate regularly without apparent problems. Z68_2020-11-30_19-30-57_CLOVERX64_5127_dc4a2617f.efi.log So I found the behavior of Clover and OC on the macOS started, big sur, in normal use practically identical 2 Link to comment Share on other sites More sharing options...
Recommended Posts