Asus X99 Deluxe funky Clover UEFI behavior (F8 key on boot) 5960X

[fullerfun]

After further review, I'm convinced the test2.efi method is the best way to go no matter what system you are using.

If it doesn't work for some reason, put your boot drive in sata port 1, boot another system, mount the efi on sata1 drive. Delete the efi folder. do a clean install of latest clover..drop in your appropriate files.

(see below for full history....)

5/23/16

 

----

Edit:

 

So, I've discovered, after much forum hunting and seeing other user's experiences, my original issue turns out to be an issue with having additional PCI cards installed.

I'm sure my MOTU audio pci & thunderbolt PCI are the causes....

 

By adding the the test2.efi file from the below link to clover, I can boot from any menu!

Credit to Nick Woodhams.

https://nickwoodhams.com/x99-hackintosh-solving-osxaptiofixdrv-allocaterelocblock-can-not-allocate-relocation-block/

 

This is an X99 firmware problem known by the developers, and they have yet to make a good solution.

Talked about extensively here:

https://sourceforge.net/p/cloverefiboot/tickets/125/

 

I recently updated to the latest clover v2.3k_r3526, dropped in test2.efi  and boom, I can boot any way I like without errors or problems!

This works perfectly for me on 10.10.5

 

Edit..

Turns out on my system, it totally depends on which sata port your boot drive containing clover is. test2.efi seems to work more consistently, even on 10.11.5 when I'm booting with a drive on one of the 1st sata ports.  When I have a drive on port 5, the h4x's method worked(below)...so weird...

 

If running 10.11.4 or 10.11.5 on an X99 board with additional PCI cards, and your boot drive is not in one of the 1st sata ports, and you get the errors, try need a different efi file

h4x has created one that works for me 

scroll down to the comments section on this page

https://nickwoodhams.com/x99-hackintosh-solving-osxaptiofixdrv-allocaterelocblock-can-not-allocate-relocation-block/

This is his comment and instructions.

"h4x  Good news! I merged OsxAptioFix2Drv with OsxLowMemFixDrv and it works with 10.11.4. 

Compiled binary (compiled with Clover r3354, tested with r3354 and r3469):
https://drive.google.com/file/d/0B2-g_coTRnPnaU53Z1F6Qm5OeXM/view?usp=sharing

Put it into your driverUEFI64 folder and delete "OsxAptioFixDrv.efi", "OsxAptioFIx2Drv.efi", "OsxLowMemFixDrv.efi" and "test2.efi" if presented.

I can also share the patch file after some cleanup if you need it. Basically I moved DoFixes() from LowMemFix to AptioFix2 and call it before RunImageWithOverrides(). This binary will free low memory up to 0x20000000 (512M) before run actual AptioFix2."

 

Awesome fix, credit to h4x & Nick Woodhams! You guys rock!

 

Bill

edited 5/18/16

-------------

 

Hello all.

As always, thank you to everyone in this community. I’ve learned a ton from too many to mention by name!

 

I've finally gotten to the bottom of some funky Clover UEFI booting behavior on my system.

When booting(cold or from restart), If I press-hold F8 to get the BIOS boot device selection menu, and select any UEFI Drive, clover boots normally to clover menu.

But no matter which system I select with clover and try to boot, I get the dreaded “AllocateReloceBlock(): can not allocate relocation block (0X16405 pages below 0x10000000)”  error that many people have been seeing on X99 systems.

 

However, if I boot into the BIOS (press & hold F2 on boot), go to the boot section, make the very first boot device option the UEFI drive that has the drive I want to run clover from, save & exit (F10),( it saves even though it says no changes).  Wait for the computer to re-boot. press nothing, clover boots, viola, I can boot any system from within clover I want without any errors.

 

Also, in the BIOS boot section, if I try the boot override options on UEFI Drives, it does not work either. Same error as above.

 

 

I have Asus X99 Deluxe (BIOS 1203 (also had the issue with BIOS 1103)),   5960X, running latest Clover build Clover_v2k_r3128 (although I experienced the same issues with older versions of clover also).  I suspect the issue is specific to my ASUS X99 Deluxe board..but I suppose it could be that in combination with the 5960X specifically.(maybe some of you with the same or similar boards and/or other processors can confirm this bug)

This issue happens on my system no matter which OS I'm trying to boot 10.9.5 or 10.10.1...both on different SSD's hooked up to SATA ports 1 & 2. Clover boot loader installed on both SSD’s. No matter which drive I boot to clover menu, same error.   This error also appears if I try booting via a USB jump drive using UEFI with the same methods above. So for any of you attempting to use a USB device with clover to install or boot a system..it won’t work unless it’s the 1 item in your boot list in the BIOS. (pressing F8 or boot override will give you the same errors)

 

 

 

Also on another note, for my system, getting 10.10 to boot properly after installing OS 10.10 for the 1st time, there was some more Clover funkiness.

The first time I tried to boot clover it just refused to boot and would stop at “uuid   ++++++++++”

Which normally you get that error when kernel is not properly patched, but mine was properly patched.   I weirdly discovered if I simply go into the bios(F2), go to the boot menu, select my 10.9.5 system with clover as the 1st boot device(instead of the clover I had just installed on my 10.10 drive), reboot the computer, select my 10.10 system in clover menu and it booted right up. very strange. ever since doing that once, now both my clover boot loaders work normally without any errors, as long as I do as above(let the system boot to the 1st UEFI item in the boot list item in BIOS)

 

 

FYI, One final note, for my specific system using OsxAptioFixDrv-64.efi  was the only one that works.  OsxAptioFix2Drv-64.efi does not work on my system.

 

Hope this helps save someone hours of time…took me forever to narrow this down…

 

Bill

Edited May 23, 2016 by fullerfun

[AllenRoss]

Very interesting i have seen some funky behavior as well with Clover... Many say slices new builds are unstable... I see you have the same board and CPU as me, please try my guide, as it is a Vanilla method

 

Dont know why aptiofix 1 is working but 2 isn't I can boot with Aptiofix 2 for both mavericks and Yosemite.

[AsusFreak]

None of these are "Funky Clover" issues, they're user issues. Set the correct boot device in the bios. If other UEFI drives don't boot from bios boot menu or bios over-ride then the EFI partitions for those drives aren't properly configured. OsxAptioFix-64 works for 95% of the systems out there, OsxAptioFix2-64 was only added for the few it doesn't work for.

[jamiethemorris]

None of these are "Funky Clover" issues, they're user issues. Set the correct boot device in the bios. If other UEFI drives don't boot from bios boot menu or bios over-ride then the EFI partitions for those drives aren't properly configured. OsxAptioFix-64 works for 95% of the systems out there, OsxAptioFix2-64 was only added for the few it doesn't work for.

Well the system should still boot properly by choosing the drive from the F8 menu, so I wouldn't say it's necessarily user error, although it's not a big deal to just set it as the first boot device in the BIOS. I would say it's more of a system-specific BIOS issue, so it's not really Clover's fault or the user's fault.

[fullerfun]

Yes totally agree. Sorry if my use of funky implied the wrong message. I am not intending to place blame for any of this behavior. Simply stating how my system is responding...and trying to help others if they experience the same. Definitely something about the boot loader menu and boot bypass in the bios of this board that is very different. And therefore, Clover is responding differently. Luckily if you know how the boards responds, you can work around it, and it's not an issue at all.

 

I can guarantee my EFI partitions for my drives are properly configured. I've been using clover on may hackintosh's and configured and setup many drives, never had an issue till this board came along.

[fullerfun]

Very interesting i have seen some funky behavior as well with Clover... Many say slices new builds are unstable... I see you have the same board and CPU as me, please try my guide, as it is a Vanilla method

 

 

Dont know why aptiofix 1 is working but 2 isn't I can boot with Aptiofix 2 for both mavericks and Yosemite.

 

Allen, that's totally weird. Which bios, which version of clover, what ram do you have, and how much ram do you have?

Did you compile clover, or just use the installer?

 

I have 32GB(4 x8GB) crucial ddr4 2133 (PC4-17000), I just used the installer for clover.

 

There's got to be be some variable that makes your system different form mine....making yours happy with Aptiofix 2.

 

Can you boot any system via clover using the Press&Hold F8 boot loader menu?

[jamiethemorris]

So are you the only one that's having this problem with the X99-Deluxe? That seems to be the most common X99 board around here so I'm surprised I haven't seen any other reports of it happening

 

You guys should also compare BIOS settings, maybe it's something to do with that. Fast boot perhaps?.

[fullerfun]

Post #270  @onmybikedrunk mentions the same issue. He has same exact setup as me.
http://www.insanelymac.com/forum/topic/301585-haswell-e-mach-kernel-patch-5960x-5830k-and-5820k-yosemite/page-14
 
Then user @jackfoley says he has success after in next post.
 
so I think it's safe to say both those users have the same issue I do.

I just went through and tried every possible change of setting in the BIOS boot menu settings..fast boot, on off, csm, on/off, and everything in between. UEFI on/off vs legacy etc...

 

No matter how I set BIOS boot menu settings, same issue. F8 & boot override do not work..same error message...

I suspect there might be some more people having this problem, but don't realize what's causing it. It took me several days of messing around to realize this is what was causing it....  I also think some poeple are just using a legacy bootloader so they don't have this issue either.

[AllenRoss]

So are you the only one that's having this problem with the X99-Deluxe? That seems to be the most common X99 board around here so I'm surprised I haven't seen any other reports of it happening

 

You guys should also compare BIOS settings, maybe it's something to do with that. Fast boot perhaps?.

Jamie Why are you so cool?

[jamiethemorris]

Jamie Why are you so cool?

I live in a refridgerator

[AllenRoss]

Me too im a cucumber =D

[fullerfun]

Hey guys, please stay on topic, or don't post here...much appreciated.
----------

Today I noticed some more weirdness on my computer.

If I booted cold, and select my 10.10.1 system in clover, it refused to boot.
But, if I booted 10.9.5 first, then restarted and then booted 10.10.1, it booted right up.

this was repeatable and happened everytime until I realized....

Adding the NeverHibernate option

<key>NeverHibernate</key>
<true/>

to the boot section of my clover config.plist, solved this problem! 10.10.1 boots right up every time for me now!

 

Thought this might help someone else with this issue....

 

 

On another note, playing around with clover 3132 's ability to patch the kernal on the fly. So far working great for me on 10.9.5 & 10.10.1!  Comes in really handy when using clover on a system installer USB!

Oh and BTW, my BIOS still no happy with Clover 3132 either...same issue as in above..post #1

[AllenRoss]

Sorry No one was trying to get off topic. Just pointing out Remotely that respect is important.

 

Do you have CSM off. I have seen funny things with which type of option it will boot from with CSM on. Sometimes it quickly boots to your legacy in which case you don't have one. Take a close look under Boot Options and set everything to UEFI only if you Don't have the GOP to run CSM off.

[fullerfun]

Yes I've tried every possible setting for CSM, makes no difference.

[PM!]

Same problem here Asus X99 Deluxe (BIOS Version 1305) and Clover r3168.

My graphic card is a GTX 970 4Gb (Gainward)

I boot with slide=0 dart=0 npci=0x2000 nvda_drv=1 kext-dev-mode=1

 

I get the AptioFixDrv error as well since I use Nvidia Web drivers.

 

On my other workstation Asus Z87X Deluxe Dual I had the same problem and it got solved by changing AptioFixDrv to AptioFixDrv2 but on the X99 it doesn't work I get an other similar error.

 

CMS is enabled in BIOS i'll try to disable it but i'm pretty shure it won't change anything.

 

I manage to boot back to the OS by deleting kextcaches from another computer (some do so from recovery partition) but the problem comes back every time I install a driver and sometimes without changing anything...

 

I think the AptioFixDrv2 should be adapted to X99 motherboards but who am I to complain and give orders :-)

[fullerfun]

Yep, I just updated my BIOS to 1305, and clover to r3185.  same error still. can't use BIOS bootloader menu(F8) or boot override in BIOS.

I had to RMA my motherboard, the new one has the same issue.

 

 

On another note, on my 10.9.5 system for this same computer, I made a huge discovery that using this USB kext from REHABMAN https://bitbucket.org/RehabMan/os-x-generic-usb3/downloads  suddenly allowed The CPU package mulitplier to change while running 10.9.5!  Before this, It only worked in 10.10 on my system.  So now, finally,  have some nice power saving running both my 10.9 & 10.10 systems.

[AsusFreak]

Yep, I just updated my BIOS to 1305, and clover to r3185.  same error still. can't use BIOS bootloader menu(F8) or boot override in BIOS.

I had to RMA my motherboard, the new one has the same issue.

 

 

On another note, on my 10.9.5 system for this same computer, I made a huge discovery that using this USB kext from REHABMAN https://bitbucket.org/RehabMan/os-x-generic-usb3/downloads  suddenly allowed The CPU package mulitplier to change while running 10.9.5!  Before this, It only worked in 10.10 on my system.  So now, finally,  have some nice power saving running both my 10.9 & 10.10 systems.

I've seen this posted over and over again that this usb kext enables speed step, how ridiculous is this... I ignored it until yesterday, installed the kext on my X99 box and improve GB scores by 3500, how can this be? It would be great to find a dev that could analyze and explain... 

 

Makes you wonder what non-related kext we need to make sleep work. lol

[jamiethemorris]

I've seen this posted over and over again that this usb kext enables speed step, how ridiculous is this... I ignored it until yesterday, installed the kext on my X99 box and improve GB scores by 3500, how can this be? It would be great to find a dev that could analyze and explain... 

 

Makes you wonder what non-related kext we need to make sleep work. lol

I'm not a dev, but from what I've read it's because of xhci interrupts. If you run sudo powermetrics -s interrupts you should see a ridiculous number of them. That's why VMs run so slow.Obviously it also affects general performance as well. However, I have no idea why exactly what causes it. Perhaps someone could disassemble the vanilla XHCI kext and compare it to the GUXHCI source and see if there's a way to patch the vanilla one. I'm curious whether this would still happen if you disable XHCI in the BIOS.

[AsusFreak]

I'm not a dev, but from what I've read it's because of xhci interrupts. If you run sudo powermetrics -s interrupts you should see a ridiculous number of them. That's why VMs run so slow.Obviously it also affects general performance as well. However, I have no idea why exactly what causes it. Perhaps someone could disassemble the vanilla XHCI kext and compare it to the GUXHCI source and see if there's a way to patch the vanilla one. I'm curious whether this would still happen if you disable XHCI in the BIOS.

Well this is interesting. I disabled XHCI in bios and had the same results as adding the USB kext (with out the negative side affects), 3500 gain in GB (26310), 75 gain in Cinebench CPU (1271)...

[jamiethemorris]

Well this is interesting. I disabled XHCI in bios and had the same results as adding the USB kext (with out the negative side affects), 3500 gain in GB (26310), 75 gain in Cinebench CPU (1271)...

Well that's something - so the problem is XHCI itself (driver specifically, obviously) and NOT the ports. Here's what I'm curious about - if a natively supported PCIe 3.0 card would have the same interrupts.

 

EDIT: I'm curious - do you have any XHCI related errors during boot or in the Console?

[AsusFreak]

Well that's something - so the problem is XHCI itself (driver specifically, obviously) and NOT the ports. Here's what I'm curious about - if a natively supported PCIe 3.0 card would have the same interrupts.

 

EDIT: I'm curious - do you have any XHCI related errors during boot or in the Console?

@jamiethemorris and @andrew: No XHCI errors and here's some interesting stuff, disable XHCI in bios and load NullCPUPowerManagement.kext and Benchies increase 15% and speedstep now works... This is soooo weird. 

EDIT: More weird, sleep is working...

 

 

[NOTNlCE]

Absolutely incredible. I've done as described with the same results on my X99-UD4 and 5820K. What a world of a difference.

[amgclk65]

@jamiethemorris and @andrew: No XHCI errors and here's some interesting stuff, disable XHCI in bios and load NullCPUPowerManagement.kext and Benchies increase 15% and speedstep now works... This is soooo weird. 

EDIT: More weird, sleep is working...

If I may ask, what version of HWsensors are you using ? I tried the latest build I beliveve. But its only showing 1 cpu core temp and 1 fan rpm out of 3 I have. Any help would be awesoem.

[AsusFreak]

http://sourceforge.net/projects/hwsensors/

[fullerfun]

So, I've discovered, after much forum hunting and seeing other user's experiences, my original issue turns out to be an issue with having additional PCI cards installed.

I'm sure my MOTU audio pci & thunderbolt PCI are the causes....

 

By adding the the test2.efi file from the below link to clover, I can boot from any menu!

Credit to Nick Woodhams.

https://nickwoodhams...location-block/

 

This is a x99 firmware problem known by the clover developers, and they have yet to make a good solution.

Talked about extensively here:

https://sourceforge....ot/tickets/125/

 

I recently updated to the latest clover v2.3k_r3526, dropped in test2.efi  and boom, I can boot any way I like without errors or problems!

This works perfectly for me on 10.10.5

 

Edit..

Turns out on my system, it totally depends on which sata port your boot drive containing clover is. test2.efi seems to work more consistently, even on 10.11.5 when I'm booting with a drive on one of the 1st sata ports.  When I have a drive on sata port 5, the h4x's method worked(below)...so weird...

 

If running 10.11.4 or 10.11.5 on an X99 board with additional PCI cards, and your boot drive is not in one of the 1st sata ports, and you get the errors, try need a different efi file

h4x has created one that works for me 

scroll down to the comments section on this page

https://nickwoodhams...location-block/

This is his comment and instructions.

"h4x  Good news! I merged OsxAptioFix2Drv with OsxLowMemFixDrv and it works with 10.11.4. 

Compiled binary (compiled with Clover r3354, tested with r3354 and r3469):
https://drive.google...iew?usp=sharing

Put it into your driverUEFI64 folder and delete "OsxAptioFixDrv.efi", "OsxAptioFIx2Drv.efi", "OsxLowMemFixDrv.efi" and "test2.efi" if presented.

I can also share the patch file after some cleanup if you need it. Basically I moved DoFixes() from LowMemFix to AptioFix2 and call it before RunImageWithOverrides(). This binary will free low memory up to 0x20000000 (512M) before run actual AptioFix2."

 

Awesome fix, credit to h4x & Nick Woodhams! You guys rock!

 

Bill

 

I edited the header of this topic with these comments...hope it helps others....thank you to all...