When using Nvidia Web Driver with a MSI GeForce GTX 970 GAMING 4G, I randomly get Error allocating 0x800 pages at boot time in verbose mode.
I couldn't fix this issue.
I've tried to replace OsxAptioFix2Drv-64.efi by OsxAptioFixDrv-64.efi without success: I then always get Error - requested memory exceeds our allocated relocation block.
Tested using Clover v2.3k r3766 (packaged with MultiBeast 9.0.1), r3974 and r3998.
I have a fix for the 0x800 boot issue you've been having. Use driver OsxAptioFixDrv-64.efi but to fix the issue with that you need to edit RtVariables in your config.plist and change the value which reads 0x3 to 0x67 - that will fully turn off SIP but fix the memory allocation issue.
Received by email:
My solution is different but I also tried what is mentioned in the post you linked and it did not work on my config (P8Z68-V/GEN3 /w 16GB of RAM).
I had to edit source code of OsxAptioFixDrv and compile it manually with changed just one variable:
line 188 in latest release: Addr = 0x100000000; // max address
changed to: Addr = 0x300000000; // max address
What I've read about the origin of this problem is that latest graphic cards (I have a 970 as well) allocate a lot of memory for VBios and thus Clover is not able to find empty spot up to address 0x100000000 for kernel (you can read a bit about it there: link).
If you have plenty of RAM you can simply adjust the max address that the allocation method will look for empty block.
For me it works 100/100 times. Note that I tripled the initial value. It might be the case that you will need to simply double it. I have 6 HDDs connected and plenty of USB devices and from what I've read each of these devices may allocate some RAM for their firmware during boot-up process.
When using Nvidia Web Driver with a MSI GeForce GTX 970 GAMING 4G, I randomly get
Error allocating 0x800 pages
at boot time in verbose mode. I couldn't fix this issue.I've tried to replace
OsxAptioFix2Drv-64.efi
byOsxAptioFixDrv-64.efi
without success: I then always getError - requested memory exceeds our allocated relocation block
.Tested using Clover v2.3k r3766 (packaged with MultiBeast 9.0.1), r3974 and r3998.
Suggestions to fix this issue: