Nice one. Let's see if anyone have time to further develop this script to provide options for multiple drives etc
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP 1,1-5,1 Activate AMD hardware acceleration
- Thread starter h9826790
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
Nice concept, but this script assumes you only have one HD and one ESP, which is almost never the case especially on the cMP.
I'm still not (yet) using OpenCore, but I do mount my EFI partitions frequently. Here's a shell script to simplify that process. You specify a target (in my case, it's the REFIND.efi program; for OC, specify a required OC file), and it loops through all EFI partitions until it either finds the target (and leaves that partition mounted at /Volumes/EFI) or runs out of EFI partitions to try.
Code:
#!/bin/sh
#
#### Edit the following if you need to:
#
# Set to true for progress messages, false for quiet (errors only)
VERBOSE=false
# Path to where EFI partitions should be mounted
EFI_MOUNT_POINT=/Volumes/EFI
# Path to file telling us we've got the right partition
TARGET_FILE=$EFI_MOUNT_POINT/EFI/refind/REFIND.efi
# Test to perform (-f ::= file, -d ::= directory, -x ::= executable)
TARGET_TEST=-f
# how long to sleep between mount/unmount (seconds, can be fractional)
# (repeatedly mounting and unmounting without a sleep usually results
# in failures.)
SLEEP_TIME=0.50
###
### Don't edit the following unless your system is non-standard!
###
DISKUTIL=/usr/sbin/diskutil
GREP=/usr/bin/grep
AWK=/usr/bin/awk
MOUNT=/sbin/mount
SLEEP=/bin/sleep
###
###### End of items you should need to edit
###
if ($VERBOSE == "true"); then ECHO=echo; else ECHO=true; fi
# Unmount the EFI partition if one is currently mounted
$MOUNT | $GREP $EFI_MOUNT_POINT > /dev/null && ($SLEEP $SLEEP_TIME; $DISKUTIL unmount $EFI_MOUNT_POINT)
# Loop through all EFI partitions until we find TARGET_FILE
for i in `$DISKUTIL list | $GREP EFI | $AWK '{print $NF}'`;
do
$ECHO "Trying $i"
[ -d $EFI_MOUNT_POINT ] || mkdir $EFI_MOUNT_POINT
$MOUNT -t msdos /dev/$i $EFI_MOUNT_POINT
[ $TARGET_TEST $TARGET_FILE ] && break;
$SLEEP $SLEEP_TIME
$DISKUTIL unmount $EFI_MOUNT_POINT > /dev/null
done
if [ $TARGET_TEST $TARGET_FILE ] ;
then
$ECHO "Found target, mounted /dev/"$i" on "$EFI_MOUNT_POINT
exit 0
else
echo $TARGET_FILE" not found on any available EFI partition."
exit 1
fiI am trying to incorporate this EFI mounter in the first script, but it comes with an error:
Then I tried:
But I got the same error. Does somebody know why the command fails through Apple script?
Code:
on run {input, parameters}
do shell script "'/Applications/reconfigure OpenCore/Reconfigure OpenCore/MountEFI-update/MountEFI.command'"
return input
end runThen I tried:
Code:
on run {input, parameters}
do shell script "export TERM=xterm-color;'/Applications/reconfigure OpenCore/Reconfigure OpenCore/MountEFI-update/MountEFI.command'"
return input
end run
Nice work. I will adopt and amend for my Refind -> OpenCore chain loading set up.
The helper apps i wrote for the chain loading follow the same approach to locate the relevant EFI partition. Basically mount each in turn and check whether a target is present.
Thank you! I had to uncheck the "Allow Frame Ordering" and now the HEVC encode is flying.
Actually, I created to a partition. Followed the instructions, but when reboot to windows installer (from OC), I got this error:
INACCESSIBLE BOOT DEVICE
Yes, but I booted it only one time in order to start windows installation. When reboot it for the first time, I turned the vm off, following the instructions.
What if I finish windows installation under the vm? I’ll try this.
Pay attention to this:
cMP5,1 and Apple BOOTCAMP assistant for creating Windows 10 partition
One thing to make clear about raw disk access in VirtualBox: On every reboot the disk ID might change. You can verify that in the Disk Utility. Unfortunately the .vmdk created for raw disk access holds the original disk number. You can inspect the file with Hex Fiend: So when you try to open...
forums.macrumors.com
Thank you! It worked.🙂
I used the second input of the same display. Now I can migrate to Catalina.
The question is why this happens? Why the second display solved this performance issue?
May it be a faulty cable or some display issue? By the way, mine is a LG HDR 4K. Is yours the same?
Yea that is not going to work. The partition only has no access to the EFI of the whole disk, therefore Windows boot loader will not be installed in the ESP partition where Windows and OC will look for it.
I guess it is Apple driver issue, some people using egpu enclosure have the same. They need to connect the graphic card to more than one displays
Yeah, I removed the 0.5.7 package and updated the package to 0.5.8.
Quite a few guys including me keep testing the latest beta. So far, all newer versions (official release) are as stable as the old one, and bring us more functions or bug fix.
I also expect to update the package once 0.5.9 release and switch the default picker to OpenCanopy. It will look like this.
Last edited:
Very nice. I just found the 0.5.8 release you posted, so I deleted my comment, and then realized you already responded to me. Thanks!
Oh, I didn't realise that I forget to update the link that inside the steps. It's now fixed. Thanks for pointing that out to me.
ndkbootpicker is not working with the latest commits and it seems that the developer has abandoned it.
We use NdkBootPicker mainly because the GUI boot picker from OpenCore wasn't working on cMP.
In fact, OpenCore developers don't care if NdkBootPicker work or not. The NdkBootPicker developer have to keep tracking the OpenCore development, and keep updating the NdkBootPicker for newer OpenCore release. And if he hit any difficulty, he must find the work around on his own.
By moving to OpenCanopy, we can guarantee it will be compatible to the latest OpenCore in future.
Now, we are testing latest OpenCore 0.5.9 beta with OpenCanopy. It works quite well so far, at least same performance level as the NdkBootPicker (if not better).
Even though the NdkBootPicker still has some advantage. e.g. Can use our own wallpaper, or make screen capture, etc. However, as I mentioned in post #1314, my aim is to provide a "one package fit all" solution. Since OpenCanopy should provide better reliability and compatibility in future. And this is more important than customisable wallpaper. Therefore, my plan is to switch to OpenCanopy.
TBH, I didn't check if NdkBootPicker still works with the latest OpenCore 0.5.9 or not. Users are free to try, and customise their own OpenCore config / boot picker, but just not in my default package.
In fact, OpenCore developers don't care if NdkBootPicker work or not. The NdkBootPicker developer have to keep tracking the OpenCore development, and keep updating the NdkBootPicker for newer OpenCore release. And if he hit any difficulty, he must find the work around on his own.
By moving to OpenCanopy, we can guarantee it will be compatible to the latest OpenCore in future.
Now, we are testing latest OpenCore 0.5.9 beta with OpenCanopy. It works quite well so far, at least same performance level as the NdkBootPicker (if not better).
Even though the NdkBootPicker still has some advantage. e.g. Can use our own wallpaper, or make screen capture, etc. However, as I mentioned in post #1314, my aim is to provide a "one package fit all" solution. Since OpenCanopy should provide better reliability and compatibility in future. And this is more important than customisable wallpaper. Therefore, my plan is to switch to OpenCanopy.
TBH, I didn't check if NdkBootPicker still works with the latest OpenCore 0.5.9 or not. Users are free to try, and customise their own OpenCore config / boot picker, but just not in my default package.
EDIT: Changed the DP slot my monitor was attached to as per this post, and I have boot screen, HWAccel... etc. Will update this post if I have any issues with Catalina (I'm going to follow this video)
Last edited: