how to rebuild kernelcache on MBP 4,1 endless loop booting ? (ASD passed)

Hello my fellow mac users, and once again pardon my french syntax.

My core2duo MBP (early 2008) fails to go thru any system boot phase, whatever input is feeded with a system disk, could it be 10.6, 10.9 or 10.10.

This happened shortly after a Yosemite update, and achieving to boot in safe mode after many attempts did not help, both ways end up rebooting the MB for ever.

First of all i reset pram/nvram and booted in SU mode to run fsck -fy, twice, to get rid of the ”filesystem was modified” message.

Still looping...

So i booted in target mode and used my Mac mini to repair permissions. That was greyed out even after in changed the status NOT to ignore that in finder info window of the disk. Still grey even though the status bar said privileges were now respected on the disk.

So i tried to repair the disk with Disk Utility, even though i just did a fsck in Single User mode

It failed claiming i should save files and reformat the disk.

I then used DriveGenius to try a rebuild and repair, which worked, and Disk Utility was then happy with it.

But it still did not go thru the (very slow) boot process and endlessly rebooting.

I suspected a kernel cache corruption, so i went back to target mode, and i surprisingly could boot my Mac Mini to use Onyx for Yosemite to erase it so that it could be rebuilded at the next straight boot. No result.

I tried to boot on an old Snow Leopard system i kept from dismantling my first MacMini (core2duo as well, but from 2009), that freezed after a couple of minutes spins of the grey daisy

I could get a version of the right ASD and formatted a USB key with both boot partitions. It passed every test on the OSD and then on EFI partition.

At that stage i did

-resize the original Yosemite partition to free about 60GB, and create a new hfs+ partition to

-install a fresh Mavericks and...

-boot my mac mini on that partition,

-create a fresh account and got straight away in Accounts pref pane to

-create another one (i always leave the first account clean and make the second one my workspace).

”re-route” the account to use the folder from original ”home” folder on the first partition of the disk, using advanced options (alt-click on account in left pane)

Mmmh... anybody around still following me?

Well, i could not believe my eyes when i rebooted the MBP on that new 10.9 partition and reached the login window and ...logged in!

The MBP was already running quite hot so i dl’d Macs Fan Control but install failed, cursor freezed and i was back to the beginning.

That second partition is also freezing on the gray daisy before cpmplete boot ever since.

Now the MBP is still passing the USB booted EFI battery of tests.

BUT i get kernel panics on every attempt with the OSD one. Too bad !

It was about time to try a verbose boot on both the original and new systems and even though i’m far from being a fluent reader of that kind of stuff i could trace some puzzling blocks of lines:

[code]Warning: could not block sleep during cache update

Warning: proceeding w/o DiskArb

rebuilding //System/Library/(...)/kernelcache

kext file://System/Library/Extensions/JmicronATA.kext/ is in hash exception list (...)

process /uisr/sbin/kextcache[8] exited due to signal 11.

error 107 rebuilding /System/(...)/kernelcache

/dev/disk0s2 on / (hfs, local, journaled)

BootCacheControl: Unable to open /var/db/BootCache.playlist: 2 No such file or directory[/code]

and further down, just before screen blacks out:

[code]WARNING - ACPI_SMC_CtrlLoop:: initCPUCtrlLoop - no match for the sub-config MacBookPro4, 1 with 8 p-states using default stepper instead [/code]

I checked

/System/Library/Extensions/IOPlatformPluginFamily.kext/Contents/PlugIns/ACPI_SMC_PlatformPlugin.kext/Contents/Resources/MacBookPro4_1.plist

and there i could find

[code]

<key>MacBookPro4,1</key>

<array>

<dict>

<key>num-states</key>

<integer>9</integer>

<key>program</key>

<string>SP4</string>

</dict>

</array>

[/code]

I can’t get any clue why the detected ”p” value is 8 and not 9. But again all that is 99% mumbo jumbo to me... don’t be misleader by my code typing, it’s only a replica, i shot pictures of my screen.

So just in case i traced that JMicronATA kext (SSD driver stuff if my search result is right) and stored it out of the kernel Extensions folder of the original system. Useless.

Now i’m down to that double question : can i force the kernel cache to be rebuilt thru firewire from a remote mac, and how to do it ?

Not quite sure that will save my *** though...

End of Chapter 1.

😁