Unlocked latest PXHCD.kext Version 1.0.11

[Henties]

I managed to unlock the latest PXHCD.kext Ver.1.0.11 My USB 3 devices now connect to the USB3 Super-Speed bus
My mobo, GA X58A-UD3R uses the NEC uPD720200 controller the firmware version of that is F3.034.08
This version 11 is a considerable improvement over version 10 however I have not completed testing yet.

In the interim I discovered the following.

1. Don't let the machine go to sleep when a USB3 drive(s) is/are connected, eject the device first and unplug it.
If you don't do this you will require to reboot your machine to get access to the drive(s) again.

2. Connect the USB3 drive(s)only after the machine has started and run up fully.

3. From the desktop hotplugging works. Unplugging must be preceded by an eject and power disconnect.

I am not sure whether I may actually attach/post this unlocked PXHCD.kext of mine. Until further notice I will therefore
hang on to it until I am sure that posting it will not be considered illegal or inappropriate.

Oh by the way I am running ML 10.8.2

Greets Henties.

 

[MacMan]

So this version is better then what is in MultiBeast 5.1.3 which is 1.0.10?

If so, then post here so I can add it to MultiBeast.

 

[Henties]

Here it is, unlocked by me today.

It is quite a significant improvement over anything previous. The latest uPD720200 ver. F3.0.34.08 firmware proves beneficial, and is
recommended.
Over the years I updated my mobo's bios a number of times just to recently discover that a mobo bios update does not include
the firmware update for the onboard NEC USB3 controller. I actually should have known better.

With this PXHCD.kext loaded, sleep remains functioning, whereas with previous versions, sleep got broken terribly.
When I boot my machine I can have a Verbatim 1 terra byte USB3 portable drive connected, without ill effects, however it does not
always show up on the desktop or in diskutil after completion of the boot process, this phenomenon manifests itself in Win 7
also, however not as frequently. It think I know the reason and will investigate. Unplugging and replugging the drive usually
makes it appear and accessible.

A self powered 2 terra byte Verbatim USB3 drive needs to be powered down prior to booting my machine. Should this drive be
powered up and connected, prior to booting, I get the following error "SerialAtapi device configuration did not complete successfully.
(failed command info =0x1) I think the reason for this is that I have too many devices connected and the PXHCD.kext, loading very
early during the boot process, grabs resources which are actually needed by "other hardware" I think I have already traced it down
to the fact that I have installed 2 DVD writers of which one is a BlueRay device. I will investigate further as soon as time permits.

My system definition is for an iMac 12.2

Sleep, the pitfalls.

With anything connected and mounted to the 2 USB3 ports, when the machine goes to sleep, day literally becomes night
It is virtually impossible to wake the machine when this happens, only a power cycle will get you going again.
I ascribe this to a poor or incomplete implementation of the USB3 specification. Win 7 does not suffer from this deficiency.

What is wrong ?

There seem to be no routines in the PXHCD.kext or elsewhere, to prevent sleep to kick in when an external USB3 drive is connected.
Once sleep kicks in, the drive gets powered down by the sleep process, the resources it occupied prior to sleeping however
remain committed, that being the reason why waking up from sleep under these conditions is so problematic. The drive light on
my self powered USB3 drive goes out when sleep kicks in, surely that should not happen at all.

Workaround.

Whilst not elegant, a workaround is to disable sleep for the time that the USB3 drive is needed, to prevent surprises, then when
done with the drive(s, ) unmount/eject them, then LOGOUT and therafter disconnect power to the USB3 drives, LOGIN again and
voila, sleep is working again as soon as one activates it through the Energy Saver in System Preferences.

Speed.

I transferred 101.78 terrabytes from a SATA3 drive to the external USB3 drive in 22 mins 8 secs, which translates to 76.67 mB/sec.
Not fantastic but acceptable, considering that the HDD in the Verbatim enclosure is a real alcheapo Seagate, spinning at only 5900 rpm.

Hoping my initial findings help others in their use of this latest kext by LaCie.

In my humble opinion there seems to me that a lot of work is still required by Apple as well as Industry, LaCie, Caldigit and others,
to make USB3 functionality in OS X comparable with that of Win 7. Progress is certainly hampered by Industry locking their drivers
so that only their peripherals are functioning. Guess we have to live with that for the time being as OS X itself is a very closed system.

In Win 7 any USB3 drive of any manufacturer, provided they adhere to the basic USB3 specification is functioning properly without
the Vendor lockout we are experiencing under OS X.

Greets Henties


LaCie-Ver 1.0.11-Unlocked-PXHCD.kext

 

[Henties]

It appears there was something wrong with sending the attachment.

here is another attempt

 

[nguyenmac]

Are you using vanilla 10.8.2 IOUSB kexts or ML DP3 kexts? Do you notice any difference between them?

 

[Henties]

All my kexts are those that Apple provided for installing Mountain Lion. So far I have had no need to replace any kexts after applying an Apple ML update.

The only non Apple kexts I use are the following:

1. Networking - I use the Lnx2Mac's RealtekRTL81xx Ethernet v0.0.90 kext. It just works great, even my USB-->Ethernet adapter from my WII console
is recognized and fully functional. That one drives my VPN to HMASS

2. Video - I use my own NVDANV50HAl.kext and NVDAResman.kext in order to be able to use a nVidia display card that Apple felt prudent to exclude from
Mountain Lion. My video subsystem on ML 10.8.2 comprises a GTX285 - still supported by ML - as well as a 8400GS not supported by ML.

3. Then I naturally also use my unlocked LaCie Ver 1.0.11 PXHCD.kext for my USB3 devices. All my boot stuff, ie. "Extra" is placed in the EFI partition,
so it is conveniently tugged away and can be forgotten that it exists.

So basically I have to look after only 4 kexts to maintain a system that really runs like a dream.

My philosophy has always been not to buy myself the functionality that I think I need or want , but rather exploit possibilities to enable the functionality
that I require, in the devices I already own.

Many Many moons/decades ago, when my son needed extra storage capacity on his Apple 2E floppy drive, he wrote an assembler routine which increased the
floppy capacity from a mere 180 kbytes to 360 kBytes. Soon thereafter he cut out a notch at the "backside" of the floppies available then, and voila
he ended up with 720 kByte total floppy storage. Man that was a lot in those days. Steve Jobs subsequently amply rewarded him for his endeavors. Bless you
Steve, and thank you for the paths that you have paved for all of us.

Coming back to USB3 and my unlocked LaCie Version 1.0.11 PXHCD.kext, after this trip to memory lane, for which I appologize, I discovered that
in my particular case, disabling "Use Kernel Cache" in the org.chameleon.Boot.plist file, or by using the -f bootup flag, allows me to boot
my machine every time successfully, with the USB3 drives connected.

I found this out by observing that my first cold boot was always successful, and only with subsequent reboots did I get this message
"SerialAtapi device configuration did not complete successfully. (failed command info =0x1) After which one of my DVD drives was inaccessible.

I figured that some resources did not get released between boots, probably in memory, or wherever, resources that the "failed" DVD drive desperately
needed. Applying the -f flag during boot, or setting the Use Kernel Cache flag to No in the org.chameleon.boot.plist file fixed this issue.

To me me this was the last straw, realizing that this whole USB3 thingy in ML is still not really usable, the software foundation on which this
USB3 functionality is built is rather shaky, and appears to be supported by spaghetti code. So I pass for the time being and await the arrival
of 2 Fresco Logic USB3 cards, which it is rumored, integrates somewhat better with ML. At least it appears that those cards will work properly,
as the code to make them work is supposedly embedded by Apple in ML, or so I understand. Where I live, it will take about another 2 months
for those items to arrive. In the meantime I will use my system as is and enjoy what it can do for me without any issues. In the interim Apple
may well have fixed some issues on their side in a 10.8.3 update, which I expect rather soon to be forthcoming.

My system is working extremely well, is stable and exceptionally fast, OCed to 4.2 Gigs. At this stage I do not want to compromise my computing
experience with half backed eggs. So the current incarnation of Apple's USB3 is on the back burner for me now. until such time that it's
integration and functionality approaches that which Microsoft has already achieved a long long time ago.
Actually wondering whether Apple will ever get there with the way things are going at the moment at Apple. Presently their priorities seem to
be elsewhere, somewhere in the "Cloud" I think, and perhaps maybe rightly so, only the future can tell.

Greets Henties

 

[Henties]

Good to hear makes me feel proud that I provided something usefull after I myself have obtained so much
valuable information on this forum helping me make my hacks as successful as they are

 

[nguyenmac]

Thank you for your contribution, but there are still problems in this new kex (and the kext from osx86.net too):
- Unplugging a mouse usb will cause KP instantly (like previous versions).
- After ejecting a flash drive from USB 3.0 ports, plugging it again and the drives will not be mounted (it still shows up in System Information). However, this issue does not happen with USB 2.0 flash drives.
- The new kext produce tons of logs at start up, and while the device is plugged, it continues create logs non-top . This does not happen with <= 1.0.10

So I think it's better that you try the 1.0.10 kext to get rid of the console flooding. Have fun hackintosh!

 

[dawilliams660]

I have a HP DV7 6143CL I installed this kext now when I connect my drive too the 3.0 usb it locks up I have to reboot it to get it to work......how do I get this to work

 

[sean2e]

I managed to unlock the latest PXHCD.kext Ver.1.0.11 My USB 3 devices now connect to the USB3 Super-Speed bus
My mobo, GA X58A-UD3R uses the NEC uPD720200 controller the firmware version of that is F3.034.08
This version 11 is a considerable improvement over version 10 however I have not completed testing yet.

In the interim I discovered the following.

1. Don't let the machine go to sleep when a USB3 drive(s) is/are connected, eject the device first and unplug it.
If you don't do this you will require to reboot your machine to get access to the drive(s) again.

2. Connect the USB3 drive(s)only after the machine has started and run up fully.

3. From the desktop hotplugging works. Unplugging must be preceded by an eject and power disconnect.

I am not sure whether I may actually attach/post this unlocked PXHCD.kext of mine. Until further notice I will therefore
hang on to it until I am sure that posting it will not be considered illegal or inappropriate.

Oh by the way I am running ML 10.8.2

Greets Henties.

I'm running 10.8.5 and seem to be the only person who can't upgrade to 10.9 from a unibeast drive....worked plenty of times before. ive found a cheap 2 port pci card...with this C/set. Think It might work after running MB?