235 replies to this topic

[MacUser2525]

Going Bald, on Nov 10 2010, 03:57 PM, said:

It shows up in profiler as "Unknown AHCI controller"

I have the same controller on an add in PCI-e card as Cartri says it works OOB all I had to do was plug in the card and the drive connected to it boot and start transferring data it shows the up same as yours in my profiler so that is just a cosmetic display issue, OS X won't boot from it now mind you which is a shame as I wanted to put my SSD on it otherwise works fine. Is yours bootable Cartri?

[cartri]

MacUser2525, on Nov 10 2010, 05:21 PM, said:

I have the same controller on an add in PCI-e card as Cartri says it works OOB all I had to do was plug in the card and the drive connected to it boot and start transferring data it shows the up same as yours in my profiler so that is just a cosmetic display issue, OS X won't boot from it now mind you which is a shame as I wanted to put my SSD on it otherwise works fine. Is yours bootable Cartri?

Never tried to boot from it, but it has a bridge and says osx that the ports are eSata so maybe is why you can't boot, if this is your problem, this is our same on the orange icon problem in gigabyte motherboards, and sits on how to write the capabilities we want:  we know even how to generate it in 64 instead of 32bit format, but we did not find where the bar0 capabilities can be forced. Anyway, if the IOAHCIBlockStorage kext gives us a internal status to the disk, some kext like this one could do the same to your ahci: even better because marvell will be matchable in a different kext if you desire. The problem is that - "oh no" - our asus expansion has a bios, and a undefined bridge.
We can create this bridge definition in dsdt, even force them by the early booted dsdt inside the mainboard bios, but that would "eat" us a slot. (you would never be able to have another thing but this board working on the hacked slot)

[MacUser2525]

cartri, on Nov 10 2010, 04:30 PM, said:

Never tried to boot from it, but it has a bridge and says osx that the ports are eSata so maybe is why you can't boot, if this is your problem, this is our same on the orange icon problem in gigabyte motherboards, and sits on how to write the capabilities we want:  we know even how to generate it in 64 instead of 32bit format, but we did not find where the bar0 capabilities can be forced. Anyway, if the IOAHCIBlockStorage kext gives us a internal status to the disk, some kext like this one could do the same to your ahci: even better because marvell will be matchable in a different kext if you desire. The problem is that - "oh no" - our asus expansion has a bios, and a undefined bridge.
We can create this bridge definition in dsdt, even force them by the early booted dsdt inside the mainboard bios, but that would "eat" us a slot. (you would never be able to have another thing but this board working on the hacked slot)

Ok I think I take most of what your saying here in. The gist of it being if I can get it recognized as internal (IORegistryExplorer shows it attached to PEX2@1c,2 on PRT0,0) where in the device tree the PRT0,0 shows AHCI-Built-In True but the actual IOAHCIDevice@0 that attached to the port as Physical Interconnect Location of External. It would be this last part of the output that needs to be changed to Internal right and would a modification/new entry in my DSDT declaring the PEX2 as built-in 0x01 like for a network card do this? Looking further down the device tree everything else shows as Physical Interconnect Location of Internal for the AppleAHCIDiskDriver, IOAHCIBlockStorageDevice and I do get an internal icon as I Have the IOAHCIBlockStorage.kext installed. Although now I look at the rest of my disks connected on the motherboard ports the attached IOAHCIDevice@0 for them shows Physical Interconnect Location of External with the same Internals on the the same entries below them in their device tree as shows for the Marvell controller and they can be booted from. So I think I am lost again on where to start doing this modification don't suppose you would have any sample snippet of code for the DSDT mod(s) needed. Sorry for hijacking the thread with this definitely off topic posting.

[Rx4Mac]

cartri, on Nov 10 2010, 06:09 AM, said:

ditto, can you explain in which port are you using the seagate?
As the bios is emulating the whole macpro 4,1 ich10 sata ahci, id suggest you to use the third port (PRT2) as the 1 1st ports could be reserved for dvds.

I will have to check tomorrow which port they were connected to; but I did use the same port for all SATA tests - this probably eliminates the question of the MB - My bet is on a firmware issue with the drives as they were in the Seagate batch that had all the firmware issues. My guess at this point is that Seagate locked the drives to SATA I when they issued the replacement firmware. This is a guess, but it does look like the drive to me. I will see if I can borrow an SSD to test in the next few weeks too.

[Going Bald]

cartri, on Nov 10 2010, 02:02 PM, said:

And your hard disks dont work in them?
I have a marvell ahci controller on the Asus U3S6 expansion card and it works out of the box

Hard disks work in them, they just don't work at 6Gb/s in OS X, and the driver just says
Unknown AHCI Standard Controller
   Marvell Virtuall
in profiler

Been watching the price on these - http://www.newegg.co...6-557-_-Product    and when it comes down to at least half what it is now, I will get 2 to use the Marvell ports. If SL doesn't have SATA III support by then, I'll probably format them NTFS and use them strictly with Win7.

[cartri]

Going Bald, on Nov 11 2010, 11:37 PM, said:

Hard disks work in them, they just don't work at 6Gb/s in OS X, and the driver just says
Unknown AHCI Standard Controller
   Marvell Virtuall
in profiler

Been watching the price on these - http://www.newegg.co...6-557-_-Product    and when it comes down to at least half what it is now, I will get 2 to use the Marvell ports. If SL doesn't have SATA III support by then, I'll probably format them NTFS and use them strictly with Win7.

AHCI is a universal standard, in theory to give the controller the capabilities and just report as a disk type to the system, what matters sometimes to the system is the AHCI-CAP but there is no exclusivity to sata2 speeds, sata3 ahci should be a faster generic port then a "named" interface - can you test and see if you get better results?

[stellarola]

My MSI X58 Big Bang X-Power has the Marvell® 88SE9128 chipset which allows for RAID 0/ 1 mode OOB in OS X. It's also completely bootable without any sort of trickery. It's badass.

-Stell

[Rx4Mac]

stellarola, on Nov 12 2010, 03:39 PM, said:

My MSI X58 Big Bang X-Power has the Marvell® 88SE9128 chipset which allows for RAID 0/ 1 mode OOB in OS X. It's also completely bootable without any sort of trickery. It's badass.

-Stell

Did you just need to use Chameleon +/- the boot file from PC EFI 10.6 ?

No kexts ?

( this is OT... Should this be discussed elsewhere? )

Thanks Stell !

[stellarola]

2decadeMacUser, on Nov 12 2010, 08:51 PM, said:

Did you just need to use Chameleon +/- the boot file from PC EFI 10.6 ?

No kexts ?

( this is OT... Should this be discussed elsewhere? )

Thanks Stell !

I'll be creating a dedicated MSI X58 Big Bang thread soon. I'll discuss the information there, instead of clutter this one anymore than we already have.

-Stell

[genzai]

New trouble:

11/13/10 8:54:51 PM kernel PXHCD    0xffffff80ffcf1000::UIMAbortEndpoint: slotID is not valid
11/13/10 8:54:51 PM kernel PXHCD    0xffffff80ffcf1000::UIMDeleteEndpoint: slotID is not valid
11/13/10 8:54:53 PM kernel USBMSC Identifier (non-unique): 00000000000000000000 0x174c 0x5106 0x1
11/13/10 8:54:53 PM kernel USBF:        80058.424        Endpoint 0x81 of the USB device "AS2105" at location 0x200000:  converting Bulk MPS from 1024 to 512 (USB 2.0 Spec section 5.8.3)
11/13/10 8:54:53 PM kernel USBF:        80058.424        Endpoint 0x2 of the USB device "AS2105" at location 0x200000:  converting Bulk MPS from 1024 to 512 (USB 2.0 Spec section 5.8.3)
11/13/10 8:55:03 PM kernel PXHCD    0xffffff80ffcf1000::IssueEndpointCommand - Deconfigure Endpoint
11/13/10 8:55:03 PM kernel PXHCD    0xffffff80ffcf1000::XHCIRootHubResetPort detect device on port 2
11/13/10 8:55:03 PM kernel UIMCreateControlEndpoint - EP type 0x2
11/13/10 8:55:03 PM kernel UIMCreateControlEndpoint - EP type 0x2

The device is no longer mounting on GA-X58A-USB3

I have updated to 10.6.5 and wonder if this borked it. Anyone else made the update to 10.6.5 having issues?
Thanks,
g\

EDIT
well rebooted with the dock still attached and it mounted the disk properly on boot so i'm not sure if that was just a fluke. I'll do some more testing soon.
/g

[stellarola]

genzai, on Nov 14 2010, 12:01 AM, said:

New trouble:

11/13/10 8:54:51 PM kernel PXHCD    0xffffff80ffcf1000::UIMAbortEndpoint: slotID is not valid
11/13/10 8:54:51 PM kernel PXHCD    0xffffff80ffcf1000::UIMDeleteEndpoint: slotID is not valid
11/13/10 8:54:53 PM kernel USBMSC Identifier (non-unique): 00000000000000000000 0x174c 0x5106 0x1
11/13/10 8:54:53 PM kernel USBF:        80058.424        Endpoint 0x81 of the USB device "AS2105" at location 0x200000:  converting Bulk MPS from 1024 to 512 (USB 2.0 Spec section 5.8.3)
11/13/10 8:54:53 PM kernel USBF:        80058.424        Endpoint 0x2 of the USB device "AS2105" at location 0x200000:  converting Bulk MPS from 1024 to 512 (USB 2.0 Spec section 5.8.3)
11/13/10 8:55:03 PM kernel PXHCD    0xffffff80ffcf1000::IssueEndpointCommand - Deconfigure Endpoint
11/13/10 8:55:03 PM kernel PXHCD    0xffffff80ffcf1000::XHCIRootHubResetPort detect device on port 2
11/13/10 8:55:03 PM kernel UIMCreateControlEndpoint - EP type 0x2
11/13/10 8:55:03 PM kernel UIMCreateControlEndpoint - EP type 0x2

The device is no longer mounting on GA-X58A-USB3

I have updated to 10.6.5 and wonder if this borked it. Anyone else made the update to 10.6.5 having issues?
Thanks,
g\

EDIT
well rebooted with the dock still attached and it mounted the disk properly on boot so i'm not sure if that was just a fluke. I'll do some more testing soon.
/g

I have these same errors when device is plugged in @ system boot.

-Stell

[bearclaw92060]

does this work on an actual mac pro?

[MacUser2525]

bearclaw92060, on Nov 23 2010, 12:58 AM, said:

does this work on an actual mac pro?

For a MacPro you would get the drivers from Lacie when you purchased one of their USB 3.0 cards so you would use the original drivers unaltered.

[genzai]

MacUser2525, on Nov 22 2010, 10:12 PM, said:

For a MacPro you would get the drivers from Lacie when you purchased one of their USB 3.0 cards so you would use the original drivers unaltered.

huh?

I see no reason why this modbin patched file wouldn't work on a mac pro. Real Mac users can also use any USB3 storage device with this driver, instead of being locked to LaCie devices.

g\

[MacUser2525]

genzai, on Nov 23 2010, 04:46 PM, said:

huh?

I see no reason why this modbin patched file wouldn't work on a mac pro. Real Mac users can also use any USB3 storage device with this driver, instead of being locked to LaCie devices.

g\

Probably would but if wanting support if anything goes wrong from said manufacturers your not getting it using a hacked driver and lets face the facts someone using a MacPro is not going to cheap out saving $10 on a generic card + hacked driver even I am not that cheap and I am one tight fisted SOB...

[delmar78]

My "My Passport" have the about the exact same speeds perhaps a bit more(about 69MB/s).

Works ok until I try to use as a backup drive with Superduper, copies for about half of the root drive then it unmounts!

Why??

[ChiemY]

Hi,

Just for information:
I try the modbin patch on MacBook Pro 15 Late 2008 with:
- Storeva ExpressCard USB 3.0 2 ports (30€ from macway.com)
  (need to plug one usb2.0 on the card for more power)
- Western Digital My Passport Essential 500 Go USB 3.0 / USB 2.0 (from fnac.com)

If I insert or remove the ExpressCard with PXHCD.kext (lacie or modbin), the macbook freeze.
I need to put the ExpressCard before and use:
/sbin/kextload -v /System/Library/Extensions/PXHCD.kext
(check "sudo dmesg  | grep PXHCD" for more log)

Read Write About 30mo/s on usb2 and 60mo/s on USB3.

Regards

Edit: I try the caldigit driver below and changes values in plist, but card was not working or crash the kernel.

Attached Files

•  sc.png   80.06K   108 downloads

[Dreamwatcher]

http://www.caldigit....rive/index.html

CalDigit USB3.0 Driver (Mac 10.5, 10.6 and Win )

include kext
for 10.6 32/64bit
- IOUSBFamily.kext 4.0.2
- CalDigitUSBxHCI.kext 1.1.9

for 10.5 32bit
- IOUSBFamily.kext 3.5.2
- CalDigitUSBxHCI.kext 1.1.9

this driver is not protected by the vendor.

[cartri]

bearclaw92060, on Nov 23 2010, 02:58 AM, said:

does this work on an actual mac pro?

It was made for that - just don't use a card with a bridge or you will need kexts for that bridge (PXL for instance). Use a direct USB3 PCI-E card

[Qianlong]

So you can't use the Asus U3S6 card for example, because it has a PLX bridge?

I thought that Cartri was using this card with the hacked Lacie drivers and the sata6g ports were also working without any new drivers or kexts?

http://www.testfreak...gbps-pcie-card/