H77/i5-2500K Build to Begin

[PH]

Just as predicted, the ASRock H77M mobo arrived today, at about noon ... a few hours' more than ONE DAY service from NewEgg, which is apparently using a new delivery service within California for its "free" three-day delivery option. The product was actually in my local delivery hub in about 12 hours' time, from its departure from NewEgg. Incredible.

A long time ago, before Micro Center went belly-up, I got a really good deal from them on an i5-2500K proc. I was going to update my 7-slot H67 machine, but I became distracted with the DC7800 and DC7900 projects, which turned out the be a LOT of fun, but also took a LOT of time. The i5-2500K is "unlocked" for overclocking (which I no longer use as a matter of personal policy), and is four cores, alright, but only four threads, and is 3.2 GHz with a relatively HUGE 6 MB cache.

Now, this is my first i3 or i5 proc which has on-chip HD3000 video (all others had HD2000 video, which is useless for a Hack), and I will indeed be trying HD3000 with the ASRock H77M mobo. I wouldn't be trying this particular proc in a P77 mobo as those have no support for the proc's graphics processor. Should the HD3000 video not work to my complete satisfaction, I will be using a tried-and-true 8400GS in the ASRock's PCI-e 16x 2.0/3.0 slot (2.0 for Ivy Bridge, such as the i5-2500K, 3.0 for Sandy Bridge, such as the current Intel CPUs).

I won't need the ASRock's CCO (compatible cooler option) as the i5-2500K comes with an i-series cooler. "Push pins", of course, which I have come to respect, although I far prefer screw mounting, as on most after-market coolers, and on BTX coolers, such as on the DC7800 and DC7900 series coolers.

ASRock H77M mobo link:

http://www.newegg.com/Product/Product.aspx?Item=N82E16813157303

Intel i5-2500K proc link:

http://www.newegg.com/Product/Product.aspx?Item=N82E16819115072

NewEgg is still selling the i5-2500K for nearly $220. I paid something like $180 for the same at Micro Center, while its Santa Clara (CA) store was still in operation. I guess they were clearing-out their stock of Ivy Bridge procs for the forthcoming Sandy Bridge procs, but bankruptcy and reorganization came first. Sad, as they were previously the only alternative to Fry's Electronics, at least in my area. I guess Santa Clara, CA, was "A Bridge too Far" for Micro Center, which has apparently elected to focus on its Kansas City (KS)/Overand Park(MO) locations. I guess eventually all folks' efforts MUST come to an end, perhaps even Randy and John Fry, who are presently so wealthy that they can own and operate their own Boeing 767 jet transport (used for transportation for their individually owned league teams), and a world-class, but completely private, 18-hole golf course within San Jose (America's tenth largest city), which is now open to the public only for the so-called "Fry's PGA Tour Open".

I intend to start out with 2 x 2 GB = 4 GB of fast Corsair DDR3 RAM, but as this mobo supports 16GB, I may eventually move on to 2 x 8 GB = 16 GB, if needed. Most probably also Corsair.

I really like the fact that this mobo supports USB 3.0 OOTB as I have become tired of fiddling and farting around with various mobo manufacturers' or add-on manufacturers' USB 3.0 "solutions".

Not that I am all that committed to USB 3.0, but it is becoming an interesting option. One MUST keep up with the times.

[Mark Leebove]

Peter,

I rarely question your information but I must regarding Micro Center. The following excerpt from the Wikipedia gives the reason for closure. 

Santa Clara Store Closure

On the 23rd of July, 2012, Micro Center's Santa Clara location was announced to be closed due to "unsuccessful attempts to negotiate an economically viable extension of [the] store lease."^[3] The location is planned to remain open for the purpose of servicing existing customers through September 21, 2012.

[edit]

Thanks for all the great info and support you have provide for novices and experts on this list.

Mark Leebove

Lurking member.

[pete...@cruzio.com]

> I rarely question your information but I must regarding Micro Center. The
> following excerpt from the Wikipedia gives the reason for closure.

I DID state that their KS/MO store(s) were remaining open, and presumably
others as well.

Micro Center's INITIAL e-mail to its Santa Clara customers certainly
IMPLIED that they were having great difficulties, without specifying a
reason. When a reason is not specified, one usually assumes the worst.

Should they open another store in the Santa Clara area, I would not
hesitate to purchase from them.

The specific location they had was just far enough away from several Fry's
Electronics locations, and just close enough to a major freeway off-ramp,
to make the drive to them worthwhile. Plus, I could use Tasman to access
them from certain North San Jose locations, from which access to Fry's
Electronics was significantly more difficult.

That, plus there was a second route OUT of their store which gave easy
access to Fry's Electronics' Sunnyvale store OR to Weird Stuff Warehouse.

It usually comes down to Location, Location, Location, and for computer
geeks in the Santa Clara/Sunnyvale area, Micro Center's former location
was just about ideal.

[pete...@cruzio.com]

Some progress has been accomplished.

The i5-2500K proc has been installed and the Intel cooler has been mounted.

Also, a 20-pin header to dual USB 3.0 cable and connector set has been
ordered from HK ...

http://www.ebay.com/itm/Internal-USB-Motherboard-Connection-2-Port-USB-3-0-Female-to-20-Pin-Baffle-Cable-/170881907758

... only US$4.79 shipped, and usable on either the rear OR, perhaps more
particularly, the front.

The H77M mobo includes two USB 3.0 connectors on the rear, so the obvious
choice is to incorporate the two additional USB 3.0 ports on the front by
modifying the case's floppy drive insert, which would otherwise go
completely unused.

As with all 7-series mobos, there is NO chipset support for legacy PCI
slots, but the ASRock H77M mobo includes an ASMedia ASM1083 chip, which is
a PCI-e to PCI "bridge" chip, with a maximum of three legacy PCI slots
(two 33/66 MHz slots are actually provided in this particular mobo; there
is an ASMedia ASM1085 chip which actually provides for FIVE legacy PCI
33/66 MHz slots).

The provided Intel "push pin" proc cooler was actually pretty easy to
install ... I got it right the FIRST time. Incredible, as I usually have
one Hell of a time with Intel's "push pins".

The next step is to remove the ASUS P67 mobo and to install the ASRock
H77M mobo in its place.

THEN, to grab the ASROCK H77M'S DSDT for eventual OS X modification.

[PH]

Some significant progress has been made.

I setup a primitive test stand using the ASRock's own packaging as an electrical insulator, connected a handy PSU, temporarily installed 2 GB of DDR3 RAM, and connected a spare SVGA/USB cable from my KVM switch.

Also, I connected a USB DVD drive with Ubuntu 9.04 Desktop, a flash drive which had bootable DOS and the DOS version of the ASRock BIOS update (the board was shipped at the 1.20 level, but 1.30, for improved processor support, is already out), and then I flashed the BIOS to 1.30.

Then, I booted Ubuntu 9.04 and grabbed the raw DSDT and wrote it to a USB flash drive for examination on another system using DSDTSE.

Finally, I unhooked the kludge and proceeded to examine the H77M micro-ATX's DSDT.

The usual errors, of course, but surprisingly few.

Mute 0x0fff, of course, also _T_0 and _T_1, plus a couple of "method must return a value" errors.

But that was about all.

Within a half hour I had those errors corrected plus all the USB devices properly renamed. Also the SBUS device amended.

Still more work to be done.

Someone on tonymacx86's site has, he says, completed an OS X-compatible DSDT for the mini-ITX version of this mobo.

I'll be looking at his for inspiration.

Because this is a 4-slot mobo and all my cases are my preferred 7-slot variety, I'll probably be purchasing a new 4-slot case just for this mobo.

Oh, yes, I DID use the i5-2500K's HD3000 video for accessing this board. Seemed to work OK after Ubuntu finally figured out what it was working with.

[paaguti]

I had some (bad) experience with Ubuntu distros prior to Precise (12.04) when using Intel HD3000 graphics. Have you considered updating? Of course, getting the DSDT is a bit different than in 09.04 but you have all tools you need to do that. I actually built a boot stick with a larger  (4G) USB stick, updated and installed the necessary dpkg's on it. Then I grabbed the Linux version of the DSDT Patcher. This is my main DSDT debugging environment.

Just my 0.001 cents,

/PA

[pete...@cruzio.com]

I ALWAYS value other's suggestions.

My SUPER-kludge installation, intended solely for grabbing the AMI BIOS's
DSDT, was completely successful. It was NOT intended to do any more than
that.

I soon was to discover that the ASRock H77M Micro-ATX (my unit) had a
sligtly different DSDT than the previously reviewed ASRock H77M Mini-ITX,
although they are otherwise identical as to major devices.

The handling of the audio in to Mini-ITX proved to be compatible with the
Micro-ATX version.

We'll eventually see how that works out to be.

[faithie999]

what commands do you issue in Ubuntu 12.04 to write the DSDT to the desktop?

thanks

[PH]

Well, that was NOT nice ... the ASRock H77M (micro ATX) has a locked MSR, or whatever it is which is locked.

I had updated the BIOS from 1.20 to 1.30, and ONLY THEN did I discover that this is a problem with MANY boards, other than Gigabytes.

But, all MAY NOT be lost.

A hacker who goes by the handle samisnake modded the 1.20 BIOS, so I may try downgrading my board from 1.30 (unmodified) to 1.20 (modified).

Supposedly, this eliminates, FOREVER, the KP issue as the KP is actually occurring when the kernel tries to update the MSR.

The dual internal USB 3.0 cable still has not arrived from HK. Doesn't really matter at this point as the ASRock H77M is bricked until the MSR is unlocked.

A word to the wise: apparently Intel-branded H77 boards CANNOT be unlocked, no matter what one does.

Another word to the wise: stick with Gigabyte!

[mosslack]

On Nov 17, 2012, at 2:26 PM, PH wrote:

Well, that was NOT nice ... the ASRock H77M (micro ATX) has a locked MSR, or whatever it is which is locked.

Confusion, does MSR refer to model-specific registers?

I had updated the BIOS from 1.20 to 1.30, and ONLY THEN did I discover that this is a problem with MANY boards, other than Gigabytes.

But, all MAY NOT be lost.

A hacker who goes by the handle samisnake modded the 1.20 BIOS, so I may try downgrading my board from 1.30 (unmodified) to 1.20 (modified).

Supposedly, this eliminates, FOREVER, the KP issue as the KP is actually occurring when the kernel tries to update the MSR.

I don't understand well enough to see how the BIOS has anything to do with this unless it unlocks the registers?

The dual internal USB 3.0 cable still has not arrived from HK. Doesn't really matter at this point as the ASRock H77M is bricked until the MSR is unlocked.

A word to the wise: apparently Intel-branded H77 boards CANNOT be unlocked, no matter what one does.

Another word to the wise: stick with Gigabyte!

Good to know.

From the main system of mosslack...

______________________________

Alt-OS <+> GG <+> TBIE <+> Hack List

[pete...@cruzio.com]

>> Supposedly, this eliminates, FOREVER, the KP issue as the KP is actually
>> occurring when the kernel tries to update the MSR.
>
> I don't understand well enough to see how the BIOS has anything to do with
> this unless it unlocks the registers?

After reading NUMEROUS web sites, this is what I BELIEVE I know ...

The OS X kernel accesses the MSR (machine status register) for WRITE, but
the BIOS has already written to another register a bit which INHIBITS any
subsequent access to WRITE to that register, perhaps most particularly the
WRITE from the OS X kernel.

The purpose of the modded BIOS is to allow ANYONE, but perhaps most
particularly the power manager, to update the MSR.

I HAVE the modded 1.20 BIOS, and I have provided samisnake with the URL of
the unmodded 1.30 BIOS, so ... perhaps ... I will eventually gain some
traction on my ASRock H77M (mico-ATX) project.

Alas, most folks seem to be concentrating on the ASRock H77M (mini-ITX)
project.

[mosslack]

On Nov 17, 2012, at 3:33 PM, pete...@cruzio.com wrote:

Supposedly, this eliminates, FOREVER, the KP issue as the KP is actually

occurring when the kernel tries to update the MSR.

I don't understand well enough to see how the BIOS has anything to do with

this unless it unlocks the registers?

After reading NUMEROUS web sites, this is what I BELIEVE I know ...

The OS X kernel accesses the MSR (machine status register) for WRITE, but
the BIOS has already written to another register a bit which INHIBITS any
subsequent access to WRITE to that register, perhaps most particularly the
WRITE from the OS X kernel.

This sounds like what I read about also, but my confusion came from the machine status register referring to IBM Power PC CPUs. The similar function for Intel is model-specific register. As long as we are on the same page then I suppose it doesn't matter. I was just attempting to clear up any confusion for others trying to follow this.

The purpose of the modded BIOS is to allow ANYONE, but perhaps most
particularly the power manager, to update the MSR.

Does this locked MSR also apply to UEFI boards?

I HAVE the modded 1.20 BIOS, and I have provided samisnake with the URL of
the unmodded 1.30 BIOS, so ... perhaps ... I will eventually gain some
traction on my ASRock H77M (mico-ATX) project.

Alas, most folks seem to be concentrating on the ASRock H77M (mini-ITX)
project.

Is it the easier/better choice for a Hackintosh?

[pete...@cruzio.com]

>> The OS X kernel accesses the MSR (machine status register) for WRITE,
>> but
>> the BIOS has already written to another register a bit which INHIBITS
>> any
>> subsequent access to WRITE to that register, perhaps most particularly
>> the
>> WRITE from the OS X kernel.
>
> This sounds like what I read about also, but my confusion came from the
> machine status register referring to IBM Power PC CPUs. The similar
> function for Intel is model-specific register. As long as we are on the
> same page then I suppose it doesn't matter. I was just attempting to clear
> up any confusion for others trying to follow this.

Yes, there appears to be a similarity in the names of certain control
registers, which are usually kernel-level ONLY, and are not intended to be
accessed by, say, an application-level program.

>> The purpose of the modded BIOS is to allow ANYONE, but perhaps most
>> particularly the power manager, to update the MSR.
>
> Does this locked MSR also apply to UEFI boards?

This issue does, indeed, have to do with UEFI boards. And, apparently, to
UEFI boards, ONLY.

>> I HAVE the modded 1.20 BIOS, and I have provided samisnake with the URL
>> of
>> the unmodded 1.30 BIOS, so ... perhaps ... I will eventually gain some
>> traction on my ASRock H77M (mico-ATX) project.
>>
>> Alas, most folks seem to be concentrating on the ASRock H77M (mini-ITX)
>> project.
>
> Is it the easier/better choice for a Hackintosh?

I suspect that much effort has gone into the new format boards (Mini-ITX)
and that less effort has gone into the old format boards (Micro-ATX).

An obvious advantage of the Micro-ATX board is four traditional slots,
including, at least, a PCI-e 16x slot and a PCI-e 1x slot, and, perhaps
two legacy PCI slots.

There ARE alternatives in this same Micro-ATX format which offer a PCI-e
16x slot and THREE legacy PCI slots.

The Micro-ITX boards tend to have only a single PCI-e 16x slot.

[phubai]

Wow....good-luck with the samisnake hack...and I look forward to your progress

[pete...@cruzio.com]

> Wow....good-luck with the samisnake hack...and I look forward to your
> progress

I send a couple of e-mails to samisnake, giving him the URL of the 1.30
BIOS, and he modded it. This is the most up-to-date BIOS for the ASRock
H77M Micro-ATX.

Should you be having trouble downloading from his site, try purging your
browser's cache.

Seems he changed the name from <something>-<something> to
<something>_<something> and the browser was getting all confused.

[phubai]

I had just come across a guide at tonymacx86 where the BIOS mod was mentioned:

Link

I'm sure you've seen it, and the link that it provides is a direct DL link.

From the article by Buckeyes1995:

BIOS and Settings

There is a great thread in the DSDT forum for modified ASRock bioses. These modifications, done by samisnake, remove the lock in the BIOS that deals with cstates. Therefore, using the modified Z77E-ITX bios (at the time of this writing, BIOS version 1.30), allow us to use native power management without the need for the nullcpupowermanagement kext. Download the modified 1.30 bios here: Z77E-ITX 1.30 Modified BIOS

I did not have to modify any BIOS settings to get OSX to install or run. 

Note if you use HD4000, set the on board video ram to 32MB (to use the EFI injection below).

[pete...@cruzio.com]

> I had just come across a guide at tonymacx86 where the BIOS mod was
> mentioned:
>

> Link<http://www.tonymacx86.com/248-building-customac-customac-mini-2012/comments4.html#comments>


Here ...

http://biosrepo.wordpress.com/

... is the main page where samisnake (or whomever) describes the tools
needed.

However ...

MANY UEFI mobos already have modded BIOSes available.

Use the sub-pages labeled with your mobo manufacurer to see what is
already available.

I quickly found that the ASRock H77M Micro-ATX 1.20 modified BIOS was
available and samisnake modded the 1.30 BIOS within a day or so.

This particular H77 mobo is four slots, not one slot as with the ASRock
H77M-ITX.

I got ML reinstalled on the HD.

Note: the SATA III ports are not seen by the installer, for some reason,
so just move the HD to a SATA II port (obviously, the optical drive should
be on a SATA II port).

Gets pretty far into booting OS X and then gets a Stopping Job ... message
from -v.

Hmmm.

[Christian Wacker]

Just wondering if you've compared the modified bios to the unmodified one in a hex or binary editor to see what changes were made? I assume they are minimal.

-----Original Message-----
From: pete...@cruzio.com
Sent: Sunday, November 18, 2012 10:34
To: hq...@googlegroups.com
Subject: Re: [HQ-A] Re: H77/i5-2500K Build to Begin


> Wow....good-luck with the samisnake hack...and I look forward to your
> progress

I send a couple of e-mails to samisnake, giving him the URL of the 1.30
BIOS, and he modded it. This is the most up-to-date BIOS for the ASRock
H77M Micro-ATX.

Should you be having trouble downloading from his site, try purging your
browser's cache.

Seems he changed the name from <something>-<something> to
<something>_<something> and the browser was getting all confused.

--
To unsubscribe: hq-a+uns...@googlegroups.com
List options: http://groups.google.com/group/hq-a/subscribe

HQ-A homepage: http://hq-a.weebly.com/
Group Files Page: http://bit.ly/ia8E2l

[pete...@cruzio.com]

> Just wondering if you've compared the modified bios to the unmodified one
> in a hex or binary editor to see what changes were made? I assume they
> are minimal.

I have NOT done a comparison, but samisnake explains the modification by
changing a SINGLE instruction in the BIOS so as to ALWAYS allow the MSR to
remain UNLOCKED.

The process involves changing this single instruction, and then
recomputing the checksum so that the mobo initializer will not reject the
modified ROM.

(I believe that the mobo initializer is located in the BOOT Block within
the ROM image. Usually, this block MAY NOT be modified in any way, but I
suppose there are ways of modifying the BOOT block. Based upon my
experience with modifying ATA 66 cards to support MacOS X, this BOOT block
is the first "page", often the first 256 bytes of the subject ROM image,
and special authority is needed to modify it.)

The modified UEFI BIOS is then loaded during mobo initialization, wherein
the BIOS, which is actually stored serially in an eight-pin electrically
Eraseable/Programmable ROM (EEPROM), usually of about an 8 Mbyte capacity,
although it is stored serially, and is subsequently copied, bit to byte,
into "high" RAM where it is executed, and this "high" RAM ROM image WILL
support MSR modifications by AppleIntelCPUPowerManagement.

Apparently, Gigabyte UEFI BIOSes (and, presumably also those which utilize
Award UEFI BIOSes) DO NOT have this issue; yet ALL AMI BIOSes DO have this
issue; whereas NO Intel UEFI BIOSes can be modified in any way.

[PH]

I have had LIMITED success with the ASRock H77M mobo using iATKOS ML2. This mobo costs about $69 from NewEgg, so it is just about the lowest-cost 7-series mobo out there.

Airport, using a Broadcom mini-PCI card in a mini-PCI-to-PCI adapter, is working.

Built-in Ethernet is NOT working ... yet.

I am presently updating the 10.8.2 to the "latest and greatest", using WiFi as my sole E-net connection.

HD 3000 DID NOT WORK. I thereafter substituted the "usual suspect" video card, an nVidia-branded 8400GS, with positive results.

(After this FAILURE trying to use HD 3000 video, I am understandably cautious about trying to use HD 4000 video, say, after Legend Micro returns the faulty GBT AIO PC, which they JUST NOW received.)

Word to the wise ... when trying iATKOS ML2, definitely directly connect your monitor and keyboard/mouse DIRECTLY to your Hack. OR ... be prepared to cycle your monitor and keyboard/mouse SEVERAL TIMES between your normal Hack and the H77M.

I will say this much ... the iATKOS ML2, using the DEFAULTS, WILL produce a fully bootable system, but you must be prepared to hand edit the DSDT, should you initially try the "no DSDT" option.

As I said before, wired E-net, even if iATKOS correctly determines your E-net NIC, MAY NOT WORK.

However, should that be some comfort, iATKOS WILL correctly determine your sound subsystem, and substitute the correct version of Voodoo.

I am writing this from my MAIN SYSTEM, my hp/Compaq DC7900 E8400.

As I am presently running SOFTWARE UPDATE on my ASRock H77M, other functions on my ASRock H77M are inhibited.

The on-mobo USB 3.0 ports ARE correctly identified as Super-speed USB ports, most probably using Apple's drivers as I DID NOT specify using a third-party USB 3.0 driver in the Installer options menu.

Oh, well ... always SOME issue with a Hack.

[mosslack]

On Jan 23, 2013, at 3:37 PM, PH wrote:

I have had LIMITED success with the ASRock H77M mobo using iATKOS ML2. This mobo costs about $69 from NewEgg, so it is just about the lowest-cost 7-series mobo out there.

Airport, using a Broadcom mini-PCI card in a mini-PCI-to-PCI adapter, is working.

Built-in Ethernet is NOT working ... yet.

what chipset is used?

I am presently updating the 10.8.2 to the "latest and greatest", using WiFi as my sole E-net connection.

HD 3000 DID NOT WORK. I thereafter substituted the "usual suspect" video card, an nVidia-branded 8400GS, with positive results.

(After this FAILURE trying to use HD 3000 video, I am understandably cautious about trying to use HD 4000 video, say, after Legend Micro returns the faulty GBT AIO PC, which they JUST NOW received.)

The Intel HD gfx have been proven to work on certain Hackintosh models, especially Gigabyte, so perhaps it may have more to do with the OEM in this case.

Word to the wise ... when trying iATKOS ML2, definitely directly connect your monitor and keyboard/mouse DIRECTLY to your Hack. OR ... be prepared to cycle your monitor and keyboard/mouse SEVERAL TIMES between your normal Hack and the H77M.

I will say this much ... the iATKOS ML2, using the DEFAULTS, WILL produce a fully bootable system, but you must be prepared to hand edit the DSDT, should you initially try the "no DSDT" option.

As I said before, wired E-net, even if iATKOS correctly determines your E-net NIC, MAY NOT WORK.

However, should that be some comfort, iATKOS WILL correctly determine your sound subsystem, and substitute the correct version of Voodoo.

Interesting, so this is done for you, no need to choose the correct kext in the initial customize settings?

I am writing this from my MAIN SYSTEM, my hp/Compaq DC7900 E8400.

As I am presently running SOFTWARE UPDATE on my ASRock H77M, other functions on my ASRock H77M are inhibited.

The on-mobo USB 3.0 ports ARE correctly identified as Super-speed USB ports, most probably using Apple's drivers as I DID NOT specify using a third-party USB 3.0 driver in the Installer options menu.

Oh, well ... always SOME issue with a Hack.

Thanks for the update on the iAtKOS ML2 distro. I had negative results on my attempts using it. Nice to see it works, at least somewhat for others.

[pete...@cruzio.com]

>> I have had LIMITED success with the ASRock H77M mobo using iATKOS ML2.
>> This mobo costs about $69 from NewEgg, so it is just about the
>> lowest-cost 7-series mobo out there.
>>
>> Airport, using a Broadcom mini-PCI card in a mini-PCI-to-PCI adapter, is
>> working.
>>
>> Built-in Ethernet is NOT working ... yet.
>
> what chipset is used?

Intel H77.

The wired E-net NIC is recognized, but it always returns a 169.x.x.x
address, which is useless.

However, the Broadcom WiFi NIC always returns a 192.x.x.x address, which
works.

The 7-series just may be a "bridge too far".

Whatever the reality, I will eventually sort it out.

[mosslack]

On Jan 23, 2013, at 4:18 PM, pete...@cruzio.com wrote:

I have had LIMITED success with the ASRock H77M mobo using iATKOS ML2.

This mobo costs about $69 from NewEgg, so it is just about the

lowest-cost 7-series mobo out there.

Airport, using a Broadcom mini-PCI card in a mini-PCI-to-PCI adapter, is

working.

Built-in Ethernet is NOT working ... yet.

what chipset is used?

Intel H77.

Actually I meant the ethernet chipset. 

[pete...@cruzio.com]

>>>> Built-in Ethernet is NOT working ... yet.
>>>
>>> what chipset is used?
>>
>> Intel H77.
>
> Actually I meant the ethernet chipset.

The usual Realtek ... RTL8168E-VL/8111E-VL.

Should work, but just returns 169.x.x.x, which is as if the NIC is not
connected to anything.


I was VERY disappointed that the HD 3000 video did not work (-v said the
frame buffer failed to load), but the 8400GS was successful.

The first installation was not successful, most probably because I
installed it over another installation.

The second installation was on a cleaned drive, and worked.

iAKTOS ML2 in both cases. From the DVD.

Incidentally, USB 3.0 appears to be supported using AppleUSBXHCI, NOT the
usual USB 3.0 driver from LaCie.

Right now I am using my "best effort" DSDT, Rev. 0.

I'm sure there will be mods forthcoming before Rev. A.


The GBT INC H61MT was delivered today at Legend Micro in Fort Lauderdale.
We'll soon see if a mech replacement or a repair is provided. Or, if the
return is rejected.

In future, I would wait until Amazon or another retailer with a nationwide
presence decides to carry such an item.

[pete...@cruzio.com]

> The on-mobo USB 3.0 ports ARE correctly identified as Super-speed USB
> ports, most probably using Apple's drivers as I DID NOT specify using a
> third-party USB 3.0 driver in the Installer options menu.

Using IORegistryExplorer, the USB 3.0 system under Mountain Lion appears
as follows:

XHCI Root Hub Super-Speed Simulation@14 USB Vendor Apple Inc

XHCI Root Hin USB 2.0 Simulation@14 USB Vendor Apple Inc

The mobo's DSDT has:

Device (XHC)
{
Name (_ADR, 0x00140000)
OperationRegion (XPRT, PCI_Config, 0xD0, 0x10)
Field (XPRT, DWordAcc, NoLock, Preserve)
{
PR2, 32,
PR2M, 32,
PR3, 32,
PR3M, 32
}
...

The associated kext is: AppleUSBXHCI version 1.4.0

So, although Apple may not have USB 3.0 on an official basis, they appear
to be supporting it on ML, most probably because they intend on migrating
to Intel's 7-series platform which has USB 3.0 builtin.

Intel's 6-series platform still requires an add-on chip, such as
NEC/Renesas, Etron, ASMedia or other.

And, Intel's 4-series platform definitely requires an add-on kext to
support the add-on USB 3.0 chip, as does the 5-series.

[faithie999]

peter--i was wondering why you are using iATKOS on this project.  did the normal ihackintosh/unibeast/etc process fail due to the bios lock issue?

FYI, you were wondering about HD 4000 graphics.

my NUC has an i3-3217U proc and its HD4000 worked immediately and properly after i normally installed ML.

i guess the intel mobo in the NUC doesn't have a locked bios, since i installed the "modded-for-10.8.2" appleintelcpupowermanagement.kext and i am not using nullcpupowermanagement.kext.  sleep/wake is working fine, including ability to be reawakened with my BT mouse or keyboard.

[pete...@cruzio.com]

> peter--i was wondering why you are using iATKOS on this project. did the
> normal ihackintosh/unibeast/etc process fail due to the bios lock issue?

Because of the complexities of the 7-series, and my previous problems
using a modified iHackintosh ML installer (USB stick) on it.

I used the iATKOS DVD as I was unable to create a USB stick from the
iATKOS DVD.

I had had this 7-series project hanging fire for so long, I just wanted
the quickest path to ML. Period.

> FYI, you were wondering about HD 4000 graphics.
>
> my NUC has an i3-3217U proc and its HD4000 worked immediately and properly
> after i normally installed ML.

Excellent.

I wonder, then, why my i5-2500K (HD 3000) did not work at the INSTALLED OS
X level, although it certainly worked at the INSTALLER OS X level (which
may be Lion).

The 7-series Intel mobos were reported to be unmodifiable for the BIOS hack.

Perhaps this assertion applies only to certain 7-series Intel mobos?

It reportedly DOES NOT apply to Gigayte 7-series mobos.

> i guess the intel mobo in the NUC doesn't have a locked bios, since i
> installed the "modded-for-10.8.2" appleintelcpupowermanagement.kext and i
> am not using nullcpupowermanagement.kext. sleep/wake is working fine,
> including ability to be reawakened with my BT mouse or keyboard.

Excellent!

[pete...@cruzio.com]

>> peter--i was wondering why you are using iATKOS on this project. did
>> the
>> normal ihackintosh/unibeast/etc process fail due to the bios lock issue?
>
> Because of the complexities of the 7-series, and my previous problems
> using a modified iHackintosh ML installer (USB stick) on it.
>
> I used the iATKOS DVD as I was unable to create a USB stick from the
> iATKOS DVD.
>
> I had had this 7-series project hanging fire for so long, I just wanted
> the quickest path to ML. Period.

Wired E-net is now working, without any effort on my part except an
extended power down (from 1 am to about 8:30 am).

I have, therefore, turned off WiFi for the present.

Alas, into every day (including this one) a little rain must fall (and it
really is, too). The formerly silver disk icons are now yellow.

Oh, well.

[mosslack]

Funny how just leaving the system off for an extended time period fixed the problem, although I bet it was working before but just needed maybe one more restart to get all your ducks in a row. 8^)

At least the orange disk icons are an easy fix. I think you are on a good path with this one.

[pete...@cruzio.com]

> Funny how just leaving the system off for an extended time period fixed
> the problem, although I bet it was working before but just needed maybe
> one more restart to get all your ducks in a row. 8^)
>
> At least the orange disk icons are an easy fix. I think you are on a good
> path with this one.

I applied all recommended maintenance from the Apple Store.

Applied successfully, but now the cursor is a non-rotating "beach ball"
and it does nothing except move upon command. Won't select windows.
Keyboard may or may not work, too.

Good thing I CCC'd the main partition to the backup partition before I
applied the maintenance.

[mosslack]

Just another example of a good reason to make a backup, especially when you are in the development phases of a new system. It saves countless hours in the long run. I'm sure you will figure it out.

[faithie999]

my most valuable hacking tools are a couple of smallish capacity 2.5 inch sata HD's (a 40 and a 160) that i picked up cheaply on LEM swap list.

one of those and an external usb adapter is all i need to make a CCC of whatever HD i'm working on.  then i can install the clone as the main HD and quickly swap it and the orig HD back and forth during the troubleshooting process.

[pete...@cruzio.com]

> my most valuable hacking tools are a couple of smallish capacity 2.5 inch
> sata HD's (a 40 and a 160) that i picked up cheaply on LEM swap list.
>
> one of those and an external usb adapter is all i need to make a CCC of
> whatever HD i'm working on. then i can install the clone as the main HD
> and quickly swap it and the orig HD back and forth during the
> troubleshooting process.

Agreed.

However, in this day and age, one really needs TWO: a USB 2.0 and a USB 3.0.

They aren't that expensive, particularly at Fry's which carries a number
of brands, including Inland, which is often a value leader.

[Kris Tilford]

On Jan 24, 2013, at 6:28 PM, pete...@cruzio.com wrote:

> However, in this day and age, one really needs TWO: a USB 2.0 and a
> USB 3.0.

With no incompatibility in any combination of USB 2.0 and USB 3.0, why
would one really need both?

[pete...@cruzio.com]

>> However, in this day and age, one really needs TWO: a USB 2.0 and a
>> USB 3.0.
>
> With no incompatibility in any combination of USB 2.0 and USB 3.0, why
> would one really need both?

I was thinking of my own situation, where I have an old USB 2.0 emergency
boot drive, yet all my new machines, including my DC7900, are fitted with
a USB 3.0 card or have USB 3.0 built-in (either in the 7-series chip set
or a 6-series with an on mobo NEC/Renesas-compatible chip).

My seldom used Shuttle SH67 has two ASMedia USB 3.0 chips, one for the
rear panel and one for the front panel. With the latest firmware update
from ASUS (the corporate owner of ASMedia), these chips are compatible
with the LaCie hacked driver and possibly others.

[pete...@cruzio.com]

Some more issues on the ASRock H77M ...

After turning kernel cache ON, the disk icons are, again, appearing as
silver icons.

However, the machine is taking a LOT of time to, finally, enable the
cursor for its customary functions.

However, eventually, the cursor WILL become active and become usable.

My presently used org.chameleon.Boot.plist is as follows:

<dict>
<key>Kernel</key>
<string>mach_kernel</string>
<key>Kernel Flags</key>
<string>-v npci=0x2000</string> <--- For debugging purposes, only
<key>Boot Graphics</key>
<string>Yes</string>
<key>Wait</key>
<string>Yes</string> <--- For debugging purposes, only
<key>Quiet Boot</key>
<string>No</string>
<key>Timeout</key>
<string>5</string>
<key>UseKernelCache</key>
<string>Yes</string>
<key>GraphicsEnabler</key>
<string>Yes</string>
<key>EthernetBuiltIn</key>
<string>Yes</string>
</dict>

AFTER the cursor becomes active, which may take several minutes, the
machine behaves as one might expect of a Hack.

I cannot complain about the performance of the i5-2500K (unlocked
multiplier), but the boot loader will not enable the multiplier, which for
me is not a deal breaker. While I paid extra for the unlocked multiplier
(namely, the K-series CPU), the boot loader WILL NOT enable its function!

The HD 3000 still does not work, and I am using a 512 MB DDR3 VRAM 8400GS
as my sole video card.

Audio is working using the iATKOS ML2-supplied AppleHDA1082_892.kext.

All-in-all, I think the iATKOS group did a pretty good job of creating
their 10.8.2 "distro".

I still have the Apple Store 10.8 distribution, should I want to go back
to "basics", but I have had precious little luck using a USB flash drive
installation method with Mountain Lion, save my previous very positive
experience with the iHackintosh method on my DC7900 CMT, which is still my
main system.

[faithie999]

My seldom used Shuttle SH67 has two ASMedia USB 3.0 chips, one for the 

rear panel and one for the front panel. With the latest firmware update
from ASUS (the corporate owner of ASMedia), these chips are compatible
with the LaCie hacked driver and possibly others.

Peter--do you have a link to the ASUS firmware update?

i have a usb3 expresscard for my hackbook that has never worked at usb3 speed, regardless of which kext i have tried.  i wonder if it has the asmedia chip in it, that a firmware update would fix?

thanks 

[pete...@cruzio.com]

> Peter--do you have a link to the ASUS firmware update?
>
> i have a usb3 expresscard for my hackbook that has never worked at usb3
> speed, regardless of which kext i have tried. i wonder if it has the
> asmedia chip in it, that a firmware update would fix?

If the USB 3.0 chips are working at 2.0 speed, then the firmware update is
not required.

The early ASMedia USB 3.0 chips were not working, for unstated reasons.
ASUS came out with a firmware update which was executable only under
Windows which changed the firmware. This was a three part process, the
first and second parts were necessary for the USB 3.0 chip firmware
update, and the third was applicable to the ASUS mobo. So, you got the
first two parts to properly install and then ignored the failure on the
third, if your mobo was not an ASUS.

I applied this to my Shuttle SH67 system and it made the ASMedia USB 3.0
chips on the rear (one chip) and on the front (another chip) work with the
LaCie driver under OS X.

After all the hassle with ASMedia USB 3.0 chips, and a few issues with
Etron Tech USB 3.0 chips, I decided to select only NEC/Renesas chips for
future purchases, or built-in USB 3.0 for 7-series.

[pete...@cruzio.com]

>> However, in this day and age, one really needs TWO: a USB 2.0 and a
>> USB 3.0.
>
> With no incompatibility in any combination of USB 2.0 and USB 3.0, why
> would one really need both?

I have read where the 7-series (H77M, for example, but there are others)
mobo's USB 3.0 ports (the blue ones) only operate when a true USB 3.0
device is connected to these, and these USB 3.0 ports are non-functional
when USB 2.0 or USB 1.1 devices are connected to these.

This would seem to be true, as my ASRock H77M mobo will not recognize a
USB 2.0/1.1 flash drive on its USB 3.0 ports.

This, not withstanding the fact that Apple's USB 3.0 drivers are indeed
loaded and the ports are indeed recognized by IORegistryExplorer.

[PH]

I performed a combination DSDT Editor/DSDTSE compilation of the "raw" ASRock H77M 1.30 DSDT using a modified H67 modification file.

The result produced a Rev. 1 DSDT which I believe will support HD 3000. DSDT Editor produced a DSDT with a few compiler errors, and the manual use of DSDTSE eliminated those and produced a fully working DSDT.aml.

I am advised that support of native USB 3.0 AND native USB 3.0 with a USB 2.0 compatible device requires additional DSDT edits, but the person which stated that claim has as yet failed to produce examples of the required edits.

Not withstanding the above, I am now running my very low cost H77M using my Rev. 1 DSDT, but still with an 8400GS, and no USB 3.0 or USB 3.0 with USB 2.0 compatible devices.

This ...

http://www.newegg.com/Product/Product.aspx?Item=N82E16813157303

... is the mobo.

At just under $70, it is a really economical way to get into 7-series computing.

I am using it with an i5-2500K, but not with the unlocked multiplier which is implied by the K suffix. I only got the K proc because the Santa Clara Micro Center store (before it closed) was having a very attractive sale on those procs. Were I to do it again, I would prefer an i3.

Again, HD 3000 was tested, and is working, at least for BIOS and boot loader functions.

HD 3000 support was added with the Rev. 1 DSDT, but I have not tested it in that mode under OS X.

Too many projects, too little time ... plus, I just came off of a three week bout with the flu, not withstanding the facts that I ALWAYS get the seasonal flu inoculations.

Everything is verified to work with the exception of HD 3000 (not re-tested after the DSDT update) and USB 3.0. I am presently supporting WiFi access using a Broadcom 4318 mini-PCI card in a mini-PCI to PCI adapter, both from China.

My board is at BIOS level 1.30, with Samisnake's mods for the MSR issue. NewEgg shipped the board at BIOS level 1.20.

The installation method was iATKOS ML2.

ALC892 audio is working perfectly.

A DSDT-less installation is indeed possible, but I prefer a custom DSDT, particularly as support for some of the new functions requires DSDT edits.

[PH]

Here's the deal on Ivy Bridge USB 3.0 ports ...

These are theoretically adaptable to USB 3.0 native or USB 2.0 compatibility (which, I suppose, also provides USB 1.1 compatibility).

However ...

For OS X, the DSDT usually requires significant modifications even to get USB 3.0 function working, and it requires even more modifications to get USB 2.0 on USB 3.0 ports working. There is one example out there, for the ASUS P877M Pro 1708.

A hacker who goes by the name Mieze states that she has got the compatibility function to work, without providing her fix.

[pete...@cruzio.com]

USB seems to be a general issue on 7-series mobos.

For example, after the transition from the boot loader to the Desktop, the
USB keyboard and mouse will generally be ignored for almost precisely 2
minutes. Thereafter, the keyboard and mouse will work correctly, even if
passed through a KVM-A switch.

While I don't consider that particular behavior desirable, I can tolerate
it as I have several Hacks and one Mac which are connected through the
KVM-A switch, and I can get some meaningful work done on the alternate
systems while I am awaiting the H77M to come alive.

I have had no KPs on this system, save the initial one which was a
consequence of the MSR updating issue, which Samisnake fixed. Both the
1.20 and 1.30 BIOSes are fixed.

[pete...@cruzio.com]

My review on NewEgg ...


5 out of 5 eggsExcellent, Low-cost Hackintosh

Pros: Low cost platform for LGA 1155 procs and OS X Mountain Lion.

HD 3000 works great at the BIOS and boot loader levels (HD 4000 was not
tested).

ALC892 for great sound.

Presently running Mountain Lion 10.8.2

Cons: Requires the BIOS modification for MSR updating, from Samisnake, for
OS X compatibility, but Samisnake has produced updates for both the 1.20
and 1.30 BIOSes [ the only ones which are out there ] .

HD 3000 doesn't work at the OS X Desktop level without having the required
DSDT edits, hence a DSDT-less installation precludes HD 3000 (HD 4000 was
not tested).

USB 3.0 doesn't work at the OS X Desktop level without having the required
DSDT edits.

USB 2.0 compatibility on USB 3.0 ports doesn't work at the OS X desktop
level without MAJOR DSDT edits, which, so far, haven't been produced for
ASRock boards, but have for certain ASUS boards [ B75, for example ] .

Other Thoughts: Very reliable board at a very attractive price.

I used a "distro" [ iATKOS ML2 ] , but pedal-to-the-metal installations of
Mountain Lion are possible using the "usual suspect" tools and a USB flash
drive [ UniBeast and MultiBeast for ML ] .

If you need FireWire, use one of the PCI slots (this board has two).

If you truly need USB 3.0, and USB 2.0 compatibility on a USB 3.0 port,
use the PCI-e 1x slot. A NEC/Renesas-based card is highly recommended.