MSI B75 Series With Special DSDT Fully Supports USB 3.0/USB 2.0/USB 1.1 Compatibility

[PH]

A truly exceptional hacker who goes under the "handle" Mieze has come up with a heavily modified DSDT implementation which provides true USB 3.0/USB 2.0/USB 1.1 compatibility, for MSI B75 series mobos (and, apparently, ONLY those).

Her preferred implementation is for a four-slot (with four DDR3 memory slots) MSI B75M-P45 mobo, but I can conform that her DSDT also works ... unmodified ... for a seven-slot (also with four DDR3 memory slots) MSI B75A-G43 mobo. I don't know if it also works for a four-slot (with two DDR3 memory slots) MSI mobo. I am basically a seven-slot PCI/four-slot RAM type of guy!

The MSI B75A-G43 mobo is presently selling at a price which is reduced to $72.99 with "free" shipping at NewEgg. For today, for sure. And, perhaps, for the next few days.

I have had my B75A-G43-based system working for several days now, and it is ROCK SOLID. I purchased an extra year warranty for $0.99 (the regular price is $9.99) just because MSI ... at least in my book ... has had occasional bouts with product quality issues, and the discounted extra year of warranty was priced too low to ignore.

My B75A-G43 system presently has a i3-level G850 proc, but I intend on updating it to an i5 soon. I am presently using an nVidia 8400GS 512 MB DDR3 video card as the HD 2000 video of the G850 is not supported by OS X.

I am presently running iATKOS ML2 with some, but not all, software updates beyond the 10.8.2 which the iATKOS group has provided, most particularly the Brother laser printer update. I DO own the retail version of ML, but I have, for the present, elected to go with a "distro", due to lack of time available to create the required FLASH drive installer.

So far, there are NO NEGATIVES associated with the MSI B75A-G43 mobo. Everything works as it is supposed to.

Now, the REALLY BIG NEWS: Mieze's DSDT does, indeed, provide COMPLETE USB 3.0/USB 2.0/USB 1.1 compatibility, even on my B75A-G43 mobo, whereas her DSDT was designed for her B75M-P45 mobo.

This is really not a stretch, as her mods are largely in the ECH1 and XCH1 (renamed from XHC) DSDT devices, and some additional methods which are utilized for both devices.

Indeed, her mods were adapted from a MacBook Pro's DSDT. And a VERY GOOD job of adaptation she did, too!

With either mobo and her DSDT, you may freely insert a USB 2.0 device (a FLASH drive, for example) into a USB 3.0 port and have it accepted as a USB 2.0 device. Also, you may freely insert a USB 1.1 device (an Apple Pro keyboard, for example) into a USB 3.0 port and have it accepted as a USB 1.1 device.

Indeed, you may unplug your keyboard and mouse from its normal USB 2.0/1.1 port and plug it into a USB 3.0 port, and it will automagically be accepted as USB 1.1, even though it is plugged into a USB 3.0 port.

7-Series Intel USB 3.0 support actually requires some DSDT mods to be even accepted as USB 3.0 under OS X, for, otherwise, the Intel USB 3.0 ports are basically useless with OS X.

With Maize's DSDT, you may freely use any USB port with ANY USB device.

For otherwise, even with some DSDT mods for USB 3.0 support, USB 3.0 is forever USB 3.0, and USB 2.0/1.1 is forever USB 2.0/1.1.

The MSI B75 mobos (as do others) have a special connector for dual USB 3.0 front panel ports, and her implementation will allow for USB 3.0/USB 2.0/USB 1.1 compatibility on those front panel ports, too. Appropriate cables and brackets are available on The Bay for about $10, with "free" shipping from HK.

So, with Mieze's DSDT, it is indeed possible to obtain the architectural maximum number of USB ports, and have these completely compatible across USB 3.0/USB 2.0/USB 1.1.

BTW, you WILL need to update the A01 BIOS to incorporate the usual MSR update mod, and, in the process, you might as well go to the A07 BIOS.

MFLASH BIOS updating is a little less flexible than Gigabyte's QFLASH BIOS updating. With MFLASH, the target BIOS image MUST reside in the root of your FLASH drive ... paging through various folders, as is supported by QFLASH, is not supported by MSI's MFLASH.

[faithie999]

Peter--great report, nicely detailed, as always.

when you update your proc to an i5, which model will you go for, and what kind of price are you finding?

thanks

ken

[pete...@cruzio.com]

> Peter--great report, nicely detailed, as always.

Thank you.

It was a VERY long road to true USB 3.0/USB 2.0/USB 1.1 compatibility, and
I finally decided that I lacked the necessary skills to properly execute
this on a 7-Series, although I obviously have been successful in the past
on a 6-Series and earlier.

> when you update your proc to an i5, which model will you go for, and what
> kind of price are you finding?

I have a number of i5s in various test systems.

I am right now trying to conserve costs as I prepare for a costly "end
game" in an intra-family dispute.

I will be reusing i5s from presently disused 61- or 77-based systems, and
moving my HD 3000- and HD 4000-based procs, which are i3s, to those
systems which cannot support PCI- e 16x video.

The $79.99 price ($72.99 for the next few days) of the MSI B75A-G43 mobo
moved me towards this new strategy.

My educated guess that Mieze's MSI B75M-P45 DSDT would work on an MSI
B75A-G43 proved to be true.

I have ordered a second MSI B75A-G43 mobo and will most likely be
completing a semi-matched pair, with one dedicated to 'Net support and the
other dedicated to media duplication, with i5s in both.

Anyway, that is the plan.

[PH]

I have created a folder which contains everything which is necessary to implement either a B75MA-P45 (4 PCI slots/4 DDR3 slots) or a B75A-G43 (7 PCI slots/4 DDR3 slots) system.

Included are the manuals for the respective mobos, the patched BIOSes (for the PM KP hack) for the respective mobos and the required DSDT which works with both.

It is 28 and a tad MBs.

If the folder is unZipped and the contents are placed in the root of a USB FLASH stick, you may directly patch the mobo's BIOS using MFLASH. MFLASH will figure out which BIOS image is appropriate for your mobo.

Available upon request.

[PH]

There have been several B75-based mobos from MSI, but only two are presently available: the B75MA-P45 (three PCI/PCI-e slots in a 4-slot format, plus four DDR3 slots; 32 GB maximum capacity) and the B75A-G43 (seven PCI/PCI-e slots in a 7-slot format, plus four DDR3 slots; also 32 GB maximum capacity).

Both have a Realtek R8111E gigabit LAN port.

The B75MA-P45 has a Realtek ALC887 codec and a three jack rear panel connection; the B75A-G43 has a Realtek ALC892 codec and a six jack rear panel connection.

While I far prefer the flexibility of a 7/4 mobo, should you prefer a 4 (really 3)/4 mobo, both are available, and both are supported by Mieze's DSDT.

[faithie999]

do you have the ALC892 audio working?

also, i know you don't use sleep, but i'm wondering if you've tried it, for research purposes <G>

ken

[pete...@cruzio.com]

> do you have the ALC892 audio working?

Yes, working, but I am using 2.0 sound out of the green jack, only. The
ALC892 supports 7.1 sound, I think.

> also, i know you don't use sleep, but i'm wondering if you've tried it,
> for research purposes <G>

The power button works to place the machine in Suspend, should you want to
do that.

A second power button action brings the machine back from Suspend.

You may view that as a form of "sleep", if you choose.

It is an STR action which is going on ... "Suspend To RAM", and the
machine does power itself down, EXCEPT for RAM, which remains powered.

[PH]

Using Mieze's MSI B75MA-P45 DSDT as a guide, I have come up with my own DSDT for my MSI B75A-G43 mobo.

Mieze's B75MA-P45 DSDT works with a MSI B75A-G43, alright, and was certainly a good way to "bootstrap" the process of making my own DSDT for my mobo.

[pete...@cruzio.com]

As I could not get Ubuntu to load on my B75 system, I installed ML2 with
the no DSDT option and then grabbed the "raw" DSDT using DSDTSE.

There were several issues with that "raw" DSDT, such as the usual _T_0
names, and an unreachable statement, but those were corrected using long
familiar techniques for AMI BIOS DSDTs.

The corrected DSDT compiled OK, and I used it for a little while before
embarking on the USB 3.0 compatibility project.

Using DSDTSE and two separate windows, the one on the left for the
B75A-G43 DSDT (starting without any USB 3.0 stuff) and the one on the
right for the B75MA-P45 DSDT (with Mieze's mods), I moved back and forth
between screens adding her code into my DSDT, as required.

Once a good compile had been achieved, I made my new DSDT the active one
on my B75A-G43 system, and tested it.

The EHC1 and XHC1 sections were taken from her DSDT as those are the areas
which address the "switch" function. There were some other areas which
needed attention, too, and her DSDT was used as a guide.

[faithie999]

that satisfies my definition of sleep.

thanks

[pete...@cruzio.com]

> that satisfies my definition of sleep.

This is accomplished by a SINGLE character change in the PWRB (power
button ?) DSDT method.

Normally, all PWRB methods have:

Name (_HID, EisaId ("PNP0C0C"))

By simply changing this to:

Name (_CID, EisaId ("PNP0C0C"))

you enable the Suspend To Ram function.

You may initiate an STR by depressing the power button momentarily.

Should the machine go to sleep on its own, you may return from STR simply
by depressing the power button momentarily.

The machine is, of course, setup to "sleep never" in the Energy Saver
control panel.

Yet, by this deceptively simple change in the DSDT, you enable Suspend to
RAM, and also returning from STR.

I have incorporated this feature in my DSDTs for quite a while.

Certainly from well before the DC7800/DC7900 project.

[PH]

Herein I want to document the steps which are necessary to incorporate Mieze's USB 3.0/USB 2.0/USB 1.1 compatibility DSDT mods into my B75A-G43 DSDT. These steps, in general, should apply to many other AMI BIOS DSDTs, possibly including other than B75 motherboards, of the 7-Series.

It is a long road, but please bear with me.

DSDTSE is used as the tool. Two windows will be used. The left window will be assigned to my B75A-G43 DSDT. The right window will be assigned to Mieze's DSDT-MSI-B75MA-P45-V1.4c DSDT.

1) Change all:

XHC. --> XHC1.

2) Change all:

(XHC) --> (XHC1)

3) Change all:

XHC, --> XHC1,

4) Change all:

_T_0 --> T_0

5) Change all:

_T_1 --> T_1

6) Insert:

Return (Zero)

before end of Method (_HID in Device (_SB.PCI0.LPCB.TPM)

7) Insert:

_OSI ("Darwin") from Mieze's to mine in _INI Method

8) Replace entire EHC1 and EHC2 devices with Mieze's

9) Copy Method (OSDW from Mieze's to just before end of DSDT

10) Copy DTGP there, too

11) Replace entire XHC1 device with Mieze's

12) Replace entire Method (_OSC with Mieze's

13) Change:

\_SB.PCI0.XHC1.XWAK() --> \_SB.PCI0.XHC1._INI()

14) Replace entire GFX0  device with IGPU device with Mieze's

15) At this point the DSDT should compile without any warnings; if not, correct the errors as required

16) Searching Meize's DSDT for instances of calls to DTGP, copy those into the corresponding devices

17) At this point the DSDT should compile without any errors of any kind.

[PH]

Revision 1

Herein I want to document the steps which are necessary to incorporate Mieze's USB 3.0/USB 2.0/USB 1.1 compatibility DSDT mods into my B75A-G43 DSDT. These steps, in general, should apply to many other AMI BIOS DSDTs, possibly including other than B75 motherboards, of the 7-Series.

It is a long road, but please bear with me.

DSDTSE is used as the tool. Two windows will be used. The left window will be assigned to the "target", my B75A-G43 DSDT. The right window will be assigned to the "source", Mieze's DSDT-MSI-B75MA-P45-V1.4c DSDT.

You may obtain the "raw" DSDT for your mobo by using Ubuntu, but an alternative is to do a "DSDT-less" install using, say, iATKOS ML2, and thereafter launching DSDTSE and pressing the "Extract DSDT" button.

Phase 1 changes: make the "raw" DSDT compile without errors

1) Search for MUT0, 0x0FFF and change it to MUT0, 0xFFFF; in this instance there is only one such occurrence, but in general, there could be several such occurrences

2) Change all:

_T_0 --> T_0 (and also any _T_1s, _T_2s, etcetera)

3) In Device (_SB.PCI0.LPCB.TPM) and within Method (_HID, 0, NotSerialized), insert Return (Zero) before the end of that Method

At this point, the DSDT should compile without any errors; do so and save the resulting .aml and .dsl files as Rev. 0

Phase 2 changes: make the power button work for "Suspend to RAM" (STR) and return from STR, thereby simulating "sleep" (it is assumed that Energy Saver has been set to "Computer sleep Never")

4) Search for Device (PWRB) and change Name (_HID, EisaId ("PNP0C0C")) to Name (_CID, EisaId ("PNP0C0C"))

Phase 3 changes: preparing for copying the "source" devices and methods to the "target"

5) Search for Windows 2001 and insert

        If (_OSI ("Darwin"))
                {
                    Store (0x2710, OSYS)
                }

immediately before that; you may also copy that code from the "source" to the "target" if you want

6) Change all XHC. to XHC1. (note the positions of the periods)

7) Change all  (XHC) to (XHC1)

8) Change all XHC, to XHC1, (note the positions of the commas)

9) Change all GFX0. to IGPU. (note the positions of the periods)

Phase 4 changes: replace the "target" devices and methods with the "source" devices and methods to add the actual USB 3.0/USB 2.0/USB 1.1 compatibility code

10) Replace entire EHC1 and EHC2 devices with those from the "source"

11) Replace entire XHC1 device with that from the "source"

12) Replace entire Method (_OSC, 4, Serialized) with that from the "source"

13) Search the "source" for Method (OSDW, 0, NotSerialized) and copy the entire Method from the "source" to just before end of the "target"

14) Search the "source" for Method (DTGP, 5, NotSerialized) and copy the entire Method from the "source" to just before beginning of the "target" Method (OSDW, 0, NotSerialized)

15) Change \_SB.PCI0.XHC1.XWAK() to \_SB.PCI0.XHC1._INI()

16) Replace entire "target" GFX0  device with "source" IGPU device

17) At this point the DSDT should compile without any warnings; if not, correct the errors as required

Phase 5 changes: add the calls to DTGP to make OS X aware of the devices so the initialization code will assign the correct device driver in all cases

18) Search the "source" DSDT for instances of calls to DTGP, and where there is no corresponding call to DTGP in the "target" device, copy the "source" call to DTGP to the "target"

At this point, the DSDT should compile without any errors; do so and save the resulting .aml and .dsl files as Rev. 1

[PH]

Revision 2 (Bold type added to emphasize "targe" and "source" text)

Herein I want to document the steps which are necessary to incorporate Mieze's USB 3.0/USB 2.0/USB 1.1 compatibility DSDT mods into my B75A-G43 DSDT. These steps, in general, should apply to many other AMI BIOS DSDTs, possibly including other than B75 motherboards, of the 7-Series.

It is a long road, but please bear with me.

DSDTSE is used as the tool. Two windows will be used. The left window will be assigned to the "target", my B75A-G43 DSDT. The right window will be assigned to the "source", Mieze's DSDT-MSI-B75MA-P45-V1.4c DSDT.

You may obtain the "raw" DSDT for your mobo by using Ubuntu, but an alternative is to do a "DSDT-less" install using, say, iATKOS ML2, and thereafter launching DSDTSE and pressing the "Extract DSDT" button.

Phase 1 changes: make the "raw" DSDT compile without errors

1) Search for MUT0, 0x0FFF and change it to MUT0, 0xFFFF; in this instance there is only one such occurrence, but in general, there could be several such occurrences

2) Change all:

_T_0 to T_0 (and also any _T_1s, _T_2s, etcetera)

[PH]

Source for Mieze's MSI B75MA-P45 DSDT (reference model for USB 3.0/2.0/1.1 compatibility) is ...

http://www.tonymacx86.com/dsdt/61894-dsdt-msi-b75ma-p45.html

... enjoy!

[PH]

Revision 3 (Bold type added to emphasize "target" and "source" text; saving additional versions; other corrections)

Herein I want to document the steps which are necessary to incorporate Mieze's USB 3.0/USB 2.0/USB 1.1 compatibility DSDT mods into my B75A-G43 DSDT. These steps, in general, should apply to many other AMI BIOS DSDTs, possibly including other than B75 motherboards, of the 7-Series.

It is a long road, but please bear with me.

DSDTSE is used as the tool. Two windows will be used. The left window will be assigned to the "target", my B75A-G43 DSDT. The right window will be assigned to the "source", Mieze's DSDT-MSI-B75MA-P45-V1.4c DSDT.

You may obtain the "raw" DSDT for your mobo by using Ubuntu, but an alternative is to do a "DSDT-less" install using, say, iATKOS ML2, and thereafter launching DSDTSE and pressing the "Extract DSDT" button.


Phase 1 changes: make the "raw" DSDT compile without errors

1) Search for MUT0, 0x0FFF and change it to MUT0, 0xFFFF; in this instance there is only one such occurrence, but in general, there could be several such occurrences

2) Change all:

_T_0 to T_0 (and also any _T_1s, _T_2s, etcetera)

3) In Device (_SB.PCI0.LPCB.TPM) and within Method (_HID, 0, NotSerialized), insert Return (Zero) before the end of that Method

At this point, the DSDT should compile without any errors; do so and save the resulting .aml and .dsl files as Rev. 0


Phase 2 changes: make the power button work for "Suspend to RAM" (STR) and return from STR, thereby simulating "sleep" (it is assumed that Energy Saver has been set to "Computer sleep Never")

4) Search for Device (PWRB) and change Name (_HID, EisaId ("PNP0C0C")) to Name (_CID, EisaId ("PNP0C0C"))

At this point, the DSDT should compile without any errors; do so and save the resulting .aml and .dsl files as Rev. 1

Phase 3 changes: preparing for copying the "source" devices and methods to the "target"

5) Search for Windows 2001 and insert

        If (_OSI ("Darwin"))
        {
            Store (0x2710, OSYS)
        }

immediately before that; you may also copy that code from the "source" to the "target" if you want

6) Change all XHC. to XHC1. (note the positions of the periods)

7) Change all  (XHC) to (XHC1)

8) Change all XHC, to XHC1, (note the positions of the commas)

9) Change all GFX0. to IGPU. (note the positions of the periods)


Phase 4 changes: replace the "target" devices and methods with the "source" devices and methods to add the actual USB 3.0/USB 2.0/USB 1.1 compatibility code

10) Replace entire EHC1 and EHC2 devices with those from the "source"

11) Replace entire XHC1 device with that from the "source"

12) Replace entire Method (_OSC, 4, Serialized) with that from the "source"

13) Search the "source" for Method (OSDW, 0, NotSerialized) and copy the entire Method from the "source" to just before end of the "target"

14) Search the "source" for Method (DTGP, 5, NotSerialized) and copy the entire Method from the "source" to just before beginning of the "target" Method (OSDW, 0, NotSerialized)

15) Change \_SB.PCI0.XHC1.XWAK () to \_SB.PCI0.XHC1._INI () (note space before parentheses)

16) Replace entire "target" GFX0  device with "source" IGPU device

17) At this point the DSDT should compile without any warnings; if not, correct the errors as required

Phase 5 changes: add the calls to DTGP to make OS X aware of the devices so the initialization code will assign the correct device driver in all cases (in many cases, there will be no additional DTGPs)

18) Search the "source" DSDT for instances of calls to DTGP, and where there is no corresponding call to DTGP in the "target" device, copy the "source" call to DTGP to the "target"

At this point, the DSDT should compile without any errors; do so and save the resulting .aml and .dsl files as Rev. 2

[PH]

The USB 3.0 front panel kit arrived today from Shanghai, China.

It is very well made and includes mounting screws.

About $10, delivered.

I have been using a loose  USB 3.0 cable on the subject mobo for about a week now and am firmly committed to this approach.

You can get two more USB 2.0/1.1 ports on 77 series mobos, but you will still need this front panel kit to get the two USB 3.0 ports accessible externally.

[faithie999]

peter--just looked at ebay for this item.  hard to tell from the picture, but it looks like it is made for a 3 1/2 inch (FDD) bay.  is that what you bought, or does yours fit in a 5 1/4 inch bay?

thanks

ken

[pete...@cruzio.com]

> peter--just looked at ebay for this item. hard to tell from the picture,
> but it looks like it is made for a 3 1/2 inch (FDD) bay. is that what you
> bought, or does yours fit in a 5 1/4 inch bay?

Floppy bay, with floppy positioned mounting screws (included).

Nice kit ... professionally executed.

I have yet to install it. We'll see in the next day or so.

http://www.ebay.com/itm/281013233732?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649

[pete...@cruzio.com]

Installed the USB 3.0 front panel cable kit early this morning.

The supplied screws are inadequate (right thread, too small in head
diameter).

You will need four standard diameter M3-0.5 screws of the type generally
used to install 5.25" optical drives or 3.5" floppy drives (large head,
metric thread).

The cable is extra long, so it should reach even the 6-series mobos which
have a second USB 3.0 chip and an industry standard front panel USB 3.0
header at the very bottom of the mobo.

On my B75-G43, which has internal USB 3.0 and a header along the front of
the mobo, there is a LOT of extra cable, so I looped it.

Worked the first time.

On a machine with TWO 3.5" floppy slots, I would install the USB 3.0
adapter in the lower of the two positions so that the media bay in the
upper of the two positions will not be interfered with by cables.

[pete...@cruzio.com]

Updated DSDT.

No new function. A compiler error which DSDTSE did not catch, but DSDT
Editor DID catch was corrected.

DSDT has a new name which correctly reflects the A suffix to the B75.

[pete...@cruzio.com]

[PH]

Based upon suggestions from Ken Baker and others, the MSI B75A-G43 DSDT has been "rolled" a couple of more times.

It is now at Rev. 13.

Both the latest DSDT and the patched A07 BIOS may be found ...

http://www.tonymacx86.com/dsdt/89426-msi-b75a-g43-a07-dsdt-miezes-usb-3-0-usb2-0-usb1-1-compatibility-support.html#post550811

... there.


With the failure of the 8400GS cooling fan in my DC7900 CMT (the lubrication procedure which I disclosed added SOME life to the bearing, but, alas, it was NOT a permanent fix), I am forced to convert to the MSI B75A-G43-based Hack as my only normally on-line Hack.

To that end, I have added to the already present dual USB 3.0 font panel bracket and cable kit the hp/Compaq USB 2.0 front panel media reader kit which was formerly in the DC7900 CMT.

Both the USB 3.0 and USB 2.0 font panel devices, in addition to the usual USB 2.0/1.1 font panel connections are working perfectly.


The four black M3-0.5 screws which were used as locators in a DC7800/DC7900 were used to mount BOTH the USB 3.0 bracket AND the USB 2.0 media reader.

The cable on the media reader reaches the remaining USB 2.0 header, with some, but not much difficulty.