ASUS RAMPAGE III EXTREME BIOS - V1601 - Updated OROM MOD

[RPro]

I am not sure if a microcode patch for 106A4 will ever come out.  As the supported CPU's are the low end i7 and I would bet few if any are still in use. 

Pir8man,
     Thank you for the continued support in finding, adding, and releasing the updated BIOS versions.  One note regarding 106A4 / Extreme Edition i7-975 / i7-920, 930, 950, 960 - Within Intel's Microcode Update Guide April 2018, they have stopped supporting the 106A4 line - listed as "Stopped" (They no longer even list the processors in the latest guides)



But the Xeon's (206C2) are still supported and listed as production in the latest Intel Microcode Update Guidance July 2018:


We will have to keep an eye out for the 0x1F code as I would guess that further refines existing Spectre (V2) and hopefully might address later Spectre-NG (3A,4)  vulnerabilities recently found (they just keep finding more and more vuls)


Thanks again!

[RPro]

Pir8man,
   I wanted to let you know that for 1F microcode for CPUID 206C2 (Westmere Xeon X56XX i.e. X5690) is actually workable on https://github.com/platomav/CPUMicrocodes.  The issue why MMTool can not add it appears to be that it is missing the end padding.  I manually downloaded the latest Dell R610 bios that contains the microcode and used MCExtractor to extract the microcode.  It has the same SHA256 hash/checksum as the file out on platomav.  I then followed the steps at http://wp.xin.at/archives/4397 (Step 6) to add a 1k padding file (could have likely had it smaller) and then concatenate the orig and padding to make the new final microcode that MMTool can use.
example (note had to use cmd, powershell didn't like the "+" to join the files)
     fsutil File CreateNew .\1kiB-zeropadding.bin 1024
     copy /b cpu206C2_plat03_ver0000001F_2018-05-08_PRD_77DADA73.bin + 1kiB-zeropadding.bin 0206C2-rev-1F.bin

$ sha256sum *
65f74e4eba9e29f3b0be3d3c07a8da7a0143f4d2dad886043c67d18a31563793 *0206C2-rev-1F.bin
5f70bf18a086007016e948b04aed3b82103a36bea41755b6cddfaf10ace3c6ef *1kiB-zeropadding.bin
6d39860414c836e949761ae152c8cf53da9519ffc0636efbacd5800eb3fa6cf8 *cpu206C2_plat03_ver0000001F_2018-05-08_PRD_77DADA73.bin

Then I just used the 0206C2-rev-1F.bin microcode file to add with MMTool 3.26
I currently have the microcode running for about a day and no issues so far:

$ coreinfo  | egrep "Intel|code|PCID" | egrep -v "VMX|SMX|EIST"
Coreinfo v3.31 - Dump information on system CPU and memory topology
Copyright © 2008-2014 Mark Russinovich
Sysinternals - www.sysinternals.com
Intel® Xeon® CPU           X5690  @ 3.47GHz
Intel64 Family 6 Model 44 Stepping 2, GenuineIntel
Microcode signature: 0000001F
PCID            *       Supports PCIDs and settable CR4.PCIDE
INVPCID         -       Supports INVPCID instruction

[pir8man]

Pir8man,
I wanted to let you know that for 1F microcode for CPUID 206C2 (Westmere Xeon X56XX i.e. X5690) is actually workable on https://github.com/platomav/CPUMicrocodes. 

Great info!  I was reading up on this padding on a different site.  The instructions were too dangerous for me to attempt on my own.  The page you found has a very easy method.  With this new info, I cooked up an S21 release with 206C2 Rev.1F.  Check OP for the update.

Also of note.... I did confirmed that 106A4 will NOT receive a microcode update, as Intel has deemed them unsupported.

EDIT: I did a test install on a spare SSD of Win10 Pro using S21i_r3e_1601.rom. Updated base Intel AHCI driver from 10.x to 11.2.0.1006 and it has passed my testing. I tested PXE boot and then disabled it again. i used inspectre and it is safe from Spectre/Meltdown bugs on my W3680 CPU. Both Marvell and jMicron pass basic I/O drive tests.

[Miklsoft]

Hi!
The I thirst downloaded and installed new bios s21_r3e_1601.rom. My system with win 10 works good! My recommendation ...

[rattyto]

Thanks to pir8man. I have update my BIOS to s21i and i have no Problems and pxe works fine. R3E with i7980x

[CaTcHmG]

Hi Guys back after a long hiatus & getting the RE3 back in action from being in a storage black hole & hopefully resurrected :O (wOw ZioGTS,sinders,SkOrPn rings a bell) cant believe dev still going after all these years just want to make sure best bios for raid 0 980x is s19-final or S20-TEST2? thanks in advance keep up the good work guys.

TEST2 is still in testing. so it is up to you to jump into that or not.

... specifically which version is best for your RAID0?  Well, nobody can answer, as you did not provide enough info. READ the thread and you will find the answer!

It was early morning reading with too much info, that's my story n sticking to it, lol just downloaded s21_r3e_1601 will let yall know results, thanks

[rattyto]

Hi pir8man ,

i have 2x Samsung 840 SSD and connect it to the red Sata Port. With the original BIOS i see under Drive Xpert both drives and in Marvel BIOS too.

With this Version i see only one and under Marvel i can create a Raid0 with both SSD's

Is this Bug correct? How i can solve it? I have try all 3x BIOS versions from here and only the 1502 Show both under Drive Xpert

[pir8man]

Hi pir8man ,

i have 2x Samsung 840 SSD and connect it to the red Sata Port. With the original BIOS i see under Drive Xpert both drives and in Marvel BIOS too.

With this Version i see only one and under Marvel i can create a Raid0 with both SSD's

Is this Bug correct? How i can solve it? I have try all 3x BIOS versions from here and only the 1502 Show both under Drive Xpert

Well, first.... I am color blind, so I have no idea what you mean by the red sata port.
please use the controller to define which port you are using.  

Second, why are you connecting SSD drives to the Marvel controller?  That will greatly hinder their speed.  This is repeatedly stated throughout the thread.

Rephrase your question and I will attempt to assist.

[Stahab]

@ pir8man,

The red dual SATA connector rattyto is referring to, is the first in the row of SATA connectors on R3E and is connected to the Marvel controller.
I have included a picture of my board showing these connectors in a circle.

According to rattyto, all modified BIOSes show only one drive even if two drives are connected to the Marvel controller, while when using BIOS 1502 both connected drives show normally.


@ rattyto,

Well, I've just connected two SATA drives (non SSD) to the Marvel controller to replicate your situation. While both drives were shown on the Marvel BIOS page during POST (black screen with white letters before R3E BIOS menu), only the first drive was shown in the Drive Xpert Configuration menu in BIOS. So you are right about this behavior.
But if you look carefully, the default "Normal Mode" for Drive Xpert states the following: "The mode only allows one hard drive to function in the SATA connector", so this behavior is normal and predicted. Maybe the bug you refer to was actually present in BIOS 1502 and was fixed in the later versions.

As pir8man said, the Marvel controller is of bad design and is not recommended for SSD drives, especially in RAID 0. You will have faster performance and much better stability if you connected your SSD drives (RAID 0 or not) on the Intel controller.
This is confirmed numerus times from testing by individual users, myself included.
You should better disable the Marvel controller from BIOS, completely uninstall the driver from the OS and have less wasted resources.

If you still insist to use the Marvel controller for RAID 0, go to Drive Xpert Configuration in BIOS and select "Update to Super Speed". I can't test it right now (it's 1:50 AM here), but I believe this way you should be able to use both drives and configure them as RAID 0.

[rattyto]

Hi Stahab,

thanks for support. You describe my Situation correclty.

Any Change in Drive Xpert, Nprmal, Backup or Super Speed Mode Change Nothing in de Drive Xpert view. Only Sata 0 is in the list.

With 1502 i see both drives.

Thanks a lot for supporting.