S3 Standby does not work with Marvell SATA

[ubiquityman]

I found that on my D975XBX2 (Bad Axe 2) if I have any devices connected to the Marvell SATA ports (DVD or HD), the computer will not resume from an S3 Standby state properly.

It wakes, but some programs hang, etc.
If I disable the SATA devices connected in Device Manager (they get marked with a red "X"), and S3 Standby then works properly.

Does anyone out there have devices connected to the Marvell SATA ports and is also using S3 Standby?

[BlueMagic]

Hi ubiquityman:

In BIOS under 'POWER, ACPI SUSPEND STATE', I'm using <S1 STATE>,
and I'm not having any trouble with my SATA DVD drive that's connected
to the Marvell controller.

I see your property windows; but in my computer under 'Details', my Device Manager does not show the heading 'Power State Mappings' nor any of the 'S' states (maybe because I'm using Windows XP Pro) that your computer
shows.

[ubiquityman]

Would you mind having a 2nd look?
What you should see under "Details" is a drop down box.
It should say "Device Instance Id"
If you click the drop down, it should show a long list of items.
The "Power State Mappings" should be at the bottom of that list.

Also, as a 2nd request (sorry, I know I'm asking a lot here), can you set your BIOS to S3 and se if Standby and then resume works properly?

After resuming, you will have to run your computer for just a minute or so, doing tasks to see if it's behaving properly.

I would really appreciate it if you could do this.

[hamishnorton]

I have a D975XBX2 with 3 hard drives on the Marvell controller. I run Vista 64 Ultimate. My power state mappings show up as "unspecified" for every sleep state for the drives on the Marvell controller, but not the Intel controller. When my computer recovers from s3 sleep, several programs hang and the system cannot shut down normally. Intel has a new BIOS, 2745 whcih claims to address this issue. I tried flashing this BIOS. Successful flashing was reported, but my system behaved like a brick when I tried to restart - no sign of life. I used a recovery floppy and success was again reported, but no sign of even beginning to post. I had to use a recovery floppy for BIOS 2692, and I am back up and running. However, I still need the fix that 2745 claims to provide.

[Apoptosis]

what revision board do you have? Maybe you should try a different BIOS updating method... they have three... the best is the floppy method though and it sounds like you did that one already.

[ubiquityman]

Intel tech support told me to do the jumper->floppy method.
Takes about 5 minutes before the flashing starts. I was ready to give up on it, but it eventually started flashing.

The bad news is that it does not solve the problem. Unless BIOS is flashed improperly, the floppy flashing doesn't make any difference.

I posted a similar thread over at XtremeSystems and someone there indicated the behavior similar to what I'm seeing.

FYI, my board revision is: D53350-505

Same deal here. SATA devices connected to Intel ports work OK.
SATA devices conntected to Marvell ports will cause system instability after resume.

[Dandruff]

i can confirm, that version 2745 will kill your board! also had to go back to 2692 via recovery flash ...

[ubiquityman]

The latest BIOS 2745 crashed my machine too. Endless boot.
Get the old version of the BIOS on a floppy using the BIOS recovery method before attempting to flash.

I reverted back to the previous version, but maybe I should have tried clearing the BIOS or using the failsafe boot via jumpers.

May be related to ECC Memory. I'm using 2x1GB of ECC.

[rwolds9]

Does anyone know if the current bios does fix the S3 standby issues?

[ubiquityman]

According to the readme, the latest bios does fix the S3 Standby problem.
(if you don't have ECC memory.)

[mickrussom]

2747 fixes the s3 issue and will work with ECC. Im running it now.

2745 did brick my machine, but 2747 fixed it.