When I mentioned that the ID change was known, I didn't implied that it was an exclusive discovery, but rather that it was already posted here numerous times. But with this Marvell headache, I suppose it doesn't hurt to bring it up front from time to time. If you don't have a Marvell controller setting under Intel controller in main BIOS, then using Marvell's Ctrl+M is the only way. Although I doubt that they will put a Marvell RAID controller without adding some related settings: either is placed in another section (read the manual), or it has something to do with the current Marvell firmware. To change it from Ctrl+M screen, use Arrows or Tab or Enter until you highlight "Configure SATA as:". By reading your second line, I can assume you did all the above, but the actual RAID mode is missing as an option. This comes from the flashed firmware:



As you can see from the screen, you have the so called IDE firmware, RAID is disabled. Yes, you might say that you can switch to AHCI mode, but this is Marvell's way of complicating even the most simplest things. More than that, you have some leftovers from a previous 2.1.0.1407 firmware. To fix both issues, I suggest you flash the image from the package you used for dumping, it was released for an onboard 9128 controller. You can see the components on the right section from <this screen>. If you are wondering about _extcard on Autoload, it means that the Autoload supports external cards with Marvell controller. Until this version, there were separate releases of the same Autoload, one for onboard and another for external. Marvell addressed this nonsense with a common Autoload, or they screwed up again. Since there haven't been any reports and I vaguely remember that someone used this firmware on an external card, let's hope it is the former.

Edit:
You should check your board and its specifications, maybe you don't have 9128 or you have some limitations in selecting RAID mode. Second warning is for the PATA connector: if it is controlled by Marvell, it might became unavailable with the new Firmware, as you can see from Gaudi tests.

Wow this info is really crucial to me. The thing is that I do not actually remember what I did nearly 4 years ago, but I definately remember that I wanted to update the firmware. However I have doubted that I have actually done so.

If you say there are some remnants of an older firmware, it looks like I had actually attempted an update on the firmware. This would also explain my statement in the mentioned post:

Zitat

---

ok, habe jetzt herausgefunden das SE889128 nur dann angezeigt wird wenn man im BIOS den SATA3 auf RAID (und nicht auf AHCI) schaltet.

---

Translation:

Zitat

---

Ok I found out that the ID 9128 only gets displayed when I set my SATA3 controller to RAID Mode from BIOS (not on AHCI)

---

It is something I did not understand when I read this post yesterday, again after many years. But now it does make sense: I have actually been able to set my Marvell controler to RAID, but my attempted flash must have screwed this up and I left it at the state that it is today.
Btw it is 100% sure that I have an SE9128; that's what the official ASRock specs say (link), SATA3 RAID feature was advertized for this mobo, too and addtitionally I checked the text on the chip itself. I dont think my PATA gets controlled by Marvell EDIT: It actually does get controlled by the Marvell chip, but I dont use it anyways so no big deal.

Think I will try to flash the firmware then.

Btw do you know what the comment under the download on station-drivers means:

Zitat

---

When using on 9128 flashing block 4 gets error wrong device id requests 9123.

---

Seems somebody had an issue flashing it onto an 9128??

That report is missing a few details, like what type of controller (onboard/external), the way it determined it truly was 9128, the preparations for flashing. I'm guessing the card must be in AHCI mode when flashing, or that the user had different conditions then what he suspected. You can go <here> and <here> to see that the firmware worked on an external card, which supports my first claim that this latest Autoload is compatible with both onboard and external controllers. Then you can start <here> and read the next page, especially <one>, <two> and <three>. Please note that around that time I wasn't yet sure about the ID change. Then there is the report from Gaudi, the original tester of this new firmware and also the provider. I'm sure you can also find other reports, if you search for tags "9128" or "2.3.0.1007". What I'm trying to say is that this firmware has been tested before.

Just have your Marvell controller in AHCI mode (which is already acomplished) and flash it with "go" and confirmations or "go -y" without confirmations. A full shutdown is needed. Test the result against your expectations, also test the IDE connector, in case it was something specific to Gaudi's board. Please provide feedback of your tests, others will benefit from them.

Hi Lordkag, it seems the flash has not gone very well. The proccess ended with the message "Flash successful!!" and without any errors, however I am not able to enter the Marvel configuration utility through BIOS anymore (when I try to enter it the computer just restarts). Fortunately the controller is still alive and recognized by Windows - well apart from the PATA ports which indeed seem to be dead now. Marvell... what is wrong with you.

My MSU utility now reports:

BIOS Version 	        1.0.0.1038
Firmware Version 	2.3.0.1007
Boot Loader Version 	1.0.1.0002
Driver Version 	        1.2.0.1047
Chip Revision ID 	B1
Vendor ID 	        1B4B
Sub Vendor ID 	        1849
Device ID 	        9123
Sub Device ID 	        9123
Port Count 	        2
Max PCIe Speed 	        5Gb/s
Current PCIe Speed 	5Gb/s
Max PCIe Link 	        1X
Current PCIe Link 	1X
Supported RAID Modes 	RAID0, RAID1
Supported Stripe Size 	32K 64K
Maximum PD per Adapter 	6
Supported VD Cache 	None
Maximum VD per Adapter 	2
Maximum PD per VD 	2
Maximum Port Multiplier 1
 

Apart from the changed version numbers of course, I have noticed one additional change: Maximum PD per Adapter (PD = Physical Devices btw) is now at the value "6" while it has been at "16" before. Any idea why?

I have also uploaded the current dump of my firmware. Maybe it helps in analysing.

Thanks in advance and merry Christmas! :)
Nuc!eoN

Zitat von lordkag im Beitrag #250

---

I am running out of ideas and you should definitely backup the firmware from your new card, for your own personal safety, but also for my research. Only thing I can think of is that your card needs a special Autoload or a special Loader, which I'm unable to find in my collection or elsewhere. Sadly, no one is backing up those firmwares, when they ask for help it is usually too late, the original firmware is gone.

---

@Nuc!eoN

Do you mean that Marvell controller is not accessible anymore from within main BIOS, Ctrl+M, or both? You should reset BIOS to defaults and maybe also clear CMOS, even disable Marvell controller for a while - if you have that option. Since this firmware update has worked for others, I suspect there is a communication problem with your mainboard. The PATA problem mirrors the bug Gaudi encountered with his board, which could point to Marvell or ASRock being responsible for this one. I'm not sure what Gaudi selected in the end, but you can read his posts to decide. The field "Maximum PD per Adapter" was already reported <here>. Either Marvell was too optimistic in the first place, or it is a typo in one of the cases. Other than that, you still have some leftovers from previous firmware. If the configuration bug is not solved with toggling settings and resetting, you can try to erase the chip before flashing. This can be done with "go -e", then immediately after success flash with "go". DON'T reboot in between, the controller might disappear for good.



@pezku

The construction seems to be the same as with the image from SD, but there is a subtle difference in Autoload. Having a second sample for comparison, it would seem that inner Firmware and possible Loader is not needed for non-RAID controllers, at least for 92xx series. What I can say for sure is that Autoload is needed and is a critical component. Beyond that, Marvell's mercy!



Although the Autoload is at the same version as with the images I have uploaded, there are two fields missing, one of them containing the ID of the controller. I'm not sure why they were removed (your card has generic 1B4B-9235 ID, it shouldn't have made a difference for the second field with ID, probably only the first field is the key), why they use the same version for different Autoloads. I don't think it has something to do with onboard/external, because I have samples from both sides, they are identical. Yet another Marvell mystery.

Zitat von lordkag im Beitrag #261

---

Do you mean that Marvell controller is not accessible anymore from within main BIOS, Ctrl+M, or both?

---

Zitat von lordkag im Beitrag #261

---

You should reset BIOS to defaults and maybe also clear CMOS, even disable Marvell controller for a while - if you have that option.
[...]
Other than that, you still have some leftovers from previous firmware. If the configuration bug is not solved with toggling settings and resetting, you can try to erase the chip before flashing. This can be done with "go -e", then immediately after success flash with "go". DON'T reboot in between, the controller might disappear for good.

---

Zitat von lordkag im Beitrag #261

---

Since this firmware update has worked for others, I suspect there is a communication problem with your mainboard.

---

The image is cleaned now. I don't know how the junk appeared, as the components have fixed offsets in chip and dump image. Maybe the small image you flashed (or was flashed by OEM) had wrong offsets.



I don't know if I can fix your issue, as it clearly stands on ASRock's shoulders. They have been nice before with other users, even with old boards (unlike other OEMs), maybe you can ask a fix from them. There is also the possibility to flash an older firmware, like 2.2.0.1125 from Station-Drivers. The image is cleaned and appropriate for onboard controllers.

Hi All,


I've SYBA SI-PEX40071 8-port SATA III PCIe External Controller (de facto it is a IO CREST IO-PCE9705-8I - 8-port SATA III PCIe External Controller).

I've tried to flash its Firmware to Marvell MV-92xx SATA 6G Controller Firmware Version 2.3.1063 from station-drivers.com.

The flash program said that everything went OK. After that I've switched OFF and then ON the PC and the controller presence was no more detected. Even by the BIOS. The Controller seems completely DEAD for the PC system!


Any suggestions?


Regards,
mswg

@lordkag Is there a Marvell image for the Marvell flasher that has the ID corrected for the 9128 external card? When I updated the ID changed from 9128 to 9123 would like to change it back. Thanks

@mswg

It says it is based on 88SE9705 (i.e. 9705) non-RAID controller, which I'm pretty sure needs a different firmware. If you don't have a backup somewhere, it is impossible to produce an Autoload from thin air. The inner Firmware is not needed for non-RAID controllers, the ROM is possible not needed (I don't have one either way), the Loader might not make a difference. But the Autoload is crucial.

@lurch228

If you would stop and read the posts that are in this very page and previous one, you would know that the ID change is normal and recommended, you wouldn't ask the same question over and over again in different threads and PM.

hey how yuo know what the difference is in the 9123 or 9128? from what i heard, there isnt a 9128. just a 9123 right? So a 9128 is the driver it is. whatever driver for the 9128 that makes a 9123 thats not one, a 9128 is what a 9128 is, just the driver. So, in the files you have to update for a marvell chip, which one runs 9128? or what does, a 9128 do?

@mrtrishterd

"Master, forgive me for being lazy and not noticing your posts earlier, forgive us, all your slaves for being such putrid slow machines during the holidays! I shall go and destroy myself, I hope the rest of the world will soon follow me into eternal shame." Seriously now, do you think that by registering to a forum you are entitled to an answer, you are entitled to eternal gratitude? If someone has an answer, it will be written. If no answer is given, you still don't have any right or reason to complain, this is not a support tech forum for a product you paid. And I seriously doubt they will put up with such behaviour either.

Do you call <this> or <this> questions? I call them random noise. What is your question in that maze, what does IDE, TRIM, X58 ICH10R, SATA2 and the rest have to do with this topic? The only question related to this topic is "9123 vs 9128 vs ...", which was answered no further <than here>, among the many times it has been brought up. For the rest of your "questions", I simply don't give a damn: they are not part of this topic, they are not related to each other (you just cheated, trying to get as many answers as possible), your attitude has zero respect for this forum and its members.

@lordkag

No, but its easy for you to just sit on a forum and insult people right?

"For the rest of your "questions", I simply don't give a damn: they are not part of this topic, they are not related to each other (you just cheated, trying to get as many answers as possible), your attitude has zero respect for this forum and its members.

you are entitled to eternal gratitude? If someone has an answer, it will be written. If no answer is given, you still don't have any right or reason to complain, this is not a support tech forum for a product you paid. And I seriously doubt they will put up with such behaviour either."

All your comments, quoted. wooooowww man. you can speak of money then insult me and stuff. thats about all this boils down too.
and another thing to add to this, i dont understand anything on this forum. and i came here not knowing anything about anything. there are no leads here only dead ends.

TAKe for example @gaudi topic they had. whatever came of that? they blew it up they done something else. whatever happened? i cant tell. then random other people come in and ask other stuff and you randomnly give them bits of information. Nothing is put togetherable here. I dont know anything, you havent even addressed me. your saying id be kicked out of a paid forum but im not mentioning it, a paid forum but you are. I'm not even on one. I wouldnt pay you to tell me help anyway. I got on the internet to help other people and to help myself. I have put some useful information here, but havent gotten any input. You'll use that to your benefit though, wont you.

@mrtrishterd

Who is the one insulting, Bozo? It took you ten days of brewing just to quote my answer and provide nothing in return, not even a documented case of my rude attitude. I don't understand the "money" reference, has anyone on this forum asked you anything? What I said was "this is not a support tech forum for a product you paid", which means you are not entitled to any answer. This is true for all user supported forums, not just this one. When I write a question on forums such as this one, I don't expect others rushing in to provide an answer; I just wait a few days and move to another if no answer is given, simple as that. Your entire attitude can be summarized by one quote: "i dont understand anything on this forum". You have no valid questions, you have no helpful details, you have no knowledge - all you have is the firm believe that someone here must fix a problem that either you or Marvell caused it.

Do you see anyone else complaining that no help is given to them? Do you see anyone else trying to help them? I put my time and knowledge to assist users, even when the issue is caused by them. What I asked in return was specific details as a starting point (type of card, Marvell controller ID, chip ID, a dump of the original firmware) and further informations (screenshots, Device Manager IDs, behaviour) to help debugging process. What have you offered from those? All I see is: SATA, IDE, AHCI, RAID, SATA, slower write speed, Intel, SATA, IDE, "Y U no help? Waah! Waah!", SATA, IDE, SATA, SATA, SATA... Contrary to what you believe, if you browse the two Marvell threads, you will see quite a few people who received help for their Marvell issues, in the limits set by Marvell's quality and design. You know what is the difference between them and you? It's your attitude, Bozo! And one last thing: your grammar sucks! If mine is rusty, yours is dead in the water.

^^^ Now that's what I would call insulting!

@Fernando

I don't have enough patience and resources to help users who expect an answer like it is their God given right. I simply refuse to help someone with the above attitude. If someone has a magic wand to solve all their problems, so be it. But I am done with this user, I don't have the time to argue with "knows it none, expects it all" type of users.