F120 SSD Random Kernel-Power 41 BSOD; Not Detected on Boot

[biatche]

Critical 2010-07-08 9:02:00 AM Kernel-Power 41 (63)

Critical 2010-07-08 8:22:35 PM Kernel-Power 41 (63)

Critical 2010-07-09 8:26:56 AM Kernel-Power 41 (63)

Critical 2010-07-20 10:52:00 AM Kernel-Power 41 (63)

Critical 2010-07-24 11:06:18 PM Kernel-Power 41 (63)

Critical 2010-07-26 1:56:40 PM Kernel-Power 41 (63)

 

OK Fine. I'm experiencing the same issue as the topic starter. (Do you need this in my first post? I mean we're facing the same thing) It usually happens during idle.. but I've seen it happen in front of me with my own eyes.

 

HDD missing after "auto restart" remains a mystery. Windows disconnected the drive?

http://forum.corsair.com/v3/showthread.php?t=88212

 

Is there something that we're doing wrong because we've tried a lot of stuff and I'm beginning to get the sensation that it's being made that the issue is not Corsair and rather something else. With sleep, hibernation and hybrid sleep all disabled, I don't see how suspend to ram or disk actually helps.

 

Sorry if you think this is all about spreading rumors. I'm looking for a solution - FAST.

 

There are about 10-15 threads regarding this, and you feel that this issue should be handled individually??

 

I'll play your game and stick to topic from now.

[Wired]

there's an updated Intel RST 9.6.4.1002 floating on Google (not yet on the Intel website).

http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=18667&lang=eng

[RAM GUY]

biatche

This is no game to us or any one else so please don't treat it that way.

We are looking into this and if there is any more information we will provide it ASAP.

[dikai_yang]

That looks to be a Windows issue.

Please see KB156669

 

When this only happens on the Corsair SSD and not my mechanical HDD from the same disk image, I think this is a Corsair issue indeed. I am still convinced that the Corsair SSD controller stops communicating to the mobo controller randomly and thus causing the BSOD. I just don't see why if it's only a Windows problem, it doesn't happen on a real hard drive.

[RAM GUY]

It may be but when exactly does that happen and do you have suspend to RAM enabled?

[wqvong]

http://downloadcenter.intel.com/Detail_Desc.aspx?agr=Y&DwnldID=18667&lang=eng

 

 

 

Tried those drivers still the same problem.

[dikai_yang]

Hibernation (suspend to disk) disabled

Sleep (suspend to ram) disabled

computer randomly restarts. The longest my computer has run is about a day and a half. The shortest a couple of hours. It's really random. Usually when the computer is idle. I did check my task scheduler on active scheduled events. Nothing is scheduled close to the time of crash (crash time recorded as per event viewer).

 

When I talked about the sleep / wake problem it's a separate issue which does not concern me a lot since I don't mind (too much) disabling sleep.

 

If you do tech support for Corsair and you've been reading other posts on this forum I guess you can understand that there are people who are having the same issue and that we are rather frustrated. The SSD is VERY FAST. But with the randomreatart problem it's a big show stopper.

[biatche]

RAM_GUY: I and many here aren't playing games nor trying to spread rumors. We're like babies trying to tell Corsair mommy we are hungry or unhappy or whatever. It's quite difficult to explain, no less can we be understood. Just trying to fix problems. Not trying to outsmart anybody here but this is not a sleep issue as I've said. dikai_yang said he cloned it to a mechanical disk and it doesn't happen. That already says so much. Even so, sleep and such should be working properly but how many times have we said sleep, hibernation and all that is disabled.

 

You realize there are F-Series topics here about 10 of them all together and we're still making a terrible mistake somewhere?

 

Take a look at my join date. I've been loyal to Corsair for a very long time now. I sell a lot of Corsair products to my friends and clients and I'm thought to be playing games and spreading rumors. Very infuriating.

[RAM GUY]

No worries biatche we are looking into the issues diligently. And I am sorry if I offended you that was not my intent at all! :o:

 

Please post a screen shot from Crystal Disk Info on your drive in your thread.

[Xenn0X]

have same issue, drive is a week old f120 sandforce

 

Windows 7 X64, tried rst and ahci drive but almost everyday my system fails with an 0x000000F4 bsod.

kinda dissapointed in this ssd and that there is no solution. paid 300 euro for it and cannot even run for a day.

have an *2 sandforce too and that one works just fine so gotta be a firmware issue..

[biatche]

what sandforce do you have that works without issues? Mind telling me

[Xenn0X]

what sandforce do you have that works without issues? Mind telling me

 

it's blocked by this forum but and V3rtex2 120 gb sandforce.

if i use that ssd in my system it runs fine without any issues..

only issue with the v3rtex2 is that i need to have rst drivers installed to use trim. ms ahci does not work.

and still using it for now until either the f120 gets fixed or i might just sell it..

the f120 is now for me 300 euro unreliable paperweight.. use this system for my one man company too.. downtime costs money

[dikai_yang]

RAM_GUY + biatche and everyone having this problem, I think I have identified the possible cause of our BSOD and a possible (temporary) solution:

 

Upon reading much on the support forum on a competitor's site (another SF-1200 based SSD), I learned that the SF-1200 has trouble working the TRIM command with the MS AHCI driver. And blah blah blah. This is interesting. Since I don't have any issue with the same disk image running on a HDD, now that I think of it, other than the different way read/writes are done, the additional difference is that SSD's TRIM (i.e. additional operation not applicable to HDD's) (requires Windows 7 + compatible driver such as MS AHCI or Intel RST 9.6).

 

I installed Intel Matrix Storage Manager driver (IMSM 8.9.0.1023), which does NOT support trim, the random restart problem seems to have gone away (system up for 11 hours right now). And even better, this driver has no sleep/wake issue like the Intel RST 9.6. And on the Asus P6X58D-E site, this version is the listed driver for the Intel SATA controller and NOT the latest 9.6.

 

Try disable TRIM and see if it works. If it does, then it IS a firmware issue and apparently Sandforce knows about it, not sure about Corsair. So, Corsair should offer some kind of a manual-trim utility like the Intel's equivalent of the SSD toolbox so we can TRIM manually. For me, stability is more important than performance. I want something that works not something that works well but only sometimes.

 

Note, reading more on the IMSM 8.9.0.1023, it has some known issues on dropping drives. It hasn't happened to me yet. So I will probably also try using the MS AHCI with TRIM disabled (follow the sticky thread on how to enable/disable TRIM on this forum).

[dikai_yang]

Xenn0X: I posted the previous message before I saw your post. But yeah... the TRIM issue with *that* controller is what I'm talking about.

 

I guess when improper TRIM commands are sent to the SSD controller, the controller might actually drop the drive and that's probably why our SSD's aren't detected following the random restart.

 

This discovery explains a lot of things...

[Xenn0X]

RAM_GUY + biatche and everyone having this problem, I think I have identified the possible cause of our BSOD and a possible (temporary) solution:

 

Upon reading much on the support forum on a competitor's site (another SF-1200 based SSD), I learned that the SF-1200 has trouble working the TRIM command with the MS AHCI driver. And blah blah blah. This is interesting. Since I don't have any issue with the same disk image running on a HDD, now that I think of it, other than the different way read/writes are done, the additional difference is that SSD's TRIM (i.e. additional operation not applicable to HDD's) (requires Windows 7 + compatible driver such as MS AHCI or Intel RST 9.6).

 

I installed Intel Matrix Storage Manager driver (IMSM 8.9.0.1023), which does NOT support trim, the random restart problem seems to have gone away (system up for 11 hours right now). And even better, this driver has no sleep/wake issue like the Intel RST 9.6. And on the Asus P6X58D-E site, this version is the listed driver for the Intel SATA controller and NOT the latest 9.6.

 

Try disable TRIM and see if it works. If it does, then it IS a firmware issue and apparently Sandforce knows about it, not sure about Corsair. So, Corsair should offer some kind of a manual-trim utility like the Intel's equivalent of the SSD toolbox so we can TRIM manually. For me, stability is more important than performance. I want something that works not something that works well but only sometimes.

 

Note, reading more on the IMSM 8.9.0.1023, it has some known issues on dropping drives. It hasn't happened to me yet. So I will probably also try using the MS AHCI with TRIM disabled (follow the sticky thread on how to enable/disable TRIM on this forum).

 

just before u posted i had the same thought. will test with trim disabled.

or i might just hook it up on a JMicron® 363 sata port i have on my msi mobo and test if it happens there too (does not support trim so then it should not trim at all).

still think that it's an firmware issue on passing trough the trim command on ahci and rst drivers. cause it works fine on same kind of sandforce v3rtex drive, same chip.

so F120 gets request to trim and boom fails.

 

Xenn0X: I posted the previous message before I saw your post. But yeah... the TRIM issue with *that* controller is what I'm talking about.

 

I guess when improper TRIM commands are sent to the SSD controller, the controller might actually drop the drive and that's probably why our SSD's aren't detected following the random restart.

 

This discovery explains a lot of things...

that's what i'm thinking too. improper trim usage on F120 sandforce.

i would really like a firmware fix for this cause i have no problems at all with other corsair products but this problem is a big boo boo..

the possibility for data loss or even the drive might get defective by this.

[dikai_yang]

Yeah... try installing IMSM 8.9.0.1023. I think using the set deletenotify 1 might not work on the MS AHCI driver, TRIM might not get disabled.

 

It's making sense, the random restart typically happens when I'm away (i.e. TRIM works when idle).

 

But what I'm afraid of is I did try my F120 on the Marvell 9xxx controller (probably no TRIM support), and still randomly restarted...

 

Keep our fingers crossed.

[Xenn0X]

Yeah... try installing IMSM 8.9.0.1023. I think using the set deletenotify 1 might not work on the MS AHCI driver, TRIM might not get disabled.

 

It's making sense, the random restart typically happens when I'm away (i.e. TRIM works when idle).

 

But what I'm afraid of is I did try my F120 on the Marvell 9xxx controller (probably no TRIM support), and still randomly restarted...

 

Keep our fingers crossed.

 

 

same here, few hours of inactivity, trim kicks in and boom, system crashes.

most of the time it happens at night when my system idles.

so i think we found the error, now a fix by corsair asap before they gonna get a lot of bricked drives back on rma due this issue. so please check this issue corsair and fix it :)

potential drive killer or data corruption..

 

gonna try the IMSM 8.9.0.1023 drivers first.. thats a sure fix to disable trim. still trim is there for a reason on ssd and should work..

kinda need raid drivers for optimal usage of my other raid 0 config i have in my system. so will install the IMSM 8.9.0.1023 and remove the rst first.

[biatche]

Good work fellas. Things DO make sense now. Haha, you fellas should be given jobs in the support department. I thought most modern marvell controllers support trim, no?

 

Currently testing with trim disabled by using:

fsutil behavior set disabledeletenotify 1

[Xenn0X]

Good work fellas. Things DO make sense now. Haha, you fellas should be given jobs in the support department. I thought most modern marvell controllers support trim, no?

 

Currently testing with trim disabled by using:

fsutil behavior set disabledeletenotify 1

 

same here, if that does not work i'm installing older intel sata drivers.

and if that fails i can use jmicron chip onboard of my mobo.:roll:

[dikai_yang]

Make sure you restart the computer and "fsutil behavior query disabledeletenotify" again to make sure it's set. Because when I did that without restart, the disabledeletenotify stays at 0. I am really hoping that this will work out.

 

If this works, Corsair should officially enter a new sticky thread to inform people of this. Because TRIM = BSOD, no BSOD = no TRIM. People need to know. Then, we need a SSD Toolbox for manual trimming.

 

It'd be pretty funny if we ask some Corsair SSD owner, "do you get random BSOD?", he says "no", then we'll say "Good, because TRIM is not working on your computer".

[biatche]

Mine stays at 1 without restarting. Strange.

[Mummy]

So puting 2 of F60 in Raid0, was good choice? ..as Trim dont work, for Raid array :laughing:

 

Edit; Dont know why, but on firsth w7 install 8.9 drivers, didnt work on sleep mod. Now at secend clean install, they seems to work.

[biatche]

perhaps my experience with raid and black screen of death was coincidentally caused by something else.

[Mummy]

@dikai_yang; If its that trim, where that 10hours + get to do with this?

 

Edit: You worked so hard that it didnt get idel even for coffee breack?

[dikai_yang]

I'm actually at work now. I Every so often vnc connect to my computer remotely to see if it's still running. It's always a good thing to know it has Not restarted on its own.

 

About 8.9.0.1023 not having sleep/wake issue, I can't guarentee that one. I justed tested it a few times and it worked. Because with rst 9.6, it doesn't work not even once. If I do find 8.9.0.1023 bsod on me on wake I'll just set disabledeletenotify to 1 on ms ahci. But that's a secondary issue at this time. I'm still trying to make sure the random restart problem is fixed.

 

If you're using raid on rst 9.6 trim does get sent through. So in theory your comp should bsod eventually. But on 8.9 or ms ahci with trim disabled, it would probably work. But don't take my words for it. We're working on a single ssd set up in this thread.