I/O Error on all my HDD/SSD (possible CPU issue? or Windows latest update...) Need Help.

[Foxiol]

Hello guys, first of all thanks for your patience and attention, hope you read in detail what I did before posting.

PC Specifications:

-Case: Cooler Master HAF 932 Advanced
-CPU: AMD FX8350 @4.0Ghz (stock)
-CPU Cooler: Cooler Master Hyper 412S
-Motherboard: Gigabyte 990FXA-UD5 Rev 3.0 BIOS version FB (old mobo that gave me the issue for the first time: AsRock FX990 Extreme 3)
-Memory (RAM): Kyngston Hyper X 1866mhz 4+4GB (2 slots = 8GB total system RAM)
-GPU: Asus Strix GTX970 4GB
-SSD: AMD Radeon R7 240GB (primary where the OS is installed of course)
-HDD's: Western Digital 1TB / Seagate 350GB (just as a backup)
-Sound Card: Onboard
-PSU (power supply): Corsair RM850 (850W) full modular
-OS: Windows 10Pro/Windows 7 Ultimate 64 bit (made tests in both, and both same results, fully updated)


Introduction:

Well in order to help you guys you need to know that I had issues in the past that you can check in this thread (almost 2 years ago): https://www.techpowerup.com/forums/...problem-need-help-solved.218791/#post-3393626. In resume, after I upgraded my GPU the CPU was throttling because of the bad heat disipation with the VRM's in that mobo (AsRock 990FX Extreme 3), in order to fix it I had to undervolt my CPU (from 1.380 vCore down to 1.275vCore) so the temperatures stay in place. Problem solved and a happy end. PC working fine up until the end of 2017, some issues here and there, upgraded from Windows 7 to 10, had some driver issues with an Asus Xonar DG sound card (it died literally) and then had some minor errors like Windows rebooting for no apparent reason and such. Things that you can live with if hardware seems to hold and work (mostly in what I do which is gaming a lot).


Issue:

The main issue as of now has something to do with an I/O operation with all of my SSD and HDD's. I think this started to happen once Fall Creators Update from Windows 10 came out (October, 2017), because I was having some weird reboots (Windows stopped working with no sign of an issue, the PC turned off and on alone, and Windows 10 started like nothing happened...also no messages or any other sign of a possible hardware related issue since this happened randomly from time to time even without doing anything on the PC -I mean while not using any program or playing games just to be clear, it just happened randomly-).

10 Days ago the issue got worse as it seems, I had a reboot and after Windows 10 started I got a message telling me that there were issues with the drive unit (I'm having one now as I'm writing this) so I had to check the unit for errors. I did the test, nothing was found, unit was okay. (or it seems)

After that the thing got worse, Windows started to block my drives as it seems or something is bad with either drivers, controllers or the CPU itself...and let me tell you why so pay attention to the things and tests I did in the past week.


Things I did to try to fix the issue (programs used and CMD):

I always try to find a solution with what I know until my brain melts literally, then see if the same issue is on the web (or similar issues) and try things.

Used programs to test unit disks such as:
-HD Tune Pro (benchmark SSD and HDDs): this program has some tests, it passes all of them (even the Benchmark) but the one that is giving an I/O error is the "File Benchmark" one. Sometimes it let me complete that test but once it fails, you can't even start the test again. Sometimes if the program is restarted it may work but then it fails again.

HD Tune Pro I/O Error in "File Benchmark" test (this is from now...the image below was taken one minute before this one):

Image when the SSD is working and not giving I/O errors (this is from now):

-CrystaldiskInfo: which shows no errors so far.

-SeaTools (for Seagate drives, yet it works with others): Tried all my 3 drives, both HDD's passed all the tests but the SSD couldn't finish the DST Short test but passed the SMART and the other 3 tests. The HDD's passed and one is a SATA 2 1TB (WD) and the other an old SATA 350GB (SeaGate) that is probably 12 years old as of now and still working.

SeaTools Basic Test 2 Short DST:

Now have in mind the things I had to do in order to do tests. Since Windows 10 won't start, neither recover, neither let me do anything other than installing it again from scratch (at this point I did it close to 15 times, and in different units). At first I thought that the AsRock motherboard was death (either SATA controller, BIOS corrupted, etc) so I finally bought a new one that came 4 days ago (solved VRM issues, RAM worked properly at 1866mhz with all 4 modules, etc).
But I still could do some tests with the AsRock.

Things get weird from here...With the new mothearboard (Gigabyte 990FXA-UD5) at first had no issues but after a day or so (downloaded a couple of games during the night, with Windows fully updated and no signs of an issue) started to play, games worked fine and all. Stopped gaming and told to myself "why not reboot the PC since it was working for almost 2 days straight?", just did it and the probelm appeared again. Windows won't start, it showed me the repair start and such but it won't let me do it. It always fails.

So then with the new motherboard now I know that the old one was not the issue (SATA controller, SATA connections, cables or whatever)...so the AsRock one still works.

With the new motherboard I did all this:

-I removed and changed all SATA cables to reconnect all units again
-tried to boot one unit at a time with a fresh install (tried with the SSD and HDD's one by one, formatting and installing Windows in all of them)
-tried to change SATA 0, 1, 2, 3 connections to see if one of them was "malfunctioning" and tested with all units
-tested RAM (no errors)
-changed RAM slots (all working)
-tried with one RAM (one slot) in single channel (it works)
-tested 2 RAM sticks in dual channel (it works)
-tested each memory in separate and combined (changing positions with the slots in the motherboard to see if one of the slots was malfunctioning), thought BIOS showed all of them working.
-got a new HDD that was never used on this PC, installed Windows in it and at first worked and then the same issue appeared after upgrading Windows 10 to the very latest build. Weird isn't it?
-tested the units in AHCI, ATA and RAID modes (the ones that worked with these configurations), no change.
-changed power cables for the drives from the modular PSU just in case the PSU wasn't giving them enough power. (Cooler Master RM850 Gold Modular)
-checked all connections in the motherboard more than once
-checked all BIOS configurations that I can think of to try to fix the issue
-Since I saw some weird partitions in the disk I did a DiskPart, deleted partitions, formated disk and even deleted the software (from the SSD) to do a full fresh Windows install in case of a Virus.

Partitions look like this now in the SSD (I think now is how it should appear, Fresh Windows 10 Install with all updates up to date):

Facts:

-Motherboard was not the main issue since I got a new one 4 days ago and the problem seemed to carried on with the drives (all of them and the new one that wasn't plugged into the previous motherboard before).
-RAM is not the issue and all are working properly since I did MEMTEST and checked one by one for errors. All 4 slots are working at 1866mhz.
-SSD had some weird partitions set as "Active", sometimes it showed like there was more than one disk. For example Disk 0, Partition 1, Partition 2 and then Disk 0 was C: but one of those partitions was E: or F:. This was shown even with a fresh install until I deleted the entire SSD and partitions with DiskPart through console.
-HDD's seems to have the same issue with partitions.


Thoughts:

-It can be either a Virus that carried on with the drives and then went straight into the BIOS of the motherboard/s. That's why I'm getting the same error with the new motherboard?
-Or can it be a Windows 10 Upgrade that likes to mess around with not so new hardware so you have to upgrade your god damn PC? Tinfoil-hat on...
-Or the last thing that the only way to know is to change it...the CPU (AMD FX8350) is malfunctioning giving this awful I/O operation error?

Just in case also tested the CPU with CPU-Z Benchmark and this is the result and did a stress test for 10 minutes as well. It is working better than expected, no throttling, all linear all the time:

Important to read as well since this also can give you a sign of what's goin on:

Something that I almost forgot, I tested with Windows 7 Ultimate as well. At first it worked fine but then after a reboot the screen went black and could only move the mouse but Windows was death. Impossible to recover Windows, same as happened with Windows 10. It is like I can update Windows but up to some point it says "enough" and done.

And the weirdest of things...I told to my father to use his old PC that he uses with a couple of not demanding programs, he is running Windows 7 Ultimate. So I asked him to use HD Tune Pro to see if he had I/O operation issues while doing the "File benchmark" test with its Kingstone SSD so then I could try with my SSD.
His SSD worked fine, he did like 20 tests and all fine, no I/O operations of course. Then he did like 10 tests with my SSD and no I/O issues at all... Mind blown and not at the same time.
But after he unplugged my SSD and turn his PC on again, it failed to boot Windows with a blue screen. He had to recover (he could do it) and then Windows 7 was working but some weird partitions appeared on both of his disks. He had Disk 0 and Disk 1 with 2 partitions each, Disk 0 as C: and D:, and Disk 1 as E: and F:.
I had to eliminate the D: and F: letters from the partitions so Windows won't mess the boot after turning off and on the PC. Isn't this weird enough? Check this out I have more...

Now my father's PC can't upgrade to the latest Windows 7 update which is from the 4th of January 2017. That "security" update once it installs and all, when Windows is trying to boot there is a blue screen and he has to recover Windows 7 again. I told him to hide that upgrade and now his PC is working fine. Tinfoil-hat engaged...or my SSD is a living virus spreading its "love" wherever he goes...


One thing left to do or 2 as of now:

Now I have something more to do but I'm not sure if doing it since the motherboard is new. Shall I upgrade the BIOS from FB to FC? The thing is that I don't like to mess so much with the BIOS since everything is working properly at the moment. (aside from this weird thing with my SSD/HDD)
But if I introduced something with the SSD/HDD's into the BIOS perhaps it could be a "solution" to start with a fresh BIOS again.

The other thing is to get a new CPU...but I'm not quite sure that this is a 100% hardware related issue at the moment...This has to be software.


If you came to this point thanks for the reading...hope you had some fun and have your mind as blown as mine with this...never seen this happen before. I have PC's since early 90's and had lots of components...you know when one failed and then just changed it and problem solved. Now it seems that problems can come from everywhere or could be generated somewhere (tinfoil-hat on).

Of course I accept suggestions, hope the information given is enough, I may be missing something but I'm not sure if Windows is going to hold since this is my 4th try to write this thread (Yeah Windows failed on me 3 times before because of this I/O thing, now it seems stable but I/O is still being detected and I have a message to recover the disk "Reboot to repair unit". I guess if I do that Windows won't be fixed anymore and I have to install it again...

Happy 2018 to everyone, not such a good start to me but at least I'm learning some new stuff here.


PS: Disclaimer, if I made some mistakes or explained like sh*t, English is not my native language.

 

[TheoneandonlyMrK]

So I've a similar pc and last night i swapped oc profile between 4core mining setup and oc 8 core gaming ,all on the latest patch d windows, i had many many many failed reboots as it will not run my memory beyond 1333 atm odd, ive put it down to chip ageing of the Cpus imc, yours like mine has four dimm slotts full, AMD's spec rates our cpu as only capable of 1333 max speed with four slots filled , i could get 1866 working on a prior 8350 i still own but it gave Os and drive errors over time.
Something to consider might be lowering its speed plus it's not giving much FPS in game's anyway i assure you I've tested this at length.
Secondly while I was in reboot hell i noticed variances between the CSM settings , basically the secure uefi boot module was failing due to a changed setting between the two profiles , your settings might have changed between boots requiring the Os to load drivers different or possibly to request drivers or uefi certificates that aren't there so check those settings.
These uefi things can be a pain I've had my gpu reinstall mid game a few times for god knows why.

Third a normal windows Os makes three i think partitions on the C drive, your main c drive, a volume info partition and another im less sure about, it needs all these and damaging one will hurt the system, seams like yours somehow got labelled, they usually aren't but you then deleted them ,im not sure that was wise it might require a restore or reinstall.
Few things to think about

I got mine booting fine by wiping the bios clcmoss style(15minute clear) then loading and adjusting the last known working profile (4core mode and the last it worked on) back up to an oc, though its not a very good oc, ill be messing further but i was eager to game at the time.

 

[Foxiol]

So I've a similar pc and last night i swapped oc profile between 4core mining setup and oc 8 core gaming ,all on the latest patch d windows, i had many many many failed reboots as it will not run my memory beyond 1333 atm odd, ive put it down to chip ageing of the Cpus imc, yours like mine has four dimm slotts full, AMD's spec rates our cpu as only capable of 1333 max speed with four slots filled , i could get 1866 working on a prior 8350 i still own but it gave Os and drive errors over time.
Something to consider might be lowering its speed plus it's not giving much FPS in game's anyway i assure you I've tested this at length.
Secondly while I was in reboot hell i noticed variances between the CSM settings , basically the secure uefi boot module was failing due to a changed setting between the two profiles , your settings might have changed between boots requiring the Os to load drivers different or possibly to request drivers or uefi certificates that aren't there so check those settings.
These uefi things can be a pain I've had my gpu reinstall mid game a few times for god knows why.

Third a normal windows Os makes three i think partitions on the C drive, your main c drive, a volume info partition and another im less sure about, it needs all these and damaging one will hurt the system, seams like yours somehow got labelled, they usually aren't but you then deleted them ,im not sure that was wise it might require a restore or reinstall.
Few things to think about

I got mine booting fine by wiping the bios clcmoss style(15minute clear) then loading and adjusting the last known working profile (4core mode and the last it worked on) back up to an oc, though its not a very good oc, ill be messing further but i was eager to game at the time.

Going to try lowering memories down to 1333mhz and see what happens. With my previous mobo (AsRock 990FX Extreme 3) I could make all 16GB run at 1600mhz, impossible to run them at 1866mhz but with this mobo (Gigabyte 990FXA-UD5) all 16GB are working without issues. (aside from this mess of course)

If you are right and the FX8350 controller can't handle those RAM speeds, I hope is not too late. But if it is at least I know that it can be the CPU as it is the only physical thing that could be causing this issue.

Hope I can reboot safely to do this because I'm affraid this is about to fail on me again.

Going to wait for a couople more comments and then try some things. Better wait while this works so I can see some possible solutions. Yours is the first I'm gonna try.

Thanks in advance mate.

 

[Vya Domus]

I consider it unlikely that the CPU is the cause of this.

 

[TheoneandonlyMrK]

I consider it unlikely that the CPU is the cause of this.

Well I gave two other options and given i experienced similar corruption once the core heated up the Imc on mine(not once i have tried literally hundreds of different Ocs, higher memory,Nb ,Ht ,core or variances of them on two chips 3x mobo, crosshairV and gigabyte 990x and ud3 ) it's possible and certainly the easiest to check , thats why it's first but hell i wouldn't put money on it.

 

[Foxiol]

Set the memory to 1333mhz, of course it is working and still getting I/O errors (not surprised). Not sure if reinstall again Windows with the memories like this just to see what happens or update the BIOS first with Windows now running (at the moment) and also see if something changes.

But I think the damage was done and the CPU might be the cause at this point.

I'll update once I do some more tests.

Quick Edit: I rebooted and let Windows search and fix my disk for errors and this time it let it happen. At least it went up to 100% and then Windows launched again. Going to try some more stuff.

 

[Vya Domus]

Well I gave two other options and given i experienced similar corruption once the core heated up the Imc on mine(not once i have tried literally hundreds of different Ocs, higher memory,Nb ,Ht ,core or variances of them on two chips 3x mobo, crosshairV and gigabyte 990x and ud3 ) it's possible and certainly the easiest to check , thats why it's first but hell i wouldn't put money on it.

Set the memory to 1333mhz, of course it is working and still getting I/O errors (not surprised). Not sure if reinstall again Windows with the memories like this just to see what happens or update the BIOS first with Windows now running (at the moment) and also see if something changes.

But I think the damage was done and the CPU might be the cause at this point.

I'll update once I do some more tests.

Well , I use three 4 sticks of RAM on my 6300 and while OC'ed to 4.5 Ghz I can still push the RAM from default 1600 to 1866 (though not 100% stable sometimes I get system freezes) so that puts a fair amount of stress on the IMC and things are still fine.

I just find this sort of thing very unlikely. Keep in mind Windows does all sort of weird shit with regards to storage and drives. Once both of my drives would get corrupted and fail no matter what I did , turned out my HDD was dying and every time I would install Windows for some fucked up reason it would decide to make the system reserved partition on the HDD even though the OS was to be installed on my SSD. And so because the HDD was dying it would also affect that partition and the whole thing would go to shit every time.

 

[TheoneandonlyMrK]

Set the memory to 1333mhz, of course it is working and still getting I/O errors (not surprised). Not sure if reinstall again Windows with the memories like this just to see what happens or update the BIOS first with Windows now running (at the moment) and also see if something changes.

But I think the damage was done and the CPU might be the cause at this point.

I'll update once I do some more tests.

Many more things to try first mate if the io errors remain then it's possible other things are impacting it , have you tried new sata cables on varied ports, are all the ports full if so and the boards old is the southbridge tim dried out, or has it cracked, twice while fitting boards ive accidentally fecked the Tim of a board heatsink(including the southbridge one) through man handelling it ,re timming fixed that, plus the partition thing is relevant ,why did it label the partitions, seams like some form of sata drivers swap caused the drives to be re aquired and re listed somehow ,in part that's why i thought memory could cause it, though as i said i have had that happen.

Well , I use three 4 sticks of RAM on my 6300 and while OC'ed to 4.5 Ghz I can still push the RAM from default 1600 to 1866 (though not 100% stable sometimes I get system freezes) so that puts a fair amount of stress on the IMC and things are still fine.

I just find this sort of thing very unlikely. Keep in mind Windows does all sort of weird shit with regards to storage and drives. Once both of my drives would get corrupted and fail no matter what I did , turned out my HDD was dying and every time I would install Windows for some fucked up reason it would decide to make the system reserved partition on the HDD even though the OS was to be installed on my SSD. And so because the HDD was dying it would also affect that partition and the whole thing would go to shit every time.

@Vya Domus chips vary in all areas and with time i assure you, my old 8350(lives as a miner now, i might swap back this chips crap and not lapped) once did 5.5ghz/2600Nb-Ht and 1866 at the same time, but it aged as will yours and my replacement but they're hardy at stock clocks and all still work in spec so i can't complain, mine were all used to fold or crunch heavily loaded 24/7 for years though so i aged them young

 

[Foxiol]

Many more things to try first mate if the io errors remain then it's possible other things are impacting it , have you tried new sata cables on varied ports, are all the ports full if so and the boards old is the southbridge tim dried out, or has it cracked, twice while fitting boards ive accidentally fecked the Tim of a board heatsink(including the southbridge one) through man handelling it ,re timming fixed that, plus the partition thing is relevant ,why did it label the partitions, seams like some form of sata drivers swap caused the drives to be re aquired and re listed somehow ,in part that's why i thought memory could cause it, though as i said i have had that happen.

Yes, I am using new SATA cables, the motherboard is new (remember the issues started on my old motherboard so I bought the new one this very same week). The only things that remain are the memory sticks, CPU, GPU and the HHD's/SSD.

I even changed the cable managment from my modular PSU just in case.

This problem was carried on either by the CPU, or the drives since once I tested my SSD on the other PC even though it had no I/O error, then that PC started to have issues as I said in the main post. That's why I suspect it could also be a virus, a corruption inside the drive causing a mess with the BIOS in the motherboard...or Windows (the other virus).

I don't know.

 

[TheoneandonlyMrK]

Yes, I am using new SATA cables, the motherboard is new (remember the issues started on my old motherboard so I bought the new one this very same week). The only things that remain are the memory sticks, CPU, GPU and the HHD's/SSD.

I even changed the cable managment from my modular PSU just in case.

This problem was carried on either by the CPU, or the drives since once I tested my SSD on the other PC even though it had no I/O error, then that PC started to have issues as I said in the main post. That's why I suspect it could also be a virus, a corruption inside the drive causing a mess with the BIOS in the motherboard...or Windows (the other virus).

I don't know.

I personally think it was a slught drive corruption on your dads due to a slught config issue or something like ahci on off but the uefi version of that issue and a unlucky one off, for a virus to jump from a drive to a motherboard bios realm or some of its firmware is isuper high level hacking and uncommon if even possible mate , i think your letting the gremlins get to you , treat each issue as an item in a list and not a definition of the list ,your drive showed no errors on his you said so it would seam likely to be your systems use of it thats odd atm.

Is it behaving ok at lower memory clocks??

 

[Foxiol]

I personally think it was a slught drive corruption on your dads due to a slught config issue or something like ahci on off but the uefi version of that issue and a unlucky one off, for a virus to jump from a drive to a motherboard bios realm or some of its firmware is isuper high level hacking and uncommon if even possible mate , i think your letting the gremlins get to you , treat each issue as an item in a list and not a definition of the list ,your drive showed no errors on his you said so it would seam likely to be your systems use of it thats odd atm.

Is it behaving ok at lower memory clocks??

At the moment it is ok, the same as with higher clocks. But I'm not doing too much of a thing as of now either aside from trying to stress the hard drive until it fails. I'm not surprised that Windows is working now, it may fail after downloading a file, play a game for a couple of hours, installing updates, etc. The system stops working without any sign, it happens randomly as far as I can tell.
After the full software wipe I did to the SSD a few hours ago, it seems more stable since partitions look right. I/O issues still with HD Tune Pro but it needs to fail while doing regular stuff not just a benchmark. It is so random that a couple of times happened while writing this thread and other times after a couple of days of work. (I installed Windows 10, 3 or 4 times today alone)

 

[TheoneandonlyMrK]

At the moment it is ok, the same as with higher clocks. But I'm not doing too much of a thing as of now either aside from trying to stress the hard drive until it fails. I'm not surprised that Windows is working now, it may fail after downloading a file, play a game for a couple of hours, installing updates, etc. The system stops working without any sign, it happens randomly as far as I can tell.
After the full software wipe I did to the SSD a few hours ago, it seems more stable since partitions look right. I/O issues still with HD Tune Pro but it needs to fail while doing regular stuff not just a benchmark. It is so random that a couple of times happened while writing this thread and other times after a couple of days of work. (I installed Windows 10, 3 or 4 times today alone)

Well good luck mate keep us in the loop though as these threads can help those who follow.

 

[Regeneration]

Sounds like RAM or Northbridge problem to me. Stress test each component, including CPU, until you figure out the cause.

 

[Sora]

seeing as nobody replied with a factual conclusion to this issue, the issue was HD Tune itself, its fixed in 5.75