Solved Recent Windows Update KB3064209 causes Windows 7 to not boot

[optiplex gx520]

run a system repair in recovery?

 

[jewishnigra]

Ok, I have it working now

I never said quoted text. I got spooked because there was multiple people in here with the same cpu and the problem of windows bricking. I generally hate system restore because that's where maleware can hide and it eats up ssd space. The "winsxs" folder is bad enough on an ssd. The only thing I did was made sure I had the precautions on, installed the update, then everything is fine. Like I said, just spooked because of similar hardware and MS's tendency to never properly explain what an update does -- to any sufficient degree. Wanna know my hidden updates so Win10 nagware never appears? Well here they are:

• kb2952664
• kb2990214
• kb3021917
• kb3050265
• kb3065987
• kb3068708
• any "Windows Malicious Software Removal Tools" because I doubt they're still scanning for worms from 2004 (MS Blaster, Sasser, Mydoom) as claimed in every month's description of the update.

^Some of those might be MS/NSA spyware as well versus simply win10 nagware. I apologize for detracting from how solve your problem of a bricked windows boot. Maybe windows isn't the OS for me.

 

[Layback Bear]

Members are just making suggestion on how to get someone back in working order before the problem with the KB.
Restore point and other methods posted may or may not help. One just has to decide to try the suggestion or not. I personally have lots of restore point that are handy for such problems.


Take a think on this.

When was the last time Microsoft had a update that messed with how a Intel processor did it's job before this one.? The best I can remember they never have.

Why would anybody want Microsoft to be messing with a Intel or AMD cpu in any fashion?
Because Microsoft is pushing Windows 10 they are doing a lot of strange KB's to Windows 7 updates so one must be watchful of all Windows 7 updates.

When I want to investigate my Intel cpu I do it the old fashion way.
I will check with Intel and/or Asus my motherboard manufacture for bios information. Surely not Microsoft. That was my thinking so I didn't install the KB.

Now here comes more guidance from a simple person, ME.

When you see a update that sound a little strange or you don't understand just post and ask a question or read through the threads and see if any other member is having problems or questions with the KB update.

This complete post is just my opinion nothing else. Some will agree and some will not.

Layback Bear

 

[Cetus35]

I never said quoted text. I got spooked because there was multiple people in here with the same cpu and the problem of windows bricking. I generally hate system restore because that's where maleware can hide and it eats up ssd space. The "winsxs" folder is bad enough on an ssd. The only thing I did was made sure I had the precautions on, installed the update, then everything is fine. Like I said, just spooked because of similar hardware and MS's tendency to never properly explain what an update does -- to any sufficient degree. Wanna know my hidden updates so Win10 nagware never appears?^Some of those might be MS/NSA spyware as well versus simply win10 nagware. I apologize for detracting from how solve your problem of a bricked windows boot. Maybe windows isn't the OS for me.

@ jewishnigra

I apologize for misquoting you. It won't happen again.

However, I at least partially stand behind my paraphrasing because although you didn't say word for word "Ok I have it working now" you did say "after installing this update, my system appears to be fine."

On this last post you're talking about the NSA and Win10 nagware.

Are you saying that you installed the problematic update (KB3064209) on Windows 10?

This thread is concerned with the combination of Windows 7, Windows Update KB3064209, and the G3258 Intel Pentium processor causing the system to boot to a black screen. (Although it does in fact successfully boot to the Desktop because you can hear all the normal Windows sounds going on when it does. The screen is just black.)

When I first posted the orginal question concerning this matter I kind of expected someone to say "You have to install KB so-and-so first" and that would be that.

Instead, come to find out that it's a widespread problem that a lot of people are having with this update.

Per Layback Bear, I'm just simply not going to install it again. Like I've said before, I'm just concerned with two matters:
1 - What exactly does Microsoft mean that the operating system may be unreliable without it and,
2 - That a future critical update will fail to install because this one is not installed.

Until then, heck with it.


@Layback Bear
Microsoft has released at least 3 updates/hotfixes regarding processors since 2010, that I know of. I've been scolded for misquoting so I'll put in a disclaimer that you said "updates" and "Intel processors".
1 - KB2493989 - Intel microcode - Windows 7 & Windows Server 2008 R2
2 - KB936357 - Intel microcode - Windows XP, Windows Server 2003, and Vista (all versions)
3 - KB2818604 - AMD microcode - Windows Server 2012, Windows 7 & 8.

I fully agree with what you said about not allowing Microsoft to fiddle around with processor microcode but
here's their apparent take on that issue (https://support.microsoft.com/en-us/kb/936357 bottom of the
page, right above where it says "Properties").
Per Microsoft:
"The processor manufacturer may use microcode updates to address errata and to make sure that operational behavior is correct. Windows operating systems include a microcode update mechanism that supports Intel processors. This mechanism lets the operating system apply the latest microcode updates to all processors in the system.

Operating system updates make it possible to apply microcode updates that are later than the versions in the system BIOS. If the processor uses the most recent microcode version, the Windows microcode update mechanism takes no action."

I'm not 100% sure that I like the idea of that. Microsoft applying microcode newer than the latest BIOS?

If that were a good idea, why do we check our motherboard manufacturers websites all the time for BIOS updates when Microsoft can just do it with Windows Updates?

AND, as all of us here know from personal experience with this KB3064209 microcode update BS, it's NOT a good idea.

As you suggest, the websites of Intel, AMD, and your particular motherboard manufacturer are the legitimate places to check for any "microcode" updates.

If they don't have any and Microsoft is saying "We have one for you" it is just a little suspect, like you mentioned.

Thanks for your input. Valuable information for all.


.

 

[detailer]

I passed on this firstly because there was no mention of specifically which Intel processors the update applied to,and because I never install any updates without searching online first.So,damn good thing I Googled up this little gem,huh?
Now,it might fly for me because I'm running an older Pentium T-4300,but I'm not going there because "If it ain't broke..."

Cheers

 

[jonny88]

Same problem here, I have a MSI H81m E33 with Pentium G3258 overclocked 4.2@ 1.120v and today it wouldn't boot to windows so I was thinking I needed more vcore so I booted to bios raised vcore in like 0.02v increments right upto 1.25v but still not boot to windows, Reset bios to defaults and windows booted no problem, then I remembered yesterday KB3064209 said CPU microcode update, When I removed KB3064209 windows booted with my overclock no problem, so I wont be putting KB3064209 back.
Just adding my 2 cents also
jonny

 

[Tone Ranger]

Fixed on MSI Z97 PC-Mate MOBO

Found a fix, at least on the MSI Z97 PC-Mate MOBO. This board was exhibiting the exact same issue as the original poster - G3258 in a death loop during windows boot after installing the vile KB3064209 update. This one had me stripping out all hardware (though not any hair, thank goodness), even attempting a PCIe card to bypass the on-chip GPU, all to no avail. To make matters worse, I had already done a lot of configuration and installation that I did not want to repeat, but had not gotten around to backing up the system (I know, I know... face-palming).

A couple of bright spots that gave me hope: the system would boot into UEFI BIOS (which is really very nice, BTW), and I was able to boot into the memtest86 utility - which passed with flying colors. At that point I looked and saw that the BIOS was a bit out of date. So as a last-ditch attempt before reformatting, I downloaded the latest recommended BIOS (4.9 - the latest one) onto a USB stick and installed it through the (brilliant) M-Flash utility built right inside the BIOS.

After successfully completing the flash, it rebooted, and Eureka! Windows came back. I now leave you all with the hope of similar success - either with this board or one with a BIOS update available that also has the same sort of fix.

 

[jonny88]

Nice fix

Nice fix, that wouldn't have worked for me because I am already on latest Bios, removing the update will do for me though!

 

[SkrillZilla]

Pentium G3258 Win 7 Boot loop After Update 06-07/15

I figured it out guys!!



so im not sure when the updates came out though it was around this summer 15 and not sure if any one here has figured it out yet sorry no time to read 3pgs of replies..

any ways i have had this issue with 2 systems both with the pentium g3258 and 270x gpu installed some updates after messing with some fans then bam continuous cycle ITS YOUR OVERCLOCK GUYS

disabled my oc and was able to boot into windows and remove and hide the updates that caused the issue and then loaded my OC profile and was up and going, luckily for me (first thing i tried) i figured the OC was keeping it from finishing the install after the restart (possibly unstable) so i took it off, later i found that with the update installed, the pc refused to boot with an OC sad i know lemme know if you guys have any ? or what not like i said i have @ of the pent gs and can run test and what ever hope this helped

 

[Cetus35]

Hi all,

@ Tone Ranger
"Found a fix, at least on the MSI Z97 PC-Mate MOBO. ... After successfully completing the flash [BIOS update], it rebooted, and Eureka! Windows came back."

I thought an updated UEFI BIOS update was what I might need but when I checked, I already had the newest one.

@SkrillZilla
"ITS YOUR OVERCLOCK GUYS ... disabled my oc and was able to boot into windows and remove and hide the updates that caused the issue and then loaded my OC profile and was up and going."

That's the very first thing I thought was causing the problem since OC'ing can cause ALL sorts of weird things to happen if it's not completely stable. SO I removed all overclocking, rebooted, and installed the update again and boom, same problem right back again. So removing my OC didn't help.
In fact, having Windows run right (i.e. not installing that freaking update), I have set all my OC back with no problems whatsoever. (4.30GHz (up from the stock 3.20GHz), 1.125V, with nothing else changed.) I evidently lost the RAM lottery because even though it's name brand with a great reputation, if you even THINK about a RAM OC, it won't boot.

For anyone who is doing like some others and, for TL;DR reasons, skipping straight to the end of these posts without reading what has happened in the meantime, here's what this original post concerns:
If you have a combination of an Intel G3258 Pentium (Anniversary Edition) and Windows 7 and install a Windows Update that came out this June (2015), specifically KB3064209 (Intel Microcode Update), Windows will not boot to the Desktop. It will show a black screen instead which happens immediately after the 'Starting Windows' screen.
Saying it won't boot to the Desktop is kind of a misnomer because it DOES in fact continue to boot to the Desktop, you just can't see it.

Apparently a couple of folks here have gotten their systems to work with it by 1 person turning off all overclocking settings then installing the update then reapplying the OC settings and 1 person by updating their BIOS then applying the update. All others have failed to find a fix.


.

 

[Tone Ranger]

"I thought an updated UEFI BIOS update was what I might need but when I checked, I already had the newest one. ... Apparently a couple of folks here have gotten their systems to work ... 1 person by updating their BIOS then applying the update."

Sorry to see that you are still out of a fix. In the meantime, I thought I'd interject a couple small clarifying points:

The MS update was already in place when I performed the BIOS update. I could not roll the MS update back because I had no restore points, nor had I imaged my drive prior to the MS update. So the MS-KB update was still in place when the BIOS update was applied, after which the system came back up AOK with the MS-KB update still in place.

One way for you to more closely reproduce my scenario would be for you to roll back your BIOS, install the update, then after the Windows boot-crash, re-update your BIOS. Not sure how constructive or effective that would be, unless there is something peculiar about the order of operations. Weird, yes, especially if it works.

 

[jewishnigra]

Hah, I didn't even touch my 4.6GHz OC when I installed this update successfully. Manual vcore at 1.345v that passes a 48-hour prime95 blend validation. PLL/VTT voltages are on auto. "Adaptive" vcore won't even post on my z87 sabertooth (pre-this update). Ram is on XMP just below rated 1866. Other hype settings like "asus multicore enhancement" and "SVID control" are on auto (and are very poorly explained I might add.) LLC=level1, cpu spread spectrum disabled (I don't live underneath a power line,) EIST enabled (because it's summer herpa-derpa), and the rest of the c-states disabled (not gonna deal with increased ssd latency.)

The one downside of owning an Asus board is they simply refuse to release a bios with an "ata password" that would allow for hardware ssd encryption instead of relying on MS's bitlocker which I mean, come on. But that's beside the point here and I'll probably get misquoted again about how this all relates to a bricked boot so I'll stop again.

 

[SteveK1]

I have a G3258 on a Gigabyte H81M-DS2V latest BIOS F5, not overclocking, and optimized defaults loaded, and installing this update breaks Windows 7 Enterprise x64 for me. It reboots while booting (during the Windows splash screen), and then prompts to try a repair or try booting again, the repair is unable to repair startup. I booted to the Windows install DVD and its repair function also failed to repair. Getting back into Windows is possible with a revert to the preupdate restore point, assuming System Protection (aka System Restore) is on.

Another thread had some talk that if you disable core #2 it might work, and the problem is something to do with the this cpu's unlocked multiplier, and how windows throttles the CPU to save power and turbo's (supported) CPUs under load.

 

[gdfsgdfg]

So does this update actually improve anything for my 3570k? otherwise I am ready to remove it.

 

[jdorje]

Great thread.

I've got this problem on windows 7, and worse still it's carried over to windows 10 - since build 10240 - without any recourse.

https://www.reddit.com/r/buildapc/comments/3f3mro/discussion_psa_windows_10_currently_not/

I have a g3258 and gb b85m-ds3h r3. Stock or overclock does not matter. Setting the number of cores to 1 does allow windows (7 and 10) to work; this could allow you to at least get into win7 to try to uninstall the update.

I encourage everyone to report this bug to either microsoft or intel. It's a pretty significant one, but very hard to pin down.

 

[Cetus35]

I knew it!

As you can see from my previous posts I have predicted the the freaking Update KB3064209 update would cause problems in the 'future' (which is now).

I say cause problems in the future because most of us agreed here before that simply not installing that Update was a 'fix' to the problem of it causing a G3258 processor with Windows 7 to boot to a black screen.

NOW, the wonderful, shiny, new, Windows 10 INCLUDES THAT UPDATE BY DEFAULT! It's not even an update you can opt out of, IT'S INTERNAL TO WINDOWS 10 !!!!

Thus, if you try to install Windows 10 on a system with a Pentium G3258 CPU, you will see the much- talked-about 'booting to a black screen'.

During the attempted Windows 10 install, you only see that problem for a second then the install will fail and revert back to the original Windows 7 (which you have hopefully left the Update 3064209 off of, or it won't boot either).

The installation fails with this error:
"We couldn't install Windows 10
0xC1900101 - 0x20017
The installation failed in the SAFE_OS phase with an error during boot operations"

There is much talk on other forums which say that most motherboard companies have known about this for months and have not released an updated BIOS which will reportedly fix the problem. Intel is blaming Microsoft for the problem and Microsoft is blaming Intel, neither of which have issued a fix.

Reportedly, Intel has agreed to "check into it, but it may take weeks."

There are only two 'fixes' that occur with some frequency on the forums. Some say that removing ALL overclocking will allow Windows 10 to install successfully (mine still fails the Windows 10 install using this 'fix', by the way) OR to go into UEFI BIOS and disable 1 core of the CPU.

ALL that have tried this particular bullshit fix say that Windows 10 will install 100% of the time doing this.

Re-enabling the 2nd core or trying to re-apply any overclock will immediately cause Windows 10 to not boot.


Did I not say that simply not installing the ****ing update may be a fix for now but it would later come back to bite us in the ass?

G3258 owners, if you want Windows 10 and you wish to run with more than 1 core of a dual core CPU or to apply an OC to said CPU, you are shit out of luck.




.

 

[ratchet26]

I just want to post that I have encountered the exact same scenario as the OP on one of my kids' PC. G3258 with ASRock H81M, the KB3064209 update causes Windows boot failure. This caused a major failure for me about a month ago, since I run nightly backups on all PC's I have system restore disabled. For some reason the image wouldn't restore and I ended up doing a clean install, leaving restore enabled. Twice I have run this update and both times it requires a system restore to boot again.

I also tried to update to Windows 10 but encountered the "We couldn't install Windows 10
0xC1900101 - 0x20017" installation errror.

I agree with jdorje, we should all be reporting this error. Maybe we could all report via the same channel and post the link here?

 

[jdorje]

GIGABYTE - eSupport

You can also search "g3258" in the "windows feedback" app on win10 to find a number of related tickets. Most likely simply upvoting them all is the right move.

Possible fix/workaround for windows 10...for win7 you're better off just not installing that update though. https://www.reddit.com/r/buildapc/comments/3f3mro/discussion_psa_windows_10_currently_not/ctnuduk

 

[archp2008]

Windows 7 installed this update on my AMD notebook (no Intel CPU). I am having the "install failed in the safe os phase" while attempting Windows 10 updates on Win7. Could this update be the culprit. Will uninstalling the update cause problems with my upgrade. I tried it 3 times manually getting this boot error each time. I have the upgrade reserved now.

 

[cpthomas]

I have the same problem as described, except that when the "throbbing" welcome item appears on the screen, it stays there and does not proceed to the next step. The only way I can shut down is manually.

I am not a tech person, but I did figure out that the KB3064209 update was causing the problem and was able to go back to a point prior to its installation, using the "last successful restart" approach. Interestingly, that worked when using system restore would not.

This is similar to a problem I have had with another update, one related to something that happened with the Russian ruble. I have that update number somewhere. Because it keeps showing up on the update list, I have set my computer to tell me when updates are available, but to install only ones I have selected. Then, I have put that update into "hide" and it stops bothering me. The only problem is that when I do certain maintenance functions, the hiding command gets cleaned out and I have to hide the update again the next time my computer checks for updates. Now, I'm going to have to add KB3064209 to Russian ruble update.

I use Windows 7 Home Premium, with SP1. I update regularly. My processor is: Intel (R) Core (TM) i5-2450M CPU @ 2.50 GHz. Not being a tech person, that means nothing to me, so I don't know if this is the same type of processor others have talked about or not.