backtop


Print E-mail del.icio.us 51 comment(s) - last by KuhnKat.. on May 14 at 1:12 PM

A wide variety of boot problems have been reported with Windows XP SP3, fortunately many of them are very fixable

Windows XP remains a standard throughout much of the IT community, and remains popular among consumers as well.  Thus many consumers are pleased that Windows XP Service Pack 3 is back in action, after being pulled a week for a software fix.  The new service pack provides additional useful features, numerous bugfixes, and minor performance improvements.

Unfortunately some users are also finding that it provides their computer with an endless reboot loop.  First, to dispel a common misconception, the reboot itself has nothing to do with a problem with XP SP3.  Rather, the problem is during the boot, which results in a crash.  In the case of the crash, Windows XP behaves correctly -- it reboots the computer and asks the user if they want to boot into safe mode, defaulting to a normal boot if no option is selected.

Users are not happy about the developments.  Michael Faklis posting on the Windows XP discussion board, vents, "My external disks are having trouble starting up, which results in Windows not starting up.  After three attempts [to install XP SP3] with different configurations each time, System Restore was the only way to get me out of deep s**t."

The source of many of the problems has been traced to manufacturers, which takes a bit of heat off Microsoft.  Perhaps the single biggest problem appears to be caused by Hewlett Packard's configuration choices.  By default HP deploys the driver intelppm.sys on all their computers, including those with AMD processors.  The driver provides power-management, but only for Intel machines.  On AMD machines a second driver, amdk8.sys is also added, which performs the same functionality for AMD processors.

If your machine's HP part number ends in a 'z', you probably have an AMD processor.  Or you could just peek inside.  Either way, if you have an AMD machine, you will likely experience crashes when you install Windows XP SP3.  This is really not Microsoft's fault as HP is installing an unsupported configuration by adding both drivers.  To fix the problem, if you haven't added SP3 yet, just type "disable intelppm" in the command prompt or run "sc config intelppm start= disabled" if you already installed it and can only boot to safe mode.

Another problem seems to occurring on certain motherboards, which affects USB devices.  Users found a simple fix to this problem -- some report that by plugging in a USB storage device their computer will boot normally, but without the device attached it will crash.  For non-AMD/HP users, this remains an option if dealing with crashes.  Also some have found that switching the mouse from USB to PS/2 port (via the adapter to the round PS/2 port) fixes the problem, indicating it may be an issue with USB mouse drivers.

Additionally, users of AMD's Catalyst 8.4 drivers have also reported some crashes.  To see if the video driver is the issue, boot to VGA mode.  If it works, the video driver may be to blame.  One isolate report of an NVIDIA related crash also occurred, but it was unreported what driver was used.  Other miscellaneous crashes appear to be due to systems with multiple hard disks. 

A very helpful resource to deal with the problems is provided by Microsoft MVP in Windows Security, Jesper Johansson, who offers additional helpful details to these and other potential problems and their fixes on his blog.

As Johansson points out, the most extreme solution when none other can be found is simply to uninstall XP SP3.  To do this, refer to the Microsoft Knowledge Base article on the topic.

While Windows XP is receiving some bad press due to the crashes, again, it appears that most of the crashes are due to hardware issues stemming from unsupported configurations, and thus the blame fall largely to the PC manufacturers, and the makers of component drivers. Fortunately, the majority of the problems have easy fixes that do not even requiring uninstalling the Service Pack.

Similar problems occurred with Windows Vista SP1, though in that case the blame ended up resting with a Microsoft pre-install update.  It is fairly typical for a Service Pack to take some computers out of commission, particularly one for an OS with as large an operating base and as varied a hardware environment as Windows XP.  Nonetheless, such problems are serious concerns for users affected, and those potentially at risk.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

My findings
By FITCamaro on 5/12/2008 10:21:43 AM , Rating: 3
I had to remove SP3 because it messed up my Windows extender to stream things to my 360. I tried to reinstall the extender software but it isn't compatible with SP3.




RE: My findings
By FITCamaro on 5/12/2008 10:30:33 AM , Rating: 2
There wasn't any issue with me having 5 hard drive though.

Could be an issue with a certain chipset.

I haven't installed it on my main desktop yet though. It has the Catalyst 8.4 drivers installed so maybe tonight I'll install it there and see what happens. It also has 5 hard drives.


RE: My findings
By oab on 5/12/2008 12:58:35 PM , Rating: 2
I have catalyst 8.4 drivers installed on my "game" desktop and SP3 installed without any problems.

Core2 Duo
Asus P5B-E
ATI Radeon x1900
no-name wireless lan card


RE: My findings
By StevoLincolnite on 5/12/08, Rating: -1
RE: My findings
By FITCamaro on 5/12/2008 10:42:50 AM , Rating: 3
There will always be bugs in a product as massive as an OS. It's just a fact.


RE: My findings
By StevoLincolnite on 5/12/2008 10:55:15 AM , Rating: 2
Any large product can introduce bugs, even your Graphics cards have had bugs in the past, which required a re-spin of the silicon, or the feature blatantly disabled.

But what I don't understand is how could they miss something like USB Mice compatibility?
I'm sure they have an army of 50 thousand strapping lads testing everything as soon as they are given it.


RE: My findings
By virtuallyserved on 5/12/2008 11:26:26 AM , Rating: 2
Acknowledging that OS's have bugs and any large installed base will introduce some issues, it does seem that code produced by Microsoft in the post Vista era seems to be very underwhelming, even if it all isn't Microsoft's fault. Yet and still, they seem to be missing simple things that affect the largest number of people.

Even after looking at Server 2008 reviews (in particular the Powershell implementation) I'm not terribly impressed. Some of the features seem compelling, self healing NTFS, and maybe their new hypervisor (when it is finally released), but it does seem as if staying on what some consider legacy operating systems is the way to go for now.


RE: My findings
By mikefarinha on 5/12/2008 11:59:02 AM , Rating: 2
Although I can't claim this as fact, it reminds me of when Microsoft introduced better HTML and CSS compatibility in IE7.

There were many web sites that would detect IE and use workarounds due to flaws in IE6. Once IE7 was introduced, many of those flaws were fixed... but subsequently screwed up web sites that depended on the flaws in IE6 to render properly.

This could be a similar phenomenon with XP SP3 and Vista. Applications that weren't necessarily developed using best practices get bent out of shape when bugs are fixed.

I do know this is true with applications that were developed with poor security are run on Vista. Since Vista was developed from the start with security in mind; applications and drivers that disregarded security had major issues with Vista... however most problems seem to be fixed now.


RE: My findings
By Omega215D on 5/12/2008 9:52:20 PM , Rating: 1
Yet the annoying Apple commercials make their return. The recent one is where PCs were living with Vista with them going how it's not working how it should. What a load of garbage.

I have a MacBook and it runs Vista just fine and my XP SP3 installation hasn't had a problem yet.


RE: My findings
By MrBlastman on 5/12/2008 10:42:58 AM , Rating: 2
You are correct but given the complexity of Windows (I hate to admit it but windows IS very complex), any time you introduce new code to an already extremely large base of code, you are bound to introduce many new and often unintended bugs.

I think I'll just sit back and use SP 2 for the time being and let the rest of people test out SP 3.


RE: My findings
By StevoLincolnite on 5/12/2008 10:52:20 AM , Rating: 1
I understand this, as I did state windows was "Complex". However, I was reading the long list of bugs that were fixed, which actually amazed me to some extent, It should all become fine in the following weeks, as I remember Windows 95, 98, ME, NT, 2K had there share of Service pack issues. (Most notable ME).


RE: My findings
By thornburg on 5/12/2008 11:17:01 AM , Rating: 2
quote:
Windows 95, 98, ME, NT, 2K had there share of Service pack issues. (Most notable ME)


Ummm... There never were any Service Packs for ME. I think there were even any Service Releases (basically an update-rollup that came slipstreamed on an installation disk).

Windows ME was horrid, but Service Pack issues were not involved. Now NT on the other hand...


RE: My findings
By mikefarinha on 5/12/2008 12:03:55 PM , Rating: 2
quote:
Ummm... There never were any Service Packs for ME.


Correct.

There were never any service packs for the 16-bit/32-bit hybrid Windows OSes... Win95, Win98, or WinME.

Service packs were only for the NT based version of windows (NT, 2k, XP, ,2003, Vista). However there has been service packs for various Microsoft applications such as IE and Office.

I think the youngsters get confused and think there were service packs with Win98 and WinME because there were IE service packs released when these OSes were still around.


RE: My findings
By Bender 123 on 5/12/2008 1:52:32 PM , Rating: 2
You mean back in the old days? When the BSD was clean and pure and we had to manually configure jumpers on boards to get the darn computers to do what we wanted. Back then service packs would have driven me insane...I was just happy when my system would POST and Windows would boot without death spiraling.
I still have nightmares about updates from back then...It was always a crapshoot, hoping a driver change/security fix/update did not screw with your carefully configured IRQs and boot sequence.

I actually think (not certain) that you could consider 98/ME to be "Service packs" for the base OS in 95. Check me if I am wrong, but aren't they even the same version number, Windows 4.x?


RE: My findings
By Bender 123 on 5/12/2008 8:44:25 PM , Rating: 2
Just to reply to myself/add on...
95/98/ME were all version Windows 4.x. Despite the changes, they all seemed to follow the same basic incremental steps of what is now service packs.

I have not needed to look at those numbers in a looooooooong time.


RE: My findings
By Devo2007 on 5/12/2008 1:54:18 PM , Rating: 2
While they weren't called "Service Packs" they still did exist.

Windows 95 OSR2, OSR2.5
Windows 98 Second Edition

...though I don't think there was any for Windows ME.


RE: My findings
By darkpaw on 5/12/2008 1:57:13 PM , Rating: 2
Nope there weren't, ME really was a paid beta test product for some of the new features that made it into XP like system restore. I'm generally a pro MS guy, but ME was probably the single worst product they ever released (well Bob might have been worse, but at least it wasn't sold long).


RE: My findings
By mikefarinha on 5/12/2008 2:23:39 PM , Rating: 2
quote:
While they weren't called "Service Packs" they still did exist.

Windows 95 OSR2, OSR2.5
Windows 98 Second Edition


The difference here was you couldn't buy Win98 and then download the updates to get Win98SE. The same goes for Win95a Win95b, and Win95c. You had to buy the updated CD to get the updated version of windows.

I believe the big brouhaha with Win95c was that it had USB support.

Also Win95c was often referred to as Win97.