Keyboard/mouse system freezup on N100 running Vista [ Edited ]

[patamia]

This is the second problem posted by me regarding my N100.  The first involves need for an Nvidia driver update.  This one is abut keyboard/mouse system freezeups.

I have been running Vista on my N100 0768A2U model since about April. I keep up with all Vista and Lenovo updates.  The system, in general works well and is stable.  However, several months ago I began to experience keyboard/mouse system freezeups in the course of rapid typing.  The system stops responsing to any and all keyboard, hotkey and mouse input -- but in the background I can see that the system continues to operate.  I can recover without data loss by pressing the power key to initiate a standby and when I come back on the recovery from standby is complete and correct and the problem is gone -- except that it is porne to repeat once it has appeared.  In short, the problem appears to be an uncleared mask of the interrupts from keyboad and mouse inputs that does not otherwise affect the running of the system.

I have experimented extensively with this problem.  I have searched for any Vista specific reports of this in any context without getting a firm match.  I have described this problem on the Design Maters blog without anyone else recognizing it as having seen it themselves.  I cannot find a strong correlate with configuration of the system or some things running -- but I do have some weak ones.  The intermittancy is troubling -- I can go days without seeing it and yet see it persistently on other days.  The only constant is that it clearly happens during rapid typing and it does not seem to matter what application I am typing into.

There is a chance that the problem is my keyboard.  My warranty runs out on Jan 20 2008.  I will accept a replacement keyboard and am qualified to do a field replacement -- I have the manual.  I would love to know if anyone else running Vista or not has seen this phenomenon. 

I have run all available diagnostics without finding any flaws of any kind.  I have checked that the keyboard driver matches what is current by Microsoft and Lenovo standards.  I cannot rule out a software issue, but the evidence so far points to a hardware one at the keyboard level.  It would be easy to find out by replacing the keyboard -- one of the least expensive components of the system.  I like the system, but find the keyboard to be a compromise anyway, but it did work flawlessly until a few months ago.  This MAY mean that some BIOS or driver update from Lenovo obtained about that time is a factor, but if this were true I'd expect it to have affected other users -- and if it has I have yet to hear about it.

I would appreciate any corroboration from any other user or from Lenovo on this before I make a formal bid for customer support.  I will seek the keyboard replacement if I get this or can ever find a firm way to trigger the problem -- or if it gets too close to January 20th 2008. 

Thanks.

Message Edited by patamia on 12-05-2007 10:48 PM

[RealBlackStuff]

From reading this (slightly convoluted) post on TechRepublic, it looks like a BIOS update could fix your problem

As an afterthought, on regular PCs there are settings for keyboard buffer size, typematic (=repeat rate) etc. Don't know if these are available on your laptop, let alone in Vista.

There are also settings in the Registry, if you trust yourself that far.
See this tip for fast typists here. It's for NT, but XP and Vista should still use the same flags.



Message Edited by RealBlackStuff on 12-06-2007 07:02 AM

[nima4u]

Hello

this problem alsoo for me,

my Labtop model 300 N100 Q4G-0768 
each buttom that I click on my keyboard , My USB gets on and off, and what I beleive is since I update the Biso this problem accure , please let me know what is this problem coming from?  how can I fix it ?

do you think the problem is from hardware ?

Thank's

[patamia]

Thanks for the pointer to TechRepublic.  It looks like he did a flash update of the previous BIOS -- cannot tell if that was a step forward or backward for him.  He wasn't so much convoluted as struggling with his English.  Anyway, it may be a clue since I already suspected that this crept in around the time of the last BIOS update -- which by the way is the one AFTER the one he says he flashed.

Problem is I am reluctant to go backwards on a BIOS update.  I'd like more info or guidance before doing that.  I am fearful of what might regress if I go back a version of the BIOS.

Thanks for the help.

[patamia]

To: RealBlackStuff:  and yes I tried to play with repeat rates for the keyboard etc.  I do not claim to have manipulated these variables extensively, but so far there has been no consequence.  Since the problem is intermittant it would take forever to carefully try many settings before you knew which ones had any impact.

[RealBlackStuff]

Methinks the registry change to increase the buffer size would be the most effective (and simple) one.

[patamia]

Okay ... I'm perfectly fine with registry edits so I tried it.  I doubled the buffer (in hex that's x64 to xc4).

I agree that something like this is a classic suspect -- buffer overruns are a common vulnerability of tightly coded system objects.  Nevertheless, I type fast sometimes, but not THAT fast!.  Still, its worth a shot.  Even if it works it may take days to know.  If it does not work, the result will be decisive, but it could still take days to know. 

I will report the outcome.  Thanks for the help.

[patamia]

I can now report that increasing the buffer size is not a solution.  It is not possible to determine as of yet whether it ameliorates the problem, but I can be decisive about whether it cures it.  It does not.

Along the way, though, I strengthened a previous observation which I don't think I mentioned before.  Namely, the freezup is most likely to occur when rapid keyboard activity lines up with disk activity -- possibly limited to things like autosaves.  Its anecdotal at this point and I still haven't been able to force it to happen, but perhaps by lowering the save delay interval for an application I can confirm that its a factor.  I can say that, when trying to provoke the problem by aggressive typing, that saves can come and go without effect.  Yet, when it happens I can often see that a save was in progress or just completed much of the time.  A save in the same activity might cause the keyboard buffer to start filling -- but not by an amount anything close to its present size (200 bytes).

Anybody got more information on the BIOS factor?  I haven't checked, but is the prior BIOS still available and are there clear risks stepping back one update?

Thanks!

[RealBlackStuff]

If that were my laptop, I would revert back to Windows XP.
Alternatively, see if you can get Vista SP1 already and see if that solves the problem. Make a backup or take an image before you install SP1.
I don't know if previous BIOS versions are available, check with your Type/Model here.

Just found the older BIOS versions here. Check that you have the proper BIOS for Vista, and not one for XP!


Message Edited by RealBlackStuff on 12-13-2007 10:52 AM

[patamia]

RealBlackStuff:

You have helped me stumble onto an important anomaly that may indeed point to the solution!

Bear in mind that the last BIOS update on this machine was performed more or less automatically when I allowed the Lenovo Care software to scan for and select appropriate updates.  Prior to today, all I did more recently was to check that the file name of the update it retrieved corresponded to the then most recently available BIOS update. 

BUT I did not look harder than that at the material. 

What you just did was refer me back to the general web page (which I could have found on my own, but did not go looking for) describing the BIOS updates from June this year.  More importantly, you questioned whether, per chance, I had the XP rather than Vista update running.  Terrific insight!  When I compare the system information to the web page, my  system IS running the XP version instead of the related Vista version!  I blame myself for not noticing this, but I blame the Lenovo Care software (no doubt the version of some months ago as I notice it was recently updated) for picking the wrong one in the first place.

Stay tuned.  I will double check all this and carefully make sure I have flashed in the latest VISTA compatible BIOS.  Then we shall see what happens.

Thank you for paying attention to this problem thread and for thining a bit outside the box.  We are all victims of our own assumptions -- and I am as guilty as anyone.

P.S. I am not, like some people, unhappy with VIsta and would not look forward to going back to XP.  Vista arrived with some problems (big surprise!), but I decided I would, for a change, ride the wave on this OS rather than play catchup later.  I find Vista overall to have better organized a lot of information and features -- which I greatly appreciate.  I can't say I like ALL of it, but on balance I find it to be a good experience that is getting better.  Heck, I even have used the 3D cascaded windows to figure out what I was doing when the desktop got messy.  I'm no poster child for Microsoft I assure you, but I have not had any genuinely bad experiences with Vista as many seem to have. I find it runs fast and clean, and I like some of the interface features.  So, Ohkay, they stole ideas from the Mac OSX -- par for the course.  Still, it is a step in the right direction.  Not a giant leap as they want us to believe -- but a positive step nonetheless.  Presumably SP-1 will stabilize things even more and perhaps deliver a few goodies.  Geee -- maybe they will stop being chincey and put direct faxing software back in most of the editions so that people who had it in 2000 and XP will still have it without having to upgrade to the fanciest Vista edition!