Posted on 08/06/2009 10:08:36 AM PDT by ShadowAce
A contributor to the online forum TheHotfix.net has provided visual evidence of what appears to be a serious memory leak caused by the CHKDSK hard disk integrity checking utility included with a build of Windows 7 that has been tagged for RTM. The bug appears to occur during phase 4 of the disk check, and can push resource usage to the 96% level.
Tester Jordan M. Jacob provides a picture of the memory leak in progress, as depicted in Windows Task Manager, a portion of which is excerpted here. (It doesn't bode well for the integrity of the test that Apple's iTunes drivers also appear to be running.) Jacob goes on to warn that the leak is capable of sending the OS into the dreaded blue screen of death (BSOD).
Jacob called the alleged bug a "showstopper," saying it's big enough to warrant Microsoft suspending its release to manufacturing for Windows 7 -- a process which has already begun. Tomorrow is the anticipated date when MSDN and TechNet subscribers will have their first opportunity to download the finalized code, which is already being installed on new PCs by participating OEMs.
But should the bug's identity and severity be confirmed by Microsoft, it actually has other means available to it to quell the problem. One such mechanism is, quite ironically, the very thing which the Web site discovering the bug is named after: a hotfix, distributed via Windows Update and conceivably downloadable and installable during the Windows 7 setup process.
That almost too-obvious fact didn't stop one outside source, who confirmed the presence of the bug for himself, to sound the red-alert klaxon of doom, declaring the launch process already derailed and leading the story with the saliva-dripping words, "Oh boy!"
One Hotfix.net user answered Jacob by saying the problem may not actually be Microsoft's at all, but rather an issue with motherboard manufacturers that have yet to issue updates to their chipset drivers for the RTM builds. Nonetheless, the sheer volume of the virtual pow-wow among enthusiasts' sites today managed to rise to such a level that the president of Microsoft's Windows Division himself, Steven Sinofsky, responded to Chris123NT's forum in person this afternoon (Microsoft verified to Betanews this afternoon that the respondent is, in fact, Sinofsky).
Sinofsky said that Microsoft has been unable to reproduce the crash, though it has seen increased memory usage by CHKDSK. He then said that behavior was by design, intended "to speed things up" rather than slow things down. Later, he went on to say the following:
While we appreciate the drama of "critical bug" and then the pickup of "showstopper" that I've seen, we might take a step back and realize that this might not have that defcon level. Bugs that are so severe as to require immediate patches and attention would have to have no workarounds and would generally be such that a large set of people would run across them in the normal course of using their PC.We appreciate the kind words that such a bug as above is "out of place" with Windows 7 -- we're working hard. We are certainly going to continue to look for, monitor, and address issues as they arise if required. So far this is not one of those issues.
Some have reported...that this specific issue repros and then goes away with updated drivers. We haven't yet confirmed that either but continue to try. We just kicked off overnight stress testing of 40 machines of variants as reported by [forum user] FireRx. We'll see.
Let's see if we can work on this one and future issues together. Deep breath.
Been using the RTM for a while now - no problems.
Cool—Have you checked out the mem usage of CHKDSK? I’m not saying it’s a problem, but I’d like a FReeper test of this issue as well.
If he’s claiming it produces a blue screen it would be nice to know the error.
Although they claim updated drivers fix it, I’m going to try it within my VM, but if it is a memory leak it is probably related to the size of the disk and the quantity of data. Which means mine won’t show anything.
My laptop is at home, I’m at work. Maybe later.
I have a TB drive over half full - I’ll try it later (when I’m at home) and post the results.
....Waiting for Mac users to show up and start talking about how much PCs suck.....
Windows Hogs Computer Resources — more folloing out other breaking news items: Politicians Lie, Benedict XVI Affirms Catholicism, Bear Defecates In Woods.
This sucks, an erro like this that could happen to anybody showing up right before RTM. Most people (who even know it exists) forget about chkdsk until they actually need it, so it probably didn’t get as much testing as the other components during beta.
I hope MS can patch it and keep the schedule.
Interesting.
The article was posted ...
1 posted on 08/06/2009 10:08:36 AM PDT by ShadowAce
You posted your snarky comment an hour later...
8 posted on 08/06/2009 11:03:35 AM PDT by wbill
An hour later, there are still no such comments. Methinks you might owe the Freeper Mac community an apology.
I *am* surprised that it has taken this long, though. Usually the flames come out on PC vs Mac threads. Must be a busy day elsewhere in FReeperland.
It took less than 5 seconds with absolutely no hiccups in the music, I didn't even have time to find it in the Task Manager to see what the mem usage was (and I had it open already).
FWIW, I'm running 64bit.
My 32-bit VM with a 16gb “drive” took about 20 seconds though the chkdsk process did get up to 256mb.
I did the same to a XP box with about 300gb occupied and it never got above 21mb memory for the chkdsk process.
I would guess it’s by design rather than a bug, to go out on a limb.
However, on a 32bit system if that process gets up to 2gb or 3gb you could expect issues.
I think they’re confusing a large memory footprint with a memory leak. Just because it hogs all the memory for a bit doesn’t mean there’s a leak, if it lets go again then it’s just a bit of a pig.
I have 4GB of memory, that may have made a difference too. I really don’t this is much of an issue either.
Close to two hours later and no nasty comments from the Mac folks. Not that it -won't- happen; by all means keep waiting and watching, it keeps you out of trouble ;-)
But I hope this means the stupid, juvenile, trollish cr@p comments from both the anti-Mac and anti-Windows fanboys are finally going down a few notches. It's gotten really annoying lately and FR is better than that.
Real computer people don't trash other computer people's choices. Instead, they work on making their own choice better. Races are won by improving one's own entry, not by hurling cr@p at the other guy.
And for some of us serious computer folks, who've been doing it for over 30 years, it isn't even really a race, because different systems appear to different people and different uses. And that's as it should be.
That's what I get for waxing philosophical after 5PM on a Friday...
Damn straight. One of the grat things the free market has brought us are wildly different computing platforms, from mainframes, to tiny embedded systems that fill niches for certain functions. Personally, I find the power and flexibility of Unix suits the way I like to use computers. That doesn't mean someone else is a fool for using something else. It just means they have different needs and/or different levels of experience than I do.
Personally, I too would prefer to see less acrimony on the tech threads here at FR, as I think those of us more technically oriented folks could really be a great resource to other conservatives if we could keep the flame-fests at a minimum.
That's not to say a good fight is always bad. For instance if ya really want, we could have a spirited discussion on the relative merits of vi and emacs! Geek sparks are bound to fly!
Pretty much the same here - I tend to think in the Unix model, and on a new system the first thing I do is get some Unix-like tools on it (e.g. Cygwin) so I don't have to fight with the lack of native tools or screwing with overly-friendly GUI tools. At present, Mac OS-X is keeping its Unix side totally available, so until/if Apple decides to hide it, that's the best of both worlds for me. But Windows+Cygwin does pretty well also.
> we could have a spirited discussion on the relative merits of vi and emacs!
Can't help ya rumble there I'm afraid -- I use 'em both all the time. I was mainly emacs for years (starting ~1984) but then got on systems with only vi, and found it entirely usable. Now I'm using vi a lot more for sys-config stuff and emacs for text documentation.
So sorry, no fight here -- love/hate 'em both!
OTH, code indent style -- well, don't get me started... ;-)
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.