This is the mail archive of the mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RE: Cygwin causes 0x00000024 Stop Error (BLUE SCREEN)


I suppose you're right: I was too "enthusiastic" (does that qualify as a euphemism in this context?). Also, I must confess I didn't read the KB entry before my previous post.

However, if this is really the result of that known problem in Windows, why have we heard so little of it here? Have we gotten to the point where people are afraid to report bugs, or too ready to blame problems on anything _but_ Cygwin? I wouldn't have thought so. Maybe we've been too emphatic in hammering home the "Blue Screen == SEP" equation.

If KB 195857 really is the culprit here, then I suppose you could well make the case that opening a lot of system handles and then just abandoning them is not good programming practice and really should be resolved on its own merits.

I take it hardware configuration (available RAM and / or paging space) is one a parameter in the likelihood of triggering this problem.

Randall Schulz
Mountain View, CA USA

At 12:40 2002-11-24, you wrote:
On Mon, 2002-11-25 at 08:27, Randall R Schulz wrote:

> Likewise, it cannot be the locus of a fix for such a problem. While I
> certainly cannot speak for others who read this list, I doubt anyone is
> going to a) be able to help you much to solve this situation, since it is
> surely idiosyncratic and specific to your system; b) be motivated to do so.

Randall, I think this is a little to enthusiastic! While setup.exe is
not the root cause of the problem, it may be triggering the fault - and
because (assuming that kb 195857 is the cause (*))
a) cygwin is for windows
b) this fault is apparently endemic to windows
c) MS haven't fixed it

The onus is on us to have cygwin work around the fault. We've done that
before, we'll do it again. Daniels email was informed, researched, and
pleasant - so I'll certainly be assisting (but not admitting to a setup
'bug' per se :}) and I suspect Max will too.


(*) The technet article doesn't list XP as suffering from the fault,
which I read to mean that the fault is fixed in XP, and (inferring here)
is a kernel issue which would change current published 2K and NT
behaviour, thus will not be fixed on those kernels.

Unsubscribe info:
Bug reporting:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]