Re: Problems on Itanium: Found the Cause, What's Next?

["Alex Alexandrov" <alex_alexandrov at fromru dot com>]

Hi, Corinna Vinschen, you wrote

> It also works fine on a x86_64 machine in 32 bit mode.

Yes, I've checked it too - it really works on Intel EM64T (didn't test it on 
AMD64).

> No, because this would also exclude x86_64 machines.

Probably, a better approach is needed. It is possible to check 
PROCESSOR_ARCHITECTURE envvar but it looks a little clumsy.

> Actually I'm somewhat in doubt that CreateFileMapping/MapVieOfFile
> functionality should actually be so broken on the ia64 platform.
> Are you really sure that the culprit isn't somewhere else?
> Your test application uses a very common idiom, so I'd guess that
> a lot more applications would fail miserably on ia64 in 32 bit mode.

Belieive it or not, it is! I've double checked it (I mean I've done that on 
2 computers :-) ). It is easily reproducible with cygwin - any malloc for a 
big memory chunk causes segmentation fault (if I understand correctly malloc 
requests for big memory areas lead to mmap call).

I've posted the bug report to public.win32.programming.kernel and 
private.windowsserver_64bit msft mailing lists - no answer so far...

With best regards, Alex Alexandrov.  E-mail: alex1_alexandrov2@fromru3.com4 
(remove numbers before e-mailing me) 


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/