Re: Strange crash

[gerard patel]

At 10:30 AM 20/04/2001 +0200, you wrote:
<snip>
> The lower 6 bits are the new floating point exception masks. Mine are
> all set and yours are cleared, therefore the crash.
>
> Why that is the case, I haven't got a clue.

I'm not sure I followed everything, but it could come from a problem that
occurred a lot of time before, no? The bits could be already cleared on entry of the subroutine.
If the exception handling bits are cleared and the app don't do anything with floating point, 
nothing will break, AFAICT. If this is the case a way could be to patch the code to
print the mask with every +relay operation.

BTW, I wonder if Dmitry has done some classic tricks like

- getting a fresh tree from Winehq
- compiling everything with -O0
- trying the same app on another computer with a (as much as possible) different
configuration.

Gerard

Strange crash

[Rein Klazes]

On Fri, 20 Apr 2001 14:14:17 +0200, you wrote:
> If this is the case a way could be to patch the code to
> print the mask with every +relay operation.

I think that is a good idea. The function fegetenv(3) can be used for
this, the lower 6 bits of member __control_word of the returned
structure is the one that we need.

Rein