Re: [Bug-xnee] cnee causing segfault

[Henrik Sandklef]

Thanks for your report. It's now fixed in CVS (HEAD). A tarball* should 
be available in a few hours.

Since the code we're dealing with here is something that soon will be 
removed (once the RECORD bug is fixed) I will not spend time 
investigating this further.

/hesa

*) http://itupw056.itu.chalmers.se/xnee/nightly-dists/


On 12/31/2009 12:42 AM, Elizabeth B. wrote:
> Hi,
>
> Using the CVS version of libxnee, cnee kept segfaulting:
>
> address@hidden [17:26]:~/src/xnee/cnee/src$  ./cnee --keyboard --record --str 5
> Sleep workaround a strange RECORD/Xtest problem around 2009 (ignore it)
> Workaround: Adding your device events as delivered events to get them recorded.
>              You can ignore this message
> X info:   The X.Org Foundation 1 6 5
> RET: 1
>
> [snip]
>
> error-range                   0-0
> X info:   The X.Org Foundation 1 6 5
> RET: 1
> Workaround: Creating context on data display instead of control
>              You can ignore this message
> X info:   The X.Org Foundation 1 6 5
> RET: 1
> Workaround: Creating context on data display instead of control
>              You can ignore this message
> X info:   The X.Org Foundation 1 6 5
> RET: 1
> Workaround: Disabling context on data display instead of control
>              You can ignore this message
> Segmentation fault
>
> anyway the offending line of code is:
>
> xnee_record:821       else if (strstr(xd->x_vendor_name, "Sun Microsystems") )
>
> for some reason (because the code is exiting?) this function is called
> but there is nothing in xd->x_vendor_name (xd->x_vendor_name == NULL);
> maybe its a weird race condition.
>
> Attached is a patch that corrects this issue. Also attached is the
> complete output of cnee before the patch.
>
> Best wishes, Elizabeth B.
>
>
>
> _______________________________________________
> Bug-xnee mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/bug-xnee