That's an oversight. I'll push an improvement. Even with the improvement I have now, checking is not as complete as you might like. The check ensures that an array value has (at least) the expected number of elements and that each element's layout (in the sense of `ctype->layout`) is as expected. Checking only the element layout is weaker than checking the element type, but requiring the types to be `eq?` ctypes seems too strong, and the ctype API doesn't enable a more flexible notion of equality.
At Tue, 10 May 2016 18:03:07 +0200, Berthold Bäuml wrote: > When copying a C array into a a field of a C struct there seems to be no > “type > checking”, i.e., at least a check for element type and array length/size. > This > leads to a Segmentation fault when running the code below. Could such a check > be added? The relevant information should be available to (make-ct or > set-ct-arr …). > > > #lang racket > (require ffi/unsafe) > > (define _at2 (_array _double 5000000)) > (define _at1 (_array _double 5)) > > (define-cstruct _ct ([arr _at2])) > (define a (ptr-ref (malloc _at1) _at1)) > (define c (make-ct a)) > > > Best, > Berthold > > -- > ----------------------------------------------------------------------- > Berthold Bäuml -- Head of Autonomous Learning Robots Lab > DLR, Robotics and Mechatronics Center (RMC) > Münchner Str. 20, D-82234 Wessling > Phone +49 8153 282489 > http://www.robotic.de/Berthold.Baeuml -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
