[PD] another free c lesson: realloc

[Jonathan Wilkes]

Looked at this again and I can't seem to induce an error or crash.  But I also cannot understand how gp_w can be used to reliably refer to an element in a dynamically allocated byte array.
Doesn't this stackoverflow discussion apply here?http://stackoverflow.com/questions/2172127/how-to-update-other-pointers-when-realloc-moves-the-memory-block
Thanks,Jonathan
 


     On Wednesday, August 19, 2015 11:33 AM, Jonathan Wilkes via Pd-list <pd-list at lists.iem.at> wrote:
   

 Hi list,So I'm looking at t_gpointer in m_pd.h, and I'm noticing that for arrays it has gp_w which is a union _word pointer.  This is the raw data that I was asking about in my previous request for a free lesson. :)
However, in the code for [setsize] the data is resized dynamically using realloc.
So...
Suppose I have a [pointer] in a patch which is pointing to a nested array, and a [setsize] which is pointing to the container array.  If I send an enormous float to [setsize], isn't it likely that realloc will move the data and my [pointer] data will silently become stale?  After all, gp_w has no idea if realloc moves the data.  Seems like it'd be pointing to the wrong place and likely to cause a crash.
I don't have access to a running Pd instance atm, otherwise I'd test...
Thanks,Jonathan


_______________________________________________
Pd-list at lists.iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list


  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20150820/b8920bc1/attachment.html>