c++11 - Is this union compatible with strict-aliasing rules? -


is ok use both parts of union if know parts don't overlap? in example, ok use both buf[31] ps?

struct ptrsize {   const char *data;   size_t size; };  class smallstringornot {   union {     ptrsize  ps;     char     buf[32];   } pb;  public:   bool issmallstring() const {     return pb.buf[31] != 0;   }   smallstringornot(const char *str) {     size_t len = strlen(str);     if (len && len < 31) {       memcpy(pb.buf, str, len);       pb.buf[31] = len;     } else {       pb.ps.data = str;       pb.ps.size = len;       pb.buf[31] = 0;    // ok, accessing buf right after ps?     }   }   ptrsize asptrsize() const {     if (issmallstring()) {       return ptrsize{pb.buf, pb.buf[31]};     } else {       return pb.ps;     }   } }; 

unfortunately code not ok: @ least not in "undefined behaviour"-zone, since in c++ legal access union through char member, have no guarantee modifying buf[31] not altering ps.data or ps.size. in 128-bit machine surely doing it.

on more normal architectures, code should fine 100% guarantee should refer compiler documentation, since size_t in principle bigger void*. example, on 64-bit machine theoretically have 192-bit ps.size member (which summed 64 bit of ps.data pointer make ptrsize overlap buffer.


Comments

Popular posts from this blog

jquery - How do you format the date used in the popover widget title of FullCalendar? -

asp.net mvc - SSO between MVCForum and Umbraco7 -

Python Tkinter keyboard using bind -