DO NOT REPLY [Bug 10648] New: - IDDocumentImpl misaligned allocations on machines with a 64 bits 'long' type

[bu...@apache.org]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10648>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=10648

IDDocumentImpl misaligned allocations on machines with a 64 bits 'long' type

           Summary: IDDocumentImpl misaligned allocations on machines with a
                    64 bits 'long' type
           Product: Xerces-C++
           Version: 1.7.0
          Platform: Other
        OS/Version: Other
            Status: NEW
          Severity: Blocker
          Priority: Other
         Component: DOM
        AssignedTo: xerces-c-dev@xml.apache.org
        ReportedBy: benjamin.mironer@zslide.com


void * IDDocumentImpl::allocate(size_t amount)
aligns the adress it return to the size of void* (ie 32bits, generally)
see IDDocumentImpl.cpp line 892

size_t sizeOfPointer = sizeof(void *);
if (amount%sizeOfPointer!=0)
   amount = amount + (sizeOfPointer - (amount % sizeOfPointer));

On some MIPS or IA64 machines, long is a 64-bits-long, 64-bits-aligned type
while void * is a 32-bits-long, 32-bits-aligned type.

Hence, if the object being allocated contains members of the type long (such as 
an IDTreeWalkerImpl), these member will only be 64-bits aligned if the instance 
itself is 64-bits aligned (the compiler takes care of the internal alignement 
but the class needs to be aligned as well). If not, the constructor crashes 
horribly.

So, I suggest using sizeof(long) instead of sizeof(void*) in 
IDDocumentImpl::allocate(size_t amount) so it can run on 64 bits platforms.

This shouldn't cause any trouble on other platform (as long and int usually are 
both 32 bits, and if not we'll only lose a little space).

---------------------------------------------------------------------
To unsubscribe, e-mail: xerces-c-dev-unsubscribe@xml.apache.org
For additional commands, e-mail: xerces-c-dev-help@xml.apache.org