[ https://issues.apache.org/jira/browse/STDCXX-1057?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13202609#comment-13202609 ]
Travis Vitek commented on STDCXX-1057: -------------------------------------- The question here is not what value should {{string::max_size()}} or {{allocator::max_size()}} return. The question is should the {{std::basic_string<T>}} constructors use {{basic_string<T>::max_size()}} or {{allocator::max_size()}} at all, or should they just try to make the allocation request and allow the exception to propagate out to the caller? I believe that Martin is suggesting the constructors should consult {{basic_string<T>::max_size()}} before making the allocation, and the Perennial testcase seems to think that this behavior is forbidden. I read through all of 21.3.1 and didn't see anything that indicates there is a requirement either way. There are requirements relating to this for functions that grow the string, but I don't see anything for the string constructors. So, as Martin suggests, this is _unspecified behavior_, and as such, the actual behavior is _implementation defined_. That said, I did find [LWG#83|http://www.open-std.org/Jtc1/sc22/wg21/docs/lwg-defects.html#83], which indicates the addition of the following paragraph to the requirements of {{std::basic_string}} {quote} For any string operation, if as a result of the operation, size() would exceed max_size() then the operation throws length_error. {quote} A quick check of the C++11 standard shows that as [string.require] p1. > attempting to create a std::string of size 65535 or greater fails with > Perennial CPPVS V8.1 > ------------------------------------------------------------------------------------------- > > Key: STDCXX-1057 > URL: https://issues.apache.org/jira/browse/STDCXX-1057 > Project: C++ Standard Library > Issue Type: Bug > Components: 21. Strings > Affects Versions: 4.2.1, 4.2.x, 4.3.x, 5.0.0 > Environment: Solaris 10 and 11, RedHat Linux, OpenSuSE Linux > SUN C++ Compilers 12.1, 12.2, 12.3 > Defect is independent of compiler and platform > Reporter: Stefan Teleman > Labels: conformance, features, standards, test > Fix For: 4.2.x, 4.3.x, 5.0.0 > > Attachments: stdcxx-1057.patch, test.cc > > > in member function: > size_type basic_string<_CharT, _Traits, _Allocator>::max_size(); > the maximum size of a basic_string is restricted to less than 65535 bytes. > The Standard is ambiguous as to what the max_size() of a std::string should > actually be (see LWG Core Issue 197). However, less than 65535 bytes for > the max_size of a std::string is rather small. GNU libstdc++ and stlport4 > set std::string::max_size to (SIZE_MAX / 4) (i.e. 1GB). Solaris sets it > to SIZE_MAX. > Perennial CPPVS explicitly tests for the creation of a std::string of size > greater than 65535. In the current stdcxx implementation, this test fails. > The max_size of a std::string should be significantly greater than 65535 > bytes. > Test to reproduce the defect: > {code:title=test.cc|borderStyle=solid} > #include <iostream> > #include <string> > const size_t maxlen = 65536U; > char array[maxlen]; > struct test_traits : public std::char_traits<char> > { }; > template<class T> > struct test_alloc : public std::allocator<T> > { > typedef typename std::allocator<T>::size_type size_type; > template<class Y> > struct rebind > { > typedef test_alloc<Y> other; > }; > test_alloc() throw() { } > test_alloc(const test_alloc& rhs) throw() { } > template<class Y> > test_alloc(const test_alloc<Y>& y) throw() { } > ~test_alloc() throw() { } > size_type max_size() const throw() { return maxlen; } > }; > int main() > { > typedef > std::basic_string<char, test_traits, test_alloc<char> > test_string; > int ret = 0; > size_t i, j; > for (i = 0; i < maxlen; i++) > array[i] = '*'; > array[maxlen - 1] = '\0'; > for (i = 0; i < maxlen - 1; i+= 8) > { > array[i] = '\0'; > test_string s(array); > j = s.size(); > array[i] = '-'; > if (i != j) > { > std::cerr << "i = " << i << " j = " << j << " expected i == j" > << std::endl; > ret = 1; > break; > } > } > return ret; > } > {code} > 1. Output from GCC 4.5.0: > {noformat} > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:16:34][2162]>> ./test-gcc > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:16:48][2163]>> echo $status > 0 > {noformat} > 2. Output from Sun C++ 12.2 with stlport: > {noformat} > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:16:50][2164]>> ./test-ss122-stlport > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:16:58][2165]>> echo $status > 0 > {noformat} > 3. Output from Sun C++ 12.2 with our patched stdcxx: > {noformat} > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:17:00][2166]>> ./test-ss122-stdcxx > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:17:06][2167]>> echo $status > 0 > {noformat} > 4. Output from Pathscale 4.0.12.1 (which did not patch stdcxx): > {noformat} > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:17:08][2168]>> ./test-pathscale > Terminating due to uncaught exception 0x614240 of type std::length_error > Abort (core dumped) > [steleman@darthvader][/src/steleman/programming/stdcxx-ss122/bugfixes-sunw/6889771][02/06/2012 > 10:17:13][2169]>> echo $status > 134 > {noformat} > Patch for 4.2.1 to follow shortly. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira