fine by me. We should probably also stop using inet_ntoa as it is not thread-safe (I don't know what is the safe alternative off-hand).
pl On 13 October 2016 at 19:00, Zachary Turner via lldb-dev <lldb-dev@lists.llvm.org> wrote: > I believe you are correct. We should create an in_addr on the stack and > memcpy into it. > > On Thu, Oct 13, 2016 at 10:52 AM Ted Woodward via lldb-dev > <lldb-dev@lists.llvm.org> wrote: >> >> TCPSocket::Connect has this line: >> >> host_str = ::inet_ntoa (*(struct in_addr >> *)*host_entry->h_addr_list); >> >> >> >> host_entry->h_addr_list is a char**, while struct in_addr contains a >> uint32_t. Casting like this (char * to uint_32t *) could cause a bus error >> on systems that don’t allow non-aligned loads. I think we need to memcpy the >> data into a struct in_addr variable. >> >> >> >> Anyone have any thoughts on this? >> >> >> >> -- >> >> Qualcomm Innovation Center, Inc. >> >> The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, a >> Linux Foundation Collaborative Project >> >> >> >> _______________________________________________ >> lldb-dev mailing list >> lldb-dev@lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev > > > _______________________________________________ > lldb-dev mailing list > lldb-dev@lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev > _______________________________________________ lldb-dev mailing list lldb-dev@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
