Hi Alan, Volker, Thanks a lot! I've pushed the change.
Best regards - Jonathan On Fri, Jun 6, 2014 at 8:27 PM, Volker Simonis <volker.simo...@gmail.com> wrote: > Hi Jonathan, > > I just wanted to let you know that I've build your changes on AIX 5..3 and > 7.1. > I've also run the jdk regression tests without specific issues. > > So thumbs up from me! > > Regards, > Volker > > > On Fri, Jun 6, 2014 at 12:03 PM, Jonathan Lu <luc...@linux.vnet.ibm.com> > wrote: > > Hi Alan, > > > > On Fri, Jun 6, 2014 at 1:53 AM, Alan Bateman <alan.bate...@oracle.com> > > wrote: > > > >> On 05/06/2014 11:37, Jonathan Lu wrote: > >> > >> > >> If getsockopt(SO_ERROR) failed, I did not find any explicit docs about > >> the behavior. > >> but as I tested with some C code snippet, the value of sockopt_arg would > >> not be changed if getsockopt(SO_ERROR) failed. > >> So I prefer to keep the current approach, does it make sense to you ? > >> > >> The case that I was wondering about is the common case where > >> getsockopt(SO_ERROR) succeeds and I was wondering if the code should > >> actually be: > >> > >> if (sockopt_arg != 0 ) { > >> errno = sockopt_arg; > >> return -1; > >> } > >> > >> That way the caller of NET_Connect will have errno set so that > >> XXX_ThrowByNameWithLastError can create an appropriate exception > message. > >> > > > > > > You are right! errno will be checked by other code if NET_Connect() > failed, > > I've updated the patch, please help to review. > > > > http://cr.openjdk.java.net/~luchsh/JDK-8043954.3/ > > > > > > > >> > >> -Alan. > >> > > > > > > Many thanks > > - Jonathan >