On Fri, 2014-08-22 at 18:50 +0200, Nikos Mavrogiannopoulos wrote: > On Fri, Aug 22, 2014 at 5:04 PM, David Woodhouse <dwmw2 at infradead.org> wrote: > > On Fri, 2014-08-22 at 11:00 +0200, Nikos Mavrogiannopoulos wrote: > >> The attached patches fix some issues in windows. Of them patch-0002 is > >> not critical, but switches to windows types to avoid casts from the > >> caller. > > Thanks. Can you say *what* valid input inet_addr() fails on? Other than > > "255.255.255.255" which is handled as a special case? > > It failed both for IP address 192.168.1.0 and 255.255.255.0 on my > windows 7 installation. I didn't bother to debug that issue, so I just > used pton. Fair enough. But still we can just use inet_pton() unconditionally. I think Windows was the only platform that *didn't* have inet_aton() anyway. I've pushed a simple version of that. > I don't think createprocess() can handle scripts, that's the reason I > got through cmd.exe. Anyway that patch is now quite improved. It can't handle scripts, but surely it can use cscript.exe instead of cmd.exe ? For the second patch... I'm not sure we should be polluting the global namespace on POSIX systems with that '#define SOCKET int'. Can we do something like #ifdef _WIN32 SOCKET #else int #endif openconnect_setup_cmd_pipe(struct openconnect_info *vpninfo); It's horrible but at least it's isolated. -- dwmw2 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 5745 bytes Desc: not available URL: <http://lists.infradead.org/pipermail/openconnect-devel/attachments/20140822/73602e88/attachment.bin>
