-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Guilhem Lavaux wrote: > Hi, > > Here is a patch to merge NATIVE-LAYER branch to HEAD. I expect that the > new code will have some troubles to compile on some random OS different > from linux but hopefully it is sufficiently flexible to be
Yeah, it won't compile on darwin/x86: - -> gcc -DHAVE_CONFIG_H -I. -I../../../../native/jni/native-lib - -I../../../include -I../../../../include - -I../../../../native/jni/classpath -I../../../../native/jni/native-lib - -pedantic -W -Wall -Wmissing-declarations -Wwrite-strings - -Wmissing-prototypes -Wno-long-long -Wstrict-prototypes -g -O2 -MT cpnet.lo -MD -MP -MF .deps/cpnet.Tpo -c ../../../../native/jni/native-lib/cpnet.c -fno-common -DPIC -o .libs/cpnet.o - -> ../../../../native/jni/native-lib/cpnet.c: In function 'cpnet_connect': - -> ../../../../native/jni/native-lib/cpnet.c:183: warning: unused variable 'theaddr' - -> ../../../../native/jni/native-lib/cpnet.c: In function 'cpnet_send': - -> ../../../../native/jni/native-lib/cpnet.c:252: error: 'MSG_NOSIGNAL' undeclared (first use in this function) - -> ../../../../native/jni/native-lib/cpnet.c:252: error: (Each undeclared identifier is reported only once - -> ../../../../native/jni/native-lib/cpnet.c:252: error: for each function it appears in.) - -> ../../../../native/jni/native-lib/cpnet.c: In function 'cpnet_sendTo': - -> ../../../../native/jni/native-lib/cpnet.c:268: error: 'MSG_NOSIGNAL' undeclared (first use in this function) - -> ../../../../native/jni/native-lib/cpnet.c: In function 'cpnet_getHostByName': - -> ../../../../native/jni/native-lib/cpnet.c:603: warning: implicit declaration of function 'gethostbyname_r' > autoconfiscated. I have checked there is one less regression on my > system (I have not tested the AWT part because this system is a bit out > of date and I could not compile the peers but I have not touched > anything in that part so there should not be any regression). At the > moment the ChangeLog is not yet merged. The branch changelog is still > held in ChangeLog.native (Mark what to do with this ?). I would like > some of you test this on a linux workstation and see if it works ok. I > will leave for (real !) vacations tomorrow so I may not be joinable. > Casey can you take this in charge if I am not responsive ? You told me > you can help... > I've been hacking on our NIO stuff a bit, and I will need to modify our java.net code, too. I'd like to do those modifications after this goes in. In general I'll say that the approach looks pretty good. I'd prefer that we stick closer to a POSIX API, and make the cp* functions just simple wrappers where they need to be (as in: blocking syscalls). I don't have a Linux machine to test this on right now, though. It is probably OK to commit this now, then fix the issues on darwin (et al) in CVS. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iQEVAwUBRODsGeRILCRAfKHCAQL9TggAtCy675oOCUPz3x8BdOXy1i8q62aI8Twb 3qSen0TD8HKrhpv/RT5yVSl8rq1oVXxByhyI7C4kQFsZY+4DlUmMuBK9tyQBACSx tdBTTkPAti3Q16it+HZeuFOgGLFaLuGaGox09R/ke3ZOU/1b6ShNMix6W4bcoWID +UEMvwE5AEblPAXvB5YIhPKNHhPv45DP8rEVdFaDh9SMUP7v4+08qeLVuRLC+LhW OHF5clPPupdQEfFWzt/PUfobZzY/s7raJRkg7SIjYZA7Owmo8r1WZJ77QG89SKmI 7CIPFXbvvAvh3KHFxzBsmgAMsV4Vzt6N9thg9eBmR5/YKHIeAt1vQA== =mXZB -----END PGP SIGNATURE-----
