Thanks for the fix. I ended up fixing some problems with our test cases
so that they run properly on OpenBSD 6.0+, and your change checks out
OK. It's all at <https://github.com/mobile-shell/mosh/pull/862>.
Keith, Jeremie, would you mind looking at that?
regards,
--jh
On 03/20/2017 06:31, Jeremie Courreges-Anglas wrote:
john hood <cg...@glup.org> writes:
Hi all,
I'm pleased to announce a release candidate for Mosh 1.3.0. We have
brought you mostly bugfixes and release engineering improvements-- the
version number jump is because we have switched to semantic versioning
(<http://semver.org/>). The changes mostly address regressions in the
previous release, CPU usage in certain scenarios, and platform/port issues.
As already reported by Dan Chen in this thread, mosh-1.3.0-rc2 works
fine on OpenBSD-current, both amd64 and armv7.
A few days ago, a change happened in OpenBSD. The "ioctl" promise was
removed from pledge(2) on 2017/03/18. This means that this code:
#ifdef HAVE_PLEDGE
/* OpenBSD pledge() syscall */
if ( pledge( "stdio inet ioctl tty", NULL )) {
perror( "pledge() failed" );
exit( 1 );
}
#endif
now errors out.
pledge("ioctl") is deprecated since some time already; this removal will
end up in the upcoming OpenBSD 6.1 release.
mosh-client and mosh-server seem to work fine here with just
"stdio inet tty". Patch attached.
_______________________________________________
mosh-devel mailing list
mosh-devel@mit.edu
http://mailman.mit.edu/mailman/listinfo/mosh-devel
_______________________________________________
mosh-devel mailing list
mosh-devel@mit.edu
http://mailman.mit.edu/mailman/listinfo/mosh-devel
