> Here's a patch, althought I left libtrivfs/io-map.c alone since it > returned EOPNOTSUPP.
Actually I am favoring returning EOPNOTSUPP instead of ENODEV. The ENODEV behavior is specific POSIXism for mmap, while EOPNOTSUPP is the generic Hurd protocol for any RPC that a server doesn't support. I'm already putting in the libc change to have mmap translate to ENODEV. > Are there any patch submission guidlines? (i.e. do we update the > ChangeLog for every trivial change? Yes. Follow the coding conventions you see, and make a good log entry for each and every change. No change ever goes in without a log entry. > Do we send patches to any one person, or just the mailing-list? etc.) bug-hurd is fine. If you get no response for a while, you can bug individuals about merging your patch. _______________________________________________ Bug-hurd mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-hurd
