Martin Pool [Wed, 5 Dec 2001 12:48:46 +0000 (12:48 +0000)]
Change back to using sockaddr_storage rather than sockaddr_in. If
sockaddr_storage is not defined, then supply our own definition that
will hopefully satisfy RFC2553 but also compile on all supported
platforms.
Thankyou to YOSHIFUJI Hideaki, SUMIKAWA Munechika and Jun-ichiro
"itojun" Hagino.
David Dykstra [Mon, 3 Dec 2001 18:37:33 +0000 (18:37 +0000)]
Don't print out the directory name twice in verbose mode; it was being
printed a second time when the modification time of the directory was
being set, and that time around recv_files() calls recv_generator() with
an f_out of -1 so check that before printing the directory name.
Martin Pool [Sun, 2 Dec 2001 08:38:51 +0000 (08:38 +0000)]
Progress indicator now shows estimated rate of transfer (e.g. kB/s).
Based on a patch from Rik Faith, but modified to make sure we do only
one rprintf call, and that we never end up with two copies of the line
printed out.
Martin Pool [Sun, 2 Dec 2001 08:16:15 +0000 (08:16 +0000)]
Revert change from 1.39, because it causes a crash because of
attempting to free a static string. (Thankyou to Paul Mackerras.)
There's still a small leak here.
Martin Pool [Thu, 29 Nov 2001 00:04:48 +0000 (00:04 +0000)]
Show version when configuring.
If we don't seem to have an ANSI compiler, then omit a warning as soon
as that is discovered, because it is likely to break later configure
tests. This doesn't seem to catch the particular HP-UX compiler I was
after, which is non-ANSI but only emits a warning on this configure
test. Nevertheless probably better to have it in.
Martin Pool [Wed, 28 Nov 2001 06:52:04 +0000 (06:52 +0000)]
Put back the --disable-ipv6 option. This should only be needed if
your platform seems to support ipv6, but actually it breaks. This
seems to be the case for "powerpc-apple-darwin1.4".
David Dykstra [Mon, 26 Nov 2001 19:15:53 +0000 (19:15 +0000)]
Better fix for case of excluded symlinks that point nowhere when using
--copy-links. The readlink_stat() does need to be done in the normal case
before checking the exclude patterns because it needs to know whether or
not a file is a directory in order to properly handle a trailing slash
in an exclude pattern. This fix makes make_file() go ahead and call
readlink_stat() but then if the latter returns an ENOENT and copy_links is
on then it will only print an error if the path is not excluded.
Martin Pool [Mon, 26 Nov 2001 07:10:06 +0000 (07:10 +0000)]
Try to be better about handling the results of getaddrinfo(): when
opening an inbound socket, we might get several address results,
e.g. for the machine's ipv4 and ipv6 name.
If binding a wildcard, then any one of them should do. If an address
was specified but it's insufficiently specific then that's not our
fault.
However, some of the advertized addresses may not work because e.g. we
don't have IPv6 support in the kernel. In that case go on and try all
addresses until one succeeds.
Martin Pool [Mon, 26 Nov 2001 00:41:11 +0000 (00:41 +0000)]
Remove highly suspicious redefinition of sockaddr structure -- there
seems no point using anything but the platform's native definition,
and if we can't get that from the system headers we're hosed anyhow.
Martin Pool [Fri, 23 Nov 2001 07:35:49 +0000 (07:35 +0000)]
Merge KAME.net IPv6 patch: you can now (in theory) pass IPv6 hostnames
or literal IP addresses to rsync, and if your platform supports them
they will be used. Also there are -4 and -6 command-line options to
choose the default address type. Thankyou!
David Dykstra [Fri, 16 Nov 2001 19:19:44 +0000 (19:19 +0000)]
Fix to make_file() to exit earlier if a file is excluded, because doing
readlink_stat() on an excluded file can be a problem when using the
--copy-links option (also known as -L) and the excluded file is a symlink
that points nowhere.