Wayne Davison [Thu, 9 Apr 2020 22:11:37 +0000 (15:11 -0700)]
Set Copyright years and make them easier to update
I replaced git-set-file-times with an improved version that I wrote
recently (in python3). A new script uses it to figure out the
last-modified year for each *.[ch] file and updates its copyright.
It also puts the latest year into the latest-year.h file for the
output of --version.
Wayne Davison [Wed, 8 Apr 2020 01:08:05 +0000 (18:08 -0700)]
Fix the default skip-compress list.
The default value of the skip-compress list actually comes from the
daemon's default lp_dont_compress() value, but a while back the vars
stopped getting default values in a non-daemon run. I added a call to
reset_daemon_vars() so that the "Vars" values get set from "Defaults".
Wayne Davison [Sun, 5 Apr 2020 23:41:15 +0000 (16:41 -0700)]
Don't throw an error if a potential fuzzy dir isn't a dir
Add a flag for calling get_dirlist() and for send_directory() that
indicates that the dirname is allowed to not be a directory. Based
on a patch by Ben Rubson. Fixes bug #13445.
Wayne Davison [Sun, 5 Apr 2020 21:52:32 +0000 (14:52 -0700)]
Add progress output via SIGINFO and SIGVTALRM
On BSD-ish systems you can type Ctrl+T to see the current file and
the progress output (in --info=progress2 format). On hosts w/o
SIGINFO, use something like "killall -VTALRM rsync" or a more
targetted "kill -VTALRM PID ..." call (as needed).
Wayne Davison [Sun, 5 Apr 2020 18:21:02 +0000 (11:21 -0700)]
Added the --write-devices option.
This is a fleshed out version of the old one in the patches repo with
documentation & proper handling of the implied --inplace option for a
daemon's option-rufusing considerations. I ommitted the -w short option
as I would hate for someone to turn this on accidentally.
Wayne Davison [Sun, 29 Mar 2020 20:01:13 +0000 (13:01 -0700)]
Add a --copy-as=USER[:GROUP] option
This can be used by a root-run rsync to try to make reading or writing
files safer in a situation where you can't run the whole rsync command
as a non-root user.
Wayne Davison [Sat, 16 Mar 2019 18:12:53 +0000 (11:12 -0700)]
Fix bug in try_dests_reg that Florian Zumbiehl pointed out.
If the alternate-destination code was scanning multiple alt dirs and it
found the right size/mtime/checksum info but not the right xattrs, it
would keep scanning the other dirs for a better xattr match, but it
would omit the unchanged-file check that needs to happen first.
Jeriko One [Mon, 20 Nov 2017 22:42:30 +0000 (14:42 -0800)]
Ignore --protect-args when already sent by client
In parse_arguments when --protect-args is encountered the function exits
early. The caller is expected to check protect_args, and recall
parse_arguments setting protect_args to 2. This patch prevents the
client from resetting protect_args during the second pass of
parse_arguments. This prevents parse_arguments returning early the
second time before it's able to sanitize the arguments it received.
Wayne Davison [Mon, 4 Sep 2017 21:20:16 +0000 (14:20 -0700)]
Add "daemon chroot|uid|gid" parameters.
This allows the daemon to run chrooted as any uid+gid you like
(prior to the transfer possibly changing the chroot and/or the
uid+gid further). Based on the patch in #12817.
Wayne Davison [Sat, 29 Oct 2016 21:47:58 +0000 (14:47 -0700)]
More fixes for --progress quirks.
This patch avoids inconsistent evaluation of options in the
show_filelist_p() function by turning it into a var. We
also avoid setting "output_needs_newline" if --quiet was
specified.
Wayne Davison [Mon, 10 Oct 2016 18:49:50 +0000 (11:49 -0700)]
Support --sparse combined with --preallocate or --inplace.
The new code tries to punch holes in the destination file using newer
Linux fallocate features. It also supports a --whole-file + --sparse +
--inplace copy on any filesystem by truncating the destination file.
xattrs: maintain a hashtable in order to speed up find_matching_xattr()
As a testcase I've used one directory on gpfs with 1000000 files,
each with an xattr called 'name$i' having a value of 'value$i'.
So we also have 1000000 unique xattrs. The source and dest directories
are already in sync before. So the rsync command is basically a noop,
just verifying that everything is already in sync.
The results before this patchset are:
[gpfs]# time rsync -a -P -X -q source-xattr/ dest-with-xattr/
real 8m46.191s
user 6m29.016s
sys 0m24.883s
[gpfs]# time rsync -a -P -q source-xattr/ dest-without-xattr/
real 1m58.462s
user 0m0.957s
sys 0m11.801s
With the patchset I got:
[gpfs]# time /gpfs/rsync.install/bin/rsync -a -P -X -q source-xattr/ dest-with-xattr/
real 2m4.150s
user 0m1.917s
sys 0m17.077s
[gpfs]# time /gpfs/rsync.install/bin/rsync -a -P -q source-xattr/ dest-without-xattr/
real 1m59.534s
user 0m0.924s
sys 0m11.599s
It means the time in userspace dropped from 6m29.016s down to 0m1.917s!
Without -X we get ~ 0m0.9s with or without the patch.