66dd25ebc4
when scandir gets called with an FD, dirent.path is not usable
...
if scandir does not get a path, it can't prefix it in front of the
filename in the direntries it returns, so dirent.path == dirent.name.
thus, we just only use dirent.name and construct the full path.
2019-02-14 09:20:04 +01:00
833c49f834
use *at style functions (e.g. openat, statat) to avoid races
...
races via changing path components can be avoided by opening the
parent directory and using parent_fd + file_name combination with
*at style functions to access the directories' contents.
2019-02-14 09:20:04 +01:00
ad5b9a1dfd
_process / process_*: change to kwargs only
...
we'll add/remove some args soon, so many pos args would be just bad.
2019-02-14 09:20:03 +01:00
8220c6eac8
move/refactor Archive._open_rb function to helpers.os_open
...
also:
- add and use OsOpen context manager
- add O_NONBLOCK, O_NOFOLLOW, O_NOCTTY (inspired by gnu tar)
2019-02-14 09:20:03 +01:00
677102f292
process_file: avoid race condition: stat data vs. content
...
always open the file and then do all operations with the fd:
- fstat
- read
- get xattrs, acls, bsdflags
2019-02-14 09:20:03 +01:00
dca04ba538
_process: remove an infrequently used micro-opt
...
st param was only given at the root paths of the recursion.
we can just drop that and make the code simpler.
2019-02-14 09:20:03 +01:00
ce69fc1e19
Merge pull request #4238 from ThomasWaldmann/fixed-block-chunker
...
implement a fixed block size chunker, fixes #1086
2019-02-14 09:04:02 +01:00
7f46eb99aa
update docs about fixed chunker and chunker algo spec needed
2019-02-13 06:30:13 +01:00
ac0803fe0b
chunker algorithms: use constants to avoid typos
2019-02-13 04:36:09 +01:00
be2c061733
chunker params parsing: add more validation
...
avoiding too large chunks that the repository can not store.
avoiding too small chunks that would create excessively many chunks
and way to much storage and management overhead. we only disallow
extreme cases, this does not mean that everything that is allowed
also makes sense in practice (and does not eat lots of memory and
storage space).
2019-02-13 04:24:14 +01:00
80e0b42f7d
add fixed blocksize chunker, fixes #1086
2019-02-13 04:24:14 +01:00
c4ffbd2a17
prepare to support multiple chunkers
2019-02-13 04:24:14 +01:00
de55d763a4
Merge pull request #4357 from ThomasWaldmann/xattr-tests
...
re-add the code that checks if we run under fakeroot, fixes #4291
2019-02-12 14:37:39 +01:00
12a18b955e
re-add the code that checks if we run under fakeroot, fixes #4291
...
code taken from 1.1-maint.
running as a user, with or without fakeroot does not have the test
fails in test_extract_capabilities any more.
2019-02-12 05:08:06 +01:00
9041a315ba
Merge pull request #4349 from donnyward/master
...
Update bundled lz4 code to 1.8.3, fixes #4209
2019-02-12 02:43:12 +01:00
095a109af4
Merge pull request #4347 from ThomasWaldmann/fix-fuse-test-fakeroot
...
test_mount_hardlinks: get rid of fakeroot-caused test fails, fixes #3389
2019-02-12 00:13:38 +01:00
e878ca5189
test_mount_hardlinks: get rid of fakeroot-caused test fails, fixes #3389
2019-02-11 17:47:34 +01:00
6de90d9460
Merge pull request #4354 from ThomasWaldmann/fuse-default-options-master
...
security fix: configure FUSE with "default_permissions", fixes #3903
2019-02-11 17:46:30 +01:00
672c2c99a7
security fix: configure FUSE with "default_permissions", fixes #3903
...
"default_permissions" is now enforced by borg by default to let the
kernel check uid/gid/mode based permissions.
"ignore_permissions" can be given to not enforce "default_permissions".
note: man mount.fuse explicitly tells about the security issue:
default_permissions
By default FUSE doesn't check file access permissions, ...
This option enables permission checking, restricting access
based on file mode.
This option is usually useful together with the allow_other
mount option.
We consider this a pitfall waiting for someone to fall into and this is
why we chose to change the default behaviour for borg.
2019-02-11 13:01:24 +01:00
af449cdd1f
Update bundled lz4 code to 1.8.3, fixes #4209
2019-02-10 20:36:29 -08:00
fa9d3263de
Merge pull request #4345 from ThomasWaldmann/update-scripts-master
...
signing binaries with Qubes OS support, docs
2019-02-11 02:45:46 +01:00
305efb5b01
signing binaries with Qubes OS support, docs
2019-02-11 02:44:44 +01:00
ba5185cbdb
Merge pull request #4340 from SanskritFritz/master
...
More complete shell completions for borg mount -o
2019-02-11 01:16:40 +01:00
bb98a1dae4
Option arguments for zsh tab completion
2019-02-10 20:33:23 +01:00
7becc03608
More complete shell completions for borg mount -o
2019-02-10 20:33:15 +01:00
10a4e011c0
add warnings on repository copies to avoid future problem like #4272 ( #4285 )
...
docs: add warnings about repo copies to avoid problems like #4272
2019-02-10 20:23:59 +01:00
9259000d73
Merge pull request #4335 from SanskritFritz/master
...
master - Shell completions.
2019-02-08 20:05:04 +01:00
8435206cc3
Shell completions for borg compact
2019-02-08 16:44:36 +01:00
0a97274bd1
New shell completions for borg 1.1.9
2019-02-08 14:27:14 +01:00
ff52d4225b
Added shell completions for borg help
2019-02-06 16:49:31 +01:00
2b1c223917
Merge pull request #4327 from ThomasWaldmann/forward-ports-16
...
master forward ports (16)
2019-02-05 17:08:38 +01:00
a9d562de17
work around some Microsoft WSL issues
...
see there:
https://github.com/borgbackup/borg/issues/1961
and especially there (not implemented sync_file_range):
https://github.com/Microsoft/WSL/issues/645
2019-02-05 15:15:50 +01:00
4fc23f532e
work around the backslash issue, see #4280
...
thanks to AJ Jordan @strugee for finding this.
while using \\ gives a single backslash when using "borg help pattern"
on the shell, our rendered html docs did not show the single backslash
at all.
i was lazy do dig into that issue, so i just used the word "backslash".
:-)
2019-02-05 15:08:11 +01:00
9a3fbdd751
Merge pull request #4316 from ThomasWaldmann/init-make-parentdirs
...
borg init --make-parent-dirs parent1/parent2/repo_dir, fixes #4235
2019-02-05 04:49:22 +01:00
b79c26a45f
Merge pull request #4319 from ThomasWaldmann/msgpack-061
...
support msgpack 0.6.1
2019-02-05 03:59:21 +01:00
f479850652
add comment about msgpack >= 0.6.1 features for later
2019-02-05 03:08:19 +01:00
940e107add
cleanup msgpack requirements
...
we do not expect 0.5.x with x > 6 any more.
also remove outdated comments from setup.py.
2019-02-05 02:37:22 +01:00
0ebfaa5b61
allow msgpack 0.6.1, fixes #4308
2019-02-05 02:32:36 +01:00
600e798201
borg init --make-parent-dirs parent1/parent2/repo_dir, fixes #4235
2019-02-04 17:12:11 +01:00
b204201fb5
Merge pull request #4302 from ThomasWaldmann/repair-output
...
add archive name to check --repair output, fixes #3447
2019-02-04 03:29:58 +01:00
c3f40de606
cache_sync: compute size/count stats, borg info: consider part files ( #4286 )
...
cache_sync: compute size/count stats, borg info: consider part files
fixes #3522
2019-02-04 03:26:45 +01:00
50666c3e44
Merge pull request #4315 from ThomasWaldmann/check-msgpack-master
...
check msgpack version, terminate for unsupported versions
2019-02-03 04:34:51 +01:00
de288cd6b2
Merge pull request #4314 from ThomasWaldmann/fix-tests-master
...
fix duplicate test method name, fixes #4311
2019-02-03 04:10:15 +01:00
18c9feb7e3
check msgpack version, terminate for unsupported versions
2019-02-03 03:54:41 +01:00
f4e1ba5995
fix duplicate test method name, fixes #4311
...
thanks to Anthony Sottile for finding this!
2019-02-03 02:37:49 +01:00
422d9cf170
Merge pull request #4275 from ThomasWaldmann/fix-empty-segment-crash-master
...
recover_segment: handle too small segment files correctly, see #4272
2019-02-02 00:06:44 +01:00
2d74e8e58d
Merge pull request #4301 from ThomasWaldmann/twine
...
update release workflow using twine (docs, scripts), see #4213
2019-02-01 23:46:43 +01:00
18b62f63a6
add archive name to check --repair output, fixes #3447
...
so it does not look like duplicated and also informs the user about
affected archives.
2019-02-01 23:30:45 +01:00
f7525156c1
update release workflow using twine (docs, scripts), see #4213
2019-02-01 22:07:49 +01:00
266fb02b9d
Merge pull request #4276 from ThomasWaldmann/fix-short-hostname
...
make "hostname" short, even on misconfigured systems, fixes #4262
2019-02-01 11:10:03 +01:00
Thomas Waldmann