52f1cd4bc6
argparse: work around bad default in py 3.7.0a/b/rc, fixes #3996
2018-07-30 01:04:14 +01:00
eb6481e8d3
Merge pull request #3998 from ThomasWaldmann/py37
...
remove loggerDict.clear() from tearDown method, fixes #3805
2018-07-29 14:11:05 +01:00
a7dd2560ba
remove loggerDict.clear() from tearDown method, fixes #3805
...
It causes problems with the new caching in the py37 logger module.
Removing loggerDict.clear() fixes this and makes the tests work again.
Also, it does not seem to have any negative effect, neither on py36
nor on py37.
See also: https://bugs.python.org/issue34269
2018-07-29 12:29:26 +01:00
34227430d4
Installation from source - setting python interpreter version ( #3995 )
...
explicit instructions for setting python interpreter when installing from source via git
2018-07-29 09:50:09 +01:00
2a4ed3f0e8
Note regarding borg version ( #3987 )
...
add a note regarding in-file checkpointing and borg versions
2018-07-20 16:25:16 +02:00
b5e11389fc
Merge pull request #3984 from ThomasWaldmann/locking-fixes-master
...
Locking: better logging, some asserts (fwd port to master)
2018-07-19 22:04:13 +02:00
3417de85e0
locking: more logging
...
(cherry picked from commit f86bf8a0a8
2018-07-17 19:06:45 +02:00
a51fca49bf
process_alive: add some assertions
...
(cherry picked from commit 2bf1b29e8e
2018-07-17 19:06:26 +02:00
a7e408db61
Merge pull request #3980 from ThomasWaldmann/update-coafile
...
coala: exclude some more stuff, move ignores to "all" section
2018-07-17 01:16:22 +02:00
ffd78493aa
Merge pull request #3983 from ThomasWaldmann/fix-cache-lock-timeout-master
...
cache lock: use lock_wait everywhere to fix infinite wait (fwd port to master)
2018-07-17 01:15:45 +02:00
89c11f45ce
cache lock: use lock_wait everywhere to fix infinite wait
...
also: clarify docs
(cherry picked from commit 2f3e60d9d5
2018-07-16 23:50:04 +02:00
9115ad58d9
Merge pull request #3970 from ThomasWaldmann/compact-commits2
...
compact commits / separate compaction
2018-07-14 23:47:47 +02:00
318cfffde1
update CHANGES about compaction changes
2018-07-14 19:15:09 +02:00
36cbc2d18f
coala: ignore some stuff it warns about
2018-07-14 18:40:50 +02:00
d5152d8f2b
coala: exclude some more stuff, move ignores to "all" section
2018-07-14 18:28:18 +02:00
a8d52351bb
update man pages / usage docs
2018-07-14 14:56:29 +02:00
e6fcf4ea42
update docs about separated compaction
2018-07-14 14:53:23 +02:00
a221ca16ad
remote: introduce a dontcare flag (all values acceptable)
...
in some cases the value of a new parameter does not really matter,
as long as it works / there is no malfunction.
without this change, a new borg client could not commit to an old
borg server (and would fail late when trying to commit):
- when using compact=False as it does at most places
- when using cleanup_commits=True as a user might trigger by running
borg compact --cleanup-commits
we never want a borg commit "fail late", leaving the repo in a state
with uncommitted data, thus the server shall just commit, no matter
how it handles compaction or whether it can do the cleanup.
2018-07-14 14:29:28 +02:00
3715d2da3e
slightly refactor write_commit using new "want_new" flag
2018-07-14 14:29:28 +02:00
1f387d911a
start new segment file for put/del to MANIFEST_ID
...
specialcase deleting / writing the manifest to be in a separate, new
segment file, so that when we supersede and compact it later, less
segment data has to be shuffled around - compaction can then just
delete this segment file and that's all.
2018-07-14 14:29:28 +02:00
755eaeec0a
borg compact --cleanup-commits to get rid of leftover 17byte segments
...
see #2850 .
2018-07-14 14:29:28 +02:00
efe60bd640
test fixes
2018-07-14 14:29:28 +02:00
aeef082483
repository: track commits in hints
2018-07-14 14:29:28 +02:00
de4afa097c
separate borg compact command, fixes #2195
2018-07-14 14:29:28 +02:00
d2a816d0d4
Merge pull request #3978 from ThomasWaldmann/fix-nanorst
...
nanorst: add missing blank to exception message
2018-07-13 11:38:52 +02:00
0d06407087
nanorst: add missing blank to exception message
2018-07-13 10:41:32 +02:00
77f8a82d3b
Merge pull request #3976 from intelfx/xattr-fix-include-path
...
xattr: use sys/xattr.h as per setxattr(2)
2018-07-10 14:28:24 +02:00
074a984111
xattr: use sys/xattr.h as per setxattr(2), attr/xattr.h is absent on my system
2018-07-10 13:08:59 +03:00
1f75b6b9b5
Merge pull request #3971 from ThomasWaldmann/more-xattr-stuff
...
xattrs: use follow_symlinks=False as default, more tests
2018-07-08 15:41:55 +02:00
44b4f9645d
xattrs: use follow_symlinks=False as default, more tests
...
it's like we do it everywhere else because this is what we usually need.
2018-07-08 15:08:01 +02:00
2855a3798b
Merge pull request #3972 from ThomasWaldmann/fix-ext-version-check
...
fix logic bug in platform module API version check
2018-07-08 15:07:01 +02:00
2bb23f00af
fix logic bug in platform module API version check
2018-07-08 14:48:16 +02:00
ff3d75fed2
Merge pull request #3969 from ThomasWaldmann/update-changes-master
...
Update CHANGES (master)
2018-07-08 00:25:09 +02:00
3fd6bd8ca7
update CHANGES: added 1.1 history (master)
2018-07-08 00:18:09 +02:00
a89aadc9f2
update CHANGES (master)
2018-07-08 00:13:13 +02:00
20abc9d687
Merge pull request #3963 from ThomasWaldmann/chunker-fd-only
...
Chunker: give os-level fd only
2018-07-07 18:38:30 +02:00
13e6970437
create: do not give chunker a py file object, it is not needed
...
the os level file handle is enough, the chunker will prefer it if
valid and won't use the file obj, so we can give None there.
this saves these unneeded syscalls:
fstat(5, {st_mode=S_IFREG|0664, st_size=227063, ...}) = 0
ioctl(5, TCGETS, 0x7ffd635635f0) = -1 ENOTTY (Inappropriate ioctl for device)
lseek(5, 0, SEEK_CUR) = 0
2018-07-07 18:06:57 +02:00
8e91694c54
Merge pull request #3961 from ThomasWaldmann/flags-fd-based
...
make bsdflags linux code fd-based
2018-07-07 18:05:58 +02:00
018b62c845
bsdflags: use fd instead of path
...
this optimization is only needed for linux, the bsd-like platforms
do not need an open file to run a ioctl against, but have bsdflags
in the stat result already.
on linux, this optimization saves 1 file open/close per input file.
2018-07-07 17:30:17 +02:00
2dd22df751
Merge pull request #3960 from ThomasWaldmann/acl-fd-based
...
ACLs fd-based
2018-07-07 17:28:11 +02:00
7e47e68e29
acls: use fd instead of path
2018-07-07 17:02:37 +02:00
5bf5f12be1
acls: only calls os.fsencode for str paths
2018-07-07 16:54:19 +02:00
34cd1b22ec
Merge pull request #3918 from ThomasWaldmann/platform-xattr
...
xattr: move to platform pkg, use cython, use bytes, use fd
2018-07-07 16:50:08 +02:00
34a51eb958
xattr: fix dummy base getxattr, must raise for any given name
2018-07-07 15:47:56 +02:00
113b0eabec
xattr: use fd for get_all
...
when processing regular files, use a fd to query xattrs.
when the file was modified and we chunked it, we have it open anyways.
if not, we open the file once and then query xattrs, in the hope that
this is more efficient than the path based calls.
guess it is less prone to race conditions in any case.
2018-07-07 15:47:56 +02:00
389e806b68
xattr: refactor/fix exception handling in get_all
...
if there is EPERM just for a single attr, it now still collects the
other attrs (previous behaviour: just return empty xattrs dict).
2018-07-07 15:47:56 +02:00
394d59e6d8
xattr: implement set_all to complement get_all
...
also: follow_symlinks param defaults to False (we do never use True)
fix tests, xattrs are set via FD now.
2018-07-07 15:47:56 +02:00
c29c3063b0
xattr: use bytes typed path for listxattr, getxattr, setxattr
2018-07-07 15:47:56 +02:00
9deb90db71
xattr: use bytes typed names for listxattr, getxattr, setxattr
2018-07-07 15:47:56 +02:00
b5a9ac5682
xattr: use bytes typed values for listattr, getxattr, setxattr
...
- getxattr should only return bytes, not None
- setxattr should not get a None value, just bytes
- remove unneeded tmp vars
2018-07-07 15:47:56 +02:00
Thomas Waldmann