Commit Graph

103 Commits

Author SHA1 Message Date
Tobias Klauser
c390565eef
lib/fs: Use file clone ioctl wrappers and types from golang.org/x/sys/unix (#7000)
Use the IoctlFileClone and IoctlFileCloneRange ioctl wrappers and the
FileCloneRange type provided by golang.org/x/sys/unix instead of
locally implementing them. This also allows to re-enable the code for
ppc/ppc64/ppc64le again (see commit 758a1a6a37 ("lib/fs: Disable ioctl
on ppc (fixes #6898) (#6901)")) since golang.org/x/sys/unix internally
uses the correct FICLONE and FICLONERANGE values depending on $GOARCH.
2020-09-24 10:29:32 +02:00
greatroar
3e24d82513
lib/fs: Optimize UnicodeLowercase (#6979)
Most notably, it now detects all-lowercase files and returns these
as-is. The tests have been expanded with two cases and are now used
as a benchmark (admittedly a rather trivial one).

name                           old time/op    new time/op    delta
UnicodeLowercaseMaybeChange-8    4.59µs ± 2%    4.57µs ± 1%    ~     (p=0.197 n=10+10)
UnicodeLowercaseNoChange-8       3.26µs ± 1%    3.09µs ± 1%  -5.27%  (p=0.000 n=9+10)
2020-09-11 09:16:10 +02:00
Jakob Borg
780fb3bac1
lib/fs: More efficient casefs cache (#6974)
This changes the cache to cache less things, yet retain the required
efficiency for our walk usecase. This uses less memory.

Specifically, instead of keeping result and child caches for each path
level, only keep a single cached child. In practice our operations are
depth-first, or almost depth-first, and then we retain the same hit
ratio for a smaller cache size.

I improved the benchmark so that it counts the Lstat and DirNames
operations performed, and they do not change significantly. The amount
of allocated memory is reduced by 20% and the walk itself is actually
slightly faster.

This also removes the clear based on number of cached names (as that is
not a thing any more) and the timer based clear (which was unused). This
means we'll retain the last cache state forever until it's cleared by a
write operation, but we did that before too and that state is now a lot
smaller...

The overhead compared to not using a casefs, for our typical "double
walk" (walk the tree then stat everything again) is 2x the dirnames we
would otherwise call, and no overhead on the stats (unchanged from old
implementation)

```
name                         old time/op         new time/op         delta
WalkCaseFakeFS100k/rawfs-8           306ms ± 1%          305ms ± 2%     ~     (p=0.182 n=9+10)
WalkCaseFakeFS100k/casefs-8          579ms ± 5%          557ms ± 1%   -3.77%  (p=0.000 n=10+10)

name                         old B/entry         new B/entry         delta
WalkCaseFakeFS100k/rawfs-8             590 ± 0%            590 ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8          1.09k ± 0%          0.87k ± 0%  -19.98%  (p=0.000 n=10+10)

name                         old DirNames/entry  new DirNames/entry  delta
WalkCaseFakeFS100k/rawfs-8            0.51 ± 0%           0.51 ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8           1.02 ± 0%           1.02 ± 0%     ~     (all equal)

name                         old DirNames/op     new DirNames/op     delta
WalkCaseFakeFS100k/rawfs-8           51.2k ± 0%          51.2k ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8           102k ± 0%           102k ± 0%     ~     (all equal)

name                         old Lstat/entry     new Lstat/entry     delta
WalkCaseFakeFS100k/rawfs-8            3.02 ± 0%           3.02 ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8           3.02 ± 0%           3.02 ± 0%     ~     (all equal)

name                         old Lstat/op        new Lstat/op        delta
WalkCaseFakeFS100k/rawfs-8            302k ± 0%           302k ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8           302k ± 0%           302k ± 0%     ~     (all equal)

name                         old allocs/entry    new allocs/entry    delta
WalkCaseFakeFS100k/rawfs-8            15.7 ± 0%           15.7 ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8           27.5 ± 0%           26.1 ± 0%   -5.09%  (p=0.000 n=10+10)

name                         old ns/entry        new ns/entry        delta
WalkCaseFakeFS100k/rawfs-8           2.02k ± 1%          2.02k ± 2%     ~     (p=0.163 n=9+10)
WalkCaseFakeFS100k/casefs-8          3.83k ± 5%          3.68k ± 1%   -3.77%  (p=0.000 n=10+10)

name                         old alloc/op        new alloc/op        delta
WalkCaseFakeFS100k/rawfs-8          89.2MB ± 0%         89.2MB ± 0%     ~     (p=0.364 n=9+10)
WalkCaseFakeFS100k/casefs-8          164MB ± 0%          131MB ± 0%  -19.97%  (p=0.000 n=10+10)

name                         old allocs/op       new allocs/op       delta
WalkCaseFakeFS100k/rawfs-8           2.38M ± 0%          2.38M ± 0%     ~     (all equal)
WalkCaseFakeFS100k/casefs-8          4.16M ± 0%          3.95M ± 0%   -5.05%  (p=0.000 n=10+10)
```
2020-09-09 14:38:39 +02:00
Simon Frei
e3cd9219b8
lib/model: Don't fail over case-conflict on tempfile (fixes #6973) (#6975) 2020-09-09 11:47:14 +02:00
Simon Frei
b628ec5054
lib/fs: Use generic case resolver on windows (fixes #6968) (#6969) 2020-09-08 13:10:26 +02:00
Audrius Butkevicius
d507d932b8
all: Use protobuf to generate config structs (fixes #6734) (#6900) 2020-08-25 08:11:14 +02:00
Simon Frei
fc2c46e82f
lib/config, lib/fs: Make junction behaviour configurable (ref #6606) (#6907) 2020-08-19 19:58:51 +02:00
Audrius Butkevicius
758a1a6a37
lib/fs: Disable ioctl on ppc (fixes #6898) (#6901) 2020-08-19 07:56:35 +02:00
Audrius Butkevicius
e9bb17307d
lib/fs: Unwrap mtimeFile, get fd the "correct" way (ref #6875) (#6877) 2020-08-07 07:47:48 +02:00
Jakob Borg
5e1f39b6f6 lib/fs: Fix WatchRename test for FreeBSD (fixes #6613) 2020-08-03 23:24:01 +02:00
Simon Frei
932d8c69de lib/fs: Properly handle case insensitive systems (fixes #1787, fixes #2739, fixes #5708)
With this change we emulate a case sensitive filesystem on top of
insensitive filesystems. This means we correctly pick up case-only renames
and throw a case conflict error when there would be multiple files differing
only in case.

This safety check has a small performance hit (about 20% more filesystem
operations when scanning for changes). The new advanced folder option
`caseSensitiveFS` can be used to disable the safety checks, retaining the
previous behavior on systems known to be fully case sensitive.

Co-authored-by: Jakob Borg <jakob@kastelo.net>
2020-07-28 11:15:11 +02:00
Simon Frei
ebead944b5
lib/fs: Pass infinite recursion error on instead of warning (#6846)
Prompted by https://forum.syncthing.net/t/infinite-filesystem-recursion-detected/15285. In my opinion the filesystem shouldn't throw warnings but pass on errors for the caller to decide what's to be happening with it. Right now in this PR an infinite recursion is a normal scan error, i.e. folder is in failed state and displays failed items, but no warning. I think that's appropriate but if deemed appropriate an additional warning can be thrown in the scanner.
2020-07-22 22:10:24 +02:00
Audrius Butkevicius
deaccc7f8d
lib/fs: Add support for Windows duplicate extents (#6764) 2020-06-18 22:32:26 +01:00
Audrius Butkevicius
4812fd3ec1
all: Add copy-on-write filesystem support (fixes #4271) (#6746) 2020-06-18 08:15:47 +02:00
Simon Frei
6976219d6d
lib/model: Check dir before deletion when pulling (#6741) 2020-06-16 15:20:08 +02:00
xarx00
f05ccd775a
lib/fs: Set execute bits on junctions converted to dirs (ref #6606) (#6645) 2020-05-14 08:09:58 +02:00
Simon Frei
074097a8e7
lib/fs: Prevent race-detector triggering in tests (fixes #6608) (#6642) 2020-05-14 00:30:09 +02:00
xarx00
ee445e35a0
lib/fs: Treat Windows junctions as normal directories (#6606)
Fixes #1830, presumably.
2020-05-13 21:46:24 +02:00
Jakob Borg
0e2a07d71a lib/fs: Avoid dirty offset read in fakefs (fixes #6584) 2020-04-28 09:58:31 +02:00
Jakob Borg
7956e7d0ef
lib/{fs,scanner}: gofmt from Go 1.14 (#6509) 2020-04-07 09:31:29 +02:00
Simon Frei
123941cf62
lib/fs: Basic with empty root shouldn't point at / (#6483) 2020-03-31 13:56:07 +02:00
Jakob Borg
dd92b2b8f4
all: Tweak error creation (#6391)
- In the few places where we wrap errors, use the new Go 1.13 "%w"
  construction instead of %s or %v.

- Where we create errors with constant strings, consistently use
  errors.New and not fmt.Errorf.

- Remove capitalization from errors in the few places where we had that.
2020-03-03 22:40:00 +01:00
Jakob Borg
0df39ddc72
lib/fs, lib/model: Rewrite RecvOnly tests (#6318)
During some other work I discovered these tests weren't great, so I've
rewritten them to be a little better. The real changes here are:

- Don't play games with not starting the folder and such, and don't
  construct a fake folder instance -- just use the one the model has. The
  folder starts and scans but the folder contents are empty at this point
  so that's fine.

- Use a fakefs instead of a temp dir.

- To support the above, implement a fakefs option `?content=true` to
  make the fakefs actually retain written content. Use sparingly,
  obviously, but it means the fakefs can usually be used instead of an
  on disk real directory.
2020-02-12 07:47:05 +01:00
Jakob Borg
e82a7e3dfa
all: Propagate errors from NamespacedKV (#6203)
As foretold by the prophecy, "once the database refactor is merged, then
shall appear a request to propagate errors from the store known
throughout the land as the NamedspacedKV, and it shall be good".
2019-11-30 13:03:24 +01:00
Evgeny Kuznetsov
1c277fc096 lib/fs: Add case-insensitive fakefs (#6074) 2019-11-29 09:17:42 +01:00
Jakob Borg
c71116ee94
Implement database abstraction, error checking (ref #5907) (#6107)
This PR does two things, because one lead to the other:

- Move the leveldb specific stuff into a small "backend" package that
defines a backend interface and the leveldb implementation. This allows,
potentially, in the future, switching the db implementation so another
KV store should we wish to do so.

- Add proper error handling all along the way. The db and backend
packages are now errcheck clean. However, I drew the line at modifying
the FileSet API in order to keep this manageable and not continue
refactoring all of the rest of Syncthing. As such, the FileSet methods
still panic on database errors, except for the "database is closed"
error which is instead handled by silently returning as quickly as
possible, with the assumption that we're anyway "on the way out".
2019-11-29 09:11:52 +01:00
Simon Frei
90d85fd0a2
lib: Replace done channel with contexts in and add names to util services (#6166) 2019-11-21 08:41:15 +01:00
Jakob Borg
1d99e5277a all: Cleanups enabled by Go 1.12 2019-11-10 10:16:10 +01:00
Lukas Lihotzki
96bb1c8e29 all, lib/logger: Refactor SetDebug calls (#6054) 2019-10-04 13:03:34 +02:00
Jakob Borg
52d80d8144 lib/fs: Improve root check (#6033)
The root check would allow things like c:\foobar\baz if the root was
c:\foo, because string wise that's a prefix. Now it doesn't.
2019-09-29 23:38:11 +08:00
Simon Frei
35b699dc77 lib/fs: Check events against both the user and eval root (#6013) 2019-09-22 08:03:22 +01:00
Simon Frei
1d182e4631 lib/fs: Use gopsutils for disk usage (#5905) 2019-07-29 20:06:17 +02:00
Simon Frei
5de9b677c2
lib/fs: Fix kqueue event list (fixes #5308) (#5885) 2019-07-23 14:11:15 +02:00
Audrius Butkevicius
b7c70a9817 lib/fs: Enhance mtimefs, use everywhere (fixes #5777) (#5776) 2019-06-11 08:27:12 +02:00
Mingxuan Lin
eb4fe808c5 lib/fs: Fallback EvalSymlinks method on windows (fixes #5609) (#5611) 2019-06-10 14:33:53 +02:00
Simon Frei
486230768e lib/fs, lib/model: Add error channel to Watch to avoid panics (fixes #5697) (#5734)
* lib/fs, lib/model: Add error channel to Watch to avoid panics (fixes #5697)

* forgot unsupported watch

* and more non(-standard)-unixy fixes

* and windows test

* review
2019-05-25 20:08:26 +01:00
Simon Frei
dfbbb286fc lib/fs: Consider win83 for root path as well when watching (ref #5706) (#5709) 2019-05-11 10:06:04 +02:00
Simon Frei
2558b021e5 lib/fs: Remove \\?\ for drive letters when watching (fixes #5578) (#5701) 2019-05-10 09:09:58 +02:00
Jakob Borg
62a6d619e7 Merge branch 'release'
* release:
  lib/fs: Revert removal of \\?\ for drive letters (fixes #5695)
2019-05-08 17:38:08 +02:00
Jakob Borg
a04fcfe749 lib/fs: Revert removal of \\?\ for drive letters (fixes #5695)
This reverts commit ca823bd591 from #5633.

This un-fixes bug #5578.
2019-05-08 17:31:52 +02:00
Audrius Butkevicius
0ca1f26ff8
lib/versioner: Restore for all versioners, cross-device support (#5514)
* lib/versioner: Restore for all versioners, cross-device support

Fixes #4631
Fixes #4586
Fixes #1634
Fixes #5338
Fixes #5419
2019-04-28 23:30:16 +01:00
Simon Frei
e302ccf4b4
lib/fs: Improve IsParent (#5658) 2019-04-22 11:12:32 +02:00
Simon Frei
79360e2205
lib/fs: Add test that symlinks are skipped on walk (#5644) 2019-04-10 22:36:37 +02:00
Simon Frei
ca823bd591
lib/fs: When watching remove \\?\ for drive letters (fixes #5578) (#5633) 2019-04-09 09:02:04 +02:00
Simon Frei
ad5a046843
lib/fs: Rename fsFile* to basicFile* (#5546) 2019-02-24 18:02:02 +01:00
Jakob Borg
c2ddc83509 all: Revert the underscore sillyness 2019-02-02 12:16:27 +01:00
Jakob Borg
9fd270d78e
all: A few more interesting linter fixes (#5502)
A couple of minor bugs and simplifications
2019-02-02 12:09:07 +01:00
Jakob Borg
df5c1eaf01
all: Bunch of more linter fixes (#5500) 2019-02-02 11:02:28 +01:00
Jakob Borg
75dcff0a0e
all: Copy owner/group from parent (fixes #5445) (#5479)
This adds a folder option "CopyOwnershipFromParent" which, when set,
makes Syncthing attempt to retain the owner/group information when
syncing files. Specifically, at the finisher stage we look at the parent
dir to get owner/group and then attempt a Lchown call on the temp file.
For this to succeed Syncthing must be running with the appropriate
permissions. On Linux this is CAP_FOWNER, which can be granted by the
service manager on startup or set on the binary in the filesystem. Other
operating systems do other things, but often it's not required to run as
full "root". On Windows this patch does nothing - ownership works
differently there and is generally less of a deal, as permissions are
inherited as ACLs anyway.

There are unit tests on the Lchown functionality, which requires the
above permissions to run. There is also a unit test on the folder which
uses the fake filesystem and hence does not need special permissions.
2019-01-25 09:52:21 +01:00
Audrius Butkevicius
96bd691f55 lib/fs: Skip some tests on OpenBSD (fixes #5077) (#5466) 2019-01-19 08:28:57 +01:00