Age | Commit message (Collapse) | Author |
|
The files are apparently processed in alphabetic order, so cli_args.go is
processed before main.go. In order to run before the go-fuse imports, put
the 'ensure fds' code in a separate package. Debug messages are omitted
to avoid additional imports (that might contain other code messing up our
file descriptors).
|
|
The Go stdlib, as well as the gocryptfs code, relies on the fact
that fds 0,1,2 are always open.
See https://github.com/rfjakob/gocryptfs/issues/320 for details.
|
|
Setting/removing extended attributes on directories was partially fixed with
commit eff35e60b63331e3e10f921792baa10b236a721d. However, on most file systems
it is also possible to do these operations without read access (see tests).
Since we cannot open a write-access fd to a directory, we have to use the
/proc/self/fd trick (already used for ListXAttr) for the other operations aswell.
For simplicity, let's separate the Linux and Darwin code again (basically revert
commit f320b76fd189a363a34bffe981aa67ab97df3362), and always use the
/proc/self/fd trick on Linux. On Darwin we use the best-effort approach with
openBackingFile() as a fallback.
More discussion about the available options is available in
https://github.com/rfjakob/gocryptfs/issues/308.
|
|
|
|
As the dirCache now has 3 entries, the tests should accept
up to 3 extra fds without declaring an fd leak.
|
|
The missing break meant that we may find a second
hit in the cache, Dup() a second fd, and leak the first
one.
Thanks @slackner for finding this.
|
|
Otherwise this can happen, as triggered by xfstests generic/011:
go-fuse: can't convert error type: openat failed: too many open files
The app then gets a misleading "Function not implemented" error.
|
|
We alread have this warning in Open(), but xfstests generic/488
causes "too many open files" via Create. Add the same message so
the user sees what is going on.
|
|
Directories cannot be opened read-write. Retry with RDONLY.
|
|
This was inadvertedly kept enabled after benchmarking.
|
|
3 entries should work well for up to three parallel users.
It works well for extractloop.bash (two parallel tar extracts).
|
|
Go version go1.10.7 linux/amd64 complains with:
internal/fusefrontend_reverse/rfs.go:333: declaration of "entries" shadows
declaration at internal/fusefrontend_reverse/rfs.go:327
|
|
Fixes https://github.com/rfjakob/gocryptfs/issues/286 :
While the actual file is properly excluded, the * .name file is still leaked in the directory listing:
```
drwxr-xr-x 2 sebastian sebastian 4,0K Dez 17 14:58 .
drwxr-xr-x 7 sebastian sebastian 4,0K Dez 17 14:45 ..
-r-------- 1 sebastian sebastian 408 Dez 17 14:56 gocryptfs.conf
-r--r--r-- 1 sebastian sebastian 16 Dez 17 14:58 gocryptfs.diriv
-r--r--r-- 1 sebastian sebastian 320 Dez 17 14:58 gocryptfs.longname.3vZ_r3eDPb1_fL3j5VA4rd_bcKWLKT9eaxOVIGK5HFA.name
```
|
|
Excluded files showed up in directory listing like this:
drwxr-xr-x 2 sebastian sebastian 4,0K Dez 17 14:48 .
drwxr-xr-x 7 sebastian sebastian 4,0K Dez 17 14:45 ..
-????????? ? ? ? ? ? abcd
-r-------- 1 sebastian sebastian 366 Dez 17 14:45 gocryptfs.conf
Fixes https://github.com/rfjakob/gocryptfs/issues/285
|
|
When the old size is zero, there are no existing blocks to merge the
new data with. Directly use Ftruncate if the size is block-aligned.
Fixes https://github.com/rfjakob/gocryptfs/issues/305
|
|
Found with the 'codespell' utility.
|
|
When O_PATH is specified in flags, flag bits other than O_CLOEXEC, O_DIRECTORY,
and O_NOFOLLOW are ignored.
|
|
|
|
|
|
|
|
We already do 'defer fs.dirCache.Clear()', so this is no longer required.
|
|
|
|
Also remove some unnecessary flags: When O_PATH is specified in flags, flag
bits other than O_CLOEXEC, O_DIRECTORY, and O_NOFOLLOW are ignored.
|
|
This code was accidentially added in 4f66d66755da63c78b09201c6c72353009251cf2.
|
|
Bug looked like this:
$ ls -l .
total 0
drwxrwxr-x. 2 jakob jakob 60 Jan 3 15:42 foo
-rw-rw-r--. 1 jakob jakob 0 Jan 3 15:46 x
$ ls -l .
ls: cannot access '.': No such file or directory
(only happened when "" was in the dirCache)
|
|
|
|
Un-spaghettify the function and let the callers open
the directory.
|
|
Copy-paste error.
https://github.com/rfjakob/gocryptfs/issues/308
|
|
This function is in all fastpaths, will get a cache, and needs
its own file.
renamed: internal/fusefrontend/names.go -> internal/fusefrontend/openbackingdir.go
renamed: internal/fusefrontend/names_test.go -> internal/fusefrontend/openbackingdir_test.go
|
|
Finally allows us to delete EncryptPathDirIV.
|
|
* listxattr is fixed via the /proc/self/fd trick
* setxattr,removexattr are fixed by opening the file O_WRONLY
Fixes https://github.com/rfjakob/gocryptfs/issues/308
|
|
|
|
Darwin now also has these functions, use them. Simplifies
the code and makes it symlink-safe on Darwin as well.
|
|
These take care of buffer sizing and parsing.
|
|
Reported by @slackner at
https://github.com/rfjakob/gocryptfs/commit/932efbd4593fe6be6c86f0dafeaea32910b7c246#r31813373
thanks!
|
|
Pread() needs retry logic, so instead of implementing it ourselves,
use os.File.
Reported by @slackner at
https://github.com/rfjakob/gocryptfs/commit/c09bf1f2284706232642431c75fa1f3d8500a9d0#r31813394
|
|
|
|
Thanks @slackner!
Fixes https://github.com/rfjakob/gocryptfs/issues/306
|
|
An Open() a fifo blocks until it is opened for writing.
This meant that xattr operations on FIFOs would block.
Pass O_NONBLOCK to fix that, and add a test.
|
|
This function is NOT symlink-safe. Darwin needs it because it lacks
fgetxattr(2) and friends.
|
|
Uses /proc/self/fd.
|
|
Uses /proc/self/fd on Linux.
|
|
Uses the /proc/self/fd trick.
|
|
Uses the /proc/self/fd trick, which does not work
on Darwin.
|
|
unix.UtimesNanoAt now also exists on Darwin, yay!
|
|
Failure was:
+ GOOS=darwin
+ GOARCH=amd64
+ go build -tags without_openssl
# github.com/rfjakob/gocryptfs/internal/fusefrontend
internal/fusefrontend/fs_dir.go:159:60: cannot use origMode | 448 (type uint16) as type uint32 in argument to syscallcompat.Fchmodat
internal/fusefrontend/fs_dir.go:170:33: cannot use origMode (type uint16) as type uint32 in argument to syscallcompat.Fchmodat
|
|
No changes needed.
|
|
Now uses Unlinkat.
|
|
And also rename DeleteLongName() -> DeleteLongNameAt(). The
naming follow the names open the openat() etc syscalls.
|
|
Now symlink-safe through Readlinkat().
|