Old XFS filesystems always return DT_UNKNOWN. Downgrade the message
to "info" level if we are on XFS.
Using the "warning" level means that users on old XFS filesystems
cannot run the test suite as it intentionally aborts on any
warnings.
Fixes https://github.com/rfjakob/gocryptfs/issues/267
When gocryptfs was started on a terminal and later
daemonized, the color codes stayed active in the syslog
output.
The codes are not visible in "journalctl -f", which is why
I have not noticed it yet, but they do show up in normal
syslog as the usual "#033[33m" crap.
Even though filesystem notifications aren't implemented for FUSE, I decided to
try my hand at implementing the autounmount feature (#128). I based it on the
EncFS autounmount code, which records filesystem accesses and checks every X
seconds whether it's idled long enough to unmount.
I've tested the feature locally, but I haven't added any tests for this flag.
I also haven't worked with Go before. So please let me know if there's
anything that should be done differently.
One particular concern: I worked from the assumption that the open files table
is unique per-filesystem. If that's not true, I'll need to add an open file
count and associated lock to the Filesystem type instead.
https://github.com/rfjakob/gocryptfs/pull/265
Error was:
# github.com/rfjakob/gocryptfs/internal/fusefrontend
internal/fusefrontend/fs.go:179: cannot use perms | 256 (type uint16) as type uint32 in argument to syscall.Fchmod
internal/fusefrontend/fs.go:185: cannot use perms (type uint16) as type uint32 in argument to syscall.Fchmod
Rename openBackingPath to openBackingDir and use OpenDirNofollow
to be safe against symlink races. Note that openBackingDir is
not used in several important code paths like Create().
But it is used in Unlink, and the performance impact in the RM benchmark
to be acceptable:
Before
$ ./benchmark.bash
Testing gocryptfs at /tmp/benchmark.bash.bYO: gocryptfs v1.6-12-g930c37e-dirty; go-fuse v20170619-49-gb11e293; 2018-09-08 go1.10.3
WRITE: 262144000 bytes (262 MB, 250 MiB) copied, 1.07979 s, 243 MB/s
READ: 262144000 bytes (262 MB, 250 MiB) copied, 0.882413 s, 297 MB/s
UNTAR: 16.703
MD5: 7.606
LS: 1.349
RM: 3.237
After
$ ./benchmark.bash
Testing gocryptfs at /tmp/benchmark.bash.jK3: gocryptfs v1.6-13-g84d6faf-dirty; go-fuse v20170619-49-gb11e293; 2018-09-08 go1.10.3
WRITE: 262144000 bytes (262 MB, 250 MiB) copied, 1.06261 s, 247 MB/s
READ: 262144000 bytes (262 MB, 250 MiB) copied, 0.947228 s, 277 MB/s
UNTAR: 17.197
MD5: 7.540
LS: 1.364
RM: 3.410
The function used to do two things:
1) Walk the directory tree in a manner safe from symlink attacks
2) Open the final component in the mode requested by the caller
This change drops (2), which was only used once, and lets the caller
handle it. This simplifies the function and makes it fit for reuse in
forward mode in openBackingPath(), and for using O_PATH on Linux.
These were silently ignored until now (!) but
are rejected by Go 1.11 stdlib.
Drop the flags so the tests work again, until
we figure out a better solution.
https://github.com/golang/go/issues/20130
Show enable_trezor in the version string if we were compiled
with `-tags enable_trezor`. And hide the `-trezor` flag from
the help output if we were not.
As uncovered by xfstests generic/465, concurrent reads and writes
could lead to this,
doRead 3015532: corrupt block #1039: stupidgcm: message authentication failed,
as the read could pick up a block that has not yet been completely written -
write() is not atomic!
Now writes take ContentLock exclusively, while reads take it shared,
meaning that multiple reads can run in parallel with each other, but
not with a write.
This also simplifies the file header locking.
xfstests generic/083 fills the filesystem almost completely while
running fsstress in parallel. In fsck, these would show up:
readFileID 2580: incomplete file, got 18 instead of 19 bytes
This could happen when writing the file header works, but writing
the actual data fails.
Now we kill the header again by truncating the file to zero.
If the underlying filesystem is full, writing to gocryptfs.diriv may
fail, and later fsck show this:
OpenDir "xyz": could not read gocryptfs.diriv: wanted 16 bytes, got 0
Uncovered by xfstests generic/083.
Also fixes a fd leak in the error path.
If the underlying filesystem is full, it is normal get ENOSPC here.
Log at Info level instead of Warning.
Fixes xfstests generic/015 and generic/027, which complained about
the extra output.
O_DIRECT accesses must be aligned in both offset and length. Due to our
crypto header, alignment will be off, even if userspace makes aligned
accesses. Running xfstests generic/013 on ext4 used to trigger lots of
EINVAL errors due to missing alignment. Just fall back to buffered IO.
The trezor libraries are not yet stable enough to build
gocryptfs with trezor support by default.
It does not even compile at the moment:
$ ./build.bash -tags enable_trezor
# github.com/conejoninja/tesoro/vendor/github.com/trezor/usbhid
../../conejoninja/tesoro/vendor/github.com/trezor/usbhid/hid.go:32:11: fatal error: os/threads_posix.c: No such file or directory
#include "os/threads_posix.c"
^~~~~~~~~~~~~~~~~~~~
compilation terminated.
https://github.com/conejoninja/tesoro/issues/9
"gocryptfs -fsck" will need access to helper functions,
and to get that, it will need to cast a gofuse.File to a
fusefrontend.File. Make fusefrontend.File exported to make
this work.
TrezorPayload stores 32 random bytes used for unlocking
the master key using a Trezor security module. The randomness makes sure
that a unique unlock value is used for each gocryptfs filesystem.
We are clean again.
Warnings were:
internal/fusefrontend/fs.go:443:14: should omit type string from declaration
of var cTarget; it will be inferred from the right-hand side
internal/fusefrontend/xattr.go:26:1: comment on exported method FS.GetXAttr
should be of the form "GetXAttr ..."
internal/syscallcompat/sys_common.go:9:7: exported const PATH_MAX should have
comment or be unexported
Reading system.posix_acl_access and system.posix_acl_default
should return EOPNOTSUPP to inform user-space that we do not
support ACLs.
xftestest essientially does
chacl -l | grep "Operation not supported"
to determine if the filesystem supports ACLs, and used to
wrongly believe that gocryptfs does.
mv is unhappy when we return EPERM when it tries to set
system.posix_acl_access:
mv: preserving permissions for ‘b/x’: Operation not permitted
Now we return EOPNOTSUPP like tmpfs does and mv seems happy.
Values a binary-safe, there is no need to base64-encode them.
Old, base64-encoded values are supported transparently
on reading. Writing xattr values now always writes them binary.
We previously returned EPERM to prevent the kernel from
blacklisting our xattr support once we get an unsupported
flag, but this causes lots of trouble on MacOS:
Cannot save files from GUI apps, see
https://github.com/rfjakob/gocryptfs/issues/229
Returning ENOSYS triggers the dotfiles fallback on MacOS
and fixes the issue.
OpenDir and ListXAttr skip over corrupt entries,
readFileID treats files the are too small as empty.
This improves usability in the face of corruption,
but hides the problem in a log message instead of
putting it in the return code.
Create a channel to report these corruptions to fsck
so it can report them to the user.
Also update the manpage and the changelog with the -fsck option.
Closes https://github.com/rfjakob/gocryptfs/issues/191
Most corruption cases except xattr should be covered.
With test filesystem.
The output is still pretty ugly. xattr support will
be added in the next commits.
This should not happen via FUSE as the kernel caps the size,
but with fsck we have the first user that calls Read directly.
For symmetry, check it for Write as well.
A few places have called tlog.Warn.Print, which directly
calls into log.Logger due to embedding, losing all features
of tlog.
Stop embedding log.Logger to make sure the internal functions
cannot be called accidentially and fix (several!) instances
that did.
We now print the number in a debug message, so define
the numeric values explicitely instead of using iota.
This way you don't have to understand how iota works
to find out what the number means. Lack of understanding
of how iota works is also the reason why the numbers
start at 3 (to keep the current behavoir).
Overwrite the masterkey with zeros once we
have encrypted it, and let it run out of scope.
Also get rid of the password duplicate in
readpassword.Twice.
As soon as we don't need them anymore, overwrite
keys with zeros and make sure they run out of scope
so we don't create a risk of inadvertedly using all-zero
keys for encryption.
https://github.com/rfjakob/gocryptfs/issues/211
Both fusefrontend and fusefrontend_reverse were doing
essentially the same thing, move it into main's
initFuseFrontend.
A side-effect is that we have a reference to cryptocore
in main, which will help with wiping the keys on exit
(https://github.com/rfjakob/gocryptfs/issues/211).
$ go.gcc build
# github.com/rfjakob/gocryptfs/internal/syscallcompat
internal/syscallcompat/unix2syscall_linux.go:32:13: error: incompatible types in assignment (cannot use type int64 as type syscall.Timespec_sec_t)
s.Atim.Sec = u.Atim.Sec
^
On mips64le, syscall.Getdents() and struct syscall.Dirent do
not fit together, causing our Getdents implementation to
return garbage ( https://github.com/rfjakob/gocryptfs/issues/200
and https://github.com/golang/go/issues/23624 ).
Switch to unix.Getdents which does not have this problem -
the next Go release with the syscall package fixes is too
far away, and will take time to trickle into distros.
Steps to reproduce:
* Create a regular reverse mount point
* Create a file "test" in the original directory
* Access the corresponding encrypted directory in the mount point (ls <encrypted dir>)
* Quickly delete the file in the original data - instead create a device node
* Access the file again, it will access the device node and attempt to read from it
Fixes https://github.com/rfjakob/gocryptfs/issues/187
Unfortunately, faccessat in Linux ignores AT_SYMLINK_NOFOLLOW,
so this is not completely atomic.
Given that the information you get from access is not very
interesting, it seems good enough.
https://github.com/rfjakob/gocryptfs/issues/165
Add faccessat(2) with a hack for symlink, because the
kernel does not actually looks at the passed flags.
From man 2 faccessat:
C library/kernel differences
The raw faccessat() system call takes only the first three argu‐
ments. The AT_EACCESS and AT_SYMLINK_NOFOLLOW flags are actually
implemented within the glibc wrapper function for faccessat().
...when opening intermedia directories to give us an
extra layer of safety.
From the FreeBSD man page:
This flag can be used to prevent applications with elevated
privileges from opening files which are even unsafe to open with O_RDONLY,
such as device nodes.
...by using the new OpenNofollow helper.
The benchmark shows a small but acceptable performance loss:
$ ./benchmark-reverse.bash
LS: 2.182
CAT: 18.221
Tracking ticket: https://github.com/rfjakob/gocryptfs/issues/165
Now that we have Fstatat we can use it in Getdents to
get rid of the path name.
Also, add an emulated version of getdents for MacOS. This allows
to drop the !HaveGetdents special cases from fusefrontend.
Modify the getdents test to test both native getdents and the emulated
version.
In PlaintextNames mode the "gocryptfs.longname." prefix does not have any
special meaning. We should not attempt to delete any .name files.
Partially fixes https://github.com/rfjakob/gocryptfs/issues/174
This is already done in regular mode, but was missing when PlaintextNames mode
is enabled. As a result, symlinks created by non-root users were still owned
by root afterwards.
Fixes https://github.com/rfjakob/gocryptfs/issues/176
In PlaintextNames mode the "gocryptfs.longname." prefix does not have any
special meaning. We should not attempt to read the directory IV or to
create special .name files.
Partially fixes https://github.com/rfjakob/gocryptfs/issues/174
If the user manages to replace the directory with
a symlink at just the right time, we could be tricked
into chown'ing the wrong file.
This change fixes the race by using fchownat, which
unfortunately is not available on darwin, hence a compat
wrapper is added.
Scenario, as described by @slackner at
https://github.com/rfjakob/gocryptfs/issues/177 :
1. Create a forward mount point with `plaintextnames` enabled
2. Mount as root user with `allow_other`
3. For testing purposes create a file `/tmp/file_owned_by_root`
which is owned by the root user
4. As a regular user run inside of the GoCryptFS mount:
```
mkdir tempdir
mknod tempdir/file_owned_by_root p &
mv tempdir tempdir2
ln -s /tmp tempdir
```
When the steps are done fast enough and in the right order
(run in a loop!), the device file will be created in
`tempdir`, but the `lchown` will be executed by following
the symlink. As a result, the ownership of the file located
at `/tmp/file_owned_by_root` will be changed.
If the symlink target gets too long due to base64 encoding, we should
return ENAMETOOLONG instead of having the kernel reject the data and
returning an I/O error to the user.
Fixes https://github.com/rfjakob/gocryptfs/issues/167
Fixes https://github.com/rfjakob/gocryptfs/issues/168
Steps to reproduce the problem:
* Create a regular reverse mount point
* Create files with the same very long name in multiple directories - so far
everything works as expected, and it will appear with a different name each
time, for example, gocryptfs.longname.A in directory A and
gocryptfs.longname.B in directory B
* Try to access a path with A/gocryptfs.longname.B or B/gocryptfs.longname.A -
this should fail, but it actually works.
The problem is that the longname cache only uses the path as key and not the
dir or divIV. Assume an attacker can directly interact with a reverse mount and
knows the relation longname path -> unencoded path in one directory, it allows
to test if the same unencoded filename appears in any other directory.
Fixes https://github.com/rfjakob/gocryptfs/issues/171
Steps to reproduce:
* Create a regular forward mount point
* Create a new directory in the mount point
* Manually delete the gocryptfs.diriv file from the corresponding ciphertext
directory
* Attempt to delete the directory with 'rmdir <dirname>'
Although the code explicitly checks for empty directories, it will still attempt
to move the non-existent gocryptfs.diriv file and fails with:
rmdir: failed to remove '<dirname>': No such file or directory
Fixes https://github.com/rfjakob/gocryptfs/issues/170
Steps to reproduce the problem:
* Create a regular forward mount point
* Create a file with a shortname and one with a long filename
* Try to run 'mv <shortname> <longname>'
This should actually work and replace the existing file, but instead it
fails with:
mv: cannot move '<shortname>' to '<longname>': File exists
The problem is the creation of the .name file. If the target already exists
we can safely ignore the EEXIST error and just keep the existing .name file.
Allows to use /dev/random for generating the master key instead of the
default Go implementation. When the kernel random generator has been
properly initialized both are considered equally secure, however:
* Versions of Go prior to 1.9 just fall back to /dev/urandom if the
getrandom() syscall would be blocking (Go Bug #19274)
* Kernel versions prior to 3.17 do not support getrandom(), and there
is no check if the random generator has been properly initialized
before reading from /dev/urandom
This is especially useful for embedded hardware with low-entroy. Please
note that generation of the master key might block indefinitely if the
kernel cannot harvest enough entropy.
...to account for unaligned reads.
I have not seen this happen in the wild because the kernel
always seems to issue 4k-aligned requests. But the cost
of the additional block in the pool is low and prevents
a buffer overrun panic when an unaligned read does happen.
Our byte cache pools are sized acc. to MAX_KERNEL_WRITE, but the
running kernel may have a higher limit set. Clamp to what we can
handle.
Fixes a panic on a Synology NAS reported at
https://github.com/rfjakob/gocryptfs/issues/145
A file with a name of exactly 176 bytes length caused this error:
ls: cannot access ./tmp/dsg/sXSGJLTuZuW1FarwIkJs0w/b6mGjdxIRpaeanTo0rbh0A/QjMRrQZC_4WLhmHI1UOBcA/gocryptfs.longname.QV-UipdDXeUVdl05WruoEzBNPrQCfpu6OzJL0_QnDKY: No such file or directory
ls: cannot access ./tmp/dsg/sXSGJLTuZuW1FarwIkJs0w/b6mGjdxIRpaeanTo0rbh0A/QjMRrQZC_4WLhmHI1UOBcA/gocryptfs.longname.QV-UipdDXeUVdl05WruoEzBNPrQCfpu6OzJL0_QnDKY.name: No such file or directory
-????????? ? ? ? ? ? gocryptfs.longname.QV-UipdDXeUVdl05WruoEzBNPrQCfpu6OzJL0_QnDKY
-????????? ? ? ? ? ? gocryptfs.longname.QV-UipdDXeUVdl05WruoEzBNPrQCfpu6OzJL0_QnDKY.name
Root cause was a wrong shortNameMax constant that failed to
account for the obligatory padding byte.
Fix the constant and also expand the TestLongnameStat test case
to test ALL file name lengths from 1-255 bytes.
Fixes https://github.com/rfjakob/gocryptfs/issues/143 .
The encrypt and decrypt path both had a copy that were equivalent
but ordered differently, which was confusing.
Consolidate it in a new dedicated function.
MacOS creates lots of these files, and if the directory is otherwise
empty, we would throw an IO error to the unsuspecting user.
With this patch, we log a warning, but otherwise pretend we did not
see it.
Mitigates https://github.com/rfjakob/gocryptfs/issues/140
...and if Getdents is not available at all.
Due to this warning I now know that SSHFS always returns DT_UNKNOWN:
gocryptfs[8129]: Getdents: convertDType: received DT_UNKNOWN, falling back to Lstat
This behavoir is confirmed at http://ahefner.livejournal.com/16875.html:
"With sshfs, I finally found that obscure case. The dtype is always set to DT_UNKNOWN [...]"
The benchmark that supported the decision for 512-byte
prefetching previously lived outside the repo.
Let's add it where it belongs so it cannot get lost.
The Readdir function provided by os is inherently slow because
it calls Lstat on all files.
Getdents gives us all the information we need, but does not have
a proper wrapper in the stdlib.
Implement the "Getdents()" wrapper function that calls
syscall.Getdents() and parses the returned byte blob to a
fuse.DirEntry slice.
Remove the "Masterkey" field from fusefrontend.Args because it
should not be stored longer than neccessary. Instead pass the
masterkey as a separate argument to the filesystem initializers.
Then overwrite it with zeros immediately so we don't have
to wait for garbage collection.
Note that the crypto implementation still stores at least a
masterkey-derived value, so this change makes it harder, but not
impossible, to extract the encryption keys from memory.
Suggested at https://github.com/rfjakob/gocryptfs/issues/137
* extend the diriv cache to 100 entries
* add special handling for the immutable root diriv
The better cache allows to shed some complexity from the path
encryption logic (parent-of-parent check).
Mitigates https://github.com/rfjakob/gocryptfs/issues/127
Dir is like filepath.Dir but returns "" instead of ".".
This was already implemented in fusefrontend_reverse as saneDir().
We will need it in nametransform for the improved diriv caching.
A directory with a long name has two associated virtual files:
the .name file and the .diriv files.
These used to get the same inode number:
$ ls -di1 * */*
33313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw
1000000000033313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw/gocryptfs.diriv
1000000000033313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw.name
With this change we use another prefix (2 instead of 1) for .name files.
$ ls -di1 * */*
33313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw
1000000000033313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw/gocryptfs.diriv
2000000000033313535 gocryptfs.longname.2togDFouca9mrTwtfF1RNW5DZRAQY8alaR7wO_Xd5Zw.name
On MacOS, building and testing without openssl is much easier.
The tests should skip tests that fail because of missing openssl
instead of aborting.
Fixes https://github.com/rfjakob/gocryptfs/issues/123
Fixed by including the correct header. Should work on older openssl
versions as well.
Error was:
locking.go:21: undefined reference to `CRYPTO_set_locking_callback'
Due to RMW, we always need read permissions on the backing file. This is a
problem if the file permissions do not allow reading (i.e. 0200 permissions).
This patch works around that problem by chmod'ing the file, obtaining a fd,
and chmod'ing it back.
Test included.
Issue reported at: https://github.com/rfjakob/gocryptfs/issues/125
Previously we ran through the decryption steps even for an empty
ciphertext slice. The functions handle it correctly, but returning
early skips all the extra calls.
Speeds up the tar extract benchmark by about 4%.
We use two levels of buffers:
1) 4kiB+overhead for each ciphertext block
2) 128kiB+overhead for each FUSE write (32 ciphertext blocks)
This commit adds a sync.Pool for both levels.
The memory-efficiency for small writes could be improved,
as we now always use a 128kiB buffer.
128kiB = 32 x 4kiB pages is the maximum we get from the kernel. Splitting
up smaller writes is probably not worth it.
Parallelism is limited to two for now.
Spawn a worker goroutine that reads the next 512-byte block
while the current one is being drained.
This should help reduce waiting times when /dev/urandom is very
slow (like on Linux 3.16 kernels).
On my machine, reading 512-byte blocks from /dev/urandom
(same via getentropy syscall) is a lot faster in terms of
throughput:
Blocksize Throughput
16 28.18 MB/s
512 83.75 MB/s
For a single-threaded streaming write, this drops the CPU usage of
nonceGenerator.Get to almost 1/3:
flat flat% sum% cum cum%
Before 0 0% 95.08% 0.35s 2.92% github.com/rfjakob/gocryptfs/internal/cryptocore.(*nonceGenerator).Get
After 0.01s 0.092% 92.34% 0.13s 1.20% github.com/rfjakob/gocryptfs/internal/cryptocore.(*nonceGenerator).Get
This change makes the nonce reading single-threaded, which may
hurt massively-parallel writes.
This check would need locking to be multithreading-safe.
But as it is in the fastpath, just remove it.
rand.Read() already guarantees that the value is random.
