The existing format was unreadable; putting the pid and tid first helps
somewhat. Also remove the unused qemu tracing which wasn't called anywhere.
Change-Id: I37ef3c556fe17b237ba1d8ca3216e2155ce5d0de
Currently run_transport_disconnects() are called twice. One is in
handle_offline(), another is before destroying transport.
The users of disconnect callback are listener, adb_auth_client, and
remote_sockets. All of them need only to be called once. And after
handle_offline, no new listeners, adb_auth_client, or remote_sockets
can be connected to the offlined transport. So I think we can remove
the second call to run_transport_disconnects().
Change-Id: I1ef8b6b7b5ab7ae1bad109be107c85973d65a2e3
adb push was not returning a bad exit code when write_data_file() or
write_data_link() failed. I encountered this when running the unittest
on Windows which can get into situations where stat() succeeds, but
open() fails due to pre-existing exclusive file access (which typically
doesn't exist on unix).
The same code is used by adb install, so this also fixes its error
handling.
Fixed some fd leaks and propagation of errors when reading a file.
Fixed a unittest to close temp files before reading them.
Change-Id: Ieba0026fa4c79eb0484676e4f2faaac9603ef584
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
The Python 2 subprocess class doesn't use Unicode, so as a work-around
write the command line to a UTF-8 batch file and run that.
I modified the test to use u'blah' without .encode('utf-8') because the
Python docs recommend dealing with string variables like that. When
formatting a string with a unicode parameter, use u'foo' on the constant
string to make it unicode.
I also tested this on Linux and it seems to work fine (I did ls in the
middle of the test to make sure the filenames came out right, etc.).
I had to close the temporary files before adb tries to read/write them
because filesystem semantics are different on Windows (technically I
might be able to modify adb to try to open files with more permissive
share flags, but then I'm not sure if Python uses the right share flags.
Basically, I'd be opening another can of worms.).
Fixed the test to delete a temp file on the device once it is done.
Change-Id: Id0c34e26d7697fbbb47a44ae45298bed5e8c59d6
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
The function of remote_socket_disconnect() is to make sure
the local_sockets and remote_sockets are closed when the binded
transport is disconnected. However, as we call close_all_sockets()
in handle_offline(), we don't need remote_socket_disconnect() any more.
Change-Id: I575f632d9f8703149f34e0210eb698a56e2516a9
Transport atransport objects are semi-reference counted: the input and
output threads each hold a reference. The adb disconnect command was
calling transport_unref to release a reference that it never had in the
first place. This meant that the refcount dropped to zero and the object
was deleted before either the input or output thread released its
reference. When that last thread released its reference, it wrote to
freed memory and also sometimes crashed.
This fix is to not release any unheld reference, instead it just kicks
the transport to break remote_read in output_thread. So all transport
close flow goes the following way:
output_thread (exit) -> main thread (offline the transport) ->
input thread (exit) -> main thread (destroy the transport)
Change-Id: Iad1fe718acc8716f3a79c8c22b426a1b2450452c
When launching the adb server (typically from adb start-server),
redirect stdout/stderr to anonymous pipes which are read by threads in
the parent process, to make error diagnosis easier.
If there is an error during adb start-server, the output looks like:
> adb start-server
* daemon not running. starting it now on port 5037 *
error: could not blah # from server process
could not read ok from ADB Server # from launch_server
* failed to start daemon * # from adb_connect
error: cannot connect to daemon # from adb_commandline
Fix handle-leaks in launch_server by using new unique_handle class
that is based on std::unique_ptr.
In the server, close stdin and redirect to adb.log *before* sending the
ACK, so that any errors are reported early instead of after the ACK.
Change-Id: I943881210a0ea9458fc36851339f916c3d6a0830
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
It is reported by tsan as a double checked locking. But I think
it is not a real data race. Because I think the old code is able
to make sure t->kick() is only called once, and the caller of
kick_transport is not relying on the side-effect of calling
t->kick().
But as it is not perf critical, I don't mind breaking the double
checked locking pattern.
Bug: 23385662
Change-Id: Ie3597dd56bb514117c3865d2afcfd7c115731a78
If s->peer->enqueue() failed, s may be freed. So we should use
saved_xxx instead of s->xxx before verifying the return value.
Change-Id: I6c072406dceb98e2d02798d0dcdc428fa99e66fb
As far as I can see, all asockets operations happen in fdevent_loop()
in the main thread, excepting close_all_sockets(). Instead of adding
lock and ref_count for each asocket, a simpler way would be moving
close_all_sockets() from input_thread to the main thread.
In input_thread(), there are two path to break the loop and call
close_all_sockets(). One path is when receiving offline A_SYNC, which
is sent by the main thread. The other path is when read_packet
fails, which I believe is almost not possible and doesn't matter
(Because t->fd is closed just before t is freed.). So I move
close_all_sockets() to handle_offline() in the main thread.
the socket_list_lock in sockets.cpp could be removed. But I prefer
to leave it for the following changes.
Bug: 6558362
Change-Id: I5da23f60a67a331262c62693b9b127fe2689c799
The error was this:
system/core/adb/usb_osx.cpp:203:74: error: values of type 'UInt32' should not
be used as format arguments; add an explicit cast to 'unsigned int' instead
[-Werror,-Wformat]
snprintf(devpathBuf, sizeof(devpathBuf), "usb:%" PRIu32 "X", locationId);
~~~ ^~~~~~~~~~
(unsigned int)
Which seems to be because on LP64 UInt32 is "unsigned int" but on LP32 it was
"unsigned long". We don't have to care about LP32, so -- if we can -- we're
probably better off just using uint32_t instead of UInt32.
Change-Id: I576f76cf2016ee59caccbc317ef74b6e8d71d722
In an earlier code review it was pointed out that there was something
very weird about fail_errno. It didn't seem to make sense that we'd
often try to continue after reporting failure. This patch cleans up
all that and assumes that if we've reported failure to the client,
we should stop what we're doing.
Bug: http://b/23437039
Change-Id: I39c38650ed9f9d5e30adbf68a7545c9e4a6ab812
We can double the speed of "adb sync" (on N9) if we increase SYNC_DATA_MAX
from 64KiB to 256KiB. This change doesn't do that, because I still haven't
managed to plumb through the information about whether we're a new adb/adbd
to file_sync_client.cpp and file_sync_service.cpp. But this is already a big
change with a lot of cleanup, so let's do the cleanup and worry about the
intended change another day...
This change does improve performance somewhat by halving the number of
lstat(2) calls made on the client side, and ensuring that most packets are
sent with a single write. This has the pleasing result of making the null
sync on an AOSP N9 go from just over 300ms to around 100ms, which means it
now seems instantaneous (https://en.wikipedia.org/wiki/Mental_chronometry).
Change-Id: If9f6d4c1f93ec752b95f71211bbbb1c513045166
adb on Windows uses \r\n line-endings, so take that into account when
parsing output for the exit code.
Change-Id: I6a3d7c5ca455b0f0f7dae174866857e0aeee9926
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
This allows us to test for features explicitly rather than relying on
the protocol version number, allowing us to fall back gracefully if a
feature is not supported.
This will be needed for the upcoming shell upgrades for stdout/stderr
separation and exit code reporting.
Change-Id: Ibb1d8ad2611f7209901ee76d51346b453e9c5873
This CL clears both the host and device endpoints right at the
beginning when the bulk endpoints are identified. This is in general
a "good idea", but more specifically for us, it fixes the issue
that sometimes when adb quits, it clears the endpoint on the host,
but not on the device which resulted in a subsequent invocation of
adb was seeing a stall.
Bug: https://code.google.com/p/android/issues/detail?id=182151
Change-Id: I331fa6805c40d1f50c153c010ceecd2f6a4045eb
- handle_host_request
- When the host:kill command comes in, shutdown the socket before
calling exit(). If we don't do this, the client will output error info
even though everything is working ok.
- adb_connect()
- If we can't parse the version string, explain this in error output
and don't goto error which would try to close an fd we already closed.
- If host:kill doesn't work, output error info. Don't try to close
already closed fd.
- adb_main()
- If writing the ACK somehow has an error, output error info (I doubt
this will ever get hit).
- adb_commandline()
- Fix typo about max port number.
- Make 'adb kill-server' and 'adb start-server' output any detailed
error info.
Change-Id: Id1a309cc1bf516f7f49bd332b34d30f148b406da
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
adb can hang at shutdown due to a deadlock relating to WSACleanup().
This works around the issue by not calling WSACleanup() which shouldn't
be done anyway since threads aren't done using Winsock at shutdown.
A quick way to reproduce the original problem is to run many instances
of adb, many of which will call exit() soon:
for /l %i in (1,1,20) do @start adb nodaemon server
You may have to boost the 20 to 200, or set ADB_TRACE=1 or use Windows
10 instead of Windows 7, to affect the timing, but eventually there
should be hung adb processes with that repro.
A more complete fix to prevent problems like this from occuring in the
future, would be to additionally do the following:
- Investigate all static destructors that are called when exit() is
called.
- If they don't do anything important, switch all calls to exit() to
instead call _exit() and then ban exit() from being called.
Change-Id: Id1be3bf0053809a45f2eca4461e4c35b5ef9388d
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
Change `adb shell` so that interactive sessions use a PTY but
non-interactive do not. This matches `ssh` functionality better
and also enables future work to split stdout/stderr for
non-interactive sessions.
A test to verify this behavior is added to test_device.py with
supporting modifications in device.py.
Bug: http://b/21215503
Change-Id: Ib4ba40df85f82ddef4e0dd557952271c859d1c7b
Make these fatal errors:
- Win32 GetTempPathW() failures.
- Errors opening /dev/null (and don't use LOG(FATAL) for this error
since that will do a crash-dump on Windows which isn't appropriate for a
transient runtime error).
- Errors with dup2.
- Errors opening adb.log.
Change-Id: Ided76a5436d8c6f059d8f6799c49ba04c87181ae
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
- handle_forward_request
- Because we have detailed info about which syscall failed (at least
on Win32), use a more generic prefix of "cannot bind listener" followed
by the detailed info.
- install_listener
- Return string errors for a few errors even though I don't think any
callers actually output the string for those errors.
- Remove the printf since the callers print the message themselves.
- adb_main
- LOG(FATAL) calls abort() which on Windows calls the Windows Error
Reporting service which pops up a dialog asking if you want a
crashdump to be uploaded to Microsoft. So really, abort() is
designed for app bugs. Windows isn't the only one doing this, Chromium
also makes LOG(FATAL) crashdump-ready. Since an error here is not
necessarily an app-bug, use a 'normal' error output API like fatal()
which prints an error and just uses exit().
- sysdeps_win32.cpp
- When Winsock APIs fail, make the string clarify which API failed.
Use terse unix-style descriptions (like what you'd get from
cp/mv/dd/etc.).
- Don't trace WSAEWOULDBLOCK from recv() which is a normal occurrence.
- Add a comment about WSAEWOULDBLOCK => EAGAIN.
Change-Id: I58e47f49fa2f6c1b4b92a36d0c4bfe369b456f2a
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
This is a follow-up to https://android-review.googlesource.com/153623
which prevented android::base::InitLogging() from being called when
tracing was disabled.
It is ok to call InitLogging() on a device or host because calling it
does not imply that a logging file is used, which was the reason for
not calling it on a device.
So this change should preserve the device behavior of not using a
logging file when tracing is disabled, plus it will call InitLogging()
all the time in case logging APIs are called.
Change-Id: I3fd6ba2c567f67a2f111a85f174893fbf866ec57
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
First, HOST is always 0 in adbd, which matches ADB_HOST=0.
Second, HOST is always 1 when adb_main is called, which matches ADB_HOST=1.
For adb client that doesn't call adb_main, it never touches local_init(),
init_transport_registration() and fdevent_loop(). So the changes in adb.cpp,
services.cpp and transport_local.cpp do nothing with it.
As a conclusion, I think we can remove HOST and use ADB_HOST instead.
Change-Id: Ide0e0eca7468b6c3c130f6b50974406280678b2e
The win32 version of 9f2d1a9cfc. The big
technique is to fit a Win32 HANDLE value in an int because it only uses
32-bits. This allows most of the other adb code to stay the same.
Also, fix a regression in the 'adb server nodaemon' command that was
erroneously returning an error when --reply-fd was not used, which
should not be necessary for this particular command.
Change-Id: I37e9c609014b813af93bf0d6c12f665b59c93c41
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
When "adb start-server" is issued, and a server needs to be launched,
adb client forks itself and the child process runs the server routine.
Once the server initializes its various components, it sends an "OK\n"
back to the client via its stderror (or stdout on Windows).
This sequence breaks down if before sending the "OK\n", the server
happens to log something on its stderr. In order to avoid this, the
client now expects the ack to come on a different fd rather than one
of the standard streams.
Bug: https://code.google.com/p/android/issues/detail?id=182150
Change-Id: I9d58a08068d71eb3b77e8a7377e934631c016466
- My recent change with -DUNICODE=1 required changing
GetProfilesDirectory() to GetProfilesDirectoryA() for the ANSI version
of the API.
- enh's edit to my previous change deleted a test that used
/proc/version, but I think another test was missed. Merge that test into
another.
Change-Id: Ic748549848e7be922bcbf218d5b0c3fca2a90704
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
The original code was:
if (strcmp(__adb_error, "unknown host service") != 0)
But that was changed by 078f0fcf4c to:
if (*error == "unknown host service") {
I think the comparison should be != so that "unknown host service"
falls-through and kills the server, and so if it is some other error,
that the other error is returned immediately.
Change-Id: Ia490a4a870d1d123a3c5ab258dd5fa0930e8032d
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
This CL adds --apk-dir option, which specifies the directory
that contains APK files to be installed before measuring
boot time.
BUG: 22207911
Change-Id: Ifeacf34c779248686443a9ef02485272c140a456
Not seeing the full output from the failed adb command is probably
the biggest issue when debugging a test failure, but this doesn't
help either.
Change-Id: Ic42cbced8be252185a799b27c210a744188a4201
Relative paths were being prefixed with OS_PATH_SEPARATOR on unix and
win32 causing adb to incorrectly try to make directories at the root.
Plus, absolute paths didn't work on win32 (C: got prefixed into \C:).
This fix is to use dirname (available on win32 via mingw's crt) to do
the messy parsing.
I added a test for the relative path case.
Change-Id: Ibb0a4a8ec7756351d252a4d267122ab18e182858
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
adb_test:
* Fix adb_utils directory_exists test for Windows. The test actually
fails because directory_exists() is not aware of junctions or symlinks,
but I'm not really sure if that is a bad thing (since these are rare on
Windows to begin with).
* Fix crash during transport tests due to mutex not being initialized.
* io tests fail for various reasons (see adb_io_test.cpp for more info).
libbase_test:
* Get it building on Win32 by implementing mkstemp() and mkdtemp().
* Run StringPrintf %z test on Windows because it passes because we build
with __USE_MINGW_ANSI_STDIO which implements %z.
* I didn't fixup the logging tests: some logging tests fail because when
abort() is called on Windows, by default it pops up UI asking whether a
crash dump should be sent to Microsoft. To some degree this makes sense,
as I think LOG(FATAL) does crash dumping in Chromium. This should be
revisited in the future.
Change-Id: Iaa2433e5294ff162e0b2aa9fe6e4ec09a6893f7a
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
Define the UNICODE and _UNICODE preprocessor symbols to make passing
char* to Ansi/Unicode-agnostic Windows and C Runtime APIs break the
build. The solution is to call wide Windows and C Runtime APIs and use
widen(utf8).c_str(). Most code was already calling wide APIs. Defining
these symbols makes a call to CreateEvent() (which previously mapped to
CreateEventA()) turn into a call to CreateEventW().
Make SystemErrorCodeToString() use Unicode.
Add various comments.
Change-Id: I9b212412348a29826718e897a486489e1f142d16
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
After resuming Windows from sleep or hibernation, USB connections are not
automatically disconnected. Writing to the USB connections does not return any
errors, but read never returns. My theory is that the device saw the host
sleep/hibernation as a disconnect, so the device is waiting for re-auth from the
host as if the host was just connected.
To solve this, detect resume from sleep/hibernation, disconnect all USB
connections and let the device poll thread re-detect the USB devices in 1 sec.
This is done by using a hidden window that receives power notifications. The
hidden window code is based on Chromium's similar code (platform-tools already
includes the Chromium Authors license).
This depends on a change to AdbWinUsbApi.dll that makes AdbCloseHandle(endpoint)
abort any pending IOs and wait for those IOs to be aborted.
I expect that this should solve many adb and Android Studio related bugs
regarding hangs or errors.
Also in this change:
- Add D() logging for any errors from AdbWinApi.dll API calls.
- Stop setting errno to Win32 error values which the caller can't really do
anything with. Stop calling SetLastError() because the callers don't check
GetLastError() anyway.
- Check the return value from writing zero length packets.
- If the full amount of data isn't written, return an error.
- Upon any usb_read/usb_write error, kick the connection instead of only
kicking when ERROR_INVALID_HANDLE is encountered.
- Restructure some code from nested-if-trees to goto-fail to make it easier
to follow.
- Delete usb_name() since it isn't thread-safe and it isn't used.
Change-Id: Iffcf5315ad8593d0c7e93012afaabe6fae354ac1
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
When repeatedly opening and closing a file descriptor, the sequence of
fds returned was: 100...227,100,100,100,100,100... Basically, the first
wave was constantly increasing fds, but after the entire fd table was
traversed once, the alloc algorithm would switch to returning the first
free fd. This is sub-optimal for reliability because use-after-free bugs
would be more likely to be hit because right after a close, the same fd
would be given out next.
This change makes the alloc algorithm use a persistent clock hand that
walks forward through the fd table (wrapping around if necessary),
searching for a free fd.
This change adds locking for fd closing:
- This prevents multiple concurrent closes of the same fd.
- There was a race between alloc and close that wasn't guaranteed to be
correct: close would set f->clazz to NULL last, but without any
preceding memory barrier/fence, then the alloc thread would check for
NULL. It probably worked out ok in practice, but it is probably best
to fix this up with a lock (as in this change) or a memory barrier/fence
(but this code isn't about performance, so why go with a complicated
barrier/fence?)
Also in this change:
- Use errno = EMFILE for the out of fds case.
- Clear FH->name
Change-Id: Ic11d2a1a9d53996edfc1ca13566a2f46de4a4316
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
Initial support for Unicode file/dir names. Unicode paths can be passed
on the command line, directory enumeration can enumerate Unicode paths,
Unicode paths are used for file access, and Unicode paths can be output
on the console correctly.
Also Unicode environment variable access.
Initial support for Unicode output from adb shell (which uses
adb_fwrite()). This is partial because the corner case of an
adb_fwrite() call with an incomplete UTF-8 multi-byte sequence does not
output correctly, but this should be uncommon, is better than what we
had before (*always* incorrect UTF-8 multi-byte sequences) and can be
fixed in the future.
Calls to Windows APIs with char strings were changed to pass wchar_t
strings to the FooW() variants.
For more details, see the giant comment in sysdeps_win32.cpp.
https://code.google.com/p/android/issues/detail?id=8185
Change-Id: I7ebf6713bb635638b986ccee97b354428837c9c5
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
Call getaddrinfo() for connecting to IPv6 destinations.
Winsock APIs do not set errno. WSAGetLastError() returns Winsock errors
that are more numerous than BSD sockets, so it really doesn't make sense
to map those to BSD socket errors. Plus, even if we did that, the
Windows C Runtime (that mingw binaries use) has a strerror() that does
not recognize BSD socket error codes.
The solution is to wrap the various libcutils socket_* APIs with
sysdeps.h network_* APIs. For POSIX, the network_* APIs just call
strerror(). For Windows, they call SystemErrorCodeToString() (adapted
from Chromium).
Also in this change:
- Various other code was modified to return errors in a std::string*
argument, to be able to surface the error string to the end-user.
- Improved error checking and use of D() to log Winsock errors for
improved debuggability.
- For sysdeps_win32.cpp, added unique_fh class that works like
std::unique_ptr, for calling _fh_close().
- Fix win32 adb_socketpair() setting of errno in error case.
- Improve _socket_set_errno() D() logging to reduce confusion. Map
a few extra error codes.
- Move adb_shutdown() lower in sysdeps_win32.cpp so it can call
_socket_set_errno().
- Move network_connect() from adb_utils.cpp to sysdeps.h.
- Merge socket_loopback_server() and socket_inaddr_any_server() into
_network_server() since most of the code was identical.
Change-Id: I945f36870f320578b3a11ba093852ba6f7b93400
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
This removes adb_dirstart and adb_dirstop. It also fixes a couple of memory
leaks by switching to std::string. This also fixes the bug in the previous
change --- mkdirs is given input like "/system/bin/sh" and only expected to
create "/system/bin". In a later change, we should remove mkdirs and only
expose the intended "unlink && mkdirs && create" functionality.
Change-Id: I30289dc1b3dff575cc1b158d993652178f587552
~ Rewrote mkdirs to be in C++ style.
~ Replaced adb_dir{start,stop} with std::string params and (r)find.
+ Added test for mkdirs.
Also make base/test_utils.h public and support temporary directories
as well as files.
Change-Id: I6fcbdc5e0099f3359d3aac6b00c436f250ca1329
Host and target are split here because the target can really only use
ubsan (most sanitizers don't support static executables).
Change-Id: I8a5a5adeeef5c27aaaa3d8145b1570760b764ce3
If ANDROID_SERIAL was not set, test_device.py was failing.
Use posixpath.join instead of os.path.join for make_random_device_files.
Change-Id: I24bfa43ba2a89a9a768f505fc0bba9d873082b2f
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
perfboot.py repeats the record of each event log during Android
boot specified times. By default, interval between measurements
is adjusted in such a way that CPUs are cooled down sufficiently
to avoid boot time slowdown caused by CPU thermal throttling.
This script also works around the issue of dropbox slowing down
boot time on userdebug build (http://b/20890386) by limiting
the number of files to be created by dropbox.
The result is output in a tab-separated value format.
BUG: 22207911
Change-Id: I0ddbac5d1c941efda87bc6db6388d8194d4bb3dd
adb forward --no-rebind (and the matching reverse command) seem to have
been broken by 424af02f36. I fixed-up the
string parsing and also fixed the error message not to include
strerror(errno) since that does not apply because it is an
application-level error that doesn't have anything to do with the
OS/libc.
Change-Id: Iba7286283bfcf1782076355edcd9c355c0a0edfd
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
Had messed this up while refactoring before I submitted and neglected
to add a test. Thanks to mazda@ for catching this.
Also disabled the root/unroot tests for user builds.
Change-Id: Icb819a820a2afa227d548d678ae471d5195f0b96
This causes build failures in google3 where they use GCC. glibc only
provides const-correct overloads for string functions for GCC >= 4.4,
but clang -- which is what we use -- pretends to be GCC 4.2.
Change-Id: I2a054823ea6201ebcea46d5e77b80a975eefc622
This is mostly just the AdbWrapper that we used in our tests, but I've
cleaned up the API to be a little more Pythonic (mostly in the sense
that commands are passed as lists rather than strings that are
shlex.split() by the shell command), and implemented the workaround
error checking for adb shell.
Move the tests up a directory. Having them buried a level down has
only been annoying.
There are now two files containing Python tests. test_device.py
contains tests specifically checking the AndroidDevice API, and
test_adb.py checks the ADB client program. To run both, use
python -m unittest discover [-v]
Change-Id: Ibd158c528d31126a5b048bd00bc93039dbc468bc
Elliott Hughes