The reason for the crash is that ADB on Windows uses WaitForMultipleObjects to
wait on connection events. When number of connections exceeds 64, ADB crashes,
because WaitForMultipleObjects API doesn't support more than 64 handles. This
CL contains a fixer routine that allows waiting on an arbitrary number of
handles.
(cherry picked from commit ac52833e48)
Change-Id: I8ad264765e5b38d01a31e42b445f97ea86e49948
The reason for the crash is that ADB on Windows uses WaitForMultipleObjects to
wait on connection events. When number of connections exceeds 64, ADB crashes,
because WaitForMultipleObjects API doesn't support more than 64 handles. This
CL contains a fixer routine that allows waiting on an arbitrary number of
handles.
Change-Id: I83f712e552018df308318154c27df184015a16ee
Normally only bandwidth stats for the uid reading .../stats is returned.
With the extra group, adb shell will be able to read stats from
all UIDs.
This is to be used to offer data usage profiling to developers.
Change-Id: If3d2941ce5aa4dbb1a23947b97b893149ba224f7
Documentation-only change that supports the new framework-side
feature to omit system packages when you're using adb backup -all.
Bug 5361503.
Change-Id: I86bca8883a7fb8c713ca352ad5980e92fd640d18
A command line flag with an argument was checked in the Pm.java code,
but it wasn't being checked by "adb install" so attempts to use it
failed.
Change-Id: I0b84a4203a416f7323fa823c0f1f1750670d0c76
The host side wasn't properly checking for argument-list sufficiency
*after* removing any [-f filename] sequence.
Fixes bug 5164135
Change-Id: I7bc49e37ef168182088e0e664b6897dd2a088ebf
We now use "backup.ab" as the default backup archive filename, and no longer
refer to "tar" or "tarfiles" in the help text. The underlying format may
be tar, but we're certainly not interoperable even with ustar/pax thanks
to our compression & encryption layers and our custom header.
Change-Id: I3e74af96cfc102e94848c969eb36af54304bfd9b
Allow "adb install" to transfer the file for the verification argument
to the package manager "pm install" command.
Change-Id: I4834f45019eb1387a5d2b205b53a67e91d5fa67e
Fixed memory leak when using transport:serial. Code was previously
using strdup() to make a copy of the serial number which was not
necessary. Instead, just set serial=service.
Change-Id: Ifa8fe7999c6fb88955d2c6bdfcd3cd724ec1f020
Signed-off-by: Tom Marlin <tomm@bsquare.com>
* commit 'f75a2ed2138cd8b74597bd57e13d12e1f0b815c2':
adb: Add vendor IDs for iRiver, Compal and T & A Mobile Phones
Add Gigabyte's USB vendor ID to adb.
Add Funai's USB vendor ID to adb.
The adb gadget driver used to reset the USB bus when the adbd daemon exited,
and the host side adb relied on this behavior to force it to reconnect
with the new adbd instance after init relaunches it. The new gadget
drivers no longer automatically reset the USB bus when adbd is restarted
which caused adb to hang since it was no longer forced to reconnect with the
device. We attempted to work around this on the host side adb, but that
work around has not been reliable.
This change adds a property trigger on the service.adb.root system property
which will reset the USB bus and restart the adbd daemon when adbd sets
the property to 1. This should be much closer to the previous behavior
and will hopefully fix some problems with automated testing.
Change-Id: I177c37400009a3d83f21a5f9431f94fd1cc19b9b
Signed-off-by: Mike Lockwood <lockwood@android.com>
This reverts commit d15e6ac95d.
Also increment adb version to 1.0.29
Change-Id: I890643f0c18a2fe90f170134df039c54116cecda
Signed-off-by: Mike Lockwood <lockwood@android.com>
The buggy behavior was that the 'adb backup ....' host command line
invocation would hang forever, even after the backup finished and
the forked device-side subprocess had been reaped. The reason for
this is that the device adbd end of the socketpair used to send
the data back from the forked subprocess was still stuck readable
even after the remote end of it had been closed.
With this patch, the thread whose job it is to waitpid() in order
to harvest the forked child process also closes the local (adbd)
end of the socketpair. This makes the fdevent logic notice that
the socket is dead, at which point it cleans up everything including
the communication back to the host.
Change-Id: I90e7e4e5db36c5a6f7363708b29a6d2c56d1250e
* Increase transfer buffer size to 32K
* Add logging about error conditions and fd teardown
* Pass the fd number as a command line option to the 'bu' subprocess
* Properly harvest the 'bu' subprocess after it's done
Change-Id: Id44dde25778ecf43c5604fd9d01d726ba58861e5
This is needed because the USB driver on the device no longer resets the bus
when exits and restarts.
Since the host side adb no longer detects that adbd has restarted we need
to reset the connection on the host side now.
Change-Id: I1a8eabd292e8b45c34fa07a8a8b8c609cc15648d
Signed-off-by: Mike Lockwood <lockwood@android.com>
Reset peers to NULL after closing them down. This prevents
other code from attempting to access that freed memory
(which prevents crashes). Previously, it left pointers to
freed memory and the "if (s->peer)" guards could not block
the attempt to access that memory later. Resolves many
crashes seen while taking repeated screenshots on WinXP.
Change-Id: I32553f4d19f6ddc9b05b6ab4dc1e9efe69e5be4f
* "adb backup" as the entire command line no longer crashes
* The "-f filename" option can now appear anywhere in the command line.
A trailing "-f" at EOL prompts an error message and usage summary.
Change-Id: I040ed73c2ca3687e265e35600eb3ab2b3c879695
It won't actually do anything until the 'bu' tool and framework are
updated to respond properly, but this is the adb side of the
necessary infrastructure: we copy the tarfile into the socket pointed
at the device, using the existing mechanisms.
Change-Id: Ic3b5779ade256bd1ad989a94b0685f7b1a7d59d2
Includes a version bump to make it possible to determine if a given
binary has this fix or not.
Bug: 4111855
Change-Id: I86d902a7fa0c4b906264141d2467b7bdddb61801
* Add support for correctly handling subprocess termination in shell service (b/3400254 b/3482112 b/2249397)
- have a waitpid() track the subprocess, then notify the fdevent via a socket
- force an eof on the pty master in fdevent's new subproc handler.
- modify fdevent to force-read the pty after an exit.
* Migrate the "shell:blabla" handling to "#if !ADB_HOST" sections, where it
belongs.
* Fix the race around OOM adjusting.
- Do it in the child before exec() instead of the in the parent as the
child could already have started or not (no /proc/pid/... yet).
* Allow for multi-threaded D() invocations to not clobber each other.
- Allow locks across object files.
- Add lock within D()
- Make sure sysdesp init (mutex init also) is called early.
* Add some missing close(fd) calls
- Match similar existing practices near dup2()
* Add extra D() invocations related to FD handling.
* Warn about using debugging as stderr/stdout is used for protocol.
* Fix some errno handling and make D() correctly handle it.
* Add new adb trace_mask: services.
* Make fdevent_loop's handle BADFDs more gracefully (could occur some subproc closed its pts explicitely).
* Remove obsolete commandline args reported in help. (b/3509092)
Change-Id: I928287fdf4f1a86777e22ce105f9581685f46e35
Ext4 filesystems like to be unmounted before rebooting. The Android system
doesn't have a traditional Linux init setup, and shutting down the system
was not much more than calling sync(2) and reboot(2). This adds a new
function to libcutils called android_reboot(). By default, it calls sync()
and then remounts all writable filesystems as read-only and marks them clean.
There is a flag parameter in which the caller can ask for sync() not to be
called, or to not remount the filesystems as read-only. Then it will call
reboot(2) as directed by the other parameters. This change also updates
adb, init and toolbox to call the new android_reboot() function.
Fixes bugs 3350709 and 3495575.
Change-Id: I16d71ffce3134310d7a260f61ec6f4dd204124a7
This is for http://b/3482112 "adb interactions with device causing test harness failures".
This reverts commit 69c5c4c45b.
Change-Id: I630bf2e04d2ecf0223bd2af4e87136754ff880d3
* Handling of the subprocess and its FD.
This fixes http://b/3400254 "Many bugreports getting hung at the end in monkey"
- Start up a service thread that waits on the subprocess to terminate,
then closes the FD associated with it.
- Have the event handler select() with a timeout so that it can
detect the closed FD. Select() with no timeout does not return when an FD is closed.
- Have the event handler force a read on the closed FD to trigger the close sequence.
- Migrate the "shell:blabla" handling to "#if !ADB_HOST" sections.
* Fix the race around OOM adjusting.
- Do it in the child before exec() instead of the in the parent as the
child could already have started or not (no /proc/pid/... yet).
* Allow for multi-threaded D() invocations to not clobber each other.
- Allow locks across object files.
- Add lock within D()
* Add some missing close(fd) calls
- Match similar existing practices near dup2()
* Add extra D() invocations related to FD handling.
* Warn about using debugging as stderr/stdout is used for protocol.
Change-Id: Ie5c4a5e6bfbe3f22201adf5f9a205d32e069bf9d
Signed-off-by: JP Abgrall <jpa@google.com>
* commit 'd4894f4be608c28ef0f39c7794ba865ac812dff9':
Fix adb leaking file descriptors to forked processes
adb: Fix command-line parser.
adb: Increase device descriptor buffer size in Linux host USB support
adb: improve debug traces readability.
adb: Don't report negative number of bytes after pushing file > 2 gigabytes
Adding Texas Instruments to the VID list.
Support an additional alias for 'adb shell.'
accept() creates a new file descriptor that should be closed on exec so
that forked processes don't keep a fd opened on the socket.
This also fixes b/3297070 where adb hangs after running adb on the
target.
Change-Id: I8df511289e5549ae49b4824c9dfb71a3bf85eae8
This fixes the command-line parser to accept both "daemon server"
and "server nodaemon". Before the patch, the second string would
ignore the "nodaemon" flag.
Fixes b/2191598
Change-Id: Ie922b3e7bf57a6e334fc448cec33fb340ca6abc4
256 bytes wasn't big enough for some complicated USB configurations
Change-Id: I31f7841953d73b06958f44040f58166b159faff4
Signed-off-by: Mike Lockwood <lockwood@android.com>
This patch makes the traces easier to read. For example transports are
displayed by name/serial instead of their hex address.
Change-Id: I7e8df44ddbec19754d63d989bd56485998b4627b
256 bytes wasn't big enough for some complicated USB configurations
Change-Id: I31f7841953d73b06958f44040f58166b159faff4
Signed-off-by: Mike Lockwood <lockwood@android.com>
Merge commit '6713a2e0c659be9d992e75ab420da5680f522658' into gingerbread-plus-aosp
* commit '6713a2e0c659be9d992e75ab420da5680f522658':
Fix USB endianness bugs on Linux.
Merge commit '81347d2764b32ec7224d4e2ab3348103816400b2' into gingerbread-plus-aosp
* commit '81347d2764b32ec7224d4e2ab3348103816400b2':
adb: Improved detection of big endian architecture
Merge commit 'a8b8c0e21ac4bdb7838c8c1b8e7d60f632eda79b'
* commit 'a8b8c0e21ac4bdb7838c8c1b8e7d60f632eda79b':
adb: Add support for Philips's USB vendor ID
The fields device->idVendor and device->idProduct should not be
converted from little endian to native byteorder, because the
kernel has in fact done so already!
On the other hand, the descriptors read using raw ioctl:s in
register_device() do need to be converted.
Change-Id: I5fe08b626b14ead56a592b68d026690e343c2656
Instead of testing for the (compiler specific) define
__ppc__ explicitly, use the define HAVE_BIG_ENDIAN provided by
AndroidConfig.h. That way, it should work on all big endian
systems.
Change-Id: Ic4d62afcefce4c8ad5716178ebfcb2b055ac73ce
Merge commit 'd02e521d79a49471b3f1dc3717e842948df64789' into gingerbread-plus-aosp
* commit 'd02e521d79a49471b3f1dc3717e842948df64789':
adb: Add support for Philips's USB vendor ID
Merge commit '89aff1d22662c64cc2de384e7e2c699e4d1d5d47' into gingerbread-plus-aosp
* commit '89aff1d22662c64cc2de384e7e2c699e4d1d5d47':
Add Asus USB vendor ID to adb.
Added a new target for adb, so that the adb
host executable is built for the target as
well. This allows the target to connect to
any Android devices which are attached to it.
Bug: 3022194
Change-Id: Ib01983e70b75cec40a9ee161da7f4cf1343eecf2
Merge commit 'f62b23f2c1540221621024c850fdf615dda0c903' into gingerbread-plus-aosp
* commit 'f62b23f2c1540221621024c850fdf615dda0c903':
Add new USB vendor IDs.
Merge commit '4309d602b562eb9f181b0314837e6096577c6c18' into gingerbread-plus-aosp
* commit '4309d602b562eb9f181b0314837e6096577c6c18':
Added Toshiba's USB VID (0x0930) upon their request.
to print msg.command as some odd decimal number is just not helping in debugging;
instead print msg.command as a string
Change-Id: Id1ec770c20c86b8bdae40a0262bf4b63c0213b0a
The ADB (Android Debug Bridge) daemon suffers from a file descriptor leak.
When transfering file descriptors, opened by the adbd daemon, to a JDWP
process for debugging through sendmsg, the file descriptor is duplicated
by the kernel and a copy is transfered to the JDWP process.
Adbd neglects to close it's copy of the descriptor after the sendmsg call
which causes a file descriptor leak in the process.
Eventually this causes the daemon to hit its maximum roof of opened file
descriptors thus making it impossible for the process to serve any new
commands from the adb server.
Jdwp_service.c has been modified to close the file descriptor after the
send message call has succeded.
Also, the socket used for transfering the file descriptor is set to blocking
mode before the sendmsg call since the adbd processes needs to know when the
duplication has been completed by the kernel.
The flags are reset to its original state after the sendmsg call.
Change-Id: I76b15460301ac01e432f5ce8678ff58c658c16f2
Signed-off-by: Christian Bejram <christian.bejram@stericsson.com>
... as memory was already obtained zero'ed by calloc().
Change-Id: Ic73bad09b54cb778fd40bdf86bb40888ea504c5f
Signed-off-by: André Goddard Rosa <andre.goddard@gmail.com>
Port number is now optional. Will use default port 5555 if not specified.
"adb disconnect" with no additional arguments will disconnect all TCP devices.
Change-Id: I7fc26528ed85e66a73b8f6254cea7bf83d98109f
Signed-off-by: Mike Lockwood <lockwood@android.com>
The linenoise library is from http://github.com/antirez/linenoise
This patch also disables command-line editing and history from adb. The
adb implementation was shadowing the Android shell's implementation.
The adb implementation was also shadowing the editing and history
implementation in alternative shells such as BusyBox's ash.
Change-Id: I7ebd4cb391d0ce966c0ce0e707d80ecd659f9079
- adb can now connect to an emulator configured with an arbitrary
pair of <console port, adb port>. These two ports do not have to be
adjacent.
This can be done from the commandline at any time using
adb connect emu:<console_port>,<adb_port>
- Emulators running on ports outside the normal range
(5554/5555-5584/5585) register themselves on startup if they follow
the convention "console port+1==abd port".
- Emulators outside the normal port range will not be auto-detected on
adb startup as these ports are not probed.
- The index into local_transports[] array in transport_local.c does no
longer indicate the port number of the local transport. Use the altered
atransport struct to get the port number.
- I have chosen not to document the adb connect emu:console_port,adb_port
syntax on adb's help screen as this might be confusing to most readers
and useful to very few.
- I don't expect this to introduce any (backwards) compatibility issues.
Change-Id: Iad3eccb2dcdde174b24ef0644d705ecfbff6e59d
Signed-off-by: Mike Lockwood <lockwood@android.com>
Joe Onorato