GitLab

Change-Id: I23b1281a63031a7481ea7b33c9ddbdbe7d3d6174

Change-Id: I130b250a663cdfb379def24583523d0287ec31dd

This is the first in a series of changes that are designed to
introduce better support for dynamic block devices.

It starts by defining a new Volume object which represents a storage
endpoint that knows how to mount, unmount, and format itself.  This
could be a filesystem directly on a partition, or it could be an
emulated FUSE filesystem, an ASEC, or an OBB.

These new volumes can be "stacked" so that unmounting a volume will
also unmount any volumes stacked above it.  Volumes that provide
shared storage can also be asked to present themselves (through bind
mounts) into user-specific mount areas.

This change also adds a Disk class which is created based on block
kernel netlink events.  Instead of waiting for partition events from
the kernel, it uses gptfdisk to read partition details and creates
the relevant Volume objects.

Change-Id: I0e8bc1f8f9dcb24405f5e795c0658998e22ae2f7

Bug: 19624276
Change-Id: I84eeb029114d6ba2a26a53d4155d123d8acd7502

* commit '58819cde':
  fstrim.c: use open(O_DIRECTORY) instead of stat

* commit 'cdcffc0d':
  fstrim.c: use open(O_DIRECTORY) instead of stat

* commit '3393bb65':
  fstrim.c: use open(O_DIRECTORY) instead of stat

open(O_DIRECTORY) returns an error if the open attempt
is against anything other than a directory. This basically
duplicates the check that the stat() call was trying to do.
Eliminate the unnecessary stat() call and use O_DIRECTORY
instead.

Change-Id: I1821abbed325f29a7214fdc41ed27cd9e26817d0

* commit '42c79333':
  Rename keymaster_device_t to keymaster0_device_t.

* commit '47bc0ffa':
  Rename keymaster_device_t to keymaster0_device_t.

* commit '5054f7ee':
  Rename keymaster_device_t to keymaster0_device_t.

* commit 'af057b52':
  vold: acquire wakelock on fstrim thread

* commit 'aac44855':
  vold: acquire wakelock on fstrim thread

* commit 'c19c6738':
  vold: acquire wakelock on fstrim thread

* commit 'b87264fb':
  Rename keymaster_device_t to keymaster0_device_t.

If fstrim thread creation or detachment is failed, wakelock can be leaked.
So move wakelock acquire to do_fstrim_filesystems function

Change-Id: I4da3164343af83fae3e5b01700f43d1752661276
Signed-off-by: Young-ho Cha <ganadist@gmail.com>

This is to accomodate the new keymaster1_device_t, which has an entirely
different interface.

Soon I'll provide a libkeymaster which provides a unified (and nicer)
interface for dealing with both v0 and v1 keymaster implementations
using a v1 keymaster API.  For now this change is just so that vold will
build and run.

Change-Id: I5c54282c12d1c4b8b22ed4929b6e6c724a94ede4

This is to accomodate the new keymaster1_device_t, which has an entirely
different interface.

Soon I'll provide a libkeymaster which provides a unified (and nicer)
interface for dealing with both v0 and v1 keymaster implementations
using a v1 keymaster API.  For now this change is just so that vold will
build and run.

Change-Id: I5c54282c12d1c4b8b22ed4929b6e6c724a94ede4

This is to accomodate the new keymaster1_device_t, which has an entirely
different interface.

Soon I'll provide a libkeymaster which provides a unified (and nicer)
interface for dealing with both v0 and v1 keymaster implementations
using a v1 keymaster API.  For now this change is just so that vold will
build and run.

Change-Id: I5c54282c12d1c4b8b22ed4929b6e6c724a94ede4

Changing the device lock (even from swipe to none) will cause the
master key to be re-encrypted.
If at that point keymaster fails (e.g. due to an incompatible keymaster update)
cryptfs will write back the now-incomplete crypto metadata.
Upon next reboot, userdata can't be decrypted.

Now we don't bother writing on keymaster failure.

Bug: 19301883
Change-Id: I2b9a1278f8b4d333ac8d567e17e2263005e99409

* commit '33b6de4b':
  prevent ioctl_init() to write outside buffer

* commit 'd773fc11':
  Change strncpy to strlcpy

* commit '38394c7d':
  Revert "Make encryption configurable"

* commit 'd55d8dac':
  prevent ioctl_init() to write outside buffer

* commit '8c0a0db1':
  Change strncpy to strlcpy

* commit '223fd1ca':
  Revert "Make encryption configurable"

This reverts commit 6a69cfc4.

The original fix seems to have led to boot failures in QA. Rather than
risk shipping, revert the change. Bug 18764230 reopened.

Requires change
    https://googleplex-android-review.git.corp.google.com/#/c/629950/

Bug: 19278390
Bug: 19199624

Change-Id: Ia858c4db0abb917f9364ec8048f59ca4fb48e233

* commit 'cf694550':
  VolumeManager: don't use faccessat(AT_SYMLINK_NOFOLLOW)

* commit 'd9f2082f':
  VolumeManager: don't use faccessat(AT_SYMLINK_NOFOLLOW)

Don't use faccessat(AT_SYMLINK_NOFOLLOW). In Android, AT_SYMLINK_NOFOLLOW
is ignored. In glibc, it returns counter intuitive results when a
symbolic link is encountered, returning true all the time even though
an open(O_NOFOLLOW) will eventually fail.

Instead, stat the file and check to see if it's a regular file,
not a directory or symlink or some other weirdness.

In addition, fix a bug where isAsecInDirectory would return
true ("-1") if the asec directory didn't exist. It should return
false.

Bug: 18867827
Change-Id: I33d90e9095fad36ce0f83fde105b70f72e4eaef4

Change all function calls to strncpy to strlcpy.

Change-Id: I2bbefb7829d715847c5b26f4b9f0faddbd4c89d0

The strncpy operation does not write a 0 termination
if the name is larger than the target buffer.

Ensure that zero termination is always written using
safe strlcpy function.

Change-Id: Idb68cdff7cd1a860c1dfac7494fa99f3d382cb91