am: c6160d2a0e
* commit 'c6160d2a0ef648ccb3d217c589c60b5c00b80387':
Add public libs from an environment variable
Change-Id: I2be47c2e9fc02860e7ddd79b97f666c9ea6e881b
This is enabled only for builds with ro.debuggable=1
It is intended for use only in tests using dalvikvm
and needing access to platform libraries.
Bug: http://b/28449304
Change-Id: I402457d0da542996ccf265aeaa305f09881e4333
am: f34304019d
* commit 'f34304019d464de0c03fe62ca7fd9e158afb598c':
Set groups before dropping privileges.
Change-Id: If5396c913cfed13c01bd06d6a5d781439e5f6876
The code for dumping tombstones tries to read the log data. This was
silently failing after the change to drop root privileges. Fix this
by explicitly setting the groups allowed before dropping privileges.
Bug: 28590884
(cherry picked from commit edc2380109)
Change-Id: I225365edccb11f8752027566fd42f3b10ce10260
In the function dump_thread_info, if /proc/<PID>/comm is unreadable, then
threadname will be used in a strncmp causing a crash. The fix is to
avoid the check if threadname is null.
Bug: 28615417
(cherry picked from commit 039976e553)
Change-Id: I1e61431b1549ecfdc49a72cbf193a364069cda79
The previous false positive fix (df9c4a01) is enough to pass tests, and
it doesn't appear that there are any remaining issues.
Change-Id: Ib9812f1201ff0cd2ae8c8371737754fc328765b5
When someone enables wait_for_gdb, activity manager will kill the
stopped process before a developer can attach to the process. To
allow debugging in this case, change the code to only contact the
activity manager right before continuing the process that is
crashing.
Also, modify the conditions under which to do a gdb attach. The previous
code did a partial attach if perform_dump failed. The new version simply
allows an attach regardless of whether perform_dump passes or fails.
Bug: 28409358
(cherry picked from commit 9818bd2bbe)
Change-Id: I42f464b69332748e16b07d9d00f44b3aa26ce8b7
This method is called by art on UnloadNativeLibraries()
Bug: http://b/28406866
Change-Id: Ibccff93c83d6ccd3d0a1c0294a64b09281b4e05e
(cherry picked from commit 09a516bf16)
Not to be confused with <cutils/ashmem.h> or <linux/ashmem.h>, one or other
of which is what you should be using instead.
Change-Id: Ie158530591b28b94c2cda9e2686cae56b7aeb1a3
The trusty-base.mk should be included by devices that use
Trusty TEE to pull in the baseline set of Trusty specific modules.
Change-Id: I47c2095a21f47a40d390c9d5426380ad9507a708
In the case when VM is restarted all weak references
from the old VM become invalid. In such event NativeLoader
needs to clear the list of classloaders from old VM.
Bug: http://b/28453840
Bug: http://b/28449304
Change-Id: I2268c1e21cf940c57ddc5f0312f56b71aa702134
(cherry picked from commit be4ca3afc0)
Vendor blobs on ryu mprotect heap pages, causing a single chunk mapping
to appear as multiple mappings. The heap iterator has to expand the
requested range to cover the beginning of the chunk to find the chunk
metadata, which will lead to duplicate identical allocations being
reported from iterating over each of the split mappings. Silently
ignore identical allocations, and only warn on non-identical allocations
that overlap.
Bug: 28269332
Change-Id: Ied2ab9270f65d00a887c7ce1a93fbf0617d69be0
(cherry picked from commit cecd64012d)
Vendor blobs on ryu mprotect heap pages, causing segfaults when dumping
unreachable memory. Handle segfaults within HeapWalker by mapping a
zero page over any unreadable pages. HeapWalker runs in the forked
process, so the mapping will not affect the original process.
Bug: 28269332
Change-Id: I16245af722123f2ad467cbc6f245a70666c55544
(cherry picked from commit ba5d9ff6d9)
