DDRCompare currently performs these tasks:
* Checks the magic values of all the statically initialized locks
* Scans the doubly linked list of the task list to find corruptions

Change-Id: Ie4904c2434f073d5711bb439029ffec35919846e

Add support for new shared memory address on target msm8939.

Change-Id: Ib1df74561bfbfbcc97a3b5154b1680d926c168fc

Add support for t32 mmu config for 8994 since it is different
from 8916/8939.

Change-Id: I2e666f50d34187ae259e35e4b941669c984132aa

Fix the cpu values for 8916, 8939 and 8994 so that the cpu_type
used while creating the t32_startup_script.cmm is correct.

Change-Id: Ib43aebebde5492a3aa6cbf39467ccae8a94a5fe7

Add autodump parsing support for DDR sizes greater than 2GB.

Change-Id: I9b008d08612f08fdbc9ec486e08fc6b00e7d3357

Use the correct link register to display cpu context dump for
64 bit.

Change-Id: Ic5d443bf568bdf9a9116bb061650e2d9bfda0d9c

rtb_data_offset doesn't exist anymore but it was still being passed to a
function (that did nothing with it). Remove it.

Change-Id: I4f3b03ce16fdb43b608b24ed8c61b7497d6f3867

Add parsing support for L2 indirect accesses and IRQ logging.

Change-Id: I029fbc43077813234a0c3d7436bc9b7d165c39d6

Add support for detecting 8939 target and parsing the ramdumps.

Change-Id: If96baa1759dbd450fef3a0db1be0ff5e4fb9654a

Some binaries may be compiled with a different page offset. Add
an option to allow the page offset to be adjusted properly.

Change-Id: I50cf0241f8a52aa0f9cd070da827775b5cb4764c

Add support to detect socinfo id correctly for different memory
maps of 8916.

Change-Id: Ic2d83288539e58bcaae1541f57d633a4ba66f169

Add new entries to client dump data id table to take into
account additional clients registering for v2 memory dump.

Change-Id: Ic6d155b2270c3659f5c7337b25c980c28562490c

Add support for detecting 8994 target and parsing the ramdumps

Change-Id: I8a5c039b4fd3a7552c44766f1a459024146cd0e4

Commit 4eb3aa20
(linux-ramdump-parser-v2: Update RTB for 64-bit) added support for
64-bit RTB parsing but broke backwards compatibility for old branches
that don't have the 64-bit changes. Update the parser to support
both old and new formats.

Change-Id: I6b9d834e1073d97dd3ffe43c1cc1f753d40e91b8

Some parsers may need to use the out file for their output. Currently,
when writing to stdout, we have 'None' as a parser which blows up
with dmesg parsing:

Traceback (most recent call last):
File "ramparse.py", line 318, in <module>
    p.cls(dump).parse()
File "dmesg.py", line 20, in parse
    dmesglib.DmesgLib(self.ramdump, print_out.out_file).extract_dmesg()
File "dmesglib.py", line 79, in extract_dmesg
    self.extract_dmesg_binary()
File "dmesglib.py", line 72, in extract_dmesg_binary
    self.outfile.write(f)
AttributeError: 'NoneType' object has no attribute 'write'

Make a default file of sys.stdout to do something reasonable when writing
to stdout. --stdout is mostly a debugging feature.

Change-Id: I8af80e1e6d79c4ef6660a536df02d26c28089760

The arm32 version of stack unwinding was missing a function. Add it
back.

Change-Id: I46f769938601867fd7d331e01b200071630696af

If a zillion billion differences are found in the RO area, just print to
the log once to notify the operator of that fact, instead of printing a
zillion billion times.

Change-Id: Ie61afe5b8bd2cb207f594b8ee46ebac1af95a08a

Sometimes it's useful to force a flush of the global outfile. Provide a
method to do so and use it in the main parser loop.

Change-Id: I18e9a07345c852d498b131b442df5a0c95bbcf98

It can be useful to know which parser is currently running and
how long each parser takes to run. Add some progress status
reporting to this end.

Change-Id: Ifc063feb4769fee02f0622d7a2c5ddf176ae00dc