Implement core dump to minidump conversion.

Implement core dump to minidump conversion.

This patch is part of a bigger patch that helps merging the breakpad code
with the modified version in Chromium OS.

Specifically, this patch makes the following changes:
1. Turn the LinuxDumper class into a base class and move ptrace related
   code into a new derived class, LinuxPtraceDumper.
2. Add a LinuxCoreDumper class, which is derived from LinuxDumper, to
   extract information from a crashed process via a core dump file instead
   of ptrace.
3. Add a WriteMinidumpFromCore function to
     src/client/linux/minidump_writer/minidump_writer.h,
   which uses LinuxCoreDumper to extract information from a core dump file.
4. Add a core2md utility, which simply wraps WriteMinidumpFromCore, for
   converting a core dump to a minidump.

BUG=455
TEST=Tested the following:
1. Build on 32-bit and 64-bit Linux with gcc 4.4.3 and gcc 4.6.
2. Build on Mac OS X 10.6.8 with gcc 4.2 and clang 3.0 (with latest gmock).
3. All unit tests pass.
4. Run Chromium OS tests to test core2md.

[kmixter]

LGTM to with a few nits and assuming that the functions refactored out of
linux_dumper and into linux_ptrace_dumper didn't change.  If they did change,
please comment on where to look in those files.

http://breakpad.appspot.com/343001/diff/1/4
File src/client/linux/minidump_writer/linux_core_dumper.cc (right):

http://breakpad.appspot.com/343001/diff/1/4#newcode81
Line 81: // If the data segment is not found in the core dump, fill the result
One question I originally had was if we should deal with CopyFromProcess'es that
were partially matching a map, or matching two separate adjacent maps.  Neither
of those cases is handled by this code, but it also never seemed to be
necessary.  May be worth a comment.

http://breakpad.appspot.com/343001/diff/1/6
File src/client/linux/minidump_writer/linux_core_dumper_unittest.cc (right):

http://breakpad.appspot.com/343001/diff/1/6#newcode104
Line 104: EXPECT_EQ(0, dumper.crash_address());
may want to say that this is testing a limitation of the core dumper.

http://breakpad.appspot.com/343001/diff/1/7
File src/client/linux/minidump_writer/linux_dumper.cc (left):

http://breakpad.appspot.com/343001/diff/1/7#oldcode138
Line 138: bool LinuxDumper::ThreadsSuspend() {
I'm assuming these are being copied exactly the same over to the ptrace dumper.

http://breakpad.appspot.com/343001/diff/1/11
File src/client/linux/minidump_writer/linux_ptrace_dumper.h (right):

http://breakpad.appspot.com/343001/diff/1/11#newcode51
Line 51: // result.|node| is the final node without any slashes. Returns true on
spacing

http://breakpad.appspot.com/343001/diff/1/12
File src/client/linux/minidump_writer/linux_ptrace_dumper_unittest.cc (right):

http://breakpad.appspot.com/343001/diff/1/12#newcode1
Line 1: // Copyright (c) 2009, Google Inc.
assuming this is a renamed file.

[Ben Chan]

Most of the code was copied from linux_dumper.cc to linux_ptrace_dumper.cc
except the changes made to BuildProcPath.

http://breakpad.appspot.com/343001/diff/1/4
File src/client/linux/minidump_writer/linux_core_dumper.cc (right):

http://breakpad.appspot.com/343001/diff/1/4#newcode81
Line 81: // If the data segment is not found in the core dump, fill the result
On 2012/01/18 01:28:57, kmixter wrote:
> One question I originally had was if we should deal with CopyFromProcess'es
that
> were partially matching a map, or matching two separate adjacent maps. 
Neither
> of those cases is handled by this code, but it also never seemed to be
> necessary.  May be worth a comment.

I had the same question. Added a TODO for further investigation.

http://breakpad.appspot.com/343001/diff/1/6
File src/client/linux/minidump_writer/linux_core_dumper_unittest.cc (right):

http://breakpad.appspot.com/343001/diff/1/6#newcode104
Line 104: EXPECT_EQ(0, dumper.crash_address());
On 2012/01/18 01:28:57, kmixter wrote:
> may want to say that this is testing a limitation of the core dumper.

Done.

http://breakpad.appspot.com/343001/diff/1/7
File src/client/linux/minidump_writer/linux_dumper.cc (left):

http://breakpad.appspot.com/343001/diff/1/7#oldcode138
Line 138: bool LinuxDumper::ThreadsSuspend() {
On 2012/01/18 01:28:57, kmixter wrote:
> I'm assuming these are being copied exactly the same over to the ptrace
dumper.

Yes, except for the BuildProc changes.

http://breakpad.appspot.com/343001/diff/1/11
File src/client/linux/minidump_writer/linux_ptrace_dumper.h (right):

http://breakpad.appspot.com/343001/diff/1/11#newcode51
Line 51: // result.|node| is the final node without any slashes. Returns true on
On 2012/01/18 01:28:57, kmixter wrote:
> spacing

Done.

http://breakpad.appspot.com/343001/diff/1/12
File src/client/linux/minidump_writer/linux_ptrace_dumper_unittest.cc (right):

http://breakpad.appspot.com/343001/diff/1/12#newcode1
Line 1: // Copyright (c) 2009, Google Inc.
On 2012/01/18 01:28:57, kmixter wrote:
> assuming this is a renamed file.

Yes, it's a renamed as indicated in the comment below. So I kept the copyright
year unchanged.

[kmixter]

Still LGTM