Lock-free GamepadSeqLock (try 2)

Lock-free GamepadSeqLock (try 2)
The change
- provides an improved lock-free SeqLock implementation which eliminates any potential blocking of readers.
- provides a higher-level and simpler API as was suggested by Darin.
- ThreadSanitizer report suppressions are replaced with correct synchronization. 
- eliminates nasty kMaximumContentionCount and associated histogram.
Oritignal review URL:
http://codereview.chromium.org/8772004
TBR=darin@chromium.org

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=146792

[dvyukov, 2012-07-04 12:56:40]

Thi is essentially:
http://codereview.chromium.org/8772004
which was committed and then reverted. The only change is:

GamepadSeqLockBase::GamepadSeqLockBase(BaseEntry* entries, size_t size)
    : size_(size)
    , current_(0) {
-  const size_t ent_size = sizeof(BaseEntry) + sizeof(Atomic32) * (size + 1);
+  const size_t ent_size = sizeof(BaseEntry) + sizeof(Atomic32) * size;

[dvyukov, 2012-07-04 13:02:30]

mac_valgrind, linux_tsan, linux_rel failures seem to be unrelated.
linux_asan is now green (it pointed to the bug).

[commit-bot: I haz the power, 2012-07-04 13:08:17]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer or
a lowly provisional committer, _not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[dvyukov, 2012-07-13 11:05:35]

On 2012/07/04 13:08:17, I haz the power (commit-bot) wrote:
> No LGTM from a valid reviewer yet. Only full committers are accepted.
> Even if an LGTM may have been provided, it was from a non-committer or
> a lowly provisional committer, _not_ a full super star committer.
> See http://www.chromium.org/getting-involved/become-a-committer
> Note that this has nothing to do with OWNERS files.

Hi,

I need a new LGTM for this CL.
Since it's essentially the same as http://codereview.chromium.org/8772004, I
think the single LGTM from darin will do.

[darin (slow to review), 2012-07-13 16:35:28]

OK, LGTM

[commit-bot: I haz the power, 2012-07-16 11:56:43]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/dvyukov@chromium.org/10704041/2003

[commit-bot: I haz the power, 2012-07-16 12:55:45]

Change committed as 146792

[Alexander Potapenko, 2014-11-24 15:22:48]

On 2012/07/16 12:55:45, I haz the power (commit-bot) wrote:
> Change committed as 146792

This CL has been reverted soon after that
(https://chromiumcodereview.appspot.com/10785024) and has never been revisited
since then.
Today the original seqlock code has been moved to
content/common/one_writer_seqlock.cc, parts of it are also used in
ppapi/proxy/gamepad_resource.cc (where they do not make much sense, IIUC).
I can try applying this CL to one_writer_seqlock.cc, not sure what to do with
gamepad_resource.cc.