Implement pthread_rwlock functions for NaCl newlib

Implement pthread_rwlock functions for NaCl newlib

This change was reverted at 26f5754 due to a bug
in pthread_rwlock_unlock.  This change includes
the original change along with a fix for the issue
new test case.

BUG= https://code.google.com/p/nativeclient/issues/detail?id=4105
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3960

Committed: https://chromium.googlesource.com/native_client/src/native_client/+/5995d5f911d908a92f892f8f9f2941ae8872fe0b

[Sam Clegg, 2015-02-23 23:25:53]

sbc@chromium.org changed reviewers:
+ mseaborn@chromium.org

[Sam Clegg, 2015-02-23 23:25:53]

Sending for review prior to revert of rwlocks so that
the delta can be seen clearly.

[Sam Clegg, 2015-02-26 18:42:04]

ptal

Now includes includes original change + fix.  To see the fix in isolation see
the first patchset.

[Nick Bray (chromium), 2015-03-11 19:48:10]

ncbray@chromium.org changed reviewers:
+ ncbray@chromium.org

[Nick Bray (chromium), 2015-03-11 19:48:11]

For my Mojo porting stuff I either need this change or I need to hack the Chrome
build.  ETA?

[Mark Seaborn, 2015-03-11 19:58:03]

Sam, could you possibly find another reviewer, since I'm rather busy, and it
would be good to have another person's eyes on the code.  I'd suggest Roland,
since he's very familiar with pthread semantics.

[Sam Clegg, 2015-03-11 21:28:05]

sbc@chromium.org changed reviewers:
+ mcgrathr@chromium.org

[Sam Clegg, 2015-03-11 21:28:06]

Hey Roland,

Could you take a look at this please?  Most of this code is already reviewed. 
The delta from the original is a bugfix which can be seen in patchset 1.

cheers,
sam

[Sam Clegg, 2015-03-17 23:58:37]

On 2015/03/11 21:28:06, Sam Clegg wrote:
> Hey Roland,
> 
> Could you take a look at this please?  Most of this code is already reviewed. 
> The delta from the original is a bugfix which can be seen in patchset 1.
> 
> cheers,
> sam

Gentle ping.. I'd really like to get this landed in nacl, so it can make its way
to chromium in plenty of time for the branch point so we can get some testing in
naclports.

[Roland McGrath, 2015-03-23 20:34:07]

LGTM with some comment fixes

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c
File tests/threads/rwlock_test.c (right):

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:211: * First, aquire the write lock so that all
readers
typo: "acquire"

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:220: * Now start two new threads that which will
both
s/that which/that/

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:232: * Sleep for 20ms to allow both readers to
become blocked on rdlock()
Put this duration in a macro so it's clear it can be tuned.  Then add a comment
explaining why the antipattern of an arbitrary delay in lieu of real
synchronization is the only way available to test this case.

[Sam Clegg, 2015-03-23 20:55:21]

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c
File tests/threads/rwlock_test.c (right):

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:211: * First, aquire the write lock so that all
readers
On 2015/03/23 20:34:07, Roland McGrath wrote:
> typo: "acquire"

Done.

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:220: * Now start two new threads that which will
both
On 2015/03/23 20:34:07, Roland McGrath wrote:
> s/that which/that/

Done.

https://codereview.chromium.org/951583004/diff/1/tests/threads/rwlock_test.c#...
tests/threads/rwlock_test.c:232: * Sleep for 20ms to allow both readers to
become blocked on rdlock()
On 2015/03/23 20:34:07, Roland McGrath wrote:
> Put this duration in a macro so it's clear it can be tuned.  Then add a
comment
> explaining why the antipattern of an arbitrary delay in lieu of real
> synchronization is the only way available to test this case.

Done.

[Sam Clegg, 2015-03-26 00:49:09]

The CQ bit was checked by sbc@chromium.org

[Sam Clegg, 2015-03-26 00:49:10]

The patchset sent to the CQ was uploaded after l-g-t-m from
mcgrathr@chromium.org
Link to the patchset: https://codereview.chromium.org/951583004/#ps60001 (title:
" ")

[commit-bot: I haz the power, 2015-03-26 00:49:53]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/951583004/60001

[commit-bot: I haz the power, 2015-03-26 04:40:26]

Committed patchset #4 (id:60001) as
https://chromium.googlesource.com/native_client/src/native_client/+/5995d5f91...