Refactor StringPiece hash.

Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
Committed: https://crrev.com/63c013002ffbbb4c7eac172097e16ea907403e08
Cr-Commit-Position: refs/heads/master@{#374005}

[Bence, 2016-02-04 13:45:01]

Patchset #2 (id:20001) has been deleted

[Bence, 2016-02-04 16:31:59]

bnc@chromium.org changed reviewers:
+ aboxhall@chromium.org, brettw@chromium.org, danakj@chromium.org,
davidben@chromium.org, yukawa@chromium.org

[Bence, 2016-02-04 16:32:00]

danakj: PTAL at base/
yukawa: PTAL at media/midi/
davidben: PTAL at net/
brettw: PTAL at tools/gn/
aboxhall: PTAL at ui/accessibility/

Thank you all.

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
File net/quic/quic_blocked_writer_interface.h (left):

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
net/quic/quic_blocked_writer_interface.h:29: #if defined(COMPILER_GCC)
I believe that this hash<> specialization was originally added only to GCC
because of library issues, then it was just a plain hash<size_t>(ptr).  Then it
was improved to bring more entropy to the lower bits, see
https://crrev.com/467963002/diff/20001/net/quic/quic_dispatcher.h#newcode31.  It
makes sense to have this improvement on all platforms.

[dmazzoni, 2016-02-04 16:42:14]

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

[dmazzoni, 2016-02-04 16:42:14]

lgtm for ui/accessibility

[Bence, 2016-02-04 16:56:23]

Description was changed from

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
==========

to

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
==========

[Bence, 2016-02-04 16:56:23]

bnc@chromium.org changed reviewers:
- aboxhall@chromium.org

[yukawa, 2016-02-04 17:37:16]

media/midi/midi_manager_alsa.h: lgtm

[davidben, 2016-02-04 19:45:09]

davidben@chromium.org changed reviewers:
+ rch@chromium.org

[davidben, 2016-02-04 19:45:09]

davidben@chromium.org changed required reviewers:
+ rch@chromium.org

[davidben, 2016-02-04 19:45:10]

net lgtm, modulo the couple of QUIC files which rch should look at.

Thanks for doing this!

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
File net/quic/quic_blocked_writer_interface.h (left):

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
net/quic/quic_blocked_writer_interface.h:29: #if defined(COMPILER_GCC)
On 2016/02/04 16:32:00, Bence wrote:
> I believe that this hash<> specialization was originally added only to GCC
> because of library issues, then it was just a plain hash<size_t>(ptr).  Then
it
> was improved to bring more entropy to the lower bits, see
> https://crrev.com/467963002/diff/20001/net/quic/quic_dispatcher.h#newcode31. 
It
> makes sense to have this improvement on all platforms.

+rch should probably review the QUIC bits, especially because of the complex
relationship with the internal codebase. Also note that this change requires
updating all the consumers of this hash. It looks like this patch gets all of
them, but rch should probably double-check.

[danakj, 2016-02-04 19:45:52]

base LGTM but pls fix the name

https://codereview.chromium.org/1666843002/diff/40001/base/strings/string_pie...
File base/strings/string_piece.h (right):

https://codereview.chromium.org/1666843002/diff/40001/base/strings/string_pie...
base/strings/string_piece.h:454: struct StringPieceHash16 {
StringPiece16Hash

[Ryan Hamilton, 2016-02-04 20:15:14]

LGTM

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
File net/quic/quic_blocked_writer_interface.h (left):

https://codereview.chromium.org/1666843002/diff/40001/net/quic/quic_blocked_w...
net/quic/quic_blocked_writer_interface.h:29: #if defined(COMPILER_GCC)
On 2016/02/04 19:45:09, davidben (catching up) wrote:
> On 2016/02/04 16:32:00, Bence wrote:
> > I believe that this hash<> specialization was originally added only to GCC
> > because of library issues, then it was just a plain hash<size_t>(ptr).  Then
> it
> > was improved to bring more entropy to the lower bits, see
> > https://crrev.com/467963002/diff/20001/net/quic/quic_dispatcher.h#newcode31.

> It
> > makes sense to have this improvement on all platforms.
> 
> +rch should probably review the QUIC bits, especially because of the complex
> relationship with the internal codebase. Also note that this change requires
> updating all the consumers of this hash. It looks like this patch gets all of
> them, but rch should probably double-check.

Interestingly, the internal version does not need this specialization so this
code is only in the chromium version.

[brettw, 2016-02-04 20:26:54]

lgtm

[Bence, 2016-02-04 20:28:20]

Thank you all for the quick review.

https://codereview.chromium.org/1666843002/diff/40001/base/strings/string_pie...
File base/strings/string_piece.h (right):

https://codereview.chromium.org/1666843002/diff/40001/base/strings/string_pie...
base/strings/string_piece.h:454: struct StringPieceHash16 {
On 2016/02/04 19:45:52, danakj wrote:
> StringPiece16Hash

Done.

[Bence, 2016-02-04 20:28:38]

The CQ bit was checked by bnc@chromium.org

[Bence, 2016-02-04 20:28:38]

The patchset sent to the CQ was uploaded after l-g-t-m from yukawa@chromium.org,
davidben@chromium.org, rch@chromium.org, danakj@chromium.org,
dmazzoni@chromium.org, brettw@chromium.org
Link to the patchset: https://codereview.chromium.org/1666843002/#ps50001
(title: "s/StringPieceHash16/StringPiece16Hash/")

[commit-bot: I haz the power, 2016-02-04 20:30:30]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1666843002/50001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1666843002/50001

[commit-bot: I haz the power, 2016-02-04 23:12:51]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-02-04 23:12:53]

Try jobs failed on following builders:
  win8_chromium_ng on tryserver.chromium.win (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.win/builders/win8_chromium_ng/...)

[Bence, 2016-02-05 16:15:23]

bnc@chromium.org changed reviewers:
+ erg@chromium.org

[Bence, 2016-02-05 16:15:24]

Elliot: PTAL at components/mus/.  Thank you.

[Elliot Glaysher, 2016-02-05 18:27:03]

lgtm

[Bence, 2016-02-06 03:53:33]

The CQ bit was checked by bnc@chromium.org

[Bence, 2016-02-06 03:53:33]

The patchset sent to the CQ was uploaded after l-g-t-m from yukawa@chromium.org,
davidben@chromium.org, rch@chromium.org, danakj@chromium.org,
dmazzoni@chromium.org, brettw@chromium.org
Link to the patchset: https://codereview.chromium.org/1666843002/#ps90001
(title: "Sort includes.")

[commit-bot: I haz the power, 2016-02-06 03:53:46]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1666843002/90001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1666843002/90001

[commit-bot: I haz the power, 2016-02-06 03:58:25]

Description was changed from

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
==========

to

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
==========

[commit-bot: I haz the power, 2016-02-06 03:58:26]

Committed patchset #5 (id:90001)

[commit-bot: I haz the power, 2016-02-06 03:59:46]

Description was changed from

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864
==========

to

==========
Refactor StringPiece hash.

Rename struct BASE_HASH_NAMESPACE::hash<base::StringPiece> to struct
base::StringPieceHash, and modify existing code as necessary.  This is motivated
by https://crrev.com/1660273002/#msg6.

* Remove #include "base/containers/hash_tables.h" where not used any more.
* Add this include for hash_map where necessary (string_piece.h does not include
  it any longer, so this might be necessary for the code to compile).
* Change net::linked_hash_map from base::hash_map to std::unordered_map and add
  Hash template parameter.  This way BASE_HASH_NAMESPACE::hash<Key> will not be
  implicitly used any more, so move QuicBlockedWriterInterfacePtrHash out of
  that namespace.

BUG=576864

Committed: https://crrev.com/63c013002ffbbb4c7eac172097e16ea907403e08
Cr-Commit-Position: refs/heads/master@{#374005}
==========

[commit-bot: I haz the power, 2016-02-06 03:59:47]

Patchset 5 (id:??) landed as
https://crrev.com/63c013002ffbbb4c7eac172097e16ea907403e08
Cr-Commit-Position: refs/heads/master@{#374005}