[base] Remove PointersMatch, making a separate std::equals hashmap

Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.

Committed: https://crrev.com/050bf39d85e592f86bfb89c7ea6819a3e152dd00
Cr-Commit-Position: refs/heads/master@{#39920}

[Leszek Swirski, 2016-09-26 16:33:43]

leszeks@chromium.org changed reviewers:
+ verwaest@chromium.org

[Leszek Swirski, 2016-09-26 16:33:44]

Hi Toon, this is my CL for specialising hashmaps that only check pointer
equality. Right now it won't save memory yet, because of the whole empty base
thing, and I haven't measured the performance impact (if any), but I thought you
might want to see it anyway.

[Leszek Swirski, 2016-09-26 17:12:32]

Description was changed from

==========
Remove PointersMatch, making a separate std::equals hashmap
==========

to

==========
Instead of having a hashmap with a function pointer defining its matching
function, creates a hashmap the specialises the case of keys that simply check
pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

[Toon Verwaest, 2016-09-27 12:46:36]

lgtm

[Leszek Swirski, 2016-09-28 15:40:00]

Description was changed from

==========
Instead of having a hashmap with a function pointer defining its matching
function, creates a hashmap the specialises the case of keys that simply check
pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

to

==========
Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

[Leszek Swirski, 2016-09-28 15:42:20]

leszeks@chromium.org changed reviewers:
+ jpp@chromium.org, marja@chromium.org, mstarzinger@chromium.org

[Leszek Swirski, 2016-09-28 15:42:20]

jpp@chromium.org: Please review changes in 
asmjs

marja@chromium.org: Please review changes in 
parsing (and maybe ast, though that got an lgtm from toon already)

mstarzinger@chromium.org: Please review changes in 
heap and compiler

[Michael Starzinger, 2016-09-29 09:28:32]

LGTM (rubber-stamp of "compiler" and "heap", didn't look at the rest).

[marja, 2016-09-30 10:27:33]

lgtm % style nits below

I was for a moment thinking about the parameter order, since now some hashmap
related funcs / ctors take them as (match func, capacity) and some (capacity,
match func) but but... I guess it's needed for being able to give
TemplateHashMapImpl just the capacity, not the match func. And for
CustomMatcherTemplateHashMapImpl we want to be able to just give the match func
and get the default capacity. So can't really help it.

https://codereview.chromium.org/2369963002/diff/1/src/profiler/heap-snapshot-...
File src/profiler/heap-snapshot-generator.cc (right):

https://codereview.chromium.org/2369963002/diff/1/src/profiler/heap-snapshot-...
src/profiler/heap-snapshot-generator.cc:359: :
next_id_(kFirstAvailableObjectId), entries_map_(), heap_(heap) {
Since entries_map_ can now use the default ctor, can you remove it from this
line?

https://codereview.chromium.org/2369963002/diff/1/test/cctest/test-hashmap.cc
File test/cctest/test-hashmap.cc (right):

https://codereview.chromium.org/2369963002/diff/1/test/cctest/test-hashmap.cc...
test/cctest/test-hashmap.cc:43: explicit IntSet(IntKeyHash hash) : hash_(hash),
map_() {}
Ditto

[Leszek Swirski, 2016-09-30 10:44:28]

Thanks Marja! I hear you about the constructor order problem -- I entertained
the idea of replacing the constructors with builders, but I started getting Java
PTSD flashbacks before I could even start writing the code. Plus builders
(probably?) don't play well with member initialisers. My kingdom for named
parameters...

https://codereview.chromium.org/2369963002/diff/1/src/profiler/heap-snapshot-...
File src/profiler/heap-snapshot-generator.cc (right):

https://codereview.chromium.org/2369963002/diff/1/src/profiler/heap-snapshot-...
src/profiler/heap-snapshot-generator.cc:359: :
next_id_(kFirstAvailableObjectId), entries_map_(), heap_(heap) {
On 2016/09/30 10:27:33, marja wrote:
> Since entries_map_ can now use the default ctor, can you remove it from this
> line?

Done.

https://codereview.chromium.org/2369963002/diff/1/test/cctest/test-hashmap.cc
File test/cctest/test-hashmap.cc (right):

https://codereview.chromium.org/2369963002/diff/1/test/cctest/test-hashmap.cc...
test/cctest/test-hashmap.cc:43: explicit IntSet(IntKeyHash hash) : hash_(hash),
map_() {}
On 2016/09/30 10:27:33, marja wrote:
> Ditto

Done.

[Leszek Swirski, 2016-09-30 12:01:39]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-09-30 12:01:42]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 12:06:11]

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

[commit-bot: I haz the power, 2016-09-30 12:06:11]

Dry run: Try jobs failed on following builders:
  v8_linux_arm_rel_ng on master.tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel_ng/builds/...)

[Leszek Swirski, 2016-09-30 13:00:04]

The CQ bit was checked by leszeks@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-09-30 13:00:08]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 13:28:56]

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

[commit-bot: I haz the power, 2016-09-30 13:28:57]

Dry run: This issue passed the CQ dry run.

[John, 2016-09-30 16:13:34]

The CQ bit was checked by jpp@chromium.org

[John, 2016-09-30 16:13:35]

lgtm

[John, 2016-09-30 16:13:36]

The patchset sent to the CQ was uploaded after l-g-t-m from
verwaest@chromium.org, mstarzinger@chromium.org, marja@chromium.org
Link to the patchset: https://codereview.chromium.org/2369963002/#ps60001
(title: "Fix the other simulators")

[commit-bot: I haz the power, 2016-09-30 16:13:47]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 16:16:54]

Description was changed from

==========
Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

to

==========
Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

[commit-bot: I haz the power, 2016-09-30 16:16:55]

Committed patchset #4 (id:60001)

[commit-bot: I haz the power, 2016-09-30 16:17:15]

Description was changed from

==========
Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.
==========

to

==========
Instead of having a hashmap with a function pointer defining its
matching function, creates a hashmap the specialises the case of keys
that simply check pointer equality.

I measure an average ~1% improvement on Octane code-load.

Committed: https://crrev.com/050bf39d85e592f86bfb89c7ea6819a3e152dd00
Cr-Commit-Position: refs/heads/master@{#39920}
==========

[commit-bot: I haz the power, 2016-09-30 16:17:16]

Patchset 4 (id:??) landed as
https://crrev.com/050bf39d85e592f86bfb89c7ea6819a3e152dd00
Cr-Commit-Position: refs/heads/master@{#39920}