[WeakMemoryCache] Remove LRU lists, prune order control and live/dead distinction

[WeakMemoryCache] Remove LRU lists, prune order control and live/dead distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462
Committed: https://crrev.com/a756f82e21bd1af1025d373c099c8a7e7b589514
Cr-Commit-Position: refs/heads/master@{#432972}

[hiroshige, 2016-10-13 09:37:47]

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

[commit-bot: I haz the power, 2016-10-13 09:37:56]

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, 2016-10-13 09:47:36]

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

[commit-bot: I haz the power, 2016-10-13 09:47:37]

Dry run: Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[hiroshige, 2016-10-19 05:48:22]

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

[commit-bot: I haz the power, 2016-10-19 05:49:08]

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

[hiroshige, 2016-10-19 07:25:24]

Description was changed from

==========
Remove LRU list from MemoryCache

BUG=
==========

to

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

BUG=603462
==========

[hiroshige, 2016-10-19 07:29:48]

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

[hiroshige, 2016-10-19 07:36:59]

The CQ bit was unchecked by hiroshige@chromium.org

[hiroshige, 2016-10-19 07:37:28]

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

[hiroshige, 2016-10-19 08:31:31]

The CQ bit was unchecked by hiroshige@chromium.org

[hiroshige, 2016-10-19 11:40:58]

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

[commit-bot: I haz the power, 2016-10-19 11:41:33]

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, 2016-10-19 13:29:07]

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

[commit-bot: I haz the power, 2016-10-19 13:29:07]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[hiroshige, 2016-10-20 05:02:40]

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

[commit-bot: I haz the power, 2016-10-20 05:03:00]

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, 2016-10-20 06:38:56]

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

[commit-bot: I haz the power, 2016-10-20 06:38:57]

Dry run: This issue passed the CQ dry run.

[hiroshige, 2016-10-20 06:41:26]

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

[commit-bot: I haz the power, 2016-10-20 06:41:42]

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, 2016-10-20 08:15:40]

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

[commit-bot: I haz the power, 2016-10-20 08:15:41]

Dry run: This issue passed the CQ dry run.

[hiroshige, 2016-10-21 05:40:13]

Description was changed from

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

BUG=603462
==========

to

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462
==========

[hiroshige, 2016-10-21 05:40:13]

hiroshige@chromium.org changed reviewers:
+ japhet@chromium.org, yhirano@chromium.org

[hiroshige, 2016-10-21 07:04:48]

PTAL.

[Nate Chapin, 2016-10-21 18:48:20]

:D !!!

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/ImageResource.cpp (left):

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/ImageResource.cpp:483: void
ImageResource::didDraw(const blink::Image* image) {
Looks like this callback is ignored by all ImageObserver subclasses. We should
get rid fo it.

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCache.h (right):

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.h:133: void
setCapacities(size_t totalBytes);
Nit: there's only one variable, so rename to setCapacity().

[hiroshige, 2016-10-24 05:29:35]

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

[commit-bot: I haz the power, 2016-10-24 05:29:44]

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, 2016-10-24 08:04:17]

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

[commit-bot: I haz the power, 2016-10-24 08:04:17]

Dry run: This issue passed the CQ dry run.

[yhirano, 2016-10-24 08:14:00]

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/css/CSSStyleSheetResourceTest.cpp (right):

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/css/CSSStyleSheetResourceTest.cpp:63:
TEST_F(CSSStyleSheetResourceTest, PruneCanCauseEviction) {
According to https://codereview.chromium.org/1537343002 this test is meaningless
without the deleted lines. Maybe you can remove the entire test case.

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCache.cpp (right):

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:195: if (!resource ||
resource->url().isNull() || resource->url().isEmpty())
String::isNull() == true implies String::isEmpty() == true

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:253: DCHECK(resource);
Does this always hold?

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:340: DCHECK(resource);
Does this always hold?

[hiroshige, 2016-10-24 10:35:28]

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/ImageResource.cpp (left):

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/ImageResource.cpp:483: void
ImageResource::didDraw(const blink::Image* image) {
On 2016/10/21 18:48:20, Nate Chapin wrote:
> Looks like this callback is ignored by all ImageObserver subclasses. We should
> get rid fo it.

Created a separate patch: https://codereview.chromium.org/2447623002/
(to prevent this CL from becoming larger)

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCache.h (right):

https://codereview.chromium.org/2411243004/diff/140001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.h:133: void
setCapacities(size_t totalBytes);
On 2016/10/21 18:48:20, Nate Chapin wrote:
> Nit: there's only one variable, so rename to setCapacity().

Done.

[hiroshige, 2016-10-24 10:48:18]

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

[hiroshige, 2016-10-24 10:48:25]

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/css/CSSStyleSheetResourceTest.cpp (right):

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/css/CSSStyleSheetResourceTest.cpp:63:
TEST_F(CSSStyleSheetResourceTest, PruneCanCauseEviction) {
On 2016/10/24 08:14:00, yhirano wrote:
> According to https://codereview.chromium.org/1537343002 this test is
meaningless
> without the deleted lines. Maybe you can remove the entire test case.

Done.

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCache.cpp (right):

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:195: if (!resource ||
resource->url().isNull() || resource->url().isEmpty())
On 2016/10/24 08:14:00, yhirano wrote:
> String::isNull() == true implies String::isEmpty() == true

Done.

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:253: DCHECK(resource);
On 2016/10/24 08:14:00, yhirano wrote:
> Does this always hold?

I expect so, because clearResourceWeak() is called when resource is garbage
collected and thus MemoryCacheEntry is evicted from the map before resource()
becomes null.

https://codereview.chromium.org/2411243004/diff/160001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCache.cpp:340: DCHECK(resource);
On 2016/10/24 08:14:00, yhirano wrote:
> Does this always hold?

ditto.

[commit-bot: I haz the power, 2016-10-24 10:48:29]

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, 2016-10-24 12:28:25]

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

[commit-bot: I haz the power, 2016-10-24 12:28:26]

Dry run: This issue passed the CQ dry run.

[yhirano, 2016-10-26 09:47:51]

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp (right):

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:149: // Next task: now,
the live resource was evicted.
"the resources were pruned"?

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:153: static void
TestLiveResourceEvictionAtEndOfTask(Resource* resource1,
I'mt not sure what "eviction" means. IIUC The resources will not be evicted
automatically as we have strong references, right?

[hiroshige, 2016-11-04 07:50:47]

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

[commit-bot: I haz the power, 2016-11-04 07:50:59]

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

[hiroshige, 2016-11-04 07:50:59]

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp (right):

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:149: // Next task: now,
the live resource was evicted.
On 2016/10/26 09:47:51, yhirano wrote:
> "the resources were pruned"?

Done.

https://codereview.chromium.org/2411243004/diff/180001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:153: static void
TestLiveResourceEvictionAtEndOfTask(Resource* resource1,
On 2016/10/26 09:47:51, yhirano wrote:
> I'mt not sure what "eviction" means. IIUC The resources will not be evicted
> automatically as we have strong references, right?

Fixed the function name and test names ("pruning" rather than "eviction").
This CL removes explicit Resource eviction because we merely have weak
references to Resource (and let unused Resources evicted due to Oilpan GC).

[commit-bot: I haz the power, 2016-11-04 09:03:36]

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

[commit-bot: I haz the power, 2016-11-04 09:03:38]

Dry run: This issue passed the CQ dry run.

[yhirano, 2016-11-14 05:59:55]

Sorry for the long delay. LGTM.

https://codereview.chromium.org/2411243004/diff/220001/third_party/WebKit/Sou...
File third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp (right):

https://codereview.chromium.org/2411243004/diff/220001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:153: static void
TestResourcePruningAtEndOfTask(Resource* resource1,
camelCase

https://codereview.chromium.org/2411243004/diff/220001/third_party/WebKit/Sou...
third_party/WebKit/Source/core/fetch/MemoryCacheTest.cpp:222: static void
TestClientRemoval(Resource* resource1, Resource* resource2) {
camelCase

[hiroshige, 2016-11-16 09:24:05]

Patchset #16 (id:300001) has been deleted

[hiroshige, 2016-11-16 09:24:13]

Patchset #13 (id:240001) has been deleted

[hiroshige, 2016-11-16 09:27:49]

hiroshige@chromium.org changed reviewers:
+ mkwst@chromium.org

[hiroshige, 2016-11-16 09:27:50]

+mkwst@, could you take a look, especially at core/css and web/ changes as a
core/ and web/ OWNER?

[hiroshige, 2016-11-16 09:29:02]

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

[commit-bot: I haz the power, 2016-11-16 09:29:17]

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, 2016-11-16 11:20:36]

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

[commit-bot: I haz the power, 2016-11-16 11:20:37]

Dry run: This issue passed the CQ dry run.

[Mike West, 2016-11-17 14:32:40]

LGTM. Thanks!

You mention `web_cache:effective_size_avg` as being not _actually_ affected: are
there any metrics you think will actually be affected? What should we keep our
eye on when this lands?

[hiroshige, 2016-11-17 15:42:35]

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

[commit-bot: I haz the power, 2016-11-17 15:42:53]

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

[hiroshige, 2016-11-17 15:47:06]

On 2016/11/17 14:32:40, Mike West wrote:
> LGTM. Thanks!
> 
> You mention `web_cache:effective_size_avg` as being not _actually_ affected:
are
> there any metrics you think will actually be affected? What should we keep our
> eye on when this lands?

- Memory usage: other memory metrics other than web_cache (because I expect they
do not depend on Resource enumeration by MemoryCache),
- Performance (execution time) metrics (this CL might cause e.g. more (or less)
re-decoding of images), and
- behavioral changes in the wild.

[hiroshige, 2016-11-17 15:55:47]

The CQ bit was unchecked by hiroshige@chromium.org

[hiroshige, 2016-11-17 15:55:53]

The CQ bit was checked by hiroshige@chromium.org

[hiroshige, 2016-11-17 15:55:53]

The patchset sent to the CQ was uploaded after l-g-t-m from yhirano@chromium.org
Link to the patchset: https://codereview.chromium.org/2411243004/#ps320001
(title: "Reflect yhirano's comment")

[commit-bot: I haz the power, 2016-11-17 15:56:25]

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, 2016-11-17 17:30:15]

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

[commit-bot: I haz the power, 2016-11-17 17:30:17]

Try jobs failed on following builders:
  linux_chromium_asan_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[hiroshige, 2016-11-17 17:52:57]

The CQ bit was checked by hiroshige@chromium.org

[commit-bot: I haz the power, 2016-11-17 17:53:45]

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, 2016-11-17 21:43:53]

Description was changed from

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462
==========

to

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462
==========

[commit-bot: I haz the power, 2016-11-17 21:43:56]

Committed patchset #15 (id:320001)

[commit-bot: I haz the power, 2016-11-17 21:48:46]

Description was changed from

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462
==========

to

==========
[WeakMemoryCache] Remove LRU lists, prune order control and live/dead
distinction

This CL removes most of the control mechanisms for prune ordering of Resources
(except one using |m_lastDecodedAccessTime|), including those using the
LRU order, |m_lastDecodedAccessTime|, |m_maxDeferredPruneDeadCapacity|, and
|m_accessCount|.
This enable us to:
- Remove updateDecodedResource(), updateForAccess() and thus
  didAccessDecodedData().
- Make update() to do nothing except for |m_size| update.
- Remove |m_liveDecodedResources| and |m_allResources| and make Resources
  managed only by the simple hashmaps in |m_resourceMaps|.
- Remove live/dead distinction in MemoryCache.
  - Remove min/max dead capacities.
  - Merge liveSize() and deadSize() into size().

Removal of min/max dead capacities and live/dead distinction causes more code
outside core/fetch to be unused.
https://codereview.chromium.org/2435603002/ will remove such unused code.

This CL also
- Removes a prune() call in didAccessDecodedData().
- Adds |const| to some methods.

This CL increases web_cache size reported in memory-infra (e.g.
memory.top_10_mobile performance test's web_cache:effective_size_avg), but
this doesn't affect the real memory usage.
Previously, dead Resources were evicted during pruneDeadResources() and such
Resources were not reported to memory-infra.
This CL make such Resources to be reported to memory-infra because this CL
doesn't evict Resources during pruneResources().

BUG=603462

Committed: https://crrev.com/a756f82e21bd1af1025d373c099c8a7e7b589514
Cr-Commit-Position: refs/heads/master@{#432972}
==========

[commit-bot: I haz the power, 2016-11-17 21:48:48]

Patchset 15 (id:??) landed as
https://crrev.com/a756f82e21bd1af1025d373c099c8a7e7b589514
Cr-Commit-Position: refs/heads/master@{#432972}