Avoid large inlined automatic methods in IndexedDB code.

Avoid large inlined automatic methods in IndexedDB code.

The chromium base policy is to not inline code but since automatically
generated code and templates easily become inlined it can happen
anyway. This explicitly outlines destructors, copy constructors and
assignment operators in the IndexedDB code for a win of
11 KB of x64 machine code. Most of it (maybe 8 KB) from IndexedDBKey.

clang x64 numbers:
Total change: -11795 bytes
==========================
  42 added, totalling +3270 bytes across 13 sources
  25 removed, totalling -5883 bytes across 14 sources
  9 grown, for a net change of +1556 bytes (7936 bytes before, 9492 bytes after) across 6 sources
  60 shrunk, for a net change of -10738 bytes (54277 bytes before, 43539 bytes after) across 17 sources

BUG=

Committed: https://crrev.com/1027bb83a75224af86c8ab8c3526a1dba5fc7243
Cr-Commit-Position: refs/heads/master@{#317090}

[Daniel Bratell, 2015-02-17 14:09:47]

bratell@opera.com changed reviewers:
+ jsbell@chromium.org

[Daniel Bratell, 2015-02-17 14:11:01]

jsbell, can you please take a look.

IndexedDB classes still turn up at the top of code that suffers (size-wise) from
much inlining. This was an attempt to make things better, now when we can use
=default and don't need to worry about getting hard-to-maintain code.

[cmumford, 2015-02-17 16:34:49]

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

[cmumford, 2015-02-17 16:34:50]

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_backing_store.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_backing_store.cc:4409: operator=(const
WriteDescriptor& other) = default;
Space before }  // namespace

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_backing_store.h (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_backing_store.h:182: WriteDescriptor&
operator=(const WriteDescriptor& other);
operator= is never called, so you could suppress this if you want.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_blob_info.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_blob_info.cc:55: const IndexedDBBlobInfo&
other) = default;
space between methods.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_metadata.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_metadata.cc:9:
IndexedDBIndexMetadata::IndexedDBIndexMetadata() = default;
These methods are a little tight. Can we have a space in between each of them?

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_metadata.h (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_metadata.h:26: ~IndexedDBIndexMetadata();
Now that these functions can't be inlined you need to add CONTENT_EXPORT so that
all targets can link.

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
File content/public/browser/indexed_db_info.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
content/public/browser/indexed_db_info.cc:20: IndexedDBInfo&
IndexedDBInfo::operator=(IndexedDBInfo& other) = default;
Need space before closing brace.

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
File content/public/browser/indexed_db_info.h (right):

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
content/public/browser/indexed_db_info.h:23: IndexedDBInfo&
operator=(IndexedDBInfo& other);
"other" param must be const.

[Daniel Bratell, 2015-02-18 13:55:45]

Thanks for the review. I think this version should be better formatted (though
long class names and narrow max widths don't work well together. :-/ ).

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_backing_store.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_backing_store.cc:4409: operator=(const
WriteDescriptor& other) = default;
On 2015/02/17 16:34:50, cmumford wrote:
> Space before }  // namespace

Done.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_backing_store.h (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_backing_store.h:182: WriteDescriptor&
operator=(const WriteDescriptor& other);
On 2015/02/17 16:34:50, cmumford wrote:
> operator= is never called, so you could suppress this if you want.

You might be right (I did notice that some of the assignment operators were not
present in the resulting binary) but it's almost impossible to know when and if
a container uses that operator so it's best to leave them be and let the
compiler strip/ignore them if they are not used.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_blob_info.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_blob_info.cc:55: const IndexedDBBlobInfo&
other) = default;
On 2015/02/17 16:34:50, cmumford wrote:
> space between methods.

Done.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_metadata.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_metadata.cc:9:
IndexedDBIndexMetadata::IndexedDBIndexMetadata() = default;
On 2015/02/17 16:34:50, cmumford wrote:
> These methods are a little tight. Can we have a space in between each of them?

Done.

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
File content/browser/indexed_db/indexed_db_metadata.h (right):

https://codereview.chromium.org/933823002/diff/1/content/browser/indexed_db/i...
content/browser/indexed_db/indexed_db_metadata.h:26: ~IndexedDBIndexMetadata();
On 2015/02/17 16:34:50, cmumford wrote:
> Now that these functions can't be inlined you need to add CONTENT_EXPORT so
that
> all targets can link.

Done.

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
File content/public/browser/indexed_db_info.cc (right):

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
content/public/browser/indexed_db_info.cc:20: IndexedDBInfo&
IndexedDBInfo::operator=(IndexedDBInfo& other) = default;
On 2015/02/17 16:34:50, cmumford wrote:
> Need space before closing brace.

Done.

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
File content/public/browser/indexed_db_info.h (right):

https://codereview.chromium.org/933823002/diff/1/content/public/browser/index...
content/public/browser/indexed_db_info.h:23: IndexedDBInfo&
operator=(IndexedDBInfo& other);
On 2015/02/17 16:34:50, cmumford wrote:
> "other" param must be const.

Right. Funny that the compiler didn't mention it. It protested violently if I
had the return type wrong (I tried returnine void from the assignment
operators).

Fixed.

[jsbell, 2015-02-18 22:29:38]

lgtm

Some of the impls are still not spaced out, but clang-format doesn't complain so
I won't either.

[Daniel Bratell, 2015-02-19 09:33:00]

On 2015/02/18 22:29:38, jsbell wrote:
> lgtm
> 
> Some of the impls are still not spaced out, but clang-format doesn't complain
so
> I won't either.

It is what git cl format created. I've decided to stop fight it and take
whatever ugliness it generates, since it saves time for everyone.

Thanks for the review!

[Daniel Bratell, 2015-02-19 09:33:29]

The CQ bit was checked by bratell@opera.com

[commit-bot: I haz the power, 2015-02-19 09:33:41]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/933823002/40001

[commit-bot: I haz the power, 2015-02-19 09:37:10]

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

[commit-bot: I haz the power, 2015-02-19 09:37:11]

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

[Daniel Bratell, 2015-02-19 13:42:28]

bratell@opera.com changed reviewers:
+ jam@chromium.org

[Daniel Bratell, 2015-02-19 13:42:28]

jam, can you please take a look at the changes in content/public? jsbell has
done a review but he's not OWNER of that directory.

[jam, 2015-02-19 19:05:51]

lgtm

[jsbell, 2015-02-19 19:06:19]

The CQ bit was checked by jsbell@chromium.org

[commit-bot: I haz the power, 2015-02-19 19:08:19]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/933823002/40001

[commit-bot: I haz the power, 2015-02-19 19:13:06]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-02-19 19:14:21]

Patchset 3 (id:??) landed as
https://crrev.com/1027bb83a75224af86c8ab8c3526a1dba5fc7243
Cr-Commit-Position: refs/heads/master@{#317090}