[serializer] encode recent long-encoded root list items as hot objects.

[serializer] encode recent long-encoded root list items as hot objects.

We have a lot of long-encoded root list items in type feedback vectors.

Committed: https://crrev.com/734898a7b2d7f2f03ec9bdfe9bd6484d3c04fa32
Cr-Commit-Position: refs/heads/master@{#37237}

[Yang, 2016-06-22 14:09:47]

yangguo@chromium.org changed reviewers:
+ vogelheim@chromium.org

[Yang, 2016-06-22 14:09:52]

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

[commit-bot: I haz the power, 2016-06-22 14:09:59]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2090563002/1

[commit-bot: I haz the power, 2016-06-22 14:38:51]

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

[commit-bot: I haz the power, 2016-06-22 14:38:52]

Dry run: This issue passed the CQ dry run.

[vogelheim, 2016-06-23 13:39:15]

lgtm

https://codereview.chromium.org/2090563002/diff/1/src/snapshot/serializer.cc
File src/snapshot/serializer.cc (right):

https://codereview.chromium.org/2090563002/diff/1/src/snapshot/serializer.cc#...
src/snapshot/serializer.cc:145: if (how_to_code == kPlain && where_to_point ==
kStartOfObject) {
[Feel free to ignore, since this is not even about your CL... :) ]

I was trying to parse the logic here, and stumbled a bit over the control flow.
For methods like these (SerializeHotObject + SerializeBackReference below),
where we effectively have an 'does this even apply?' condition plus the core
work of the method, I find it somewhat easier to read to handle all of the
conditions as early-out conditions at the top.

That's totally a matter of taste, though, and as said, it's not even about your
change... :-/

E.g. SerializeHotObject:
if (! ...how_to_code + where_to_point...)
  return false;
int index = ...
if (index == ...)
  return false;

[work being done]
return true;

[Yang, 2016-06-24 05:27:33]

The CQ bit was checked by yangguo@chromium.org

[Yang, 2016-06-24 05:27:33]

The patchset sent to the CQ was uploaded after l-g-t-m from
vogelheim@chromium.org
Link to the patchset: https://codereview.chromium.org/2090563002/#ps20001
(title: "address comments")

[commit-bot: I haz the power, 2016-06-24 05:27:37]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2090563002/20001

[Yang, 2016-06-24 06:00:58]

https://codereview.chromium.org/2090563002/diff/1/src/snapshot/serializer.cc
File src/snapshot/serializer.cc (right):

https://codereview.chromium.org/2090563002/diff/1/src/snapshot/serializer.cc#...
src/snapshot/serializer.cc:145: if (how_to_code == kPlain && where_to_point ==
kStartOfObject) {
On 2016/06/23 13:39:14, vogelheim wrote:
> [Feel free to ignore, since this is not even about your CL... :) ]
> 
> I was trying to parse the logic here, and stumbled a bit over the control
flow.
> For methods like these (SerializeHotObject + SerializeBackReference below),
> where we effectively have an 'does this even apply?' condition plus the core
> work of the method, I find it somewhat easier to read to handle all of the
> conditions as early-out conditions at the top.
> 
> That's totally a matter of taste, though, and as said, it's not even about
your
> change... :-/
> 
> E.g. SerializeHotObject:
> if (! ...how_to_code + where_to_point...)
>   return false;
> int index = ...
> if (index == ...)
>   return false;
> 
> [work being done]
> return true;
> 

Done.

[commit-bot: I haz the power, 2016-06-24 06:55:49]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-06-24 06:57:40]

Description was changed from

==========
[serializer] encode recent long-encoded root list items as hot objects.

We have a lot of long-encoded root list items in type feedback vectors.
==========

to

==========
[serializer] encode recent long-encoded root list items as hot objects.

We have a lot of long-encoded root list items in type feedback vectors.

Committed: https://crrev.com/734898a7b2d7f2f03ec9bdfe9bd6484d3c04fa32
Cr-Commit-Position: refs/heads/master@{#37237}
==========

[commit-bot: I haz the power, 2016-06-24 06:57:41]

Patchset 2 (id:??) landed as
https://crrev.com/734898a7b2d7f2f03ec9bdfe9bd6484d3c04fa32
Cr-Commit-Position: refs/heads/master@{#37237}