[serializer] prepare attached references for general use.

[serializer] prepare attached references for general use.

Currently attached references are only used for global proxy, source
string and code stubs. Mid-term future we want to use attached
references for arbitrary objects (in fixed order) provided from outside.

This change renames BackReference to SerializerReference to include both
back references and attached references.

R=mtrofin@chromium.org, vogelheim@chromium.org

Committed: https://crrev.com/735fa0c478c5bba3c33d0509dd3263c745a50180
Cr-Commit-Position: refs/heads/master@{#36318}

[Mircea Trofin, 2016-05-18 10:47:42]

lgtm

[Yang, 2016-05-18 10:58:53]

The CQ bit was checked by yangguo@chromium.org

[commit-bot: I haz the power, 2016-05-18 10:58:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1992723002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1992723002/1

[vogelheim, 2016-05-18 11:18:48]

lgtm

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h
File src/address-map.h (right):

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode63
src/address-map.h:63: explicit SerializerReference(uint32_t bitfield) :
bitfield_(bitfield) {}
Why is this a public constructor?

(It assumes the caller knows the bitfield encodings, which this class does not
make public.)
(I think this is used from the friend class ...Map. As that's a friend, this
could be pr[ivate|otected].)

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode138
src/address-map.h:138: // value, or a attached reference index.
super nitpick: "a attached" -> "an attached"

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode164
src/address-map.h:164: STATIC_ASSERT(SpaceBits::kNext == 32);
Maybe assert for ValueIndexBits::kNext == ChunkIndexBits::kNext ?

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode203
src/address-map.h:203: HashMap* map_;
Why not a regular (non-pointer) member?

(From what I'm seeing, this is always [co|de]nstructed in the [co|de]nstructor,
so it could as well be a regular member.)

[Yang, 2016-05-18 11:28:50]

The CQ bit was unchecked by yangguo@chromium.org

[commit-bot: I haz the power, 2016-05-18 11:29:02]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-05-18 11:30:42]

Description was changed from

==========
[serializer] prepare attached references for general use.

Currently attached references are only used for global proxy, source
string and code stubs. Mid-term future we want to use attached
references for arbitrary objects (in fixed order) provided from outside.

This change renames BackReference to SerializerReference to include both
back references and attached references.

R=mtrofin@chromium.org, vogelheim@chromium.org
==========

to

==========
[serializer] prepare attached references for general use.

Currently attached references are only used for global proxy, source
string and code stubs. Mid-term future we want to use attached
references for arbitrary objects (in fixed order) provided from outside.

This change renames BackReference to SerializerReference to include both
back references and attached references.

R=mtrofin@chromium.org, vogelheim@chromium.org

Committed: https://crrev.com/735fa0c478c5bba3c33d0509dd3263c745a50180
Cr-Commit-Position: refs/heads/master@{#36318}
==========

[commit-bot: I haz the power, 2016-05-18 11:30:42]

Patchset 1 (id:??) landed as
https://crrev.com/735fa0c478c5bba3c33d0509dd3263c745a50180
Cr-Commit-Position: refs/heads/master@{#36318}

[Yang, 2016-05-18 11:58:42]

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h
File src/address-map.h (right):

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode63
src/address-map.h:63: explicit SerializerReference(uint32_t bitfield) :
bitfield_(bitfield) {}
On 2016/05/18 11:18:47, vogelheim wrote:
> Why is this a public constructor?
> 
> (It assumes the caller knows the bitfield encodings, which this class does not
> make public.)
> (I think this is used from the friend class ...Map. As that's a friend, this
> could be pr[ivate|otected].)

The deserializer reconstructs references from serialized bitfields. I added an
explicit SerializerReference::FromBitfield.

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode138
src/address-map.h:138: // value, or a attached reference index.
On 2016/05/18 11:18:47, vogelheim wrote:
> super nitpick: "a attached" -> "an attached"

Done.

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode164
src/address-map.h:164: STATIC_ASSERT(SpaceBits::kNext == 32);
On 2016/05/18 11:18:47, vogelheim wrote:
> Maybe assert for ValueIndexBits::kNext == ChunkIndexBits::kNext ?

Done.

https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode203
src/address-map.h:203: HashMap* map_;
On 2016/05/18 11:18:47, vogelheim wrote:
> Why not a regular (non-pointer) member?
> 
> (From what I'm seeing, this is always [co|de]nstructed in the
[co|de]nstructor,
> so it could as well be a regular member.)

Good point. Done.

[Yang, 2016-05-18 11:59:28]

On 2016/05/18 11:58:42, Yang wrote:
> https://codereview.chromium.org/1992723002/diff/1/src/address-map.h
> File src/address-map.h (right):
> 
> https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode63
> src/address-map.h:63: explicit SerializerReference(uint32_t bitfield) :
> bitfield_(bitfield) {}
> On 2016/05/18 11:18:47, vogelheim wrote:
> > Why is this a public constructor?
> > 
> > (It assumes the caller knows the bitfield encodings, which this class does
not
> > make public.)
> > (I think this is used from the friend class ...Map. As that's a friend, this
> > could be pr[ivate|otected].)
> 
> The deserializer reconstructs references from serialized bitfields. I added an
> explicit SerializerReference::FromBitfield.
> 
> https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode138
> src/address-map.h:138: // value, or a attached reference index.
> On 2016/05/18 11:18:47, vogelheim wrote:
> > super nitpick: "a attached" -> "an attached"
> 
> Done.
> 
> https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode164
> src/address-map.h:164: STATIC_ASSERT(SpaceBits::kNext == 32);
> On 2016/05/18 11:18:47, vogelheim wrote:
> > Maybe assert for ValueIndexBits::kNext == ChunkIndexBits::kNext ?
> 
> Done.
> 
> https://codereview.chromium.org/1992723002/diff/1/src/address-map.h#newcode203
> src/address-map.h:203: HashMap* map_;
> On 2016/05/18 11:18:47, vogelheim wrote:
> > Why not a regular (non-pointer) member?
> > 
> > (From what I'm seeing, this is always [co|de]nstructed in the
> [co|de]nstructor,
> > so it could as well be a regular member.)
> 
> Good point. Done.

Crap. It landed the first patch set in the meantime :/