Unify HTTP/2 and SPDY stream schedulers under common WriteScheduler interface.

Unify HTTP/2 and SPDY stream schedulers under common WriteScheduler interface.

This is to make it possible for SpdyDispatcher to maintain a single set of
codepaths for dealing with stream scheduling, with differences in scheduling
schemes hidden behind the WriteScheduler interface.  The actual wiring of
WriteScheduler into SpdyDispatcher will be in a follow-on CL.

This CL lands server change 115262424 by mpw.

BUG=488484
Committed: https://crrev.com/28130cb63807bc4bc81a176ada9d22ab3f5daa9e
Cr-Commit-Position: refs/heads/master@{#393758}

[Bence, 2016-04-21 18:56:36]

[email protected] changed reviewers:
+ [email protected]

[Bence, 2016-04-21 18:56:37]

Ryan: PTAL.  Thank you.

[Ryan Hamilton, 2016-04-21 21:09:23]

lgtm

[Bence, 2016-04-21 23:27:55]

The CQ bit was checked by [email protected]

[commit-bot: I haz the power, 2016-04-21 23:28:22]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1900403004/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1900403004/20001

[commit-bot: I haz the power, 2016-04-22 00:10:08]

The CQ bit was unchecked by [email protected]

[commit-bot: I haz the power, 2016-04-22 00:10:09]

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

[Bence, 2016-04-22 14:49:36]

Ryan: What's the story with net_nacl_common_sources versus net_non_nacl_sources
in net/net.gypi?  I see a comment saying "Most files in net/quic are in
net_nacl_common_sources, except for the files that have dependency on SPDY
(net/spdy) or NSS [that are in net_non_nacl_sources].".  But that comment seems
to be out of sync.  For example, quic/quic_write_blocked_list.h is in
net_nacl_common_sources, even though it includes
net/spdy/priority_write_scheduler.h.  If I wanted to make the comment true, I
would probably need to move all quic files over to net_non_nacl_sources.

This CL has some build errors that I am able to reproduce locally when I build
the remoting/webapp target.  I can fix them by moving
spdy/write_scheduler.{h,cc} from net_non_nacl_sources to
net_nacl_common_sources, which is (1) probably wrong because there are currently
no spdy/ files in net_nacl_common_sources, (2) confusing because
spdy/priority_write_scheduler.h (through which quic/quic_write_blocked_list.h
indirectly includes spdy/write_scheduler.h) is still in net_non_nacl_sources,
and so is spdy/spdy_protocol.h that spdy/write_scheduler.h includes, yet it
compiles.

[Ryan Hamilton, 2016-04-22 17:59:59]

On 2016/04/22 14:49:36, Bence wrote:
> Ryan: What's the story with net_nacl_common_sources versus
net_non_nacl_sources
> in net/net.gypi?  I see a comment saying "Most files in net/quic are in
> net_nacl_common_sources, except for the files that have dependency on SPDY
> (net/spdy) or NSS [that are in net_non_nacl_sources].".  But that comment
seems
> to be out of sync.  For example, quic/quic_write_blocked_list.h is in
> net_nacl_common_sources, even though it includes
> net/spdy/priority_write_scheduler.h.  If I wanted to make the comment true, I
> would probably need to move all quic files over to net_non_nacl_sources.
> 
> This CL has some build errors that I am able to reproduce locally when I build
> the remoting/webapp target.  I can fix them by moving
> spdy/write_scheduler.{h,cc} from net_non_nacl_sources to
> net_nacl_common_sources, which is (1) probably wrong because there are
currently
> no spdy/ files in net_nacl_common_sources, (2) confusing because
> spdy/priority_write_scheduler.h (through which quic/quic_write_blocked_list.h
> indirectly includes spdy/write_scheduler.h) is still in net_non_nacl_sources,
> and so is spdy/spdy_protocol.h that spdy/write_scheduler.h includes, yet it
> compiles.

I don't know the answer to that, other than that NaCL builds some subset of the
net stack, but I don't really understand the details. perhaps git blame can find
someone with more clue :(

[Bence, 2016-04-25 13:56:22]

[email protected] changed reviewers:
+ [email protected]

[Bence, 2016-04-25 13:56:22]

Sergey, please help with the NaCL issue described in comment #8.  git blame
points to https://crrev.com/1265283003, which you authored.  Thank you.

[Sergey Ulanov, 2016-05-12 19:01:31]

On 2016/04/25 13:56:22, Bence OOO until May 23 wrote:
> Sergey, please help with the NaCL issue described in comment #8.  git blame
> points to https://crrev.com/1265283003, which you authored.  Thank you.

We no longer need to compile QUIC for NaCl, so all quic files can be moved
net_non_nacl_sources. I'll send out a CL for review today to fix this.

[Bence, 2016-05-12 20:19:40]

On 2016/05/12 19:01:31, Sergey Ulanov wrote:
> On 2016/04/25 13:56:22, Bence OOO until May 23 wrote:
> > Sergey, please help with the NaCL issue described in comment #8.  git blame
> > points to https://crrev.com/1265283003, which you authored.  Thank you.
> 
> We no longer need to compile QUIC for NaCl, so all quic files can be moved
> net_non_nacl_sources. I'll send out a CL for review today to fix this.

Awesome.  Please cc me on that CL, and I'll rebase this one once that lands (and
once I'm back from leave).  Thank you.

[Bence, 2016-05-13 15:58:21]

The CQ bit was checked by [email protected]

[Bence, 2016-05-13 15:58:21]

The patchset sent to the CQ was uploaded after l-g-t-m from [email protected]
Link to the patchset: https://codereview.chromium.org/1900403004/#ps60001
(title: "Rebase.")

[commit-bot: I haz the power, 2016-05-13 15:58:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1900403004/60001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1900403004/60001

[commit-bot: I haz the power, 2016-05-13 16:21:15]

The CQ bit was unchecked by [email protected]

[commit-bot: I haz the power, 2016-05-13 16:21:16]

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

[Bence, 2016-05-14 21:29:53]

The CQ bit was checked by [email protected]

[Bence, 2016-05-14 21:29:54]

The patchset sent to the CQ was uploaded after l-g-t-m from [email protected]
Link to the patchset: https://codereview.chromium.org/1900403004/#ps80001
(title: "Rebase on https://crrev.com/1976803002.")

[commit-bot: I haz the power, 2016-05-14 21:30:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1900403004/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1900403004/80001

[commit-bot: I haz the power, 2016-05-15 01:24:01]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2016-05-15 01:25:16]

Description was changed from

==========
Unify HTTP/2 and SPDY stream schedulers under common WriteScheduler interface.

This is to make it possible for SpdyDispatcher to maintain a single set of
codepaths for dealing with stream scheduling, with differences in scheduling
schemes hidden behind the WriteScheduler interface.  The actual wiring of
WriteScheduler into SpdyDispatcher will be in a follow-on CL.

This CL lands server change 115262424 by mpw.

BUG=488484
==========

to

==========
Unify HTTP/2 and SPDY stream schedulers under common WriteScheduler interface.

This is to make it possible for SpdyDispatcher to maintain a single set of
codepaths for dealing with stream scheduling, with differences in scheduling
schemes hidden behind the WriteScheduler interface.  The actual wiring of
WriteScheduler into SpdyDispatcher will be in a follow-on CL.

This CL lands server change 115262424 by mpw.

BUG=488484

Committed: https://crrev.com/28130cb63807bc4bc81a176ada9d22ab3f5daa9e
Cr-Commit-Position: refs/heads/master@{#393758}
==========

[commit-bot: I haz the power, 2016-05-15 01:25:17]

Patchset 5 (id:??) landed as
https://crrev.com/28130cb63807bc4bc81a176ada9d22ab3f5daa9e
Cr-Commit-Position: refs/heads/master@{#393758}