Use unique_ptrs to manage callback lifetimes in some QUIC code

Use unique_ptrs to manage callback lifetimes in some QUIC code

Follow-up to cl/132102408.

This CL eliminates the self-deleting callback pattern from a chunk of
QUIC code and replaces it with explicit unique_ptr ownership of all
callbacks.

Replace self-deleting callbacks with unique_ptr owned callbacks in some
QUIC code.  No functional change intended, not flag-protected.

Merge internal change: 132439857

[Ryan Hamilton, 2016-09-13 20:30:37]

rch@chromium.org changed reviewers:
+ zhongyi@chromium.org

[Zhongyi Shi, 2016-09-13 21:13:21]

The stateless_rejector.cc in the internal code is under quic/core/. I am
surprised to see it's not in chromium. Is that by intention?

https://codereview.chromium.org/2339433004/diff/1/net/quic/core/crypto/crypto...
File net/quic/core/crypto/crypto_server_test.cc (right):

https://codereview.chromium.org/2339433004/diff/1/net/quic/core/crypto/crypto...
net/quic/core/crypto/crypto_server_test.cc:1158: 
This code seems out of sync with internal. ShouldQueue() is missing here. 

The internal cl: 116790643 specifies do not merge, but then no follow-up.

[Ryan Hamilton, 2016-09-13 21:26:04]

https://codereview.chromium.org/2339433004/diff/1/net/quic/core/crypto/crypto...
File net/quic/core/crypto/crypto_server_test.cc (right):

https://codereview.chromium.org/2339433004/diff/1/net/quic/core/crypto/crypto...
net/quic/core/crypto/crypto_server_test.cc:1158: 
On 2016/09/13 21:13:21, Zhongyi Shi wrote:
> This code seems out of sync with internal. ShouldQueue() is missing here. 
> 
> The internal cl: 116790643 specifies do not merge, but then no follow-up.

Yeah, I noticed that too! No follow-up==bad!

Thankfully, ShouldQueue is unused internally and I'm removing it in my "merge
changes from chrome" CL. Sorry for not giving you the heads up.

[Zhongyi Shi, 2016-09-13 21:30:26]

lgtm