Revert of Replace StreamListenSocket with StreamSocket in HttpServer. (patchset #29 of https://codereview.chromium.org/296053012/)

Reason for revert:
looks like it caused net_unittests failures: http://build.chromium.org/p/chromium.win/builders/Vista%20Tests%20(2)/builds/44610/steps/net_unittests/logs/MultipleRequestsOnSameConnection

Original issue's description:
> Replace StreamListenSocket with StreamSocket in HttpServer.
>     
> 1) HttpServer gets ServerSocket instead of StreamListenSocket.
> 2) HttpConnection is just a container for socket, websocket, and pending read/write buffers.
> 3) HttpServer handles data buffering and asynchronous read/write.
> 4) HttpConnection has limit in data buffering, up to 1Mbytes by default.
> 5) For devtools, send buffer limit is 100Mbytes.
> 6) Unittests for buffer handling in HttpConnection.
>     
> BUG=371906
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=291447

NOTREECHECKS=true
NOTRY=true
TBR=rsleevi@chromium.org

Review URL: https://codereview.chromium.org/497223003

Cr-Commit-Position: refs/heads/master@{#291521}
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@291521 0039d316-1c4b-4281-b951-d872f2087c98


CrOS-Libchrome-Original-Commit: 0b2f33f4a88efbd203b0623324ad4114e3bb9d23
2 files changed
tree: c9e9f548da98e514e7de080542eec5cc6d13d71a
  1. base/
  2. build/
  3. components/
  4. dbus/
  5. device/
  6. ipc/
  7. mojo/
  8. testing/
  9. third_party/
  10. ui/