Commit Graph

12 Commits

Author SHA1 Message Date
klzgrad
0934618957 h2: Notify delegate about read EOF
So the delegate can close the socket instead of keeping sending data.

Read EOF or h2 half-closed (remote) state was introduced in
https://codereview.chromium.org/129543002. But StreamSocket doesnt
really supports a half closed state, so upon a read EOF the only sane
action is to close the socket immediately even if in theory more send
is possible.
2020-04-14 19:21:25 +08:00
klzgrad
15e6f27e26 h2: Reduce warnings about RST on invalid streams
Per RFC 7540#6.4:

  However, after sending the RST_STREAM, the sending endpoint MUST be
  prepared to receive and process additional frames sent on the stream
  that might have been sent by the peer prior to the arrival of the
  RST_STREAM.
2020-04-14 19:21:25 +08:00
klzgrad
a0b8192172 socket: Fix a bug in TCPClientSocket
Needed by transparent proxying.
2020-04-14 19:21:25 +08:00
klzgrad
2700ff4844 socket: Force tunneling for all sockets
After the upstream large refactor, now only WebSocket sockets
have tunneling via HTTP/1 proxies. "Raw" sockets in the normal
socket pool don't have tunneling via HTTP/1 proxies, i.e.
CONNECT headers are not sent, instead the raw payload is sent
as-is to the HTTP/1 proxy, which makes it not work.

For the reference the official code does this:

- HTTP sockets via HTTP/1 proxies: normal pool, no tunneling.
- HTTPS sockets via HTTP/1 proxies: normal pool, no tunneling
  but does its own proxy encapsulation.
- WS sockets via HTTP/1 proxies: WS pool, tunneling.

We want the normal pool because the WS pool has some extra
restrictions but we also want tunneling to expose a client socket
with the proxy built in.

Therefore we can force tunneling for all sockets. This will always
send CONNECT headers first and thus break HTTP client sockets via
HTTP/1 proxies, but since we don't use this combination this is ok.
2020-04-14 19:21:25 +08:00
klzgrad
f30518b21e socket: Add RawConnect method 2020-04-14 19:21:25 +08:00
klzgrad
c0c0d7da47 socket: Allow higher limits for proxies
As an intermediary proxy we should not enforce stricter connection
limits in addition to what the user is already enforcing.
2020-04-14 19:21:25 +08:00
klzgrad
26ea8a6d64 build: Disable use_pangocario 2020-04-14 19:21:25 +08:00
klzgrad
a960f9ee76 build: Remove icu 2020-04-14 19:21:25 +08:00
klzgrad
af7205defe build: Force determinism in official build 2020-04-14 19:21:25 +08:00
klzgrad
ac00a8a279 build: Move exclude_unwind_tables back into declare_args
There is desire to adjust this flag manually.

BUG=762629
R=thakis@chromium.org

Change-Id: I3bd134c19270cd1f729b3ea078674e734493d4ab
2020-04-14 19:21:25 +08:00
klzgrad
7903249a77 build: Don't depend on dri in //content/gpu 2020-04-14 19:18:56 +08:00
klzgrad
518142cc31 Import chromium-81.0.4044.92 2020-04-14 19:18:56 +08:00