Make sure the endpoints of the SandboxIPC mechanism cannot be misused.

Make sure the endpoints of the SandboxIPC mechanism cannot be misused.

BUG=322185
TEST=Browser starts, loads webpages.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=267183

[Jorge Lucangeli Obes, 2014-04-29 23:16:12]

Julien, Antoine: this closes the unused ends of the SandboxIPC mechanism.
Matthew: FYI.

I chose to do it here instead of in sandbox_ipc_linux.cc and
zygote_host_impl_linux.cc to keep it close to the actual socketpair call.

[jln (very slow on Chromium), 2014-04-29 23:23:20]

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
File content/browser/renderer_host/render_sandbox_host_linux.cc (right):

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:44:
shutdown(renderer_socket_, SHUT_RD);
Please, PCHECK(0 == ...);

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:49:
shutdown(browser_socket, SHUT_WR);
Same remark, please PCHECK().

[piman, 2014-04-29 23:39:19]

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
File content/browser/renderer_host/render_sandbox_host_linux.cc (right):

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:44:
shutdown(renderer_socket_, SHUT_RD);
On 2014/04/29 23:23:20, jln wrote:
> Please, PCHECK(0 == ...);

Actually, do we need HANDLE_EINTR? (and below)

[jln (very slow on Chromium), 2014-04-29 23:41:46]

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
File content/browser/renderer_host/render_sandbox_host_linux.cc (right):

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:44:
shutdown(renderer_socket_, SHUT_RD);
On 2014/04/29 23:39:19, piman wrote:
> On 2014/04/29 23:23:20, jln wrote:
> > Please, PCHECK(0 == ...);
> 
> Actually, do we need HANDLE_EINTR? (and below)

No, shutdown() doesn't require HANDLE_EINTR. It's also not considered a
cancellation point for threading.

[piman, 2014-04-29 23:50:33]

LGTM then.

[Jorge Lucangeli Obes, 2014-04-30 00:26:02]

PTAL. Thanks!

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
File content/browser/renderer_host/render_sandbox_host_linux.cc (right):

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:44:
shutdown(renderer_socket_, SHUT_RD);
On 2014/04/29 23:23:20, jln wrote:
> Please, PCHECK(0 == ...);

Done.

https://chromiumcodereview.appspot.com/253213002/diff/1/content/browser/rende...
content/browser/renderer_host/render_sandbox_host_linux.cc:49:
shutdown(browser_socket, SHUT_WR);
On 2014/04/29 23:23:20, jln wrote:
> Same remark, please PCHECK().

Done.

[jln (very slow on Chromium), 2014-04-30 01:35:54]

lgtm

[Jorge Lucangeli Obes, 2014-04-30 04:40:41]

The CQ bit was checked by jorgelo@chromium.org

[commit-bot: I haz the power, 2014-04-30 04:42:31]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jorgelo@chromium.org/253213002/20001

[commit-bot: I haz the power, 2014-04-30 11:29:21]

Change committed as 267183