Remove 'mask' from SocketData, as it's not required.

Remove 'mask' from SocketData, as it's not required.

BUG=
R=sgjesse@google.com

Committed: https://code.google.com/p/dart/source/detail?r=32794

[Søren Gjesse, 2014-02-19 15:53:25]

lgtm

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
File runtime/bin/eventhandler_linux.cc (right):

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:240: SocketData* sd) {
Maybe I am wrong, but it seems like this is now echoing EPOLLIN, EPOLLOUT,
EPOLLHUP (and EPOLLRDHUP) if there is no error, and just send error if there is
error.

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:245: // Prioritize data events over close and
error events.
This comment seems wrong. We are not prioritizing data events over close and
error events. Actually error events are prioritized, and close events are
combined with data events.

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:246: if ((events & (EPOLLIN | EPOLLHUP |
EPOLLRDHUP)) != 0) {
Is this if statement actually needed?

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:264: if ((events & EPOLLOUT) != 0 && (events &
EPOLLERR) == 0) {
Can't this move up after line 250?

[Anders Johnsen, 2014-02-19 16:08:29]

PTAL

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
File runtime/bin/eventhandler_linux.cc (right):

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:240: SocketData* sd) {
On 2014/02/19 15:53:25, Søren Gjesse wrote:
> Maybe I am wrong, but it seems like this is now echoing EPOLLIN, EPOLLOUT,
> EPOLLHUP (and EPOLLRDHUP) if there is no error, and just send error if there
is
> error.

Done.

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:245: // Prioritize data events over close and
error events.
On 2014/02/19 15:53:25, Søren Gjesse wrote:
> This comment seems wrong. We are not prioritizing data events over close and
> error events. Actually error events are prioritized, and close events are
> combined with data events.

Done.

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:246: if ((events & (EPOLLIN | EPOLLHUP |
EPOLLRDHUP)) != 0) {
On 2014/02/19 15:53:25, Søren Gjesse wrote:
> Is this if statement actually needed?

Done.

https://codereview.chromium.org/167193004/diff/1/runtime/bin/eventhandler_lin...
runtime/bin/eventhandler_linux.cc:264: if ((events & EPOLLOUT) != 0 && (events &
EPOLLERR) == 0) {
On 2014/02/19 15:53:25, Søren Gjesse wrote:
> Can't this move up after line 250?

Done.

[Anders Johnsen, 2014-02-19 16:15:06]

Committed patchset #2 manually as r32794 (presubmit successful).