Use glibc's sigaction unless we are building for nonsfi nacl.

Use glibc's sigaction unless we are building for nonsfi nacl.

BUG=532180
Committed: https://crrev.com/fd3ffac4f3421662e5e9700a7b7fa0d4330e1432
Cr-Commit-Position: refs/heads/master@{#350260}

[rickyz (no longer on Chrome), 2015-09-22 10:32:14]

rickyz@chromium.org changed reviewers:
+ mdempsky@chromium.org

[rickyz (no longer on Chrome), 2015-09-22 10:32:15]

For now, the unwinding issues may still exist on nonsfi nacl builds, but this
should at least solve the issue for the rest of chromium. Once I can find the
implementation of sigaction use in those builds, I can see whether it is
possible to steal the sa_restorer from there instead of implementing our own.

[mdempsky, 2015-09-22 10:52:29]

LGTM assuming it fixes krasin@'s failing test scenario.

ACK that the OS_NACL_NONSFI case might still need work. Maybe add a TODO comment
or something to that effect?

[rickyz (no longer on Chrome), 2015-09-22 21:29:27]

The CQ bit was checked by rickyz@chromium.org

[rickyz (no longer on Chrome), 2015-09-22 21:29:28]

The patchset sent to the CQ was uploaded after l-g-t-m from
mdempsky@chromium.org
Link to the patchset: https://codereview.chromium.org/1358503005/#ps20001
(title: "Add a TODO for nonsfi nacl unwinding")

[commit-bot: I haz the power, 2015-09-22 21:29:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1358503005/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1358503005/20001

[commit-bot: I haz the power, 2015-09-22 23:23:35]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-09-22 23:24:25]

Patchset 2 (id:??) landed as
https://crrev.com/fd3ffac4f3421662e5e9700a7b7fa0d4330e1432
Cr-Commit-Position: refs/heads/master@{#350260}