init: use a no-op signal handler instead of SIG_IGN for SIGPIPE

We want to ignore SIGPIPE within init, but if we use SIG_IGN, that
would be inherited by child processes through exec(), which we do not
want to have happen.  We instead set up a real signal handler with a
no-op handler function, that will ignore SIGPIPE within init, but will
not be inherited across exec().

This fixes c29c2baa69 ("init: Add support for native service
registration with lmkd"), when SIG_IGN was introduced.
Note that we caught this issue before shipping a release with that
change, so the major motivation here is to not cause a behavior change
in init.

Bug: 151581751
Test: children of init that don't explicitly block SIGPIPE exit when
      sent SIGPIPE
Test: children of init that do explicitly block SIGPIPE do not exit
      when sent SIGPIPE
Test: init does not exit when sent SIGPIPE
Test: init exits when sent SIGABRT
Change-Id: Ieda8555fd03836bcd672a422fe673a8369ad9beb
This commit is contained in:
Tom Cherry 2020-03-16 10:17:05 -07:00
parent 01c8eccca3
commit fd470e87cc

View file

@ -713,8 +713,15 @@ int SecondStageMain(int argc, char** argv) {
InitKernelLogging(argv);
LOG(INFO) << "init second stage started!";
// Will handle EPIPE at the time of write by checking the errno
signal(SIGPIPE, SIG_IGN);
// Init should not crash because of a dependence on any other process, therefore we ignore
// SIGPIPE and handle EPIPE at the call site directly. Note that setting a signal to SIG_IGN
// is inherited across exec, but custom signal handlers are not. Since we do not want to
// ignore SIGPIPE for child processes, we set a no-op function for the signal handler instead.
{
struct sigaction action = {.sa_flags = SA_RESTART};
action.sa_handler = [](int) {};
sigaction(SIGPIPE, &action, nullptr);
}
// Set init and its forked children's oom_adj.
if (auto result =