ok: handle concurrent crashes better

lockf() is a simple way to prevent interlaced stack traces when there
are concurrent crashes.  After a crashing process rethrows its signal
and dies for real, the OS unlocks the file for other processes.

I tested this by making SkCanvas::drawRRect() crash on Linux:
20-odd GM crashes with interlaced stack traces before, none after.

Change-Id: I99930756b8c85c552eef7c3a77778e4c00d34c42
Reviewed-on: https://skia-review.googlesource.com/10177
Reviewed-by: Herb Derby <herb@google.com>
Commit-Queue: Mike Klein <mtklein@chromium.org>

1 files changed, 3 insertions, 0 deletions

diff --git a/tools/ok.cpp b/tools/ok.cpp
index e4bf739e7d..b6d25b840c 100644
--- a/tools/ok.cpp
+++ b/tools/ok.cpp
@@ -38,6 +38,9 @@ static thread_local const char* tls_name = "";
 
         for (int sig : std::vector<int>{ SIGABRT, SIGBUS, SIGFPE, SIGILL, SIGSEGV }) {
             original_handlers[sig] = signal(sig, [](int sig) {
+                // To prevent interlaced output, lock the stacktrace log file until we die.
+                lockf(crash_stacktrace_fd, F_LOCK, 0);
+
                 auto ez_write = [](const char* str) {
                     write(crash_stacktrace_fd, str, strlen(str));
                 };