Compute the start time before registering the callback, rather than after, to avoid the possibility of a small race.

The following scenario illustrates the problem.

1. The callback is registers with a 2 second timeout at t=0ms.
2. We compute the start time after registering the callback. For the sake of argument, let's say it took 5ms to return from registering the callback and computing the current time. Start=5ms.
3. The callback fires after exactly 2 seconds, or t=2000ms.
4. We compute the difference between start and now. If it took less than 5ms to compute, then we end up with a difference that's less than 2000ms and the test fails. Let's say it took 3ms this time, then 2003ms-5ms=1998ms < 2000ms.

The actual values in the example above are arbitrary. All that matters is that it took longer to compute the start time than the end time. My theory is that this explains why this test is flaky when running under ASan in CI (which has unpredictable timing).

rdar://160956999

Full diff: https://github.com/llvm/llvm-project/pull/162197.diff

1 Files Affected:

• (modified) lldb/unittests/Host/MainLoopTest.cpp (+1-1)

diff --git a/lldb/unittests/Host/MainLoopTest.cpp b/lldb/unittests/Host/MainLoopTest.cpp
index 0bc291c26b9c9..ae16d02101819 100644
--- a/lldb/unittests/Host/MainLoopTest.cpp
+++ b/lldb/unittests/Host/MainLoopTest.cpp
@@ -424,9 +424,9 @@ TEST_F(MainLoopTest, ManyPendingCallbacks) {
 
 TEST_F(MainLoopTest, CallbackWithTimeout) {
   MainLoop loop;
+  auto start = std::chrono::steady_clock::now();
   loop.AddCallback([](MainLoopBase &loop) { loop.RequestTermination(); },
                    std::chrono::seconds(2));
-  auto start = std::chrono::steady_clock::now();
   ASSERT_THAT_ERROR(loop.Run().takeError(), llvm::Succeeded());
   EXPECT_GE(std::chrono::steady_clock::now() - start, std::chrono::seconds(2));
 }