This patch by Lynn Boger increases the time to wait for signals to be
delivered in the TestAtomicStop test in the os/signal package. When
running gccgo tests on ppc64 or ppc64le, there are intermittent
failures in this test because the wait time is too small.
Bootstrapped and ran os/signal test on x86_64-pc-linux-gnu.  Committed
to mainline.

Ian
Index: gcc/go/gofrontend/MERGE
===================================================================
--- gcc/go/gofrontend/MERGE     (revision 267033)
+++ gcc/go/gofrontend/MERGE     (working copy)
@@ -1,4 +1,4 @@
-4db15776701a7cfc1abcef08f7a98c63343d17e6
+be6dbbdbe29f39d6d2ec0d5574ced398cb6e250a
 
 The first line of this file holds the git revision number of the last
 merge done from the gofrontend repository.
Index: libgo/go/os/signal/signal_test.go
===================================================================
--- libgo/go/os/signal/signal_test.go   (revision 267026)
+++ libgo/go/os/signal/signal_test.go   (working copy)
@@ -432,12 +432,12 @@ func atomicStopTestProgram() {
 
                // At this point we should either die from SIGINT or
                // get a notification on cs. If neither happens, we
-               // dropped the signal. Give it a second to deliver,
-               // which is far far longer than it should require.
+               // dropped the signal. It is given 2 seconds to
+               // deliver, as needed for gccgo on some loaded test systems.
 
                select {
                case <-cs:
-               case <-time.After(1 * time.Second):
+               case <-time.After(2 * time.Second):
                        if !printed {
                                fmt.Print("lost signal on tries:")
                                printed = true

Reply via email to