One of the relatively few remaining libgo failures on Solaris is
--- FAIL: TestG0StackOverflow (2.17s)
crash_test.go:754: === RUN TestG0StackOverflow
(exit status signal: Segmentation Fault)
FAIL
FAIL: runtime
Given that Solaris is a pthread target just like a bunch of others, it
seems that the test should skipped there as well. Done with the
following patch, tested on i386-pc-solaris2.11 and
sparc-sun-solaris2.11.
This leaves only (in runtime)
--- FAIL: TestSelectStackAdjust (0.16s)
chan_test.go:728: failed to trigger concurrent GC
(both sparc and x86)
and
--- FAIL: TestPanicInlined (0.00s)
[...]
crash_test.go:566: expecting stack trace to contain call to point.negate
(sparc only). I haven't yet investigated those.
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
2019-02-22 Rainer Orth <r...@cebitec.uni-bielefeld.de>
* go/runtime/crash_test.go (TestG0StackOverflow): Skip on solaris.
# HG changeset patch
# Parent b4d1b955d0574477e08883eb0143e36610f99f8f
Skip runtime.TestG0StackOverflow on Solaris
diff --git a/libgo/go/runtime/crash_test.go b/libgo/go/runtime/crash_test.go
--- a/libgo/go/runtime/crash_test.go
+++ b/libgo/go/runtime/crash_test.go
@@ -741,7 +741,7 @@ func TestG0StackOverflow(t *testing.T) {
testenv.MustHaveExec(t)
switch runtime.GOOS {
- case "darwin", "dragonfly", "freebsd", "linux", "netbsd", "openbsd", "android":
+ case "darwin", "dragonfly", "freebsd", "linux", "netbsd", "openbsd", "android", "solaris":
t.Skipf("g0 stack is wrong on pthread platforms (see golang.org/issue/26061)")
}
