Hi, On Thu, Nov 26, 2020 at 5:15 AM Lucas Nussbaum <lu...@debian.org> wrote: > > The relevant part is near "FAIL:" > > > > === RUN TestIssue1374 > > proc_test.go:112: failed assertion at proc_test.go:4211: Call - > > could not restore registers: bad address > > --- FAIL: TestIssue1374 (0.62s) > > > > > > However as delve was just built by buildd 11hour ago, and on my > > desktop. So this seems a corner case triggered by your environment. > > Hence I downgrade the severity. > > > > I guess the difference between buildd, my desktop and your env is real > > hardware and VM. > > Since delve is a debug tool, like gdb, probably the failed tests are > > using some extreme methods which are restricted by VM. > > I tried to build this package in another environment (standard sbuild, > not a VM) and could reproduce a build failure (but I'm not sure it's the > same issue). Could you maybe post a build log and diff it with mine so > that we can track down what differs? > > I'm attaching a build log from today. >
It's a new issue. The error is different from the previous one === RUN TestStepIntoWrapperForEmbeddedPointer support.go:248: enabling recording for github.com/go-delve/delve/pkg/proc_test.TestStepIntoWrapperForEmbeddedPointer support.go:248: enabling recording for github.com/go-delve/delve/pkg/proc_test.TestStepIntoWrapperForEmbeddedPointer proc_test.go:488: Program did not continue to correct next location expected 19 was ifaceembcall.go:28 (0x49aa62) (testcase 3) --- FAIL: TestStepIntoWrapperForEmbeddedPointer (1.12s) === RUN TestStarlarkVariable starlark_test.go:160: for "v = eval(None, \"f1\").Variable; print(v.Value)" expected "3" got "3.0" --- FAIL: TestStarlarkVariable (0.39s) I have pushed a fix to salsa, should I just close this bug, or you prefer cloning a new one? https://salsa.debian.org/go-team/packages/delve/-/commit/5d77e7f46cb38bd2be5f9e0c12302541f0a02cb3 -- Shengjing Zhu