bug#37037: Guile 2.0.13+1-5.1 FTBFS on GNU/Hurd due to one failing test.

[Svante Signell]

Ping!

On Thu, 2019-08-15 at 14:29 +0200, Svante Signell wrote:
> Hello,
> 
> When building Debian version 2.0.13+1-5.1 one test fails making the
> whole build to fail. The problem is in the file test-
> suite/tests/statprof.test. According to that file it is known that
> ITIMER_PROF is not implemented:
> 
> ;; Throw `unresolved' upon ENOSYS.  This is used to skip tests on
> ;; platforms such as GNU/Hurd where `ITIMER_PROF' is is currently
> ;; unimplemented.
>       
> (define-syntax-rule (when-implemented body ...)
>   (catch 'system-error
>     (lambda ()
>       body ...)
>     (lambda args
>       (let ((errno (system-error-errno args)))
>         (false-if-exception (statprof-stop))
>         (if (= errno ENOSYS)
>             (throw 'unresolved)
>             (apply throw args))))))
> 
> (pass-if-equal "return values"
>     '(42 77)
>   (call-with-values
>       (lambda ()
>         (with-output-to-port (%make-void-port "w")
>           (lambda ()
>             (with-statprof
>                 (let loop ((i 10000))
>                   (if (zero? i)
>                       (values 42 77)
>                       (loop (1- i))))))))
>     list))
> 
> However, when that test is run the following is reported:
> 
> Running statprof.test
> ERROR: statprof.test: return values - arguments: ((system-error
> "setitimer" "~A" ("Function not implemented") (1073741902)))
> ERROR: statprof.test: statistical sample counts within expected range
> -
> arguments: ((misc-error #f "~A" ("Can't reset profiler while profiler
> is running.") #f))
> ERROR: statprof.test: accurate call counting - arguments: ((misc-
> error
> #f "~A" ("Can't reset profiler while profiler is running.") #f))
> 
> making that test (and package build) to fail.
> 
> An UNRESOLVED result would make that test to enable a successful
> package build. Unfortunately I don't know anything about scheme code
> so
> I cannot propose a patch to the Debian package maintainer. Maybe you
> can help me?
> 
> Please Cc: since I'm not subscribed.
> 
> Thanks!
> 
>