Re: [PATCH v3 01/17] Fix and test "checkpoint" in non-stop mode

[Pedro Alves <palves at redhat dot com>]

On 04/28/2015 10:50 PM, Doug Evans wrote:

> Another thought was that if this is just a one-off and you're manually
> splitting up non-stop from all-stop to maintain parallelizability of the tests
> then I might not mind the new file.

I don't have plans to do this in any other test, so for now at least,
this is really just a one-off.  It's special because "checkpoint"
only really works with single-threaded programs (because it uses
fork), while non-stop/all-stop mostly matters about threaded programs.

> 
> IOW, tests that want to handle both all-stop and non-stop in one
> "make check" can do so in one file. Any loss in parallelization
> is probably minimal (modulo really big tests).

Yes, and we already do that in some tests.

> 
> Btw, checkpoint.exp completes in 9 seconds on my system.
> I suspect doubling that and doing both all-stop and non-stop in
> the one file won't increase "make check-parallel" noticeably.

It makes a larger difference for me:

$ time make check RUNTESTFLAGS="checkpoint.exp"

real    0m24.716s
user    0m22.676s
sys     0m2.168s

$ time make check RUNTESTFLAGS="checkpoint.exp checkpoint-ns.exp"

real    0m49.241s
user    0m45.183s
sys     0m4.306s

$ time make check TESTS="gdb.base/checkpoint.exp gdb.base/checkpoint-ns.exp" -j

real    0m28.606s
user    0m49.684s
sys     0m5.590s

Thanks,
Pedro Alves