run-tests.py: warn when no tests were run

run-tests.py: warn when no tests were run

The previous approach to this (in 1b1de2d22) did not work well with
bots that intentionally run no tests. But I think this warning message
is still useful for developers.

Committed: https://crrev.com/7a823598f8242cdbe93199f25224cac1ff835f2a
Cr-Commit-Position: refs/heads/master@{#30116}

[adamk, 2015-08-10 20:02:12]

adamk@chromium.org changed reviewers:
+ jkummerow@chromium.org, machenbach@chromium.org

[Jakob Kummerow, 2015-08-11 12:08:38]

Frankly, I don't quite see why we need this, as the existing output conveys the
same information:

[00:00|%   0|+   0|-   0]: Done

But if you like that more explicit warning, fine, LGTM.

[caitp (gmail), 2015-08-11 14:36:09]

On 2015/08/11 12:08:38, Jakob wrote:
> Frankly, I don't quite see why we need this, as the existing output conveys
the
> same information:
> 
> [00:00|%   0|+   0|-   0]: Done
> 
> But if you like that more explicit warning, fine, LGTM.

This CL doesn't seem to have been responsible, but something landed in the past
24 hours seems to be breaking commands in the form `tools/run-tests.py --arch
<arch> --mode <mode> <specific test job>`. I'm getting the following output:

```
$ tools/run-tests.py --mode release --arch x64 cctest/test-parsing
>>> running presubmit tests
Running C++ lint check...
No changes in files detected. Skipping cpplint check.
Running copyright header, trailing whitespaces and two empty lines between
declarations check...
Total violating files: 0
Runtime/Natives name clashes: checked 773/1042 functions, all good.
>>> Running tests for x64.release
[00:00|%   0|+   0|-   0]: Done 
```

Same thing with `make x64.release.check TESTFLAGS=--noi18n
TESTJOBS=cctest/test-parsing`

I'm having trouble finding what would have broken this, and rebasing things that
affected the test runner in the past 24 hours hasn't yielded any fixes yet.

[caitp (gmail), 2015-08-11 14:54:20]

On 2015/08/11 14:36:09, caitp wrote:
> On 2015/08/11 12:08:38, Jakob wrote:
> > Frankly, I don't quite see why we need this, as the existing output conveys
> the
> > same information:
> > 
> > [00:00|%   0|+   0|-   0]: Done
> > 
> > But if you like that more explicit warning, fine, LGTM.
> 
> This CL doesn't seem to have been responsible, but something landed in the
past
> 24 hours seems to be breaking commands in the form `tools/run-tests.py --arch
> <arch> --mode <mode> <specific test job>`. I'm getting the following output:
> 
> ```
> $ tools/run-tests.py --mode release --arch x64 cctest/test-parsing
> >>> running presubmit tests
> Running C++ lint check...
> No changes in files detected. Skipping cpplint check.
> Running copyright header, trailing whitespaces and two empty lines between
> declarations check...
> Total violating files: 0
> Runtime/Natives name clashes: checked 773/1042 functions, all good.
> >>> Running tests for x64.release
> [00:00|%   0|+   0|-   0]: Done 
> ```
> 
> Same thing with `make x64.release.check TESTFLAGS=--noi18n
> TESTJOBS=cctest/test-parsing`
> 
> I'm having trouble finding what would have broken this, and rebasing things
that
> affected the test runner in the past 24 hours hasn't yielded any fixes yet.

er, "and reverting things that affected the test runner", not "rebasing".

[adamk, 2015-08-11 15:01:49]

On 2015/08/11 12:08:38, Jakob wrote:
> Frankly, I don't quite see why we need this, as the existing output conveys
the
> same information:
> 
> [00:00|%   0|+   0|-   0]: Done
> 
> But if you like that more explicit warning, fine, LGTM.

It's hard to explain rationally, but I've previously found myself in the middle
of a debugging session and been confused by this output. I think it's too many
zeros. The difference between:

 [00:00|%   0|+   0|-   0]: Done

and

 [00:00|% 100|+   4|-   0]: Done

just isn't that big for my brain. But I could believe I'm the only one that has
this problem.

[Jakob Kummerow, 2015-08-11 15:25:03]

Adam, as I wrote, landing this is fine with me.

Caitlin, you want "cctest/test-parsing/*" (or "cctest" or "cctest/test-pars*" or
...) now. Matching has been made exact, as opposed to implicitly prefix-based.
The intention is that when you have tests "foo" and "foobar", there's should be
a way to run just "foo".

[caitp (gmail), 2015-08-11 15:31:02]

On 2015/08/11 15:25:03, Jakob wrote:
> Adam, as I wrote, landing this is fine with me.
> 
> Caitlin, you want "cctest/test-parsing/*" (or "cctest" or "cctest/test-pars*"
or
> ...) now. Matching has been made exact, as opposed to implicitly prefix-based.
> The intention is that when you have tests "foo" and "foobar", there's should
be
> a way to run just "foo".

Thanks for the info --- When did this change? It seems to be very recent, but
nothing stuck out as having caused it

[Jakob Kummerow, 2015-08-11 15:51:03]

On 2015/08/11 15:31:02, caitp wrote:
> Thanks for the info --- When did this change? 

https://codereview.chromium.org/1251363002

[adamk, 2015-08-11 16:36:25]

The CQ bit was checked by adamk@chromium.org

[commit-bot: I haz the power, 2015-08-11 16:36:41]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1281313004/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1281313004/1

[commit-bot: I haz the power, 2015-08-11 17:03:29]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-08-11 17:03:36]

Patchset 1 (id:??) landed as
https://crrev.com/7a823598f8242cdbe93199f25224cac1ff835f2a
Cr-Commit-Position: refs/heads/master@{#30116}