Ensure _NT_SYMBOL_PATH is set for bot runs in cdb test

Ensure _NT_SYMBOL_PATH is set for bot runs in cdb test

Getting closer... Some tests passed on the last run, but the ones that
rely on having ntdll symbols fail on the bot. With `_NT_SYMBOL_PATH`
set, cdb will be able to download the PDBs so will be able to dump
data for `ntdll!_PEB`, etc.

R=mark@chromium.org
BUG=crashpad:46
Committed: https://chromium.googlesource.com/crashpad/crashpad/+/c3f4e2d8eb35b30494c478cd370f77916aa8e187

[scottmg, 2015-10-09 23:08:52]

Patchset #1 (id:1) has been deleted

[Mark Mentovai, 2015-10-09 23:24:34]

LGTM

We don’t have a more local cache of this?

[Mark Mentovai, 2015-10-09 23:25:11]

I guess for something checked into the tree, we wouldn’t want to use a local
thing anyway. The bots can set the environment variable if they have access to
something closer to them.

[scottmg, 2015-10-09 23:28:12]

On 2015/10/09 23:25:11, Mark Mentovai wrote:
> I guess for something checked into the tree, we wouldn’t want to use a local
> thing anyway. The bots can set the environment variable if they have access to
> something closer to them.

As far as I know, that's the recommended location. I'm not sure, we might not be
legally allowed to mirror them, at least in an externally-accessible location.

But yeah, it doesn't override, so the bots could set it to a local something if
we do have one.

[scottmg, 2015-10-09 23:28:24]

Committed patchset #2 (id:40001) manually as
c3f4e2d8eb35b30494c478cd370f77916aa8e187 (presubmit successful).

[brucedawson, 2015-10-09 23:30:54]

brucedawson@chromium.org changed reviewers:
+ brucedawson@chromium.org

[brucedawson, 2015-10-09 23:30:55]

https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
File snapshot/win/end_to_end_test.py (right):

https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
snapshot/win/end_to_end_test.py:202: 'SRV*' + symbol_dir +
'*https://msdl.microsoft.com/download/symbols')
Doesn't first-time use of Microsoft's symbol server bring up a EULA?

[scottmg, 2015-10-09 23:31:35]

On 2015/10/09 23:30:55, brucedawson wrote:
>
https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
> File snapshot/win/end_to_end_test.py (right):
> 
>
https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
> snapshot/win/end_to_end_test.py:202: 'SRV*' + symbol_dir +
> '*https://msdl.microsoft.com/download/symbols')
> Doesn't first-time use of Microsoft's symbol server bring up a EULA?

Hmm, maybe. I'll tell you in a few minutes when the bots cycle. Again. :/

[scottmg, 2015-10-09 23:33:59]

On 2015/10/09 23:31:35, scottmg wrote:
> On 2015/10/09 23:30:55, brucedawson wrote:
> >
>
https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
> > File snapshot/win/end_to_end_test.py (right):
> > 
> >
>
https://codereview.chromium.org/1402643002/diff/40001/snapshot/win/end_to_end...
> > snapshot/win/end_to_end_test.py:202: 'SRV*' + symbol_dir +
> > '*https://msdl.microsoft.com/download/symbols')
> > Doesn't first-time use of Microsoft's symbol server bring up a EULA?
> 
> Hmm, maybe. I'll tell you in a few minutes when the bots cycle. Again. :/

It looks like not. Maybe only in windbg, not
cdb?https://build.chromium.org/p/client.crashpad/builders/crashpad_win_x86_wow64_rel/builds/42/steps/run%20tests/logs/stdio
(sadly there's another failure)