harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexei Fedotov" <alexei.fedo...@gmail.com>
Subject [drlvm][tests] patch testing Was: alert
Date Fri, 09 Nov 2007 12:37:39 GMT
Hello, DRLVM gurus,

I'm looking for help from people who have a whole picture of DRLVM
development practices. I've noticed that these practices were either
changed since my last VM patch or I lost their correct understanding.

Honestly, I never followed the practices blindly and neglected
cross-platfrom testing for verfier patches since the verifer was
perfectly independent from the platform. The thread about M4 was
mentioning these x86-64 patforms as our target, and Pavel Rebriy
started forwarding me with warning messages in my code on that
platforms. Finally, I realized that there is some ongoing development
for x86-64 platforms and tried to get a computer to run 64-bit testing
even for verifier patches.

>From the other side I never neglected running Anton Luht's acceptance
test for automated Eclipse Hello World application mostly because it
is very simple and convinient to execute ./build test2 command at VM
workspace and get a good class verification practice. Moreover, this
target included a good amount of regression tests to debug subroutine
verification.

Thanks to new CruiseControl alert integrated by Stepan, I have noticed
that currently the test developed by Anton Luht crashes on Linux
x86-64 with a strange exit code [1]. I wonder if there exist any DRLVM
developers who are running acceptance tests for their patches on Linux
x86-64 platform, and if yes, do you use test2 target?

Thank you for answering!

[1] http://issues.apache.org/jira/browse/HARMONY-5102

On 11/9/07, nomail@intel.com <nomail@intel.com> wrote:
> Subject: [build-test alert] FAILED Linux x86_64: ehwa-api debug
>
> BUILD FAILED:
[...]


-- 
With best regards,
Alexei,
ESSD, Intel

Mime
View raw message