Re: Daily Xen Builds

[Anthony Liguori <aliguori@us.ibm.com>]

David F Barrera wrote:

>Test changeset:
>
>changeset:   7293:0f33cbec4e36
>tag:         tip
>user:        emellor@ewan
>date:        Mon Oct 10 13:06:14 2005 +0100
>summary:     This patch fixes an error in the xm create path when the
>
>Last known good changeset for all platforms: 7069:a172340ae3f3
>  
>
<snip>

>xm-test: This suite provides a framework for testing the Xen userspace
>tools.
>
>SUMMARY:
>            Platform | PASS | FAIL | XPASS | XFAIL |
>---------------------+------+------+-------+-------+
>              FC3pae |   64 |   30 |     0 |     1 |
> hs20.1.sles9-x86_64 |   62 |   32 |     0 |     1 |
> hs20.2.sles9-x86_64 |   62 |   32 |     0 |     1 |
>     hs20.fc4_x86_64 |   63 |   31 |     0 |     1 |
>     x235sles9nonpae |   64 |   28 |     0 |     1 |
>          x305rh4pae |   63 |   30 |     0 |     1 |
>          x335fc4pae |   53 |   39 |     0 |     1 |
>
>  
>
This is a lot of failures.

Is it perhaps wise, nearing 3.0, that we adopt a policy of running the 
various test suites on code before it gets pushed to the public tree?  
This seems like requiring passing on xm-test a sane thing to do for the 
tools related changes.  Is there anything that needs to be done to the 
test suites to make this less painful for the committers?

Just to clarify, I'm suggesting that all changes should be checked 
(external patches or Cambridge pushes) against the testsuites before 
committing.  Just seems like it would be a good way to ensure that we 
avoid regressing as we near 3.0 release.

I don't think anyone is particularly at fault for the recent 
regressions, I just think it's time we adopt a more rigorious commit policy.

Thoughts?

Regards,

Anthony Liguori

>-- Regards, 
>
>David F Barrera 
>Linux Technology Center 
>Systems and Technology Group, IBM 
>
>"The wisest men follow their own direction. " 
>				Euripides
>
>  
>