harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geir Magnusson Jr." <g...@pobox.com>
Subject Re: [drlvm] stress.Mix problems -- should it prevent committing patches?
Date Tue, 09 Jan 2007 12:48:50 GMT

On Jan 9, 2007, at 1:36 AM, Weldon Washburn wrote:

> On 1/8/07, Geir Magnusson Jr. <geir@pobox.com> wrote:
>>
>>
>> On Jan 8, 2007, at 7:55 PM, Weldon Washburn wrote:
>>
>> > Naveen,
>> >
>> > This is a big help.  It really looks like stress.Mix/MegaSpawn has
>> > uncovered
>> > a new bug.  I looked at the excludes svn log and indeed stress.Mix
>> > was taken
>> > off the excludes list on December 10.
>> >
>> > Unless someone objects in the next 24 hours, I will put stress.Mix
>> > back on
>> > the excludes list.
>>
>> And then what?
>
>
> stress.Mix reverts back to whatever status it was in on December  
> 9.  No need
> to rollback because there was never a point in time when stress.Mix  
> passed
> on all known platforms.  In other words, its a new bug not a rollback.
>
> MegaSpawn.java was attached to harmony-2803.  A simple quick to run
> reproducible test for this new bug has already been captured.  This  
> should
> help a bunch when working on the patch.
>
> I will work on the patch for MegaSpawn if the list thinks its top  
> priority.

I think he fact that we have totally broken threading should be top  
priority :)
>

> Meanwhile, there are some GCV5 related patches that should be  
> committed
> soon.  That is, once svn head is working again.

I assume that these are orthogonal to anything else, so I have no  
problem with that.

geir


>
>
> geir
>>
>>


Mime
View raw message