Mailing-List: contact dev-help@harmony.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@harmony.apache.org
Received-SPF: pass (herse.apache.org: domain of t.p.ellison@gmail.com
 designates 66.249.92.171 as permitted sender)
DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
        s=beta; d=gmail.com;
        h=received:message-id:date:from:user-agent:mime-version:to:subject:references:in-reply-to:x-enigmail-version:content-type:content-transfer-encoding;
        b=h5Can7yJ3cUm6Dz8QM98UKE7gViPYs7f675Y78X/lQwACVbrakOIe8IcYzkW1bk31BlaZI0Drode2Eu6xaTsJcHjo4D6+IPJYyXCD8ZW7RO5WbF2it000hNZlh24ATaT7RxpTbM+0Bpx9YlUVMRSnxJAnJx5PKD6P631zY0UJk8=
Message-ID: <456AAAA0.70006@gmail.com>
Date: Mon, 27 Nov 2006 09:06:40 +0000
From: Tim Ellison <t.p.ellison@gmail.com>
User-Agent: Thunderbird 1.5.0.8 (Windows/20061025)
MIME-Version: 1.0
To: dev@harmony.apache.org
Subject: Re: [classlib][awt] Test failure
References: <4565DB0D.6050602@gmail.com>
	 <26c14c2a0611240727v26026917wf0bbae5e3455432a@mail.gmail.com>
	 <45671B19.8090104@gmail.com>
 <26c14c2a0611240855u28354578uc0343d5f509be2d1@mail.gmail.com>
 <45675C0C.30304@gmail.com> <4568B98B.9030108@pobox.com>
In-Reply-To: <4568B98B.9030108@pobox.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Geir Magnusson Jr. wrote:
> Tim Ellison wrote:
>> Oleg Khaschansky wrote:
>>> Do you suggest to introduce the error handling for all the wrappers or
>>> only for this one? If for all, do you think if it will affect the
>>> performance?
>>
>> I was thinking of something like this.  In the non-error case it would
>> add an additional NULL check as the function address is being cached,
>> and no additional overhead after the cache has been populated.
>>
>> The actual exception thrown is open to debate.\
> 
> :)
> 
> I don't understand the context - in the event of this fault, is it time
> to simply shut down?  Or is this survivable?  My understanding is that
> you just don't want to see a GPF.

I don't know at this level if it is survivable or not.  The native
contract is to call that function, if I can't then I'll just throw an
exception.  If it is handled as survivable in the Java code then fine,
otherwise it will cause a sane termination.

> Either way, how about a hint - like "missing libxmu" or similar?

I don't know the cause, so the exception will just describe the effect.

Regards,
Tim

-- 

Tim Ellison (t.p.ellison@gmail.com)
IBM Java technology centre, UK.