RFR - BUG7163117: SA Agent can't connect to process on Mac OSX

[Nils Loodin]

Hey David!

On May 8, 2012, at 10:18 , David Holmes wrote:
> Changes to x86_64 seem okay though it seems a little odd to accept amd64 but report the only valid values are x86/x86_64.
I guess you mean that I've changed the exception output like this:
-            throw new DebuggerException("Bsd only supported on x86/amd64");
+            throw new DebuggerException("Bsd only supported on x86/x86_64. Current arch: " + cpu);
I interpreted the existing output to refer to bitness, and not to the actual string of the cpu.
However, since we changed the string used to refer to the cpu on mac to be "x86_64" from "amd64", I thought it best to mention the actual string we expected in the checks however. Do you think this is a bad idea?

> 
> Also I'm a little confused ... some of this is not BSD/OSX specific and we should already support x86_64 for Linux/Solaris shouldn't we? (I guess not but why not? we don't use amd64 on any platform any more).
This confuse me too. I changed some places where it seemed to break for OSX. In general, I'd say that the code checks the cpu-string far to often and is in need of some serious refactorisation.

Regards, 
Nils Loodin.

> 
> David
> 
>> Regards
>> Nils Loodin
>> 
>> 
>> On May 4, 2012, at 13:01 , Nils Loodin wrote:
>> 
>>> Built it and connected to a running java process and ran some commands.
>>> I see other issues, but the connecting part actually works now.
>>> 
>>> This will probably open up more bugs.
>>> 
>>> /Nils Loodin
>>> 
>>> On May 4, 2012, at 12:05 , Staffan Larsen wrote:
>>> 
>>>> Looks good. I can sponsor the push.
>>>> 
>>>> What kind of testing have you done?
>>>> 
>>>> /Staffan
>>>> 
>>>> On 4 maj 2012, at 11:28, Nils Loodin wrote:
>>>> 
>>>>> looking for cpu-arch "amd64" only, and not "x86_64" as mac osx is using nowadays.
>>>>> 
>>>>> http://cr.openjdk.java.net/~nloodin/7163117/webrev.00/
>>>>> 
>>>>> 
>>>>> 
>>>>> Need reviews and a  sponsor!
>>>>> Regards,
>>>>> Nils Loodin
>>>> 
>>> 
>> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/serviceability-dev/attachments/20120508/2e500d0b/attachment.html