Re: [drlvm] Ignoring Sun command line options?

["Geir Magnusson Jr." <ge...@pobox.com>]

On Feb 15, 2007, at 10:20 AM, Gregory Shimansky wrote:

> Alexey Varlamov wrote:
>> In my understanding, we face 2 issues here:
>> 1) Support "-XX:" format for internal options; this is  
>> straightforward
>> and I'll fix this shortly.
>> This way such options will not be refused but silently ignored.
>
> Ignoring them silently I think it not the right thing to do. Why  
> not add a warning like JRockit does, so that application developers  
> would pay more attention to the options that they pass to the VM?

Agreed.

geir

>
>> 2) Validate all specified internal options. More generally, VM
>> components should parse/validate related cmd-line args and report
>> back, instead of current hardcoded parsing in vmcore. This is a good
>> designing exercise, need to consider possible circular dependencies
>> between components and such. Any volunteers?
>> 2007/2/15, Zakharov, Vasily M <va...@intel.com>:
>>>
>>> Yes, I fully agree.
>>>
>>> And sure, we should track Sun options, it's a general question of
>>> compatibility.
>>>
>>>  Vasily
>>>
>>>
>>> -----Original Message-----
>>> From: Ilya Berezhniuk [mailto:ilya.berezhniuk@gmail.com]
>>> Sent: Thursday, February 15, 2007 6:16 PM
>>> To: dev@harmony.apache.org
>>> Subject: Re: [drlvm] Ignoring Sun command line options?
>>>
>>> To check Sun options we must hardcode some particular Sun  
>>> options, and
>>> update it when Sun options list is changed.
>>> But items (1) and (2) are very good :)
>>> I suggest the following sequence:
>>>
>>> 1. VM core takes the options it knows.
>>> 2. Other components take the options they know.
>>> 3. If remaining options contain only special options like -X then  
>>> print
>>> warnings, skip these options and continue.
>>> 4. If remaining options contain unknown general options then  
>>> print error
>>> and
>>> abort.
>>>
>>>
>>> 2007/2/15, Zakharov, Vasily M <va...@intel.com>:
>>> >
>>> >
>>> > Can we add checking for unsupported Sun options before throwing an
>>> > error?
>>> > I mean, the options processing logic could be like this:
>>> >
>>> > 1. VM core takes the options it knows.
>>> >
>>> > 2. Other components take the options they know.
>>> >
>>> > 3. The remaining options are checked for whether they are Sun  
>>> options.
>>> >
>>> > 4. If anything remains - print error and abort.
>>> >
>>> > Vasily
>>> >
>>> >
>>> > -----Original Message-----
>>> > From: Alexey Varlamov [mailto:alexey.v.varlamov@gmail.com]
>>> > Sent: Thursday, February 15, 2007 5:48 PM
>>> > To: dev@harmony.apache.org
>>> > Subject: Re: [drlvm] Ignoring Sun command line options?
>>> >
>>> > +1
>>> > Actually I thought we already agreed we should support this  
>>> format,
>>> > and I even promised to fix this. So thanks for the reminder :)
>>> >
>>> > Yet there might be a problem to detect if an option is not  
>>> supported,
>>> > as VM core knows nothing about options consumed by other  
>>> components
>>> > (as gc or jit) and not able to validate them. But this is a  
>>> separate
>>> > problem...
>>> >
>>> > --
>>> > Alexey
>>> >
>>> > 2007/2/15, Gregory Shimansky <gs...@gmail.com>:
>>> > > Zakharov, Vasily M wrote:
>>> > > > Hi, all,
>>> > > >
>>> > > > I'm now trying to run Netbeans on Harmony. On IBM VM it  
>>> starts,
>>> > > > though with many errors, but on DRL VM it exits immediately  
>>> with
>>> > > > the following diagnostic:
>>> > > >
>>> > > > Unknown option -XX:PermSize=32m
>>> > > > Use java -help to get help on command line options
>>> > > >
>>> > > > The problem is Netbeans startup file is .exe that makes up the
>>> > > > command line and starts java. And that .exe adds -XX:PermSize
>>> > > > and -XX:MaxPermSize options to command line, and DRL VM can't
>>> > > > handle them.
>>> > > >
>>> > > > JRockit also doesn't support these options, but it just  
>>> ignores
>>> > > > them and starts normally:
>>> > > >
>>> > > > [JRockit] WARNING:  '-XX:PermSize=32m' is not a valid VM  
>>> option.
>>> > > > Ignoring.
>>> > > > [JRockit] WARNING:  '-XX:MaxPermSize=160m' is not a valid VM
>>> option.
>>> > > > Ignoring.
>>> > > >
>>> > > > I think we should follow the JRockit/IBM way in DRL VM and  
>>> ignore
>>> > > > Sun options we do not support.
>>> > > >
>>> > > > My suggestion is to file this as an app-oriented RI- 
>>> compatibility
>>> > bug
>>> > > > in DRLVM and fix it as such.
>>> > > >
>>> > > > What do you think?
>>> > >
>>> > > +1
>>> > > Create a JIRA for this.
>>> > >
>>> > > --
>>> > > Gregory
>>> > >
>>> > >
>>> >
>>>
>
>
> -- 
> Gregory
>