On 04/21/2011 04:28 PM, Coleen Phillimore wrote:
Hi,
I'm looking at this crash
http://mail.openjdk.java.net/pipermail/mlvm-dev/2011-January/002268.html
but the test doesn't run anymore because the name of the classes have
changed from java/dyn/CallSite to java/lang/invoke.
Can you send me an u
Nice. Thanks! -- John
On Jan 4, 2011, at 5:30 PM, Rémi Forax wrote:
> On 01/05/2011 01:31 AM, John Rose wrote:
>> On Jan 2, 2011, at 6:42 AM, Rémi Forax wrote:
>>
>>> This means there is a bug in my code but also
>>> that there is a bug in the VM which should reject the bad class file
>>> inst
On 01/05/2011 01:31 AM, John Rose wrote:
On Jan 2, 2011, at 6:42 AM, Rémi Forax wrote:
This means there is a bug in my code but also
that there is a bug in the VM which should reject the bad class file
instead of crashing.
Try "java -Xverify:all". If the class file fails verification *but* th
On Jan 2, 2011, at 6:42 AM, Rémi Forax wrote:
> This means there is a bug in my code but also
> that there is a bug in the VM which should reject the bad class file
> instead of crashing.
Try "java -Xverify:all". If the class file fails verification *but* the
verifier is not used, the JVM is al
> While patching ASM in order to generate the new class file format,
> I generate a class file that crash the VM
> (see attachment).
Hi Remi,
Yes, VM definitely should not crash with invalid class files.
I filed a bug and depending on Oracle bug visibility policy, it may
appear at bugs.sun.com:
On 01/02/2011 03:42 PM, Rémi Forax wrote:
> Hi all,
> While patching ASM in order to generate the new class file format,
> I generate a class file that crash the VM
> (see attachment).
>
> This means there is a bug in my code but also
> that there is a bug in the VM which should reject the bad clas