Crash from "recursive" method handle chain

[Christian Thalinger]

On Jul 27, 2011, at 12:41 PM, Charles Oliver Nutter wrote:

> On Wed, Jul 27, 2011 at 3:21 AM, Rémi Forax <forax at univ-mlv.fr> wrote:
>> I've just read your slides [1].
>> I believe that the test before returning the result of add (slide 201)
>> is a safepoint check.
> 
> Oh really? What is it doing with eax, though, since that has the
> result of the add operation?

Right, it's a poll page read.  test on x86 only sets flags according to the test.  The Instruction Set Reference from Intel says:

"Description

Computes the bit-wise logical AND of first operand (source 1 operand) and the second operand (source 2 operand) and sets the SF, ZF, and PF status flags according to the result. The result is then discarded."

-- Christian

> 
> - Charlie
> _______________________________________________
> mlvm-dev mailing list
> mlvm-dev at openjdk.java.net
> http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev