inline...
On 12/6/19 6:12 PM, serguei.spit...@oracle.com wrote:
On 12/6/19 17:24, Daniel D. Daugherty wrote:
On 12/6/19 6:26 PM, serguei.spit...@oracle.com wrote:
On 12/6/19 13:52, Chris Plummer wrote:
On 12/6/19 1:18 PM, serguei.spit...@oracle.com wrote:
On 12/6/19 11:07, Chris Plummer wrote:
On 12/5/19 6:45 PM, David Holmes wrote:
Hi Serguei,
On 6/12/2019 11:31 am, serguei.spit...@oracle.com wrote:
Hi Chris and Alex,
(I've also included Dan, David and Dean to the mailing list)
We have to reach a consensus about this.
This is just part of a much broader issue with JVM TI that I
tried to have a discussion started based on Richard Reingruber's
proposals around Escape Analysis:
http://mail.openjdk.java.net/pipermail/serviceability-dev/2019-September/029285.html
Unfortunately that discussion did not get much traction.
Hmm. I have the emails that precede yours above, but not that
one. Not sure how what happened. Just read through it and it did
give me one thought.
Consider a model where the program is designed drive behavior of
the agent, triggering the agent to do certain things by having
the program do certain things. Normally an agent monitors the
application, but in this case the application is purposefully
controlling actions performed by the agent. If code is elided
from the program, then the agent no longer performs as expected.
It's a kind of backwards jvmti programming model, and you may ask
why would anyone do this. I'm not sure if there's a good reason
for it, but should it be expected to work given how the spec is
written?
My interpretation is that the current JVM TI PopFrame behavior
does not break this model.
The spec says: "any changes to the arguments, which occurred in
the called method, remain;"
As the code was eliminated by the compiler then no changes to this
argument occurred.
So, the PopFrame behavior follows the spec. So, I think, the
option #2 is not right. But it depends on our basic philosophy.
If the developer wants to control the agent then the program has
to be designed to do something meaningful that is not going to be
optimized out by the JIT compiler.
You misunderstood my point. What I'm saying is that someone might
do something like assign to a local with the specific intent of
having that trigger a jmvti event, with the specific intent of
having the agent perform some expected action as a result. Think of
it as being a trigger for the agent, not as the agent monitoring
the app. For example, you could right a program + agent, and
setting a specific local in the program triggers the agent to turn
on a light, and setting some other local turns it off. Absurd, but
possible, and maybe there are less absurd applications.
I think, I understood your point correctly.
Your point is that the code that can be eliminated (e.g. local++) is
not that meaningless as it seems to be.
My point is that there are still other more reliable ways to trigger
the agent.
So that relying on something that can be eliminated by JIT compilers
is not important to support.
You are making the assumption that the agent author understands what
Java code/variables *might* be eliminated by the JIT compiler. I don't
think that's a good assumption. I might have code that does a really
complicated thing in a local variable that is only useful to the
agent itself. The JIT will see that the local variable cannot escape
the function and is not used outside the function (as far as it can
see) so it will elide the local variable and the code that was used
to generated the local result in the variable.
If that local result happens to be some computation that the agent
needed to see to do its next operation...
Thank you for sharing your point.
I'm not insisting on my assumptions here, just not sure this is more
important than allowing optimizations.
Do you actually think this use case needs to be supported?
In general, to identify our philosophy about interaction between JIT
compiler
code elimination and JVM TI we need to make some assumptions.
Let's temporarily put JVM TI out of scope.
Are there any assumptions when JIT compilers eliminate some code?
Is it based on some vision what code is observable?
If it was decided some code can be eliminated then is it JVM TI only
that breaks such assumptions about observability?
If so, then such optimizations can be disabled at some level.
Then we end up debugging/profiling/monitoring, and finally, observing
a slightly different application.
Are we Okay with this? Do we need any compromises here?
Maybe we need more flags to control the JIT compiler behavior.
I would say that "in general" (not Java specific) there is an implicit
assumption that compiler optimization and "debugging" are diametrically
opposed to each other, and thus one cannot assume/expect that either can
transparently co-exist, you either optimize or you debug,
but there is a sliding scale between the two extremes, fully optimized
and no optimization (and hence fully debug-able).
The question is: where does "observe-ability" (as distinct from
debugging) lie on that continuum?
In most "ahead of time" language compilers, where all the code
generation occurs during the compilation phase, the developer can choose to
inform the compiler if their intent is to debug the resulting code (no
mutating optimizations, and full metadata retention), or to optimize for
"production" execution (maximal optimization, and no retention of debug
metadata).
The JVM moves some of this activity to r/t which is I think an
implementation detail, there is still a "contract" between the activity
of the "compilation" component and the "debug/observe-ability" component
of the runtime environment.
The debug/observe-ability component can only interact with the code that
the "compiler" generates (AOT & JIT etc).
In other language toolchains, the specification of the intent to "debug"
typically constrains the "compiler" from making mutating optimizations that
would result in an execution behavior that is not broadly equivalent to
that expressed at source level.
In short, I think we have two options; define the behavior of the
execution environment as "undefined", that is the compilers and runtime are
permitted to mutate the code generated from a form equivalent to that
expressed in source, or we add the ability to express the intent of the
code generated, such that when that intent is to debug it, that mutating
optimizations are suppressed by the compiler and runtime.
- Larry
Thanks,
Serguei
Dan
Thanks,
Serguei
Chris
We have 3 options:
Option #1:
The JIT optimization to delete a code which "looks useless"
has to be disabled if can_pop_frame capability is enabled.
Than this problem becomes a JIT compiler bug.
Option #2:
Consider to relax the JVMTI PopFrame spec by changing it to
something like:
"Note however, that the original argument values are not
preserved and can be changed by the called method;"
Than this problem becomes a JVM TI spec bug.
Option #3:
Consider it is Okay for compiler to eliminate useless code,
so the argument values can be reinitialized by the PopFrame.
Than this problem becomes just a test bug.
My preference is option #3.
The point is that if the arguments are not really used in
a method then restoring them to any values is a no-op.
It is really meaningless use case, so why should we care about it.
Is "restoring" the proper term here? I thought they were just
left on the stack and reused on the subsequent invoke.
Agreed. The term "restoring" is not accurate here.
In fact I figured the reason for the language in the spec in the
first place is to alleviate JVMTI from having to restore them to
their original values, which is probably not even possible.
Right.
Thanks for setting that out clearly.
I'd like to agree this is particular case is a test bug. If we
have a method:
int incr(int val) {
val++;
popFrameHere();
return val;
}
then the change to the argument is necessary and must be
preserved. In contrast:
void incr(int val) {
val++;
popFrameHere();
}
the change to the argument is meaningless and I would hope any
decent JIT would simply elide it.
So, this goes back to my example above where the program is
trying to elicit behavior from the agent. It's not meaningless in
that case, but that doesn't mean I think we need to support it.
Even with this model it is possible and better to do something
meaningful to control the agent.
This model is very rare use case.
It is hard to justify a need to support it. :)
But we must have a consistent approach to such things. What
would happen if a breakpoint were to be placed on the
instruction that uselessly modified the argument - would we
still see the modification or would it be elided?
Breakpoints force interpreted mode for the method, although I
suppose that's a hotspot implementation detail and not something
a VM would be required to do. A VM that allows breakpoints in
compiled methods has the potential to miss the breakpoint if code
is elided.
Also, what if you put a breakpoint in a method, the call to it is
elided. You would never hit the breakpoint. That could cause some
serious head scratching for a debugger user if they know the code
doing the method call is "executed".
If the method is not actually being called then missing
breakpoints there gives a clue what is going on.
Otherwise, it will cause cause some serious head scratching for a
debugger user.
In general, my preference would be to debug actual behavior.
It is not good we have no support breakpoints in compiled methods.
And how do C1 and C2 avoid this issue? Do they simply not
optimise away the useless assignment? Or do they actively
disable that optimization in this context?
We need, IMO, to establish the basic philosophy of how to manage
JVM TI / JIT interactions, so we know what things must remain
visible and which can be optimised away.
That said, changing the test allows us to defer having to reach
that consensus.
Agreed. I think it's ok to work around the test issue as long as
we keep this overall issue on the radar. Do we have a bug field
for that?
I thought, it is a little bit early to file a bug for it.
Also, probably, it can be an umbrella enhancement or task.
Thanks,
Serguei
thanks,
Chris
David
-----
Thanks,
Serguei
On 11/11/19 3:17 AM, serguei.spit...@oracle.com wrote:
Hi Alex,
The fix itself looks Okay.
Minor: replace in the comment: "compiler don't drop" =>
"compiler doesn't drop".
However, we still have to reach a consensus on how we treat
this issue (as Chris already commented).
Thanks,
Serguei
On 11/8/19 15:22, Alex Menkov wrote:
Hi all,
Please review the fix for
https://bugs.openjdk.java.net/browse/JDK-8215196
webrev:
http://cr.openjdk.java.net/~amenkov/jdk14/popframe_args/webrev/
Currently PopFrame is disabled with JVMCI by [1], so for
testing I reverted [1] changes.
[1] https://bugs.openjdk.java.net/browse/JDK-8218025
--alex
