Changeset: 14ffe87f27c6
Author:jrose
Date: 2011-05-25 00:15 -0700
URL: http://hg.openjdk.java.net/mlvm/mlvm/rev/14ffe87f27c6
netbeans/meth project
! netbeans/meth/build.xml
! netbeans/meth/nbproject/build-impl.xml
! netbeans/meth/nbproject/genfiles.properties
! netbeans/meth/nbproj
Changeset: 8e5985e788d6
Author:jrose
Date: 2011-05-25 01:48 -0700
URL: http://hg.openjdk.java.net/mlvm/mlvm/jdk/rev/8e5985e788d6
meth: incorporate more review comments; add security manager logic
! meth-doc-7014005.patch
! meth-review-7032323.patch
On May 25, 2011, at 5:58 AM, Ola Bini wrote:
> Hi,
>
> There are at least three problems that are still there. They might be
> connected, or not.
> (I will tell you how to reproduce these at the end)
>
> I just built a new JVM:
> openjdk version "1.7.0-internal"
> OpenJDK Runtime Environment (bui
Changeset: cc4e3c77aee3
Author:jrose
Date: 2011-05-25 02:36 -0700
URL: http://hg.openjdk.java.net/mlvm/mlvm/jdk/rev/cc4e3c77aee3
meth: add explanatory text
! meth-review-7032323.patch
+ meth-review-7032323.txt
___
mlvm-dev mailing list
m
Hi,
Just to clarify, my builds are against the current patchset in the MLVM
repository, so that might explain why you're not seeing these problems.
Cheers
On 2011-05-25 14.29, Christian Thalinger wrote:
> On May 25, 2011, at 5:58 AM, Ola Bini wrote:
>> Hi,
>>
>> There are at least three problems
On May 25, 2011, at 12:20 PM, Ola Bini wrote:
> Hi,
>
> Just to clarify, my builds are against the current patchset in the MLVM
> repository, so that might explain why you're not seeing these problems.
I know that. That's why we need to find out where the problem is (some hints
below). Can som
This is the last major bundle of changes for JDK 7.
http://cr.openjdk.java.net/~jrose/7032323/webrev.00/
(Note: This is an engineering code review. The associated API documents are
also under review and are about to be finalized by the Expert Group. Here is a
recent version: http://cr.openj
On May 25, 2011, at 3:20 AM, Ola Bini wrote:
> Just to clarify, my builds are against the current patchset in the MLVM
> repository, so that might explain why you're not seeing these problems.
Good news: The jdk7 repository just integrated a bunch of critical 292 fixes.
I hope the bsd-port can
Well, on my linux box with a JDK built this morning I don't see any of
these problems actually. (However, there are things on my master that
only works with the mlvm patches. Type conversions specifically.)
Problem 3 also seems to not be there - you would notice if it were since
it's also a total
On May 25, 2011, at 3:23 AM, Christian Thalinger wrote:
> I know that. That's why we need to find out where the problem is (some hints
> below). Can someone provide a jar file of the current MLVM JSR 292 classes
> (like John's meth.jar)?
I just posted it with the javadoc:
http://cr.openjdk.
On Wed, May 25, 2011 at 12:18 AM, John Rose wrote:
> On May 23, 2011, at 5:43 PM, Charles Oliver Nutter wrote:
>
> Well I have a prototype GWT in place, but I'm having trouble getting
> invokeExact to work. As far as I can tell, incoming arguments should
> match the handle type just fine, but I'm
On Wed, May 25, 2011 at 11:31 AM, Charles Oliver Nutter
wrote:
> Build in progress! I'll let you know how it goes. How low is low in "low
> arity"?
Ok, we're definitely back to having solid performance, but it's not as
good as it was before. I'm not sure if it's related to the inlining
threshold
Actually now that I look at it, this isn't all that solid performance.
It's just *slightly* faster than using JRuby's inline cache, which
almost never inlines...
Inline cache only, via CachingCallSite.
~/projects/jruby ➔ jruby -Xcompile.invokedynamic=false --server
bench/bench_fib_recursive.rb 5
On May 25, 2011, at 9:31 AM, Charles Oliver Nutter wrote:
> Build in progress! I'll let you know how it goes. How low is low in "low
> arity"?
About 0-9.
Let's find the next bottleneck!
-- John
___
mlvm-dev mailing list
mlvm-dev@openjdk.java.net
htt
On Wed, May 25, 2011 at 2:14 PM, John Rose wrote:
> On May 25, 2011, at 9:31 AM, Charles Oliver Nutter wrote:
>
>> Build in progress! I'll let you know how it goes. How low is low in "low
>> arity"?
>
> About 0-9.
>
> Let's find the next bottleneck!
That's the spirit! We can do it!
- Charlie
__
I've got the bimorphic method handle inline working. I was doing something
stupid at the beginning which is why it wasn't as easy as I expected to be and
then I encountered a stupid API which wasted a little more time. I was able to
leverage the existing invokedynamic call site cache code gen
On Wed, May 25, 2011 at 5:12 PM, Tom Rodriguez wrote:
> I've got the bimorphic method handle inline working. I was doing something
> stupid at the beginning which is why it wasn't as easy as I expected to be
> and then I encountered a stupid API which wasted a little more time. I was
> able t
Oh my goodness...I just realized an obvious missing API...is there still time?!
SwitchPoint has no way to query if it has been invalidated!
Is this by design? I would like to use this as both a call site guard
and as a queryable trigger...to know whether I can go forward with the
faster SwitchPoi
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 2011-05-26 10.33, Charles Oliver Nutter wrote:
> Oh my goodness...I just realized an obvious missing API...is there still
> time?!
>
> SwitchPoint has no way to query if it has been invalidated!
>
> Is this by design? I would like to use this as
Ok, onward with perf exploration, folks!
I'm running with mostly-current MLVM, with John's temporary reversion
of GWT to the older non-ricochet logic.
As reported before, "fib" has improved with the reversion, but it's
only marginally faster than JRuby's inline caching logic and easily
30-40% slo
On Thu, May 26, 2011 at 12:19 AM, Ola Bini wrote:
> Take a look at the way I use SwitchPoints in Seph. The api you're
> looking for is already there:
>
> SwitchPoint x = new SwitchPoint();
> MethodHandle valid =
> x.guardWithTest(MethodHandles.constant(boolean.class, true),
> MethodHandle.constant
On May 25, 2011, at 10:34 PM, Charles Oliver Nutter wrote:
> On Thu, May 26, 2011 at 12:19 AM, Ola Bini wrote:
>> Take a look at the way I use SwitchPoints in Seph. The api you're
>> looking for is already there:
>>
>> SwitchPoint x = new SwitchPoint();
>> MethodHandle valid =
>> x.guardWithTest
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
>
> Well, lacking a method, it's actually probably the most
> straightforward one :-) Ola wins gazillion points for
> resourcefulness.
Actually, it was Remi who came up with this gem when we discussed adding
an isInvalid() method on SwitchPoint.
The
On Thu, May 26, 2011 at 12:55 AM, Ola Bini wrote:
> Actually, it was Remi who came up with this gem when we discussed adding
> an isInvalid() method on SwitchPoint.
I don't say it's not clever. It's just not something anyone is
typically going to figure out on their own. There will be constant
qu
Ok, here we go with the macosx build from 5/13. Performance is
*substantially* better.
First tak:
user system totalreal
1.401000 0.00 1.401000 ( 0.821000)
0.552000 0.00 0.552000 ( 0.552000)
0.561000 0.00 0.561000 ( 0.561000)
0.552000 0.0
Now for something completely different: SwitchPoint-based "constant"
lookup in JRuby.
It's certainly possible I'm doing something wrong here, but using a
SwitchPoint for constant invalidation in JRuby (rather than pinging a
global serial number) is significantly slower.
Using SwitchPoint:
~/proj
On Thu, May 26, 2011 at 1:34 AM, Charles Oliver Nutter
wrote:
> Ok, here we go with the macosx build from 5/13. Performance is
> *substantially* better.
It's worth mentioning that the 5/13 build results are only slightly
slower than our "ideal", JRuby's dynopt mode. That made me water at
the mout
On May 25, 2011, at 9:39 PM, Charles Oliver Nutter wrote:
> On Wed, May 25, 2011 at 5:12 PM, Tom Rodriguez
> wrote:
>> I've got the bimorphic method handle inline working. I was doing something
>> stupid at the beginning which is why it wasn't as easy as I expected to be
>> and then I encoun
Overall it seems ok. A few minor oddities:
MethodHandle.java
variable arity is sometimes hyphenated. It seems more correct without but be
consistent.
AdapterMethodHandle.java:
The new value srcSlot appears unused.
Should this just go away:
-if (argCount <= 2) return false; // mus
29 matches
Mail list logo
