Implementing new algorithm in Deuce

15 views
Skip to first unread message

Guy Korland

unread,
Nov 24, 2009, 3:57:27 PM11/24/09
to Deuce-STM developers, mihai...@gmail.com
Hi Guy,

I just tested my algorithm in Deuce and it didn't work. I got the
following error for every thread:

Exception in thread "Thread-1" java.lang.NullPointerException
at org.deuce.transaction.swisstm.Context.init(Unknown Source)
at org.deuce.benchmark.bank.Account.transfer(Account.java)
at
org.deuce.benchmark.bank.BenchmarkThread.step(BenchmarkThread.java:77)
at org.deuce.benchmark.BenchmarkThread.run(BenchmarkThread.java:
30)
at java.lang.Thread.run(Thread.java:619)


Can you tell me while this happens?
My code doesn't execute at all. I tried putting some print statements
both in the Context constructor
and as the first statement in the init method. None of them were
executed.
What's going on? Any help would be appreciated. Thanks.

Mihai

Guy Korland

unread,
Nov 24, 2009, 3:58:07 PM11/24/09
to Deuce-STM developers
I managed to start debugging my implementation with the tests, not the
benchmarks.
I found bugs in the read, write and commit methods but not in the
init.
I can go along without solving this problem, at least for the moment.
It's still weird though.

Mihai

Guy Korland

unread,
Nov 24, 2009, 3:59:47 PM11/24/09
to Deuce-STM developers, Mihai Letia
Can you post a test case?

Guy

Mihai Letia

unread,
Nov 24, 2009, 5:16:58 PM11/24/09
to Guy Korland, Deuce-STM developers
I used the counters test provided in the framework.
This gave me more specific error messages and I could see the output of
my prints.
No errors in the init method, opposed to what the previous output would
indicate.

Mihai

Guy Korland

unread,
Nov 25, 2009, 12:33:10 AM11/25/09
to Mihai Letia, Deuce-STM developers
Mihai,

Sorry I lost you is there any problem?

Guy
--
Guy Korland

Mihai Letia

unread,
Nov 25, 2009, 4:26:25 AM11/25/09
to Guy Korland, Deuce-STM developers
Just that it's impossible to debug using the benchmarks.
Not even the print statements give output.
It's OK when using the tests.

Mihai

Guy Korland wrote:
> Mihai,
>
> Sorry I lost you is there any problem?
>
> Guy
>
> On Wed, Nov 25, 2009 at 12:16 AM, Mihai Letia <mihai...@gmail.com
> <mailto:mihai...@gmail.com>> wrote:
>
> I used the counters test provided in the framework.
> This gave me more specific error messages and I could see the
> output of
> my prints.
> No errors in the init method, opposed to what the previous output
> would
> indicate.
>
> Mihai
>
> Guy Korland wrote:
> > Can you post a test case?
> >
> > Guy
> >
> > On Nov 24, 10:58 pm, Guy Korland <gkorl...@gmail.com
> <mailto:gkorl...@gmail.com>> wrote:
> >
> >> I managed to start debugging my implementation with the tests,
> not the
> >> benchmarks.
> >> I found bugs in the read, write and commit methods but not in the
> >> init.
> >> I can go along without solving this problem, at least for the
> moment.
> >> It's still weird though.
> >>
> >> Mihai
> >>
> >> On Nov 24, 10:57 pm, Guy Korland <gkorl...@gmail.com
Reply all
Reply to author
Forward
0 new messages