HBM | High Bandwidth Memory timing model based on DRAMSim2

Here's a working example that accesses the structures mentioned with comments about corrections to the structure definitions and a port of the C code walking the structure. There is test code I used to verify the accesses were written correctly.

test.py

On a 64-bit system, return values default to c_int (32-bit). At a minimum, set the .restype to at least a c_void_p to indicate a 64-bit pointer is returned.

Ideally, set .argtypes and .restype for each function called.

The problem is caused by an HTTP->HTTPS redirect that was set lately on the hibernate.org website:

I inspected your project on GitHub, thanks for the link. There is so much wrong with it, I hardly know where to begin:

1. You committed library JARs instead of doing proper dependency management using Maven, Gradle or, if you insist to stay on Ant, something like Ivy. Dependencies should be downloaded during the build, not committed into a source code management (SCM) repository. Using Maven or Gradle also would have the advantage that IDEs like IntelliJ IDEA or Eclipse can automatically import your projects and also know where to find the dependencies and share them between projects, instead of you redundantly committing them to each of your project SCM repositories and manually adding them to your IDE project configuration. That is just ugly.

2. One of the JARs in your project is aspectj-1.9.6.jar. This is not what you want but an executable installer, the purpose of which is to install AspectJ locally. In there, you also find nested JARs such as aspectjrt.jar (runtime), aspectjtools.jar (compiler), aspectjweaver.jar (load-time weaving agent). You would have to execute that installer and then copy the libraries you need to your lib directory.

3. In order for the Java compiler to recognise your AspectJ imports, you need the AspectJ runtime on your classpath, i.e. you can download aspectjrt-1.9.6.jar from Maven Central (select "Downloads → jar" in the top right menu). Then you simply copy it to lib and add this to your Ant script:

At the end I found a configuration that seems to work as I wish. I have positioned:

• persistence.xml for tests at src/test/resources/META-INF/,
• persistence.xml for production at src/main/resources/META-INF/,
• the hbm.xml mapping files at src/main/resources/META-INF/,

Also in both test and production persistence.xml I added the tag:

While mapping the property in your code:

you are specifying type as String. Change it to AnsiString instead.

Please refer to NHibernate Documentation and search for "Table 5.3. System.ValueType Mapping Types" in there for list of all types.

Please note that using NVARCHAR (and hence string) is better way to go in many situations considering unicode support.

You shouldn't use hibernate.hbm2ddl.auto=update in the first place. This is needed only to play with Hibernate. In real projects we use Liquibase or Flyway for DB migrations. And we set hibernate.hbm2ddl.auto to validate or to none.

And in the DB migrations you can create or not create FKs as you please.

Also you can give names to FKs in JPA so that it hbm2ddl creates them with the right names.

Having a look at the output of the request:

Hibernate logs in the org.hibernate.SQL category at the DEBUG level, your configuration limits it to INFO so they are silenced. You need to lower that level to DEBUG if you want to see the logs.

This not the same thing as hibernate.show_sql, when that option is enabled it uses the console, not the logger. So it should probably be disabled