From: thomas at tada.se (Thomas Hallgren)
Date: Sun, 09 Mar 2014 12:06:37 +0100
Subject: [Pljava-dev] Is it safe to use multi-threaded PL/Java in
 single-threaded postgres process?
In-Reply-To: <30A6B4B499494CED93770BBC4835840C@maumau>
References: <30A6B4B499494CED93770BBC4835840C@maumau>
Message-ID: <531C4B3D.8020102@tada.se>

On 2014-03-08 11:21, MauMau wrote:
> Hello,
>
> Is PL/Java safe to use in terms of its threading design?  I've just 
> asked pgsql-hackers about this and am waiting for response, adding the 
> sentence:
>
> "To put the question in other words, is it safe to load a 
> multi-threaded PL library in the single-threaded backend process, if 
> the PL only calls SPI in the main thread?"
>
> My understanding is:
>
> * PL/Java (pljava.so) is linked with the JNI (Java Native Interface) 
> library, libjvm.so, in JRE.  libjvm.so is linked with libpthread.so, 
> because Java VM is multi-threaded.  SO, "ldd pljava.so" shows 
> libjvm.so and libpthread.so. pljava.so doesn't seem to be built for 
> multh-threading ---  none of -mt, -D_REENTRANT or -D_POSIX_C_SOURCE is 
> specified when building it.
That wouldn't do much good since the psql backend that it links with 
isn't built for multi-threading. This is also why the PL/Java protects 
all calls to the psql backend. Only one single thread at a time can call 
back to the backend.

>
> * When the application calls Java stored function, pljava.so calls a 
> function in libjvm.so to create a JVM in the backend process, then 
> invokes the user-defined Java method in the main thread.  The 
> user-defined Java method calls JDBC methods to access database. The 
> JDBC method calls are translated to backend SPI function calls through 
> JNI.
>
That sounds correct.

> * The main thread can create Java threads using Java Thread API, and 
> those threads can call JDBC methods.  However, PL/Java intercepts JDBC 
> method calls and serializes SPI calls. So, only one thread calls SPI 
> functions at a time.  I'm wondering if this is the reason why PL/Java 
> is safe for use.
>
Yes.

>
> What I'm concerned about is whether multi-threaded code (Java VM) can 
> run safely in a single-threaded code (postgres).  I cannot point out 
> what can be a particular problem with PL/Java, but in general, the 
> mixture of single-threaded code and multi-threaded one seems to cause 
> trouble around handling errno, memory and file handles/pointers.
>
While I understand your concern, I still believe that the PL/Java code 
base does what's needed to run together with the PostgreSQL 
single-threaded back-end. I don't think there are any problems with 
errno or pointers. If there was, such problems would become apparent 
immediately. Either as linker problems or as a general crash when the 
runtime boots up. No such problems have been reported AFAIK.

> FYI, JNI specification says that the code called from Java VM should 
> be built for multi-threading as follows.  But postgres is not.
>
> http://docs.oracle.com/javase/7/docs/technotes/guides/jni/spec/design.html#wp9502 
>
>
> [Excerpt]
> Compiling, Loading and Linking Native Methods
> Since the Java VM is multithreaded, native libraries should also be 
> compiled and linked with multithread aware native compilers. For 
> example, the -mt flag should be used for C++ code compiled with the 
> Sun Studio compiler. For code complied with the GNU gcc compiler, the 
> flags -D_REENTRANT or -D_POSIX_C_SOURCE should be used. For more 
> information please refer to the native compiler documentation.
>
All true, but here we're marrying a JVM runtime with a single threaded 
PostgreSQL backend. Normal recommendations don't really apply.

If you have time to really stress test PL/Java with respect to compiler 
flags etc. that would of course be very valuable. I haven't performed 
such tests since I first built PL/Java back in 2004. A lot has happened 
since.

- thomas

> Regards
> MauMau
>
> _______________________________________________
> Pljava-dev mailing list
> Pljava-dev at lists.pgfoundry.org
> http://lists.pgfoundry.org/mailman/listinfo/pljava-dev