my bet? Maybe char Arrays allocated during transforms? I ran some OptimizeIt profiles awhile back and seem to remember these numbers creeping up, especially during the i18n transformation.
-----Original Message----- From: Stefano Mazzocchi [mailto:[EMAIL PROTECTED] Sent: Tuesday, March 18, 2003 6:03 PM To: [EMAIL PROTECTED] Subject: Re: [FYI] heapprofile analysis Pier Fumagalli wrote: > On 18/3/03 21:00, "Stefano Mazzocchi" <[EMAIL PROTECTED]> wrote: > > >>I assume, but I'm not sure, that [xx are native internal objects, so >>there is not much we can do about those. Still, I would like to know >>what is that [10 object that accounts for so much memory. > > > They should be (might be) arrays... As that's the way in which you define > them in JNI (roughly).... And the number after that _should_ (but I ain't > sure) the type of array that you're actually looking at. Should be one of: > > #define JVMPI_NORMAL_OBJECT ((jint)0) > #define JVMPI_CLASS ((jint)2) > #define JVMPI_BOOLEAN ((jint)4) > #define JVMPI_CHAR ((jint)5) > #define JVMPI_FLOAT ((jint)6) > #define JVMPI_DOUBLE ((jint)7) > #define JVMPI_BYTE ((jint)8) > #define JVMPI_SHORT ((jint)9) > #define JVMPI_INT ((jint)10) > #define JVMPI_LONG ((jint)11) > > You never _never_ have access to the JVM internal objects from any > whatsoever possible pre-defined interface... And I tried... :-) Ah, gotcha. So, the question becomes: what allocates 500Kb of int arrays? Stefano.
