A memory leak was reported in 4317277. The submitter says that the leak is
not observable in JDK1.3 builds R,S,U (I assume FCS). This probably means that the test case submitted for 4317277 will not show the leak. However, the code and comment below are still in the Kestrel-FCS-U build. We should investigate
this for Merlin to see if it is a leak.

Java_sun_awt_motif_MComponentPeer_setFont

awt_Component.c
 
/*Leak baby leak! */
    fdata = awtJNI_GetFontData(env, f, &err);
    if (fdata == 0) {
        JNU_ThrowInternalError(env, err);
        AWT_UNLOCK();
        return;
    }


The fdata is practically never getting freed. The only
place it can be freed is in the fonts dispose method and
it is never getting called, with the Buta testcase. This
explains the constant rate of the leak.