Progress update:
Devmiyax from the Yabause forums pointed out a couple days ago that although the dynarec was compiling, he doesn't believe it is actually engaged by the app. If this is true, there would be a major speed improvement if it were engaged. To be honest, I am somewhat sceptical, because for an interpreted core it is running extremely fast for such slow hardware (talking about my Droid X here, which is comparable hardware to the Pandora). I made the changes to the code that devmiyax suggested, but now it crashes. When I have some time to look at this more closely, I'll determine for sure if the dynarec is engaged or not, and if not I'll have to figure out why it crashes (I have tried builds with the ARM5, ARM6, and ARM7 versions of the dynarec, but all crash). I seriously suck at assembly language, so if there is an issue with the dynarec, it is highly doubtful that I'll be able to fix it myself. I'll see if Ari64 will be willing to help me debug the problem (I'm pretty sure I tried his patience back when he helped me debug the Android port of his N64 dynarec..) If he decides not to tackle the problem, I'll still continue to work on the other components I have planned (audio and GLES renderer), and hopefully another dev with ARM assembly skills might join the project to help debug the dynarec.