Joe Onorato wrote: > Ahh, I see. It's not not finding the JNI functions, it's not able to > resolve some other symbol in the library. What happens if you compile the > .so with --error-unresolved-symbols?
I don't have the toolchain with me right now, so I can't try it. I would, however, not expect it to tell me anything particularly useful --- the linker can't distinguish between symbols that are resolvable at runtime and which ones aren't. (I've actually been using a system using nm and grep to list all unresolved symbols in the library, to try and figure out which ones are causing the problems.) digit wrote: > can I ask you to list the missing functions that you need ? especially the C > library ones. I don't know for sure. I can only enable or disable blocks of code and see what works and what doesn't. Given Google's comments on the use of the C library, I'm not really expecting to be able to use many libc functions. However, I am seeing problems with the C runtime --- the compiler generates references to helper functions with names beginning with __armeabi for doing things like float emulation, and at least some of these appear to be missing. Normally I'd just statically link them and not worry about it, but my CodeSourcery toolchain doesn't appear to have a static libgcc; I'll have to hunt around to see if I can find a copy elsewhere. (*And* its static libc contains fixup types that can't be put into shared libraries. Grumble.) -- ┌─── ｄｇ＠ｃｏｗｌａｒｋ．ｃｏｍ ───── http://www.cowlark.com ───── │ "I have always wished for my computer to be as easy to use as my │ telephone; my wish has come true because I can no longer figure out │ how to use my telephone." --- Bjarne Stroustrup
signature.asc
Description: OpenPGP digital signature
