question about java::lang::String::getChars

Mon Oct 20 19:25:00 GMT 2003

On 2003年10月19日, Randolph Chung wrote:
> While running the libjava testsuite on hppa, I notice that at least two
> tests were triggering unaligned accesses. The unaligned access happens
> in java::lang::String::getChars in the copy loop. Looks like sPtr is
> pointing to an unaligned address (0x50069 in this case). Is that
> supposed to happen? If so, should that method use memcpy instead?

Wild guess: your GC is broken, and the string object is overlaid with
non-String data?
(Is 0x50069 even a valid address on hppa?)
Jeff