Dan Smith wrote:
KR> + unsigned_val = value;
I think you should cast this explicitly to avoid a warning in the future. We have to compile without strict warnings because the generated parser code is riddled with issues, but I think we should try to make a routine such as this one as clean as possible.
Good catch =)
KR> + s = CMSetProperty(inst, prop, &(unsigned_val), type);
Don't you want a (CMPIValue *) cast in front of your &(unsigned_val)?
I think you can leave this off for integer types - it seems to work both with and without. The original call in eo_util_parse.y didn't have it. But I need to resend, so I'll add it for consistency =)
KR> + EOTRACE("propertyname = %s\n", $1); KR> + int rc; KR> + //uint64_t value = $3; KR> + CMPIType t = set_int_prop($3, $1, *_INSTANCE);
I think you've got some whitespace issues here, and I think we can remove the commented-out code, right?
Oops - I though I'd gone back and removed the commented out code. Sending updated patch =) -- Kaitlin Rupert IBM Linux Technology Center kaitlin@linux.vnet.ibm.com