resolver: abuse _res a little less

[vangyzen]

Summary

In the past, _res was a global variable. Now, it's multiple function calls.
Several functions in the resolver use _res multiple times and therefore
call the function(s) far more than necessary.

Fix those callers to store the result of _res in a local variable.

[jilles]

Reading the code, this looks OK, but in compilation an undesirable side effect can be seen. This is on amd64 with 11-current's default clang. The improvements to res_init() cause it to fall below a threshold and be inlined everywhere, leading to a 15% code size increase rather than the expected decrease (in the "text" column of size(1)).

Adding __noinline to the definition of res_init() fixes this, giving a code size decrease of 3%.

A further function could be extracted with the check for -1 and RES_SET_H_ERRNO(statp, NETDB_INTERNAL); for further size reduction.

__predict_false might have a similar effect (I have not tried this) but the comments in <sys/cdefs.h> discourage its use for this purpose.

[vangyzen]

Add __noinline to the definition of res_init() to avoid the code bloat
that these changes would have otherwise incurred. (Thanks, jilles.)