Handle int rank issues in in vn_getsize_locked() and vn_seek()

[kib]

Summary

In vn_getsize_locked(), when storing vattr.va_size of type u_quad_t into
off_t size, we must avoid overflow.

Then, the check for fsize < 0, introduced in the commit
f45feecfb27ca51067d6789eaa43547cadc4990b 'vfs: add vn_getsize', is nop [1].

Reported by:    jhb
Coverity CID:   1502346

[jhb]

sys/kern/vfs_subr.c

7133

Is this something we can assert should never happen?

[kib]

sys/kern/vfs_subr.c

7133

Sane filesystem would not do that indeed, but we did not required this before, and allowed full u_quad_t. My concern with adding the assert is that it would be hard to reach, esp. in non-targeted testing, so the assert would be a ticking bomb.

I prefer to have the check and error return, due to this.