2nd Set of IIO fixes for the 5.5 cycle.

* Core
  - Fix buffer alignment calculation.  The previous logic meant that scans
    could under some circumstances not have naturally aligned elements.
    In theory someone might have been working around this in userspace, but
    if so we'd have expected them to query this very odd behaviour.
    It's not hit in that many drivers as requires a mixture of channel
    resolutions broad enough to require different packing.  The one obvious
    case of timestamps was fine under existing code as they are always the
    last element if enabled and that was handled correctly.
* ad7124
  - DT channel configuration wasn't using the reg value so would lead
    to some odd mappings and readout of the wrong channels.
* pms7003
  - Kconfig didn't specify triggered buffer dependency.
* st_lsm6dsx
  - For the LSM6DS3 the id match wasn't working.
* vcnl4000
  - Scale was wrong for the vcnl4040
-----BEGIN PGP SIGNATURE-----
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=f5WE
-----END PGP SIGNATURE-----