Index for time_shift param was off by one.

Add appropriate offset to duotone calculation when time_shift is not zero

See merge request cds/software/advligorts!445

Better cleanup of netlink socket, some errors can cause a model not to restart with the old code

See merge request cds/software/advligorts!444

When time_shift on the duotone ADC (ADC 0) is n > zero, then duotone input points
are shifted later in time by n points.  This change corrects back duotone
to the value that would be calculated with time_shift == 0.

Bug would cause time_shift to be assigned to the wrong adc it some configurations with 18 bit ADCs.

Update the slow controls SDF to export its DCUID in EPICS like regular models do.

See merge request cds/software/advligorts!440

Work on issue #431

Addding better error message output in card count and card 2 slot failure cases

See merge request cds/software/advligorts!434

More consistent language/units around autocal output, and more verbosity

Closes #426

See merge request cds/software/advligorts!432

Librts integrated advligo and into rtcds build of models.

See merge request cds/software/advligorts!428

Adding mutex so we only process one mesage at once

See merge request cds/software/advligorts!430

Adding better failure checking in the dolphin proxy km

See merge request cds/software/advligorts!429

Cleaning up a userspace build warning

See merge request cds/software/advligorts!427

Mbuf spinlock deadlock, and small cdsrfm error state fix

See merge request cds/software/advligorts!426

Changing dolphin_proxy_km to match dkms dolphin-proxy-km

See merge request cds/software/advligorts!424

Adding more flexability to dolphin_proxy_km makefile for dkms builds.

See merge request cds/software/advligorts!420