Smoothing length failed to converge

Since the last few commits (ironically, the ones without @pdraper supervision), I've started getting warnings when running the SodShock example that I use for testing:

[00002.5] main: Running on 1024128 gas particles and 0 DM particles until t=1.000e-02 with 1 threads and 1 queues (dt_min=1.000e-02, dt_max=1.000e-02)...
[00002.5] engine_init_particles: Initialising particles
[00003.3] engine_print_task_counts: task counts are [ none=0 sort=14256 self=28512 pair=370656 sub=0 init=14256 ghost=14256 drift=14256 kick=14256 send=0 recv=0 grav_pp=0 grav_mm=0 grav_up=0 grav_down=0 part_sort=1 gpart_sort=1 split_cell=0 rewait=0 skipped=0 ]
[00003.3] engine_print_task_counts: nr_parts = 1024128.
[00003.3] engine_print_task_counts: nr_gparts = 0.
[00005.2] runner_doghost: Smoothing length failed to converge on 1 particles.
[00007.3] runner_doghost: Smoothing length failed to converge on 1 particles.
[00009.6] runner_doghost: Smoothing length failed to converge on 2 particles.
[00010.4] runner_doghost: Smoothing length failed to converge on 2 particles.
[00011.6] runner_doghost: Smoothing length failed to converge on 1 particles.

Does this have to do with the recent refactoring of the hydro interactions? For completeness, I'm running

./swift_fixdt -t 1 -f SodShock/sodShock.hdf5 -m 0.01 -w 5000 -c 0.01 -d 1e-7 -e 0.01