regardless if unblinding of practitioners was unavoidable or not, is blinding of patients only without the practitioners considered high or low risk of performance bias?
it would depend if blinding of the practitioner was possible. If it was and they weren't blinded, you would more than likely rate them high risk. If it wasn't, then your trigger question should be were the outcome assessors blinded. If they weren't then again you would likely rate them as medium or high risk of bias.
It is difficult to provide a categorical answer that is "high risk" or "low risk" without also considering the context of the study. For example, the answer is dependent on the type of intervention which the practitioner is unblinded to e.g. the extent to which the practitioners actions/practices may inadvertently alter the effect of the intervention, or not. Similarly, to what extent is the intervention being measured considered to be routine practice, or not?