As with Mitsuba PR 1317, the Dr.Jit stubs were similarly being generated twice on Windows.
Although less frequent, there are examples on the Windows CI of failed builds that appear to be because of race conditions as stub files are attempted to be written into by two independent processes.
As with Mitsuba PR 1317, the Dr.Jit stubs were similarly being generated twice on Windows.
Although less frequent, there are examples on the Windows CI of failed builds that appear to be because of race conditions as stub files are attempted to be written into by two independent processes.