Just FYI, I had the exact same issue as the reporter, and it turned out that one of my kernel commandline args (previously set for a no-longer present SATA controller) was causing it. I'm not sure exactly which it was, but removing the following args fixed it.
libata.noacpi=1 pci=assign-busses,ecrc=on,realloc,hpmemsize=512M,nocrs