https://bugzilla.kernel.org/show_bug.cgi?id=201609 Emil Velikov (emil.l.velikov@xxxxxxxxx) changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |emil.l.velikov@xxxxxxxxx --- Comment #5 from Emil Velikov (emil.l.velikov@xxxxxxxxx) --- The other five files ({subsystem_,}{vendor,device} and class) do not have a prefix, so the more reasonable thing was it not add one for the revision file. That said, I did spend a lot of time going through the Documentation looking for guidelines, rules or restrictions that should be applied and could not find any. The PCI subsystem maintainer did not suggest adding a prefix either, so I would not use "blatently silly" here. On a more practical side - the aic94xx driver custom revision file solves the exact same problem my patch does. Admittedly, aic94xx lacks the leading "0x" for the hexadecimal number provided. My personal inclination is that we'd want to check if existing userspace requires the leading 0x and if so to what extend: - cosmetic - remove the aic94xx code, update userspace - unused - remove the aic94xx code - major impact - the aic94xx driver removes the generic revision file, before setting it's own Does that sound reasonable, am I missing something? -- You are receiving this mail because: You are watching the assignee of the bug.
