Scott Long wrote: > Roman Kurakin wrote: > >> Scott Long: >> >>> Maksim Yevmenkin wrote: >>> >>>> Scott Long wrote: >>>> >>>>> Maksim Yevmenkin wrote: >>>>> >>>>>> Dear Hackers, >>>>>> >>>>>> any objections to the attached patch? >>>>>> >>>>> >>>>> Yes, as I stated in another email, I think that the core netgraph >>>>> module should be initialized before the SI_SUB_DRIVERS step. I >>>>> propose creating a new sysinit called SI_SUB_NETGRAPH with a value >>>>> of 0x30100000. That way it comes after SI_SUB_IF and before >>>>> SI_SUB_DRIVERS. This make fiddling with SI_ORDER_* unneccesary. >>>> >>>> >>>> >>>> >>>> how about new attached patch? >>>> >>>> thanks, >>>> max >>> >>> >>> >>> >>> Exactly what I had in mind =-) Have you tested this out to make sure >>> it fixes the problem cases? >> >> >> >> But this wouldn't save from the same problem it the future. >> >> rik >> > > What same problem? This ensures that the netgraph core gets initialized As I understand this situation we could get the same problem for any two modules if one depends on other and both of them have the same load order. If we teach "kernel" modules to check for dependences to sort modules of the same order this would solve such problems. We may also think what to do with 'buggy' dependency but this would be other step. At first we need to do this one. I agree that NETGRAPH should be treated other way and I don't object from patch that will solve this particular problem. It should be applied not only because this problem. But I believe that we also should think about other possible situations like this one. rik > before any driver. Keeping it at SI_SUB_DRIVERS and trying to order the > it via SI_ORDER_* is risky because you can't guarantee that some other > driver won't try to also take SI_ORDER_FIRST. > > ScottReceived on Thu Jan 06 2005 - 09:45:01 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:25 UTC