Subject: Re: 3.0 pcmcia regressions?
To: Christos Zoulas <christos@zoulas.com>
From: John Nemeth <jnemeth@victoria.tc.ca>
List: tech-kern
Date: 01/04/2006 21:03:02
On May 27,  3:12am, Christos Zoulas wrote:
} On Jan 4,  1:15am, jnemeth@victoria.tc.ca (John Nemeth) wrote:
} 
} | } a similar patch fixes a few other devices, but not all of them. I
} | } am ok for pulling this up to 3.0 as long as there is a big XXX comment
} | 
} |      Just 3.0, or -current as well?
} 
} Just 3.0. For current we need to find out what is wrong...

     Okay.

} | } next to it saying that this is a workaround to another problem.
} | 
} |      Documented in doc/HACKS as well?
} | 
} |      I don't know PCMCIA very well.  Can somebody provide a short
} | description of the problem and what the patch does, please?
} 
} For HACKS in 3.0? Something like: On some cards the card align field
} property value interferes with the rbus resource allocation causing
} the card not to configure properly. Setting it to 0 fixes the problem
} but it is not the proper solution. This is done on aic_pcmcia.c

     I really meant for the XXX comment, but I guess I can use a
shortened form, i.e. "gross hack to work around rbus resource
allocation problem"?  I applied the same patch to com_pcmcia.c.  Can
that be rolled in as well?  Where can I learn more about PCMCIA, both
in general and NetBSD's implementation (other then pcmcia(9))?

}-- End of excerpt from Christos Zoulas