camcontrol rescan all fails if there is no bus 0

[Scott Long]

On Oct 26, 2010, at 10:04 AM, Ed Maste wrote:

> On Tue, Oct 26, 2010 at 09:55:24AM -0600, Scott Long wrote:
> 
>> On Oct 25, 2010, at 10:59 AM, Ryan Stone wrote:
>> 
>>> ...
>>> Index: camcontrol.c
>>> ===================================================================
>>> --- camcontrol.c	(revision 214271)
>>> +++ camcontrol.c	(working copy)
>>> @@ -1519,6 +1519,7 @@ rescan_or_reset_bus(int bus, int rescan)
>>> 	bzero(&(&matchccb.ccb_h)[1],
>>> 	      sizeof(struct ccb_dev_match) - sizeof(struct ccb_hdr));
>>> 	matchccb.ccb_h.func_code = XPT_DEV_MATCH;
>>> +	matchccb.ccb_h.path_id = CAM_BUS_WILDCARD;
>>> 	bufsize = sizeof(struct dev_match_result) * 20;
>>> 	matchccb.cdm.match_buf_len = bufsize;
>>> 	matchccb.cdm.matches=(struct dev_match_result *)malloc(bufsize);
>>> ...
>> ...
>> That being said, there's a comment block right above the code that you changed that suggests that the XPT pseudo-bus handler won't behave correctly if it gets a rescan or reset command, hinting that that's why the enumeration starts specifically a '0' instead of 'CAM_BUS_WILDCARD'.  Did you see any evidence of this in your testing?
> 
> That's this comment block I presume:
> 
> /*
> * The right way to handle this is to modify the xpt so that it can
> * handle a wildcarded bus in a rescan or reset CCB.  At the moment
> * that isn't implemented, so instead we enumerate the busses and
> * send the rescan or reset to those busses in the case where the
> * given bus is -1 (wildcard).  We don't send a rescan or reset
> * to the xpt bus; sending a rescan to the xpt bus is effectively a
> * no-op, sending a rescan to the xpt bus would result in a status of
> * CAM_REQ_INVALID.
> */
> 
> I think it's still true; Ryan's change uses CAM_BUS_WILDCARD for the
> XPT_DEV_MATCH ioctl in order to enumerate the busses, and the path_id
> that gets used for the actual rescan or reset should be that returned
> from the match should it not?
> 
> -Ed

Ok, I guess I need to see if CAM_XPT_PATH_ID can get returned from the devmatch.  What happens with this patch if you run it with no real buses registered (i.e. cam module present, but no SIMs present)?

Scott