[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#339565: sun sparc ultra 1e etch beta1



(I'm aware that there are overlaps in the addressing, but still feel this 
to be correct in this case; please limit address list when replying.)

On Thursday 17 November 2005 09:18, Jurij Smakov wrote:
> discover1 had support for sbus, discover (aka discover2) does not.
> However, I was told by Colin Watson that d-i is using hotplug for
> device discovery and finding the correct drivers these days.

There seems to be some confusion about this, so here is the current 
situation:

- d-i still uses discover1 for both Sarge and Etch beta1

- we have just started implementing udev support together with kernel
  2.6.14 for daily sid_d-i builds, but without disabling discover1

- hotplug is no longer relevant for d-i as the most recent version of
  udev takes care of that itself

- discover probably will be disabled for installations using 2.6.14, if
  only to get a feel where open issues are; if there are significant open
  issues (unsupported drivers) we will have to see how these can be solved
  or worked around

- we will almost certainly keep on using discover for < 2.6.14
  installations; a change from discover1 to discover2 is definitely an
  option for beta2, now would be a good time to discuss this


For this bug this means that if there is a regression, it needs to be 
debugged; if it's just a missing device spec in discover, please provide 
the information needed to add the support (as already requested by 
joeyh).


Discover maintainers (reply only to d-boot list s.v.p.):
- is there any real interest in doing a migration from discover1 to
  discover2 for d-i
- if so, can you give an overview of the main differences and in what
  areas improvements/problems can be expected
I see that  for discover2 the program udeb if quite a bit bigger (5-7x); 
the data udeb is marginally smaller. I also note that the maintenance of 
discover1 and discover seems to be done by separate teams.

Cheers,
FJP

Attachment: pgpspS2mWRVqt.pgp
Description: PGP signature


Reply to: