Hi Keith,

The changes look OK to me.

BTW thanks for the helpful comments. :-)

-evan

On 2/19/10 1:32 PM, Keith Mitchell wrote:
> All,
>
> I'm still looking for a reviewer or two on this, if anyone has time.
> I've updated the webrev to also include the changes for 14601, since
> that bug affects similar areas of code (and exposed an issue I had with
> the original webrev) - so if you happened to start looking at the first
> set of changes and just hadn't sent out a response yet, please take a
> look at the updated files:
> http://cr.opensolaris.org/~kemitche/13924.14601/
>
> Thanks,
> Keith
>
> On 02/16/10 01:55 PM, Keith Mitchell wrote:
>> Hello all,
>>
>> I'd like to request a code review for the following text installer bug:
>> http://defect.opensolaris.org/bz/show_bug.cgi?id=13924
>> 13924 Invalid existing disk configurations result in indeterminate
>> Text UI behavior
>>
>> Webrev:
>> http://cr.opensolaris.org/~kemitche/13924/
>>
>> Summary of issues at hand:
>> The text installer UI makes the faulty assumption that disk
>> configurations it gets back from target discovery are valid "as-is".
>> This resulted in some undesirable behavior, such as installing into
>> the incorrect partition when the disk had two existing Solaris2
>> partitions. This bug fix shores up the UI understand those situations.
>> Note that this fix does not add support for creating or maintaining
>> odd disk configurations, but rather, prevents the user from attempting
>> an installation in cases where the installer's results would be
>> indeterminate, and protects the UI from crashing as a result of
>> assuming that all existing partition/slice layouts are sane.
>>
>> Testing:
>> - Ensured that UI forced user to delete Solaris2 partitions before
>> continuing, if more than one Solaris2 partition existed on the disk
>> - Created VTOC layout with overlapping slices. Ensured that values for
>> "Avail" column on slice editing screen were accurate in context.
>> Ensured that leaving such overlapping untouched and attempting an
>> install caused the new rpool to be created on the correct slice with
>> correct geometry (the overlapping slices were left untouched). Also
>> verified that the overlapping slices could be deleted and that the
>> installed system had the expected VTOC geometry.
>>
>> - Keith
>> _______________________________________________
>> caiman-discuss mailing list
>> caiman-discuss at opensolaris.org
>> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss
> _______________________________________________
> caiman-discuss mailing list
> caiman-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss

Reply via email to