Sanjay, Thanks for catching that.
In researching this and discussing it with William. This logic had been in the older post install tasks code and migrated when the new ICT was created for x86 support. As can be seen on line: 1647 the comment and underlying code was there. I hadn't realized it might not be needed for OpenSolaris when I did this work for SPARC support, good catch. I opened bug 6238 - "ICT support method get_rootdev_list not needed" to track this as a separate issue. Thanks, Joe Sanjay Nadkarni wrote: > Joe, > Can you elaborate on the comment 1565 - i.e. under what situation > would a root device be a metadevice in OpenSolaris ? > > > -Sanjay > > Joseph J VLcek wrote: >> Thank you Jack. >> >> All suggestions made built and I ran a test of the individual ICT, ICT >> set_Solaris_partition_active, on SPARC. >> >> Joe >> >> >> >> Jack Schwartz wrote: >> >>> Hi Joe. >>> >>> Looks really good. A few small comments. >>> >>> 1563: where did 1024 come from? My guess is that's MAXPATHLEN, but >>> you may want to say so in a comment, and add this to bug 5559 for >>> extracting #defines from C header files. >>> >>> 1567: Why not just use status, and get rid of return_status? >>> >>> 1571: This should be '...status = ' + str(return_status) >>> >>> These are easy, low-risk fixes to make. As such, they don't require >>> retesting except for sanity checking (i.e. that the module compiles) >>> and I won't ask for another code review once these changes are made. >>> >>> Thanks, >>> Jack >>> >>> >>> >>> On 01/26/09 12:56, Joseph J VLcek wrote: >>> >>>> * Please review the changes for Bug 6080: >>>> >>>> http://defect.opensolaris.org/bz/show_bug.cgi?id=6080 >>>> >>>> * webrev >>>> >>>> http://cr.opensolaris.org/~joev/bug6080/ >>>> >>>> * The modules affected and tested: >>>> >>>> ICT - Python code >>>> >>>> --- Testing --- >>>> >>>> * Testing done on SPARC: >>>> ------------------------ >>>> >>>> On a SPARC test system an alternate root was manually populated >>>> using pkg image-create and pkg install of SUNWcsd, SUNWcs and >>>> slim_install. >>>> >>>> Using the eeprom command the default boot-device was set to an >>>> invalid value. >>>> >>>> Then the ict.py updated Class member: set_Solaris_partition_active >>>> was exercised using the ICT Class test member: exec_ict() >>>> >>>> Than a reboot was performed >>>> >>>> * SPARC Results: >>>> ICT set_Solaris_partition_active completed successfully and system >>>> booted to the correct disk. >>>> >>>> * Testing done on x86 (to confirm no regressions:) >>>> -------------------------------------------------- >>>> >>>> [1] Booted LiveCD image on live hardware HP Pavilion dv5000 >>>> [2] mount -F lofs to use the updated ict.py module >>>> [3] Installer run >>>> >>>> * Results: >>>> All ICT completed successfully and system booted >>>> >>>> * Testing done on SPARC: (To be completed prior to push of fix) >>>> --------------------------------------------------------------- >>>> >>>> On SPARC, with help from Jan a full AI install was performed. >>>> >>>> * SPARC Results: >>>> All ICT completed successfully and system booted >>>> >>>> * Description: >>>> >>>> This code change is to set the default boot-device on SPARC. >>>> >>>> * Outstanding issue: >>>> >>>> The Python ICT code needs to be updated to share definitions between >>>> C system headers and the ICT Python code. >>>> >>>> The following bug has been filed and updated with ICT specific >>>> information to track this issue. >>>> >>>> Bug 5559 - Need better scheme for sharing definitions between C & >>>> Python code. >>>> >>>> >>>> >>>> >>>> >>>> >>>> Huge thanks for help to Tycho Nightingale, Jan Damborsky and Jack >>>> Schwartz >>>> >>>> Thank you. >>>> Joe >>>> >> >> _______________________________________________ >> caiman-discuss mailing list >> caiman-discuss at opensolaris.org >> http://mail.opensolaris.org/mailman/listinfo/caiman-discuss >> >
