anaconda in 7.2 now looks at partition contents rather than partition type to determine what to make into swap, which is a vast improvement over previous releases. Unfortunately, 1.) it still wants to turn any and all Solaris x86 slices into swap, and 2.) it mkswap'd, swapon'd, and put these partitions (all of them) in my fstab *despite* my clicking "leave this alone" for all of them and then clicking OK to a dialog that said those partitions would be left alone. (Unless anaconda never shows swap partitions to be destroyed, in which case that's another bug.) I'd file this as an RFE for "just create swapfiles and don't destroy my data," except I doubt it'd get far given the history of the required swap partition RFE. In any case, anaconda really needs to be much less aggressive about mkswap'ing everything in sight, especially when specifically told not to.
Hmm... you should have been asked to confirm formatting them as swapping. Marking for fixing for the next release.
If you have these slices and you run parted on the drive and print the partition table from inside parted, what is the output?
I no longer have access to the machine in question. Is there something specific you're looking for?
This works just fine for me here. I just did an install of Solaris x86 leaving space on the disk free, and then did another install of our current test trees, said "No" to the question about using them as swap partitions and they didn't get mkswap'd or used as swap partitions.