Executive Summary: Bad page state at prep_new_page
This might or might not be useful to anybody but I googled with the above error and tried to track down how many of these errors had actually been resolved. [It is possible that people don't come back to summarize once the problem is resolved]. I only tracked once that did not get reported on lkml. And I tried to drill down to get as much info as I can.
there were 12 in total that I tracked down starting 04/04/04 till 01/26/05
None had a good resolution reported
4 out of 12 had extensive memtest runs without any errors.
12/12 were for 2.6.7-rc3 onwards.
4/12 had mapping all zeroes [rest did not report or had non-zero mapping]
I might be barking up the wrong tree here without much knowledge, but I have a summary gnumeric spreadsheet [which I won't even think of attaching here] if it is going to be useful to anyone I would be glad to email personally. It also has html links to the original reports.
Thanks,
Jay