Including the write generation for each page in the output of

wt verify -d dump_address

would make it easier to debug certain data corruption failures.

In these scenarios it can be useful to figure out when specific blocks were (or weren't) written. One tool for this is the write generation number. It doesn't provide a wall clock time, but it does provide an ordering.

When there is a checksum mismatch, we can feed the bad data (from the hexdump in the error messages) into a tool such as wt_binary_decode.py. If there is any valid WT block header in the dump, this will provide the corresponding write generation. 

But I don't see a way to get similar information from the various "dump" options to wt verify. So we can't connect a write generation from a corrupt block to the good blocks in the table.

For this ticket, I would like to include the write generation for each block (as stored in the page header) in the output of wt verify -d dump_address. If there is a concern about increasing the size of that output, it could be a command line option.