• NetBSD
• evbmips
• broadcom

NetBSD/evbmips/broadcom

Week 1

• CFE traps a crash in initial trials of ported-to-evbmips-tree NetBSD/bcmmips code from Ada Lim.
  • Shaded suggests the fix, adding my CPU to a lookup table
• Initial console code goes through a few revisions during the week.
• sibabus driver
• bce at sibabus
• bce has 'descriptor protocol error' all week
• sibapci
• Death of 'Bob'
  • Somehow during the testing of initial sibapci(4) DMA code CFE is destroyed.
  • The spare WGT634U that replaces Bob is modified to disallow writes to flash.

Week 2

• investigated and fixed bce(4) DMA problems, bad bus_dmamap_sync(9) call.
• ubsec at sibabus
• Discovered documentation. Oh thank FSM.
• interrupt dispatch issues
  • finally solved with the documentation's help.

Still To Do

• Switch configuration
• Bus-Mastering and Memory-space accesses to ath0 reverting to bad settings.
• ehci(4) write issue
• proper multiple bus attachment code sharing for bce(4) and ubsec(4).
• all the proper sibabus goodness
  • chip common core
    • com0
    • com1
    • Flash
    • GPIO
  • ohci at sibabus (won't be able to test)
  • Timer0 in core 5
• getting boot information from CFE