r1294 - developers/werner

werner at sita.openmoko.org werner at sita.openmoko.org
Fri Mar 9 01:32:28 CET 2007 

Author: werner
Date: 2007-03-09 01:32:24 +0100 (Fri, 09 Mar 2007)
New Revision: 1294

Modified:
   developers/werner/TODO
Log:
Cleanup and update. Removed "items done" list, since it just gets confusing.
It's in SVN anyway :-)



Modified: developers/werner/TODO
===================================================================
--- developers/werner/TODO	2007-03-09 00:13:48 UTC (rev 1293)
+++ developers/werner/TODO	2007-03-09 00:32:24 UTC (rev 1294)
@@ -1,44 +1,34 @@
 IRC/neo-devel:
 - kick off bad block migration with all other phase -1 developers
-  [ done so far: mickey, thos, xora ]
+  [ done so far: mickey, thos, xora ] (obsolete ? just devirginate.)
 
 planning/design:
-- phase 0 default environment draft (neo-devel)
-- virgin system auto-install environment
+- phase 0 default environment draft (neo-devel) (obsolete ?)
 
 Mickey:
-- ask for quick walk-through review
+- ask for quick walk-through review (obsolete ?)
 
 coreteam:
 - send customs war story
 
 bugzilla:
   P0
-  - #79: suspend/resume to RAM support
-  - #178: u-boot 'factory reset' option
   P1
   - #164: improve non-SanDisk microSD support in u-boot 
-  - #179: Implement u-boot power-off timer
-  ??
-  - #202: Start using NAND hardware ECC support
 
 other:
-- test V-360 headset
 - set up automated build tests on buildhost-old
   - check: does default build yield old or new sjf ? if new, update
     Building_OpenMoko_from_scratch#Flash_boot_loader_into_NAND
 - once SVN opens, check that build passes without prompting for certificates or
   authentication. then update Building_OpenMoko_from_scratch#OpenEmbedded_build
-- local: make fs ("lab") boot into Gentoo directly
 
 u-boot:
-- set up openocd and familiarize with it (-> update wiki)
 - bbt migration & walk-through: future updates
   - update for debug v2
   - switch from wiggler to openocd
   - all boards will be v3 or later, so mention battery in
     Building_OpenMoko_from_scratch#Start_the_Neo_and_enter_the_boot_prompt
-- u-boot: fix weird "product name" after "mmc[init]"
 - write "intelligent" serial/USB console tool
 - u-boot: make fatload behave like ext2load (hex output, ${filesize})
 - idea: connect charger when off -> show battery gauge
@@ -46,76 +36,46 @@
   https://wiki.internal.openmoko.org/wiki/Bluetooth_MAC_Assignments
 - u-boot NAND update also removes environment offset. Documented it. Fix it,
   too ?
+  - also in DFU, it seems (fixed there !)
 - check that we don't have hidden power-on sequences (e.g., USB disconnect)
 - write requests for power measurements (TBD)
 - think about u-boot progress bar (e.g., time-driven)
-- review schematics before the end of Chinese new year holidays
 - verify/update
   https://wiki.openmoko.org/wiki/NAND_bad_blocks#JTAG_.2F_OpenOCD_.2F_u-boot_RAM_based
   https://wiki.openmoko.org/wiki/Bootloader#Using_JTAG_to_boot_from_RAM
 - find out why Harald's debug v2 is 6 times (!) faster than my JTAGkey
+- http://downloads.openmoko.org/snapshots/ (pester)
+- screenshots
+- compile BB wishlist (bb forget, ...)
 
---- Feb 13 --------------------------------------------------------------------
+Infrastructure:
+- get debug v2 to work
+- local: make fs ("lab") boot into Gentoo directly
 
-  - (#164) survey cards, then update Wiki
-- check (non-)use of BB_COLLECTIONS (do I have to care ?) [mickey says it's ok]
-- check walk-through exceptions [mickey says keep them]
-- put shipping instructions on Wiki
-- complete bad block migration (mickey)
+review:
+http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=231
+http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=237
 
---- Feb 14 --------------------------------------------------------------------
+- reopen #179 for console without menu
+- we need to check for an invalid block marker in the first _and_
+  +second page of each block
+  <LaF0rge> i think we only check in the first page
+- <LaF0rge> btw: some mid-term optimization: don't read the splash screen untile
+  +we've leaft board_late_init (or reached the end of it)
+  > right now, gcc doesn't know in bar.c what happens to var_set_by_isr, so it
+  +can't optimize. but when it knows what the caller does, whoopee !
+  <LaF0rge> if the power button gets pressed often accidentially (when in off
+  +mode), then we read the flash again and again and again
+block 0/OTP:
+- "sacred"
+- BT MAC
+- host key
+- as little code as possible
 
-- to whom do users report build failures ? [-> coreteam list]
-- need board hw revision identification procedure for "building from scratch"
-  [-> coreteam list]
-- change SVN/OE version for phase 0 ? (*must* have recent SVN for BBT)
-  (kernel compatibility)
-  [-> Mickey's fixing "now" and stuff ]
 
---- Feb 15 --------------------------------------------------------------------
-
-- try s3c2410 speed set 101 [my phone stil crashes]
-- complete walk-through update/corrections
-- bbt migration: describe how to undo bad "dynenv set" (already done ?)
-- bbt migration & walk-through: clone and merge
-  - nand flashing: do gradual erases (env, kernel, etc.), use partition names
-    and ${filesize}
-- buy and try battery
-
-(Not much happened for a while, 'cause some flu bug had other plans for me.)
-
---- Feb 20 --------------------------------------------------------------------
-
-  - #14: bootloader should display startup image before booting
-    (as early as possible, i.e., even before LCM is up)
-  - #40: test + debug display of image on LCM in u-boot 
-- GTA01Bv2 # dynenv set u-boot_env
-  45 4e 56 30 - 00 00 00 00
-  *OUCH* !
-
---- Feb 23 --------------------------------------------------------------------
-
-- prevent poweroff if starting from RAM
-  [ wakeup-reason-nand-only.patch, experimental, DOESN'T WORK ]
-- u-boot: start.S: boot from RAM with relocation
-  [ boot-from-ram-reloc.patch ]
-- nice to have: make u-boot detect at run-time if it's booting from RAM or NAND
-  [ boot-from-ram-and-nand.patch ]
-- board/neo1973/neo1973.c: start LCD first, to avoid white flash
-- add lowlevel.S to u-boot build
-
---- Feb 25 --------------------------------------------------------------------
-
-- force console switch for boot menu
-
---- Feb 26 --------------------------------------------------------------------
-
-boot menu finalizing:
-> ok, so i'll add some default menu items to my env draft (in preparation)
-> then i'll kill the "indefinite boot delay" and require a 20% or so presence
-+of AUX while power-up to enter the boot menu
-> last but not least, the first item, "Boot" gets a poweroff on failure, so
-+that desperate users don't get even more desperate when they can't seem to
-+turn off the device (well, most likely, they'll just pull the battery, but
-+still ...)
-
+* force non-SPI: nope
+* power cycle (SD_ON, not data): nyet
+* CMD8: rien
+* exact speed: nichts
+* manual bit-banging:
+* pessimistic speed: so-so (kingston, toshiba)

More information about the commitlog mailing list