First of all, screw you for all of these patches. Frekin' 80+ emails! Dang :). Anyway, what's an efficient way for me to try and port my AT91SAM9 NAND flash over to this new interface? I'm thinking I create a branch for your changes, pull your changes against current orign/master, and merge in my three branches that have that at91sam9 driver and the core updates that it depends on. I'm not sure on what the commands should be, so any direction would be appreciated. :)
// Dean Glazeski On Tue, Nov 24, 2009 at 4:02 PM, Zach Welch <z...@superlucidity.net> wrote: > Hi all, > > I have finished my series to convert the tree to use the improved > command registration methodology. I apologize for the number of patches > in these series; however, only a handful of patches are bigger than 10K, > and most are just boring semi-mechanical transformations. > > The following benefits can be expected from this work: > - Command registration is much simpler and easier to read and update. > - All registrations now get proper error checking. > - Eliminates redundant calls to register command: smaller binaries. > - Improves 'help' display and add new 'usage' command. > - Changes all register_commands callbacks into command_registration *'s. > > To facilitate fast and painless review, the patches have been grouped > into sequential series. These will be posted shortly under this summary > message. They will contain the following sets of changes: > > 1) Core upgrades: > - updates the core command registration routines, > - provides the new facilities used in the subsequent series, > - convert registration to use struct command_registration arrays > - allows batch registration, better error checking, and more > - extended to allow command usage information to be provided > - rewrites the 'unknown' function in C > - provides proper support for calling N-deep nested commands, > - removes the magic to try "blah blah" as "blah_blah" > > 2) Initial conversions: > - converts helper and top-level modules to use new registration, > - demonstrates command chaining possibilities in hello.c, > - updates server and pld modules, > - removes register_callbacks from pld_device, > > 3) Update JTAG module: > - convert each module to use new registration, > - remove jtag_interface callback, > > 4) Update Flash module: > - convert each moduleto use new registration, > - remove flash_driver and nand_controller callbacks, > > 5) Update Target module: > - convert each moduleto use new registration, > - remove target_type callbacks, > - remove etm_capture_driver callbacks, > > 6) Updates made possible/easier after series 1-5: > - integrate Jim command handlers with command_registration records > - add support for scripts to update command usage information > - improve formatting display help output by rewriting cmd_help in C > - moves 'hello' and 'foo' commands to dummy and faux drivers, > - add a new 'testee' target driver, which passes 'init' okay, > - cleanup log initialization; support OPENOCD_DEBUG_LEVEL env var, > > Without feedback that holds me back, I will push these when I see fit, > so please consider pulling this tree and testing it for regressions: > > Fetch URL: git://repo.or.cz/openocd/ztw.git > > Let me know what I have missed, or find out when it hits the tree.... > > Cheers, > > Zach > _______________________________________________ > Openocd-development mailing list > Openocd-development@lists.berlios.de > https://lists.berlios.de/mailman/listinfo/openocd-development >
_______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
