On Mon, Oct 26, 2009 at 3:24 AM, David Brownell <[email protected]> wrote: > On Sunday 25 October 2009, Ųyvind Harboe wrote: >> > >> > Am I wrong in my assertion that these commands are ARM-specific and do >> > not belong in target.h? That seems like one reasonable objection for >> > holding off with integrating this series. >> >> Your assumption is correct, but it is unreasonable to hold of all >> work until the interface problem can be tackled. > > Not sure I agree about "unreasonable". It's pretty routine > to avoid merging incomplete patch series. *Especially* when > releases are overdue...
The series is complete at this point. What's lacking are other non-related issues to be solved. Meanwhile the code could be tested on hardware and old code could be retired. Once various other issues have been solved, then the code could take advantage of that. The re-testing phase of the mrc/mcr fn's in targets will take longer, and is not affected by, the interface fixes that will allow mrc/mcr to be moved. W.r.t. the release, then the danger is not of regression, but that undocumented commands, that probably work could be discovered by users. Alternatively it could be that we're committing completely new stuff, but I think as long as the "stuff" has very little chance of regression, then keeping all the code synced up as much as possible(wonderful as git is) has value. -- Øyvind Harboe http://www.zylin.com/zy1000.html ARM7 ARM9 ARM11 XScale Cortex JTAG debugger and flash programmer _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
