[Crash-utility] Crash Subscription - Something Wrong

[Piet Delaney]

Dave Anderson wrote:
> Piet Delaney wrote:
>> Hi Dave [I moved]:
>>
>> I subscribed to your crash mailing list last evening and didn't receive 
>> anything yet.
>> Looks like there was a posting today.
> 
> I remember approving your subscription request.  Looking at the current
> list of subscribers, I see these two:
> 
>    Piet.Delaney at tensilica.com
>    piet at bluelane.com
> 
>> How much work do you think it would take to add the xtensa architecture 
>> to crash?
>>  From our IA64 work together I recall it mostly being the backtrace code 
>> needing work.
> 
> That's certainly a big part of it, but that's only the "gravy" that needs
> to be done after all of the foundational work is done.  New architectures
> basically require:
> 
> 1. cloning the arch-specific parts of configure.c so that there
>     can be "#ifdef XTENSA" parts of the code.
> 2. creating a new "xtensa.c" file that satisfies all of the
>     requirements for the generic machdep-><architecture-data> and
>     machdep-><architecture-function> fields used throughout the
>     common code.  The crash sources do contain a template PLATFORM.c
>     file, which is a bit dated and will most likely not have everything
>     needed, but it's a good place to start.  You would basically copy
>     PLATFORM.c to xtensa.c, and then do a "s/PLATFORM/xtensa/" on the
>     file, and then start working on the specifics.

I'll start work on xtensa.c next now that configure.c seems to be fine.

> 3. in conjunction with step 2, satisfying all of the "common" #define's
>     in defs.h.
> 4. After 1, 2 and 3 are done, update the Makefile to compile the
>     new file, and as it continually fails due to missing stuff, fix then
>     one step at a time.
> 5. At that point, you can start working on stuff like the backtrace
>     implementation.
> 
> Although -- now that I think of it -- none of the above will be worth
> a damn without gdb support of the architecture.  And gdb-6.1 certainly
> doesn't have it unless it's a clone of some pre-existing architecture.

With your new crash 5.0 and it's use of gdb-8.0 that's no longer a problem.

> 
> So that's a whole other can of worms.  Updating the version of gdb to
> "merge" with the crash sources is yet another massive, always prone-to-error,
> always finding-something-that-no-long-applies-anymore, situation.

Besides my maintaining a subscription to your mailing list I've created a crash-xtensa
mailing list at linux-xtensa.org and an associated crash-xtensa-commits for logging the
git repository changes. Your welcome to be a list administrators if your interested to to
recommend when to push changes/patches to your repository.

Should be interesting but likely a rather slow effort; lots of other stuff on my plate.

-piet

> 
> Dave
> 
>> Some the the 'vm' command functionality would be helpful right now to debug
>> a cache/tldb bug that I'm looking at.
>>
>> -piet
>