git sources vs. cvs

[Dave Nomura]

I originally was working with a snapshot that contained your task finder 
stuff, but then thought I should build the patch from the latest 
source.  That's where I mistakenly went down the CVS path.

I haven't been able to get git to work but I'm going to try and get Jim 
Keniston to get it for me.
I tried building the snapshot from June 21 but it got build errors on my 
ppc machine:
tapsets.cxx: In member function 'void 
symbol_table::mark_dwarf_redundancies(dwfl
pp*)':
tapsets.cxx:4915: error: no matching function for call to 
'Internal::hashtable_i
terator<std::pair<const std::basic_string<char, std::char_traits<char>, 
std::all
ocator<char> >, cu_function_cache_t*>, false, false>::hashtable_iterator()'

The previous snapshot I was using was May10, and I had no problems 
building it.
David Smith wrote:
> Dave Nomura wrote:
>   
>> I was starting to re-build my patch for itrace and noticed that most of
>> the utrace machinery had been removed from tapsets.cxx.  Are there any
>> functional/interface changes that I should know about?  I patterned my
>> changes off of the utrace stuff.
>>     
>
> I don't think the utrace machinery has been removed from tapsets.cxx.
> It has morphed a little.
>
> From your other email about ChangeLogs, I'm wondering if you are still
> using cvs - we've switched to git.
>
> If I'm wrong about cvs vs. git being the problem here, ask a more
> specific question (like what exact utrace machinery is gone that you
> need) and I'll be happy to try and answer.
>
>   


-- 
Dave Nomura
LTC Linux Power Toolchain