g++ 3.4.1

[Reini Urban]

Gerrit P. Haase schrieb:
> Gerrit P. Haase wrote:
>> No, it will be enabled, or better sjlj will not be enabled and the build
>> defaults to dwarf2.
> 
> I was wrong with this statement.
> 
> To clarify this:
> gcc-3.4 has DWARF2 debug informations enabled, but exception handling 
> will stay to be SjLj exceptions and not be DWARF2, because:
> 
> - there are w32api callback problems
> - there is a generic problem with stdcall fuctions and some combinations
>   of -fomit-frame-pointer/-mtune/optimisation levels
> - it is a major ABI change

So the octave folks will need to build their private dwarf2 exception 
gcc until this is fixed. Or stay with the old versions.

Wonder how this will affect postgresql.
I'll report timings, when the new beta5 is out, 1-3 days.
With the older gcc and postgresql I got 50 concurrent connections, with 
latest versions only max 2, which was way too slow. But it could have 
been some other concurrency problem also.
-- 
Reini Urban
http://xarch.tu-graz.ac.at/home/rurban/

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/