possible buglet in format

[Tobias Burnus]

Hi Dominique and FX,

FX Coudert wrote:
>> Test1 - Gauss 2000 (101x101) inverts 5.0 sec Err= 0.000000000000004
>> Test2 - Crout 2000 (101x101) inverts 8.0 sec Err= 0.000000000000028
>> Test3 - Crout 2 (1001x1001) inverts 16.0 sec Err= 0.000000000000032
>> Test4 - Lapack 2 (1001x1001) inverts 9.0 sec Err= 0.000000000000266
>> total = 38.0 sec
>>
>> i.e., always with a 0 decimal part.
> I don't understand what you call a 0 decimal part? Do you mean
> "fractional part"? If it's the case, then I don't understand either
> because it's not zero. What is the expected output?
I agree that this is quite cryptic; meant is that the time in seconds is
always ".0". But the reason is not the format statement but that
system_clock always returns a rate 1.

The reason is that HAVE_GETTIMEOFDAY is not defined.

http://gcc.gnu.org/viewcvs/trunk/libgfortran/configure.ac shows for the
backtrace patch:

-AC_CHECK_FUNCS(wait setmode getrlimit gettimeofday)
+AC_CHECK_FUNCS(wait setmode execvp pipe dup2 close fdopen strcasestr getrlimit)
+
+# Check for glibc backtrace functions
+AC_CHECK_FUNCS(backtrace backtrace_symbols)


Tobias

PS: Thanks, Dominique for finding this bug!