This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: compilation time of block data


On Fri, Mar 31, 2006 at 11:39:07AM -0800, Bud Davis wrote:
> thanks steve, the e-mail kinda scrunched out the
> spaces.  a copy of a.f is attached.
> 
> 
> --bud

Spaces are still off-by-one.  I've modified your program
to accept size via a read(*,*) statement.  I've found

 1000   0.04
 2000   0.10
 3000   0.18
 4000   0.33
 5000   0.50
 6000   0.70
 7000   0.92
 8000   1.22
 9000   1.51
10000   1.97
11000   2.75
12000   4.1
13000   5.64
14000   7.1
20000  18.00

where the 2nd column is "user time" from the time(1) command.
The behavior looks to be quadratic (note there is an uncertainty
in the time measurement).  For the 14000 case, I find

/mnt1/sgk/work/4x/libexec/gcc/amd64-unknown-freebsd7.0/4.2.0/f951 bud2.f

Execution times (seconds)
 callgraph construction:   0.01 ( 0%) usr   0.00 ( 0%) sys   0.00 ( 0%) wall       0 kB ( 0%) ggc
 parser                :   6.97 (99%) usr   0.10 (87%) sys   7.12 (99%) wall    3108 kB (63%) ggc
 varconst              :   0.02 ( 0%) usr   0.00 ( 0%) sys   0.02 ( 0%) wall     729 kB (15%) ggc
 TOTAL                 :   7.01             0.12             7.17               4947 kB

-- 
Steve


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]