| 1 |
On 2006-06-01 13:34+0000, Duncan wrote: |
| 2 |
> |
| 3 |
> BTW, the 4.1 series is IMO the biggest jump in gcc efficiency for amd64 |
| 4 |
> since the -march=k8 support got added in 3.3. |
| 5 |
|
| 6 |
Inspired by your comment, I installed 4.1.1 and did very un-scientific |
| 7 |
test: dcraw compiled [1] with gcc 3.4.5 and 4.1.1. Then convert one |
| 8 |
raw picture with it: |
| 9 |
|
| 10 |
time dcraw-3 -w test.CR2 |
| 11 |
real 0m10.338s |
| 12 |
user 0m9.969s |
| 13 |
sys 0m0.332s |
| 14 |
|
| 15 |
time dcraw-4 -w test.CR2 |
| 16 |
real 0m9.141s |
| 17 |
user 0m8.849s |
| 18 |
sys 0m0.292s |
| 19 |
|
| 20 |
This is pretty good, and that was only the dcraw, all libraries are |
| 21 |
still done by gcc 3.4.x. |
| 22 |
|
| 23 |
BR, Jani |
| 24 |
|
| 25 |
P.S. gcc -march=k8 -o dcraw -O3 dcraw.c -lm -ljpeg -llcms |
| 26 |
|
| 27 |
-- |
| 28 |
Jani Averbach |
| 29 |
-- |
| 30 |
gentoo-amd64@g.o mailing list |
Archives