Bug#833630: scalpel: performance decreases with running time

[ydirson at free.fr]

Package: scalpel
Version: 1.60-1
Severity: important

Running scalpel on a 1TB drive with 100 patterns of 1024 bytes each:
* ETA some time after starting stabilized around 11h (as I recall it, but this
  number is superfluous, the next ones talk loud enough)
* after 11h run it has done 63% and the ETA is still in 6h
* after 9h more it has done 68% and ETA is now up to 9h

That does not quite match everyone's idea of high performance, there has
to be a problem.