scan_one_extent: remove the unreadahead after benchmark results

That unreadahead used to result in a 10% hit on benchmarks.  Now it's
closer to 75%.

Signed-off-by: Zygo Blaxell <bees@furryterror.org>

diff --git a/src/bees-context.cc b/src/bees-context.cc
index e259b211693a76db1479f6aa26264e5b4ae5660a..194c124c1865ee4930da0ef505da971394efb8ed 100644 (file)
--- a/src/bees-context.cc
+++ b/src/bees-context.cc
@@ -839,8 +839,8 @@ BeesContext::scan_one_extent(const BeesFileRange &bfr, const Extent &e)
        }
        lock_seen.unlock();
 
-       // Still hurts benchmarks...or does it?
-       bees_unreadahead(bfr.fd(), bfr.begin(), bfr.size());
+       // Now causes 75% loss of performance in benchmarks
+       // bees_unreadahead(bfr.fd(), bfr.begin(), bfr.size());
 }
 
 shared_ptr<Exclusion>