diff options
author | Paul Phillips <paulp@improving.org> | 2012-09-18 00:59:00 -0700 |
---|---|---|
committer | Paul Phillips <paulp@improving.org> | 2012-09-18 02:19:12 -0700 |
commit | 0e061f420f3b6e7447658f003d00f6108fa649e0 (patch) | |
tree | 0b8593545df932f08b051441112ce2ae2f6804e3 /src/library/scala/collection/SeqLike.scala | |
parent | 66603a2c003852d39faec20a9763fb0e25049cf4 (diff) | |
download | scala-0e061f420f3b6e7447658f003d00f6108fa649e0.tar.gz scala-0e061f420f3b6e7447658f003d00f6108fa649e0.tar.bz2 scala-0e061f420f3b6e7447658f003d00f6108fa649e0.zip |
Removed many @inline annotations and final modifiers.
It is my belief that these @inlines and finals landed between
unhelpful and harmful. I am sure this will be disputed in some
cases. It's too much and too difficult to measure except in the
aggregate unless we have specific @inline sites to discuss.
I don't know upon whom the burden of proof lies. I think we
should err on the side given here, since there is no evidence
of any consistent rationale being applied and it is easy to
verify the negative impact scala compiler inlining can have on
hotspot's far more sophisticated inlining.
Diffstat (limited to 'src/library/scala/collection/SeqLike.scala')
-rw-r--r-- | src/library/scala/collection/SeqLike.scala | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/library/scala/collection/SeqLike.scala b/src/library/scala/collection/SeqLike.scala index a3ff812024..cda8b1a0e4 100644 --- a/src/library/scala/collection/SeqLike.scala +++ b/src/library/scala/collection/SeqLike.scala @@ -735,8 +735,8 @@ object SeqLike { */ private def kmpSearch[B](S: Seq[B], m0: Int, m1: Int, W: Seq[B], n0: Int, n1: Int, forward: Boolean): Int = { // Check for redundant case when target has single valid element - @inline def clipR(x: Int, y: Int) = if (x<y) x else -1 - @inline def clipL(x: Int, y: Int) = if (x>y) x else -1 + def clipR(x: Int, y: Int) = if (x < y) x else -1 + def clipL(x: Int, y: Int) = if (x > y) x else -1 if (n1 == n0+1) { if (forward) |