summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJarkko Hietaniemi <jhi@cc.hut.fi>1997-01-30 21:58:10 +0200
committerChip Salzenberg <chip@atlantic.net>1997-02-04 17:47:00 +1200
commitf2c6908735e0c2f2afe8eb6d87788a1d29d32bfb (patch)
treea763f4c374389c19e37ede04e77fbdbb6ef932e5
parent062dca251b623b6c7d3b729b1ee411d62a0ab399 (diff)
downloadperl-f2c6908735e0c2f2afe8eb6d87788a1d29d32bfb.tar.gz
hints/dec_osf.sh: polishing the comments
private-msgid: <199701301958.VAA08992@alpha.hut.fi>
-rw-r--r--hints/dec_osf.sh22
1 files changed, 20 insertions, 2 deletions
diff --git a/hints/dec_osf.sh b/hints/dec_osf.sh
index efe2c8a65b..f0b380bf8c 100644
--- a/hints/dec_osf.sh
+++ b/hints/dec_osf.sh
@@ -22,8 +22,17 @@
# executable code simply do not agree any more while in
# mid-execution, the optimiser only cares about the results.
#
-# Note: Configure will automatically
-# add the often quoted -DDEBUGGING for you)
+# Note 3: Configure will automatically add the often quoted
+# -DDEBUGGING for you if the -g is specified.
+#
+# Note 4: There is even more optimisation available in the new
+# (GEM) DEC cc: -O5 and -fast. "man cc" will tell more about them.
+# The jury is still out whether either or neither help for Perl
+# and how much. Based on very quick testing, -fast boosts
+# raw data copy by about 5-15% (-fast brings in, among other
+# things, inlined, ahem, fast memcpy()), while on the other
+# hand searching things (index, m//, s///), seems to get slower.
+# Your mileage will vary.
#
case "$optimize" in
@@ -84,6 +93,15 @@ esac
#
# History:
#
+# perl5.003_24:
+#
+# 30-Jan-1997 Jarkko Hietaniemi <jhi@iki.fi>
+#
+# * Fixing the note on -DDEBUGGING.
+#
+# * Note on -O5 -fast.
+#
+#
# perl5.003_23:
#
# 26-Jan-1997 Jarkko Hietaniemi <jhi@iki.fi>