Rules that perform index->fixnum conversions must use arithmetic left
authorChris Hanson <org/chris-hanson/cph>
Wed, 15 Nov 1989 02:40:21 +0000 (02:40 +0000)
committerChris Hanson <org/chris-hanson/cph>
Wed, 15 Nov 1989 02:40:21 +0000 (02:40 +0000)
commit7498e41630b0ea6b6802ffc3922e0a082a8d9a2d
tree7accc3c69bac05c1c22fea4ab5862fa65cc3807e
parent4f8443e4b9a127006432ff8cc2a0d1e618ea19b8
Rules that perform index->fixnum conversions must use arithmetic left
shift so that proper testing for overflow can occur.  The reason:
these rules also get invoked by generic arithmetic because the code
compressor sees them and combines instruction pairs to use them.
v7/src/compiler/machines/bobcat/rules1.scm