Bugzilla – Bug 66
[scalarrepl] Scalar Replacement of aggregates is decimating structures it shouldn't be
Last modified: 2003-10-29 11:56:24
You need to log in before you can comment on or make changes to this bug.
The following testprogram is being misoptimized by the scalarrepl pass: http://mail.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20031027/008945.html Scalar repl is decimating the structure in main, despite the fact that its address is used and passed into a function. Thanks to Bill Wending for reducing this from a monster HDF5 testcase.
This is now fixed: http://mail.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20031027/008948.html New testcase here: ScalarRepl/2003-10-29-ArrayProblem.ll