Re: [GHC] #11292: Generics unboxed types: TEST=GEq1 WAY=optasm is failing

#11292: Generics unboxed types: TEST=GEq1 WAY=optasm is failing -------------------------------------+------------------------------------- Reporter: thomie | Owner: Type: bug | Status: patch Priority: normal | Milestone: Component: Compiler | Version: 7.11 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: #11312 | Differential Rev(s): Phab:D1714 Wiki Page: | -------------------------------------+------------------------------------- Comment (by Ben Gamari ): In [changeset:"8e735fd0f88454b74fbb866a59b608925a2b3e48/ghc" 8e735fd0/ghc]: {{{ #!CommitTicketReference repository="ghc" revision="8e735fd0f88454b74fbb866a59b608925a2b3e48" Fix GEq1 when optimizations are enabled When optimizations are enabled, primitive string literals can be inlined, which can create two copies of a string constant with different addresses. We want to avoid this behavior at all costs in the `GEq1` test, since the output depends on the result of `eqAddr#`. We prevent such inlining through use of the `{-# NOINLINE #-}` pragma. Fixes #11292. Test Plan: Validate with T11292 Reviewers: thomie, austin, bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D1714 GHC Trac Issues: #11292 }}} -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/11292#comment:4 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler