Chris Lattner dbab4dc942 implement a nice little efficiency hack in the inliner. Since we're now ... running IPSCCP early, and we run functionattrs interlaced with the inliner, we often (particularly for small or noop functions) completely propagate all of the information about a call to its call site in IPSSCP (making a call dead) and functionattrs is smart enough to realize that the function is readonly (because it is interlaced with inliner). To improve compile time and make the inliner threshold more accurate, realize that we don't have to inline dead readonly function calls. Instead, just delete the call. This happens all the time for C++ codes, here are some counters from opt/llvm-ld counting the number of times calls were deleted vs inlined on various apps: Tramp3d opt: 5033 inline - Number of call sites deleted, not inlined 24596 inline - Number of functions inlined llvm-ld: 667 inline - Number of functions deleted because all callers found 699 inline - Number of functions inlined 483.xalancbmk opt: 8096 inline - Number of call sites deleted, not inlined 62528 inline - Number of functions inlined llvm-ld: 217 inline - Number of allocas merged together 2158 inline - Number of functions inlined 471.omnetpp: 331 inline - Number of call sites deleted, not inlined 8981 inline - Number of functions inlined llvm-ld: 171 inline - Number of functions deleted because all callers found 629 inline - Number of functions inlined Deleting a call is much faster than inlining it, and is insensitive to the size of the callee. :) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@86975 91177308-0d34-0410-b5e6-96231b3b80d8		2009-11-12 07:56:08 +00:00
..
2003-09-14-InlineValue.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2003-09-22-PHINodeInlineFail.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2003-09-22-PHINodesInExceptionDest.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2003-09-22-PHINodesInNormalInvokeDest.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2003-10-13-AllocaDominanceProblem.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2003-10-26-InlineInvokeExceptionDestPhi.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2004-04-15-InlineDeletesCall.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2004-04-20-InlineLinkOnce.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2004-10-17-InlineFunctionWithoutReturn.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2006-01-14-CallGraphUpdate.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2006-07-12-InlinePruneCGUpdate.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2006-11-09-InlineCGUpdate-2.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2006-11-09-InlineCGUpdate.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2007-04-15-InlineEH.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2007-06-06-NoInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2007-06-25-WeakInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2007-12-19-InlineNoUnwind.ll	Eliminate more uses of llvm-as and llvm-dis.	2009-09-09 00:09:15 +00:00
2008-03-04-StructRet.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-03-07-Inline-2.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-03-07-Inline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-09-02-AlwaysInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-09-02-NoInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-10-30-AlwaysInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2008-11-04-AlwaysInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2009-01-08-NoInlineDynamicAlloca.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2009-01-12-RecursiveInline.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2009-01-13-RecursiveInlineCrash.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
2009-05-07-CallUsingSelfCrash.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
alloca_test.ll	Eliminate uses of %prcontext.	2009-09-05 11:35:16 +00:00
alloca-in-scc.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
always_inline_dyn_alloca.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
array_merge.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
basictest.ll	Make opt default to not adding a target data string and update tests that depend on target data to supply it within the test	2009-11-03 15:29:06 +00:00
byval2.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
byval.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
callgraph-update.ll	Make these tests more interesting by using	2009-10-24 23:23:04 +00:00
casts.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
cfg_preserve_test.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
crash.ll	Simplify some code (first hunk) and fix PR5208 (second hunk) by	2009-10-17 05:39:39 +00:00
delete-call.ll	implement a nice little efficiency hack in the inliner. Since we're now	2009-11-12 07:56:08 +00:00
dg.exp	sabre brings to my attention that the 'tr' suffix is also obsolete	2008-05-20 21:00:03 +00:00
dynamic_alloca_test.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
externally_available.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
indirect_resolve.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline_cleanup.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline_constprop.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline_dce.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline_prune.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline-invoke-tail.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
inline-tail.ll	Use opt -S instead of piping bitcode output through llvm-dis.	2009-09-08 22:34:10 +00:00
invoke_test-1.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
invoke_test-2.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
invoke_test-3.ll	Change tests from "opt %s" to "opt < %s" so that opt doesn't see the	2009-09-11 18:01:28 +00:00
nested-inline.ll	When considering whether to inline Callee into Caller,	2009-10-09 00:11:32 +00:00
PR4909.ll	Eliminate more redundant llvm-as calls.	2009-09-11 18:17:12 +00:00