llvm-6502/test/Transforms/Inline
Gerolf Hoflehner 049a087d3f Suppress inlining when the block address is taken
Inlining functions with block addresses can cause many problem and requires a
rich infrastructure to support including escape analysis.  At this point the
safest approach to address these problems is by blocking inlining from
happening.

Background:
There have been reports on Ruby segmentation faults triggered by inlining
functions with block addresses like

//Ruby code snippet
vm_exec_core() {
    finish_insn_seq_0 = &&INSN_LABEL_finish;
    INSN_LABEL_finish:
      ;
}

This kind of scenario can also happen when LLVM picks a subset of blocks for
inlining, which is the case with the actual code in the Ruby environment.

LLVM suppresses inlining for such functions when there is an indirect branch.
The attached patch does so even when there is no indirect branch.  Note that
user code like above would not make much sense: using the global for jumping
across function boundaries would be illegal.

Why was there a segfault:

In the snipped above the block with the label is recognized as dead So it is
eliminated. Instead of a block address the cloner stores a constant (sic!) into
the global resulting in the segfault (when the global is used in a goto).

Why had it worked in the past then:

By luck. In older versions vm_exec_core was also inlined but the label address
used was the block label address in vm_exec_core.  So the global jump ended up
in the original function rather than in the caller which accidentally happened
to work.

Test case ./tools/clang/test/CodeGen/indirect-goto.c will fail as a result
of this commit.

rdar://17245966



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212077 91177308-0d34-0410-b5e6-96231b3b80d8
2014-07-01 00:19:34 +00:00
..
2003-09-14-InlineValue.ll
2003-09-22-PHINodeInlineFail.ll
2003-09-22-PHINodesInExceptionDest.ll
2003-09-22-PHINodesInNormalInvokeDest.ll
2003-10-13-AllocaDominanceProblem.ll
2004-04-15-InlineDeletesCall.ll
2004-04-20-InlineLinkOnce.ll
2004-10-17-InlineFunctionWithoutReturn.ll
2006-01-14-CallGraphUpdate.ll
2006-07-12-InlinePruneCGUpdate.ll
2006-11-09-InlineCGUpdate-2.ll
2006-11-09-InlineCGUpdate.ll
2007-04-15-InlineEH.ll
2007-06-25-WeakInline.ll
2007-12-19-InlineNoUnwind.ll
2008-09-02-NoInline.ll
2009-01-08-NoInlineDynamicAlloca.ll
2009-01-13-RecursiveInlineCrash.ll
2009-05-07-CallUsingSelfCrash.ll
2010-05-12-ValueMap.ll
alloca_test.ll Tests: rewrite 'opt ... %s' to 'opt ... < %s' so that opt does not emit a ModuleID 2012-12-30 01:28:40 +00:00
alloca-bonus.ll
alloca-in-scc.ll
alloca-merge-align-nodl.ll
alloca-merge-align.ll
always-inline.ll Check the alwaysinline attribute on the call as well as on the caller. 2014-05-19 18:25:54 +00:00
array_merge.ll
attributes.ll
basictest.ll
blockaddress.ll Suppress inlining when the block address is taken 2014-07-01 00:19:34 +00:00
byval_lifetime.ll Add lifetime markers for allocas created to hold byval arguments, make them 2014-04-15 18:06:46 +00:00
byval-tail-call.ll Fix PR7272 in -tailcallelim instead of the inliner 2014-04-21 20:48:47 +00:00
byval.ll Revert "Inliner: Handle readonly attribute per argument when adding memcpy" 2013-11-03 12:22:13 +00:00
callgraph-update.ll
casts.ll
cfg_preserve_test.ll
crash2.ll
crash.ll
debug-invoke.ll DebugInfo: Preserve debug location information when transforming a call into an invoke during inlining. 2014-06-30 20:30:39 +00:00
delete-call.ll
devirtualize-2.ll
devirtualize-3.ll
devirtualize.ll
dynamic_alloca_test.ll
externally_available.ll
gvn-inline-iteration.ll
ignore-debug-info.ll Revert test commit. Removed blank line. 2014-05-08 12:54:43 +00:00
inline_cleanup.ll
inline_constprop.ll
inline_dce.ll
inline_invoke.ll Fix inlining to not produce duplicate landingpad clauses 2013-12-08 00:50:58 +00:00
inline_minisize.ll
inline_prune.ll
inline_returns_twice.ll [inliner] Fix PR18206 by preventing inlining functions that call setjmp 2013-12-13 08:00:01 +00:00
inline_ssp.ll
inline-byval-bonus.ll
inline-cold.ll [inline cold threshold] Command line argument for inline threshold will 2014-04-25 17:34:55 +00:00
inline-invoke-tail.ll
inline-invoke-with-asm-call.ll Rename testing case to use - instead of _. 2013-11-04 18:52:06 +00:00
inline-optnone.ll The 'optnone' attribute means don't inline anything into this function 2013-11-18 21:44:03 +00:00
inline-optsize.ll
inline-tail.ll Don't insert lifetime.end markers between a musttail call and ret 2014-05-15 21:10:46 +00:00
inline-vla.ll Handle vlas during inline cost computation if they'll be turned 2014-04-07 13:36:21 +00:00
invoke_test-1.ll
invoke_test-2.ll
invoke_test-3.ll
invoke-cleanup.ll Fix inlining to not lose the "cleanup" clause from landingpads 2013-12-08 00:51:21 +00:00
invoke-combine-clauses.ll Fix inlining to not produce duplicate landingpad clauses 2013-12-08 00:50:58 +00:00
invoke-cost.ll [inliner] Completely change (and fix) how the inline cost analysis 2013-12-13 07:59:56 +00:00
lifetime-no-datalayout.ll
lifetime.ll
nested-inline.ll
noinline-recursive-fn.ll
noinline.ll
optimization-remarks.ll Add support for missed and analysis optimization remarks. 2014-05-22 14:19:46 +00:00
PR4909.ll
ptr-diff.ll Fix known typos 2014-01-24 17:20:08 +00:00
recursive.ll
switch.ll [inliner] Significantly improve the compile time in cases like PR19499 2014-04-28 08:52:44 +00:00