search for: lhs_alloca

Displaying 4 results from an estimated 4 matches for "lhs_alloca".

Did you mean: rhs_alloca
2013 Jan 15
2
[LLVMdev] [cfe-dev] no-alias generated as result of restrict function arguments
...Ps and determine the object types. + LHSPtr = LHSPtr->stripInBoundsOffsets(); + RHSPtr = RHSPtr->stripInBoundsOffsets(); + // For allocas and arguments, remember the function they belong to. + const Function *lhs_func = 0; + const Function *rhs_func = 0; + + bool lhs_alloca = isa<AllocaInst>(LHSPtr); + bool rhs_alloca = isa<AllocaInst>(RHSPtr); + bool lhs_global = isa<GlobalValue>(LHSPtr); + bool rhs_global = isa<GlobalValue>(RHSPtr); + + bool lhs_noaliasarg = false; + bool rhs_noaliasarg = false; + bool lhs_byvala...
2013 Jan 16
0
[LLVMdev] [cfe-dev] no-alias generated as result of restrict function arguments
...tr)) { - // If both sides are different identified objects, they aren't equal - // unless they're null. - if (LHSPtr != RHSPtr && llvm::isIdentifiedObject(RHSPtr) && - Pred == CmpInst::ICMP_EQ) - return ConstantInt::get(ITy, false); + bool lhs_alloca = isa<AllocaInst>(LHSPtr); + bool rhs_alloca = isa<AllocaInst>(RHSPtr); + bool lhs_global = isa<GlobalValue>(LHSPtr); + bool rhs_global = isa<GlobalValue>(RHSPtr); - // A local identified object (alloca or noalias call) can't equal any - // inc...
2012 Dec 12
0
[LLVMdev] [cfe-dev] no-alias generated as result of restrict function arguments
On Wed, Dec 12, 2012 at 1:26 PM, Joerg Sonnenberger <joerg at britannica.bec.de> wrote: > On Wed, Dec 12, 2012 at 11:01:01AM -0800, Dan Gohman wrote: >> > Is that >> > assumption violated if I explicitly cast away const and pass the result >> > to a function with NoAlias argument? >> >> Not immediately, no. It means that you can't access the
2012 Dec 12
3
[LLVMdev] [cfe-dev] no-alias generated as result of restrict function arguments
On Wed, Dec 12, 2012 at 11:01:01AM -0800, Dan Gohman wrote: > > Is that > > assumption violated if I explicitly cast away const and pass the result > > to a function with NoAlias argument? > > Not immediately, no. It means that you can't access the constant > pointer's pointee directly within the noalias argument's scope. Access > to that object must go