search for: regmaskbit

...true)); else addFastRegAlloc(createRegAllocPass(false)); // Run post-ra passes. addPostRegAlloc(); // Adding a new pass here which keeps register mask information across function calls. . . . } But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) I think this also applies in someway to Mehdi Amini's idea to keep a ModulePass for bookkeeping b...

...e)); > > // Run post-ra passes. > addPostRegAlloc(); > // Adding a new pass here which keeps register mask information across > function calls. > . > . > . > } > > But this also requires current register allocators to use this information > in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static > across calls. I mean I am not clear for how to propagate static info to > Intra-procedural Register allocators (if possible without disturbing their > code ) > > First, my hope is that we won't need to change the register allocators, a...

...>> addPostRegAlloc(); >> // Adding a new pass here which keeps register mask information across >> function calls. >> . >> . >> . >> } >> >> But this also requires current register allocators to use this >> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >> not static across calls. I mean I am not clear for how to propagate static >> info to Intra-procedural Register allocators (if possible without >> disturbing their code ) >> >> First, my hope is that we won't need to change the re...

...post-ra passes. >> addPostRegAlloc(); >> // Adding a new pass here which keeps register mask information across function calls. >> . >> . >> . >> } >> >> But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) >> First, my hope is that we won't need to change the register allocators, as such, in order...

...t;> // Adding a new pass here which keeps register mask information across >>> function calls. >>> . >>> . >>> . >>> } >>> >>> But this also requires current register allocators to use this >>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>> not static across calls. I mean I am not clear for how to propagate static >>> info to Intra-procedural Register allocators (if possible without >>> disturbing their code ) >>> >>> First, my hope is that we won't n...

...gt; > > > > > > > > > > > > > > > > > > > > > > > > > But this also requires current register allocators to use > > > > > > > this > > > > > > > information in someway because RegMaskBits in > > > > > > > LiveIntervalAnalysis.cpp is not static across calls. I > > > > > > > mean > > > > > > > I > > > > > > > am > > > > > > > not > > > > > > > clear for how to...

...which keeps register mask information across >>>> function calls. >>>> . >>>> . >>>> . >>>> } >>>> >>>> But this also requires current register allocators to use this >>>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>>> not static across calls. I mean I am not clear for how to propagate static >>>> info to Intra-procedural Register allocators (if possible without >>>> disturbing their code ) >>>> >>>> First, my hope is...

...; > > > > > > > > > > > > > But this also requires current register allocators to > > > > > > > > > use > > > > > > > > > this > > > > > > > > > information in someway because RegMaskBits in > > > > > > > > > LiveIntervalAnalysis.cpp is not static across calls. > > > > > > > > > I > > > > > > > > > mean > > > > > > > > > I > > > > > > > > > am >...

...mation across >>>>> function calls. >>>>> . >>>>> . >>>>> . >>>>> } >>>>> >>>>> But this also requires current register allocators to use this >>>>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>>>> not static across calls. I mean I am not clear for how to propagate static >>>>> info to Intra-procedural Register allocators (if possible without >>>>> disturbing their code ) >>>>> >>>>&g...

...across function calls. >>>>>> . >>>>>> . >>>>>> . >>>>>> } >>>>>> >>>>>> But this also requires current register allocators to use this >>>>>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>>>>> not static across calls. I mean I am not clear for how to propagate static >>>>>> info to Intra-procedural Register allocators (if possible without >>>>>> disturbing their code ) >>>>>> &g...

...ass(false)); > > // Run post-ra passes. > addPostRegAlloc(); > // Adding a new pass here which keeps register mask information across function calls. > . > . > . > } > > But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) > First, my hope is that we won't need to change the register allocators, as such, in order to...

...post-ra passes. >> addPostRegAlloc(); >> // Adding a new pass here which keeps register mask information across function calls. >> . >> . >> . >> } >> >> But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) >> First, my hope is that we won't need to change the register allocators, as such, in order...

...gt;>>>>> . >>>>>>> . >>>>>>> . >>>>>>> } >>>>>>> >>>>>>> But this also requires current register allocators to use this >>>>>>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>>>>>> not static across calls. I mean I am not clear for how to propagate static >>>>>>> info to Intra-procedural Register allocators (if possible without >>>>>>> disturbing their code ) >>>&gt...

...t;>>>> . >>>>>>>>>>> . >>>>>>>>>>> } >>>>>>>>>>> >>>>>>>>>>> But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) >>>>>>>>>>> First, my hope is that we won't need to change the re...

.... >>>>>>>> . >>>>>>>> . >>>>>>>> } >>>>>>>> >>>>>>>> But this also requires current register allocators to use this >>>>>>>> information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is >>>>>>>> not static across calls. I mean I am not clear for how to propagate static >>>>>>>> info to Intra-procedural Register allocators (if possible without >>>>>>>> disturbing their code ) &gt...

...ass(false)); > > // Run post-ra passes. > addPostRegAlloc(); > // Adding a new pass here which keeps register mask information across function calls. > . > . > . > } > > But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) > First, my hope is that we won't need to change the register allocators, as such, in order to...

...ass(false)); > > // Run post-ra passes. > addPostRegAlloc(); > // Adding a new pass here which keeps register mask information across function calls. > . > . > . > } > > But this also requires current register allocators to use this information in someway because RegMaskBits in LiveIntervalAnalysis.cpp is not static across calls. I mean I am not clear for how to propagate static info to Intra-procedural Register allocators (if possible without disturbing their code ) > First, my hope is that we won't need to change the register allocators, as such, in order to...

...> use > > > > > > > > > > > > > > > > > this > > > > > > > > > > > > > > > > > information in someway because > > > > > > > > > > > > > > > > > RegMaskBits > > > > > > > > > > > > > > > > > in > > > > > > > > > > > > > > > > > LiveIntervalAnalysis.cpp is not > > > > > > > > > > > > > > > > > static &...

...; > > > > > > } > > > > > > > > > > > > > > > > > > > > > But this also requires current register allocators to use > > > > > > this > > > > > > information in someway because RegMaskBits in > > > > > > LiveIntervalAnalysis.cpp is not static across calls. I mean > > > > > > I > > > > > > am > > > > > > not > > > > > > clear for how to propagate static info to Intra-procedural > > > &g...

> On May 10, 2016, at 6:06 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > > > From: "vivek pandya" <vivekvpandya at gmail.com> > To: "llvm-dev" <llvm-dev at lists.llvm.org>, "Tim Amini Golling" <mehdi.amini at apple.com>, "Hal Finkel" <hfinkel at anl.gov> > Cc: "Quentin Colombet" <qcolombet