Sanjiv Gupta
2010-Feb-11 05:17 UTC
[LLVMdev] adding switches to llvm-ld to disable certain optimizations.
Dan Gohman wrote:> Have you ever investigated the following approach? Define fake > register+register forms of common instructions, in addition to the > register+memory forms. Let the instruction selector work as if > everything were in registers. Then, since there's only one physical > register, the register allocator will have to spill, and the spills > and reloads can be folded in, eliminating the take register+register > forms. You might need special handling for the case where both > operands are the same. > > If this works well enough, it would allow your target to be less > strange from LLVM's perspective. Fewer things would need to be > Custom-expanded (e.g. ADD), and it may even allow you to actually > run more of the optimizer (since without mem2reg, much of the > optimizer is effectively disabled). > > Dan > >I remember that you had suggested this in one of earlier emails as well, which I lost. And I was desperately searching for that email. Glad that you put up it again. The approach actually sounds better as it will drastically simplify the back-end code. But I was clueless as to how to make register allocator fold the spills and reloads into the actual target instructions. The only interfaces that it exposes are saveRegToStackSlot and loadRegFromStackSlot, and we didn't even know for which instructions these spills are reloads are happening. All these APIs get is a frameIndex. Now that you have decided to get us to explore a better path, it would be good if you could put more light to these issues. One more thing that I feel will simplify things in a great sense is to make i16 legal (as it would make the pointer legal) and there onwards lower the types/operations ourselves to 8-bit (as type legalizer wouldn't do that). By doing that we would pretty much need to duplicate the legalizer code in our back-end as the TypeLegalizer interfaces currently are not exposed to TargetLowering. Or can a back-end just create an instance of Type Legalizer and use it? Thanks, - Sanjiv
Dan Gohman
2010-Feb-11 20:02 UTC
[LLVMdev] adding switches to llvm-ld to disable certain optimizations.
On Feb 10, 2010, at 9:17 PM, Sanjiv Gupta wrote:> Dan Gohman wrote: >> Have you ever investigated the following approach? Define fake >> register+register forms of common instructions, in addition to the >> register+memory forms. Let the instruction selector work as if >> everything were in registers. Then, since there's only one physical >> register, the register allocator will have to spill, and the spills >> and reloads can be folded in, eliminating the take register+register >> forms. You might need special handling for the case where both >> operands are the same. >> >> If this works well enough, it would allow your target to be less >> strange from LLVM's perspective. Fewer things would need to be >> Custom-expanded (e.g. ADD), and it may even allow you to actually >> run more of the optimizer (since without mem2reg, much of the >> optimizer is effectively disabled). >> >> Dan >> >> > I remember that you had suggested this in one of earlier emails as well, which I lost. And I was desperately searching for that email. Glad that you put up it again. > The approach actually sounds better as it will drastically simplify the back-end code. But I was clueless as to how to make register allocator fold the spills and reloads into the actual target instructions. The only interfaces that it exposes are saveRegToStackSlot and loadRegFromStackSlot, and we didn't even know for which instructions these spills are reloads are happening. All these APIs get is a frameIndex. > Now that you have decided to get us to explore a better path, it would be good if you could put more light to these issues.The main API hooks here are TargetInstrInfo::foldMemoryOperandImpl; there's a FrameIndex form and a generic load form. To be sure, I don't know if this kind of approach will work well. But if it does, it could help make PIC16 less different from other targets in LLVM.> > One more thing that I feel will simplify things in a great sense is to make i16 legal (as it would make the pointer legal) and there onwards lower the types/operations ourselves to 8-bit (as type legalizer wouldn't do that). By doing that we would pretty much need to duplicate the legalizer code in our back-end as the TypeLegalizer interfaces currently are not exposed to TargetLowering. Or can a back-end just create an instance of Type Legalizer and use it?I don't have anything to suggest here. Dan
Sanjiv.Gupta at microchip.com
2010-Feb-12 06:41 UTC
[LLVMdev] adding switches to llvm-ld to disable certain optimizations.
>> >> One more thing that I feel will simplify things in a great sense is to make i16 legal (as it would make the pointer legal) and there onwards lower the types/operations ourselves to 8-bit (as type legalizer wouldn't do that). By doing that we would pretty much need to duplicate the legalizer code in our back-end as the TypeLegalizer interfaces currently are not exposed to TargetLowering. Or can a back-end just create an instance of Type Legalizer and use it?>I don't have anything to suggest here.>DanDuncan, Your two cents needed here. - Sanjiv -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20100211/519824d1/attachment.html>
Alireza.Moshtaghi at microchip.com
2010-Feb-12 18:24 UTC
[LLVMdev] adding switches to llvm-ld to disable certainoptimizations.
I don't think our problem is in the way that we define our instructions nor even will it be resolved by removing Mem2Reg optimizations. As Dan says, Mem2Reg is the prerequisite for so many other optimizations that we can't afford to loose it; in fact removing Mem2Reg helps in some cases, but in few cases even increases the code size. I think the answer is in the scheduler. Currently the LLVM scheduler tries to reduce the register pressure on the aggregate of operations in one basic block and leaves the rest to the register allocator to do it magic (at least that is how I understand it); however, for an 8-bit device with only one register, there isn't much that the register allocator can do, hence increasing the number of spills. What I think we should do is to add a new scheduling mode where the scheduler tries to keep all operations on one dataflow path together; kind of like what one would do for a stack based machine. Now this stack-based scheduler mode is what I've been thinking of adding, but I need more clues into the how-to of it and what it will affect as far as other pieces of LLVM. Any kind of input with this regard is appreciated. Thanks A.> Dan Gohman wrote: >> Have you ever investigated the following approach? Define fake >> register+register forms of common instructions, in addition to the >> register+memory forms. Let the instruction selector work as if >> everything were in registers. Then, since there's only one physical >> register, the register allocator will have to spill, and the spills >> and reloads can be folded in, eliminating the take register+register >> forms. You might need special handling for the case where both >> operands are the same. >> >> If this works well enough, it would allow your target to be less >> strange from LLVM's perspective. Fewer things would need to be >> Custom-expanded (e.g. ADD), and it may even allow you to actually >> run more of the optimizer (since without mem2reg, much of the >> optimizer is effectively disabled). >> >> Dan >> >> > I remember that you had suggested this in one of earlier emails aswell, which I lost. And I was desperately searching for that email. Glad that you put up it again.> The approach actually sounds better as it will drastically simplifythe back-end code. But I was clueless as to how to make register allocator fold the spills and reloads into the actual target instructions. The only interfaces that it exposes are saveRegToStackSlot and loadRegFromStackSlot, and we didn't even know for which instructions these spills are reloads are happening. All these APIs get is a frameIndex.> Now that you have decided to get us to explore a better path, it wouldbe good if you could put more light to these issues. The main API hooks here are TargetInstrInfo::foldMemoryOperandImpl; there's a FrameIndex form and a generic load form. To be sure, I don't know if this kind of approach will work well. But if it does, it could help make PIC16 less different from other targets in LLVM.> > One more thing that I feel will simplify things in a great sense is tomake i16 legal (as it would make the pointer legal) and there onwards lower the types/operations ourselves to 8-bit (as type legalizer wouldn't do that). By doing that we would pretty much need to duplicate the legalizer code in our back-end as the TypeLegalizer interfaces currently are not exposed to TargetLowering. Or can a back-end just create an instance of Type Legalizer and use it? I don't have anything to suggest here. Dan _______________________________________________ LLVM Developers mailing list LLVMdev at cs.uiuc.edu http://llvm.cs.uiuc.edu http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
Apparently Analagous Threads
- [LLVMdev] adding switches to llvm-ld to disable certain optimizations.
- [LLVMdev] adding switches to llvm-ld to disable certain optimizations.
- [LLVMdev] adding switches to llvm-ld to disable certain optimizations.
- [LLVMdev] adding switches to llvm-ld to disable certain optimizations.
- [LLVMdev] adding switches to llvm-ld to disable certain optimizations.