search for: lowermultidimrefs

...but it can happen with any instructions (including the use of large integer >> (or any FP) constants on RISC machines, addressing globals with PIC code, >> handling extended/truncated operations (for RISC machines with a single >> integer size), etc. Note that hacks like "LowerMultiDimRefs" and the sparc >> Preselection pass sorta-kinda work around SOME of this, > > Actually, they are capable of working around much of this, and most such > optimizations are actually quite architecture-independent, e.g., > LowerMultiDimRefs, which is a major optimization for...

...this, but it can happen with any instructions > (including the use of large integer (or any FP) constants on RISC > machines, addressing globals with PIC code, handling > extended/truncated operations (for RISC machines with a single integer > size), etc. Note that hacks like "LowerMultiDimRefs" and the sparc > Preselection pass sorta-kinda work around SOME of this, Actually, they are capable of working around much of this, and most such optimizations are actually quite architecture-independent, e.g., LowerMultiDimRefs, which is a major optimization for array-intensive langua...

...particular case of this, but it can happen with any instructions (including the use of large integer (or any FP) constants on RISC machines, addressing globals with PIC code, handling extended/truncated operations (for RISC machines with a single integer size), etc. Note that hacks like "LowerMultiDimRefs" and the sparc Preselection pass sorta-kinda work around SOME of this, but they really are just as bad as the problem: they solve some cases and make other cases worse. The only way to make preselection or lowermultidimrefs work is to duplication all of the knowledge of how the instructio...

On Feb 22, 2005, at 9:09 AM, Vikram S. Adve wrote: >> The only way to make preselection or lowermultidimrefs work is to >> duplication all of the knowledge of how the instruction selector will >> select the code (e.g. the advice about allowing constant indices to >> be grouped together). > > This is why you need a separate, low-level optimization framework - > the kind you w...

Vikram S. Adve wrote: >> The only way to make preselection or lowermultidimrefs work is to >> duplication all of the knowledge of how the instruction selector will >> select the code (e.g. the advice about allowing constant indices to >> be grouped together). > > > This is why you need a separate, low-level optimization framework - > the kind...

I figured getelementptr exists as it does to facilitate data flow analysis, but it does need to be broken down before instruction selection. It's not just the missed optimization opportunities. It also introduces a huge amount of complexity into instruction selection as they deal with its complexity. It would also take care of many of the FIXMEs in LoopStrengthReduce. Vikram Adve

> > Now the problem is obvious.  A two dimensional array access is being > performed by a single instruction.  The arithmetic needed to address > the element is implicit, and therefore inaccessible to optimizations.  > The redundant calculations can not be eliminated, nor can strength > reduction be performed.  getelementptr needs to be broken down into > its constituent