Performance Tips for Frontend Authors


The intended audience of this document is developers of language frontends targeting LLVM IR. This document is home to a collection of tips on how to generate IR that optimizes well. As with any optimizer, LLVM has its strengths and weaknesses. In some cases, surprisingly small changes in the source IR can have a large effect on the generated code.

Avoid loads and stores of large aggregate type

LLVM currently does not optimize well loads and stores of large aggregate types (i.e. structs and arrays). As an alternative, consider loading individual fields from memory.

Aggregates that are smaller than the largest (performant) load or store instruction supported by the targeted hardware are well supported. These can be an effective way to represent collections of small packed fields.

Zext GEP indices to machine register width

Internally, LLVM often promotes the width of GEP indices to machine register width. When it does so, it will default to using sign extension (sext) operations for safety. If your source language provides information about the range of the index, you may wish to manually extend indices to machine register width using a zext instruction.

Other things to consider

  1. Make sure that a DataLayout is provided (this will likely become required in the near future, but is certainly important for optimization).
  2. Add nsw/nuw flags as appropriate. Reasoning about overflow is generally hard for an optimizer so providing these facts from the frontend can be very impactful.
  3. Use fast-math flags on floating point operations if legal. If you don’t need strict IEEE floating point semantics, there are a number of additional optimizations that can be performed. This can be highly impactful for floating point intensive computations.
  4. Use inbounds on geps. This can help to disambiguate some aliasing queries.
  5. Add noalias/align/dereferenceable/nonnull to function arguments and return values as appropriate
  6. Mark functions as readnone/readonly or noreturn/nounwind when known. The optimizer will try to infer these flags, but may not always be able to. Manual annotations are particularly important for external functions that the optimizer can not analyze.
  7. Use ptrtoint/inttoptr sparingly (they interfere with pointer aliasing analysis), prefer GEPs
  8. Use the lifetime.start/lifetime.end and invariant.start/invariant.end intrinsics where possible. Common profitable uses are for stack like data structures (thus allowing dead store elimination) and for describing life times of allocas (thus allowing smaller stack sizes).
  9. Use pointer aliasing metadata, especially tbaa metadata, to communicate otherwise-non-deducible pointer aliasing facts
  10. Use the “most-private” possible linkage types for the functions being defined (private, internal or linkonce_odr preferably)
  11. Mark invariant locations using !invariant.load and TBAA’s constant flags
  12. Prefer globals over inttoptr of a constant address - this gives you dereferencability information. In MCJIT, use getSymbolAddress to provide actual address.
  13. Be wary of ordered and atomic memory operations. They are hard to optimize and may not be well optimized by the current optimizer. Depending on your source language, you may consider using fences instead.
  14. If calling a function which is known to throw an exception (unwind), use an invoke with a normal destination which contains an unreachable instruction. This form conveys to the optimizer that the call returns abnormally. For an invoke which neither returns normally or requires unwind code in the current function, you can use a noreturn call instruction if desired. This is generally not required because the optimizer will convert an invoke with an unreachable unwind destination to a call instruction.
  15. If you language uses range checks, consider using the IRCE pass. It is not currently part of the standard pass order.
  16. For languages with numerous rarely executed guard conditions (e.g. null checks, type checks, range checks) consider adding an extra execution or two of LoopUnswith and LICM to your pass order. The standard pass order, which is tuned for C and C++ applications, may not be sufficient to remove all dischargeable checks from loops.
  17. Use profile metadata to indicate statically known cold paths, even if dynamic profiling information is not available. This can make a large difference in code placement and thus the performance of tight loops.
  18. When generating code for loops, try to avoid terminating the header block of the loop earlier than necessary. If the terminator of the loop header block is a loop exiting conditional branch, the effectiveness of LICM will be limited for loads not in the header. (This is due to the fact that LLVM may not know such a load is safe to speculatively execute and thus can’t lift an otherwise loop invariant load unless it can prove the exiting condition is not taken.) It can be profitable, in some cases, to emit such instructions into the header even if they are not used along a rarely executed path that exits the loop. This guidance specifically does not apply if the condition which terminates the loop header is itself invariant, or can be easily discharged by inspecting the loop index variables.
  19. In hot loops, consider duplicating instructions from small basic blocks which end in highly predictable terminators into their successor blocks. If a hot successor block contains instructions which can be vectorized with the duplicated ones, this can provide a noticeable throughput improvement. Note that this is not always profitable and does involve a potentially large increase in code size.
  20. Avoid high in-degree basic blocks (e.g. basic blocks with dozens or hundreds of predecessors). Among other issues, the register allocator is known to perform badly with confronted with such structures. The only exception to this guidance is that a unified return block with high in-degree is fine.
  21. When checking a value against a constant, emit the check using a consistent comparison type. The GVN pass will optimize redundant equalities even if the type of comparison is inverted, but GVN only runs late in the pipeline. As a result, you may miss the opportunity to run other important optimizations. Improvements to EarlyCSE to remove this issue are tracked in Bug 23333.
  22. Avoid using arithmetic intrinsics unless you are required by your source language specification to emit a particular code sequence. The optimizer is quite good at reasoning about general control flow and arithmetic, it is not anywhere near as strong at reasoning about the various intrinsics. If profitable for code generation purposes, the optimizer will likely form the intrinsics itself late in the optimization pipeline. It is very rarely profitable to emit these directly in the language frontend. This item explicitly includes the use of the overflow intrinsics.
  23. Avoid using the assume intrinsic until you’ve established that a) there’s no other way to express the given fact and b) that fact is critical for optimization purposes. Assumes are a great prototyping mechanism, but they can have negative effects on both compile time and optimization effectiveness. The former is fixable with enough effort, but the later is fairly fundamental to their designed purpose.

p.s. If you want to help improve this document, patches expanding any of the above items into standalone sections of their own with a more complete discussion would be very welcome.

Adding to this document

If you run across a case that you feel deserves to be covered here, please send a patch to llvm-commits for review.

If you have questions on these items, please direct them to llvm-dev. The more relevant context you are able to give to your question, the more likely it is to be answered.