| Philip Reames | f8bf9dd | 2015-02-27 23:14:50 +0000 | [diff] [blame] | 1 | ===================================== |
| 2 | Performance Tips for Frontend Authors |
| 3 | ===================================== |
| 4 | |
| 5 | .. contents:: |
| 6 | :local: |
| 7 | :depth: 2 |
| 8 | |
| 9 | Abstract |
| 10 | ======== |
| 11 | |
| 12 | The intended audience of this document is developers of language frontends |
| 13 | targeting LLVM IR. This document is home to a collection of tips on how to |
| 14 | generate IR that optimizes well. As with any optimizer, LLVM has its strengths |
| 15 | and weaknesses. In some cases, surprisingly small changes in the source IR |
| 16 | can have a large effect on the generated code. |
| 17 | |
| 18 | Avoid loads and stores of large aggregate type |
| 19 | ================================================ |
| 20 | |
| 21 | LLVM currently does not optimize well loads and stores of large :ref:`aggregate |
| 22 | types <t_aggregate>` (i.e. structs and arrays). As an alternative, consider |
| 23 | loading individual fields from memory. |
| 24 | |
| 25 | Aggregates that are smaller than the largest (performant) load or store |
| 26 | instruction supported by the targeted hardware are well supported. These can |
| 27 | be an effective way to represent collections of small packed fields. |
| 28 | |
| 29 | Prefer zext over sext when legal |
| 30 | ================================== |
| 31 | |
| 32 | On some architectures (X86_64 is one), sign extension can involve an extra |
| 33 | instruction whereas zero extension can be folded into a load. LLVM will try to |
| 34 | replace a sext with a zext when it can be proven safe, but if you have |
| 35 | information in your source language about the range of a integer value, it can |
| 36 | be profitable to use a zext rather than a sext. |
| 37 | |
| 38 | Alternatively, you can :ref:`specify the range of the value using metadata |
| 39 | <range-metadata>` and LLVM can do the sext to zext conversion for you. |
| 40 | |
| 41 | Zext GEP indices to machine register width |
| 42 | ============================================ |
| 43 | |
| 44 | Internally, LLVM often promotes the width of GEP indices to machine register |
| 45 | width. When it does so, it will default to using sign extension (sext) |
| 46 | operations for safety. If your source language provides information about |
| 47 | the range of the index, you may wish to manually extend indices to machine |
| 48 | register width using a zext instruction. |
| 49 | |
| 50 | |
| 51 | Adding to this document |
| 52 | ======================= |
| 53 | |
| 54 | If you run across a case that you feel deserves to be covered here, please send |
| 55 | a patch to `llvm-commits |
| 56 | <http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits>`_ for review. |
| 57 | |
| 58 | If you have questions on these items, please direct them to `llvmdev |
| 59 | <http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev>`_. The more relevant |
| 60 | context you are able to give to your question, the more likely it is to be |
| 61 | answered. |
| 62 | |