LLVM Testing Infrastructure Guide

Overview

This document is the reference manual for the LLVM testinginfrastructure. It documents the structure of the LLVM testinginfrastructure, the tools needed to use it, and how to add and runtests.

Requirements

In order to use the LLVM testing infrastructure, you will need all of thesoftware required to build LLVM, as well as Python 2.7 orlater.

LLVM Testing Infrastructure Organization

The LLVM testing infrastructure contains three major categories of tests:unit tests, regression tests and whole programs. The unit tests and regressiontests are contained inside the LLVM repository itself under llvm/unittestsand llvm/test respectively and are expected to always pass – they should berun before every commit.

The whole programs tests are referred to as the “LLVM test suite” (or“test-suite”) and are in the test-suite module in subversion. Forhistorical reasons, these tests are also referred to as the “nightlytests” in places, which is less ambiguous than “test-suite” and remainsin use although we run them much more often than nightly.

Unit tests

Unit tests are written using Google Testand Google Mockand are located in the llvm/unittests directory.

Regression tests

The regression tests are small pieces of code that test a specificfeature of LLVM or trigger a specific bug in LLVM. The language they arewritten in depends on the part of LLVM being tested. These tests are driven bythe Lit testing tool (which is part of LLVM), andare located in the llvm/test directory.

Typically when a bug is found in LLVM, a regression test containing justenough code to reproduce the problem should be written and placedsomewhere underneath this directory. For example, it can be a smallpiece of LLVM IR distilled from an actual application or benchmark.

test-suite

The test suite contains whole programs, which are pieces of code whichcan be compiled and linked into a stand-alone program that can beexecuted. These programs are generally written in high level languagessuch as C or C++.

These programs are compiled using a user specified compiler and set offlags, and then executed to capture the program output and timinginformation. The output of these programs is compared to a referenceoutput to ensure that the program is being compiled correctly.

In addition to compiling and executing programs, whole program testsserve as a way of benchmarking LLVM performance, both in terms of theefficiency of the programs generated as well as the speed with whichLLVM compiles, optimizes, and generates code.

The test-suite is located in the test-suite Subversion module.

See the test-suite Guide for details.

Debugging Information tests

The test suite contains tests to check quality of debugging information.The test are written in C based languages or in LLVM assembly language.

These tests are compiled and run under a debugger. The debugger outputis checked to validate of debugging information. See README.txt in thetest suite for more information. This test suite is located in thedebuginfo-tests Subversion module.

Quick start

The tests are located in two separate Subversion modules. The unit andregression tests are in the main “llvm” module under the directoriesllvm/unittests and llvm/test (so you get these tests for free with themain LLVM tree). Use make check-all to run the unit and regression testsafter building LLVM.

The test-suite module contains more comprehensive tests including whole Cand C++ programs. See the test-suite Guide for details.

Unit and Regression tests

To run all of the LLVM unit tests use the check-llvm-unit target:

  1. % make check-llvm-unit

To run all of the LLVM regression tests use the check-llvm target:

  1. % make check-llvm

In order to get reasonable testing performance, build LLVM and subprojectsin release mode, i.e.

  1. % cmake -DCMAKE_BUILD_TYPE="Release" -DLLVM_ENABLE_ASSERTIONS=On

If you have Clang checked out and built, youcan run the LLVM and Clang tests simultaneously using:

  1. % make check-all

To run the tests with Valgrind (Memcheck by default), use the LIT_ARGS makevariable to pass the required options to lit. For example, you can use:

  1. % make check LIT_ARGS="-v --vg --vg-leak"

to enable testing with valgrind and with leak checking enabled.

To run individual tests or subsets of tests, you can use the llvm-litscript which is built as part of LLVM. For example, to run theInteger/BitPacked.ll test by itself you can run:

  1. % llvm-lit ~/llvm/test/Integer/BitPacked.ll

or to run all of the ARM CodeGen tests:

  1. % llvm-lit ~/llvm/test/CodeGen/ARM

For more information on using the lit tool, see llvm-lit —helpor the lit man page.

Debugging Information tests

To run debugging information tests simply add the debuginfo-testsproject to your LLVM_ENABLE_PROJECTS define on the cmakecommand-line.

Regression test structure

The LLVM regression tests are driven by lit and are located in thellvm/test directory.

This directory contains a large array of small tests that exercisevarious features of LLVM and to ensure that regressions do not occur.The directory is broken into several sub-directories, each focused on aparticular area of LLVM.

Writing new regression tests

The regression test structure is very simple, but does require someinformation to be set. This information is gathered via configureand is written to a file, test/lit.site.cfg in the build directory.The llvm/test Makefile does this work for you.

In order for the regression tests to work, each directory of tests musthave a lit.local.cfg file. lit looks for this file to determinehow to run the tests. This file is just Python code and thus is veryflexible, but we’ve standardized it for the LLVM regression tests. Ifyou’re adding a directory of tests, just copy lit.local.cfg fromanother directory to get running. The standard lit.local.cfg simplyspecifies which files to look in for tests. Any directory that containsonly directories does not need the lit.local.cfg file. Read the Litdocumentation for more information.

Each test file must contain lines starting with “RUN:” that tell lithow to run it. If there are no RUN lines, lit will issue an errorwhile running a test.

RUN lines are specified in the comments of the test program using thekeyword RUN followed by a colon, and lastly the command (pipeline)to execute. Together, these lines form the “script” that litexecutes to run the test case. The syntax of the RUN lines is similar to ashell’s syntax for pipelines including I/O redirection and variablesubstitution. However, even though these lines may look like a shellscript, they are not. RUN lines are interpreted by lit.Consequently, the syntax differs from shell in a few ways. You can specifyas many RUN lines as needed.

lit performs substitution on each RUN line to replace LLVM tool nameswith the full paths to the executable built for each tool (in$(LLVM_OBJ_ROOT)/$(BuildMode)/bin). This ensures that lit doesnot invoke any stray LLVM tools in the user’s path during testing.

Each RUN line is executed on its own, distinct from other lines unlessits last character is \. This continuation character causes the RUNline to be concatenated with the next one. In this way you can build uplong pipelines of commands without making huge line lengths. The linesending in \ are concatenated until a RUN line that doesn’t end in\ is found. This concatenated set of RUN lines then constitutes oneexecution. lit will substitute variables and arrange for the pipelineto be executed. If any process in the pipeline fails, the entire line (andtest case) fails too.

Below is an example of legal RUN lines in a .ll file:

  1. ; RUN: llvm-as < %s | llvm-dis > %t1
  2. ; RUN: llvm-dis < %s.bc-13 > %t2
  3. ; RUN: diff %t1 %t2

As with a Unix shell, the RUN lines permit pipelines and I/Oredirection to be used.

There are some quoting rules that you must pay attention to when writingyour RUN lines. In general nothing needs to be quoted. lit won’tstrip off any quote characters so they will get passed to the invoked program.To avoid this use curly braces to tell lit that it should treateverything enclosed as one value.

In general, you should strive to keep your RUN lines as simple as possible,using them only to run tools that generate textual output you can then examine.The recommended way to examine output to figure out if the test passes is usingthe FileCheck tool. [The usage of grep in RUNlines is deprecated - please do not send or commit patches that use it.]

Put related tests into a single file rather than having a separate file pertest. Check if there are files already covering your feature and consideradding your code there instead of creating a new file.

Extra files

If your test requires extra files besides the file containing the RUN:lines, the idiomatic place to put them is in a subdirectory Inputs.You can then refer to the extra files as %S/Inputs/foo.bar.

For example, consider test/Linker/ident.ll. The directory structure isas follows:

  1. test/
  2.   Linker/
  3.     ident.ll
  4.     Inputs/
  5.       ident.a.ll
  6.       ident.b.ll

For convenience, these are the contents:

  1. ;;;;; ident.ll:
  2.  
  3. ; RUN: llvm-link %S/Inputs/ident.a.ll %S/Inputs/ident.b.ll -S | FileCheck %s
  4.  
  5. ; Verify that multiple input llvm.ident metadata are linked together.
  6.  
  7. ; CHECK-DAG: !llvm.ident = !{!0, !1, !2}
  8. ; CHECK-DAG: "Compiler V1"
  9. ; CHECK-DAG: "Compiler V2"
  10. ; CHECK-DAG: "Compiler V3"
  11.  
  12. ;;;;; Inputs/ident.a.ll:
  13.  
  14. !llvm.ident = !{!0, !1}
  15. !0 = metadata !{metadata !"Compiler V1"}
  16. !1 = metadata !{metadata !"Compiler V2"}
  17.  
  18. ;;;;; Inputs/ident.b.ll:
  19.  
  20. !llvm.ident = !{!0}
  21. !0 = metadata !{metadata !"Compiler V3"}

For symmetry reasons, ident.ll is just a dummy file that doesn’tactually participate in the test besides holding the RUN: lines.

Note

Some existing tests use RUN: true in extra files instead of justputting the extra files in an Inputs/ directory. This pattern isdeprecated.

Fragile tests

It is easy to write a fragile test that would fail spuriously if the tool beingtested outputs a full path to the input file. For example, opt bydefault outputs a ModuleID:

  1. $ cat example.ll
  2. define i32 @main() nounwind {
  3.     ret i32 0
  4. }
  5.  
  6. $ opt -S /path/to/example.ll
  7. ; ModuleID = '/path/to/example.ll'
  8.  
  9. define i32 @main() nounwind {
  10.     ret i32 0
  11. }

ModuleID can unexpectedly match against CHECK lines. For example:

  1. ; RUN: opt -S %s | FileCheck
  2.  
  3. define i32 @main() nounwind {
  4.     ; CHECK-NOT: load
  5.     ret i32 0
  6. }

This test will fail if placed into a download directory.

To make your tests robust, always use opt … < %s in the RUN line.opt does not output a ModuleID when input comes from stdin.

Platform-Specific Tests

Whenever adding tests that require the knowledge of a specific platform,either related to code generated, specific output or back-end features,you must make sure to isolate the features, so that buildbots thatrun on different architectures (and don’t even compile all back-ends),don’t fail.

The first problem is to check for target-specific output, for example sizesof structures, paths and architecture names, for example:

  • Tests containing Windows paths will fail on Linux and vice-versa.
  • Tests that check for x86_64 somewhere in the text will fail anywhere else.
  • Tests where the debug information calculates the size of types and structures.

Also, if the test rely on any behaviour that is coded in any back-end, it mustgo in its own directory. So, for instance, code generator tests for ARM gointo test/CodeGen/ARM and so on. Those directories contain a speciallit configuration file that ensure all tests in that directory willonly run if a specific back-end is compiled and available.

For instance, on test/CodeGen/ARM, the lit.local.cfg is:

  1. config.suffixes = ['.ll', '.c', '.cpp', '.test']
  2. if not 'ARM' in config.root.targets:
  3.   config.unsupported = True

Other platform-specific tests are those that depend on a specific featureof a specific sub-architecture, for example only to Intel chips that support AVX2.

For instance, test/CodeGen/X86/psubus.ll tests three sub-architecturevariants:

  1. ; RUN: llc -mcpu=core2 < %s | FileCheck %s -check-prefix=SSE2
  2. ; RUN: llc -mcpu=corei7-avx < %s | FileCheck %s -check-prefix=AVX1
  3. ; RUN: llc -mcpu=core-avx2 < %s | FileCheck %s -check-prefix=AVX2

And the checks are different:

  1. ; SSE2: @test1
  2. ; SSE2: psubusw LCPI0_0(%rip), %xmm0
  3. ; AVX1: @test1
  4. ; AVX1: vpsubusw LCPI0_0(%rip), %xmm0, %xmm0
  5. ; AVX2: @test1
  6. ; AVX2: vpsubusw LCPI0_0(%rip), %xmm0, %xmm0

So, if you’re testing for a behaviour that you know is platform-specific ordepends on special features of sub-architectures, you must add the specifictriple, test with the specific FileCheck and put it into the specificdirectory that will filter out all other architectures.

Constraining test execution

Some tests can be run only in specific configurations, such aswith debug builds or on particular platforms. Use REQUIRESand UNSUPPORTED to control when the test is enabled.

Some tests are expected to fail. For example, there may be a known bugthat the test detect. Use XFAIL to mark a test as an expected failure.An XFAIL test will be successful if its execution fails, andwill be a failure if its execution succeeds.

  1. ; This test will be only enabled in the build with asserts.
  2. ; REQUIRES: asserts
  3. ; This test is disabled on Linux.
  4. ; UNSUPPORTED: -linux-
  5. ; This test is expected to fail on PowerPC.
  6. ; XFAIL: powerpc

REQUIRES and UNSUPPORTED and XFAIL all accept a comma-separatedlist of boolean expressions. The values in each expression may be:

  • Features added to config.available_features byconfiguration files such as lit.cfg.
  • Substrings of the target triple (UNSUPPORTED and XFAIL only).

REQUIRES enables the test if all expressions are true.

UNSUPPORTED disables the test if any expression is true.

XFAIL expects the test to fail if any expression is true.

As a special case, XFAIL: * is expected to fail everywhere.

  1. ; This test is disabled on Windows,
  2. ; and is disabled on Linux, except for Android Linux.
  3. ; UNSUPPORTED: windows, linux && !android
  4. ; This test is expected to fail on both PowerPC and ARM.
  5. ; XFAIL: powerpc || arm

Substitutions

Besides replacing LLVM tool names the following substitutions are performed inRUN lines:

  • %%
  • Replaced by a single %. This allows escaping other substitutions.
  • %s
  • File path to the test case’s source. This is suitable for passing on thecommand line as the input to an LLVM tool.

Example: /home/user/llvm/test/MC/ELF/foo_test.s

  • %S
  • Directory path to the test case’s source.

Example: /home/user/llvm/test/MC/ELF

  • %t
  • File path to a temporary file name that could be used for this test case.The file name won’t conflict with other test cases. You can append to itif you need multiple temporaries. This is useful as the destination ofsome redirected output.

Example: /home/user/llvm.build/test/MC/ELF/Output/foo_test.s.tmp

  • %T
  • Directory of %t. Deprecated. Shouldn’t be used, because it can be easilymisused and cause race conditions between tests.

Use rm -rf %t && mkdir %t instead if a temporary directory is necessary.

Example: /home/user/llvm.build/test/MC/ELF/Output

%{pathsep}

Expands to the path separator, i.e. : (or ; on Windows).

%/s, %/S, %/t, %/T:

Act like the corresponding substitution above but replace any \character with a /. This is useful to normalize path separators.

Example: %s: C:\Desktop Files/foo_test.s.tmp

Example: %/s: C:/Desktop Files/foo_test.s.tmp

%:s, %:S, %:t, %:T:

Act like the corresponding substitution above but remove colons atthe beginning of Windows paths. This is useful to allow concatenationof absolute paths on Windows to produce a legal path.

Example: %s: C:\Desktop Files\foo_test.s.tmp

Example: %:s: C\Desktop Files\foo_test.s.tmp

LLVM-specific substitutions:

  • %shlibext
  • The suffix for the host platforms shared library files. This includes theperiod as the first character.

Example: .so (Linux), .dylib (macOS), .dll (Windows)

  • %exeext
  • The suffix for the host platforms executable files. This includes theperiod as the first character.

Example: .exe (Windows), empty on Linux.

  • %(line), %(line+<number>), %(line-<number>)
  • The number of the line where this substitution is used, with an optionalinteger offset. This can be used in tests with multiple RUN lines, whichreference test file’s line numbers.

Clang-specific substitutions:

  • %clang
  • Invokes the Clang driver.
  • %clang_cpp
  • Invokes the Clang driver for C++.
  • %clang_cl
  • Invokes the CL-compatible Clang driver.
  • %clangxx
  • Invokes the G++-compatible Clang driver.
  • %clang_cc1
  • Invokes the Clang frontend.
  • %itanium_abi_triple, %ms_abi_triple
  • These substitutions can be used to get the current target triple adjusted tothe desired ABI. For example, if the test suite is running with thei686-pc-win32 target, %itanium_abi_triple will expand toi686-pc-mingw32. This allows a test to run with a specific ABI withoutconstraining it to a specific triple.

FileCheck-specific substitutions:

  • %ProtectFileCheckOutput
  • This should precede a FileCheck call if and only if the call’s textualoutput affects test results. It’s usually easy to tell: just look forredirection or piping of the FileCheck call’s stdout or stderr.

To add more substituations, look at test/lit.cfg or lit.local.cfg.

Options

The llvm lit configuration allows to customize some things with user options:

  • llc, opt, …
  • Substitute the respective llvm tool name with a custom command line. Thisallows to specify custom paths and default arguments for these tools.Example:

% llvm-lit “-Dllc=llc -verify-machineinstrs”

  • run_long_tests
  • Enable the execution of long running tests.
  • llvm_site_config
  • Load the specified lit configuration instead of the default one.

Other Features

To make RUN line writing easier, there are several helper programs. Thesehelpers are in the PATH when running tests, so you can just call them usingtheir name. For example:

  • not
  • This program runs its arguments and then inverts the result code from it.Zero result codes become 1. Non-zero result codes become 0.

To make the output more useful, lit will scanthe lines of the test case for ones that contain a pattern that matchesPR[0-9]+. This is the syntax for specifying a PR (Problem Report) numberthat is related to the test case. The number after “PR” specifies theLLVM bugzilla number. When a PR number is specified, it will be used inthe pass/fail reporting. This is useful to quickly get some context whena test fails.

Finally, any line that contains “END.” will cause the specialinterpretation of lines to terminate. This is generally done right afterthe last RUN: line. This has two side effects:

  • it prevents special interpretation of lines that are part of the testprogram, not the instructions to the test case, and
  • it speeds things up for really big test cases by avoidinginterpretation of the remainder of the file.