forked from OSchip/llvm-project
a192ea204b
The output can be in different orders, which breaks the test in some situations. I have not yet found out what the root cause of the order difference is. This fixes our internal build. If it is not the right solution, feel free to roll back. llvm-svn: 184535 |
||
|---|---|---|
| .. | ||
| GNU.a | ||
| IsNAN.o | ||
| MacOSX.a | ||
| README.txt | ||
| SVR4.a | ||
| check_binary_output.ll | ||
| directory.ll | ||
| evenlen | ||
| extract.ll | ||
| lit.local.cfg | ||
| oddlen | ||
| toc_GNU.ll | ||
| toc_MacOSX.ll | ||
| toc_SVR4.ll | ||
| toc_xpg4.ll | ||
| very_long_bytecode_file_name.bc | ||
| xpg4.a | ||
README.txt
test/Regression/Archive ======================= This directory contains various tests of llvm-ar and to ensure compatibility reading other ar(1) formats. It also provides a basic functionality test for these tools. There are four archives accompanying these tests: GNU.a - constructed on Linux with GNU ar MacOSX.a - constructed on Mac OS X with its native BSD4.4 ar SVR4.a - constructed on Solaris with /usr/ccs/bin/ar xpg4.a - constructed on Solaris with /usr/xpg4/bin/ar Each type of test is run on each of these archive files. These archives each contain four members: oddlen - a member with an odd lengthed name and content evenlen - a member with an even lengthed name and content IsNAN.o - a Linux native binary very_long_bytecode_file_name.bc - LLVM bytecode file with really long name These files test different aspects of the archiver that should cause failures in llvm-ar if regressions are introduced.