#===----------------------------------------------------------------------===## # # Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. # See https://llvm.org/LICENSE.txt for license information. # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception # #===----------------------------------------------------------------------===## # Test that we don't remove transitive includes of public C++ headers in the library accidentally. # When we remove a transitive public include, clients tend to break because they don't always # properly include what they use. Note that we don't check which system (C) headers are # included transitively, because that is too unstable across platforms, and hence difficult # to test for. # # This is not meant to block libc++ from removing unused transitive includes # forever, however we do try to group removals for a couple of releases # to avoid breaking users at every release. # RUN: %{python} %s %{libcxx}/utils import sys sys.path.append(sys.argv[1]) from libcxx.header_information import lit_header_restrictions, public_headers import re # To re-generate the list of expected headers, temporarily set this to True, and run this test. # Note that this needs to be done for all supported language versions of libc++: # for std in c++03 c++11 c++14 c++17 c++20 c++23 c++26; do /bin/llvm-lit --param std=$std libcxx/test/libcxx/transitive_includes.gen.py; done regenerate_expected_results = False BLOCKLIT = '' # block Lit from interpreting a RUN/XFAIL/etc inside the generation script if regenerate_expected_results: print(f"""\ //--- generate-transitive-includes.sh.cpp // RUN{BLOCKLIT}: mkdir %t """) all_traces = [] for header in sorted(public_headers): if header.endswith('.h'): # Skip C compatibility or detail headers continue normalized_header = re.sub('/', '_', header) print(f"""\ // RUN{BLOCKLIT}: echo "#include <{header}>" | %{{cxx}} -xc++ - %{{flags}} %{{compile_flags}} --trace-includes -fshow-skipped-includes --preprocess > /dev/null 2> %t/trace-includes.{normalized_header}.txt """) all_traces.append(f'%t/trace-includes.{normalized_header}.txt') print(f"""\ // RUN{BLOCKLIT}: %{{python}} %{{libcxx}}/test/libcxx/transitive_includes_to_csv.py {' '.join(all_traces)} > %{{libcxx}}/test/libcxx/transitive_includes/%{{cxx_std}}.csv """) else: for header in public_headers: if header.endswith('.h'): # Skip C compatibility or detail headers continue # Escape slashes for the awk command below escaped_header = header.replace('/', '\/') print(f"""\ //--- {header}.sh.cpp {lit_header_restrictions.get(header, '')} // TODO: Fix this test to make it work with localization or wide characters disabled // UNSUPPORTED{BLOCKLIT}: no-localization, no-wide-characters, no-threads, no-filesystem, libcpp-has-no-incomplete-tzdb, no-tzdb // When built with modules, this test doesn't work because --trace-includes doesn't // report the stack of includes correctly. // UNSUPPORTED{BLOCKLIT}: clang-modules-build // This test uses --trace-includes, which is not supported by GCC. // UNSUPPORTED{BLOCKLIT}: gcc // This test is not supported when we remove the transitive includes provided for backwards // compatibility. When we bulk-remove them, we'll adjust the includes that are expected by // this test instead. // UNSUPPORTED{BLOCKLIT}: transitive-includes-disabled // TODO: Figure out why doesn't work on FreeBSD // UNSUPPORTED{BLOCKLIT}: LIBCXX-FREEBSD-FIXME // RUN{BLOCKLIT}: mkdir %t // RUN{BLOCKLIT}: %{{cxx}} %s %{{flags}} %{{compile_flags}} --trace-includes -fshow-skipped-includes --preprocess > /dev/null 2> %t/trace-includes.txt // RUN{BLOCKLIT}: %{{python}} %{{libcxx}}/test/libcxx/transitive_includes_to_csv.py %t/trace-includes.txt > %t/actual_transitive_includes.csv // RUN{BLOCKLIT}: cat %{{libcxx}}/test/libcxx/transitive_includes/%{{cxx_std}}.csv | awk '/^{escaped_header} / {{ print }}' > %t/expected_transitive_includes.csv // RUN{BLOCKLIT}: diff -w %t/expected_transitive_includes.csv %t/actual_transitive_includes.csv #include <{header}> """)