Merge pull request #304 from pjotrp/main

wfmash now builds with clang

.github/workflows/test_on_push.yml

@@ -36,16 +36,12 @@ jobs:
         run: git submodule update --init --recursive
       - name: Build wfmash
         run: cmake -H. -Bbuild -D CMAKE_BUILD_TYPE=Debug -DWFA_PNG_AND_TSV=ON && cmake --build build -- -j 2
-      - name: Test mapping coverage with 8 yeast genomes (PAF output)
-        run: ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 build/bin/wfmash data/scerevisiae8.fa.gz -p 95 -n 7 -m -L -Y '#' > scerevisiae8.paf; scripts/test.sh data/scerevisiae8.fa.gz.fai scerevisiae8.paf 0.92
+      - name: Run cmake tests
+        run: |
+          cd build
+          ctest --verbose
       - name: Test mapping+alignment with a subset of the LPA dataset (PAF output)
         run: ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 build/bin/wfmash data/LPA.subset.fa.gz -n 10 -L > LPA.subset.paf && head LPA.subset.paf
-      - name: Test mapping+alignment with a subset of the LPA dataset (SAM output)
-        run: ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 build/bin/wfmash data/LPA.subset.fa.gz -N -a -L > LPA.subset.sam && samtools view LPA.subset.sam -bS | samtools sort > LPA.subset.bam && samtools index LPA.subset.bam && samtools view LPA.subset.bam | head | cut -f 1-9
-      - name: Test mapping+alignment with short reads (500 bps) to a reference (SAM output)
-        run: ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 build/bin/wfmash data/reference.fa.gz data/reads.500bps.fa.gz -s 0.5k -N -a > reads.500bps.sam && samtools view reads.500bps.sam -bS | samtools sort > reads.500bps.bam && samtools index reads.500bps.bam && samtools view reads.500bps.bam | head
-      - name: Test mapping+alignment with short reads (255bps) (PAF output)
-        run: ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 build/bin/wfmash data/reads.255bps.fa.gz -w 16 -s 100 -L > reads.255bps.paf && head reads.255bps.paf
       - name: Test input mapping functionality
         run: |
           # First generate mappings

.gitignore

@@ -22,8 +22,14 @@ wfmash
 #build directories
 src/common/wflign/build
 build
+pgo
 
 #Others
+*.bam
+*.bai
+*.sam
+*.bed
+wfmash-*
 *.cache
 *~
 \#*
@@ -32,9 +38,11 @@ src/common/WFA2-lib/bin/
 src/wfmash_git_version.hpp
 src/common/wflign/src/wflign_git_version.hpp
 
-test/
 .idea/
 cmake-build-debug/
 result
+tmp/
 .aider*
 .env
+scerevisiae8*
+reads.255bps.paf

CMakeLists.txt

@@ -1,27 +1,49 @@
+# Usage:
+#
+#   mkdir build ; cd build
+#   cmake -DCMAKE_BUILD_TYPE=Debug ..
+#   make -j 8 VERBOSE=1
+#
+#   Build types are Release (default), Debug, Generic, RelWithDebInfo
+
 cmake_minimum_required(VERSION 3.9 FATAL_ERROR)
 project(wfmash)
 
 enable_testing()
 
-include(CheckIPOSupported) # adds lto
-check_ipo_supported(RESULT ipo_supported OUTPUT output)
-
 include(GNUInstallDirs)
 include(CheckCXXCompilerFlag)
 
 set(CMAKE_CXX_STANDARD 17)
+# set(CMAKE_CXX_STANDARD 20) # because of sequenceIds.hpp:101:23: warning: captured structured bindings are a C++20 extension [-Wc++20-extensions]
 set(CMAKE_CXX_STANDARD_REQUIRED ON)  # Falling back to different standard is not allowed.
 set(CMAKE_CXX_EXTENSIONS OFF)  # Make sure no compiler-specific features are used.
 set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
 
+
 option(BUILD_STATIC "Build static binary" OFF)
 option(BUILD_DEPS "Build external dependencies" OFF)
-option(BUILD_RETARGETABLE "Build retargetable binary" OFF)
+option(BUILD_OPTIMIZED "Build optimized binary" OFF)
+option(GPROF "Enable gprof profiling" OFF)
+option(DISABLE_LTO "Disable IPO/LTO" OFF)
+option(STOP_ON_ERROR "Stop compiling on first error" OFF)
+
+if (NOT DISABLE_LTO)
+  include(CheckIPOSupported) # adds lto
+  check_ipo_supported(RESULT ipo_supported OUTPUT output) # lto
+endif()
+
+if (STOP_ON_ERROR)
+  add_definitions( -Wfatal-errors )
+endif()
 
 if (BUILD_STATIC)
   set(CMAKE_FIND_LIBRARY_SUFFIXES ".a")
   set(BUILD_SHARED_LIBS OFF)
   set(CMAKE_EXE_LINKER_FLAGS "-static")
+  set(CMAKE_POSITION_INDEPENDENT_CODE OFF)
+else ()
+  set(CMAKE_POSITION_INDEPENDENT_CODE ON)
 endif ()
 
 find_package(PkgConfig REQUIRED)
@@ -34,43 +56,53 @@ find_package(OpenMP REQUIRED)
 
 if(NOT CMAKE_BUILD_TYPE)
     set(CMAKE_BUILD_TYPE Release CACHE STRING
-            "Choose the type of build, options are: Release Debug Generic." FORCE)
+            "Choose the type of build, options are: Release|Debug|Generic|RelWithDebInfo (for distros)." FORCE)
 endif()
 
 message(STATUS "CMAKE_BUILD_TYPE: ${CMAKE_BUILD_TYPE}")
 
 if (${CMAKE_BUILD_TYPE} MATCHES Release)
-  set(CMAKE_CXX_FLAGS_RELEASE "-DNDEBUG")
-  set(CMAKE_C_FLAGS_RELEASE "-DNDEBUG")
-  if (NOT EXTRA_FLAGS)
-    if (BUILD_RETARGETABLE)
-      set(EXTRA_FLAGS "-Ofast -march=x86-64-v3 -flto")
-    else()
-      set(EXTRA_FLAGS "-Ofast -march=native -flto")
-    endif()
+  if (BUILD_OPTIMIZED)
+    # set(EXTRA_FLAGS "-Ofast -march=native -pipe -msse4.2 -funroll-all-loops") #  -fprofile-generate=../pgo")
+    # set(EXTRA_FLAGS "-Ofast -march=x86-64-v3 -funroll-all-loops")
+    set(EXTRA_FLAGS "-O3 -march=native -funroll-all-loops")
+    # set(EXTRA_FLAGS "-O3 -march=native -funroll-all-loops -fprofile-generate=${CMAKE_BINARY_DIR}/../pgo")
+    # set(EXTRA_FLAGS "-O3 -march=native -funroll-all-loops -fprofile-use=${CMAKE_BINARY_DIR}/../pgo")
+
+  else()
+    set(EXTRA_FLAGS "-Ofast -march=native")
   endif()
+  set(CMAKE_CXX_FLAGS_RELEASE "-DNDEBUG ${EXTRA_FLAGS}")
+  set(CMAKE_C_FLAGS_RELEASE "-DNDEBUG ${EXTRA_FLAGS}")
 endif ()
 
 if (${CMAKE_BUILD_TYPE} MATCHES Generic)
   set(CMAKE_CXX_FLAGS_RELEASE "-DNDEBUG")
   set(CMAKE_C_FLAGS_RELEASE "-DNDEBUG")
-  if (NOT EXTRA_FLAGS)
-    set(EXTRA_FLAGS "-Ofast -flto")
-  endif()
 endif ()
 
 if (${CMAKE_BUILD_TYPE} MATCHES Debug)
     # Enable debug symbols and ASan with no optimizations
     set (CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -O0 -g -fsanitize=address -fno-omit-frame-pointer")
     set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -O0 -g -fsanitize=address -fno-omit-frame-pointer")
     # Ensure that ASan is linked explicitly
-    set(CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=address")
-else()
-    set (CMAKE_C_FLAGS "${OpenMP_C_FLAGS} ${PIC_FLAG} ${EXTRA_FLAGS}")
-    set (CMAKE_CXX_FLAGS "${OpenMP_CXX_FLAGS} ${PIC_FLAG} ${EXTRA_FLAGS}")
+    set (CMAKE_EXE_LINKER_FLAGS "${CMAKE_EXE_LINKER_FLAGS} -fsanitize=address")
+else ()
+    set (CMAKE_C_FLAGS "${OpenMP_C_FLAGS}")
+    set (CMAKE_CXX_FLAGS "${OpenMP_CXX_FLAGS}")
 endif ()
 
-if(WFA_PNG_AND_TSV)
+if( NOT DISABLE_LTO AND ipo_supported )
+  message(STATUS "IPO / LTO enabled")
+  set(CMAKE_INTERPROCEDURAL_OPTIMIZATION ${ipo_supported})
+endif()
+
+if(GPROF)
+  set(CMAKE_C_FLAGS "${CMAKE_CXX_FLAGS} -g -pg")
+  set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -g -pg")
+endif(GPROF)
+
+if (WFA_PNG_AND_TSV)
     set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -DWFA_PNG_TSV_TIMING")
     set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DWFA_PNG_TSV_TIMING")
 endif ()
@@ -145,13 +177,15 @@ else()
   #)
 
   target_link_libraries(wfmash
-  gsl
-  gslcblas
-  hts
-  deflate
+    gsl
+    gslcblas
+    hts
+    deflate
   )
 endif()
 
+
+
 target_include_directories(wfmash PRIVATE
   src
   src/common
@@ -162,32 +196,85 @@ target_include_directories(wfmash PRIVATE
 target_link_libraries(wfmash
   m
   pthread
-  libwflign_static
   rt
-  wfa2cpp_static
   lzma
   bz2
   z
   Threads::Threads
 )
 
+if (BUILD_STATIC)
+  target_link_libraries(wfmash
+    wflign_static
+    wfa2cpp_static
+  )
+else()
+  target_link_libraries(wfmash
+    wflign
+    wfa2cpp
+  )
+endif()
+
 # This is to disable tests defined in CTestCustom.cmake:
 configure_file(${CMAKE_SOURCE_DIR}/CTestCustom.cmake ${CMAKE_BINARY_DIR})
 
+set(INVOKE "ASAN_OPTIONS=detect_leaks=1:symbolize=1 LSAN_OPTIONS=verbosity=0:log_threads=1 ${CMAKE_BINARY_DIR}/bin/wfmash")
+set(REGRESSION_TEST_DIR "test/data/regression")
+
+add_test(
+  NAME wfmash-time-LPA
+  COMMAND bash -c "${INVOKE} data/LPA.subset.fa.gz -p 80 -n 5 -t 8 > wfmash-time-LPA.output"
+  WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+
+add_test(
+  NAME wfmash-subset-LPA-to-SAM
+  COMMAND bash -c "${INVOKE} data/LPA.subset.fa.gz -N -a -L > LPA.subset.sam && samtools view LPA.subset.sam -bS | samtools sort > LPA.subset.bam && samtools index LPA.subset.bam && samtools view LPA.subset.bam | head | cut -f 1-9"
+  WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+
+add_test(
+  NAME wfmash-mapping-coverage-with-8-yeast-genomes-to-PAF
+  COMMAND bash -c "${INVOKE} data/scerevisiae8.fa.gz -p 95 -n 7 -m -L -Y \\# > scerevisiae8.paf && ./scripts/test.sh data/scerevisiae8.fa.gz.fai scerevisiae8.paf 0.92 && head -3000 scerevisiae8.paf > scerevisiae8.paf.output && ${CMAKE_COMMAND} -E compare_files ${REGRESSION_TEST_DIR}/scerevisiae8.paf.output scerevisiae8.paf.output"
+  WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+
 add_test(
-  NAME wfmash-test
-  COMMAND wfmash data/LPA.subset.fa.gz -p 80 -n 5 -t 8
+  NAME wfmash-short-reads-500bps-to-SAM
+  COMMAND bash -c "${INVOKE} data/reference.fa.gz data/reads.500bps.fa.gz -s 0.5k -N -a > reads.500bps.sam && samtools view reads.500bps.sam -bS | samtools sort > reads.500bps.bam && samtools index reads.500bps.bam && samtools view reads.500bps.bam | head > wfmash-short-reads-500bps-to-SAM.output && ${CMAKE_COMMAND} -E compare_files ${REGRESSION_TEST_DIR}/wfmash-short-reads-500bps-to-SAM.output wfmash-short-reads-500bps-to-SAM.output"
+  WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
+
+add_test(
+  NAME wfmash-short-reads-255bps-to-PAF
+  COMMAND bash -c "${INVOKE} data/reads.255bps.fa.gz -w 16 -s 100 -L > reads.255bps.paf && head reads.255bps.paf && ${CMAKE_COMMAND} -E compare_files ${REGRESSION_TEST_DIR}/reads.255bps.paf reads.255bps.paf"
   WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR})
 
 install(TARGETS wfmash DESTINATION bin)
 
-install(TARGETS wfa2cpp_static
-        LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
-        PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+if (BUILD_STATIC)
+
+  install(TARGETS wfa2cpp_static
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+
+  install(TARGETS wflign_static
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
 
-install(TARGETS wfa2_static
-        LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
-        PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+  install(TARGETS wfa2_static
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+else()
+
+  install(TARGETS wfa2cpp
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+
+  install(TARGETS wflign
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+
+  install(TARGETS wfa2
+    LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
+    PUBLIC_HEADER DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+endif()
 
 file(MAKE_DIRECTORY ${CMAKE_SOURCE_DIR}/include)
 execute_process(COMMAND bash ${CMAKE_SOURCE_DIR}/scripts/generate_git_version.sh ${CMAKE_SOURCE_DIR}/src ${CMAKE_SOURCE_DIR}/src/common/wflign/src)