Add RealDataBench.cpp

CMakeLists.txt

@@ -189,6 +189,9 @@ if(BUILD_TESTING)
   # "-fsanitize=address,undefined,fuzzer")
   # target_link_options(conflict_set_bench PRIVATE
   # "-fsanitize=address,undefined,fuzzer")
+
+  add_executable(conflict_set_bench_e2e RealDataBench.cpp)
+  target_link_libraries(conflict_set_bench_e2e PRIVATE ${PROJECT_NAME})
 endif()
 # packaging
 

RealDataBench.cpp

@@ -0,0 +1,114 @@
+#include <ConflictSet.h>
+
+#include <chrono>
+#include <fcntl.h>
+#include <fstream>
+#include <sys/mman.h>
+#include <sys/stat.h>
+#include <unistd.h>
+
+double now() {
+  return std::chrono::duration_cast<std::chrono::nanoseconds>(
+             std::chrono::steady_clock::now().time_since_epoch())
+             .count() *
+         1e-9;
+}
+
+inline size_t getPageSize() {
+  static size_t kPageSize = sysconf(_SC_PAGESIZE);
+  return kPageSize;
+}
+
+/// Helper for rounding up to page size (or some other alignment)
+constexpr inline size_t rightAlign(size_t offset, size_t alignment) {
+  return offset % alignment == 0 ? offset
+                                 : ((offset / alignment) + 1) * alignment;
+}
+
+int main(int argc, const char **argv) {
+  // Use with this dataset https://snap.stanford.edu/data/memetracker9.html
+  // Preprocess the files with `sed -i '' '/^Q/d'`
+
+  double checkTime = 0;
+  double addTime = 0;
+  double gcTime = 0;
+
+  ConflictSet cs{0};
+  int64_t version = 0;
+  double timer = 0;
+
+  for (int i = 1; i < argc; ++i) {
+    int fd = open(argv[i], O_RDONLY);
+    struct stat st;
+    if (fstat(fd, &st) == -1) {
+      int err = errno;
+      fprintf(stderr, "stat error %s - %s\n", argv[i], strerror(err));
+      fflush(stderr);
+      abort();
+    }
+
+    int64_t size = rightAlign(st.st_size, getPageSize());
+    const uint8_t *begin =
+        (uint8_t *)mmap(0, size, PROT_READ, MAP_PRIVATE, fd, 0);
+    madvise((void *)begin, size, MADV_SEQUENTIAL);
+
+    using StringView = std::basic_string_view<uint8_t>;
+
+    StringView write;
+    std::vector<StringView> reads;
+    std::vector<ConflictSet::ReadRange> readRanges;
+    std::vector<ConflictSet::Result> results;
+
+    for (uint8_t *end = (uint8_t *)memchr(begin, '\n', size); end != nullptr;) {
+      StringView line{begin, static_cast<size_t>(end - begin)};
+      size -= end - begin + 1;
+      begin = end + 1;
+      end = (uint8_t *)memchr(begin, '\n', size);
+
+      if (line.size() > 0 && line[0] == 'P') {
+        write = std::move(line);
+      } else if (line.size() > 0 && line[0] == 'L') {
+        reads.push_back(std::move(line));
+      } else if (line.empty()) {
+        {
+          readRanges.resize(reads.size());
+          auto iter = readRanges.begin();
+          for (const auto &read : reads) {
+            iter->begin.p = (const uint8_t *)read.data();
+            iter->begin.len = read.size();
+            iter->end.len = 0;
+            iter->readVersion = version - 100;
+            ++iter;
+          }
+        }
+        results.resize(readRanges.size());
+
+        timer = now();
+        cs.check(readRanges.data(), results.data(), readRanges.size());
+        checkTime += now() - timer;
+
+        // Add unconditionally so that the load doesn't actually depend on the
+        // conflict rate
+        ConflictSet::WriteRange w;
+        w.writeVersion = ++version;
+        w.begin.p = (const uint8_t *)write.data();
+        w.begin.len = write.size();
+        w.end.len = 0;
+
+        timer = now();
+        cs.addWrites(&w, 1);
+        addTime += now() - timer;
+
+        write = {};
+        reads.clear();
+
+        timer = now();
+        cs.setOldestVersion(version - 10000);
+        gcTime += now() - timer;
+      }
+    }
+    close(fd);
+  }
+
+  printf("Check: %g, Add: %g, Gc: %g\n", checkTime, addTime, gcTime);
+}