Fix thinLTO long compile time problem

[JohnLee1243]

Summary

ThinLTO compile time is longer on aarch64 system than on x86 system. The reason is that on aarch64 the function of getting physical core number return is 1, so thinLTO cannot work parallel.

[tejohnson]

llvm/lib/Support/Unix/Threading.inc

421

Are you compiling for Android? If not, it should have used this case, which does a superset of what you have added above. I tracked the Android exclusion down to this commit: e9b213131ae9c57f4f151d3206916676135b31b0, which says:

Android is excluded because of the higher API level requirement:
`sched_getaffinity; # introduced-arm=12 introduced-arm64=21 introduced-x86=12 introduced-x86_64=21`

Is that an issue then for aarch64 which iiuc is arm64? Or should the code be restructured so aarch64 falls into this case for Android?

[JohnLee1243]

OK, I think this patch e9b213131ae9c57f4f151d3206916676135b31b0 can fix my bug.

llvm/lib/Support/Unix/Threading.inc

else if (Name == "core id") {
       Val.getAsInteger(10, CurCoreId);
       // The processor id corresponds to an index into cpu_set_t.
       if (CPU_ISSET(CurProcessor, &Affinity))
         CPU_SET(CurPhysicalId * CurSiblings + CurCoreId, &Enabled);
     }
   }
   return CPU_COUNT(&Enabled);
 }
+#elif defined(__linux__) && defined(__aarch64__)
+static int computeHostNumPhysicalCores() {
+  cpu_set_t Affinity, Enabled;
+  if (sched_getaffinity(0, sizeof(Affinity), &Affinity) == 0 &&
+      CPU_COUNT(&Affinity) != 0)
+    return CPU_COUNT(&Affinity);
+
+  // Get affinity failed, check the core number.
+  // Read /proc/cpuinfo as a stream (until EOF reached). It cannot be
+  // mmapped because it appears to have 0 size.
+  llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> Text =
+      llvm::MemoryBuffer::getFileAsStream("/proc/cpuinfo");
+  if (std::error_code EC = Text.getError()) {
+    llvm::errs() << "Can't read "
+                 << "/proc/cpuinfo: " << EC.message() << "\n";
+    return -1;
+  }
+  SmallVector<StringRef, 8> strs;
+  (*Text)->getBuffer().split(strs, "\n", -1, false);
+  int CurProcessor = -1;
+  int CurPhysicalId = -1;
+  int CurSiblings = -1;
+  int CurCoreId = -1;
+  for (StringRef Line : strs) {
+    std::pair<StringRef, StringRef> Data = Line.split(':');
+    auto Name = Data.first.trim();
+    auto Val = Data.second.trim();
+    // These fields are available if the kernel is configured with CONFIG_SMP.
+    if (Name == "processor")
+      Val.getAsInteger(10, CurProcessor);
+    else if (Name == "physical id")
+      Val.getAsInteger(10, CurPhysicalId);
+    else if (Name == "siblings")
+      Val.getAsInteger(10, CurSiblings);
+    else if (Name == "core id") {
+      Val.getAsInteger(10, CurCoreId);
+      // The processor id corresponds to an index into cpu_set_t.
+      if (CPU_ISSET(CurProcessor, &Affinity))
+        CPU_SET(CurPhysicalId * CurSiblings + CurCoreId, &Enabled);
+    }
+  }
+  return CPU_COUNT(&Enabled);
+}
 #elif defined(__linux__) && defined(__s390x__)
 static int computeHostNumPhysicalCores() {
   return sysconf(_SC_NPROCESSORS_ONLN);
 }
 #elif defined(__linux__) && !defined(__ANDROID__)
 static int computeHostNumPhysicalCores() {
   cpu_set_t Affinity;
   if (sched_getaffinity(0, sizeof(Affinity), &Affinity) == 0)
     return CPU_COUNT(&Affinity);
 
   // The call to sched_getaffinity() may have failed because the Affinity
   // mask is too small for the number of CPU's on the system (i.e. the
   // system has more than 1024 CPUs). Allocate a mask large enough for

llvm/unittests/Support/Threading.cpp

 #if LLVM_ENABLE_THREADS
   Triple Host(Triple::normalize(sys::getProcessTriple()));
 
   // Initially this is only testing detection of the number of
   // physical cores, which is currently only supported/tested on
   // some systems.
   return (Host.isOSWindows() && llvm_is_multithreaded()) || Host.isOSDarwin() ||
          (Host.isX86() && Host.isOSLinux()) ||
+         (Host.isAArch64() && Host.isOSLinux()) ||
          (Host.isOSLinux() && !Host.isAndroid()) ||
          (Host.isSystemZ() && Host.isOSzOS());
 #else
   return false;
 #endif
 }
 
 TEST(Threading, PhysicalConcurrency) {