Add working set base runtime library

Summary:
Adds the base runtime library for the working set tool.
Adds slowpath code for updating the shadow memory.

To be added in the future:
+ Scan memory and report the total size.
+ Take samples for intermediate values.

Reviewers: aizatsky

Subscribers: kubabrecka, vitalybuka, zhaoqin, kcc, eugenis, llvm-commits

Differential Revision: http://reviews.llvm.org/D20485

git-svn-id: https://llvm.org/svn/llvm-project/compiler-rt/trunk@270650 91177308-0d34-0410-b5e6-96231b3b80d8

5 files changed, 146 insertions, 6 deletions

diff --git a/lib/esan/CMakeLists.txt b/lib/esan/CMakeLists.txt
index 5e4ee107a..029c13579 100644
--- a/lib/esan/CMakeLists.txt
+++ b/lib/esan/CMakeLists.txt
@@ -11,7 +11,8 @@ set(ESAN_SOURCES
   esan.cpp
   esan_flags.cpp
   esan_interface.cpp
-  esan_interceptors.cpp)
+  esan_interceptors.cpp
+  working_set.cpp)
 
 foreach (arch ${ESAN_SUPPORTED_ARCH})
   add_compiler_rt_runtime(clang_rt.esan
diff --git a/lib/esan/esan.cpp b/lib/esan/esan.cpp
index 746575f66..f4087bcf2 100644
--- a/lib/esan/esan.cpp
+++ b/lib/esan/esan.cpp
@@ -19,6 +19,7 @@
 #include "sanitizer_common/sanitizer_common.h"
 #include "sanitizer_common/sanitizer_flag_parser.h"
 #include "sanitizer_common/sanitizer_flags.h"
+#include "working_set.h"
 
 // See comment below.
 extern "C" {
@@ -31,6 +32,15 @@ bool EsanIsInitialized;
 ToolType WhichTool;
 ShadowMapping Mapping;
 
+// Different tools use different scales within the same shadow mapping scheme.
+// The scale used here must match that used by the compiler instrumentation.
+// This array is indexed by the ToolType enum.
+static const uptr ShadowScale[] = {
+  0, // ESAN_None.
+  2, // ESAN_CacheFrag: 4B:1B, so 4 to 1 == >>2.
+  6, // ESAN_WorkingSet: 64B:1B, so 64 to 1 == >>6.
+};
+
 // We are combining multiple performance tuning tools under the umbrella of
 // one EfficiencySanitizer super-tool.  Most of our tools have very similar
 // memory access instrumentation, shadow memory mapping, libc interception,
@@ -57,6 +67,8 @@ void processRangeAccess(uptr PC, uptr Addr, int Size, bool IsWrite) {
   if (WhichTool == ESAN_CacheFrag) {
     // TODO(bruening): add shadow mapping and update shadow bits here.
     // We'll move this to cache_frag.cpp once we have something.
+  } else if (WhichTool == ESAN_WorkingSet) {
+    processRangeAccessWorkingSet(PC, Addr, Size, IsWrite);
   }
 }
 
@@ -113,10 +125,7 @@ static bool verifyShadowScheme() {
 static void initializeShadow() {
   DCHECK(verifyShadowScheme());
 
-  if (WhichTool == ESAN_CacheFrag)
-    Mapping.initialize(2); // 4B:1B, so 4 to 1 == >>2.
-  else
-    UNREACHABLE("unknown tool shadow mapping");
+  Mapping.initialize(ShadowScale[WhichTool]);
 
   VPrintf(1, "Shadow scale=%d offset=%p\n", Mapping.Scale, Mapping.Offset);
 
@@ -157,7 +166,7 @@ void initializeLibrary(ToolType Tool) {
   ::__cxa_atexit((void (*)())finalizeLibrary);
 
   VPrintf(1, "in esan::%s\n", __FUNCTION__);
-  if (WhichTool != ESAN_CacheFrag) {
+  if (WhichTool <= ESAN_None || WhichTool >= ESAN_Max) {
     Printf("ERROR: unknown tool %d requested\n", WhichTool);
     Die();
   }
@@ -165,6 +174,12 @@ void initializeLibrary(ToolType Tool) {
   initializeShadow();
   initializeInterceptors();
 
+  if (WhichTool == ESAN_CacheFrag) {
+    // FIXME: add runtime code for this tool
+  } else if (WhichTool == ESAN_WorkingSet) {
+    initializeWorkingSet();
+  }
+
   EsanIsInitialized = true;
 }
 
@@ -175,6 +190,8 @@ int finalizeLibrary() {
     // strategy for how to generate a final report.
     // We'll move this to cache_frag.cpp once we have something.
     Report("%s is not finished: nothing yet to report\n", SanitizerToolName);
+  } else if (WhichTool == ESAN_WorkingSet) {
+    return finalizeWorkingSet();
   }
   return 0;
 }
diff --git a/lib/esan/esan_interface_internal.h b/lib/esan/esan_interface_internal.h
index 577c7e3cf..83e0433c7 100644
--- a/lib/esan/esan_interface_internal.h
+++ b/lib/esan/esan_interface_internal.h
@@ -28,6 +28,8 @@ extern "C" {
 typedef enum Type : u32 {
   ESAN_None = 0,
   ESAN_CacheFrag,
+  ESAN_WorkingSet,
+  ESAN_Max,
 } ToolType;
 
 // This function should be called at the very beginning of the process,
diff --git a/lib/esan/working_set.cpp b/lib/esan/working_set.cpp
new file mode 100644
index 000000000..3cee6155b
--- /dev/null
+++ b/lib/esan/working_set.cpp
@@ -0,0 +1,90 @@
+//===-- working_set.cpp ---------------------------------------------------===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file is a part of EfficiencySanitizer, a family of performance tuners.
+//
+// This file contains working-set-specific code.
+//===----------------------------------------------------------------------===//
+
+#include "working_set.h"
+#include "esan.h"
+#include "esan_flags.h"
+#include "esan_shadow.h"
+
+// We shadow every cache line of app memory with one shadow byte.
+// - The highest bit of each shadow byte indicates whether the corresponding
+//   cache line has ever been accessed.
+// - The lowest bit of each shadow byte indicates whether the corresponding
+//   cache line was accessed since the last sample.
+// - The other bits can be used either for a single working set snapshot
+//   between two consecutive samples, or an aggregate working set snapshot
+//   over multiple sample periods (future work).
+// We live with races in accessing each shadow byte.
+typedef unsigned char byte;
+
+namespace __esan {
+
+// See the shadow byte layout description above.
+static const u32 TotalWorkingSetBitIdx = 7;
+static const u32 CurWorkingSetBitIdx = 0;
+static const byte ShadowAccessedVal =
+  (1 << TotalWorkingSetBitIdx) | (1 << CurWorkingSetBitIdx);
+
+void processRangeAccessWorkingSet(uptr PC, uptr Addr, SIZE_T Size,
+                                  bool IsWrite) {
+  if (Size == 0)
+    return;
+  SIZE_T I = 0;
+  uptr LineSize = getFlags()->cache_line_size;
+  // As Addr+Size could overflow at the top of a 32-bit address space,
+  // we avoid the simpler formula that rounds the start and end.
+  SIZE_T NumLines = Size / LineSize +
+    // Add any extra at the start or end adding on an extra line:
+    (LineSize - 1 + Addr % LineSize + Size % LineSize) / LineSize;
+  byte *Shadow = (byte *)appToShadow(Addr);
+  // Write shadow bytes until we're word-aligned.
+  while (I < NumLines && (uptr)Shadow % 4 != 0) {
+    if ((*Shadow & ShadowAccessedVal) != ShadowAccessedVal)
+      *Shadow |= ShadowAccessedVal;
+    ++Shadow;
+    ++I;
+  }
+  // Write whole shadow words at a time.
+  // Using a word-stride loop improves the runtime of a microbenchmark of
+  // memset calls by 10%.
+  u32 WordValue = ShadowAccessedVal | ShadowAccessedVal << 8 |
+    ShadowAccessedVal << 16 | ShadowAccessedVal << 24;
+  while (I + 4 <= NumLines) {
+    if ((*(u32*)Shadow & WordValue) != WordValue)
+      *(u32*)Shadow |= WordValue;
+    Shadow += 4;
+    I += 4;
+  }
+  // Write any trailing shadow bytes.
+  while (I < NumLines) {
+    if ((*Shadow & ShadowAccessedVal) != ShadowAccessedVal)
+      *Shadow |= ShadowAccessedVal;
+    ++Shadow;
+    ++I;
+  }
+}
+
+void initializeWorkingSet() {
+  // The shadow mapping assumes 64 so this cannot be changed.
+  CHECK(getFlags()->cache_line_size == 64);
+}
+
+int finalizeWorkingSet() {
+  // FIXME NYI: we need to add memory scanning to report the total lines
+  // touched, and later add sampling to get intermediate values.
+  Report("%s is not finished: nothing yet to report\n", SanitizerToolName);
+  return 0;
+}
+
+} // namespace __esan
diff --git a/lib/esan/working_set.h b/lib/esan/working_set.h
new file mode 100644
index 000000000..4b1bc1596
--- /dev/null
+++ b/lib/esan/working_set.h
@@ -0,0 +1,30 @@
+//===-- working_set.h -------------------------------------------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This file is a part of EfficiencySanitizer, a family of performance tuners.
+//
+// Header for working-set-specific code.
+//===----------------------------------------------------------------------===//
+
+#ifndef WORKING_SET_H
+#define WORKING_SET_H
+
+#include "interception/interception.h"
+#include "sanitizer_common/sanitizer_internal_defs.h"
+
+namespace __esan {
+
+void initializeWorkingSet();
+int finalizeWorkingSet();
+void processRangeAccessWorkingSet(uptr PC, uptr Addr, SIZE_T Size,
+                                  bool IsWrite);
+
+} // namespace __esan
+
+#endif // WORKING_SET_H