[llvm-branch-commits] [compiler-rt] [llvm] Write out raw profile bytes in little endian. (PR #150375)

[Snehasish Kumar via llvm-branch-commits]

https://github.com/snehasish created https://github.com/llvm/llvm-project/pull/150375

Instead of writing out in native endian, write out the raw profile bytes
in little endian. Also update the MIB data in little endian. Also clean
up some lint and unused includes in rawprofile.cpp.

>From a3475c3aa304bd8ded9888bbd39825f4cff94cc6 Mon Sep 17 00:00:00 2001
From: Snehasish Kumar <snehasishk at google.com>
Date: Thu, 24 Jul 2025 06:10:55 +0000
Subject: [PATCH] Write out raw profile bytes in little endian.

Instead of writing out in native endian, write out the raw profile bytes
in little endian. Also update the MIB data in little endian. Also clean
up some lint and unused includes in rawprofile.cpp.
---
 .../lib/memprof/memprof_rawprofile.cpp        | 17 +++++----
 llvm/lib/ProfileData/MemProfReader.cpp        | 35 +++++++++++++++++--
 2 files changed, 44 insertions(+), 8 deletions(-)

diff --git a/compiler-rt/lib/memprof/memprof_rawprofile.cpp b/compiler-rt/lib/memprof/memprof_rawprofile.cpp
index f909d78f5f36a..fbcfee3d655eb 100644
--- a/compiler-rt/lib/memprof/memprof_rawprofile.cpp
+++ b/compiler-rt/lib/memprof/memprof_rawprofile.cpp
@@ -7,10 +7,7 @@
 #include "sanitizer_common/sanitizer_allocator_internal.h"
 #include "sanitizer_common/sanitizer_array_ref.h"
 #include "sanitizer_common/sanitizer_common.h"
-#include "sanitizer_common/sanitizer_linux.h"
-#include "sanitizer_common/sanitizer_procmaps.h"
 #include "sanitizer_common/sanitizer_stackdepot.h"
-#include "sanitizer_common/sanitizer_stackdepotbase.h"
 #include "sanitizer_common/sanitizer_stacktrace.h"
 #include "sanitizer_common/sanitizer_vector.h"
 
@@ -23,7 +20,16 @@ using ::llvm::memprof::encodeHistogramCount;
 
 namespace {
 template <class T> char *WriteBytes(const T &Pod, char *Buffer) {
-  *(T *)Buffer = Pod;
+  static_assert(is_trivially_copyable<T>::value, "T must be POD");
+  const uint8_t *Src = reinterpret_cast<const uint8_t *>(&Pod);
+  for (size_t I = 0; I < sizeof(T); ++I) {
+    Buffer[I] = Src[I];
+  }
+#if defined(__BYTE_ORDER__) && __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__
+  for (size_t i = 0; i < sizeof(T) / 2; ++i) {
+    std::swap(buffer[i], buffer[sizeof(T) - 1 - i]);
+  }
+#endif
   return Buffer + sizeof(T);
 }
 
@@ -33,7 +39,6 @@ void RecordStackId(const uptr Key, UNUSED LockedMemInfoBlock *const &MIB,
   auto *StackIds = reinterpret_cast<Vector<u64> *>(Arg);
   StackIds->PushBack(Key);
 }
-} // namespace
 
 u64 SegmentSizeBytes(ArrayRef<LoadedModule> Modules) {
   u64 NumSegmentsToRecord = 0;
@@ -184,6 +189,7 @@ void SerializeMIBInfoToBuffer(MIBMapTy &MIBMap, const Vector<u64> &StackIds,
   CHECK(ExpectedNumBytes >= static_cast<u64>(Ptr - Buffer) &&
         "Expected num bytes != actual bytes written");
 }
+} // namespace
 
 // Format
 // ---------- Header
@@ -288,5 +294,4 @@ u64 SerializeToRawProfile(MIBMapTy &MIBMap, ArrayRef<LoadedModule> Modules,
 
   return TotalSizeBytes;
 }
-
 } // namespace __memprof
diff --git a/llvm/lib/ProfileData/MemProfReader.cpp b/llvm/lib/ProfileData/MemProfReader.cpp
index 2605e10da5f84..a6ad1c7ba5e3b 100644
--- a/llvm/lib/ProfileData/MemProfReader.cpp
+++ b/llvm/lib/ProfileData/MemProfReader.cpp
@@ -178,8 +178,39 @@ readMemInfoBlocksV5(const char *Ptr) {
     const uint64_t Id =
         endian::readNext<uint64_t, llvm::endianness::little, unaligned>(Ptr);
 
-    MemInfoBlock MIB = *reinterpret_cast<const MemInfoBlock *>(Ptr);
-    Ptr += sizeof(MemInfoBlock);
+    MemInfoBlock MIB;
+#define READ_MIB_FIELD(FIELD)                                                  \
+  MIB.FIELD = endian::readNext<decltype(MIB.FIELD), llvm::endianness::little,  \
+                               unaligned>(Ptr)
+
+    READ_MIB_FIELD(AllocCount);
+    READ_MIB_FIELD(TotalAccessCount);
+    READ_MIB_FIELD(MinAccessCount);
+    READ_MIB_FIELD(MaxAccessCount);
+    READ_MIB_FIELD(TotalSize);
+    READ_MIB_FIELD(MinSize);
+    READ_MIB_FIELD(MaxSize);
+    READ_MIB_FIELD(AllocTimestamp);
+    READ_MIB_FIELD(DeallocTimestamp);
+    READ_MIB_FIELD(TotalLifetime);
+    READ_MIB_FIELD(MinLifetime);
+    READ_MIB_FIELD(MaxLifetime);
+    READ_MIB_FIELD(AllocCpuId);
+    READ_MIB_FIELD(DeallocCpuId);
+    READ_MIB_FIELD(NumMigratedCpu);
+    READ_MIB_FIELD(NumLifetimeOverlaps);
+    READ_MIB_FIELD(NumSameAllocCpu);
+    READ_MIB_FIELD(NumSameDeallocCpu);
+    READ_MIB_FIELD(DataTypeId);
+    READ_MIB_FIELD(TotalAccessDensity);
+    READ_MIB_FIELD(MinAccessDensity);
+    READ_MIB_FIELD(MaxAccessDensity);
+    READ_MIB_FIELD(TotalLifetimeAccessDensity);
+    READ_MIB_FIELD(MinLifetimeAccessDensity);
+    READ_MIB_FIELD(MaxLifetimeAccessDensity);
+    READ_MIB_FIELD(AccessHistogramSize);
+    READ_MIB_FIELD(AccessHistogram);
+#undef READ_MIB_FIELD
 
     if (MIB.AccessHistogramSize > 0) {
       // The in-memory representation uses uint64_t for histogram entries.