DFSan's set label function should avoid writing to the shadow memory when the write would not change the value in memory.

[skerner]

Summary

When writing a label to shadow memory, don't write if the value is already set to the value being written. This dramatically reduces real memory consumption in programs with sparse use of labels.

[pcc]

LGTM

This is not the only place where we set labels. For example, the instrumentation pass may insert code that does it directly. But this seems like a strict improvement over the current situation.

[pcc]

Closed by commit rL215961 (authored by @pcc).

compiler-rt/trunk/lib/dfsan/dfsan.cc

dfsan_label dfsan_create_label(const char *desc, void *userdata) {
   __dfsan_label_info[label].l1 = __dfsan_label_info[label].l2 = 0;
   __dfsan_label_info[label].desc = desc;
   __dfsan_label_info[label].userdata = userdata;
   return label;
 }
 
 extern "C" SANITIZER_INTERFACE_ATTRIBUTE
 void __dfsan_set_label(dfsan_label label, void *addr, uptr size) {
-  for (dfsan_label *labelp = shadow_for(addr); size != 0; --size, ++labelp)
+  for (dfsan_label *labelp = shadow_for(addr); size != 0; --size, ++labelp) {
+    // Don't write the label if it is already the value we need it to be.
+    // In a program where most addresses are not labeled, it is common that
+    // a page of shadow memory is entirely zeroed.  The Linux copy-on-write
+    // implementation will share all of the zeroed pages, making a copy of a
+    // page when any value is written.  The un-sharing will happen even if
+    // the value written does not change the value in memory.  Avoiding the
+    // write when both |label| and |*labelp| are zero dramatically reduces
+    // the amount of real memory used by large programs.
+    if (label == *labelp)
+      continue;
+
     *labelp = label;
   }
+}
 
 SANITIZER_INTERFACE_ATTRIBUTE
 void dfsan_set_label(dfsan_label label, void *addr, uptr size) {
   __dfsan_set_label(label, addr, size);
 }
 
 SANITIZER_INTERFACE_ATTRIBUTE
 void dfsan_add_label(dfsan_label label, void *addr, uptr size) {