bpo-35214: Initial clang MemorySanitizer support (GH-10479)

Adds configure flags for msan and ubsan builds to make it easier to enable.
These also encode the detail that address sanitizer and memory sanitizer
should disable pymalloc.

Define MEMORY_SANITIZER when appropriate at build time and adds workarounds
to existing code to mark things as initialized where the sanitizer is otherwise unable to
determine that.  This lets our build succeed under the memory sanitizer.  not all tests
pass without sanitizer failures yet but we're in pretty good shape after this.

1 files changed, 44 insertions, 1 deletions

diff --git a/configure b/configure
index 3d66c1c..e9e33b3 100755
--- a/configure
+++ b/configure
@@ -819,6 +819,8 @@ enable_optimizations
 with_lto
 with_hash_algorithm
 with_address_sanitizer
+with_memory_sanitizer
+with_undefined_behavior_sanitizer
 with_libs
 with_system_expat
 with_system_ffi
@@ -1515,7 +1517,10 @@ Optional Packages:
   --with-hash-algorithm=[fnv|siphash24]
                           select hash algorithm
   --with-address-sanitizer
-                          enable AddressSanitizer
+                          enable AddressSanitizer (asan)
+  --with-memory-sanitizer enable MemorySanitizer (msan)
+  --with-undefined-behavior-sanitizer
+                          enable UndefinedBehaviorSanitizer (ubsan)
   --with-libs='lib1 ...'  link against additional libs
   --with-system-expat     build pyexpat module using an installed expat
                           library
@@ -10038,6 +10043,44 @@ if test "${with_address_sanitizer+set}" = set; then :
 $as_echo "$withval" >&6; }
 BASECFLAGS="-fsanitize=address -fno-omit-frame-pointer $BASECFLAGS"
 LDFLAGS="-fsanitize=address $LDFLAGS"
+# ASan works by controlling memory allocation, our own malloc interferes.
+with_pymalloc="no"
+
+else
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for --with-memory-sanitizer" >&5
+$as_echo_n "checking for --with-memory-sanitizer... " >&6; }
+
+# Check whether --with-memory_sanitizer was given.
+if test "${with_memory_sanitizer+set}" = set; then :
+  withval=$with_memory_sanitizer;
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $withval" >&5
+$as_echo "$withval" >&6; }
+BASECFLAGS="-fsanitize=memory -fsanitize-memory-track-origins=2 -fno-omit-frame-pointer $BASECFLAGS"
+LDFLAGS="-fsanitize=memory -fsanitize-memory-track-origins=2 $LDFLAGS"
+# MSan works by controlling memory allocation, our own malloc interferes.
+with_pymalloc="no"
+
+else
+  { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
+$as_echo "no" >&6; }
+fi
+
+
+{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for --with-undefined-behavior-sanitizer" >&5
+$as_echo_n "checking for --with-undefined-behavior-sanitizer... " >&6; }
+
+# Check whether --with-undefined_behavior_sanitizer was given.
+if test "${with_undefined_behavior_sanitizer+set}" = set; then :
+  withval=$with_undefined_behavior_sanitizer;
+{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $withval" >&5
+$as_echo "$withval" >&6; }
+BASECFLAGS="-fsanitize=undefined $BASECFLAGS"
+LDFLAGS="-fsanitize=undefined $LDFLAGS"
 
 else
   { $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5