MDEV-20377: Make WITH_MSAN more usable

MemorySanitizer (clang -fsanitize=memory) requires that all code
be compiled with instrumentation enabled. The only exception is the
C runtime library. Failure to use instrumented libraries will cause
bogus messages about memory being uninitialized.

In WITH_MSAN builds, we must avoid calling getservbyname(),
because even though it is a standard library function, it is
not instrumented, not even in clang 10.

Note: Before MariaDB Server 10.5, ./mtr will typically fail
due to the old PCRE library, which was updated in MDEV-14024.

The following cmake options were tested on 10.5
in commit 94d0bb4dbeb28a94d1f87fdd55f4297ff3df0157:

cmake \
-DCMAKE_C_FLAGS='-march=native -O2' \
-DCMAKE_CXX_FLAGS='-stdlib=libc++ -march=native -O2' \
-DWITH_EMBEDDED_SERVER=OFF -DWITH_UNIT_TESTS=OFF -DCMAKE_BUILD_TYPE=Debug \
-DWITH_INNODB_{BZIP2,LZ4,LZMA,LZO,SNAPPY}=OFF \
-DPLUGIN_{ARCHIVE,TOKUDB,MROONGA,OQGRAPH,ROCKSDB,CONNECT,SPIDER}=NO \
-DWITH_SAFEMALLOC=OFF \
-DWITH_{ZLIB,SSL,PCRE}=bundled \
-DHAVE_LIBAIO_H=0 \
-DWITH_MSAN=ON

MEM_MAKE_DEFINED(): An alias for VALGRIND_MAKE_MEM_DEFINED()
and __msan_unpoison().

MEM_GET_VBITS(), MEM_SET_VBITS(): Aliases for
VALGRIND_GET_VBITS(), VALGRIND_SET_VBITS(), __msan_copy_shadow().

InnoDB: Replace the UNIV_MEM_ macros with corresponding MEM_ macros.

ut_crc32_8_hw(), ut_crc32_64_low_hw(): Use the compiler built-in
functions instead of inline assembler when building WITH_MSAN.
This will require at least -msse4.2 when building for IA-32 or AMD64.
The inline assembler would not be instrumented, and would thus cause
bogus failures.

1 files changed, 28 insertions, 12 deletions

diff --git a/storage/innobase/row/row0log.cc b/storage/innobase/row/row0log.cc
index e46ee2c4f18..986cac54540 100644
--- a/storage/innobase/row/row0log.cc
+++ b/storage/innobase/row/row0log.cc
@@ -323,7 +323,9 @@ row_log_online_op(
 		goto err_exit;
 	}
 
-	UNIV_MEM_INVALID(log->tail.buf, sizeof log->tail.buf);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(log->tail.buf, sizeof log->tail.buf);
+#endif /* HAVE_valgrind_or_MSAN */
 
 	ut_ad(log->tail.bytes < srv_sort_buf_size);
 	avail_size = srv_sort_buf_size - log->tail.bytes;
@@ -373,7 +375,7 @@ row_log_online_op(
 			       log->tail.buf, avail_size);
 		}
 
-		UNIV_MEM_ASSERT_RW(buf, srv_sort_buf_size);
+		MEM_CHECK_DEFINED(buf, srv_sort_buf_size);
 
 		if (row_log_tmpfile(log) < 0) {
 			log->error = DB_OUT_OF_MEMORY;
@@ -407,8 +409,10 @@ write_failed:
 			index->type |= DICT_CORRUPT;
 		}
 
-		UNIV_MEM_INVALID(log->tail.block, srv_sort_buf_size);
-		UNIV_MEM_INVALID(buf, srv_sort_buf_size);
+#ifdef HAVE_valgrind_or_MSAN
+		MEM_UNDEFINED(log->tail.block, srv_sort_buf_size);
+		MEM_UNDEFINED(buf, srv_sort_buf_size);
+#endif /* HAVE_valgrind_or_MSAN */
 
 		memcpy(log->tail.block, log->tail.buf + avail_size,
 		       mrec_size - avail_size);
@@ -418,7 +422,9 @@ write_failed:
 		ut_ad(b == log->tail.block + log->tail.bytes);
 	}
 
-	UNIV_MEM_INVALID(log->tail.buf, sizeof log->tail.buf);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(log->tail.buf, sizeof log->tail.buf);
+#endif /* HAVE_valgrind_or_MSAN */
 err_exit:
 	mutex_exit(&log->mutex);
 }
@@ -450,7 +456,9 @@ row_log_table_open(
 {
 	mutex_enter(&log->mutex);
 
-	UNIV_MEM_INVALID(log->tail.buf, sizeof log->tail.buf);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(log->tail.buf, sizeof log->tail.buf);
+#endif /* HAVE_valgrind_or_MSAN */
 
 	if (log->error != DB_SUCCESS) {
 err_exit:
@@ -510,7 +518,7 @@ row_log_table_close_func(
 			memcpy(buf + log->tail.bytes, log->tail.buf, avail);
 		}
 
-		UNIV_MEM_ASSERT_RW(buf, srv_sort_buf_size);
+		MEM_CHECK_DEFINED(buf, srv_sort_buf_size);
 
 		if (row_log_tmpfile(log) < 0) {
 			log->error = DB_OUT_OF_MEMORY;
@@ -541,8 +549,10 @@ row_log_table_close_func(
 write_failed:
 			log->error = DB_ONLINE_LOG_TOO_BIG;
 		}
-		UNIV_MEM_INVALID(log->tail.block, srv_sort_buf_size);
-		UNIV_MEM_INVALID(buf, srv_sort_buf_size);
+#ifdef HAVE_valgrind_or_MSAN
+		MEM_UNDEFINED(log->tail.block, srv_sort_buf_size);
+		MEM_UNDEFINED(buf, srv_sort_buf_size);
+#endif /* HAVE_valgrind_or_MSAN */
 		memcpy(log->tail.block, log->tail.buf + avail, size - avail);
 		log->tail.bytes = size - avail;
 	} else {
@@ -551,7 +561,9 @@ write_failed:
 	}
 
 	log->tail.total += size;
-	UNIV_MEM_INVALID(log->tail.buf, sizeof log->tail.buf);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(log->tail.buf, sizeof log->tail.buf);
+#endif /* HAVE_valgrind_or_MSAN */
 err_exit:
 	mutex_exit(&log->mutex);
 
@@ -2557,7 +2569,9 @@ row_log_table_apply_ops(
 	ut_ad(new_trx_id_col > 0);
 	ut_ad(new_trx_id_col != ULINT_UNDEFINED);
 
-	UNIV_MEM_INVALID(&mrec_end, sizeof mrec_end);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(&mrec_end, sizeof mrec_end);
+#endif /* HAVE_valgrind_or_MSAN */
 
 	offsets = static_cast<rec_offs*>(ut_malloc_nokey(i * sizeof *offsets));
 	rec_offs_set_n_alloc(offsets, i);
@@ -3434,7 +3448,9 @@ row_log_apply_ops(
 	ut_ad(!index->is_committed());
 	ut_ad(rw_lock_own(dict_index_get_lock(index), RW_LOCK_X));
 	ut_ad(index->online_log);
-	UNIV_MEM_INVALID(&mrec_end, sizeof mrec_end);
+#ifdef HAVE_valgrind_or_MSAN
+	MEM_UNDEFINED(&mrec_end, sizeof mrec_end);
+#endif /* HAVE_valgrind_or_MSAN */
 
 	offsets = static_cast<rec_offs*>(ut_malloc_nokey(i * sizeof *offsets));
 	rec_offs_set_n_alloc(offsets, i);