Make improvements

- wcsstr() is now linearly complex
- strstr16() is now linearly complex
- strstr() is now vectorized on aarch64 (10x)
- strstr() now uses KMP on pathological cases
- memmem() is now vectorized on aarch64 (10x)
- memmem() now uses KMP on pathological cases
- Disable shared_ptr::owner_before until fixed
- Make iswlower(), iswupper() consistent with glibc
- Remove figure space from iswspace() implementation
- Include line and paragraph separator in iswcntrl()
- Use Musl wcwidth(), iswalpha(), iswpunct(), towlower(), towupper()

test/libc/intrin/BUILD.mk

@@ -37,15 +37,16 @@ TEST_LIBC_INTRIN_DIRECTDEPS =					\
 	LIBC_STR						\
 	LIBC_SYSV						\
 	LIBC_SYSV_CALLS						\
-	LIBC_THREAD						\
 	LIBC_TESTLIB						\
+	LIBC_THREAD						\
 	LIBC_TINYMATH						\
 	LIBC_X							\
-	TOOL_VIZ_LIB						\
 	THIRD_PARTY_COMPILER_RT					\
+	THIRD_PARTY_MUSL					\
 	THIRD_PARTY_NSYNC					\
 	THIRD_PARTY_OPENMP					\
-	THIRD_PARTY_XED
+	THIRD_PARTY_XED						\
+	TOOL_VIZ_LIB						\
 
 TEST_LIBC_INTRIN_DEPS :=					\
 	$(call uniq,$(foreach x,$(TEST_LIBC_INTRIN_DIRECTDEPS),$($(x))))

test/libc/str/memmem_test.c

@@ -17,10 +17,17 @@
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/mem/mem.h"
+#include "libc/assert.h"
+#include "libc/calls/calls.h"
 #include "libc/intrin/likely.h"
+#include "libc/intrin/safemacros.h"
 #include "libc/mem/alg.h"
+#include "libc/runtime/runtime.h"
+#include "libc/runtime/sysconf.h"
 #include "libc/stdio/rand.h"
 #include "libc/str/str.h"
+#include "libc/sysv/consts/map.h"
+#include "libc/sysv/consts/prot.h"
 #include "libc/testlib/ezbench.h"
 #include "libc/testlib/hyperion.h"
 #include "libc/testlib/testlib.h"
@@ -172,6 +179,26 @@ TEST(memmem, fuzz) {
   }
 }
 
+TEST(memmem, safety) {
+  int pagesz = sysconf(_SC_PAGESIZE);
+  char *map = (char *)mmap(0, pagesz * 2, PROT_READ | PROT_WRITE,
+                           MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
+  npassert(map != MAP_FAILED);
+  npassert(!mprotect(map + pagesz, pagesz, PROT_NONE));
+  for (int haylen = 1; haylen < 128; ++haylen) {
+    char *hay = map + pagesz - (haylen + 1);
+    for (int i = 0; i < haylen; ++i)
+      hay[i] = max(rand() & 255, 1);
+    hay[haylen] = 0;
+    for (int neelen = 1; neelen < haylen; ++neelen) {
+      char *nee = hay + (haylen + 1) - (neelen + 1);
+      ASSERT_EQ(memmem_naive(hay, haylen, nee, neelen),
+                memmem(hay, haylen, nee, neelen));
+    }
+  }
+  munmap(map, pagesz * 2);
+}
+
 /*
  *     memmem naive        l:    43,783c    14,142ns   m:    31,285c    10,105ns
  *     memmem              l:     2,597c       839ns   m:     2,612c       844ns
@@ -201,7 +228,12 @@ BENCH(memmem, bench) {
   EZBENCH2("memmem", donothing,
            __expropriate(memmem(kHyperion, kHyperionSize, "THE END", 7)));
   EZBENCH2("memmem", donothing,
-           __expropriate(memmem(
-               "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaab",
-               62, "aaaaaab", 7)));
+           __expropriate(
+               memmem("aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
+                      "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
+                      "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaab",
+                      152,
+                      "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa"
+                      "aaaaaaaaaaaaaaaaaaaaaaaab",
+                      81)));
 }

test/libc/str/strstr_test.c

@@ -17,11 +17,23 @@
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/str/str.h"
+#include "libc/assert.h"
+#include "libc/calls/calls.h"
 #include "libc/dce.h"
+#include "libc/intrin/kprintf.h"
+#include "libc/intrin/safemacros.h"
 #include "libc/mem/alg.h"
 #include "libc/mem/gc.h"
 #include "libc/mem/mem.h"
 #include "libc/nexgen32e/x86feature.h"
+#include "libc/runtime/runtime.h"
+#include "libc/runtime/sysconf.h"
+#include "libc/stdalign.h"
+#include "libc/stdio/rand.h"
+#include "libc/stdio/stdio.h"
+#include "libc/stdio/sysparam.h"
+#include "libc/sysv/consts/map.h"
+#include "libc/sysv/consts/prot.h"
 #include "libc/testlib/ezbench.h"
 #include "libc/testlib/hyperion.h"
 #include "libc/testlib/testlib.h"
@@ -48,6 +60,13 @@ char *strstr_naive(const char *haystack, const char *needle) {
   return 0;
 }
 
+TEST(strstr, special) {
+  MAKESTRING(haystack, "abc123def");
+  ASSERT_STREQ(&haystack[0], strstr(haystack, haystack));
+  ASSERT_STREQ(&haystack[0], strstr(haystack, ""));
+  free(haystack);
+}
+
 TEST(strstr, test_emptyString_isFoundAtBeginning) {
   MAKESTRING(haystack, "abc123def");
   ASSERT_STREQ(&haystack[0], strstr(haystack, gc(strdup(""))));
@@ -67,7 +86,8 @@ TEST(strstr, test_notFound1) {
 }
 
 TEST(strstr, test_middleOfString) {
-  MAKESTRING(haystack, "abc123def");
+  alignas(16) char hog[] = "abc123def";
+  MAKESTRING(haystack, hog);
   ASSERT_STREQ(&haystack[3], strstr(haystack, gc(strdup("123"))));
   free(haystack);
 }
@@ -98,6 +118,25 @@ TEST(strstr, test) {
   ASSERT_STREQ("x", strstr("x", "x"));
 }
 
+TEST(strstr, safety) {
+  int pagesz = sysconf(_SC_PAGESIZE);
+  char *map = (char *)mmap(0, pagesz * 2, PROT_READ | PROT_WRITE,
+                           MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
+  npassert(map != MAP_FAILED);
+  npassert(!mprotect(map + pagesz, pagesz, PROT_NONE));
+  for (int haylen = 1; haylen < 128; ++haylen) {
+    char *hay = map + pagesz - (haylen + 1);
+    for (int i = 0; i < haylen; ++i)
+      hay[i] = max(rand() & 255, 1);
+    hay[haylen] = 0;
+    for (int neelen = 1; neelen < haylen; ++neelen) {
+      char *nee = hay + (haylen + 1) - (neelen + 1);
+      ASSERT_EQ(strstr_naive(hay, nee), strstr(hay, nee));
+    }
+  }
+  munmap(map, pagesz * 2);
+}
+
 TEST(strstr, breakit) {
   char *p;
   p = gc(calloc(1, 32));

test/libc/str/towupper_test.c

@@ -30,7 +30,7 @@ TEST(towupper, test) {
   EXPECT_EQ(u'!', towupper(u'!'));
   EXPECT_EQ(u'A', towupper(u'a'));
   EXPECT_EQ(u'À', towupper(u'à'));
-  EXPECT_EQ(L'𝛥', towupper(L'𝛿'));
+  /* EXPECT_EQ(L'𝛥', towupper(L'𝛿')); */
   EXPECT_EQ(L'Ｂ', towupper(L'ｂ'));
   EXPECT_EQ(u'Ꭰ', towupper(u'ꭰ'));
 }
@@ -39,7 +39,7 @@ TEST(towlower, test) {
   EXPECT_EQ(u'!', towlower(u'!'));
   EXPECT_EQ(u'a', towlower(u'A'));
   EXPECT_EQ(u'à', towlower(u'À'));
-  EXPECT_EQ(L'𝛿', towlower(L'𝛥'));
+  /* EXPECT_EQ(L'𝛿', towlower(L'𝛥')); */
   EXPECT_EQ(L'ｂ', towlower(L'Ｂ'));
   EXPECT_EQ(u'ꭰ', towlower(u'Ꭰ'));
 }

test/libc/str/wcwidth_test.c

@@ -16,9 +16,11 @@
 │ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR             │
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
+#include "libc/stdio/stdio.h"
 #include "libc/str/str.h"
 #include "libc/str/strwidth.h"
 #include "libc/str/unicode.h"
+#include "libc/testlib/benchmark.h"
 #include "libc/testlib/ezbench.h"
 #include "libc/testlib/testlib.h"
 
@@ -28,6 +30,7 @@ TEST(wcwidth, test) {
   ASSERT_EQ(-1, wcwidth(-7));
   ASSERT_EQ(1, wcwidth(0x10FFFD));
   ASSERT_EQ(-1, wcwidth(0x10FFFD + 1));
+  ASSERT_EQ(2, wcwidth(L'😀'));
 }
 
 TEST(strwidth, testCjkWidesAndCombiningLowLines_withThompsonPikeEncoding) {
@@ -74,6 +77,12 @@ TEST(strwidth, testTextDelimitingControlCodes_dontHaveSubstance) {
   EXPECT_EQ(0, strwidth("\1", 0));
 }
 
+#define WCWIDTH(x) __expropriate(wcwidth(__veil("r", x)))
+
 BENCH(wcwidth, bench) {
-  EZBENCH2("wcwidth", donothing, __expropriate(wcwidth(__veil("r", u'→'))));
+  BENCHMARK(1000, 1, WCWIDTH(u'a'));
+  BENCHMARK(1000, 1, WCWIDTH(u'a'));
+  BENCHMARK(1000, 1, WCWIDTH(u'→'));
+  BENCHMARK(1000, 1, WCWIDTH(L'😀'));
+  BENCHMARK(1000, 1, WCWIDTH(0));
 }