Get --ftrace working on aarch64

This change implements a new approach to function call logging, that's
based on the GCC flag: -fpatchable-function-entry. Read the commentary
in build/config.mk to learn how it works.

libc/str/str.mk

@@ -51,7 +51,7 @@ o/$(MODE)/libc/str/wow.o: private				\
 o/$(MODE)/libc/str/wmemset.o					\
 o/$(MODE)/libc/str/memset16.o					\
 o/$(MODE)/libc/str/dosdatetimetounix.o: private			\
-		OVERRIDE_CFLAGS +=				\
+		CFLAGS +=					\
 			-O3
 
 o/$(MODE)/libc/str/getzipcdir.o					\
@@ -66,14 +66,14 @@ o/$(MODE)/libc/str/getzipcfileuncompressedsize.o		\
 o/$(MODE)/libc/str/getziplfilecompressedsize.o			\
 o/$(MODE)/libc/str/getziplfileuncompressedsize.o		\
 o/$(MODE)/libc/str/getzipcfiletimestamps.o: private		\
-		OVERRIDE_CFLAGS +=				\
+		CFLAGS +=					\
 			-Os
 
 o/$(MODE)/libc/str/iswpunct.o					\
 o/$(MODE)/libc/str/iswupper.o					\
 o/$(MODE)/libc/str/iswlower.o					\
 o/$(MODE)/libc/str/iswseparator.o: private			\
-		OVERRIDE_CFLAGS +=				\
+		CFLAGS +=					\
 			-fno-jump-tables
 
 o/$(MODE)/libc/str/bcmp.o					\
@@ -84,13 +84,13 @@ o/$(MODE)/libc/str/timevaltowindowstime.o			\
 o/$(MODE)/libc/str/timespectowindowstime.o			\
 o/$(MODE)/libc/str/windowstimetotimeval.o			\
 o/$(MODE)/libc/str/windowstimetotimespec.o: private		\
-		OVERRIDE_CFLAGS +=				\
+		CFLAGS +=					\
 			-O2
 
 # we can't use compiler magic because:
 #   kprintf() depends on these functions
 o/$(MODE)/libc/fmt/strsignal.greg.o: private			\
-		OVERRIDE_CFLAGS +=				\
+		CFLAGS +=					\
 			-fpie					\
 			-ffreestanding				\
 			$(NO_MAGIC)

libc/str/strcspn.c

@@ -16,16 +16,8 @@
 │ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR             │
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
-#include "libc/intrin/pcmpeqb.h"
-#include "libc/intrin/pmovmskb.h"
-#include "libc/intrin/pshufd.h"
-#include "libc/intrin/punpcklbw.h"
-#include "libc/intrin/punpcklwd.h"
-#include "libc/nexgen32e/hascharacter.internal.h"
 #include "libc/str/str.h"
 
-#define V(p) (void *)(p)
-
 /**
  * Returns prefix length, consisting of chars not in reject.
  * a.k.a. Return index of first byte that's in charset.
@@ -35,28 +27,24 @@
  * @asyncsignalsafe
  */
 size_t strcspn(const char *s, const char *reject) {
-  size_t i, n;
-  unsigned m;
-  char cv[16], sv[16];
-  if ((n = strlen(reject)) < 16) {
-    bzero(sv, 16);
-    memcpy(sv, reject, n);
-    for (i = 0;; ++i) {
-      cv[0] = s[i];
-      punpcklbw(V(cv), V(cv), V(cv));
-      punpcklwd(V(cv), V(cv), V(cv));
-      pshufd(V(cv), V(cv), 0);
-      pcmpeqb(V(cv), V(cv), V(sv));
-      if ((m = pmovmskb(V(cv)))) {
-        break;
-      }
-    }
-    return i;
+  int c;
+  size_t i;
+  bool lut[256];
+#ifndef TINY
+  if (!reject[0]) {
+    return strlen(s);
   }
-  for (i = 0; s[i]; ++i) {
-    if (HasCharacter(s[i], reject)) {
-      break;
+  if (!reject[1]) {
+    return strchrnul(s, reject[0]) - s;
+  }
+#endif
+  bzero(lut, sizeof(lut));
+  do {
+    lut[(c = *reject++ & 255)] = true;
+  } while (c);
+  for (i = 0;; i++) {
+    if (lut[s[i] & 255]) {
+      return i;
     }
   }
-  return i;
 }

libc/str/strspn.c

@@ -16,7 +16,6 @@
 │ TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR             │
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
-#include "libc/nexgen32e/hascharacter.internal.h"
 #include "libc/str/str.h"
 
 /**
@@ -27,11 +26,28 @@
  * @asyncsignalsafe
  */
 size_t strspn(const char *s, const char *accept) {
+  int c;
   size_t i;
-  for (i = 0; s[i]; ++i) {
-    if (!HasCharacter(s[i], accept)) {
-      break;
+  bool lut[256];
+#ifndef TINY
+  if (!accept[0]) {
+    return 0;
+  }
+  if (!accept[1]) {
+    for (i = 0;; i++) {
+      if (s[i] != accept[0]) {
+        return i;
+      }
+    }
+  }
+#endif
+  bzero(lut, sizeof(lut));
+  while ((c = *accept++ & 255)) {
+    lut[c] = true;
+  }
+  for (i = 0;; i++) {
+    if (!lut[s[i] & 255]) {
+      return i;
     }
   }
-  return i;
 }