Make improvements

- Emulator can now test the αcτµαlly pδrταblε εxεcµταblε bootloader

- Whipped up a webserver named redbean. It services 150k requests per
  second on a single core. Bundling assets inside zip enables extremely
  fast serving for two reasons. The first is that zip central directory
  lookups go faster than stat() system calls. The second is that both
  zip and gzip content-encoding use DEFLATE, therefore, compressed
  responses can be served via the sendfile() system call which does an
  in-kernel copy directly from the zip executable structure. Also note
  that red bean zip executables can be deployed easily to all platforms,
  since these native executables work on Linux, Mac, BSD, and Windows.

- Address sanitizer now works very well

libc/tinymath/copysignl.S

@@ -17,6 +17,7 @@
 │ Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA                │
 │ 02110-1301 USA                                                               │
 ╚─────────────────────────────────────────────────────────────────────────────*/
+#include "ape/lib/pc.h"
 #include "libc/macros.h"
 .source	__FILE__
 
@@ -29,7 +30,7 @@ tinymath_copysignl:
 	fnstsw
 	fstp	%st
 	fldt	16(%rbp)
-	testb	$2,%ah
+	test	$FPU_C1>>8,%ah
 	fabs
 	je	1f
 	fchs

libc/tinymath/lroundl.S

@@ -26,22 +26,14 @@ tinymath_lroundl:
 	mov	%rsp,%rbp
 	.profilable
 	push	%rax
-	push	%rax
 	fldt	16(%rbp)
-	fnstcw	-8(%rbp)
-	movzwl	-8(%rbp),%edx
-	and	$0b11110011,%dh			# RC (Rounding Control)
-	or	$0b00000100,%dh			# →-∞
-	mov	%dx,-4(%rbp)
 	fxam
 	fnstsw
 	fabs
 	test	$FPU_C1>>8,%ah
 	fadds	.Lhalf(%rip)
-	fldcw	-4(%rbp)
-	fistpq	-16(%rbp)
-	fldcw	-8(%rbp)
-	mov	-16(%rbp),%rax
+	fisttpq	-8(%rbp)
+	mov	-8(%rbp),%rax
 	je	1f
 	neg	%rax
 1:	leave

libc/tinymath/round.S

@@ -26,7 +26,7 @@
 /
 /	@param	𝑥 is double scalar in low half of %xmm0
 /	@return	double scalar in low half of %xmm0
-/	@define	round(𝑥) = copysign(floor(fabs(𝑥)+.5),𝑥)
+/	@define	round(𝑥) = copysign(trunc(fabs(𝑥)+.5),𝑥)
 /		round(𝑥) = trunc(𝑥+copysign(.5,𝑥))
 tinymath_round:
 #if !X86_NEED(SSE4_2)

libc/tinymath/roundl.S

@@ -30,19 +30,17 @@ tinymath_roundl:
 	.profilable
 	push	%rax
 	fldt	16(%rbp)
+	fnstcw	-6(%rbp)
 	fnstcw	-8(%rbp)
-	movzwl	-8(%rbp),%edx
-	and	$0b11110011,%dh			# RC (Rounding Control)
-	or	$0b00000100,%dh			# →-∞ a.k.a. floor()
-	mov	%dx,-4(%rbp)
+	orb	$0b00001100,-7(%rbp)		# RC = →0
 	fxam					# C1 is set to sign bit
 	fnstsw
 	fabs
 	test	$FPU_C1>>8,%ah
 	fadds	.Lhalf(%rip)
-	fldcw	-4(%rbp)
-	frndint
 	fldcw	-8(%rbp)
+	frndint
+	fldcw	-6(%rbp)
 	je	1f
 	fchs
 1:	leave