Introduce #include <cosmo.h> to toolchain users

This change improves the way internal APIs are being hidden behind the
`COSMO` define. The cosmo.h header will take care of defining that, so
that a separate define statement isn't needed. This change also does a
lot more to define which APIs are standard, and which belong to Cosmo.

libc/nexgen32e/loadxmm.S

@@ -17,7 +17,6 @@
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/macros.internal.h"
-.privileged
 
 //	Loads XMM registers from buffer.
 //

libc/nexgen32e/longerjmp.S

@@ -17,14 +17,16 @@
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/macros.internal.h"
-.privileged
 
 //	Loads previously saved processor state.
 //
 //	@param	rdi points to the jmp_buf
 //	@param	rsi is returned by setlongerjmp() invocation
 //	@noreturn
+	.ftrace1
 longerjmp:
+	.ftrace2
+#ifdef __x86_64__
 	mov	$1,%eax
 	mov	%rsi,%rdx
 	mov	(%rdi),%rsp
@@ -35,4 +37,22 @@ longerjmp:
 	mov	40(%rdi),%r14
 	mov	48(%rdi),%r15
 	jmp	*56(%rdi)
+#elif defined(__aarch64__)
+	ldp	x19,x20,[x0,#0]
+	ldp	x21,x22,[x0,#16]
+	ldp	x23,x24,[x0,#32]
+	ldp	x25,x26,[x0,#48]
+	ldp	x27,x28,[x0,#64]
+	ldp	x29,x30,[x0,#80]
+	ldr	x2,[x0,#104]
+	mov	sp,x2
+	ldp	d8 ,d9,[x0,#112]
+	ldp	d10,d11,[x0,#128]
+	ldp	d12,d13,[x0,#144]
+	ldp	d14,d15,[x0,#160]
+	mov	x0,x1
+	br	x30
+#else
+#error "unsupported architecture"
+#endif
 	.endfn	longerjmp,globl

libc/nexgen32e/longjmp.S

@@ -57,6 +57,8 @@ longjmp:
 	cmp	w1,0
 	csinc	w0,w1,wzr,ne
 	br	x30
+#else
+#error "unsupported architecture"
 #endif
 	.endfn	longjmp,globl
 	.alias	longjmp,_longjmp

libc/nexgen32e/savexmm.S

@@ -17,7 +17,6 @@
 │ PERFORMANCE OF THIS SOFTWARE.                                                │
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/macros.internal.h"
-.privileged
 
 //	Stores XMM registers to buffer.
 //

libc/nexgen32e/setjmp.S

@@ -27,7 +27,8 @@
 //	@note	code built w/ microsoft abi compiler can't call this
 //	@see	longjmp(), _gclongjmp()
 	.ftrace1
-setjmp:	.ftrace2
+setjmp:
+	.ftrace2
 #ifdef __x86_64__
 	lea	8(%rsp),%rax
 	mov	%rax,(%rdi)
@@ -40,6 +41,7 @@ setjmp:	.ftrace2
 	mov	(%rsp),%rax
 	mov	%rax,56(%rdi)
 	xor	%eax,%eax
+	ret
 #elif defined(__aarch64__)
 	stp	x19,x20,[x0,#0]
 	stp	x21,x22,[x0,#16]
@@ -54,7 +56,9 @@ setjmp:	.ftrace2
 	stp	d12,d13,[x0,#144]
 	stp	d14,d15,[x0,#160]
 	mov	x0,#0
-#endif
 	ret
+#else
+#error "unsupported architecture"
+#endif
 	.endfn	setjmp,globl
 	.alias	setjmp,_setjmp

libc/nexgen32e/setlongerjmp.S

@@ -24,7 +24,10 @@
 //	@return	eax contains 0 when set, and 1 if jumped
 //	@return	rdx contains value passed to longerjmp()
 //	@returnstwice
+	.ftrace1
 setlongerjmp:
+	.ftrace2
+#ifdef __x86_64__
 	lea	8(%rsp),%rax
 	mov	%rax,(%rdi)
 	mov	%rbx,8(%rdi)
@@ -38,4 +41,23 @@ setlongerjmp:
 	xor	%eax,%eax
 	xor	%edx,%edx
 	ret
+#elif defined(__aarch64__)
+	stp	x19,x20,[x0,#0]
+	stp	x21,x22,[x0,#16]
+	stp	x23,x24,[x0,#32]
+	stp	x25,x26,[x0,#48]
+	stp	x27,x28,[x0,#64]
+	stp	x29,x30,[x0,#80]
+	mov	x2,sp
+	str	x2,[x0,#104]
+	stp	d8,d9,[x0,#112]
+	stp	d10,d11,[x0,#128]
+	stp	d12,d13,[x0,#144]
+	stp	d14,d15,[x0,#160]
+	mov	x0,#0
+	mov	x1,#0
+	ret
+#else
+#error "unsupported architecture"
+#endif
 	.endfn	setlongerjmp,globl

libc/nexgen32e/zip.S

@@ -18,7 +18,7 @@
 ╚─────────────────────────────────────────────────────────────────────────────*/
 #include "libc/macros.internal.h"
 #include "ape/relocations.h"
-#include "libc/zip.h"
+#include "libc/zip.internal.h"
 
 //	ZIP Central Directory.
 	.section .zip.3,"",@progbits