kernel: sys_sleep - a cooperative delay off the DIV timer (+ ping pacing)

There was no time source at all (IRQ vectors just reti; scheduler is purely
cooperative). The DIV register (FF04) free-runs at 16384 Hz regardless of
interrupts, so sys_sleep accumulates DIV deltas across SchedYields (other procs
keep running) until the requested number of 1/64-second units elapse.

- SYS_SLEEP(34): B = 1/64s units; libc gsleep(units) / msleep(ms) wrappers.
- ping now msleep(800) between echoes, so it paces like real ping instead of
  blasting all four at once.

Verified real-time (capped emulator): replies land ~0.85s apart. In --uncapped
runs the delay is GB-time (fast wall-clock), as expected.

4 files changed, 20 insertions, 1 deletions

diff --git a/c/gbos.h b/c/gbos.h
index 6ecbd4a..390fe3a 100644
--- a/c/gbos.h
+++ b/c/gbos.h
@@ -18,6 +18,8 @@ unsigned char srecv(void);              /* link-port: receive one byte (blocks)
 int           srecv_nb(void);           /* link-port: receive, or -1 if none */
 unsigned char pollin(void);             /* poll OSK for a typed char, or 0 */
 unsigned char sys_net(void *req);       /* socket layer trap (see sock.h)  */
+void          gsleep(unsigned char units); /* delay units*(1/64)s (raw trap)  */
+void          msleep(unsigned int ms);     /* delay ~ms milliseconds          */
 
 /* helpers (c/libc.c) */
 void          putu(unsigned int n);              /* print decimal          */
diff --git a/c/libc.c b/c/libc.c
index 9da63bc..a48a64c 100644
--- a/c/libc.c
+++ b/c/libc.c
@@ -1,6 +1,15 @@
 /* gbos libc C helpers (compiled once, linked into every program). */
 #include "gbos.h"
 
+/* sleep for approximately 'ms' milliseconds (cooperative; other procs run).
+   The kernel counts in 1/64-second units (~15.6 ms), so we round ms to that. */
+void msleep(unsigned int ms) {
+    unsigned int u = ms >> 4;          /* ~ms/15.6, close enough for a delay */
+    if (u > 255) u = 255;
+    if (u == 0 && ms) u = 1;
+    gsleep((unsigned char)u);
+}
+
 /* print an unsigned int in decimal */
 void putu(unsigned int n) {
     char buf[5];
diff --git a/c/libc.s b/c/libc.s
index 724fc3b..b413f3b 100644
--- a/c/libc.s
+++ b/c/libc.s
@@ -1,7 +1,7 @@
 	.module libc
 	.globl _writes, _putc, _puts, _nl, _strlen, _readc, _sexit, _getpid, _getargs
 	.globl _open, _close, _fgetc, _fputc, _flist, _fremove, _pipe, _ssend, _srecv, _srecv_nb, _pollin
-	.globl _sys_net
+	.globl _sys_net, _gsleep
 	;; SDCC sm83 sdcccall(1): 1st arg -> A (byte) or DE (pointer); ret A / BC.
 	;; rst $30 (the trap) clobbers BC/DE/HL; SDCC treats them caller-saved, so
 	;; wrappers need not preserve them - but we compute cleanly regardless.
@@ -123,6 +123,13 @@ _sys_net:
 	rst	#0x30
 	ret
 
+	;; void gsleep(unsigned char units /*A*/) - delay units*(1/64)s
+_gsleep:
+	ld	b, a
+	ld	c, #34			; SYS_SLEEP
+	rst	#0x30
+	ret
+
 	;; unsigned char pollin(void) -> A  (typed OSK char, or 0)
 _pollin:
 	ld	c, #32			; SYS_POLLIN
diff --git a/c/ping.c b/c/ping.c
index c98f106..68c4c86 100644
--- a/c/ping.c
+++ b/c/ping.c
@@ -47,6 +47,7 @@ void main(void) {
         } else {
             puts("  seq="); putu(seq); puts(" timeout"); nl();
         }
+        if (seq < 4) msleep(800);      /* pace like real ping */
     }
     puts("-- "); putu(got); puts("/4 received"); nl();
     net_close(s);