feat: add streq

user/namedreader.c

@@ -0,0 +1,60 @@
+#include "ulib.h"
+
+/*
+ * Boot-spawned independently of namedwriter.c (see kmain.c's spawnUser
+ * calls) -- unlike fiforeader.c/fifowriter.c, which fifomaster.c spawns as
+ * its own children and hands a shared pipe.h handle to via argv, this
+ * process and namedwriter.c have no relationship at all beyond both being
+ * launched at boot: neither one's pid, argv, or fd table ever mentions the
+ * other. The only thing they share is agreeing on the path "/tmp/rendezvous"
+ * (mkfifo'd by kmain.c before either is spawned) -- proving vfsMkfifo/
+ * vfsOpen (vfs.h) really do let two unrelated processes rendezvous purely
+ * through the filesystem namespace, the gap FALA 11 explicitly left open.
+ * Opens for read immediately, well before namedwriter.c's deliberate sleep
+ * elapses, so the blocking read() below has to genuinely park across many
+ * timer ticks and be woken later -- not just win a race against data that
+ * was already sitting in the pipe.
+ */
+
+#define namedPipePath "/tmp/rendezvous"
+
+static int streq(const char* a, const char* b)
+{
+    while (*a && *a == *b) {
+        a++;
+        b++;
+    }
+    return *a == *b;
+}
+
+int main(void)
+{
+    print("namedreader: start, pid=");
+    printDec(userGetpid());
+    print("\n");
+
+    long fd = userOpen(namedPipePath, oReadOnly);
+    if (fd < 0) {
+        print("namedreader: FAIL open\n");
+        return 1;
+    }
+
+    print("namedreader: opened, blocking on read\n");
+    char buffer[64];
+    long got = userRead((int)fd, buffer, sizeof(buffer) - 1);
+    if (got <= 0) {
+        print("namedreader: FAIL read\n");
+        return 1;
+    }
+    buffer[got] = 0;
+
+    print("namedreader: got: ");
+    print(buffer);
+    print("\n");
+
+    int ok = streq(buffer, "hello via named pipe");
+    print(ok ? "namedreader: payload-ok\n" : "namedreader: payload-FAIL\n");
+
+    userClose((int)fd);
+    return ok ? 0 : 1;
+}