ww/test/wcc/989_m1union_run.c

/*
 * 989_m1union_run — M1 (#199b): a `match` on a tagged union returned by a
 * CROSS-MODULE fn must keep each nominal variant on a DISTINCT tag. The
 * canonical case is `utf8.next` (encoding.utf8), whose return
 * `(rune | done | more | invalid)` has three structurally-identical
 * void-alias variants (done/more = void, invalid = !void). Pre-#199b the
 * wwstage collapsed done/more/invalid onto one tag (a byte-id-only
 * miscompile: the doc's repro hit the rune arm and returned the same value
 * either way). This test is GATE-VISIBLE: each arm maps to a distinct
 * return code and the inputs deterministically hit each arm, so a collapse
 * mis-routes an arm and changes the runtime exit — caught on BOTH stages.
 *
 *   row          | input            | arm       | want_exit
 *   -------------+------------------+-----------+----------
 *   rune         | [65]             | rune      | 1
 *   done         | [65] next twice  | done      | 2
 *   more         | [195] (lead-only)| more      | 3
 *   invalid      | [255]            | invalid   | 4
 *   all_arms     | all four, mixed  | (1,2,4,3) | 219  (1243 % 256)
 *
 * Runtime ownership is now r989_m1union_*; this wrapper retains all_arms
 * solely for cstage.s == wwstage.s across the separate-package build.
 */
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/wait.h>

static int
runwait(const char *cmd)
{
	int rc = system(cmd);
	if (rc == -1) return -1;
	if (WIFEXITED(rc)) return WEXITSTATUS(rc);
	return -1;
}

/* classify() is shared boilerplate: a cross-module match whose four arms
 * each return a distinct code. Each row supplies only the main body. */
#define PRELUDE \
	"package main;\n" \
	"import encoding.utf8;\n" \
	"fn classify(d: *utf8.decoder) int = {\n" \
	"	match (utf8.next(d)) {\n" \
	"	case let r: rune          => return 1;\n" \
	"	case let dn: utf8.done    => return 2;\n" \
	"	case let m:  utf8.more    => return 3;\n" \
	"	case let e:  utf8.invalid => return 4;\n" \
	"	};\n" \
	"	return 0;\n" \
	"};\n"

struct row {
	const char *label;
	const char *body;	/* main() body, appended to PRELUDE */
	int want_exit;
};

static const struct row rows[] = {
	/* Runtime rows moved to r989_m1union_*; retain byte-id source only. */
	{ "all_arms",
	  "export fn main() int = {\n"
	  "\tlet g: []u8 = [65u8];\n"
	  "\tlet dg = utf8.decode(g);\n"
	  "\tlet a = classify(&dg);\n"
	  "\tlet b = classify(&dg);\n"
	  "\tlet bad: []u8 = [255u8];\n"
	  "\tlet db = utf8.decode(bad);\n"
	  "\tlet c = classify(&db);\n"
	  "\tlet tr: []u8 = [195u8];\n"
	  "\tlet dt = utf8.decode(tr);\n"
	  "\tlet e = classify(&dt);\n"
	  "\treturn a * 1000 + b * 100 + c * 10 + e;\n"
	  "};\n", 219 },
};

/* write_src — PRELUDE + row body into <path>. */
static int
write_src(const char *path, const struct row *r)
{
	FILE *f = fopen(path, "wb");
	if (!f) return -1;
	fputs(PRELUDE, f);
	fputs(r->body, f);
	fclose(f);
	return 0;
}

int
main(void)
{
	const char *bin = getenv("BIN");
	if (!bin) bin = "out/bin";
	char absbin[1024];
	if (bin[0] != '/') {
		char cwd[1024];
		if (getcwd(cwd, sizeof cwd) == NULL) return 1;
		snprintf(absbin, sizeof absbin, "%s/%s", cwd, bin);
		bin = absbin;
	}
	char cdrv[1024], wdrv[1024];
	snprintf(cdrv, sizeof cdrv, "%s/ww", bin);
	snprintf(wdrv, sizeof wdrv, "%s/ww_ww", bin);

	int n = (int)(sizeof rows / sizeof rows[0]);
	int total = 0, fail = 0;
	int pid = (int)getpid();

	/* rule-10 byte-id gate on the all_arms cross-module match. */
	if (access(wdrv, X_OK) == 0) {
		total++;
		const struct row *r = &rows[n - 1];	/* all_arms */
		char csrc[80], cstem[80], cmd[512];
		snprintf(csrc, sizeof csrc, "/tmp/m1union_bid_%d.ww", pid);
		snprintf(cstem, sizeof cstem, "/tmp/m1union_bid_c_%d", pid);
		char wstem[80];
		snprintf(wstem, sizeof wstem, "/tmp/m1union_bid_w_%d", pid);
		if (write_src(csrc, r) != 0) {
			/* a silent skip here would print "1/1 ok" having
			 * asserted nothing; nothing was created, so no
			 * cleanup is owed either. */
			fprintf(stderr, "m1union_run: cannot write %s\n",
			    csrc);
			fail++;
		} else {
			/* #93 sep layout: each stage emits per-unit asm under
			 * <stem>.sepwork/ (__root.s + encoding.utf8.s); the
			 * rule-10 byte-id gate becomes a recursive diff of the
			 * two sepwork trees. */
			snprintf(cmd, sizeof cmd,
			    "%s build -S -o '%s' '%s' "
			    "2>/dev/null", cdrv, cstem, csrc);
			int cb = runwait(cmd);
			snprintf(cmd, sizeof cmd,
			    "%s build -S -o '%s' '%s' "
			    "2>/dev/null", wdrv, wstem, csrc);
			int wb = runwait(cmd);
			if (cb != 0 || wb != 0) {
				fprintf(stderr, "m1union_run: byte-id build "
				    "failed (cstage=%d wwstage=%d)\n", cb, wb);
				fail++;
			} else {
				/* Concat each tree's per-unit *.s (sorted glob
				 * order is deterministic) and cmp — sidesteps a
				 * stray <pkg>.md5tmp the wwstage drops in the
				 * sep dir, which a full `diff -r` would flag. */
				snprintf(cmd, sizeof cmd,
				    "cat '%s.sepwork/'*.s > '%s.scat' && "
				    "cat '%s.sepwork/'*.s > '%s.scat' && "
				    "cmp -s '%s.scat' '%s.scat'",
				    cstem, cstem, wstem, wstem, cstem, wstem);
				if (runwait(cmd) != 0) {
					fprintf(stderr, "m1union_run: cstage.s "
					    "!= wwstage.s (byte-id break)\n");
					fail++;
				}
			}
			/* exact owned paths only (`build -S` writes solely
			 * under <stem>.sepwork/, plus the .scat concats) —
			 * no prefix glob; a failed rm must fail the carrier,
			 * not leak silently. */
			snprintf(cmd, sizeof cmd,
			    "rm -rf '%s' '%s.scat' '%s.sepwork' "
			    "'%s.scat' '%s.sepwork'",
			    csrc, cstem, cstem, wstem, wstem);
			if (runwait(cmd) != 0) {
				fprintf(stderr,
				    "m1union_run: cleanup rm failed\n");
				fail++;
			}
		}
	} else {
		fprintf(stderr, "m1union_run: skip byte-id (no ww_ww)\n");
	}

	if (fail) {
		fprintf(stderr, "m1union_run: %d/%d failed\n", fail, total);
		return 1;
	}
	printf("m1union_run: %d/%d ok\n", total, total);
	return 0;
}