8dc6a014

test: the scrollback fixture types only after both resizes have landed

a73x 2026-09-03 19:25

Commit message

test: the scrollback fixture types only after both resizes have landed

macOS's /bin/sh is bash, and its readline redraws the input line when
SIGWINCH arrives: carriage return, erase to end of line, prompt again. The
test typed immediately after the second attach, so the resize landed mid-echo
and the history row it selects from lost its first characters — the reply
read "q 1 100" where the test wanted "seq 1 100". Waiting for a state frame
on each fd proves seatClient already resized, since it resizes before it
answers. Linux's dash has no readline and redraws nothing.

The unread-observer test now caps the buffer on both ends. Linux holds
AF_UNIX bytes against the sender's send buffer; Darwin hands them to the
peer and charges the receiver's, so capping only the daemon's end left room
for the whole burst and no drop to grade.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SakwJEwD9dXBoRP5kWbemW

src/server/server_test_attach.zig

Old		New
@@ -539,6 +539,18 @@ test "Server: scrollback fetch is per-client and independent" {
539	const b = try dial.dialAttach(td.sock_path, 100, 30);	539	const b = try dial.dialAttach(td.sock_path, 100, 30);
540	defer b.close();	540	defer b.close();
541		541
		542	// Both attaches must have REACHED the pty before anything is typed.
		543	// `seatClient` resizes before it answers, so a state frame on each fd is
		544	// proof that the SIGWINCH for that attach is already out. Without the
		545	// wait, macOS's /bin/sh is bash and its readline takes the resize MID-ECHO
		546	// and redraws the input line over itself — carriage return, erase to end
		547	// of line, prompt again — so the history row this test selects from lost
		548	// its first characters and every column after them moved. Linux's dash has
		549	// no readline and redraws nothing, which is why the race only ever showed
		550	// on the Mac.
		551	_ = try h.firstStateFrame(alloc, a.handle, 10_000);
		552	_ = try h.firstStateFrame(alloc, b.handle, 10_000);
		553
542	// Typed by B; both clients see the history it produces.	554	// Typed by B; both clients see the history it produces.
543	try proto.writeFrame(b.handle, .input, "seq 1 100\n");	555	try proto.writeFrame(b.handle, .input, "seq 1 100\n");
544		556
@@ -1386,12 +1398,16 @@ test "Server: an observer that never reads its replies is dropped, not allowed t
1386	defer peer.close();	1398	defer peer.close();
1387	try td.srv.pumpOnce(20);	1399	try td.srv.pumpOnce(20);
1388	// Every reply is up to sessions_text_len bytes and the peer never reads	1400	// Every reply is up to sessions_text_len bytes and the peer never reads
1389	// one, so a small send buffer is what an unread socket looks like a few	1401	// one, so a small buffer is what an unread socket looks like a few frames
1390	// frames sooner. Set on the DAEMON's end: for AF_UNIX it is the sender's	1402	// sooner. Set on BOTH ends, because the two kernels put the bytes in
1391	// buffer that fills.	1403	// different places: for AF_UNIX Linux holds them against the SENDER's
		1404	// send buffer, while Darwin hands them to the peer and charges the
		1405	// RECEIVER's receive buffer, so capping only the daemon's end left the
		1406	// Mac with room for the whole burst and no drop to grade.
1392	const fd = td.srv.observers[0].?.fd;	1407	const fd = td.srv.observers[0].?.fd;
1393	const sndbuf: c_int = 4096;	1408	const small: c_int = 4096;
1394	try std.posix.setsockopt(fd, std.posix.SOL.SOCKET, std.posix.SO.SNDBUF, std.mem.asBytes(&sndbuf));	1409	try std.posix.setsockopt(fd, std.posix.SOL.SOCKET, std.posix.SO.SNDBUF, std.mem.asBytes(&small));
		1410	try std.posix.setsockopt(peer.handle, std.posix.SOL.SOCKET, std.posix.SO.RCVBUF, std.mem.asBytes(&small));
1395		1411
1396	// One write, hundreds of frames: they fit in a single 64 KB read, which	1412	// One write, hundreds of frames: they fit in a single 64 KB read, which
1397	// is what makes the whole burst one uninterrupted drain.	1413	// is what makes the whole burst one uninterrupted drain.