Pass argv to processes on a SysV entry stack; grow the user stack to 32 KiB
Processes now start with C-compatible arguments: the kernel builds the System V AMD64 entry block (argc, argv, empty envp, auxiliary vector) at the top of the stack, argv[0] is the path or initial-ramdisk name the process was spawned as, and system_spawn carries an optional NUL-separated blob that becomes argv[1..]. The runtime parses the block (runtime.argumentCount/argument) and its spawn wrappers pass arguments through. The name is also recorded on the task, so a fault report says which binary died, not just its id. The user stack grows from one page to eight (32 KiB, parameters.user_stack_pages), with the page below left unmapped as a guard so an overflow faults into a clean process kill rather than corrupting the image. Task.name_buffer is zero-initialised, not undefined: an undefined default is materialised as a 0xAA fill that moved the static task pool out of .bss and made the whole kernel ~7x slower under QEMU TCG (caught by the affinity test). Proven end to end by the new args test: args-echo respawns itself with arguments via the syscall blob, burns more stack than one page could hold, and echoes its argv intact. Full suite: 44/44.
This commit is contained in:
@@ -299,6 +299,7 @@ pub fn build(b: *std.Build) void {
|
||||
const hpet_exe = addUserBinary(b, kernel_target, runtime_module, posix_module, mmio_module, "hpet", "system/drivers/hpet/hpet.zig");
|
||||
const bus_exe = addUserBinary(b, kernel_target, runtime_module, posix_module, mmio_module, "bus", "system/drivers/bus/bus.zig");
|
||||
const device_manager_exe = addUserBinary(b, kernel_target, runtime_module, posix_module, mmio_module, "device-manager", "system/services/device-manager/device-manager.zig");
|
||||
const args_echo_exe = addUserBinary(b, kernel_target, runtime_module, posix_module, mmio_module, "args-echo", "system/services/args-echo/args-echo.zig");
|
||||
|
||||
// Pack the user binaries into the initial_ramdisk image with the host-side Python tool
|
||||
// (the container format is trivial, and Python sidesteps std API churn). Args:
|
||||
@@ -316,6 +317,8 @@ pub fn build(b: *std.Build) void {
|
||||
mk_run.addFileArg(bus_exe.getEmittedBin());
|
||||
mk_run.addArg("device-manager");
|
||||
mk_run.addFileArg(device_manager_exe.getEmittedBin());
|
||||
mk_run.addArg("args-echo");
|
||||
mk_run.addFileArg(args_echo_exe.getEmittedBin());
|
||||
|
||||
// Also install the packed binaries to their FHS homes, so zig-out is a true image
|
||||
// of the filesystem — even though at boot they arrive inside the initial-ramdisk.
|
||||
|
||||
@@ -164,8 +164,11 @@ If a class driver needs `mmio`, it has become an HCD and should be one.
|
||||
DMA is still unprotected** (the caveat below). Enforcement lands with the first DMA
|
||||
driver, which is what there is to protect and test against. Proven in the `iommu` test,
|
||||
booted with an emulated `intel-iommu`.
|
||||
- **`system_spawn`** — a user-space supervisor starts a driver: `system_spawn(name)`
|
||||
loads a binary bundled in the initial-ramdisk as a fresh ring-3 process. This is what
|
||||
- **`system_spawn`** — a user-space supervisor starts a driver:
|
||||
`system_spawn(name, arguments)` loads a binary bundled in the initial-ramdisk as a
|
||||
fresh ring-3 process; `name` becomes the child's argv[0] and the optional
|
||||
NUL-separated `arguments` blob its argv[1..], delivered on a SysV entry stack
|
||||
([sysv.md](sysv.md)). This is what
|
||||
turned the device manager from "log the match" into "run the driver": the kernel now
|
||||
spawns only `init`, `init` spawns the services, and the **device-manager** discovers
|
||||
the hardware and spawns each driver ([drivers.md](drivers.md)). Ungated for now — a
|
||||
|
||||
+5
-3
@@ -37,8 +37,9 @@ kernel ──spawns──► init (PID 1) ──spawns──► device-manag
|
||||
```
|
||||
|
||||
The kernel launches exactly one process — `init` — and hands it nothing but the raw
|
||||
ability to start more (`system_spawn(name)`, which loads a binary bundled in the
|
||||
initial-ramdisk as a fresh ring-3 process). Everything else is a user-space decision:
|
||||
ability to start more (`system_spawn(name, arguments)`, which loads a binary bundled
|
||||
in the initial-ramdisk as a fresh ring-3 process — `name` becoming its argv[0],
|
||||
the optional arguments its argv[1..], on a SysV entry stack, see sysv.md). Everything else is a user-space decision:
|
||||
|
||||
- **init** ([system/services/init](system/services/init/init.zig)) is the **service
|
||||
supervisor**. It spawns the system services danos brings up at boot — today `vfs` and
|
||||
@@ -52,7 +53,8 @@ initial-ramdisk as a fresh ring-3 process). Everything else is a user-space deci
|
||||
is a table (`driverFor`): today a static `timer → hpet` map; a fuller system reads
|
||||
what each driver *binds* (a manifest under `/system/drivers`, or the driver
|
||||
describing its own match).
|
||||
3. **Spawn** — `system_spawn(driver_name)` starts the matched driver, which then claims
|
||||
3. **Spawn** — `system_spawn(driver_name, arguments)` starts the matched driver (the
|
||||
arguments can carry *which* device it matched), which then claims
|
||||
its device and runs the event loop below.
|
||||
|
||||
So "how is a driver discovered and configured" has two halves: **discovery** is the
|
||||
|
||||
@@ -65,6 +65,36 @@ function-pointer type and the kernel's `_start` both carry
|
||||
whole reason `kernel_abi` lives in the shared contract — see [efi.md](efi.md) for
|
||||
the handoff it governs.
|
||||
|
||||
## The process-entry stack (argc/argv)
|
||||
|
||||
The SysV ABI also fixes what a *fresh process* finds on its stack — and danos
|
||||
follows it, so its own runtime and any future C libc read arguments the same way.
|
||||
At the first user instruction, `rsp` is 16-byte aligned and points at (addresses
|
||||
growing upward):
|
||||
|
||||
```
|
||||
rsp → argc u64
|
||||
argv[0] … argv[argc-1] pointers into the strings area below
|
||||
NULL argv terminator
|
||||
NULL envp terminator (no environment yet)
|
||||
{AT_PAGESZ, page size} auxiliary vector
|
||||
{AT_NULL, 0} auxiliary-vector terminator
|
||||
argv string bytes NUL-terminated
|
||||
───────────────────────── stack top (stack_top_virtual)
|
||||
```
|
||||
|
||||
The kernel builds this block at the top of the process's stack — 8 pages (32 KiB,
|
||||
`parameters.user_stack_pages`) mapped RW+NX below a fixed top, with the page below
|
||||
them left unmapped as a **guard**, so a stack overflow faults (killing only that
|
||||
process) instead of silently corrupting the image
|
||||
(`buildEntryStack` in `system/kernel/process.zig`); `argv[0]` is always the path
|
||||
or initial-ramdisk name the process was spawned as, and `system_spawn`'s optional
|
||||
argument blob becomes `argv[1..]`. The runtime's `_start`
|
||||
(`library/runtime/start.zig`) hands the block to `rt_start`, which exposes it as
|
||||
`runtime.argumentCount()` / `runtime.argument(i)`. A C runtime's `crt0` would walk
|
||||
the identical layout unmodified — that's the compatibility being bought. The
|
||||
`args` test proves the round trip.
|
||||
|
||||
## Where else it surfaces
|
||||
|
||||
- **The red zone → `red_zone = false`.** `build.zig` disables the red zone for the
|
||||
|
||||
@@ -26,5 +26,10 @@ pub const dma = @import("dma.zig");
|
||||
/// Re-exported so a user binary can `pub const panic = runtime.panic;`.
|
||||
pub const panic = start.panic;
|
||||
|
||||
/// Process arguments (argc/argv, parsed from the kernel-built entry stack):
|
||||
/// `argument(0)` is the path or name this binary was spawned as.
|
||||
pub const argumentCount = start.argumentCount;
|
||||
pub const argument = start.argument;
|
||||
|
||||
/// The heap as a `std.mem.Allocator`, for Zig `std` containers in user code.
|
||||
pub const allocator = heap.allocator;
|
||||
|
||||
@@ -5,25 +5,52 @@
|
||||
const std = @import("std");
|
||||
const system = @import("system.zig");
|
||||
|
||||
/// The kernel enters at `_start` with rsp 16-aligned, but a SystemV function expects
|
||||
/// rsp ≡ 8 (mod 16) on entry (as if reached by `call`). The `call` below pushes
|
||||
/// the 8-byte return address, satisfying the ABI before any Zig frame runs; the
|
||||
/// `ud2` is a safety net if `rt_start` ever returns.
|
||||
/// The kernel enters at `_start` with rsp 16-aligned, pointing at the System V
|
||||
/// process-entry block it built: argc, argv pointers, NULL, envp terminator, the
|
||||
/// auxiliary vector, then the strings (see system/kernel/process.zig,
|
||||
/// `buildEntryStack`). Capture that address in rdi — the first SysV argument —
|
||||
/// before `call` disturbs the stack; the call's pushed return address also puts
|
||||
/// rsp ≡ 8 (mod 16), satisfying the ABI before any Zig frame runs. The `ud2` is a
|
||||
/// safety net if `rt_start` ever returns.
|
||||
pub export fn _start() callconv(.naked) noreturn {
|
||||
asm volatile (
|
||||
\\mov %%rsp, %%rdi
|
||||
\\call rt_start
|
||||
\\ud2
|
||||
);
|
||||
}
|
||||
|
||||
/// The first Zig frame. The heap is lazy (first alloc grows it), so there is no
|
||||
/// runtime init to order here — just hand control to the program's `main`.
|
||||
export fn rt_start() callconv(.c) noreturn {
|
||||
// The process-entry block, recorded by `rt_start` for the accessors below.
|
||||
var argument_count: usize = 0;
|
||||
var argument_vector: [*]const u64 = undefined;
|
||||
|
||||
/// The first Zig frame, entered with `stack` pointing at the kernel-built entry
|
||||
/// block. Record argc/argv for the accessors, then hand control to the program's
|
||||
/// `main`. The heap is lazy (first alloc grows it), so there is no other runtime
|
||||
/// init to order here.
|
||||
export fn rt_start(stack: [*]const u64) callconv(.c) noreturn {
|
||||
argument_count = stack[0];
|
||||
argument_vector = stack + 1;
|
||||
const root = @import("root"); // the user binary's root source file
|
||||
root.main();
|
||||
system.exit(0);
|
||||
}
|
||||
|
||||
/// Number of process arguments (argc). At least 1: argument 0 is the path or
|
||||
/// name this binary was spawned as.
|
||||
pub fn argumentCount() usize {
|
||||
return argument_count;
|
||||
}
|
||||
|
||||
/// Process argument `index` (0 = the program's own path/name), or an empty slice
|
||||
/// if out of range. The bytes live in the entry block at the top of the stack
|
||||
/// page, NUL-terminated, valid for the process's lifetime.
|
||||
pub fn argument(index: usize) []const u8 {
|
||||
if (index >= argument_count) return "";
|
||||
const string: [*:0]const u8 = @ptrFromInt(argument_vector[index]);
|
||||
return std.mem.span(string);
|
||||
}
|
||||
|
||||
/// No runtime to unwind into — report a panic as a nonzero exit code.
|
||||
pub const panic = std.debug.FullPanic(struct {
|
||||
fn panic(_: []const u8, _: ?usize) noreturn {
|
||||
|
||||
@@ -44,11 +44,31 @@ pub fn exit(code: usize) noreturn {
|
||||
}
|
||||
|
||||
/// Start the binary bundled in the initial-ramdisk under `name` as a new ring-3
|
||||
/// process, returning true on success. This is how a supervisor (the device manager)
|
||||
/// launches a driver it matched — danos-native, not POSIX (a spawn/exec family comes
|
||||
/// with the process work later).
|
||||
/// process, returning true on success. The child's argv[0] is `name`. This is how
|
||||
/// a supervisor (the device manager) launches a driver it matched — danos-native,
|
||||
/// not POSIX (a spawn/exec family comes with the process work later).
|
||||
pub fn spawn(name: []const u8) bool {
|
||||
return sc.systemCall2(.system_spawn, @intFromPtr(name.ptr), name.len) == 0;
|
||||
return sc.systemCall4(.system_spawn, @intFromPtr(name.ptr), name.len, 0, 0) == 0;
|
||||
}
|
||||
|
||||
/// Like `spawn`, but hands the child command-line arguments: they arrive as
|
||||
/// argv[1..] on its System V entry stack (argv[0] is still `name`). Marshalled to
|
||||
/// the kernel as one NUL-separated blob; the combined arguments must fit
|
||||
/// `blob` (the kernel caps the blob at 256 bytes and argc at 8 anyway).
|
||||
pub fn spawnWithArguments(name: []const u8, arguments: []const []const u8) bool {
|
||||
var blob: [256]u8 = undefined;
|
||||
var len: usize = 0;
|
||||
for (arguments, 0..) |argument, i| {
|
||||
if (i != 0) {
|
||||
if (len >= blob.len) return false;
|
||||
blob[len] = 0;
|
||||
len += 1;
|
||||
}
|
||||
if (len + argument.len > blob.len) return false;
|
||||
@memcpy(blob[len..][0..argument.len], argument);
|
||||
len += argument.len;
|
||||
}
|
||||
return sc.systemCall4(.system_spawn, @intFromPtr(name.ptr), name.len, @intFromPtr(&blob), len) == 0;
|
||||
}
|
||||
|
||||
/// Grant `len` bytes (rounded up to whole pages) of fresh, zeroed, writable
|
||||
|
||||
@@ -284,7 +284,7 @@ fn kmain(boot_information: *const BootInformation) noreturn {
|
||||
if (boot_information.init_len != 0) {
|
||||
status("starting /system/services/init...\n");
|
||||
const image = @as([*]const u8, @ptrFromInt(boot_handoff.physicalToVirtual(boot_information.init_base)))[0..boot_information.init_len];
|
||||
process.spawnProcess(image, 4) catch |err| {
|
||||
process.spawnProcess(image, 4, &.{"/system/services/init"}) catch |err| {
|
||||
statusPrint("/system/services/init failed to load: {s}\n", .{@errorName(err)});
|
||||
};
|
||||
} else {
|
||||
@@ -414,7 +414,7 @@ fn recoverableFault(vector: u64) bool {
|
||||
/// is no scheduled process to kill.)
|
||||
fn onException(state: *const architecture.CpuState) noreturn {
|
||||
if (architecture.fromUser(state) and scheduler.currentIsUserProcess() and recoverableFault(state.vector)) {
|
||||
statusPrint("\ndanos: process {d} killed by {s} (vector {d}) on core {d}\n", .{ scheduler.currentId(), architecture.exceptionName(state.vector), state.vector, scheduler.currentCpuIndex() });
|
||||
statusPrint("\ndanos: process {d} ({s}) killed by {s} (vector {d}) on core {d}\n", .{ scheduler.currentId(), scheduler.current().name(), architecture.exceptionName(state.vector), state.vector, scheduler.currentCpuIndex() });
|
||||
statusPrint(" error code : 0x{x}\n", .{state.error_code});
|
||||
statusPrint(" IP : 0x{x:0>16}\n", .{architecture.instructionPointer(state)});
|
||||
if (architecture.faultAddress(state)) |address| statusPrint(" fault addr : 0x{x:0>16}\n", .{address});
|
||||
|
||||
+140
-21
@@ -24,6 +24,7 @@ const elf = std.elf;
|
||||
const boot_handoff = @import("boot-handoff");
|
||||
const abi = @import("abi");
|
||||
const device_abi = @import("device-abi");
|
||||
const parameters = @import("parameters");
|
||||
const architecture = @import("architecture");
|
||||
const pmm = @import("pmm.zig");
|
||||
const scheduler = @import("scheduler.zig");
|
||||
@@ -40,9 +41,18 @@ const SystemCall = abi.SystemCall;
|
||||
/// User virtual addresses. PML4 index 224 — a user-exclusive region, far from
|
||||
/// the identity map (low indices) and the vmm test address (index 128), so
|
||||
/// setting the U/S bit on its intermediate tables widens no kernel mapping.
|
||||
/// An ELF image may occupy [code_virtual, stack_virtual); the stack page sits above.
|
||||
/// An ELF image may occupy [code_virtual, stack_virtual); the stack sits above.
|
||||
pub const code_virtual: u64 = 0x0000_7000_0000_0000;
|
||||
pub const stack_virtual: u64 = 0x0000_7000_0020_0000;
|
||||
|
||||
/// The stack region, above the image. The page at `stack_virtual` is **never
|
||||
/// mapped** — it is the guard page: a process that overflows its stack walks into
|
||||
/// it and faults (killing only that process) rather than silently corrupting the
|
||||
/// top of its own image. The stack proper is `parameters.user_stack_pages` pages
|
||||
/// at [stack_base_virtual, stack_top_virtual), RW + NX, with the System V entry
|
||||
/// block (argc/argv) at the very top.
|
||||
pub const stack_virtual: u64 = 0x0000_7000_0020_0000; // guard page (unmapped)
|
||||
pub const stack_base_virtual: u64 = stack_virtual + page_size;
|
||||
pub const stack_top_virtual: u64 = stack_base_virtual + parameters.user_stack_pages * page_size;
|
||||
|
||||
/// The mmap grant arena: where `mmap` hands out fresh user pages, above the image
|
||||
/// and stack but still inside PML4[224] (so no kernel mapping is widened). Each
|
||||
@@ -73,6 +83,19 @@ pub const dma_arena_end: u64 = dma_arena_base + (256 << 20); // 256 MiB per proc
|
||||
/// chunks, so this bound is generous; it also caps the frame scratch array below.
|
||||
const maximum_mmap_pages = 256;
|
||||
|
||||
/// Ceiling on a process's argv entries, including argv[0]. Arguments are spawn
|
||||
/// parameters ("you are the driver for device 12"), not bulk data — IPC carries
|
||||
/// that — so the bound is small and everything fits the single stack page.
|
||||
pub const maximum_arguments = 8;
|
||||
|
||||
/// Ceiling on the `system_spawn` extra-arguments blob (argv[1..], NUL-separated).
|
||||
pub const maximum_argument_bytes = 256;
|
||||
|
||||
/// Auxiliary-vector entry types (System V AMD64 process entry). Only what the
|
||||
/// kernel emits today; a C runtime scans the vector until the null terminator.
|
||||
const auxiliary_vector_null: u64 = 0; // AT_NULL — end of the vector
|
||||
const auxiliary_vector_page_size: u64 = 6; // AT_PAGESZ
|
||||
|
||||
// The hand-assembled user program blob (isr.s, .rodata) — the isolation probe.
|
||||
const pf_start = @extern([*]const u8, .{ .name = "user_pf_start" });
|
||||
const pf_end = @extern([*]const u8, .{ .name = "user_pf_end" });
|
||||
@@ -392,28 +415,50 @@ fn systemDeviceRegister(state: *architecture.CpuState) void {
|
||||
architecture.setSystemCallResult(state, id);
|
||||
}
|
||||
|
||||
/// system_spawn(name_ptr, name_len) -> 0 on success, -1 on failure. Load the binary
|
||||
/// bundled in the initial-ramdisk under `name` as a fresh ring-3 process. This is the
|
||||
/// mechanism a user-space supervisor (the device manager) uses to start a driver it
|
||||
/// matched: discovery and policy stay in user space, the kernel only spawns.
|
||||
/// system_spawn(name_ptr, name_len, arguments_ptr, arguments_len) -> 0 on success,
|
||||
/// -1 on failure. Load the binary bundled in the initial-ramdisk under `name` as a
|
||||
/// fresh ring-3 process. `name` becomes the child's argv[0] (and its task name, so
|
||||
/// a fault report can say which binary died); `arguments` is an optional
|
||||
/// NUL-separated blob that becomes argv[1..] — how a supervisor parameterises what
|
||||
/// it starts ("you are the driver for device 12"). 0/0 means no extra arguments.
|
||||
/// This is the mechanism a user-space supervisor (the device manager) uses to start
|
||||
/// a driver it matched: discovery and policy stay in user space, the kernel only
|
||||
/// spawns.
|
||||
///
|
||||
/// Ungated for now — any process may spawn any bundled binary. A capability (only a
|
||||
/// supervisor holds the right to spawn) belongs here once the model grows one; see
|
||||
/// docs/driver-model.md. The name is bounds-checked into the user half exactly like
|
||||
/// `debug_write`, and an unknown name or a load failure returns -1.
|
||||
/// docs/driver-model.md. Both buffers are bounds-checked into the user half exactly
|
||||
/// like `debug_write`, and an unknown name or a load failure returns -1.
|
||||
fn systemSpawn(state: *architecture.CpuState) void {
|
||||
const ptr = architecture.systemCallArg(state, 0);
|
||||
const len = architecture.systemCallArg(state, 1);
|
||||
const arguments_ptr = architecture.systemCallArg(state, 2);
|
||||
const arguments_len = architecture.systemCallArg(state, 3);
|
||||
if (len == 0 or len > 64 or ptr >= user_half_end or ptr + len > user_half_end) return fail(state);
|
||||
if (arguments_len > maximum_argument_bytes) return fail(state);
|
||||
if (arguments_len != 0 and (arguments_ptr >= user_half_end or arguments_ptr + arguments_len > user_half_end)) return fail(state);
|
||||
const image = ramdisk_image orelse return fail(state);
|
||||
const rd = initial_ramdisk.Reader.init(image) orelse return fail(state);
|
||||
|
||||
const name = @as([*]const u8, @ptrFromInt(ptr))[0..len];
|
||||
var argv: [maximum_arguments][]const u8 = undefined;
|
||||
argv[0] = name;
|
||||
var argc: usize = 1;
|
||||
if (arguments_len != 0) {
|
||||
const blob = @as([*]const u8, @ptrFromInt(arguments_ptr))[0..arguments_len];
|
||||
var pieces = std.mem.tokenizeScalar(u8, blob, 0);
|
||||
while (pieces.next()) |piece| {
|
||||
if (argc == maximum_arguments) return fail(state);
|
||||
argv[argc] = piece;
|
||||
argc += 1;
|
||||
}
|
||||
}
|
||||
|
||||
var i: u32 = 0;
|
||||
while (i < rd.count) : (i += 1) {
|
||||
const item = rd.entry(i) orelse continue;
|
||||
if (!std.mem.eql(u8, item.name, name)) continue;
|
||||
spawnProcess(item.blob, 4) catch return fail(state);
|
||||
spawnProcess(item.blob, 4, argv[0..argc]) catch return fail(state);
|
||||
architecture.setSystemCallResult(state, 0);
|
||||
return;
|
||||
}
|
||||
@@ -641,15 +686,15 @@ pub fn run(blob: []const u8) RunError!void {
|
||||
@memset(code[blob.len..page_size], 0xCC);
|
||||
|
||||
architecture.mapUserPage(code_virtual, code_frame, false, true); // RO + X
|
||||
architecture.mapUserPage(stack_virtual, stack_frame, true, false); // RW + NX
|
||||
architecture.mapUserPage(stack_base_virtual, stack_frame, true, false); // RW + NX (one page; the probe barely stacks)
|
||||
resetRecords();
|
||||
|
||||
architecture.enterUser(scheduler.currentCpuIndex(), code_virtual, stack_virtual + page_size);
|
||||
architecture.enterUser(scheduler.currentCpuIndex(), code_virtual, stack_base_virtual + page_size);
|
||||
|
||||
// Back via the exit system_call; the interrupt gate left IF clear.
|
||||
architecture.enableInterrupts();
|
||||
architecture.unmapPage(code_virtual);
|
||||
architecture.unmapPage(stack_virtual);
|
||||
architecture.unmapPage(stack_base_virtual);
|
||||
pmm.free(code_frame);
|
||||
pmm.free(stack_frame);
|
||||
}
|
||||
@@ -660,6 +705,7 @@ pub const InitError = error{
|
||||
BadElf, // malformed/inapplicable image (magic, class, machine, type, bounds)
|
||||
BadSegment, // PT_LOAD unaligned, out of the user region, W&X, or overlapping
|
||||
BadEntry, // e_entry not inside an executable segment
|
||||
BadArguments, // no argv[0], too many entries, or too many bytes for the entry stack
|
||||
ProgramTooBig, // more pages than the loader's budget
|
||||
OutOfMemory,
|
||||
};
|
||||
@@ -760,13 +806,74 @@ fn loadPageInto(aspace: u64, image: []const u8, seg: Segment, page_index: u64) I
|
||||
architecture.mapUserPageInto(aspace, seg.vaddr + page_off, frame, seg.writable, seg.executable);
|
||||
}
|
||||
|
||||
/// Build the System V AMD64 process-entry block at the top of a process's stack
|
||||
/// page and return the initial user stack pointer. At the first user instruction,
|
||||
/// rsp is 16-byte aligned and points at (addresses growing upward):
|
||||
///
|
||||
/// argc, argv[0..argc-1], NULL, NULL (empty envp), auxiliary vector, strings
|
||||
///
|
||||
/// — the layout every C runtime's startup code walks, so danos's own runtime and a
|
||||
/// future libc port read arguments identically (docs/sysv.md). `page` is the kernel
|
||||
/// (physmap) view of the stack's **top** frame and `page_user_base` that frame's
|
||||
/// user address (stack_top_virtual - page_size); the pointers written into it are
|
||||
/// user addresses inside that page. The caller has validated the sizes
|
||||
/// (`entryStackBytes`), so this cannot overrun.
|
||||
fn buildEntryStack(page: [*]u8, page_user_base: u64, argv: []const []const u8) u64 {
|
||||
// The strings live at the very top of the page, packed from the end downward.
|
||||
var string_offset: usize = page_size;
|
||||
var pointers: [maximum_arguments]u64 = undefined;
|
||||
var i: usize = argv.len;
|
||||
while (i > 0) {
|
||||
i -= 1;
|
||||
string_offset -= argv[i].len + 1;
|
||||
@memcpy(page[string_offset..][0..argv[i].len], argv[i]);
|
||||
page[string_offset + argv[i].len] = 0; // NUL-terminated, as C expects
|
||||
pointers[i] = page_user_base + string_offset;
|
||||
}
|
||||
|
||||
// The vector sits below the strings: argc, the argv pointers, the argv
|
||||
// terminator, an empty envp (terminator only), then the auxiliary vector.
|
||||
const word_count = 1 + argv.len + 1 + 1 + 4;
|
||||
const vector_offset = (string_offset - word_count * 8) & ~@as(usize, 15); // entry rsp % 16 == 0
|
||||
const words: [*]u64 = @ptrCast(@alignCast(page + vector_offset));
|
||||
var w: usize = 0;
|
||||
words[w] = argv.len; // argc
|
||||
w += 1;
|
||||
for (pointers[0..argv.len]) |pointer| {
|
||||
words[w] = pointer;
|
||||
w += 1;
|
||||
}
|
||||
words[w] = 0; // argv terminator
|
||||
words[w + 1] = 0; // envp: no environment yet, just the terminator
|
||||
words[w + 2] = auxiliary_vector_page_size;
|
||||
words[w + 3] = page_size;
|
||||
words[w + 4] = auxiliary_vector_null; // end of the auxiliary vector
|
||||
words[w + 5] = 0;
|
||||
return page_user_base + vector_offset;
|
||||
}
|
||||
|
||||
/// Bytes the entry block for `argv` occupies at the top of the stack page:
|
||||
/// strings (each NUL-terminated), vector words, and the alignment slack.
|
||||
fn entryStackBytes(argv: []const []const u8) usize {
|
||||
var string_bytes: usize = 0;
|
||||
for (argv) |argument| string_bytes += argument.len + 1;
|
||||
return string_bytes + (1 + argv.len + 1 + 1 + 4) * 8 + 16;
|
||||
}
|
||||
|
||||
/// Load a user ELF image into a fresh address space and spawn it as a scheduled
|
||||
/// ring-3 process at `priority`. Returns immediately — the process runs
|
||||
/// preemptively on its own page tables alongside everything else, and its exit
|
||||
/// is handled by the system_call layer. The whole build (address space + ELF load +
|
||||
/// task) runs under the kernel lock so it appears atomically and can't race
|
||||
/// pmm/heap on another core.
|
||||
pub fn spawnProcess(image: []const u8, priority: u3) InitError!void {
|
||||
/// ring-3 process at `priority`, entered with `argv` on its stack per the System V
|
||||
/// convention (`buildEntryStack`). `argv[0]` is required — it names the process:
|
||||
/// the path or initial-ramdisk name it was spawned as. It is also recorded on the
|
||||
/// task, so a fault report can say *which* binary died, not just its id.
|
||||
/// Returns immediately — the process runs preemptively on its own page tables
|
||||
/// alongside everything else, and its exit is handled by the system_call layer.
|
||||
/// The whole build (address space + ELF load + task) runs under the kernel lock so
|
||||
/// it appears atomically and can't race pmm/heap on another core.
|
||||
pub fn spawnProcess(image: []const u8, priority: u3, argv: []const []const u8) InitError!void {
|
||||
if (argv.len == 0 or argv.len > maximum_arguments) return error.BadArguments;
|
||||
// The entry block must leave most of the page as actual stack.
|
||||
if (entryStackBytes(argv) > page_size / 2) return error.BadArguments;
|
||||
|
||||
var segs: [maximum_segments]Segment = undefined;
|
||||
const parsed = try parseSegments(image, &segs);
|
||||
|
||||
@@ -779,10 +886,22 @@ pub fn spawnProcess(image: []const u8, priority: u3) InitError!void {
|
||||
for (segs[0..parsed.count]) |seg| {
|
||||
for (0..seg.pages()) |i| try loadPageInto(aspace, image, seg, i);
|
||||
}
|
||||
const stack_frame = pmm.alloc() orelse return error.OutOfMemory;
|
||||
architecture.mapUserPageInto(aspace, stack_virtual, stack_frame, true, false); // RW + NX
|
||||
|
||||
if (!scheduler.spawnUserLocked(aspace, parsed.entry, stack_virtual + page_size, priority))
|
||||
// The stack: `user_stack_pages` zeroed pages below stack_top_virtual, RW + NX.
|
||||
// The page below them (`stack_virtual`) stays unmapped as the overflow guard.
|
||||
// The entry block goes at the top of the highest page.
|
||||
var user_sp: u64 = 0;
|
||||
for (0..parameters.user_stack_pages) |i| {
|
||||
const stack_frame = pmm.alloc() orelse return error.OutOfMemory;
|
||||
const stack_page: [*]u8 = @ptrFromInt(boot_handoff.physicalToVirtual(stack_frame));
|
||||
@memset(stack_page[0..page_size], 0); // no stale frame contents leak into user space
|
||||
const page_virtual = stack_base_virtual + i * page_size;
|
||||
if (i == parameters.user_stack_pages - 1)
|
||||
user_sp = buildEntryStack(stack_page, page_virtual, argv);
|
||||
architecture.mapUserPageInto(aspace, page_virtual, stack_frame, true, false); // RW + NX
|
||||
}
|
||||
|
||||
if (!scheduler.spawnUserLocked(aspace, parsed.entry, user_sp, priority, argv[0]))
|
||||
return error.OutOfMemory;
|
||||
}
|
||||
|
||||
|
||||
@@ -70,8 +70,24 @@ pub const Task = struct {
|
||||
ipc_send_cap: u64 = ~@as(u64, 0), // handle to transfer with this message (abi.no_cap = none)
|
||||
ipc_received_cap: u64 = ~@as(u64, 0), // client: handle the reply's transferred cap landed at (abi.no_cap = none)
|
||||
next: ?*Task = null, // ready-queue link (also the endpoint sender-FIFO link)
|
||||
// The process's name — argv[0] as it was spawned (a boot-volume path for init,
|
||||
// an initial-ramdisk name for everything else); empty for kernel tasks. Fixed
|
||||
// storage, so the fault path can name the dead without touching the heap.
|
||||
// Zero-initialised (not `undefined`): an undefined default is materialised as
|
||||
// a 0xAA fill, which would move the whole static task pool out of .bss.
|
||||
name_buffer: [maximum_task_name]u8 = .{0} ** maximum_task_name,
|
||||
name_length: u8 = 0,
|
||||
|
||||
/// The task's name (argv[0] at spawn), or empty for a kernel task.
|
||||
pub fn name(self: *const Task) []const u8 {
|
||||
return self.name_buffer[0..self.name_length];
|
||||
}
|
||||
};
|
||||
|
||||
/// Capacity of `Task.name_buffer` — matches the longest name `system_spawn`
|
||||
/// accepts, so a spawned name is never truncated.
|
||||
pub const maximum_task_name = 64;
|
||||
|
||||
/// Size of each task's IPC handle table. Kept here (not in ipc_sync.zig) because
|
||||
/// it dimensions a field of `Task`; ipc_sync.zig re-exports it.
|
||||
pub const ipc_maximum_handles = 16;
|
||||
@@ -258,12 +274,13 @@ pub fn spawnOn(entry: *const fn () void, priority: Priority, cpu: u32) bool {
|
||||
}
|
||||
|
||||
/// Spawn a **user** task: a task with its own address space (`aspace`) that starts
|
||||
/// in user mode at `entry` on `user_sp`. It gets a fresh kernel stack for
|
||||
/// syscalls/interrupts, and its first switch-in lands in `user_task_trampoline`.
|
||||
/// in user mode at `entry` on `user_sp`, recorded under `name` (its argv[0]).
|
||||
/// It gets a fresh kernel stack for syscalls/interrupts, and its first switch-in
|
||||
/// lands in `user_task_trampoline`.
|
||||
/// Returns false (creating nothing) if the table is full or out of memory.
|
||||
/// **Caller must hold the kernel lock** (the loader that builds `aspace` holds it
|
||||
/// across the whole spawn, so the address space and the task appear atomically).
|
||||
pub fn spawnUserLocked(aspace: u64, entry: u64, user_sp: u64, priority: Priority) bool {
|
||||
pub fn spawnUserLocked(aspace: u64, entry: u64, user_sp: u64, priority: Priority, task_name: []const u8) bool {
|
||||
const t = freeSlot() orelse return false;
|
||||
const stack = heap.allocator().alloc(u8, stack_size) catch return false;
|
||||
t.* = .{
|
||||
@@ -275,6 +292,9 @@ pub fn spawnUserLocked(aspace: u64, entry: u64, user_sp: u64, priority: Priority
|
||||
.user_ip = entry,
|
||||
.user_sp = user_sp,
|
||||
};
|
||||
const name_length = @min(task_name.len, maximum_task_name);
|
||||
@memcpy(t.name_buffer[0..name_length], task_name[0..name_length]);
|
||||
t.name_length = @intCast(name_length);
|
||||
next_id += 1;
|
||||
const top = @intFromPtr(stack.ptr) + stack.len;
|
||||
t.kstack_top = top;
|
||||
|
||||
+48
-8
@@ -120,6 +120,8 @@ pub fn run(case: []const u8, boot_information: *const BootInformation) void {
|
||||
userPfTest();
|
||||
} else if (eql(case, "fault-recovery")) {
|
||||
faultRecoveryTest(boot_information);
|
||||
} else if (eql(case, "args")) {
|
||||
argsTest(boot_information);
|
||||
} else if (eql(case, "init")) {
|
||||
initTest(boot_information);
|
||||
} else if (eql(case, "process")) {
|
||||
@@ -1162,8 +1164,8 @@ fn processTest(boot_information: *const BootInformation) void {
|
||||
scheduler.spawn(procWorker, 4); // kernel task at the processes' priority
|
||||
|
||||
var spawned: u32 = 0;
|
||||
if (process.spawnProcess(image, 4)) spawned += 1 else |_| {}
|
||||
if (process.spawnProcess(image, 4)) spawned += 1 else |_| {}
|
||||
if (process.spawnProcess(image, 4, &.{"/system/services/init"})) spawned += 1 else |_| {}
|
||||
if (process.spawnProcess(image, 4, &.{"/system/services/init"})) spawned += 1 else |_| {}
|
||||
|
||||
// Wait (real time) for several heartbeats across the two processes. Each
|
||||
// process sleeps ~1 s between beats, so a few seconds yields several.
|
||||
@@ -1218,9 +1220,9 @@ fn spawnFaultingProcess() bool {
|
||||
architecture.destroyAddressSpace(aspace); // frees code_frame too — it's mapped
|
||||
return false;
|
||||
};
|
||||
architecture.mapUserPageInto(aspace, process.stack_virtual, stack_frame, true, false); // RW + NX
|
||||
architecture.mapUserPageInto(aspace, process.stack_base_virtual, stack_frame, true, false); // RW + NX
|
||||
|
||||
if (!scheduler.spawnUserLocked(aspace, process.code_virtual, process.stack_virtual + abi.page_size, 4)) {
|
||||
if (!scheduler.spawnUserLocked(aspace, process.code_virtual, process.stack_base_virtual + abi.page_size, 4, "fault-probe")) {
|
||||
architecture.destroyAddressSpace(aspace);
|
||||
return false;
|
||||
}
|
||||
@@ -1243,7 +1245,7 @@ fn faultRecoveryTest(boot_information: *const BootInformation) void {
|
||||
|
||||
process.write_count = 0;
|
||||
process.fault_kill_count = 0;
|
||||
const spawned = if (process.spawnProcess(image, 4)) true else |_| false;
|
||||
const spawned = if (process.spawnProcess(image, 4, &.{"/system/services/init"})) true else |_| false;
|
||||
check("init spawned as the surviving process", spawned);
|
||||
|
||||
// A first heartbeat proves init runs before the fault.
|
||||
@@ -1287,7 +1289,7 @@ fn initTest(boot_information: *const BootInformation) void {
|
||||
}
|
||||
const image = @as([*]const u8, @ptrFromInt(boot_handoff.physicalToVirtual(boot_information.init_base)))[0..boot_information.init_len];
|
||||
process.write_count = 0;
|
||||
const spawned = if (process.spawnProcess(image, 4)) true else |err| blk: {
|
||||
const spawned = if (process.spawnProcess(image, 4, &.{"/system/services/init"})) true else |err| blk: {
|
||||
log("DANOS-INIT-ERR: {s}\n", .{@errorName(err)});
|
||||
break :blk false;
|
||||
};
|
||||
@@ -1334,7 +1336,7 @@ fn initialRamdiskTest(boot_information: *const BootInformation) void {
|
||||
var i: u32 = 0;
|
||||
while (i < rd.count) : (i += 1) {
|
||||
const item = rd.entry(i) orelse continue;
|
||||
if (process.spawnProcess(item.blob, 4)) spawned += 1 else |err| {
|
||||
if (process.spawnProcess(item.blob, 4, &.{item.name})) spawned += 1 else |err| {
|
||||
log("DANOS-INITRD-ERR: {s}: {s}\n", .{ item.name, @errorName(err) });
|
||||
}
|
||||
}
|
||||
@@ -1394,6 +1396,44 @@ fn vfsTest(boot_information: *const BootInformation) void {
|
||||
result();
|
||||
}
|
||||
|
||||
/// Process arguments, end to end: spawn args-echo bare (its argv[0] is the
|
||||
/// initial-ramdisk name). Instance 1 sees argc == 1 and respawns itself through
|
||||
/// `system_spawn` with the extra arguments "alpha beta-42" — the syscall argument
|
||||
/// blob. Instance 2 parses the kernel-built System V entry stack via the runtime
|
||||
/// and echoes its whole argv in one write, which must arrive exactly as sent.
|
||||
fn argsTest(boot_information: *const BootInformation) void {
|
||||
log("DANOS-TEST-BEGIN: args\n", .{});
|
||||
check("bootloader handed over an initial_ramdisk", boot_information.initial_ramdisk_len != 0);
|
||||
if (boot_information.initial_ramdisk_len == 0) {
|
||||
result();
|
||||
return;
|
||||
}
|
||||
const image = @as([*]const u8, @ptrFromInt(boot_handoff.physicalToVirtual(boot_information.initial_ramdisk_base)))[0..boot_information.initial_ramdisk_len];
|
||||
const rd = initial_ramdisk.Reader.init(image) orelse {
|
||||
check("initial_ramdisk image is valid", false);
|
||||
result();
|
||||
return;
|
||||
};
|
||||
process.setInitialRamdisk(image); // args-echo respawns itself through system_spawn
|
||||
|
||||
process.write_count = 0;
|
||||
process.write_from_user = false;
|
||||
check("args-echo spawned from the initial_ramdisk", spawnNamed(rd, "args-echo"));
|
||||
|
||||
// Wait for the *second* instance's echo (the first writes nothing).
|
||||
scheduler.setPriority(1);
|
||||
const deadline = architecture.millis() + 8000;
|
||||
while (process.write_count < 1 and architecture.millis() < deadline) scheduler.yield();
|
||||
scheduler.setPriority(4);
|
||||
|
||||
const expected = "args: args-echo alpha beta-42\n";
|
||||
const echoed = process.write_len == expected.len and eql(process.write_buffer[0..process.write_len], expected);
|
||||
if (!echoed and process.write_len > 0) log("DANOS-ARGS: got \"{s}\"\n", .{process.write_buffer[0..process.write_len]});
|
||||
check("argv arrived intact (argv[0] = name, argv[1..] = spawn arguments)", echoed);
|
||||
check("echo came from user mode (CPL 3)", process.write_from_user);
|
||||
result();
|
||||
}
|
||||
|
||||
/// Spawn the initial_ramdisk binary named `name` as a ring-3 process. Returns false if it
|
||||
/// isn't in the image or fails to load.
|
||||
fn spawnNamed(rd: initial_ramdisk.Reader, name: []const u8) bool {
|
||||
@@ -1401,7 +1441,7 @@ fn spawnNamed(rd: initial_ramdisk.Reader, name: []const u8) bool {
|
||||
while (i < rd.count) : (i += 1) {
|
||||
const item = rd.entry(i) orelse continue;
|
||||
if (eql(item.name, name)) {
|
||||
return if (process.spawnProcess(item.blob, 4)) true else |_| false;
|
||||
return if (process.spawnProcess(item.blob, 4, &.{item.name})) true else |_| false;
|
||||
}
|
||||
}
|
||||
return false;
|
||||
|
||||
@@ -22,6 +22,12 @@ pub const maximum_tasks = 16;
|
||||
/// Each task's kernel stack (also each AP's bring-up stack), in bytes.
|
||||
pub const kernel_stack_size = 16 * 1024;
|
||||
|
||||
/// Each user process's stack, in pages (32 KiB). Mapped just below a fixed top;
|
||||
/// the System V entry block (argc/argv) occupies the top of the highest page, and
|
||||
/// the page below the mapping is left unmapped as a guard, so an overflow faults
|
||||
/// (killing only that process) instead of silently corrupting the image.
|
||||
pub const user_stack_pages = 8;
|
||||
|
||||
/// Each core's IST (double-fault) stack, in bytes. The BSP's is static; an AP's is
|
||||
/// heap-allocated at bring-up.
|
||||
pub const ist_stack_size = 16 * 1024;
|
||||
|
||||
@@ -0,0 +1,55 @@
|
||||
//! args-echo — a test fixture for process arguments (bundled in the
|
||||
//! initial-ramdisk, spawned only by the `args` test case). Run with no arguments,
|
||||
//! it respawns itself *with* some via `spawnWithArguments` — exercising the
|
||||
//! system_spawn argument blob. Run with arguments, it burns more stack than one
|
||||
//! page could hold (proving the multi-page stack: on a single-page stack the
|
||||
//! recursion would hit the guard and the process would be killed before echoing),
|
||||
//! then echoes its whole argv in one `debug_write` the kernel test asserts on —
|
||||
//! proving the kernel-built System V entry stack (argc, argv pointers,
|
||||
//! NUL-terminated strings) and the runtime's parsing of it, end to end.
|
||||
|
||||
const runtime = @import("runtime");
|
||||
|
||||
/// Recurse with a real frame each level: `depth` levels of ~0.5 KiB, touched
|
||||
/// through a volatile pointer so no optimiser can flatten the frames away.
|
||||
fn burnStack(depth: usize) u8 {
|
||||
var frame: [512]u8 = undefined;
|
||||
const touch: *volatile [512]u8 = &frame;
|
||||
touch[0] = @truncate(depth);
|
||||
touch[511] = touch[0];
|
||||
if (depth == 0) return touch[511];
|
||||
return touch[0] +% burnStack(depth - 1);
|
||||
}
|
||||
|
||||
pub fn main() void {
|
||||
if (runtime.argumentCount() <= 1) {
|
||||
// First instance: spawn the second with real arguments, then exit.
|
||||
_ = runtime.system.spawnWithArguments("args-echo", &.{ "alpha", "beta-42" });
|
||||
return;
|
||||
}
|
||||
|
||||
// ~16 x 0.5 KiB frames: comfortably past one page, well inside the 32 KiB stack.
|
||||
_ = burnStack(16);
|
||||
|
||||
// Second instance: echo "args: <argv0> <argv1> ..." for the test to match.
|
||||
var buffer: [128]u8 = undefined;
|
||||
const prefix = "args:";
|
||||
@memcpy(buffer[0..prefix.len], prefix);
|
||||
var len: usize = prefix.len;
|
||||
for (0..runtime.argumentCount()) |i| {
|
||||
const argument = runtime.argument(i);
|
||||
if (len + 1 + argument.len + 1 > buffer.len) break;
|
||||
buffer[len] = ' ';
|
||||
len += 1;
|
||||
@memcpy(buffer[len..][0..argument.len], argument);
|
||||
len += argument.len;
|
||||
}
|
||||
buffer[len] = '\n';
|
||||
len += 1;
|
||||
_ = runtime.system.write(buffer[0..len]);
|
||||
}
|
||||
|
||||
pub const panic = runtime.panic;
|
||||
comptime {
|
||||
_ = &runtime.start._start; // pull the runtime entry shim into the image
|
||||
}
|
||||
@@ -205,6 +205,11 @@ CASES = [
|
||||
"timeout": 60,
|
||||
"expect": r"DANOS-TEST-RESULT: PASS",
|
||||
"fail": r"DANOS-TEST-RESULT: FAIL"},
|
||||
# Process arguments: argv arrives on the SysV entry stack (argv[0] = the spawned
|
||||
# name, argv[1..] = the system_spawn argument blob) and echoes back intact.
|
||||
{"name": "args",
|
||||
"expect": r"DANOS-TEST-RESULT: PASS",
|
||||
"fail": r"DANOS-TEST-RESULT: FAIL"},
|
||||
# The real user binary: the bootloader ships /system/services/init off the ESP, the
|
||||
# kernel loads the ELF and runs it in ring 3, and it writes + exits cleanly.
|
||||
{"name": "init",
|
||||
|
||||
Reference in New Issue
Block a user