1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
use super::arch::*;
use super::data::{Map, SigAction, Stat, StatVfs, TimeSpec};
use super::error::Result;
use super::flag::*;
use super::number::*;
use super::io_uring;

use core::{mem, ptr};

// Signal restorer
extern "C" fn restorer() -> ! {
    sigreturn().unwrap();
    unreachable!();
}

/// Set the end of the process's heap
///
/// When `addr` is `0`, this function will return the current break.
///
/// When `addr` is nonzero, this function will attempt to set the end of the process's
/// heap to `addr` and return the new program break. The new program break should be
/// checked by the allocator, it may not be exactly `addr`, as it may be aligned to a page
/// boundary.
///
/// On error, `Err(ENOMEM)` will be returned indicating that no memory is available
pub unsafe fn brk(addr: usize) -> Result<usize> {
    syscall1(SYS_BRK, addr)
}

/// Change the process's working directory
///
/// This function will attempt to set the process's working directory to `path`, which can be
/// either a relative, scheme relative, or absolute path.
///
/// On success, `Ok(0)` will be returned. On error, one of the following errors will be returned.
///
/// # Errors
///
/// * `EACCES` - permission is denied for one of the components of `path`, or `path`
/// * `EFAULT` - `path` does not point to the process's addressible memory
/// * `EIO` - an I/O error occurred
/// * `ENOENT` - `path` does not exit
/// * `ENOTDIR` - `path` is not a directory
pub fn chdir<T: AsRef<[u8]>>(path: T) -> Result<usize> {
    unsafe { syscall2(SYS_CHDIR, path.as_ref().as_ptr() as usize, path.as_ref().len()) }
}

#[deprecated(
    since = "0.1.55",
    note = "use fchmod instead"
)]
pub fn chmod<T: AsRef<[u8]>>(path: T, mode: usize) -> Result<usize> {
    unsafe { syscall3(SYS_CHMOD, path.as_ref().as_ptr() as usize, path.as_ref().len(), mode) }
}

/// Produce a fork of the current process, or a new process thread
pub unsafe fn clone(flags: CloneFlags) -> Result<usize> {
    syscall1_clobber(SYS_CLONE, flags.bits())
}

/// Close a file
pub fn close(fd: usize) -> Result<usize> {
    unsafe { syscall1(SYS_CLOSE, fd) }
}

/// Get the current system time
pub fn clock_gettime(clock: usize, tp: &mut TimeSpec) -> Result<usize> {
    unsafe { syscall2(SYS_CLOCK_GETTIME, clock, tp as *mut TimeSpec as usize) }
}

/// Copy and transform a file descriptor
pub fn dup(fd: usize, buf: &[u8]) -> Result<usize> {
    unsafe { syscall3(SYS_DUP, fd, buf.as_ptr() as usize, buf.len()) }
}

/// Copy and transform a file descriptor
pub fn dup2(fd: usize, newfd: usize, buf: &[u8]) -> Result<usize> {
    unsafe { syscall4(SYS_DUP2, fd, newfd, buf.as_ptr() as usize, buf.len()) }
}

/// Exit the current process
pub fn exit(status: usize) -> Result<usize> {
    unsafe { syscall1(SYS_EXIT, status) }
}

/// Change file permissions
pub fn fchmod(fd: usize, mode: u16) -> Result<usize> {
    unsafe { syscall2(SYS_FCHMOD, fd, mode as usize) }

}

/// Change file ownership
pub fn fchown(fd: usize, uid: u32, gid: u32) -> Result<usize> {
    unsafe { syscall3(SYS_FCHOWN, fd, uid as usize, gid as usize) }

}

/// Change file descriptor flags
pub fn fcntl(fd: usize, cmd: usize, arg: usize) -> Result<usize> {
    unsafe { syscall3(SYS_FCNTL, fd, cmd, arg) }
}

/// Replace the current process with a new executable
pub fn fexec(fd: usize, args: &[[usize; 2]], vars: &[[usize; 2]]) -> Result<usize> {
    unsafe { syscall5(SYS_FEXEC, fd, args.as_ptr() as usize, args.len(), vars.as_ptr() as usize, vars.len()) }
}

/// Map a file into memory
pub unsafe fn fmap(fd: usize, map: &Map) -> Result<usize> {
    syscall3(SYS_FMAP, fd, map as *const Map as usize, mem::size_of::<Map>())
}

/// Unmap a memory-mapped file
pub unsafe fn funmap(addr: usize) -> Result<usize> {
    syscall1(SYS_FUNMAP, addr)
}

/// Retrieve the canonical path of a file
pub fn fpath(fd: usize, buf: &mut [u8]) -> Result<usize> {
    unsafe { syscall3(SYS_FPATH, fd, buf.as_mut_ptr() as usize, buf.len()) }
}

/// Rename a file
pub fn frename<T: AsRef<[u8]>>(fd: usize, path: T) -> Result<usize> {
    unsafe { syscall3(SYS_FRENAME, fd, path.as_ref().as_ptr() as usize, path.as_ref().len()) }
}

/// Get metadata about a file
pub fn fstat(fd: usize, stat: &mut Stat) -> Result<usize> {
    unsafe { syscall3(SYS_FSTAT, fd, stat as *mut Stat as usize, mem::size_of::<Stat>()) }
}

/// Get metadata about a filesystem
pub fn fstatvfs(fd: usize, stat: &mut StatVfs) -> Result<usize> {
    unsafe { syscall3(SYS_FSTATVFS, fd, stat as *mut StatVfs as usize, mem::size_of::<StatVfs>()) }
}

/// Sync a file descriptor to its underlying medium
pub fn fsync(fd: usize) -> Result<usize> {
    unsafe { syscall1(SYS_FSYNC, fd) }
}

/// Truncate or extend a file to a specified length
pub fn ftruncate(fd: usize, len: usize) -> Result<usize> {
    unsafe { syscall2(SYS_FTRUNCATE, fd, len) }
}

// Change modify and/or access times
pub fn futimens(fd: usize, times: &[TimeSpec]) -> Result<usize> {
    unsafe { syscall3(SYS_FUTIMENS, fd, times.as_ptr() as usize, times.len() * mem::size_of::<TimeSpec>()) }
}

/// Fast userspace mutex
pub unsafe fn futex(addr: *mut i32, op: usize, val: i32, val2: usize, addr2: *mut i32)
                    -> Result<usize> {
    syscall5(SYS_FUTEX, addr as usize, op, (val as isize) as usize, val2, addr2 as usize)
}

/// Get the current working directory
pub fn getcwd(buf: &mut [u8]) -> Result<usize> {
    unsafe { syscall2(SYS_GETCWD, buf.as_mut_ptr() as usize, buf.len()) }
}

/// Get the effective group ID
pub fn getegid() -> Result<usize> {
    unsafe { syscall0(SYS_GETEGID) }
}

/// Get the effective namespace
pub fn getens() -> Result<usize> {
    unsafe { syscall0(SYS_GETENS) }
}

/// Get the effective user ID
pub fn geteuid() -> Result<usize> {
    unsafe { syscall0(SYS_GETEUID) }
}

/// Get the current group ID
pub fn getgid() -> Result<usize> {
    unsafe { syscall0(SYS_GETGID) }
}

/// Get the current namespace
pub fn getns() -> Result<usize> {
    unsafe { syscall0(SYS_GETNS) }
}

/// Get the current process ID
pub fn getpid() -> Result<usize> {
    unsafe { syscall0(SYS_GETPID) }
}

/// Get the process group ID
pub fn getpgid(pid: usize) -> Result<usize> {
    unsafe { syscall1(SYS_GETPGID, pid) }
}

/// Get the parent process ID
pub fn getppid() -> Result<usize> {
    unsafe { syscall0(SYS_GETPPID) }
}

/// Get the current user ID
pub fn getuid() -> Result<usize> {
    unsafe { syscall0(SYS_GETUID) }
}

/// Set the I/O privilege level
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
/// * `EINVAL` - `level > 3`
pub unsafe fn iopl(level: usize) -> Result<usize> {
    syscall1(SYS_IOPL, level)
}

/// Send a signal `sig` to the process identified by `pid`
pub fn kill(pid: usize, sig: usize) -> Result<usize> {
    unsafe { syscall2(SYS_KILL, pid, sig) }
}

/// Create a link to a file
pub unsafe fn link(old: *const u8, new: *const u8) -> Result<usize> {
    syscall2(SYS_LINK, old as usize, new as usize)
}

/// Seek to `offset` bytes in a file descriptor
pub fn lseek(fd: usize, offset: isize, whence: usize) -> Result<usize> {
    unsafe { syscall3(SYS_LSEEK, fd, offset as usize, whence) }
}

/// Make a new scheme namespace
pub fn mkns(schemes: &[[usize; 2]]) -> Result<usize> {
    unsafe { syscall2(SYS_MKNS, schemes.as_ptr() as usize, schemes.len()) }
}

/// Change mapping flags
pub unsafe fn mprotect(addr: usize, size: usize, flags: MapFlags) -> Result<usize> {
    syscall3(SYS_MPROTECT, addr, size, flags.bits())
}

/// Sleep for the time specified in `req`
pub fn nanosleep(req: &TimeSpec, rem: &mut TimeSpec) -> Result<usize> {
    unsafe { syscall2(SYS_NANOSLEEP, req as *const TimeSpec as usize,
                                     rem as *mut TimeSpec as usize) }
}

/// Open a file
pub fn open<T: AsRef<[u8]>>(path: T, flags: usize) -> Result<usize> {
    unsafe { syscall3(SYS_OPEN, path.as_ref().as_ptr() as usize, path.as_ref().len(), flags) }
}

/// Allocate frames, linearly in physical memory.
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
/// * `ENOMEM` - the system has run out of available memory
pub unsafe fn physalloc(size: usize) -> Result<usize> {
    syscall1(SYS_PHYSALLOC, size)
}

/// Allocate frames, linearly in physical memory, with an extra set of flags. If the flags contain
/// [`PARTIAL_ALLOC`], this will result in `physalloc3` with `min = 1`.
///
/// Refer to the simpler [`physalloc`] and the more complex [`physalloc3`], that this convenience
/// function is based on.
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
/// * `ENOMEM` - the system has run out of available memory
pub unsafe fn physalloc2(size: usize, flags: usize) -> Result<usize> {
    let mut ret = 1usize;
    physalloc3(size, flags, &mut ret)
}

/// Allocate frames, linearly in physical memory, with an extra set of flags. If the flags contain
/// [`PARTIAL_ALLOC`], the `min` parameter specifies the number of frames that have to be allocated
/// for this operation to succeed. The return value is the offset of the first frame, and `min` is
/// overwritten with the number of frames actually allocated.
///
/// Refer to the simpler [`physalloc`] and the simpler library function [`physalloc2`].
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
/// * `ENOMEM` - the system has run out of available memory
/// * `EINVAL` - `min = 0`
pub unsafe fn physalloc3(size: usize, flags: usize, min: &mut usize) -> Result<usize> {
    syscall3(SYS_PHYSALLOC3, size, flags, min as *mut usize as usize)
}

/// Free physically allocated pages
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
pub unsafe fn physfree(physical_address: usize, size: usize) -> Result<usize> {
    syscall2(SYS_PHYSFREE, physical_address, size)
}

/// Map physical memory to virtual memory
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
pub unsafe fn physmap(physical_address: usize, size: usize, flags: PhysmapFlags) -> Result<usize> {
    syscall3(SYS_PHYSMAP, physical_address, size, flags.bits())
}

/// Unmap previously mapped physical memory
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
/// * `EFAULT` - `virtual_address` has not been mapped
pub unsafe fn physunmap(virtual_address: usize) -> Result<usize> {
    syscall1(SYS_PHYSUNMAP, virtual_address)
}

/// Create a pair of file descriptors referencing the read and write ends of a pipe
pub fn pipe2(fds: &mut [usize; 2], flags: usize) -> Result<usize> {
    unsafe { syscall2(SYS_PIPE2, fds.as_ptr() as usize, flags) }
}

/// Read from a file descriptor into a buffer
pub fn read(fd: usize, buf: &mut [u8]) -> Result<usize> {
    unsafe { syscall3(SYS_READ, fd, buf.as_mut_ptr() as usize, buf.len()) }
}

/// Remove a directory
pub fn rmdir<T: AsRef<[u8]>>(path: T) -> Result<usize> {
    unsafe { syscall2(SYS_RMDIR, path.as_ref().as_ptr() as usize, path.as_ref().len()) }
}

/// Set the process group ID
pub fn setpgid(pid: usize, pgid: usize) -> Result<usize> {
    unsafe { syscall2(SYS_SETPGID, pid, pgid) }
}

/// Set the current process group IDs
pub fn setregid(rgid: usize, egid: usize) -> Result<usize> {
    unsafe { syscall2(SYS_SETREGID, rgid, egid) }
}

/// Make a new scheme namespace
pub fn setrens(rns: usize, ens: usize) -> Result<usize> {
    unsafe { syscall2(SYS_SETRENS, rns, ens) }
}

/// Set the current process user IDs
pub fn setreuid(ruid: usize, euid: usize) -> Result<usize> {
    unsafe { syscall2(SYS_SETREUID, ruid, euid) }
}

/// Set up a signal handler
pub fn sigaction(sig: usize, act: Option<&SigAction>, oldact: Option<&mut SigAction>) -> Result<usize> {
    unsafe { syscall4(SYS_SIGACTION, sig,
                      act.map(|x| x as *const _).unwrap_or_else(ptr::null) as usize,
                      oldact.map(|x| x as *mut _).unwrap_or_else(ptr::null_mut) as usize,
                      restorer as usize) }
}

/// Get and/or set signal masks
pub fn sigprocmask(how: usize, set: Option<&[u64; 2]>, oldset: Option<&mut [u64; 2]>) -> Result<usize> {
    unsafe { syscall3(SYS_SIGPROCMASK, how,
                      set.map(|x| x as *const _).unwrap_or_else(ptr::null) as usize,
                      oldset.map(|x| x as *mut _).unwrap_or_else(ptr::null_mut) as usize) }
}

// Return from signal handler
pub fn sigreturn() -> Result<usize> {
    unsafe { syscall0(SYS_SIGRETURN) }
}

/// Set the file mode creation mask
pub fn umask(mask: usize) -> Result<usize> {
    unsafe { syscall1(SYS_UMASK, mask) }
}

/// Remove a file
pub fn unlink<T: AsRef<[u8]>>(path: T) -> Result<usize> {
    unsafe { syscall2(SYS_UNLINK, path.as_ref().as_ptr() as usize, path.as_ref().len()) }
}

/// Convert a virtual address to a physical one
///
/// # Errors
///
/// * `EPERM` - `uid != 0`
pub unsafe fn virttophys(virtual_address: usize) -> Result<usize> {
    syscall1(SYS_VIRTTOPHYS, virtual_address)
}

/// Check if a child process has exited or received a signal
pub fn waitpid(pid: usize, status: &mut usize, options: WaitFlags) -> Result<usize> {
    unsafe { syscall3(SYS_WAITPID, pid, status as *mut usize as usize, options.bits()) }
}

/// Write a buffer to a file descriptor
///
/// The kernel will attempt to write the bytes in `buf` to the file descriptor `fd`, returning
/// either an `Err`, explained below, or `Ok(count)` where `count` is the number of bytes which
/// were written.
///
/// # Errors
///
/// * `EAGAIN` - the file descriptor was opened with `O_NONBLOCK` and writing would block
/// * `EBADF` - the file descriptor is not valid or is not open for writing
/// * `EFAULT` - `buf` does not point to the process's addressible memory
/// * `EIO` - an I/O error occurred
/// * `ENOSPC` - the device containing the file descriptor has no room for data
/// * `EPIPE` - the file descriptor refers to a pipe or socket whose reading end is closed
pub fn write(fd: usize, buf: &[u8]) -> Result<usize> {
    unsafe { syscall3(SYS_WRITE, fd, buf.as_ptr() as usize, buf.len()) }
}

/// Yield the process's time slice to the kernel
///
/// This function will return Ok(0) on success
pub fn sched_yield() -> Result<usize> {
    unsafe { syscall0(SYS_YIELD) }
}

/// Ask a scheme to attach an already-opened io_uring. The target scheme will become the consumer
/// of the submission queue, and the producer of the completion queue. This syscall is required
/// before an io_uring can operate properly.
///
/// The scheme will receive a [`recv_io_uring`] syscall ([`SYS_RECV_IORING`]), rather than this
/// syscall directly, due to the kernel having to interfere when allocating and mapping memory.
///
/// [`recv_io_uring`]: ../scheme/scheme/trait.Scheme.html#method.recv_io_uring
/// [`SYS_RECV_IORING`]: ../number/constant.SYS_RECV_IORING.html
pub fn attach_iouring(ringfd: usize, scheme_name: &[u8]) -> Result<usize> {
    unsafe { syscall3(SYS_ATTACH_IORING, ringfd, scheme_name.as_ptr() as usize, scheme_name.len()) }
}

/// Tell the kernel that the current process has nothing left to do, and is ready to wait for
/// additional io_uring entries before waking up again.
///
/// The `min_complete` parameter defines the minimum number of completion entries that must be
/// pushed onto the completion queue, before the function is able to return. This can also take a
/// value of zero, in which case the ring can wake up without any entries pushed onto the queue,
/// which can be especially useful in async executors that want to stop waiting, and tell the
/// executor to continue.
pub fn enter_iouring(ringfd: usize, min_complete: usize, flags: io_uring::IoUringEnterFlags) -> Result<usize> {
    unsafe { syscall3(SYS_ENTER_IORING, ringfd, min_complete, flags.bits()) }
}