Commit graph

133 commits

Author SHA1 Message Date
Ben Gras
c3c08d252c report driver source of grant mismatch 2006-08-04 13:31:06 +00:00
Ben Gras
18327f02a8 Introduced unprivileged getsysinfo variant, to retrieve harmless data
in formats that don't change (or is upwards compatible).
2006-07-27 16:23:01 +00:00
Ben Gras
82173212ee Introduction of a 'big' ioctl, with more bits (20) reserved for encoding the
size field. The TIOCSFON ioctl size (8192) didn't get encoded properly,
as there weren't enough bits for it (12) in the regular format.

The new format has only one type field, and an extra flag (_IOC_BIG)
turned on. FS checks for this flag and uses the alternative decoding
of the ioctl codes to determine the size when doing grants.

This unbreaks loadfont, although that still uses a phys copy in tty.
2006-07-25 09:41:40 +00:00
Philip Homburg
61337d5f7b Allow holes in executables. 2006-07-24 13:07:25 +00:00
Philip Homburg
7ce17fe655 Support for SI_CALL_STATS (counting systemcalls). 2006-07-10 12:44:43 +00:00
Ben Gras
1561067ee4 Grant system dynamic-only. 2006-06-30 14:40:29 +00:00
Ben Gras
0323892f71 3rd fix for improper behaviour on crashing driver with safe_io_conversion 2006-06-29 14:24:07 +00:00
Ben Gras
bfca7d68ba Fix another bug related to trying i/o more than once 2006-06-28 12:20:30 +00:00
Ben Gras
f9fb0ff546 Fix bug where safe conversion doesn't happen after 1st time in loop 2006-06-28 10:03:18 +00:00
Ben Gras
bd535a120b . satisfy some gcc warnings (uninitialized/unused variables)
. change cloexec mask from long to fd_set to remove 32 fd's per
   process restriction (from cloexec at least)
2006-06-27 16:47:35 +00:00
Ben Gras
a587273c56 . expanded grant manipulation interface to more direct
manipulation
. made fs allocate a grant for every ioctl, even if no data
  is being copied, in order to disambiguate concurrent ioctls
  on the same minor
2006-06-26 16:08:42 +00:00
Philip Homburg
2f50f92eb3 Move inet to safe copies. 2006-06-26 16:03:35 +00:00
Ben Gras
82855e9cf5 . leave out dead code from device.c
. don't loop doing a receive() after sendrec() - chance of recovering is not
  high, and can lead to receive()ing a notify() (which can't happen in sendrec()),
  which is terrible
. return status from device when DEV_CANCEL is done on a signal; hardcode EAGAIN to
  become EINTR though
2006-06-23 11:51:56 +00:00
Ben Gras
2384a85296 FS support for grant-based i/o.
For character device i/o, FS does a so-called 'magic' grant to let the
driver copy from or to user space. As this is done in FS address space,
the driver is told to do this in FS address space. The redirection to
the right user process then happens at copy-time in the kernel, using the
FS grant table. This also happens for DEV_READ and DEV_WRITE on block
devices.

For other block device i/o, which happens from/to FS buffers, FS does
a 'direct' grant to its own address space for the driver.

After the i/o returns, this access has to be K-I-L-L-E-D, revoked.
Sometimes this is after a SUSPEND and DEV_REVIVE, in which case the
revoking happens in pipe.c.

This conversion happens in safe_io_conversion() in device.c, called
by dev_io and dev_bio.

FS has to pre-allocate its own space for these grant tables. This happens
in main.c.
2006-06-20 10:12:09 +00:00
Philip Homburg
221e731e45 Wake up writer when selecting for read on an empty pipe.
Set fp_revived to NOT_REVIVING when decrementing reviving.
2006-06-14 13:17:41 +00:00
Philip Homburg
cb02a90b77 Fixed some select bugs related to pipes. Removed SELFD_* because they were
bogus and unused.
2006-06-07 14:41:47 +00:00
Philip Homburg
e9aabcf2f8 Disabled building rescue driver (no longer needed). Moved allocmem from
library to the memory driver. Always put output from within TTY directly on
the console. Removed second include of driver.h from tty.c. Made tty_inrepcode
bigger. First step to move PM and FS calls that are not regular (API)
system calls out of callnr.h (renumbered them, and removed them from the
table.c files). Imported the Minix-vmd uname implementation. This provides
a more stable ABI than the current implementation. Added a bit of security
checking. Unfortunately not nearly enough to get a secure system. Fixed a
bug related to the sizes of the programs in the image (in PM patch_mem_chunks).
2006-05-19 12:19:37 +00:00
Philip Homburg
f0186f4179 Do not unmap a driver when the driver dies (it interferes with restarting
disk drivers). Fixed accounting for REVIVING/reviving.
2006-05-15 11:43:06 +00:00
Philip Homburg
773844a816 New interface between PM and FS. 2006-05-11 14:57:23 +00:00
Ben Gras
461a4fafb1 Added fchmod() and fchown() 2006-04-18 11:26:04 +00:00
Ben Gras
7497a76779 More filesystems. 2006-03-28 11:34:59 +00:00
Ben Gras
2d5058685d Don't open pipes readwrite as it leads to confusion. 2006-03-28 11:28:16 +00:00
Ben Gras
df9f4f4369 . minor fix in select() logic on pipes
. fix in pipe_check when 'notouch' is on - we touched something!
2006-03-24 20:53:10 +00:00
Ben Gras
26a59eea43 readlink(); check bounds
top: add 'system' cpu time to 'kernel', 'idle' and 'user'
2006-03-17 13:34:40 +00:00
Ben Gras
6e7ba75acb Driver disappearing during dev_io() is not a fatal error. 2006-03-15 21:55:32 +00:00
Ben Gras
ff67776995 FS:
. loops checked for PID_FREE
  . exit broken down in exit and cleanup functions; when reboot happens,
    cleanup is done but not exit (as processes have not actually exited),
    this keeps drivers working
  . fixed a few uninitialized and unused variables

scripts:
  . new packaging system
2006-03-15 15:34:12 +00:00
Philip Homburg
50d805144c dmap_io now returns a status. map_driver no longer calls dev_up. 2006-03-15 13:37:20 +00:00
Ben Gras
def8c13ba8 Do a sync() before reboot()ing in FS. 2006-03-14 13:54:20 +00:00
Jorrit Herder
021e3234d8 Jorrit's ... "progress?" 2006-03-10 16:10:05 +00:00
Ben Gras
6fe89cfbe8 Hack small and big fs into a single image. Must now set BIG or SMALL label. 2006-03-10 14:11:19 +00:00
Ben Gras
25fdc7657d panic() on inconsistent endpoints 2006-03-10 12:59:46 +00:00
Ben Gras
6746ee10f5 If re-open fails, mark fd as unusable but not-reusable either (until
close()d).
2006-03-09 16:00:33 +00:00
Ben Gras
b103bf5dc7 Fix of endpointless slot message in FS
Removed some leftover '#if 0'ed debug code in PM
2006-03-03 10:25:28 +00:00
Ben Gras
7967177710 endpoint-aware conversion of servers.
'who', indicating caller number in pm and fs and some other servers, has
been removed in favour of 'who_e' (endpoint) and 'who_p' (proc nr.).

In both PM and FS, isokendpt() convert endpoints to process slot
numbers, returning OK if it was a valid and consistent endpoint number.
okendpt() does the same but panic()s if it doesn't succeed. (In PM,
this is pm_isok..)

pm and fs keep their own records of process endpoints in their proc tables,
which are needed to make kernel calls about those processes.

message field names have changed.

fs drivers are endpoints.

fs now doesn't try to get out of driver deadlock, as the protocol isn't
supposed to let that happen any more. (A warning is printed if ELOCKED
is detected though.)

fproc[].fp_task (indicating which driver the process is suspended on)
became an int.

PM and FS now get endpoint numbers of initial boot processes from the
kernel. These happen to be the same as the old proc numbers, to let
user processes reach them with the old numbers, but FS and PM don't know
that. All new processes after INIT, even after the generation number
wraps around, get endpoint numbers with generation 1 and higher, so
the first instances of the boot processes are the only processes ever
to have endpoint numbers in the old proc number range.

More return code checks of sys_* functions have been added.

IS has become endpoint-aware. Ditched the 'text' and 'data' fields
in the kernel dump (which show locations, not sizes, so aren't terribly
useful) in favour of the endpoint number. Proc number is still visible.

Some other dumps (e.g. dmap, rs) show endpoint numbers now too which got
the formatting changed.

PM reading segments using rw_seg() has changed - it uses other fields
in the message now instead of encoding the segment and process number and
fd in the fd field. For that it uses _read_pm() and _write_pm() which to
_taskcall()s directly in pm/misc.c.

PM now sys_exit()s itself on panic(), instead of sys_abort().

RS also talks in endpoints instead of process numbers.
2006-03-03 10:20:58 +00:00
Ben Gras
7d83ef196d Don't include cdprobe any more 2006-02-24 12:56:06 +00:00
Philip Homburg
c3a88d15d8 Initial root filesystem is now on a ramdisk that is part of the image. 2006-02-15 11:18:21 +00:00
Ben Gras
b9bdbcba1b Don't sync before exiting drivers.
Wipe inodes of pipes in truncate_inode(), to avoid the size remaining
at PIPE_SIZE causing pipes to break.
2006-02-13 10:28:42 +00:00
Ben Gras
040cebcefe Don't kill processes by signals in process group 0, and do a sync before
killing all processes in FS reboot.
2006-02-03 15:21:00 +00:00
Philip Homburg
f4ae126596 Cosmetic changes. 2006-01-27 13:26:42 +00:00
Ben Gras
43051c182a Safety feature for symlink() - verify length of symlink contents
buffer with strlen()
2006-01-26 14:31:00 +00:00
Ben Gras
f806a66edf Fix for fchdir(). 2006-01-20 12:43:35 +00:00
Ben Gras
3bddaaf580 Let other uids (temporarily) set their mapping. 2006-01-20 12:43:19 +00:00
Ben Gras
966905d295 Not having a driver isn't a print-worthy error. Just return ENXIO. 2006-01-17 17:55:35 +00:00
Ben Gras
50b06261b6 Implementation of truncate(), ftruncate() and the F_FREESP fcntl().
Implemented by changing write_map to accept a WMAP_FREE flag. In that
case, it doesn't update the datablock (creating indirect zones as
necessary) pointer, but it frees the datablock if present. Also it
frees the single and double indirect blocks if unused.

This makes the implementation of truncate_inode() simpler.
truncate_inode() now accepts a truncation length which makes
implementing truncate() and ftruncate() simple.

This also allowed implementing the F_FREESP fcntl().
2006-01-11 17:14:51 +00:00
Ben Gras
6a911b5284 Merge of minix-vmd symlink code.
. new_node() now returns inode of parent dir as argument that
	  has to be put_node()d again by the caller of new_node().
	  it can also return the name of the last component as last_dir()
	  did.
	. advance() takes a pointer to a pointer of an inode as the
	  parent now. This parent can change, in which case the old
	  one is put_node()d and a new one is put there.
	. eat_path() replaced by more flexible parse_path()
	. last_dir() replaced by call to parse_path().
	. do_slink(), do_readlink(), do_lstat() and slink_traverse() added.

Also added some truncate()/ftruncate()-introduction related changes.
(e.g. renamed truncate() to truncate_inode().)
2005-12-20 14:23:44 +00:00
Ben Gras
743bdb038e Backout of symlink patches. 2005-12-19 10:45:50 +00:00
Ben Gras
eb1919722d Priority explicitization. 2005-12-14 12:08:49 +00:00
Ben Gras
f6e5052597 seteuid() and setegid() 2005-12-13 09:48:29 +00:00
Ben Gras
dd99ee67e0 Rename some #defines such as MAX_BLOCK_SIZE to _MAX_BLOCK_SIZE and such
so as to not pollute application namespace.
2005-11-14 18:41:27 +00:00
Ben Gras
fa76dcb8c1 Giovanni's symlink patches. 2005-11-01 13:19:40 +00:00