Alan Cox [Wed, 30 Apr 2008 07:53:54 +0000 (00:53 -0700)]
istallion: TIOCG/SSOFTCAR handling removal
This is handled (and correctly) by the core code so does not belong
incorrectly in the driver.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:54 +0000 (00:53 -0700)]
Char: ip2, macros cleanup
- remove i2os.h -- there was only macro to macro renaming or useless
stuff
- remove another uselless stuf (NULLFUNC, NULLPTR, YES, NO)
- use outb/inb directly
- use locking functions directly
- don't define another ROUNDUP, use roundup(x, 2) instead
- some comments and whitespace cleanup
- remove some commented crap
- prepend the rest by I2 prefix to not collide with rest of the world
like in following output (pointed out by akpm)
In file included from drivers/char/ip2/ip2main.c:128:
drivers/char/ip2/i2ellis.h:608:1: warning: "COMPLETE" redefined
In file included from include/net/netns/ipv4.h:8,
from include/net/net_namespace.h:13,
from include/linux/seq_file.h:7,
from include/asm/machdep.h:12,
from include/asm/pci.h:17,
from include/linux/pci.h:951,
from drivers/char/ip2/ip2main.c:95:
include/net/inet_frag.h:28:1: warning: this is the location of the previous definition
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jon Schindler [Wed, 30 Apr 2008 07:53:53 +0000 (00:53 -0700)]
drivers/char/ip2/ip2main.c: replace init_module&cleanup_module with module_init&module_exit
Replace init_module and cleanup_module with static functions and
module_init/module_exit.
Signed-off-by: Jon Schindler <jkschind@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Harvey Harrison [Wed, 30 Apr 2008 07:53:52 +0000 (00:53 -0700)]
epca.c: static functions and integer as NULL pointer fixes
drivers/char/epca.c:926:28: warning: Using plain integer as NULL pointer
drivers/char/epca.c:1841:2: warning: Using plain integer as NULL pointer
Forward declarations were already marked static, mark the definitions too.
drivers/char/epca.c:2493:6: warning: symbol 'digi_send_break' was not declared. Should it be static?
drivers/char/epca.c:2881:12: warning: symbol 'init_PCI' was not declared. Should it be static?
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Harvey Harrison [Wed, 30 Apr 2008 07:53:52 +0000 (00:53 -0700)]
cyclades.c: fix sparse shadowed variable warnings
Nested min() macros.
drivers/char/cyclades.c:2750:7: warning: symbol '_x' shadows an earlier one
drivers/char/cyclades.c:2750:7: originally declared here
drivers/char/cyclades.c:2750:7: warning: symbol '_x' shadows an earlier one
drivers/char/cyclades.c:2750:7: originally declared here
drivers/char/cyclades.c:2750:7: warning: symbol '_y' shadows an earlier one
drivers/char/cyclades.c:2750:7: originally declared here
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Harvey Harrison [Wed, 30 Apr 2008 07:53:51 +0000 (00:53 -0700)]
char: rocket.c: fix sparse variable shadowing and int as NULL pointer
Nested min() macros shadow _x, separate into two lines.
drivers/char/rocket.c:451:7: warning: symbol '_x' shadows an earlier one
drivers/char/rocket.c:451:7: originally declared here
drivers/char/rocket.c:451:7: warning: symbol '_x' shadows an earlier one
drivers/char/rocket.c:451:7: originally declared here
drivers/char/rocket.c:451:7: warning: symbol '_y' shadows an earlier one
drivers/char/rocket.c:451:7: originally declared here
drivers/char/rocket.c:1754:7: warning: symbol '_x' shadows an earlier one
drivers/char/rocket.c:1754:7: originally declared here
drivers/char/rocket.c:1754:7: warning: symbol '_x' shadows an earlier one
drivers/char/rocket.c:1754:7: originally declared here
drivers/char/rocket.c:1754:7: warning: symbol '_y' shadows an earlier one
drivers/char/rocket.c:1754:7: originally declared here
drivers/char/rocket.c:1751:20: warning: Using plain integer as NULL pointer
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Harvey Harrison [Wed, 30 Apr 2008 07:53:50 +0000 (00:53 -0700)]
char: esp.c: fix possible double-unlock
Hitting either of the break statements in the while loop would cause a
double-unlock of info->lock.
[Jiri Slaby suggested simply returning is safe here, rather than a goto]
Noticed by sparse:
drivers/char/esp.c:2042:2: warning: context imbalance in 'rs_wait_until_sent' - unexpected unlock
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Cc: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Harvey Harrison [Wed, 30 Apr 2008 07:53:50 +0000 (00:53 -0700)]
char: fix sparse shadowed variable warnings in esp.c
flags only use was in spin_lock_irqsave/spin_lock_irgrestore pairs, no
need to redeclare for each one.
drivers/char/esp.c:1599:17: warning: symbol 'flags' shadows an earlier one
drivers/char/esp.c:1517:16: originally declared here
drivers/char/esp.c:1615:17: warning: symbol 'flags' shadows an earlier one
drivers/char/esp.c:1517:16: originally declared here
drivers/char/esp.c:1631:17: warning: symbol 'flags' shadows an earlier one
drivers/char/esp.c:1517:16: originally declared here
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:49 +0000 (00:53 -0700)]
Char: moxa, add firmware loading fix
Be more verbose on fw load fail as noted by Oyvind.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Cc: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:48 +0000 (00:53 -0700)]
Char: moxa, update credits
- update version
- update maintainers
- copyright the stuff
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:48 +0000 (00:53 -0700)]
Char: moxa, notify about board readiness
Drop a message to dmesg about card being ready.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:47 +0000 (00:53 -0700)]
Char: moxa, introduce MOXA_IS_320 macro
It allows to simplify the code, especially MoxaPortSetBaud.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:47 +0000 (00:53 -0700)]
Char: moxa, remove useless tty functions
- moxa_flush_chars -- no code; ldics handle this well
- moxa_put_char -- only wrapper to moxa_write (same code), tty does this
the same way if tty->driver->put_char is NULL
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:46 +0000 (00:53 -0700)]
Char: moxa, little cleanup
Cleanup of
- whitespace
- macros
- useless casts
- return (sth); -> return sth;
- types
- superfluous parenthesis and braces
- init tmp directly in moxa_get_serial_info
- commented defunct code
- commented prototypes
- MOXA/moxa printk case
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:45 +0000 (00:53 -0700)]
Char: moxa, rework open/close
- add locking to open/close/hangup and ioctl (tiocm)
- add pci hot-un-plug support (hangup on board remove, wait for openers)
- cleanup block_till_ready
- move close code common to close/hangup into separate function to be
able to call it from open when hangup occurs while block_till_ready
- let ldisc flush on tty layer, it will do it after we return
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:45 +0000 (00:53 -0700)]
Char: moxa, serialise timer
- del timer after we are sure it won't be fired again
- make timer scheduling atomic
- don't reschedule timer when all cards have gone
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:44 +0000 (00:53 -0700)]
Char: moxa, cleanup rx/tx
- cleanup types
- use tty_prepare_flip_string and io memcpys
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:43 +0000 (00:53 -0700)]
Char: moxa, merge 2 poll functions
- merge 2 timers into one -- one can handle the emptywait as good as the other
- merge 2 separated poll functions into one, this allows handle the actions
directly and simplifies the code
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:43 +0000 (00:53 -0700)]
Char: moxa, ioctl cleanup
- allow stats only for sys_admin
- move TCSBRK* processing to .break_ctl tty op
- let TIOCGSOFTCAR and TIOCSSOFTCAR be processed by ldisc
- remove MOXA_GET_MAJOR, MOXA_GET_CUMAJOR
- fix jiffies subtraction by time_after
- move moxa ioctl numbers into the header; still not exported to userspace,
needs _IOC and 32/64 compat cleanup anyways
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:42 +0000 (00:53 -0700)]
Char: moxa, timer cleanup
- schedule timer even after some card is installed, not after insmod
- cleanup timer functions
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:41 +0000 (00:53 -0700)]
Char: moxa, centralize board readiness
The only relevant sign of port being ready is its board->ready since now.
Remove all other flags for this purpose which are set almost on the same
place. Move ports inside the board to be sure that nobody will grab reference
to the port without being sure that it exists.
[jirislaby@gmail.com: fix unused var warning]
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:41 +0000 (00:53 -0700)]
Char: moxa, remove unused port entries
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:40 +0000 (00:53 -0700)]
Char: moxa, remove port->port
We don't need to hold a reference to port index. In most cases we need port
structure anyway and index is available in port->tty->index.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:39 +0000 (00:53 -0700)]
Char: moxa, merge c2xx and c320 firmware loading
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:39 +0000 (00:53 -0700)]
Char: moxa, add firmware loading
Substitute ioctl load firmware interface by kernel firmware api.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:38 +0000 (00:53 -0700)]
Char: moxa, fix TIOC(G/S)SOFTCAR param
according to ioctl_list, both have int * as a param, not ulong *.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:37 +0000 (00:53 -0700)]
Char: moxa, pci io space fixup
- request region before remapping pci io space
- use ioremap, iounmap istead of iomap interface, because we use
readX/writeX for accessing this space because of isa support
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:37 +0000 (00:53 -0700)]
Char: moxa, cleanup module-param passed isa init
Make the code more readable, remap the base address directly. Describe module
parameters.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jiri Slaby [Wed, 30 Apr 2008 07:53:36 +0000 (00:53 -0700)]
Char: moxa, remove static isa support
Static ISA field is empty and probably will never be filled in, remove it.
The driver still supports ISA cards passed through module parameter. This
actually fixes one bug inside the initialization of module-param passed cards
initialization.
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Tested-by: Oyvind Aabling <Oyvind.Aabling@uni-c.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Dimitri Sivanich [Wed, 30 Apr 2008 07:53:35 +0000 (00:53 -0700)]
SGI Altix mmtimer: allow larger number of timers per node
The purpose of this patch to the SGI Altix specific mmtimer (posix timer)
driver is to allow a virtually infinite number of timers to be set per
node.
Timers will now be kept on a sorted per-node list and a single node-based
hardware comparator is used to trigger the next timer.
[akpm@linux-foundation.org: mark things static]
[akpm@linux-foundation.org: fix warning]
Signed-off-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:34 +0000 (00:53 -0700)]
tty: drop the BKL for driver/ldisc ioctl methods
Now we have pushed the lock down we can stop wrapping the call with a lock in
the tty layer.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:34 +0000 (00:53 -0700)]
tty_ioctl: soft carrier handling
First cut at moving the soft carrier handling knowledge entirely into the core
code. One or two drivers still needed to snoop these functions to track
CLOCAL internally. Instead make TIOCSSOFTCAR generate the same driver calls
as other termios ioctls changing the clocal flag. This allows us to remove
any driver knowledge and special casing. Also while we are at it we can fix
the error handling.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:32 +0000 (00:53 -0700)]
tty_ioctl: locking for tty_wait_until_sent
This function still depends on the big kernel lock in some cases. Push
locking into the function ready for removal of the BKL from ioctl call paths.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:31 +0000 (00:53 -0700)]
tty_io: fix remaining pid struct locking
This fixes the last couple of pid struct locking failures I know about.
[oleg@tv-sign.ru: clean up do_task_stat()]
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Joe Peterson [Wed, 30 Apr 2008 07:53:30 +0000 (00:53 -0700)]
Resume TTY on SUSP and fix CRNL order in N_TTY line discipline
Refine these behaviors in the N_TTY line discipline:
1) Handle the signal characters consistently when received in a stopped TTY
so that SUSP (typically ctrl-Z) behaves like INTR and QUIT in resuming a
stopped TTY.
2) Adjust the order in which the IGNCR/ICRNL/INLCR processing is applied to
be more logical and consistent with the behavior of other Unix systems.
Signed-off-by: Joe Peterson <joe@skyrush.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:30 +0000 (00:53 -0700)]
redo locking of tty->pgrp
Historically tty->pgrp and friends were pid_t and the code "knew" they were
safe. The change to pid structs opened up a few races and the removal of the
BKL in places made them quite hittable. We put tty->pgrp under the ctrl_lock
for the tty.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:29 +0000 (00:53 -0700)]
tty: BKL pushdown
- Push the BKL down into the line disciplines
- Switch the tty layer to unlocked_ioctl
- Introduce a new ctrl_lock spin lock for the control bits
- Eliminate much of the lock_kernel use in n_tty
- Prepare to (but don't yet) call the drivers with the lock dropped
on the paths that historically held the lock
BKL now primarily protects open/close/ldisc change in the tty layer
[jirislaby@gmail.com: a couple of fixes]
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:28 +0000 (00:53 -0700)]
serial_core: Prepare for BKL push down
Instead of checking for the BKL in these methods, take it ourselves. That
avoids propogating it into the serial drivers and we can then fix them later
on.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:27 +0000 (00:53 -0700)]
68360serial: Note that there isn't any info->mcr locking
Noticed while auditing the code for the BKL elimination project
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:27 +0000 (00:53 -0700)]
isdn_tty: Prepare for BKL push down
Three things here
- Remove softcar handler
- Correct termios change detection logic
- Wrap break/ioctl in lock_kernel ready to drop it in the caller
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Karsten Keil <kkeil@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:26 +0000 (00:53 -0700)]
vt_ioctl: Prepare for BKL push down
This one could do with some eyeballs on it. In theory it simply wraps the
ioctl handler in lock/unlock_kernel ready for the lock/unlocks to be pushed
into specific switch values. To do that means changing the code to return via
a common exit path not all over the place as it does now, hence the big diff
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:25 +0000 (00:53 -0700)]
viocons: BKL locking
For some weird reason I can't ascertain (translation "I think its
broken") the viocons driver calls directly into the n_tty ldisc code even
if another ldisc is in use. It'll probably break if you do that but I'm
just fixing the locking and adding a comment that its horked.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:24 +0000 (00:53 -0700)]
synclink series: Prepare for BKL pushdown
As these are quite complex I've simply pushed the BKL down into the ioctl
handler not tried to do anything neater.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Paul Fulghum <paulkf@microgate.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:24 +0000 (00:53 -0700)]
sx: prepare for BKL pushdown
Wrap the ioctl handler, and in this case the break handler also in the
BKL. Remove bogus softcar handlers.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:23 +0000 (00:53 -0700)]
stallion: Prepare for BKL push down
Remove broken softcar functions, wrap ioctl handler in BKL
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:22 +0000 (00:53 -0700)]
specialix: Prepare for BKL pushdown
Lock the ioctl handlers and remove bogus softcar handling.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:22 +0000 (00:53 -0700)]
serial167: prepare to push BKL down into drivers
Kill the softcar handlers again, wrap the ioctl handler in the BKL
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:21 +0000 (00:53 -0700)]
rocket: Prepare for BKL pushdown
Wrap the ioctl code in lock_kernel calls
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:21 +0000 (00:53 -0700)]
riscom8: Prepare for BKL pushdown
Push the locking down into a couple of functions that need it and remove
bogus TIOCG/SSOFTCAR handling
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:20 +0000 (00:53 -0700)]
mxser: prepare for BKL pushdown
Push the BKL down into various internal routines in the driver ready to
remove it from the break, ioctl and other call points.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: Jiri Slaby <jirislaby@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:19 +0000 (00:53 -0700)]
isicom: istallion prepare for lock_kernel pushdown
This is an ancient driver so just wrap it in lock_kernel internally and
be done.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:19 +0000 (00:53 -0700)]
isicom: prepare for lock_kernel push down
Again lock the bits we can't trivially prove are safe without the BKL and
remove the broken TIOCS/GSOFTCAR handler.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:18 +0000 (00:53 -0700)]
esp: lock_kernel push down
Push the BKL down into a few internal bits of code in this driver.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:17 +0000 (00:53 -0700)]
epca: lock_kernel push down
Prepare epca for removing the lock from above. Most of epca is internally
locked so we can trivially push it down to a few bits of code. Drop the TIOCG/SSOFTCAR handling as that is done *properly* with locks by the mid layer.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:16 +0000 (00:53 -0700)]
cyclades: Prepare for relaxed locking in callers
Basically wrap it in lock_kernel where it is hard to prove the locking is
ok.
Signed-off-by: Alan Cox <alan@redhat.com>
Cc: "John Stoffel" <john@stoffel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alan Cox [Wed, 30 Apr 2008 07:53:16 +0000 (00:53 -0700)]
amiserial: prepare for locking relaxation in caller
Just wrap this one in a lock_kernel. As I understand it there is no M68K
SMP anyway.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:15 +0000 (00:53 -0700)]
ptrace: permit ptracing of /sbin/init
Afaics, currently there are no kernel problems with ptracing init, it can't
lose SIGNAL_UNKILLABLE flag and be killed/stopped by accident.
The ability to strace/debug init can be very useful if you try to figure out
why it does not work as expected.
However, admin should know what he does, "gdb /sbin/init 1" stops init, it
can't reap orphaned zombies or take care of /etc/inittab until continued. It
is even possible to crash init (and thus the whole system) if you wish,
ptracer has full control.
See also the long discussion: http://marc.info/?t=
120628018600001
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Roland McGrath <roland@redhat.com>
Acked-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:14 +0000 (00:53 -0700)]
ptrace: ptrace_attach: use send_sig_info() instead force_sig_specific()
Nobody can block/ignore SIGSTOP, no need to use force_sig_specific() in
ptrace_attach. Use the "regular" send_sig_info().
With this patch stracing of /sbin/init doesn't clear its SIGNAL_UNKILLABLE,
but not that this makes ptracing of init safe.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:14 +0000 (00:53 -0700)]
ptrace: __ptrace_unlink: use the ptrace_reparented() helper
Currently __ptrace_unlink() checks list_empty(->ptrace_list) to figure out
whether the child was reparented. Change the code to use ptrace_reparented()
to make this check more explicit and consistent.
No functional changes.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:13 +0000 (00:53 -0700)]
ptrace: introduce ptrace_reparented() helper
Add another trivial helper for the sake of grep. It also auto-documents the
fact that ->parent != real_parent implies ->ptrace.
No functional changes.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:12 +0000 (00:53 -0700)]
document de_thread() with exit_notify() connection
Add a couple of small comments, it is not easy to see what this code does.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:12 +0000 (00:53 -0700)]
reparent_thread: use same_thread_group()
Trivial, use same_thread_group() in reparent_thread().
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:11 +0000 (00:53 -0700)]
ptrace: introduce task_detached() helper
exit.c has numerous "->exit_signal == -1" comparisons, this check is subtle
and deserves a helper. Imho makes the code more parseable for humans. At
least it's surely more greppable.
Also, a couple of whitespace cleanups. No functional changes.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Wed, 30 Apr 2008 07:53:10 +0000 (00:53 -0700)]
signals: x86 TS_RESTORE_SIGMASK
Replace TIF_RESTORE_SIGMASK with TS_RESTORE_SIGMASK and define our own
set_restore_sigmask() function. This saves the costly SMP-safe set_bit
operation, which we do not need for the sigmask flag since TIF_SIGPENDING
always has to be set too.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Wed, 30 Apr 2008 07:53:09 +0000 (00:53 -0700)]
signals: use HAVE_SET_RESTORE_SIGMASK
Change all the #ifdef TIF_RESTORE_SIGMASK conditionals in non-arch code to
#ifdef HAVE_SET_RESTORE_SIGMASK. If arch code defines it first, the generic
set_restore_sigmask() using TIF_RESTORE_SIGMASK is not defined.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
akpm@linux-foundation.org [Wed, 30 Apr 2008 07:53:09 +0000 (00:53 -0700)]
signals: ia64 renumber TIF_RESTORE_SIGMASK
TIF_RESTORE_SIGMASK no longer needs to be in the _TIF_WORK_* masks.
Those low bits are scarce. Renumber TIF_RESTORE_SIGMASK to free one up.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Wed, 30 Apr 2008 07:53:08 +0000 (00:53 -0700)]
signals: s390: renumber TIF_RESTORE_SIGMASK
TIF_RESTORE_SIGMASK no longer needs to be in the _TIF_WORK_* masks. Those low
bits are scarce, and are all used up now. Renumber TIF_RESTORE_SIGMASK to
free one up.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Wed, 30 Apr 2008 07:53:07 +0000 (00:53 -0700)]
signals: set_restore_sigmask TIF_SIGPENDING
Set TIF_SIGPENDING in set_restore_sigmask. This lets arch code take
TIF_RESTORE_SIGMASK out of the set of bits that will be noticed on return to
user mode. On some machines those bits are scarce, and we can free this
unneeded one up for other uses.
It is probably the case that TIF_SIGPENDING is always set anyway everywhere
set_restore_sigmask() is used. But this is some cheap paranoia in case there
is an arcane case where it might not be.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Roland McGrath [Wed, 30 Apr 2008 07:53:06 +0000 (00:53 -0700)]
signals: add set_restore_sigmask
This adds the set_restore_sigmask() inline in <linux/thread_info.h> and
replaces every set_thread_flag(TIF_RESTORE_SIGMASK) with a call to it. No
change, but abstracts the details of the flag protocol from all the calls.
Signed-off-by: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:05 +0000 (00:53 -0700)]
signals: allow the kernel to actually kill /sbin/init
Currently the buggy /sbin/init hangs if SIGSEGV/etc happens. The kernel sends
the signal, init dequeues it and ignores, returns from the exception, repeats
the faulting instruction, and so on forever.
Imho, such a behaviour is not good. I think that the explicit loud death of
the buggy /sbin/init is better than the silent hang.
Change force_sig_info() to clear SIGNAL_UNKILLABLE when the task should be
really killed.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:04 +0000 (00:53 -0700)]
signals: de_thread: simplify the ->child_reaper switching
Now that we rely on SIGNAL_UNKILLABLE flag, de_thread() doesn't need the nasty
hack to kill the old ->child_reaper during the mt-exec.
This also means we can avoid taking tasklist_lock around zap_other_threads().
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:03 +0000 (00:53 -0700)]
signals: fix /sbin/init protection from unwanted signals
The global init has a lot of long standing problems with the unhandled fatal
signals.
- The "is_global_init(current)" check in get_signal_to_deliver()
protects only the main thread. Sub-thread can dequee the fatal
signal and shutdown the whole thread group except the main thread.
If it dequeues SIGSTOP /sbin/init will be stopped, this is not
right too. Note that we can't use is_global_init(->group_leader),
this breaks exec and this can't solve other problems we have.
- Even if afterwards ignored, the fatal signals sets SIGNAL_GROUP_EXIT
on delivery. This breaks exec, has other bad implications, and this
is just wrong.
Introduce the new SIGNAL_UNKILLABLE flag to fix these problems. It also helps
to solve some other problems addressed by the subsequent patches.
Currently we use this flag for the global init only, but it could also be used
by kthreads and (perhaps) by the sub-namespace inits.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:02 +0000 (00:53 -0700)]
signals: check_kill_permission: remove tasklist_lock
Now that task_session() can't return a false NULL, check_kill_permission()
doesn't need tasklist_lock.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:01 +0000 (00:53 -0700)]
signals: check_kill_permission: check session under tasklist_lock
This wasn't documented, but as Atsushi Tsuji pointed out
check_kill_permission() needs tasklist_lock for task_session_nr(). I missed
this fact when removed tasklist from the callers.
Change check_kill_permission() to take tasklist_lock for the SIGCONT case.
Re-order security checks so that we take tasklist_lock only if/when it is
actually needed. This is a minimal fix for now, tasklist will be removed
later.
Also change the code to use task_session() instead of task_session_nr().
Also, remove the SIGCONT check from cap_task_kill(), it is bogus (and the
whole function is bogus. Serge, Eric, why it is still alive?).
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Atsushi Tsuji <a-tsuji@bk.jp.nec.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:00 +0000 (00:53 -0700)]
signals: send_signal: be paranoid about signalfd_notify()
send_signal() shouldn't call signalfd_notify() if it then fails with -EAGAIN.
Harmless, just a paranoid cleanup.
Also remove the comment. It is obsolete, signalfd_notify() was simplified and
does a simple wakeup.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Davide Libenzi <davidel@xmailserver.org>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:53:00 +0000 (00:53 -0700)]
signals: document CLD_CONTINUED notification mechanics
A couple of small comments about how CLD_CONTINUED notification works.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:59 +0000 (00:52 -0700)]
signals: fold sig_ignored() into handle_stop_signal()
Rename handle_stop_signal() to prepare_signal(), make it return a boolean, and
move the callsites of sig_ignored() into it.
No functional changes for now. But it would be nice to factor out the "should
we drop this signal" checks as much as possible, before we try to fix the bugs
with the sub-namespace init's signals (actually the global /sbin/init has some
problems with signals too).
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:58 +0000 (00:52 -0700)]
signals: cleanup the usage of print_fatal_signal()
Move the callsite of print_fatal_signal() down, under "if
(sig_kernel_coredump(signr))", so we don't need to check signr != SIGKILL.
We are only interested in the sig_kernel_coredump() signals anyway, and due to
the previous changes we almost never can see other fatal signals here except
SIGKILL.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:58 +0000 (00:52 -0700)]
signals: handle_stop_signal: don't worry about SIGKILL
handle_stop_signal() clears SIGNAL_STOP_DEQUEUED when sig == SIGKILL. Remove
this nasty special case. It was needed to prevent the race with group stop
and exit caused by thread-specific SIGKILL. Now that we use complete_signal()
for private signals too this is not needed, complete_signal() will notice
SIGKILL and abort the soon-to-begin group stop.
Except: the target thread is dead (has PF_EXITING). But in that case we
should not just clear SIGNAL_STOP_DEQUEUED and nothing more. We should either
kill the whole thread group, or silently ignore the signal.
I suspect we are not right wrt zombie leaders, but this is another issue which
and should be fixed separately. Note that this check can't abort the group
stop if it was already started/finished, this check only adds a subtle side
effect if we race with the thread which has already dequeued sig_kernel_stop()
signal and temporary released ->siglock.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:57 +0000 (00:52 -0700)]
signals: join send_sigqueue() with send_group_sigqueue()
We export send_sigqueue() and send_group_sigqueue() for the only user,
posix_timer_event(). This is a bit silly, because both are just trivial
helpers on top of do_send_sigqueue() and because the we pass the unused
.si_signo parameter.
Kill them both, rename do_send_sigqueue() to send_sigqueue(), and export it.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:56 +0000 (00:52 -0700)]
signals: unify send_sigqueue/send_group_sigqueue completely
Suggested by Pavel Emelyanov.
send_sigqueue/send_group_sigqueue are only differ in how they lock ->siglock.
Unify them. send_group_sigqueue() uses spin_lock() because it knows the task
can't exit, but in that case lock_task_sighand() can't fail and doesn't hurt.
Note that the "sig" argument is ignored, it is always equal to ->si_signo.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pavel Emelyanov [Wed, 30 Apr 2008 07:52:55 +0000 (00:52 -0700)]
signals: fold complete_signal() into send_signal/do_send_sigqueue
Factor out complete_signal() callsites. This change completely unifies the
helpers sending the specific/group signals.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:55 +0000 (00:52 -0700)]
signals: use __group_complete_signal() for the specific signals too
Based on Pavel Emelyanov's suggestion.
Rename __group_complete_signal() to complete_signal() and use it to process
the specific signals too. To do this we simply add the "int group" argument.
This allows us to greatly simply the signal-sending code and adds a useful
behaviour change. We can avoid the unneeded wakeups for the private signals
because wants_signal() is more clever than sigismember(blocked), but more
importantly we now take into account the fatal specific signals too.
The latter allows us to kill some subtle checks in handle_stop_signal() and
makes the specific/group signal's behaviour more consistent. For example,
currently sigtimedwait(FATAL_SIGNAL) behaves differently depending on was the
signal sent by kill() or tkill() if the signal was not blocked.
And. This allows us to tweak/fix the behaviour when the specific signal is
sent to the dying/dead ->group_leader.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:54 +0000 (00:52 -0700)]
signals: change send_signal/do_send_sigqueue to take "boolean group" parameter
send_signal() is used either with ->pending or with ->signal->shared_pending.
Change it to take "int group" instead, this argument will be re-used later.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:53 +0000 (00:52 -0700)]
signals: move the definition of __group_complete_signal() up
Move the unchanged definition of __group_complete_signal() so that send_signal
can see it. To simplify the reading of the next patches.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:52 +0000 (00:52 -0700)]
signals: microoptimize the usage of ->curr_target
Suggested by Roland McGrath.
Initialize signal->curr_target in copy_signal(). This way ->curr_target is
never == NULL, we can kill the check in __group_complete_signal's hot path.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:51 +0000 (00:52 -0700)]
signals: send_sig_info: don't take tasklist_lock
The comment in send_sig_info() is wrong, tasklist_lock can't help.
The caller must ensure the task can't go away, otherwise ->sighand can be NULL
even before we take the lock.
p->sighand could be changed by exec(), but I can't imagine how it is possible
to prevent exit(), but not exec().
Since the things seem to work, I assume all callers are correct. However,
drm_vbl_send_signals() looks broken. block_all_signals() which is solely used
by drm is definitely broken.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:51 +0000 (00:52 -0700)]
signals: do_tkill: don't use tasklist_lock
Convert do_tkill() to use rcu_read_lock() + lock_task_sighand() to avoid
taking tasklist lock.
Note that we don't return an error if lock_task_sighand() fails, we pretend
the task dies after receiving the signal. Otherwise, we should fight with the
nasty races with mt-exec without having any advantage.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:50 +0000 (00:52 -0700)]
signals: move handle_stop_signal() into send_signal()
Move handle_stop_signal() into send_signal(). This factors out a couple of
callsites and allows us to do further unifications.
Also, with this change specific_send_sig_info() does handle_stop_signal().
Not that this is really important, we never send STOP/CONT via send_sig() and
friends, but still this looks more consistent.
The only (afaics) special case is get_signal_to_deliver(). If the traced task
dequeues SIGCONT, it can re-send it to itself after ptrace_stop() if the
signal was blocked by debugger. In that case handle_stop_signal() is
unnecessary, but hopefully not a problem.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:49 +0000 (00:52 -0700)]
signals: send_group_sigqueue: don't take tasklist_lock
handle_stop_signal() was changed, now send_group_sigqueue() doesn't need
tasklist_lock.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:49 +0000 (00:52 -0700)]
signals: __group_complete_signal: cache the value of p->signal
Cosmetic, cache p->signal to make the code a bit more readable.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:48 +0000 (00:52 -0700)]
signals: send_sigqueue: don't forget about handle_stop_signal()
send_group_sigqueue() calls handle_stop_signal(), send_sigqueue() doesn't.
This is not consistent and in fact I'd say this is (minor) bug.
Move handle_stop_signal() from send_group_sigqueue() to do_send_sigqueue(),
the latter is called by send_sigqueue() too.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:48 +0000 (00:52 -0700)]
signals: send_sigqueue: don't take rcu lock
lock_task_sighand() was changed, send_sigqueue() doesn't need rcu_read_lock()
any longer.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:47 +0000 (00:52 -0700)]
get_signal_to_deliver: use the cached ->signal/sighand values
Cache the values of current->signal/sighand. Shrinks .text a bit and makes
the code more readable. Also, remove "sigset_t *mask", it is pointless
because in fact we save the constant offset.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:46 +0000 (00:52 -0700)]
handle_stop_signal: use the cached p->signal value
Cache the value of p->signal, and change the code to use while_each_thread()
helper.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:46 +0000 (00:52 -0700)]
handle_stop_signal: unify partial/full stop handling
Now that handle_stop_signal() doesn't drop ->siglock, we can't see both
->group_stop_count && SIGNAL_STOP_STOPPED. Merge two "if" branches.
As Roland pointed out, we never actually needed 2 do_notify_parent_cldstop()
calls.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:45 +0000 (00:52 -0700)]
kill_pid_info: don't take now unneeded tasklist_lock
Previously handle_stop_signal(SIGCONT) could drop ->siglock. That is why
kill_pid_info(SIGCONT) takes tasklist_lock to make sure the target task can't
go away after unlock. Not needed now.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:44 +0000 (00:52 -0700)]
signals: re-assign CLD_CONTINUED notification from the sender to reciever
Based on discussion with Jiri and Roland.
In short: currently handle_stop_signal(SIGCONT, p) sends the notification to
p->parent, with this patch p itself notifies its parent when it becomes
running.
handle_stop_signal(SIGCONT) has to drop ->siglock temporary in order to notify
the parent with do_notify_parent_cldstop(). This leads to multiple problems:
- as Jiri Kosina pointed out, the stopped task can resume without
actually seeing SIGCONT which may have a handler.
- we race with another sig_kernel_stop() signal which may come in
that window.
- we race with sig_fatal() signals which may set SIGNAL_GROUP_EXIT
in that window.
- we can't avoid taking tasklist_lock() while sending SIGCONT.
With this patch handle_stop_signal() just sets the new SIGNAL_CLD_CONTINUED
flag in p->signal->flags and returns. The notification is sent by the first
task which returns from finish_stop() (there should be at least one) or any
other signalled thread from get_signal_to_deliver().
This is a user-visible change. Say, currently kill(SIGCONT, stopped_child)
can't return without seeing SIGCHLD, with this patch SIGCHLD can be delayed
unpredictably. Another difference is that if the child is ptraced by another
process, CLD_CONTINUED may be delivered to ->real_parent after ptrace_detach()
while currently it always goes to the tracer which doesn't actually need this
notification. Hopefully not a problem.
The patch asks for the futher obvious cleanups, I'll send them separately.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Oleg Nesterov [Wed, 30 Apr 2008 07:52:42 +0000 (00:52 -0700)]
signals: cleanup security_task_kill() usage/implementation
Every implementation of ->task_kill() does nothing when the signal comes from
the kernel. This is correct, but means that check_kill_permission() should
call security_task_kill() only for SI_FROMUSER() case, and we can remove the
same check from ->task_kill() implementations.
(sadly, check_kill_permission() is the last user of signal->session/__session
but we can't s/task_session_nr/task_session/ here).
NOTE: Eric W. Biederman pointed out cap_task_kill() should die, and I think
he is very right.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Serge Hallyn <serue@us.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>
Cc: David Quigley <dpquigl@tycho.nsa.gov>
Cc: Eric Paris <eparis@redhat.com>
Cc: Harald Welte <laforge@gnumonks.org>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Stephen Smalley <sds@tycho.nsa.gov>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pavel Emelyanov [Wed, 30 Apr 2008 07:52:41 +0000 (00:52 -0700)]
signals: consolidate send_sigqueue and send_group_sigqueue
Both functions do the same thing after proper locking, but with
different sigpending structs, so move the common code into a helper.
After this we have 4 places that look very similar: send_sigqueue: calls
do_send_sigqueue and signal_wakeup send_group_sigqueue: calls
do_send_sigqueue and __group_complete_signal __group_send_sig_info:
calls send_signal and __group_complete_signal specific_send_sig_info:
calls send_signal and signal_wakeup
Besides, send_signal performs actions similar to do_send_sigqueue's
and __group_complete_signal - to signal_wakeup.
It looks like they can be consolidated gracefully.
Oleg said:
Personally, I think this change is very good. But send_sigqueue() and
send_group_sigqueue() have a very subtle difference which I was never able
to understand.
Let's suppose that sigqueue is already queued, and the signal is ignored
(the latter means we should re-schedule cpu timer or handle overrruns). In
that case send_sigqueue() returns 0, but send_group_sigqueue() returns 1.
I think this is not the problem (in fact, I think this patch makes the
behaviour more correct), but I hope Thomas can take a look and confirm.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>