Add support for MOXA:0x1120 pci device. It's a 2-port device and differs
in no way from the others. So this turns out to be a trivial
pci_device_id change.
Stanse found a tty refcnt leak in read_int_callback. In fact
it's handled wrong altogether. tty_port_tty_get can return NULL
and it's not checked in that manner.
Fix that by checking the tty_port_tty_get retval and put tty kref
properly.
Jiri Slaby [Tue, 11 Aug 2009 21:20:41 +0000 (23:20 +0200)]
tty: Power: fix suspend vt regression
vt_waitactive no longer accepts console parameter as console-1
since commit "vt: add an event interface". It expects console
number directly (as viewed by userspace -- counting from 1).
Fix a deadlock suspend regression by redefining adding one
to vt in vt_move_to_console.
Arnd Bergmann [Thu, 6 Aug 2009 13:09:28 +0000 (15:09 +0200)]
tty: handle VT specific compat ioctls in vt driver
The VT specific compat_ioctl handlers are the only ones
in common code that require the BKL. Moving them into
the vt driver lets us remove the BKL from the other handlers
and cleans up the code.
Joe Peterson [Wed, 9 Sep 2009 21:03:47 +0000 (15:03 -0600)]
n_tty: move echoctl check and clean up logic
Check L_ECHOCTL before insertting a character in the echo buffer
(rather than as the buffer is processed), to be more consistent with
when all other L_ flags are checked. Also cleaned up the related logic.
Note that this and the previous patch ("n_tty: honor opost flag for echoes")
were verified together by the reporters of the bug that patch addresses
(http://bugs.linuxbase.org/show_bug.cgi?id=2692), and the test now passes.
Joe Peterson [Wed, 9 Sep 2009 21:03:13 +0000 (15:03 -0600)]
n_tty: honor opost flag for echoes
Fixes the following bug:
http://bugs.linuxbase.org/show_bug.cgi?id=2692
Causes processing of echoed characters (output from the echo buffer) to
honor the O_OPOST flag, which is consistent with the old behavior.
Note that this and the next patch ("n_tty: move echoctl check and
clean up logic") were verified together by the bug reporters, and
the test now passes.
Alan Cox [Sat, 19 Sep 2009 20:13:33 +0000 (13:13 -0700)]
tty: USB serial termios bits
Various drivers have hacks to mangle termios structures. This stems from
the fact there is no nice setup hook for configuring the termios settings
when the port is created
Set proper console speed on resume if console suspend is disabled
Commit b5b82df6, from May 2007, breaks no_console_suspend on the OLPC
XO laptop. Basically what happens is that upon returning from resume,
serial8250_resume_port() will reconfigure the port for high speed
mode and all console output will be garbled, making debug of the
resume path painful. This patch modifies uart_resume_port() to
reset the port to the state it was in before we suspended.
Original patch by Marcelo Tosatti
Second patch by Deepak then reworked by Alan to fit with the tty changes
before it got submitted. Also fixed the console path to set c_i/ospeed as
some drivers require the termios fields are valid
Alan Cox [Sat, 19 Sep 2009 20:13:32 +0000 (13:13 -0700)]
serial: kill USF_CLOSING_* definitions
The serial layer for some reason uses different defines for the special
case close delays and then conditionally switches to/from the normal ones
in the ioctls.
Alan Cox [Sat, 19 Sep 2009 20:13:30 +0000 (13:13 -0700)]
serial: move the flags into the tty_port field
Fortunately the serial layer was designed to use the same flag values but
with different names. It has its own SUSPENDED flag which is a free slot in
the ASYNC flags so we allocate it in the ASYNC flags instead.
Alan Cox [Sat, 19 Sep 2009 20:13:29 +0000 (13:13 -0700)]
serial: use tty_port pointers in the core code
Extract out a lot of the x.port. uses and also show up where there are
things left to be isolated that prevent use using the port helpers in the
serial layer at this point
Alan Cox [Sat, 19 Sep 2009 20:13:25 +0000 (13:13 -0700)]
vt: remove power stuff from kernel/power
In the past someone gratuitiously borrowed chunks of kernel internal vt
code and dumped them in kernel/power. They have all sorts of deep relations
with the vt code so put them in the vt tree instead
Alan Cox [Sat, 19 Sep 2009 20:13:24 +0000 (13:13 -0700)]
vt: add an event interface
This is needed and requested in various forms for ConsoleKit, screenblank
handling and the like so do the job with a single interface. Also build the
interface so that unlike VT_WAITACTIVE and friends it won't miss events.
FIXME: Should this be a waitactive ioctl or a new device file you can poll
and read events from. We need the code anyway to fix up the existing broken
wait for console switch logic but the ConsoleKit people would prefer the
new device to the ioctl we have here
Alan Cox [Sat, 19 Sep 2009 20:13:23 +0000 (13:13 -0700)]
tty: USB hangup is racy
The USB layer uses tty_hangup to deal with unplugs of the physical hardware
(analogous to loss of carrier) and then frees the resources. However the
tty_hangup is asynchronous. As the hangup can sleep we can use tty_vhangup
which is the non async version to avoid freeing resources too early.
Alan Cox [Sat, 19 Sep 2009 20:13:22 +0000 (13:13 -0700)]
tty: make the kref destructor occur asynchronously
We want to be able to sleep in the destructor for USB at least. It isn't a
hot path so just pushing it to a work queue doesn't really cause any
difficulty.
Alan Cox [Sat, 19 Sep 2009 20:13:20 +0000 (13:13 -0700)]
tty: Add a full port_close function
Now we are extracting out methods for shutdown and the like we can add a
proper tty_port_close method that knows all the innards of the tty closing
process and hides the lot from the caller.
At some point in the future this will be paired with a similar open()
helper and the drivers can stick to hardware management.
Anton Vorontsov [Sat, 19 Sep 2009 20:13:20 +0000 (13:13 -0700)]
8250: Now honours baud rate lower bounds
A platform clock drives 8250 ports in most SOC systems, the clock
might run at high frequencies, and so it's not always possible to
downscale uart clock to a desired value.
Currently the 8250 uart driver accepts not supported baud rates, and
what is worse, it is doing this silently, and then passes not accepted
values to a new termios, so userspace has no chance to catch this kind
of errors (userspace verifies that settings were accepted by reading
back and comparing the settings).
This patch fixes the issue by passing minimum baud rate to the
uart_get_baud_rate() call, the call should take care of all bounds,
so userspace should now report:
# stty -F /dev/ttyS0 speed 300
115200
stty: /dev/ttyS0: unable to perform all requested operations
p.s. uart_get_baud_rate() falls back to 9600, which still might be too
low for some 10 GHz platforms, but that's a separate issue, and
we can wait with fixing this till we find such a platform.
Alan Cox [Sat, 19 Sep 2009 20:13:17 +0000 (13:13 -0700)]
kfifo: Use "const" definitions
Currently kfifo cannot be used by parts of the kernel that use "const"
properly as kfifo itself does not use const for passed data blocks which
are indeed const.
Alan Cox [Sat, 19 Sep 2009 20:13:17 +0000 (13:13 -0700)]
slip: Clean up create and destroy
The network layer now has a destructor we can hook to clean up the slip
devices array. That needs us to initiate unregister events in the right
places which with the current tty layer we can do, and with network
refcounting is safe to do.
Add helpers for io operations, so that we can eliminate huge
amount of supporting code. It is now centralized in those
helpers and used values are precomputed in the init phase.
- save one indent level by inverting !fw_loaded condition
- read rs_status on Z and write it after we change all the flags,
don't do that separately
- remove Y inverted rts/dtr branching, precompute registers and use
them
- add a cy_ prefix to functions with changed prototypes
- cy_get_serial_info: initialize serial_struct by initializer,
save a memset
- inline simple functions (get_mon_info, {s,g}et_default_threshold,
{s,g}et_default_timeout) directly in the ioctl handler
- add a cy_cflags_changed helper to not copy its code by
wait_event_interruptible
- remove some ret_val = 0 assignments, it's preset to 0
- TIOCGICOUNT: don't do many put_user's, do one copy_to_user
- remove changelog from the file. we don't care about ancient
history
- update copyright year
- update version
- constify some stuff
- empty lines removal
- unused variables and macros removal
- remove some asm/ includes, they are sucked by linux/ variants
Use new tty helpers for close, which allows much code removal.
The only real change is locking. card_lock for protecting was
used inappropriately (just to have a critical section, no matter
which lock is used), so the change to port->lock is fine.
Remove also useless debug printks while being there.
Do not duplicate common tty_port_hangup code. Use it instead.
Also do not unset ASYNC_NORMAL_ACTIVE and wake up from the
tty_hangup() caller. It makes no sense since we don't check that
flag in sleepers. tty_port_hangup() performed later will do the
right job.
While this is not problem for Y card handlers (they are protected
by card_lock), Z handlers and other functions may dereference NULL
at any point after hangup/close. Even if (tty == NULL) was already
performed in the handler.
Note that it's not an issue for Y cards just for now. After
switching to tty_port_close_* et al. this will be a problem. So
add refcounting to them all.
Also proc .show doesn't take a tty reference and it should (along
with a ldisc one).
While at it and changing prototypes (adding tty param), prepend
cy_ to functions which don't have it yet.
Linus Torvalds [Mon, 3 Aug 2009 23:01:28 +0000 (16:01 -0700)]
tty-ldisc: get rid of tty_ldisc_try_get() helper function
Now that the /proc/tty/ldiscs handling doesn't play games with 'struct
ldisc' any more, the only remaining user of 'tty_ldisc_try_get()' is
'tty_ldisc_get()' (note the lack of 'try').
And we're actually much better off folding the logic directly into that
file, since the 'try' part was always about trying to get the ldisc
operations, not the ldisc itself: and making that explicit inside of
'tty_ldisc_get()' clarifies the whole semantics.
Linus Torvalds [Mon, 3 Aug 2009 23:00:15 +0000 (16:00 -0700)]
tty-ldisc: make /proc/tty/ldiscs use ldisc_ops instead of ldiscs
The /proc/tty/ldiscs file is totally and utterly un-interested in the
"struct tty_ldisc" structures, and only cares about the underlying ldisc
operations.
So don't make it create a dummy 'struct ldisc' only to get a pointer to
the operations, and then destroy it. Instead, we split up the function
'tty_ldisc_try_get()', and create a 'get_ldops()' helper that just looks
up the ldisc operations based on the ldisc number.
That makes the code simpler to read (smaller and more well-defined
helper functions), and allows the /proc functions to avoid creating that
useless dummy only to immediately free it again.
Mike Frysinger [Thu, 6 Aug 2009 22:20:05 +0000 (15:20 -0700)]
serial: bfin_5xx: fix building as module when early printk is enabled
Since early printk only makes sense/works when the serial driver is built
into the kernel, disable the option for this driver when it is going to be
built as a module. Otherwise we get build failures due to the ifdef
handling.
Tilman Schmidt [Fri, 7 Aug 2009 03:33:33 +0000 (20:33 -0700)]
tty: gigaset: really fix chars_in_buffer
The tty_operation chars_in_buffer() is not allowed to return a negative
value to signal an error. Corrects the problem flagged by commit 23198fda7182969b619613a555f8645fdc3dc334, "tty: fix chars_in_buffers".
Kay Sievers [Fri, 18 Sep 2009 21:01:12 +0000 (23:01 +0200)]
Driver-Core: extend devnode callbacks to provide permissions
This allows subsytems to provide devtmpfs with non-default permissions
for the device node. Instead of the default mode of 0600, null, zero,
random, urandom, full, tty, ptmx now have a mode of 0666, which allows
non-privileged processes to access standard device nodes in case no
other userspace process applies the expected permissions.
This also fixes a wrong assignment in pktcdvd and a checkpatch.pl complain.
1. some cpu feature in early stage too, like cpu_has_x2apic
2. will break built-in-command line
3. will break other memmap= and mem=
4. early_dbgp and early_console that will use early_ioremap to access mmio (?)
Mike Galbraith [Sat, 19 Sep 2009 14:52:35 +0000 (16:52 +0200)]
sched: Re-add lost cpu_allowed check to sched_fair.c::select_task_rq_fair()
While doing some testing, I pinned mplayer, only to find it
following X around like a puppy. Looking at commit c88d591, I found
a cpu_allowed check that went AWOL. I plugged it back in where it
looks like it needs to go, and now when I say "sit, stay!", mplayer
obeys again.
'c88d591 sched: Merge select_task_rq_fair() and
sched_balance_self()' accidentally dropped the check, causing
wake_affine() to pull pinned tasks - put it back.
Arjan van de Ven [Sat, 12 Sep 2009 05:53:05 +0000 (07:53 +0200)]
perf: Add the timechart tool
timechart is a tool to visualize what is going on in the system.
The user makes a trace of what is going on with
> perf record --timechart /usr/bin/some_command
and then can turn the output of this into an svg file
> perf timechart
which then can be viewed with any SVG view; inkscape works well
enough for me.
The idea behind timechart is to create a "infinitely zoomable"
picture; something that has high level information on a 1:1 zoom
level, but which exposes more details every time you zoom into a
specific area.
Arjan van de Ven [Sat, 12 Sep 2009 05:52:51 +0000 (07:52 +0200)]
perf: Store trace event name/id pairs in perf.data
The trace event name<->id mapping is dynamic for each kernel
compile. In order for perf.data to be useable outside the actual
system, we thus need to store a table of this mapping for later
use.
This patch adds this table to perf.data, and provides helper
functions for lookup up fields from this table.
To avoid mistakes, lookup-from-table is kept completely seprate
from lookup-from-local-debugfs.
x86, apic: Use logical flat on intel with <= 8 logical cpus
On Intel platforms, we can use logical flat mode if there are <= 8
logical cpu's (irrespective of physical apic id values). This will
enable simplified and efficient IPI and device interrupt routing on
such platforms.
Fix the relevant comments while we are at it.
We can clean up default_setup_apic_routing() by using apic->probe()
but that is a different item.
Paul E. McKenney [Fri, 18 Sep 2009 16:50:19 +0000 (09:50 -0700)]
rcu: Fix thinko, actually initialize full tree
Commit de078d8 ("rcu: Need to update rnp->gpnum if preemptable RCU
is to be reliable") repeatedly and incorrectly initializes the root
rcu_node structure's ->gpnum field rather than initializing the
->gpnum field of each node in the tree. Fix this. Also add an
additional consistency check to catch this in the future.
Paul E. McKenney [Fri, 18 Sep 2009 16:50:18 +0000 (09:50 -0700)]
rcu: Apply results of code inspection of kernel/rcutree_plugin.h
o Drop the calls to cpu_quiet() from the online/offline code.
These are unnecessary, since force_quiescent_state() will
clean up, and removing them simplifies the code a bit.
o Add a warning to check that we don't enqueue the same blocked
task twice onto the ->blocked_tasks[] lists.
o Rework the phase computation in rcu_preempt_note_context_switch()
to be more readable, as suggested by Josh Triplett.
o Disable irqs to close a race between the scheduling clock
interrupt and rcu_preempt_note_context_switch() WRT the
->rcu_read_unlock_special field.
o Add comments to rnp->lock acquisition and release within
rcu_read_unlock_special() noting that irqs are already
disabled.
projects / linux.git / log