License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 17:07:57 +03:00
// SPDX-License-Identifier: GPL-2.0
2005-04-17 02:20:36 +04:00
/*
* This file contains the procedures for the handling of select and poll
*
* Created for Linux based loosely upon Mathius Lattner ' s minix
* patches by Peter MacDonald . Heavily edited by Linus .
*
* 4 February 1994
* COFF / ELF binary emulation . If the process has the STICKY_TIMEOUTS
* flag set in its personality we do * not * modify the given timeout
* parameter to reflect time remaining .
*
* 24 January 2000
* Changed sys_poll ( ) / do_poll ( ) to use PAGE_SIZE chunk - based allocation
* of fds to overcome nfds < 16390 descriptors limit ( Tigran Aivazian ) .
*/
2007-05-08 11:29:02 +04:00
# include <linux/kernel.h>
2017-02-08 20:51:30 +03:00
# include <linux/sched/signal.h>
# include <linux/sched/rt.h>
2005-04-17 02:20:36 +04:00
# include <linux/syscalls.h>
2011-11-17 08:57:37 +04:00
# include <linux/export.h>
2005-04-17 02:20:36 +04:00
# include <linux/slab.h>
# include <linux/poll.h>
# include <linux/personality.h> /* for STICKY_TIMEOUTS */
# include <linux/file.h>
2008-04-24 15:44:08 +04:00
# include <linux/fdtable.h>
2005-04-17 02:20:36 +04:00
# include <linux/fs.h>
2005-09-10 00:04:14 +04:00
# include <linux/rcupdate.h>
2008-08-31 19:26:40 +04:00
# include <linux/hrtimer.h>
2013-05-07 03:50:17 +04:00
# include <linux/freezer.h>
2013-07-10 18:13:17 +04:00
# include <net/busy_poll.h>
2016-10-11 23:51:14 +03:00
# include <linux/vmalloc.h>
2005-04-17 02:20:36 +04:00
2016-12-24 22:46:01 +03:00
# include <linux/uaccess.h>
2005-04-17 02:20:36 +04:00
2008-09-02 02:55:35 +04:00
/*
* Estimate expected accuracy in ns from a timeval .
*
* After quite a bit of churning around , we ' ve settled on
* a simple thing of taking 0.1 % of the timeout as the
* slack , with a cap of 100 msec .
* " nice " tasks get a 0.5 % slack instead .
*
* Consider this comment an open invitation to come up with even
* better solutions . .
*/
2009-09-23 03:43:30 +04:00
# define MAX_SLACK (100 * NSEC_PER_MSEC)
2016-05-20 03:09:05 +03:00
static long __estimate_accuracy ( struct timespec64 * tv )
2008-09-02 02:55:35 +04:00
{
2008-09-08 03:08:55 +04:00
long slack ;
2008-09-02 02:55:35 +04:00
int divfactor = 1000 ;
2009-09-23 03:43:30 +04:00
if ( tv - > tv_sec < 0 )
return 0 ;
2008-09-08 02:31:39 +04:00
if ( task_nice ( current ) > 0 )
2008-09-02 02:55:35 +04:00
divfactor = divfactor / 5 ;
2009-09-23 03:43:30 +04:00
if ( tv - > tv_sec > MAX_SLACK / ( NSEC_PER_SEC / divfactor ) )
return MAX_SLACK ;
2008-09-02 02:55:35 +04:00
slack = tv - > tv_nsec / divfactor ;
slack + = tv - > tv_sec * ( NSEC_PER_SEC / divfactor ) ;
2009-09-23 03:43:30 +04:00
if ( slack > MAX_SLACK )
return MAX_SLACK ;
2008-09-08 03:08:55 +04:00
2008-09-02 02:55:35 +04:00
return slack ;
}
2016-05-20 03:09:05 +03:00
u64 select_estimate_accuracy ( struct timespec64 * tv )
2008-09-02 02:55:35 +04:00
{
timer: convert timer_slack_ns from unsigned long to u64
This patchset introduces a /proc/<pid>/timerslack_ns interface which
would allow controlling processes to be able to set the timerslack value
on other processes in order to save power by avoiding wakeups (Something
Android currently does via out-of-tree patches).
The first patch tries to fix the internal timer_slack_ns usage which was
defined as a long, which limits the slack range to ~4 seconds on 32bit
systems. It converts it to a u64, which provides the same basically
unlimited slack (500 years) on both 32bit and 64bit machines.
The second patch introduces the /proc/<pid>/timerslack_ns interface
which allows the full 64bit slack range for a task to be read or set on
both 32bit and 64bit machines.
With these two patches, on a 32bit machine, after setting the slack on
bash to 10 seconds:
$ time sleep 1
real 0m10.747s
user 0m0.001s
sys 0m0.005s
The first patch is a little ugly, since I had to chase the slack delta
arguments through a number of functions converting them to u64s. Let me
know if it makes sense to break that up more or not.
Other than that things are fairly straightforward.
This patch (of 2):
The timer_slack_ns value in the task struct is currently a unsigned
long. This means that on 32bit applications, the maximum slack is just
over 4 seconds. However, on 64bit machines, its much much larger (~500
years).
This disparity could make application development a little (as well as
the default_slack) to a u64. This means both 32bit and 64bit systems
have the same effective internal slack range.
Now the existing ABI via PR_GET_TIMERSLACK and PR_SET_TIMERSLACK specify
the interface as a unsigned long, so we preserve that limitation on
32bit systems, where SET_TIMERSLACK can only set the slack to a unsigned
long value, and GET_TIMERSLACK will return ULONG_MAX if the slack is
actually larger then what can be stored by an unsigned long.
This patch also modifies hrtimer functions which specified the slack
delta as a unsigned long.
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Oren Laadan <orenl@cellrox.com>
Cc: Ruchi Kandoi <kandoiruchi@google.com>
Cc: Rom Lemarchand <romlem@android.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Android Kernel Team <kernel-team@android.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-18 00:20:51 +03:00
u64 ret ;
2016-05-20 03:09:05 +03:00
struct timespec64 now ;
2008-09-02 02:55:35 +04:00
/*
* Realtime tasks get a slack of 0 for obvious reasons .
*/
2008-09-08 02:31:39 +04:00
if ( rt_task ( current ) )
2008-09-02 02:55:35 +04:00
return 0 ;
2016-05-20 03:09:05 +03:00
ktime_get_ts64 ( & now ) ;
now = timespec64_sub ( * tv , now ) ;
2008-09-02 02:55:35 +04:00
ret = __estimate_accuracy ( & now ) ;
if ( ret < current - > timer_slack_ns )
return current - > timer_slack_ns ;
return ret ;
}
2005-04-17 02:20:36 +04:00
struct poll_table_page {
struct poll_table_page * next ;
struct poll_table_entry * entry ;
struct poll_table_entry entries [ 0 ] ;
} ;
# define POLL_TABLE_FULL(table) \
( ( unsigned long ) ( ( table ) - > entry + 1 ) > PAGE_SIZE + ( unsigned long ) ( table ) )
/*
* Ok , Peter made a complicated , but straightforward multiple_wait ( ) function .
* I have rewritten this , taking some shortcuts : This code may not be easy to
* follow , but it should be free of race - conditions , and it ' s practical . If you
* understand what I ' m doing here , then you understand how the linux
* sleep / wakeup mechanism works .
*
* Two very simple procedures , poll_wait ( ) and poll_freewait ( ) make all the
* work . poll_wait ( ) is an inline - function defined in < linux / poll . h > ,
* as all select / poll functions have to call it to add an entry to the
* poll table .
*/
2005-05-06 03:16:09 +04:00
static void __pollwait ( struct file * filp , wait_queue_head_t * wait_address ,
poll_table * p ) ;
2005-04-17 02:20:36 +04:00
void poll_initwait ( struct poll_wqueues * pwq )
{
init_poll_funcptr ( & pwq - > pt , __pollwait ) ;
2009-01-07 01:40:59 +03:00
pwq - > polling_task = current ;
2009-08-15 21:30:24 +04:00
pwq - > triggered = 0 ;
2005-04-17 02:20:36 +04:00
pwq - > error = 0 ;
pwq - > table = NULL ;
2006-03-28 13:56:33 +04:00
pwq - > inline_index = 0 ;
2005-04-17 02:20:36 +04:00
}
EXPORT_SYMBOL ( poll_initwait ) ;
2006-03-28 13:56:33 +04:00
static void free_poll_entry ( struct poll_table_entry * entry )
{
2007-05-09 09:10:02 +04:00
remove_wait_queue ( entry - > wait_address , & entry - > wait ) ;
2006-03-28 13:56:33 +04:00
fput ( entry - > filp ) ;
}
2005-04-17 02:20:36 +04:00
void poll_freewait ( struct poll_wqueues * pwq )
{
struct poll_table_page * p = pwq - > table ;
2006-03-28 13:56:33 +04:00
int i ;
for ( i = 0 ; i < pwq - > inline_index ; i + + )
free_poll_entry ( pwq - > inline_entries + i ) ;
2005-04-17 02:20:36 +04:00
while ( p ) {
struct poll_table_entry * entry ;
struct poll_table_page * old ;
entry = p - > entry ;
do {
entry - - ;
2006-03-28 13:56:33 +04:00
free_poll_entry ( entry ) ;
2005-04-17 02:20:36 +04:00
} while ( entry > p - > entries ) ;
old = p ;
p = p - > next ;
free_page ( ( unsigned long ) old ) ;
}
}
EXPORT_SYMBOL ( poll_freewait ) ;
2009-01-07 01:40:59 +03:00
static struct poll_table_entry * poll_get_entry ( struct poll_wqueues * p )
2005-04-17 02:20:36 +04:00
{
struct poll_table_page * table = p - > table ;
2006-03-28 13:56:33 +04:00
if ( p - > inline_index < N_INLINE_POLL_ENTRIES )
return p - > inline_entries + p - > inline_index + + ;
2005-04-17 02:20:36 +04:00
if ( ! table | | POLL_TABLE_FULL ( table ) ) {
struct poll_table_page * new_table ;
new_table = ( struct poll_table_page * ) __get_free_page ( GFP_KERNEL ) ;
if ( ! new_table ) {
p - > error = - ENOMEM ;
2006-03-28 13:56:33 +04:00
return NULL ;
2005-04-17 02:20:36 +04:00
}
new_table - > entry = new_table - > entries ;
new_table - > next = table ;
p - > table = new_table ;
table = new_table ;
}
2006-03-28 13:56:33 +04:00
return table - > entry + + ;
}
2017-06-20 13:06:13 +03:00
static int __pollwake ( wait_queue_entry_t * wait , unsigned mode , int sync , void * key )
2009-01-07 01:40:59 +03:00
{
struct poll_wqueues * pwq = wait - > private ;
DECLARE_WAITQUEUE ( dummy_wait , pwq - > polling_task ) ;
/*
* Although this function is called under waitqueue lock , LOCK
* doesn ' t imply write barrier and the users expect write
* barrier semantics on wakeup functions . The following
* smp_wmb ( ) is equivalent to smp_wmb ( ) in try_to_wake_up ( )
2015-05-12 11:51:55 +03:00
* and is paired with smp_store_mb ( ) in poll_schedule_timeout .
2009-01-07 01:40:59 +03:00
*/
smp_wmb ( ) ;
pwq - > triggered = 1 ;
/*
* Perform the default wake up operation using a dummy
* waitqueue .
*
* TODO : This is hacky but there currently is no interface to
* pass in @ sync . @ sync is scheduled to be removed and once
* that happens , wake_up_process ( ) can be used directly .
*/
return default_wake_function ( & dummy_wait , mode , sync , key ) ;
}
2017-06-20 13:06:13 +03:00
static int pollwake ( wait_queue_entry_t * wait , unsigned mode , int sync , void * key )
2009-06-17 02:33:36 +04:00
{
struct poll_table_entry * entry ;
entry = container_of ( wait , struct poll_table_entry , wait ) ;
if ( key & & ! ( ( unsigned long ) key & entry - > key ) )
return 0 ;
return __pollwake ( wait , mode , sync , key ) ;
}
2006-03-28 13:56:33 +04:00
/* Add a new entry */
static void __pollwait ( struct file * filp , wait_queue_head_t * wait_address ,
poll_table * p )
{
2009-01-07 01:40:59 +03:00
struct poll_wqueues * pwq = container_of ( p , struct poll_wqueues , pt ) ;
struct poll_table_entry * entry = poll_get_entry ( pwq ) ;
2006-03-28 13:56:33 +04:00
if ( ! entry )
return ;
2012-08-27 22:48:26 +04:00
entry - > filp = get_file ( filp ) ;
2006-03-28 13:56:33 +04:00
entry - > wait_address = wait_address ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
entry - > key = p - > _key ;
2009-01-07 01:40:59 +03:00
init_waitqueue_func_entry ( & entry - > wait , pollwake ) ;
entry - > wait . private = pwq ;
2007-05-09 09:10:02 +04:00
add_wait_queue ( wait_address , & entry - > wait ) ;
2005-04-17 02:20:36 +04:00
}
2009-01-07 01:40:59 +03:00
int poll_schedule_timeout ( struct poll_wqueues * pwq , int state ,
ktime_t * expires , unsigned long slack )
{
int rc = - EINTR ;
set_current_state ( state ) ;
if ( ! pwq - > triggered )
2013-10-30 02:43:08 +04:00
rc = schedule_hrtimeout_range ( expires , slack , HRTIMER_MODE_ABS ) ;
2009-01-07 01:40:59 +03:00
__set_current_state ( TASK_RUNNING ) ;
/*
* Prepare for the next iteration .
*
2015-05-12 11:51:55 +03:00
* The following smp_store_mb ( ) serves two purposes . First , it ' s
2009-01-07 01:40:59 +03:00
* the counterpart rmb of the wmb in pollwake ( ) such that data
* written before wake up is always visible after wake up .
* Second , the full barrier guarantees that triggered clearing
* doesn ' t pass event check of the next iteration . Note that
* this problem doesn ' t exist for the first iteration as
* add_wait_queue ( ) has full barrier semantics .
*/
2015-05-12 11:51:55 +03:00
smp_store_mb ( pwq - > triggered , 0 ) ;
2009-01-07 01:40:59 +03:00
return rc ;
}
EXPORT_SYMBOL ( poll_schedule_timeout ) ;
2008-08-31 19:16:57 +04:00
/**
* poll_select_set_timeout - helper function to setup the timeout value
2016-05-20 03:09:05 +03:00
* @ to : pointer to timespec64 variable for the final timeout
2008-08-31 19:16:57 +04:00
* @ sec : seconds ( from user space )
* @ nsec : nanoseconds ( from user space )
*
* Note , we do not use a timespec for the user space value here , That
* way we can use the function for timeval and compat interfaces as well .
*
* Returns - EINVAL if sec / nsec are not normalized . Otherwise 0.
*/
2016-05-20 03:09:05 +03:00
int poll_select_set_timeout ( struct timespec64 * to , time64_t sec , long nsec )
2008-08-31 19:16:57 +04:00
{
2016-05-20 03:09:05 +03:00
struct timespec64 ts = { . tv_sec = sec , . tv_nsec = nsec } ;
2008-08-31 19:16:57 +04:00
2016-05-20 03:09:05 +03:00
if ( ! timespec64_valid ( & ts ) )
2008-08-31 19:16:57 +04:00
return - EINVAL ;
/* Optimize for the zero timeout value here */
if ( ! sec & & ! nsec ) {
to - > tv_sec = to - > tv_nsec = 0 ;
} else {
2016-05-20 03:09:05 +03:00
ktime_get_ts64 ( to ) ;
* to = timespec64_add_safe ( * to , ts ) ;
2008-08-31 19:16:57 +04:00
}
return 0 ;
}
2016-05-20 03:09:05 +03:00
static int poll_select_copy_remaining ( struct timespec64 * end_time ,
void __user * p ,
2008-08-31 19:16:57 +04:00
int timeval , int ret )
{
2017-08-05 07:12:31 +03:00
struct timespec64 rts ;
2008-08-31 19:16:57 +04:00
struct timeval rtv ;
if ( ! p )
return ret ;
if ( current - > personality & STICKY_TIMEOUTS )
goto sticky ;
/* No update for zero timeout */
if ( ! end_time - > tv_sec & & ! end_time - > tv_nsec )
return ret ;
2017-08-05 07:12:31 +03:00
ktime_get_ts64 ( & rts ) ;
rts = timespec64_sub ( * end_time , rts ) ;
if ( rts . tv_sec < 0 )
rts . tv_sec = rts . tv_nsec = 0 ;
2016-05-20 03:09:05 +03:00
2008-08-31 19:16:57 +04:00
if ( timeval ) {
2011-01-13 04:00:00 +03:00
if ( sizeof ( rtv ) > sizeof ( rtv . tv_sec ) + sizeof ( rtv . tv_usec ) )
memset ( & rtv , 0 , sizeof ( rtv ) ) ;
2017-08-05 07:12:31 +03:00
rtv . tv_sec = rts . tv_sec ;
rtv . tv_usec = rts . tv_nsec / NSEC_PER_USEC ;
2008-08-31 19:16:57 +04:00
if ( ! copy_to_user ( p , & rtv , sizeof ( rtv ) ) )
return ret ;
2017-08-05 07:12:31 +03:00
} else if ( ! put_timespec64 ( & rts , p ) )
2008-08-31 19:16:57 +04:00
return ret ;
/*
* If an application puts its timeval in read - only memory , we
* don ' t want the Linux - specific update to the timeval to
* cause a fault after the select has completed
* successfully . However , because we ' re not updating the
* timeval , we can ' t restart the system call .
*/
sticky :
if ( ret = = - ERESTARTNOHAND )
ret = - EINTR ;
return ret ;
}
2017-04-08 23:50:24 +03:00
/*
* Scalable version of the fd_set .
*/
typedef struct {
unsigned long * in , * out , * ex ;
unsigned long * res_in , * res_out , * res_ex ;
} fd_set_bits ;
/*
* How many longwords for " nr " bits ?
*/
# define FDS_BITPERLONG (8*sizeof(long))
# define FDS_LONGS(nr) (((nr)+FDS_BITPERLONG-1) / FDS_BITPERLONG)
# define FDS_BYTES(nr) (FDS_LONGS(nr)*sizeof(long))
/*
* We do a VERIFY_WRITE here even though we are only reading this time :
* we ' ll write to it eventually . .
*
* Use " unsigned long " accesses to let user - mode fd_set ' s be long - aligned .
*/
static inline
int get_fd_set ( unsigned long nr , void __user * ufdset , unsigned long * fdset )
{
nr = FDS_BYTES ( nr ) ;
if ( ufdset )
return copy_from_user ( fdset , ufdset , nr ) ? - EFAULT : 0 ;
memset ( fdset , 0 , nr ) ;
return 0 ;
}
static inline unsigned long __must_check
set_fd_set ( unsigned long nr , void __user * ufdset , unsigned long * fdset )
{
if ( ufdset )
return __copy_to_user ( ufdset , fdset , FDS_BYTES ( nr ) ) ;
return 0 ;
}
static inline
void zero_fd_set ( unsigned long nr , unsigned long * fdset )
{
memset ( fdset , 0 , FDS_BYTES ( nr ) ) ;
}
2005-04-17 02:20:36 +04:00
# define FDS_IN(fds, n) (fds->in + n)
# define FDS_OUT(fds, n) (fds->out + n)
# define FDS_EX(fds, n) (fds->ex + n)
# define BITS(fds, n) (*FDS_IN(fds, n)|*FDS_OUT(fds, n)|*FDS_EX(fds, n))
static int max_select_fd ( unsigned long n , fd_set_bits * fds )
{
unsigned long * open_fds ;
unsigned long set ;
int max ;
2005-09-10 00:04:10 +04:00
struct fdtable * fdt ;
2005-04-17 02:20:36 +04:00
/* handle last in-complete long-word first */
2012-07-25 18:40:34 +04:00
set = ~ ( ~ 0UL < < ( n & ( BITS_PER_LONG - 1 ) ) ) ;
n / = BITS_PER_LONG ;
2005-09-10 00:04:10 +04:00
fdt = files_fdtable ( current - > files ) ;
2012-02-16 21:49:54 +04:00
open_fds = fdt - > open_fds + n ;
2005-04-17 02:20:36 +04:00
max = 0 ;
if ( set ) {
set & = BITS ( fds , n ) ;
if ( set ) {
if ( ! ( set & ~ * open_fds ) )
goto get_max ;
return - EBADF ;
}
}
while ( n ) {
open_fds - - ;
n - - ;
set = BITS ( fds , n ) ;
if ( ! set )
continue ;
if ( set & ~ * open_fds )
return - EBADF ;
if ( max )
continue ;
get_max :
do {
max + + ;
set > > = 1 ;
} while ( set ) ;
2012-07-25 18:40:34 +04:00
max + = n * BITS_PER_LONG ;
2005-04-17 02:20:36 +04:00
}
return max ;
}
# define POLLIN_SET (POLLRDNORM | POLLRDBAND | POLLIN | POLLHUP | POLLERR)
# define POLLOUT_SET (POLLWRBAND | POLLWRNORM | POLLOUT | POLLERR)
# define POLLEX_SET (POLLPRI)
2009-06-17 02:33:36 +04:00
static inline void wait_key_set ( poll_table * wait , unsigned long in ,
2013-06-24 11:28:03 +04:00
unsigned long out , unsigned long bit ,
unsigned int ll_flag )
2009-06-17 02:33:36 +04:00
{
2013-06-24 11:28:03 +04:00
wait - > _key = POLLEX_SET | ll_flag ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
if ( in & bit )
wait - > _key | = POLLIN_SET ;
if ( out & bit )
wait - > _key | = POLLOUT_SET ;
2009-06-17 02:33:36 +04:00
}
2017-04-08 23:50:24 +03:00
static int do_select ( int n , fd_set_bits * fds , struct timespec64 * end_time )
2005-04-17 02:20:36 +04:00
{
2008-08-31 19:26:40 +04:00
ktime_t expire , * to = NULL ;
2005-04-17 02:20:36 +04:00
struct poll_wqueues table ;
poll_table * wait ;
2008-08-31 19:26:40 +04:00
int retval , i , timed_out = 0 ;
timer: convert timer_slack_ns from unsigned long to u64
This patchset introduces a /proc/<pid>/timerslack_ns interface which
would allow controlling processes to be able to set the timerslack value
on other processes in order to save power by avoiding wakeups (Something
Android currently does via out-of-tree patches).
The first patch tries to fix the internal timer_slack_ns usage which was
defined as a long, which limits the slack range to ~4 seconds on 32bit
systems. It converts it to a u64, which provides the same basically
unlimited slack (500 years) on both 32bit and 64bit machines.
The second patch introduces the /proc/<pid>/timerslack_ns interface
which allows the full 64bit slack range for a task to be read or set on
both 32bit and 64bit machines.
With these two patches, on a 32bit machine, after setting the slack on
bash to 10 seconds:
$ time sleep 1
real 0m10.747s
user 0m0.001s
sys 0m0.005s
The first patch is a little ugly, since I had to chase the slack delta
arguments through a number of functions converting them to u64s. Let me
know if it makes sense to break that up more or not.
Other than that things are fairly straightforward.
This patch (of 2):
The timer_slack_ns value in the task struct is currently a unsigned
long. This means that on 32bit applications, the maximum slack is just
over 4 seconds. However, on 64bit machines, its much much larger (~500
years).
This disparity could make application development a little (as well as
the default_slack) to a u64. This means both 32bit and 64bit systems
have the same effective internal slack range.
Now the existing ABI via PR_GET_TIMERSLACK and PR_SET_TIMERSLACK specify
the interface as a unsigned long, so we preserve that limitation on
32bit systems, where SET_TIMERSLACK can only set the slack to a unsigned
long value, and GET_TIMERSLACK will return ULONG_MAX if the slack is
actually larger then what can be stored by an unsigned long.
This patch also modifies hrtimer functions which specified the slack
delta as a unsigned long.
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Oren Laadan <orenl@cellrox.com>
Cc: Ruchi Kandoi <kandoiruchi@google.com>
Cc: Rom Lemarchand <romlem@android.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Android Kernel Team <kernel-team@android.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-18 00:20:51 +03:00
u64 slack = 0 ;
2013-07-08 17:20:34 +04:00
unsigned int busy_flag = net_busy_loop_on ( ) ? POLL_BUSY_LOOP : 0 ;
2017-03-24 20:08:18 +03:00
unsigned long busy_start = 0 ;
2005-04-17 02:20:36 +04:00
2005-09-10 00:04:14 +04:00
rcu_read_lock ( ) ;
2005-04-17 02:20:36 +04:00
retval = max_select_fd ( n , fds ) ;
2005-09-10 00:04:14 +04:00
rcu_read_unlock ( ) ;
2005-04-17 02:20:36 +04:00
if ( retval < 0 )
return retval ;
n = retval ;
poll_initwait ( & table ) ;
wait = & table . pt ;
2008-08-31 19:26:40 +04:00
if ( end_time & & ! end_time - > tv_sec & & ! end_time - > tv_nsec ) {
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
wait - > _qproc = NULL ;
2008-08-31 19:26:40 +04:00
timed_out = 1 ;
}
2008-09-08 03:08:55 +04:00
if ( end_time & & ! timed_out )
2010-10-28 02:34:53 +04:00
slack = select_estimate_accuracy ( end_time ) ;
2008-09-02 02:55:35 +04:00
2005-04-17 02:20:36 +04:00
retval = 0 ;
for ( ; ; ) {
unsigned long * rinp , * routp , * rexp , * inp , * outp , * exp ;
2013-07-08 17:20:34 +04:00
bool can_busy_loop = false ;
2005-04-17 02:20:36 +04:00
inp = fds - > in ; outp = fds - > out ; exp = fds - > ex ;
rinp = fds - > res_in ; routp = fds - > res_out ; rexp = fds - > res_ex ;
for ( i = 0 ; i < n ; + + rinp , + + routp , + + rexp ) {
2017-07-04 05:25:56 +03:00
unsigned long in , out , ex , all_bits , bit = 1 , j ;
2005-04-17 02:20:36 +04:00
unsigned long res_in = 0 , res_out = 0 , res_ex = 0 ;
2017-07-04 05:25:56 +03:00
__poll_t mask ;
2005-04-17 02:20:36 +04:00
in = * inp + + ; out = * outp + + ; ex = * exp + + ;
all_bits = in | out | ex ;
if ( all_bits = = 0 ) {
2012-07-25 18:40:34 +04:00
i + = BITS_PER_LONG ;
2005-04-17 02:20:36 +04:00
continue ;
}
2012-07-25 18:40:34 +04:00
for ( j = 0 ; j < BITS_PER_LONG ; + + j , + + i , bit < < = 1 ) {
2012-08-28 20:52:22 +04:00
struct fd f ;
2005-04-17 02:20:36 +04:00
if ( i > = n )
break ;
if ( ! ( bit & all_bits ) )
continue ;
2012-08-28 20:52:22 +04:00
f = fdget ( i ) ;
if ( f . file ) {
const struct file_operations * f_op ;
f_op = f . file - > f_op ;
2005-04-17 02:20:36 +04:00
mask = DEFAULT_POLLMASK ;
2013-09-23 00:27:52 +04:00
if ( f_op - > poll ) {
2013-06-24 11:28:03 +04:00
wait_key_set ( wait , in , out ,
2013-07-08 17:20:34 +04:00
bit , busy_flag ) ;
2012-08-28 20:52:22 +04:00
mask = ( * f_op - > poll ) ( f . file , wait ) ;
2009-06-17 02:33:36 +04:00
}
2012-08-28 20:52:22 +04:00
fdput ( f ) ;
2005-04-17 02:20:36 +04:00
if ( ( mask & POLLIN_SET ) & & ( in & bit ) ) {
res_in | = bit ;
retval + + ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
wait - > _qproc = NULL ;
2005-04-17 02:20:36 +04:00
}
if ( ( mask & POLLOUT_SET ) & & ( out & bit ) ) {
res_out | = bit ;
retval + + ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
wait - > _qproc = NULL ;
2005-04-17 02:20:36 +04:00
}
if ( ( mask & POLLEX_SET ) & & ( ex & bit ) ) {
res_ex | = bit ;
retval + + ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
wait - > _qproc = NULL ;
2005-04-17 02:20:36 +04:00
}
2013-06-24 11:28:03 +04:00
/* got something, stop busy polling */
2013-07-08 17:20:34 +04:00
if ( retval ) {
can_busy_loop = false ;
busy_flag = 0 ;
/*
* only remember a returned
* POLL_BUSY_LOOP if we asked for it
*/
} else if ( busy_flag & mask )
can_busy_loop = true ;
2005-04-17 02:20:36 +04:00
}
}
if ( res_in )
* rinp = res_in ;
if ( res_out )
* routp = res_out ;
if ( res_ex )
* rexp = res_ex ;
2008-06-22 23:23:15 +04:00
cond_resched ( ) ;
2005-04-17 02:20:36 +04:00
}
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
wait - > _qproc = NULL ;
2008-08-31 19:26:40 +04:00
if ( retval | | timed_out | | signal_pending ( current ) )
2005-04-17 02:20:36 +04:00
break ;
2008-04-22 02:15:06 +04:00
if ( table . error ) {
2005-04-17 02:20:36 +04:00
retval = table . error ;
break ;
}
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2013-07-08 17:20:34 +04:00
/* only if found POLL_BUSY_LOOP sockets && not out of time */
2013-07-09 14:09:21 +04:00
if ( can_busy_loop & & ! need_resched ( ) ) {
2017-03-24 20:08:18 +03:00
if ( ! busy_start ) {
busy_start = busy_loop_current_time ( ) ;
2013-07-09 14:09:21 +04:00
continue ;
}
2017-03-24 20:08:18 +03:00
if ( ! busy_loop_timeout ( busy_start ) )
2013-07-09 14:09:21 +04:00
continue ;
}
busy_flag = 0 ;
2013-06-24 11:28:03 +04:00
2008-08-31 19:26:40 +04:00
/*
* If this is the first loop and we have a timeout
* given , then we convert to ktime_t and set the to
* pointer to the expiry value .
*/
if ( end_time & & ! to ) {
2016-05-20 03:09:05 +03:00
expire = timespec64_to_ktime ( * end_time ) ;
2008-08-31 19:26:40 +04:00
to = & expire ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
2008-08-31 19:26:40 +04:00
2009-01-07 01:40:59 +03:00
if ( ! poll_schedule_timeout ( & table , TASK_INTERRUPTIBLE ,
to , slack ) )
2008-08-31 19:26:40 +04:00
timed_out = 1 ;
2005-04-17 02:20:36 +04:00
}
poll_freewait ( & table ) ;
return retval ;
}
/*
* We can actually return ERESTARTSYS instead of EINTR , but I ' d
* like to be certain this leads to no problems . So I return
* EINTR just for safety .
*
* Update : ERESTARTSYS breaks at least the xview clock binary , so
* I ' m trying ERESTARTNOHAND which restart only when you want to .
*/
2008-04-23 22:05:15 +04:00
int core_sys_select ( int n , fd_set __user * inp , fd_set __user * outp ,
2016-05-20 03:09:05 +03:00
fd_set __user * exp , struct timespec64 * end_time )
2005-04-17 02:20:36 +04:00
{
fd_set_bits fds ;
2006-04-11 09:52:46 +04:00
void * bits ;
2006-12-10 13:21:12 +03:00
int ret , max_fds ;
2016-10-11 23:51:14 +03:00
size_t size , alloc_size ;
2005-09-10 00:04:10 +04:00
struct fdtable * fdt ;
2006-03-28 13:56:33 +04:00
/* Allocate small arguments on the stack to save memory and be faster */
2006-03-31 20:18:57 +04:00
long stack_fds [ SELECT_STACK_ALLOC / sizeof ( long ) ] ;
2005-04-17 02:20:36 +04:00
ret = - EINVAL ;
if ( n < 0 )
goto out_nofds ;
2006-12-10 13:21:12 +03:00
/* max_fds can increase, so grab it once to avoid race */
2005-09-10 00:04:14 +04:00
rcu_read_lock ( ) ;
2005-09-10 00:04:10 +04:00
fdt = files_fdtable ( current - > files ) ;
2006-12-10 13:21:12 +03:00
max_fds = fdt - > max_fds ;
2005-09-10 00:04:14 +04:00
rcu_read_unlock ( ) ;
2006-12-10 13:21:12 +03:00
if ( n > max_fds )
n = max_fds ;
2005-04-17 02:20:36 +04:00
/*
* We need 6 bitmaps ( in / out / ex for both incoming and outgoing ) ,
* since we used fdset we need to allocate memory in units of
* long - words .
*/
size = FDS_BYTES ( n ) ;
2006-04-11 09:54:08 +04:00
bits = stack_fds ;
if ( size > sizeof ( stack_fds ) / 6 ) {
/* Not enough space in on-stack array; must use kmalloc */
ret = - ENOMEM ;
2016-10-11 23:51:14 +03:00
if ( size > ( SIZE_MAX / 6 ) )
goto out_nofds ;
alloc_size = 6 * size ;
2017-05-09 01:57:27 +03:00
bits = kvmalloc ( alloc_size , GFP_KERNEL ) ;
2006-04-11 09:54:08 +04:00
if ( ! bits )
goto out_nofds ;
}
2006-04-11 09:52:46 +04:00
fds . in = bits ;
fds . out = bits + size ;
fds . ex = bits + 2 * size ;
fds . res_in = bits + 3 * size ;
fds . res_out = bits + 4 * size ;
fds . res_ex = bits + 5 * size ;
2005-04-17 02:20:36 +04:00
if ( ( ret = get_fd_set ( n , inp , fds . in ) ) | |
( ret = get_fd_set ( n , outp , fds . out ) ) | |
( ret = get_fd_set ( n , exp , fds . ex ) ) )
goto out ;
zero_fd_set ( n , fds . res_in ) ;
zero_fd_set ( n , fds . res_out ) ;
zero_fd_set ( n , fds . res_ex ) ;
2008-08-31 19:26:40 +04:00
ret = do_select ( n , & fds , end_time ) ;
2005-04-17 02:20:36 +04:00
if ( ret < 0 )
goto out ;
if ( ! ret ) {
ret = - ERESTARTNOHAND ;
if ( signal_pending ( current ) )
goto out ;
ret = 0 ;
}
if ( set_fd_set ( n , inp , fds . res_in ) | |
set_fd_set ( n , outp , fds . res_out ) | |
set_fd_set ( n , exp , fds . res_ex ) )
ret = - EFAULT ;
out :
2006-03-28 13:56:33 +04:00
if ( bits ! = stack_fds )
2016-10-11 23:51:14 +03:00
kvfree ( bits ) ;
2005-04-17 02:20:36 +04:00
out_nofds :
return ret ;
}
2009-01-14 16:14:25 +03:00
SYSCALL_DEFINE5 ( select , int , n , fd_set __user * , inp , fd_set __user * , outp ,
fd_set __user * , exp , struct timeval __user * , tvp )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
{
2016-05-20 03:09:05 +03:00
struct timespec64 end_time , * to = NULL ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
struct timeval tv ;
int ret ;
if ( tvp ) {
if ( copy_from_user ( & tv , tvp , sizeof ( tv ) ) )
return - EFAULT ;
2008-08-31 19:26:40 +04:00
to = & end_time ;
2008-10-25 23:41:41 +04:00
if ( poll_select_set_timeout ( to ,
tv . tv_sec + ( tv . tv_usec / USEC_PER_SEC ) ,
( tv . tv_usec % USEC_PER_SEC ) * NSEC_PER_USEC ) )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return - EINVAL ;
}
2008-08-31 19:26:40 +04:00
ret = core_sys_select ( n , inp , outp , exp , to ) ;
ret = poll_select_copy_remaining ( & end_time , tvp , 1 , ret ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return ret ;
}
2009-01-14 16:13:57 +03:00
static long do_pselect ( int n , fd_set __user * inp , fd_set __user * outp ,
fd_set __user * exp , struct timespec __user * tsp ,
const sigset_t __user * sigmask , size_t sigsetsize )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
{
sigset_t ksigmask , sigsaved ;
2017-08-05 07:12:31 +03:00
struct timespec64 ts , end_time , * to = NULL ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
int ret ;
if ( tsp ) {
2017-08-05 07:12:31 +03:00
if ( get_timespec64 ( & ts , tsp ) )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return - EFAULT ;
2008-08-31 19:26:40 +04:00
to = & end_time ;
2017-08-05 07:12:31 +03:00
if ( poll_select_set_timeout ( to , ts . tv_sec , ts . tv_nsec ) )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return - EINVAL ;
}
if ( sigmask ) {
/* XXX: Don't preclude handling different sized sigset_t's. */
if ( sigsetsize ! = sizeof ( sigset_t ) )
return - EINVAL ;
if ( copy_from_user ( & ksigmask , sigmask , sizeof ( ksigmask ) ) )
return - EFAULT ;
sigdelsetmask ( & ksigmask , sigmask ( SIGKILL ) | sigmask ( SIGSTOP ) ) ;
sigprocmask ( SIG_SETMASK , & ksigmask , & sigsaved ) ;
}
2009-01-14 00:14:48 +03:00
ret = core_sys_select ( n , inp , outp , exp , to ) ;
2008-08-31 19:26:40 +04:00
ret = poll_select_copy_remaining ( & end_time , tsp , 0 , ret ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
if ( ret = = - ERESTARTNOHAND ) {
/*
* Don ' t restore the signal mask yet . Let do_signal ( ) deliver
* the signal on the way back to userspace , before the signal
* mask is restored .
*/
if ( sigmask ) {
memcpy ( & current - > saved_sigmask , & sigsaved ,
sizeof ( sigsaved ) ) ;
2008-04-30 11:53:06 +04:00
set_restore_sigmask ( ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
} else if ( sigmask )
sigprocmask ( SIG_SETMASK , & sigsaved , NULL ) ;
return ret ;
}
/*
* Most architectures can ' t handle 7 - argument syscalls . So we provide a
* 6 - argument version where the sixth argument is a pointer to a structure
* which has a pointer to the sigset_t itself followed by a size_t containing
* the sigset size .
*/
2009-01-14 16:14:34 +03:00
SYSCALL_DEFINE6 ( pselect6 , int , n , fd_set __user * , inp , fd_set __user * , outp ,
fd_set __user * , exp , struct timespec __user * , tsp ,
void __user * , sig )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
{
size_t sigsetsize = 0 ;
sigset_t __user * up = NULL ;
if ( sig ) {
if ( ! access_ok ( VERIFY_READ , sig , sizeof ( void * ) + sizeof ( size_t ) )
2006-02-01 13:26:09 +03:00
| | __get_user ( up , ( sigset_t __user * __user * ) sig )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
| | __get_user ( sigsetsize ,
2006-02-01 13:26:09 +03:00
( size_t __user * ) ( sig + sizeof ( void * ) ) ) )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return - EFAULT ;
}
2009-01-14 16:13:57 +03:00
return do_pselect ( n , inp , outp , exp , tsp , up , sigsetsize ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
2010-03-11 02:21:13 +03:00
# ifdef __ARCH_WANT_SYS_OLD_SELECT
struct sel_arg_struct {
unsigned long n ;
fd_set __user * inp , * outp , * exp ;
struct timeval __user * tvp ;
} ;
SYSCALL_DEFINE1 ( old_select , struct sel_arg_struct __user * , arg )
{
struct sel_arg_struct a ;
if ( copy_from_user ( & a , arg , sizeof ( a ) ) )
return - EFAULT ;
return sys_select ( a . n , a . inp , a . outp , a . exp , a . tvp ) ;
}
# endif
2005-04-17 02:20:36 +04:00
struct poll_list {
struct poll_list * next ;
int len ;
struct pollfd entries [ 0 ] ;
} ;
# define POLLFD_PER_PAGE ((PAGE_SIZE-sizeof(struct poll_list)) / sizeof(struct pollfd))
2006-06-23 13:05:16 +04:00
/*
* Fish for pollable events on the pollfd - > fd file descriptor . We ' re only
* interested in events matching the pollfd - > events mask , and the result
* matching that mask is both recorded in pollfd - > revents and returned . The
* pwait poll_table will be used by the fd - provided poll handler for waiting ,
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
* if pwait - > _qproc is non - NULL .
2006-06-23 13:05:16 +04:00
*/
2013-06-24 11:28:03 +04:00
static inline unsigned int do_pollfd ( struct pollfd * pollfd , poll_table * pwait ,
2013-07-08 17:20:34 +04:00
bool * can_busy_poll ,
unsigned int busy_flag )
2005-04-17 02:20:36 +04:00
{
2017-07-04 05:25:56 +03:00
__poll_t mask ;
2006-06-23 13:05:16 +04:00
int fd ;
mask = 0 ;
fd = pollfd - > fd ;
if ( fd > = 0 ) {
2012-08-28 20:52:22 +04:00
struct fd f = fdget ( fd ) ;
2006-06-23 13:05:16 +04:00
mask = POLLNVAL ;
2012-08-28 20:52:22 +04:00
if ( f . file ) {
2006-06-23 13:05:16 +04:00
mask = DEFAULT_POLLMASK ;
2013-09-23 00:27:52 +04:00
if ( f . file - > f_op - > poll ) {
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
pwait - > _key = pollfd - > events | POLLERR | POLLHUP ;
2013-07-08 17:20:34 +04:00
pwait - > _key | = busy_flag ;
2012-08-28 20:52:22 +04:00
mask = f . file - > f_op - > poll ( f . file , pwait ) ;
2013-07-08 17:20:34 +04:00
if ( mask & busy_flag )
* can_busy_poll = true ;
2009-06-17 02:33:36 +04:00
}
2006-06-23 13:05:16 +04:00
/* Mask out unneeded events. */
mask & = pollfd - > events | POLLERR | POLLHUP ;
2012-08-28 20:52:22 +04:00
fdput ( f ) ;
2005-04-17 02:20:36 +04:00
}
}
2006-06-23 13:05:16 +04:00
pollfd - > revents = mask ;
return mask ;
2005-04-17 02:20:36 +04:00
}
2016-01-06 08:41:53 +03:00
static int do_poll ( struct poll_list * list , struct poll_wqueues * wait ,
2016-05-20 03:09:05 +03:00
struct timespec64 * end_time )
2005-04-17 02:20:36 +04:00
{
poll_table * pt = & wait - > pt ;
2008-08-31 19:26:40 +04:00
ktime_t expire , * to = NULL ;
int timed_out = 0 , count = 0 ;
timer: convert timer_slack_ns from unsigned long to u64
This patchset introduces a /proc/<pid>/timerslack_ns interface which
would allow controlling processes to be able to set the timerslack value
on other processes in order to save power by avoiding wakeups (Something
Android currently does via out-of-tree patches).
The first patch tries to fix the internal timer_slack_ns usage which was
defined as a long, which limits the slack range to ~4 seconds on 32bit
systems. It converts it to a u64, which provides the same basically
unlimited slack (500 years) on both 32bit and 64bit machines.
The second patch introduces the /proc/<pid>/timerslack_ns interface
which allows the full 64bit slack range for a task to be read or set on
both 32bit and 64bit machines.
With these two patches, on a 32bit machine, after setting the slack on
bash to 10 seconds:
$ time sleep 1
real 0m10.747s
user 0m0.001s
sys 0m0.005s
The first patch is a little ugly, since I had to chase the slack delta
arguments through a number of functions converting them to u64s. Let me
know if it makes sense to break that up more or not.
Other than that things are fairly straightforward.
This patch (of 2):
The timer_slack_ns value in the task struct is currently a unsigned
long. This means that on 32bit applications, the maximum slack is just
over 4 seconds. However, on 64bit machines, its much much larger (~500
years).
This disparity could make application development a little (as well as
the default_slack) to a u64. This means both 32bit and 64bit systems
have the same effective internal slack range.
Now the existing ABI via PR_GET_TIMERSLACK and PR_SET_TIMERSLACK specify
the interface as a unsigned long, so we preserve that limitation on
32bit systems, where SET_TIMERSLACK can only set the slack to a unsigned
long value, and GET_TIMERSLACK will return ULONG_MAX if the slack is
actually larger then what can be stored by an unsigned long.
This patch also modifies hrtimer functions which specified the slack
delta as a unsigned long.
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Oren Laadan <orenl@cellrox.com>
Cc: Ruchi Kandoi <kandoiruchi@google.com>
Cc: Rom Lemarchand <romlem@android.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Android Kernel Team <kernel-team@android.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-18 00:20:51 +03:00
u64 slack = 0 ;
2013-07-08 17:20:34 +04:00
unsigned int busy_flag = net_busy_loop_on ( ) ? POLL_BUSY_LOOP : 0 ;
2017-03-24 20:08:18 +03:00
unsigned long busy_start = 0 ;
2005-04-17 02:20:36 +04:00
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
/* Optimise the no-wait case */
2008-08-31 19:26:40 +04:00
if ( end_time & & ! end_time - > tv_sec & & ! end_time - > tv_nsec ) {
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
pt - > _qproc = NULL ;
2008-08-31 19:26:40 +04:00
timed_out = 1 ;
}
2007-10-17 10:26:18 +04:00
2008-09-08 03:08:55 +04:00
if ( end_time & & ! timed_out )
2010-10-28 02:34:53 +04:00
slack = select_estimate_accuracy ( end_time ) ;
2008-09-02 02:55:35 +04:00
2005-04-17 02:20:36 +04:00
for ( ; ; ) {
struct poll_list * walk ;
2013-07-08 17:20:34 +04:00
bool can_busy_loop = false ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2006-06-23 13:05:16 +04:00
for ( walk = list ; walk ! = NULL ; walk = walk - > next ) {
struct pollfd * pfd , * pfd_end ;
pfd = walk - > entries ;
pfd_end = pfd + walk - > len ;
for ( ; pfd ! = pfd_end ; pfd + + ) {
/*
* Fish for events . If we found one , record it
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
* and kill poll_table - > _qproc , so we don ' t
2006-06-23 13:05:16 +04:00
* needlessly register any other waiters after
* this . They ' ll get immediately deregistered
* when we break out and return .
*/
2013-07-08 17:20:34 +04:00
if ( do_pollfd ( pfd , pt , & can_busy_loop ,
busy_flag ) ) {
2006-06-23 13:05:16 +04:00
count + + ;
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
pt - > _qproc = NULL ;
2013-07-08 17:20:34 +04:00
/* found something, stop busy polling */
busy_flag = 0 ;
can_busy_loop = false ;
2006-06-23 13:05:16 +04:00
}
}
2005-04-17 02:20:36 +04:00
}
2006-06-23 13:05:16 +04:00
/*
* All waiters have already been registered , so don ' t provide
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
* a poll_table - > _qproc to them on the next loop iteration .
2006-06-23 13:05:16 +04:00
*/
poll: add poll_requested_events() and poll_does_not_wait() functions
In some cases the poll() implementation in a driver has to do different
things depending on the events the caller wants to poll for. An example
is when a driver needs to start a DMA engine if the caller polls for
POLLIN, but doesn't want to do that if POLLIN is not requested but instead
only POLLOUT or POLLPRI is requested. This is something that can happen
in the video4linux subsystem among others.
Unfortunately, the current epoll/poll/select implementation doesn't
provide that information reliably. The poll_table_struct does have it: it
has a key field with the event mask. But once a poll() call matches one
or more bits of that mask any following poll() calls are passed a NULL
poll_table pointer.
Also, the eventpoll implementation always left the key field at ~0 instead
of using the requested events mask.
This was changed in eventpoll.c so the key field now contains the actual
events that should be polled for as set by the caller.
The solution to the NULL poll_table pointer is to set the qproc field to
NULL in poll_table once poll() matches the events, not the poll_table
pointer itself. That way drivers can obtain the mask through a new
poll_requested_events inline.
The poll_table_struct can still be NULL since some kernel code calls it
internally (netfs_state_poll() in ./drivers/staging/pohmelfs/netfs.h). In
that case poll_requested_events() returns ~0 (i.e. all events).
Very rarely drivers might want to know whether poll_wait will actually
wait. If another earlier file descriptor in the set already matched the
events the caller wanted to wait for, then the kernel will return from the
select() call without waiting. This might be useful information in order
to avoid doing expensive work.
A new helper function poll_does_not_wait() is added that drivers can use
to detect this situation. This is now used in sock_poll_wait() in
include/net/sock.h. This was the only place in the kernel that needed
this information.
Drivers should no longer access any of the poll_table internals, but use
the poll_requested_events() and poll_does_not_wait() access functions
instead. In order to enforce that the poll_table fields are now prepended
with an underscore and a comment was added warning against using them
directly.
This required a change in unix_dgram_poll() in unix/af_unix.c which used
the key field to get the requested events. It's been replaced by a call
to poll_requested_events().
For qproc it was especially important to change its name since the
behavior of that field changes with this patch since this function pointer
can now be NULL when that wasn't possible in the past.
Any driver accessing the qproc or key fields directly will now fail to compile.
Some notes regarding the correctness of this patch: the driver's poll()
function is called with a 'struct poll_table_struct *wait' argument. This
pointer may or may not be NULL, drivers can never rely on it being one or
the other as that depends on whether or not an earlier file descriptor in
the select()'s fdset matched the requested events.
There are only three things a driver can do with the wait argument:
1) obtain the key field:
events = wait ? wait->key : ~0;
This will still work although it should be replaced with the new
poll_requested_events() function (which does exactly the same).
This will now even work better, since wait is no longer set to NULL
unnecessarily.
2) use the qproc callback. This could be deadly since qproc can now be
NULL. Renaming qproc should prevent this from happening. There are no
kernel drivers that actually access this callback directly, BTW.
3) test whether wait == NULL to determine whether poll would return without
waiting. This is no longer sufficient as the correct test is now
wait == NULL || wait->_qproc == NULL.
However, the worst that can happen here is a slight performance hit in
the case where wait != NULL and wait->_qproc == NULL. In that case the
driver will assume that poll_wait() will actually add the fd to the set
of waiting file descriptors. Of course, poll_wait() will not do that
since it tests for wait->_qproc. This will not break anything, though.
There is only one place in the whole kernel where this happens
(sock_poll_wait() in include/net/sock.h) and that code will be replaced
by a call to poll_does_not_wait() in the next patch.
Note that even if wait->_qproc != NULL drivers cannot rely on poll_wait()
actually waiting. The next file descriptor from the set might match the
event mask and thus any possible waits will never happen.
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Reviewed-by: Jonathan Corbet <corbet@lwn.net>
Reviewed-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Davide Libenzi <davidel@xmailserver.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: David Miller <davem@davemloft.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-03-24 02:02:27 +04:00
pt - > _qproc = NULL ;
2007-10-17 10:26:18 +04:00
if ( ! count ) {
count = wait - > error ;
if ( signal_pending ( current ) )
count = - EINTR ;
}
2008-08-31 19:26:40 +04:00
if ( count | | timed_out )
2005-04-17 02:20:36 +04:00
break ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2013-07-08 17:20:34 +04:00
/* only if found POLL_BUSY_LOOP sockets && not out of time */
2013-07-09 14:09:21 +04:00
if ( can_busy_loop & & ! need_resched ( ) ) {
2017-03-24 20:08:18 +03:00
if ( ! busy_start ) {
busy_start = busy_loop_current_time ( ) ;
2013-07-09 14:09:21 +04:00
continue ;
}
2017-03-24 20:08:18 +03:00
if ( ! busy_loop_timeout ( busy_start ) )
2013-07-09 14:09:21 +04:00
continue ;
}
busy_flag = 0 ;
2013-06-28 16:59:35 +04:00
2008-08-31 19:26:40 +04:00
/*
* If this is the first loop and we have a timeout
* given , then we convert to ktime_t and set the to
* pointer to the expiry value .
*/
if ( end_time & & ! to ) {
2016-05-20 03:09:05 +03:00
expire = timespec64_to_ktime ( * end_time ) ;
2008-08-31 19:26:40 +04:00
to = & expire ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
2009-01-07 01:40:59 +03:00
if ( ! poll_schedule_timeout ( wait , TASK_INTERRUPTIBLE , to , slack ) )
2008-08-31 19:26:40 +04:00
timed_out = 1 ;
2005-04-17 02:20:36 +04:00
}
return count ;
}
2006-03-28 13:56:33 +04:00
# define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \
sizeof ( struct pollfd ) )
2017-04-08 23:50:24 +03:00
static int do_sys_poll ( struct pollfd __user * ufds , unsigned int nfds ,
2016-05-20 03:09:05 +03:00
struct timespec64 * end_time )
2005-04-17 02:20:36 +04:00
{
struct poll_wqueues table ;
2007-10-17 10:26:17 +04:00
int err = - EFAULT , fdcount , len , size ;
2006-03-31 20:18:57 +04:00
/* Allocate small arguments on the stack to save memory and be
faster - use long to make sure the buffer is aligned properly
on 64 bit archs to avoid unaligned access */
long stack_pps [ POLL_STACK_ALLOC / sizeof ( long ) ] ;
2007-10-17 10:26:17 +04:00
struct poll_list * const head = ( struct poll_list * ) stack_pps ;
struct poll_list * walk = head ;
unsigned long todo = nfds ;
2005-04-17 02:20:36 +04:00
2010-03-06 00:42:42 +03:00
if ( nfds > rlimit ( RLIMIT_NOFILE ) )
2005-04-17 02:20:36 +04:00
return - EINVAL ;
2007-10-17 10:26:17 +04:00
len = min_t ( unsigned int , nfds , N_STACK_PPS ) ;
for ( ; ; ) {
walk - > next = NULL ;
walk - > len = len ;
if ( ! len )
break ;
2005-04-17 02:20:36 +04:00
2007-10-17 10:26:17 +04:00
if ( copy_from_user ( walk - > entries , ufds + nfds - todo ,
sizeof ( struct pollfd ) * walk - > len ) )
goto out_fds ;
todo - = walk - > len ;
if ( ! todo )
break ;
2005-04-17 02:20:36 +04:00
2007-10-17 10:26:17 +04:00
len = min ( todo , POLLFD_PER_PAGE ) ;
size = sizeof ( struct poll_list ) + sizeof ( struct pollfd ) * len ;
walk = walk - > next = kmalloc ( size , GFP_KERNEL ) ;
if ( ! walk ) {
err = - ENOMEM ;
2005-04-17 02:20:36 +04:00
goto out_fds ;
}
}
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2007-10-17 10:26:17 +04:00
poll_initwait ( & table ) ;
2016-01-06 08:41:53 +03:00
fdcount = do_poll ( head , & table , end_time ) ;
2007-10-17 10:26:17 +04:00
poll_freewait ( & table ) ;
2005-04-17 02:20:36 +04:00
2007-10-17 10:26:17 +04:00
for ( walk = head ; walk ; walk = walk - > next ) {
2005-04-17 02:20:36 +04:00
struct pollfd * fds = walk - > entries ;
int j ;
2007-10-17 10:26:17 +04:00
for ( j = 0 ; j < walk - > len ; j + + , ufds + + )
if ( __put_user ( fds [ j ] . revents , & ufds - > revents ) )
2005-04-17 02:20:36 +04:00
goto out_fds ;
}
2007-10-17 10:26:17 +04:00
2005-04-17 02:20:36 +04:00
err = fdcount ;
out_fds :
2007-10-17 10:26:17 +04:00
walk = head - > next ;
while ( walk ) {
struct poll_list * pos = walk ;
walk = walk - > next ;
kfree ( pos ) ;
2005-04-17 02:20:36 +04:00
}
2007-10-17 10:26:17 +04:00
2005-04-17 02:20:36 +04:00
return err ;
}
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2007-10-17 10:27:18 +04:00
static long do_restart_poll ( struct restart_block * restart_block )
{
2008-08-31 19:26:40 +04:00
struct pollfd __user * ufds = restart_block - > poll . ufds ;
int nfds = restart_block - > poll . nfds ;
2016-05-20 03:09:05 +03:00
struct timespec64 * to = NULL , end_time ;
2007-10-17 10:27:18 +04:00
int ret ;
2008-08-31 19:26:40 +04:00
if ( restart_block - > poll . has_timeout ) {
end_time . tv_sec = restart_block - > poll . tv_sec ;
end_time . tv_nsec = restart_block - > poll . tv_nsec ;
to = & end_time ;
}
ret = do_sys_poll ( ufds , nfds , to ) ;
2007-10-17 10:27:18 +04:00
if ( ret = = - EINTR ) {
restart_block - > fn = do_restart_poll ;
ret = - ERESTART_RESTARTBLOCK ;
}
return ret ;
}
2009-01-14 16:14:25 +03:00
SYSCALL_DEFINE3 ( poll , struct pollfd __user * , ufds , unsigned int , nfds ,
sys_poll: fix incorrect type for 'timeout' parameter
The 'poll()' system call timeout parameter is supposed to be 'int', not
'long'.
Now, the reason this matters is that right now 32-bit compat mode is
broken on at least x86-64, because the 32-bit code just calls
'sys_poll()' directly on x86-64, and the 32-bit argument will have been
zero-extended, turning a signed 'int' into a large unsigned 'long'
value.
We could just introduce a 'compat_sys_poll()' function for this, and
that may eventually be what we have to do, but since the actual standard
poll() semantics is *supposed* to be 'int', and since at least on x86-64
glibc sign-extends the argument before invocing the system call (so
nobody can actually use a 64-bit timeout value in user space _anyway_,
even in 64-bit binaries), the simpler solution would seem to be to just
fix the definition of the system call to match what it should have been
from the very start.
If it turns out that somebody somehow circumvents the user-level libc
64-bit sign extension and actually uses a large unsigned 64-bit timeout
despite that not being how poll() is supposed to work, we will need to
do the compat_sys_poll() approach.
Reported-by: Thomas Meyer <thomas@m3y3r.de>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-02-22 05:24:20 +04:00
int , timeout_msecs )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
{
2016-05-20 03:09:05 +03:00
struct timespec64 end_time , * to = NULL ;
2007-10-17 10:27:18 +04:00
int ret ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
2008-08-31 19:26:40 +04:00
if ( timeout_msecs > = 0 ) {
to = & end_time ;
poll_select_set_timeout ( to , timeout_msecs / MSEC_PER_SEC ,
NSEC_PER_MSEC * ( timeout_msecs % MSEC_PER_SEC ) ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
2008-08-31 19:26:40 +04:00
ret = do_sys_poll ( ufds , nfds , to ) ;
2007-10-17 10:27:18 +04:00
if ( ret = = - EINTR ) {
struct restart_block * restart_block ;
2008-08-31 19:26:40 +04:00
2015-02-13 02:01:14 +03:00
restart_block = & current - > restart_block ;
2007-10-17 10:27:18 +04:00
restart_block - > fn = do_restart_poll ;
2008-08-31 19:26:40 +04:00
restart_block - > poll . ufds = ufds ;
restart_block - > poll . nfds = nfds ;
if ( timeout_msecs > = 0 ) {
restart_block - > poll . tv_sec = end_time . tv_sec ;
restart_block - > poll . tv_nsec = end_time . tv_nsec ;
restart_block - > poll . has_timeout = 1 ;
} else
restart_block - > poll . has_timeout = 0 ;
2007-10-17 10:27:18 +04:00
ret = - ERESTART_RESTARTBLOCK ;
}
return ret ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
2009-01-14 16:14:34 +03:00
SYSCALL_DEFINE5 ( ppoll , struct pollfd __user * , ufds , unsigned int , nfds ,
struct timespec __user * , tsp , const sigset_t __user * , sigmask ,
size_t , sigsetsize )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
{
sigset_t ksigmask , sigsaved ;
2017-08-05 07:12:31 +03:00
struct timespec64 ts , end_time , * to = NULL ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
int ret ;
if ( tsp ) {
2017-08-05 07:12:31 +03:00
if ( get_timespec64 ( & ts , tsp ) )
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return - EFAULT ;
2008-08-31 19:26:40 +04:00
to = & end_time ;
if ( poll_select_set_timeout ( to , ts . tv_sec , ts . tv_nsec ) )
return - EINVAL ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
if ( sigmask ) {
/* XXX: Don't preclude handling different sized sigset_t's. */
if ( sigsetsize ! = sizeof ( sigset_t ) )
return - EINVAL ;
if ( copy_from_user ( & ksigmask , sigmask , sizeof ( ksigmask ) ) )
return - EFAULT ;
sigdelsetmask ( & ksigmask , sigmask ( SIGKILL ) | sigmask ( SIGSTOP ) ) ;
sigprocmask ( SIG_SETMASK , & ksigmask , & sigsaved ) ;
}
2008-08-31 19:26:40 +04:00
ret = do_sys_poll ( ufds , nfds , to ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
/* We can restart this syscall, usually */
if ( ret = = - EINTR ) {
/*
* Don ' t restore the signal mask yet . Let do_signal ( ) deliver
* the signal on the way back to userspace , before the signal
* mask is restored .
*/
if ( sigmask ) {
memcpy ( & current - > saved_sigmask , & sigsaved ,
sizeof ( sigsaved ) ) ;
2008-04-30 11:53:06 +04:00
set_restore_sigmask ( ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
}
ret = - ERESTARTNOHAND ;
} else if ( sigmask )
sigprocmask ( SIG_SETMASK , & sigsaved , NULL ) ;
2008-08-31 19:26:40 +04:00
ret = poll_select_copy_remaining ( & end_time , tsp , 0 , ret ) ;
[PATCH] Add pselect/ppoll system call implementation
The following implementation of ppoll() and pselect() system calls
depends on the architecture providing a TIF_RESTORE_SIGMASK flag in the
thread_info.
These system calls have to change the signal mask during their
operation, and signal handlers must be invoked using the new, temporary
signal mask. The old signal mask must be restored either upon successful
exit from the system call, or upon returning from the invoked signal
handler if the system call is interrupted. We can't simply restore the
original signal mask and return to userspace, since the restored signal
mask may actually block the signal which interrupted the system call.
The TIF_RESTORE_SIGMASK flag deals with this by causing the syscall exit
path to trap into do_signal() just as TIF_SIGPENDING does, and by
causing do_signal() to use the saved signal mask instead of the current
signal mask when setting up the stack frame for the signal handler -- or
by causing do_signal() to simply restore the saved signal mask in the
case where there is no handler to be invoked.
The first patch implements the sys_pselect() and sys_ppoll() system
calls, which are present only if TIF_RESTORE_SIGMASK is defined. That
#ifdef should go away in time when all architectures have implemented
it. The second patch implements TIF_RESTORE_SIGMASK for the PowerPC
kernel (in the -mm tree), and the third patch then removes the
arch-specific implementations of sys_rt_sigsuspend() and replaces them
with generic versions using the same trick.
The fourth and fifth patches, provided by David Howells, implement
TIF_RESTORE_SIGMASK for FR-V and i386 respectively, and the sixth patch
adds the syscalls to the i386 syscall table.
This patch:
Add the pselect() and ppoll() system calls, providing core routines usable by
the original select() and poll() system calls and also the new calls (with
their semantics w.r.t timeouts).
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-01-19 04:44:05 +03:00
return ret ;
}
2017-04-08 23:50:24 +03:00
# ifdef CONFIG_COMPAT
# define __COMPAT_NFDBITS (8 * sizeof(compat_ulong_t))
static
2017-08-05 07:12:31 +03:00
int compat_poll_select_copy_remaining ( struct timespec64 * end_time , void __user * p ,
2017-04-08 23:50:24 +03:00
int timeval , int ret )
{
2017-08-05 07:12:31 +03:00
struct timespec64 ts ;
2017-04-08 23:50:24 +03:00
if ( ! p )
return ret ;
if ( current - > personality & STICKY_TIMEOUTS )
goto sticky ;
/* No update for zero timeout */
if ( ! end_time - > tv_sec & & ! end_time - > tv_nsec )
return ret ;
2017-08-05 07:12:31 +03:00
ktime_get_ts64 ( & ts ) ;
ts = timespec64_sub ( * end_time , ts ) ;
2017-04-08 23:50:24 +03:00
if ( ts . tv_sec < 0 )
ts . tv_sec = ts . tv_nsec = 0 ;
if ( timeval ) {
struct compat_timeval rtv ;
rtv . tv_sec = ts . tv_sec ;
rtv . tv_usec = ts . tv_nsec / NSEC_PER_USEC ;
if ( ! copy_to_user ( p , & rtv , sizeof ( rtv ) ) )
return ret ;
} else {
2017-08-05 07:12:31 +03:00
if ( ! compat_put_timespec64 ( & ts , p ) )
2017-04-08 23:50:24 +03:00
return ret ;
}
/*
* If an application puts its timeval in read - only memory , we
* don ' t want the Linux - specific update to the timeval to
* cause a fault after the select has completed
* successfully . However , because we ' re not updating the
* timeval , we can ' t restart the system call .
*/
sticky :
if ( ret = = - ERESTARTNOHAND )
ret = - EINTR ;
return ret ;
}
/*
* Ooo , nasty . We need here to frob 32 - bit unsigned longs to
* 64 - bit unsigned longs .
*/
static
int compat_get_fd_set ( unsigned long nr , compat_ulong_t __user * ufdset ,
unsigned long * fdset )
{
if ( ufdset ) {
2017-06-04 10:24:26 +03:00
return compat_get_bitmap ( fdset , ufdset , nr ) ;
2017-04-08 23:50:24 +03:00
} else {
fs/select: Fix memory corruption in compat_get_fd_set()
Commit 464d62421cb8 ("select: switch compat_{get,put}_fd_set() to
compat_{get,put}_bitmap()") changed the calculation on how many bytes
need to be zeroed when userspace handed over a NULL pointer for a fdset
array in the select syscall.
The calculation was changed in compat_get_fd_set() wrongly from
memset(fdset, 0, ((nr + 1) & ~1)*sizeof(compat_ulong_t));
to
memset(fdset, 0, ALIGN(nr, BITS_PER_LONG));
The ALIGN(nr, BITS_PER_LONG) calculates the number of _bits_ which need
to be zeroed in the target fdset array (rounded up to the next full bits
for an unsigned long).
But the memset() call expects the number of _bytes_ to be zeroed.
This leads to clearing more memory than wanted (on the stack area or
even at kmalloc()ed memory areas) and to random kernel crashes as we
have seen them on the parisc platform.
The correct change should have been
memset(fdset, 0, (ALIGN(nr, BITS_PER_LONG) / BITS_PER_LONG) * BYTES_PER_LONG);
which is the same as can be archieved with a call to
zero_fd_set(nr, fdset).
Fixes: 464d62421cb8 ("select: switch compat_{get,put}_fd_set() to compat_{get,put}_bitmap()"
Acked-by:: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-08-23 23:37:00 +03:00
zero_fd_set ( nr , fdset ) ;
2017-06-04 10:24:26 +03:00
return 0 ;
2017-04-08 23:50:24 +03:00
}
}
static
int compat_set_fd_set ( unsigned long nr , compat_ulong_t __user * ufdset ,
unsigned long * fdset )
{
if ( ! ufdset )
return 0 ;
2017-06-04 10:24:26 +03:00
return compat_put_bitmap ( ufdset , fdset , nr ) ;
2017-04-08 23:50:24 +03:00
}
/*
* This is a virtual copy of sys_select from fs / select . c and probably
* should be compared to it from time to time
*/
/*
* We can actually return ERESTARTSYS instead of EINTR , but I ' d
* like to be certain this leads to no problems . So I return
* EINTR just for safety .
*
* Update : ERESTARTSYS breaks at least the xview clock binary , so
* I ' m trying ERESTARTNOHAND which restart only when you want to .
*/
static int compat_core_sys_select ( int n , compat_ulong_t __user * inp ,
compat_ulong_t __user * outp , compat_ulong_t __user * exp ,
2017-08-05 07:12:31 +03:00
struct timespec64 * end_time )
2017-04-08 23:50:24 +03:00
{
fd_set_bits fds ;
void * bits ;
int size , max_fds , ret = - EINVAL ;
struct fdtable * fdt ;
long stack_fds [ SELECT_STACK_ALLOC / sizeof ( long ) ] ;
if ( n < 0 )
goto out_nofds ;
/* max_fds can increase, so grab it once to avoid race */
rcu_read_lock ( ) ;
fdt = files_fdtable ( current - > files ) ;
max_fds = fdt - > max_fds ;
rcu_read_unlock ( ) ;
if ( n > max_fds )
n = max_fds ;
/*
* We need 6 bitmaps ( in / out / ex for both incoming and outgoing ) ,
* since we used fdset we need to allocate memory in units of
* long - words .
*/
size = FDS_BYTES ( n ) ;
bits = stack_fds ;
if ( size > sizeof ( stack_fds ) / 6 ) {
bits = kmalloc ( 6 * size , GFP_KERNEL ) ;
ret = - ENOMEM ;
if ( ! bits )
goto out_nofds ;
}
fds . in = ( unsigned long * ) bits ;
fds . out = ( unsigned long * ) ( bits + size ) ;
fds . ex = ( unsigned long * ) ( bits + 2 * size ) ;
fds . res_in = ( unsigned long * ) ( bits + 3 * size ) ;
fds . res_out = ( unsigned long * ) ( bits + 4 * size ) ;
fds . res_ex = ( unsigned long * ) ( bits + 5 * size ) ;
if ( ( ret = compat_get_fd_set ( n , inp , fds . in ) ) | |
( ret = compat_get_fd_set ( n , outp , fds . out ) ) | |
( ret = compat_get_fd_set ( n , exp , fds . ex ) ) )
goto out ;
zero_fd_set ( n , fds . res_in ) ;
zero_fd_set ( n , fds . res_out ) ;
zero_fd_set ( n , fds . res_ex ) ;
ret = do_select ( n , & fds , end_time ) ;
if ( ret < 0 )
goto out ;
if ( ! ret ) {
ret = - ERESTARTNOHAND ;
if ( signal_pending ( current ) )
goto out ;
ret = 0 ;
}
if ( compat_set_fd_set ( n , inp , fds . res_in ) | |
compat_set_fd_set ( n , outp , fds . res_out ) | |
compat_set_fd_set ( n , exp , fds . res_ex ) )
ret = - EFAULT ;
out :
if ( bits ! = stack_fds )
kfree ( bits ) ;
out_nofds :
return ret ;
}
COMPAT_SYSCALL_DEFINE5 ( select , int , n , compat_ulong_t __user * , inp ,
compat_ulong_t __user * , outp , compat_ulong_t __user * , exp ,
struct compat_timeval __user * , tvp )
{
2017-08-05 07:12:31 +03:00
struct timespec64 end_time , * to = NULL ;
2017-04-08 23:50:24 +03:00
struct compat_timeval tv ;
int ret ;
if ( tvp ) {
if ( copy_from_user ( & tv , tvp , sizeof ( tv ) ) )
return - EFAULT ;
to = & end_time ;
if ( poll_select_set_timeout ( to ,
tv . tv_sec + ( tv . tv_usec / USEC_PER_SEC ) ,
( tv . tv_usec % USEC_PER_SEC ) * NSEC_PER_USEC ) )
return - EINVAL ;
}
ret = compat_core_sys_select ( n , inp , outp , exp , to ) ;
ret = compat_poll_select_copy_remaining ( & end_time , tvp , 1 , ret ) ;
return ret ;
}
struct compat_sel_arg_struct {
compat_ulong_t n ;
compat_uptr_t inp ;
compat_uptr_t outp ;
compat_uptr_t exp ;
compat_uptr_t tvp ;
} ;
COMPAT_SYSCALL_DEFINE1 ( old_select , struct compat_sel_arg_struct __user * , arg )
{
struct compat_sel_arg_struct a ;
if ( copy_from_user ( & a , arg , sizeof ( a ) ) )
return - EFAULT ;
return compat_sys_select ( a . n , compat_ptr ( a . inp ) , compat_ptr ( a . outp ) ,
compat_ptr ( a . exp ) , compat_ptr ( a . tvp ) ) ;
}
static long do_compat_pselect ( int n , compat_ulong_t __user * inp ,
compat_ulong_t __user * outp , compat_ulong_t __user * exp ,
struct compat_timespec __user * tsp , compat_sigset_t __user * sigmask ,
compat_size_t sigsetsize )
{
sigset_t ksigmask , sigsaved ;
2017-08-05 07:12:31 +03:00
struct timespec64 ts , end_time , * to = NULL ;
2017-04-08 23:50:24 +03:00
int ret ;
if ( tsp ) {
2017-08-05 07:12:31 +03:00
if ( compat_get_timespec64 ( & ts , tsp ) )
2017-04-08 23:50:24 +03:00
return - EFAULT ;
to = & end_time ;
if ( poll_select_set_timeout ( to , ts . tv_sec , ts . tv_nsec ) )
return - EINVAL ;
}
if ( sigmask ) {
if ( sigsetsize ! = sizeof ( compat_sigset_t ) )
return - EINVAL ;
2017-09-04 04:45:17 +03:00
if ( get_compat_sigset ( & ksigmask , sigmask ) )
2017-04-08 23:50:24 +03:00
return - EFAULT ;
sigdelsetmask ( & ksigmask , sigmask ( SIGKILL ) | sigmask ( SIGSTOP ) ) ;
sigprocmask ( SIG_SETMASK , & ksigmask , & sigsaved ) ;
}
ret = compat_core_sys_select ( n , inp , outp , exp , to ) ;
ret = compat_poll_select_copy_remaining ( & end_time , tsp , 0 , ret ) ;
if ( ret = = - ERESTARTNOHAND ) {
/*
* Don ' t restore the signal mask yet . Let do_signal ( ) deliver
* the signal on the way back to userspace , before the signal
* mask is restored .
*/
if ( sigmask ) {
memcpy ( & current - > saved_sigmask , & sigsaved ,
sizeof ( sigsaved ) ) ;
set_restore_sigmask ( ) ;
}
} else if ( sigmask )
sigprocmask ( SIG_SETMASK , & sigsaved , NULL ) ;
return ret ;
}
COMPAT_SYSCALL_DEFINE6 ( pselect6 , int , n , compat_ulong_t __user * , inp ,
compat_ulong_t __user * , outp , compat_ulong_t __user * , exp ,
struct compat_timespec __user * , tsp , void __user * , sig )
{
compat_size_t sigsetsize = 0 ;
compat_uptr_t up = 0 ;
if ( sig ) {
if ( ! access_ok ( VERIFY_READ , sig ,
sizeof ( compat_uptr_t ) + sizeof ( compat_size_t ) ) | |
__get_user ( up , ( compat_uptr_t __user * ) sig ) | |
__get_user ( sigsetsize ,
( compat_size_t __user * ) ( sig + sizeof ( up ) ) ) )
return - EFAULT ;
}
return do_compat_pselect ( n , inp , outp , exp , tsp , compat_ptr ( up ) ,
sigsetsize ) ;
}
COMPAT_SYSCALL_DEFINE5 ( ppoll , struct pollfd __user * , ufds ,
unsigned int , nfds , struct compat_timespec __user * , tsp ,
const compat_sigset_t __user * , sigmask , compat_size_t , sigsetsize )
{
sigset_t ksigmask , sigsaved ;
2017-08-05 07:12:31 +03:00
struct timespec64 ts , end_time , * to = NULL ;
2017-04-08 23:50:24 +03:00
int ret ;
if ( tsp ) {
2017-08-05 07:12:31 +03:00
if ( compat_get_timespec64 ( & ts , tsp ) )
2017-04-08 23:50:24 +03:00
return - EFAULT ;
to = & end_time ;
if ( poll_select_set_timeout ( to , ts . tv_sec , ts . tv_nsec ) )
return - EINVAL ;
}
if ( sigmask ) {
if ( sigsetsize ! = sizeof ( compat_sigset_t ) )
return - EINVAL ;
2017-09-04 04:45:17 +03:00
if ( get_compat_sigset ( & ksigmask , sigmask ) )
2017-04-08 23:50:24 +03:00
return - EFAULT ;
sigdelsetmask ( & ksigmask , sigmask ( SIGKILL ) | sigmask ( SIGSTOP ) ) ;
sigprocmask ( SIG_SETMASK , & ksigmask , & sigsaved ) ;
}
ret = do_sys_poll ( ufds , nfds , to ) ;
/* We can restart this syscall, usually */
if ( ret = = - EINTR ) {
/*
* Don ' t restore the signal mask yet . Let do_signal ( ) deliver
* the signal on the way back to userspace , before the signal
* mask is restored .
*/
if ( sigmask ) {
memcpy ( & current - > saved_sigmask , & sigsaved ,
sizeof ( sigsaved ) ) ;
set_restore_sigmask ( ) ;
}
ret = - ERESTARTNOHAND ;
} else if ( sigmask )
sigprocmask ( SIG_SETMASK , & sigsaved , NULL ) ;
ret = compat_poll_select_copy_remaining ( & end_time , tsp , 0 , ret ) ;
return ret ;
}
# endif