micscif/micscif_select.c

/*
 * Implementation of select and poll
 *
 * Copyright 2011-2012 Intel Corporation.
 *
 * This file is a derivative of fs/select.c from within the Linux kernel
 * source distribution, version 2.6.34; it has been modified (starting
 * in May 2011) to work within the context of the SCIF driver.
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License, version 2, as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301
 * USA.
 *
 * Initial comment from fs/select.c:
 *
 * 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).
 */

#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/file.h>
#include <linux/hrtimer.h>
#include <linux/module.h>

#include "mic/micscif.h"

#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,10,0))
#include <linux/sched/rt.h>
#endif

struct poll_table_page {
        struct poll_table_page *next;
        struct poll_table_entry *entry;
        struct poll_table_entry entries[0];
};

/*
 * 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..
 */

#define MAX_SLACK       (100 * NSEC_PER_MSEC)

static long __estimate_accuracy(struct timespec *tv)
{
        long slack;
        int divfactor = 1000;

        if (tv->tv_sec < 0)
                return 0;

        if (task_nice(current) > 0)
                divfactor = divfactor / 5;

        if (tv->tv_sec > MAX_SLACK / (NSEC_PER_SEC/divfactor))
                return MAX_SLACK;

        slack = tv->tv_nsec / divfactor;
        slack += tv->tv_sec * (NSEC_PER_SEC/divfactor);

        if (slack > MAX_SLACK)
                return MAX_SLACK;

        return slack;
}

static long estimate_accuracy(struct timespec *tv)
{
        unsigned long ret;
        struct timespec now;

        /*
         * Realtime tasks get a slack of 0 for obvious reasons.
         */

        if (rt_task(current))
                return 0;

        ktime_get_ts(&now);
        now = timespec_sub(*tv, now);
        ret = __estimate_accuracy(&now);
        if (ret < current->timer_slack_ns)
                return current->timer_slack_ns;
        return ret;
}

#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.
 */
static void __pollwait(struct file *filp __attribute__((unused)), wait_queue_head_t *wait_address,
                       poll_table *p);

static void scif_poll_initwait(struct poll_wqueues *pwq)
{
        init_poll_funcptr(&pwq->pt, __pollwait);
        pwq->polling_task = current;
        pwq->triggered = 0;
        pwq->error = 0;
        pwq->table = NULL;
        pwq->inline_index = 0;
}

static void free_poll_entry(struct poll_table_entry *entry)
{
        remove_wait_queue(entry->wait_address, &entry->wait);
}

static void scif_poll_freewait(struct poll_wqueues *pwq)
{
        struct poll_table_page * p = pwq->table;
        int i;
        for (i = 0; i < pwq->inline_index; i++)
                free_poll_entry(pwq->inline_entries + i);
        while (p) {
                struct poll_table_entry *entry;
                struct poll_table_page *old;

                entry = p->entry;
                do {
                        entry--;
                        free_poll_entry(entry);
                } while (entry > p->entries);
                old = p;
                p = p->next;
                free_page((unsigned long) old);
        }
}

static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p)
{
        struct poll_table_page *table = p->table;

        if (p->inline_index < N_INLINE_POLL_ENTRIES)
                return p->inline_entries + p->inline_index++;

        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;
                        return NULL;
                }
                new_table->entry = new_table->entries;
                new_table->next = table;
                p->table = new_table;
                table = new_table;
        }

        return table->entry++;
}

static int __pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key)
{
        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()
         * and is paired with set_mb() in poll_schedule_timeout.
         */
        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);
}

static int pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key)
{
        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);
}

/* Add a new entry */
static void __pollwait(struct file *filp __attribute__((unused)), wait_queue_head_t *wait_address,
                                poll_table *p)
{
        struct poll_wqueues *pwq = container_of(p, struct poll_wqueues, pt);
        struct poll_table_entry *entry = poll_get_entry(pwq);
        if (!entry)
                return;
        entry->filp = NULL;
        entry->wait_address = wait_address;
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,10,0))
        entry->key = p->_key;
#else
        entry->key = p->key;
#endif
        init_waitqueue_func_entry(&entry->wait, pollwake);
        entry->wait.private = pwq;
        add_wait_queue(wait_address, &entry->wait);
}

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)
                rc = schedule_hrtimeout_range(expires, slack, HRTIMER_MODE_ABS);
        __set_current_state(TASK_RUNNING);

        /*
         * Prepare for the next iteration.
         *
         * The following set_mb() serves two purposes.  First, it's
         * 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.
         */
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(4,2,0))
        smp_store_mb(pwq->triggered, 0);
#else
        set_mb(pwq->triggered, 0);
#endif

        return rc;
}

static unsigned int scif_poll_kernel(poll_table *pwait, struct endpt *ep)
{
        return __scif_pollfd(NULL, pwait, ep);
}

/*
 * 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,
 * if non-NULL.
 */
static inline unsigned int do_pollfd(struct scif_pollepd *pollfd, poll_table *pwait)
{
        unsigned int mask;
        scif_epd_t epd;

        mask = 0;
        epd = pollfd->epd;
        if (epd) {
                mask = POLLNVAL;
                mask = DEFAULT_POLLMASK;
                if (pwait)
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,10,0))
                        pwait->_key = pollfd->events | POLLERR | POLLHUP;
#else
                        pwait->key = pollfd->events | POLLERR | POLLHUP;
#endif
                mask = scif_poll_kernel(pwait, epd);
                /* Mask out unneeded events. */
                mask &= pollfd->events | POLLERR | POLLHUP;
        }
        pollfd->revents = mask;

        return mask;
}

static int do_poll(unsigned int nfds,  struct scif_pollepd *ufds,
                   struct poll_wqueues *wait, struct timespec *end_time)
{
        poll_table* pt = &wait->pt;
        ktime_t expire, *to = NULL;
        int timed_out = 0, count = 0, i = 0;
        unsigned long slack = 0;

        /* Optimise the no-wait case */
        if (end_time && !end_time->tv_sec && !end_time->tv_nsec) {
                pt = NULL;
                timed_out = 1;
        }

        if (end_time && !timed_out)
                slack = estimate_accuracy(end_time);

        for (;;) {
                for (i = 0; i < nfds; i++) {
                        /*
                         * Fish for events. If we found one, record it
                         * and kill the poll_table, so we don't
                         * needlessly register any other waiters after
                         * this. They'll get immediately deregistered
                         * when we break out and return.
                         */
                        if (do_pollfd(ufds + i, pt)) {
                                count++;
                                pt = NULL;
                        }
                }
                /*
                 * All waiters have already been registered, so don't provide
                 * a poll_table to them on the next loop iteration.
                 */
                pt = NULL;
                if (!count) {
                        count = wait->error;
                        if (signal_pending(current))
                                count = -EINTR;
                }
                if (count || timed_out)
                        break;

                /*
                 * 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) {
                        expire = timespec_to_ktime(*end_time);
                        to = &expire;
                }

                if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack))
                        timed_out = 1;
        }
        return count;
}

static int do_scif_poll(struct scif_pollepd *ufds, unsigned int nfds,
                struct timespec *end_time)
{
        struct poll_wqueues table;
        int epdcount;

        scif_poll_initwait(&table);
        epdcount = do_poll(nfds, ufds, &table, end_time);
        scif_poll_freewait(&table);

        return epdcount;
}

/*
 * Add two timespec values and do a safety check for overflow.
 * It's assumed that both values are valid (>= 0)
 */
static struct timespec scif_timespec_add_safe(const struct timespec lhs,
                                  const struct timespec rhs)
{
        struct timespec res;

        set_normalized_timespec(&res, lhs.tv_sec + rhs.tv_sec,
                                lhs.tv_nsec + rhs.tv_nsec);

        if (res.tv_sec < lhs.tv_sec || res.tv_sec < rhs.tv_sec)
                res.tv_sec = TIME_T_MAX;

        return res;
}
/**
 * poll_select_set_timeout - helper function to setup the timeout value
 * @to:         pointer to timespec variable for the final timeout
 * @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.
 */
static int scif_poll_select_set_timeout(struct timespec *to, long sec, long nsec)
{
        struct timespec ts = {.tv_sec = sec, .tv_nsec = nsec};

        if (!timespec_valid(&ts))
                return -EINVAL;

        /* Optimize for the zero timeout value here */
        if (!sec && !nsec) {
                to->tv_sec = to->tv_nsec = 0;
        } else {
                ktime_get_ts(to);
                *to = scif_timespec_add_safe(*to, ts);
        }
        return 0;
}

int scif_poll(struct scif_pollepd *ufds, unsigned int nfds, long timeout_msecs)
{
        struct timespec end_time, *to = NULL;
        if (timeout_msecs >= 0) {
                to = &end_time;
                scif_poll_select_set_timeout(to, timeout_msecs / MSEC_PER_SEC,
                        NSEC_PER_MSEC * (timeout_msecs % MSEC_PER_SEC));
        }

        return do_scif_poll(ufds, nfds, to);
}
EXPORT_SYMBOL(scif_poll);