[AIO] add support for POSIX AIO completion signals This patch adds POSIX AIO completion notification event by adding an aio_sigevent field to the aiocb. The sigevent structure is filled in by the user application as part of the AIO request preparation. Upon request completion, the kernel notifies the application using those sigevent parameters. The original patch was by Sébastien Dugué with modifications by Benjamin LaHaise. Signed-off-by: Sébastien Dugué Signed-off-by: Benjamin LaHaise diff -purN --exclude=description 816_iocb_run_deadlock/fs/aio.c 817_sigevent/fs/aio.c --- 816_iocb_run_deadlock/fs/aio.c 2005-08-17 13:46:21.000000000 -0400 +++ 817_sigevent/fs/aio.c 2005-08-17 13:46:48.000000000 -0400 @@ -404,6 +404,7 @@ static struct kiocb fastcall *__aio_get_ req->ki_cancel = NULL; req->ki_retry = NULL; req->ki_dtor = NULL; + req->ki_signo = 0; req->private = NULL; INIT_LIST_HEAD(&req->ki_run_list); @@ -932,6 +933,97 @@ void fastcall kick_iocb(struct kiocb *io } EXPORT_SYMBOL(kick_iocb); +static void __aio_send_signal(struct kiocb *iocb) +{ + struct siginfo info; + struct task_struct *p; + unsigned long flags; + int ret = -1; + + memset(&info, 0, sizeof(struct siginfo)); + + info.si_signo = iocb->ki_signo; + info.si_errno = 0; + info.si_code = SI_ASYNCIO; + info.si_pid = 0; + info.si_uid = 0; + info.si_value = iocb->ki_sigev_value; + + read_lock(&tasklist_lock); + p = find_task_by_pid(iocb->ki_pid); + if (!p || !p->sighand) + goto out_unlock; + + /* Do we have permission to signal this task? */ + if ((iocb->ki_euid ^ p->suid) && (iocb->ki_euid ^ p->uid) + && (iocb->ki_uid ^ p->suid) && (iocb->ki_uid ^ p->uid)) + goto out_unlock; /* No. */ + + spin_lock_irqsave(&p->sighand->siglock, flags); + + switch(iocb->ki_notify) { + case IO_NOTIFY_SIGNAL: + ret = __group_send_sig_info(iocb->ki_signo, &info, p); + break; + case IO_NOTIFY_THREAD_ID: + //ret = specific_send_sig_info(iocb->ki_signo, &info, p); + ret = __group_send_sig_info(iocb->ki_signo, &info, p); + break; + } + + spin_unlock_irqrestore(&p->sighand->siglock, flags); + + if (ret) + printk(KERN_DEBUG "__aio_send_signal: failed to send signal %d to %d\n", + iocb->ki_signo, iocb->ki_pid); + +out_unlock: + read_unlock(&tasklist_lock); +} + +static void __aio_write_evt(struct kioctx *ctx, struct io_event *event) +{ + struct aio_ring_info *info; + struct aio_ring *ring; + struct io_event *ring_event; + unsigned long tail; + + info = &ctx->ring_info; + + /* add a completion event to the ring buffer. + * must be done holding ctx->ctx_lock to prevent + * other code from messing with the tail + * pointer since we might be called from irq + * context. + */ + + ring = kmap_atomic(info->ring_pages[0], KM_IRQ1); + + tail = info->tail; + ring_event = aio_ring_event(info, tail, KM_IRQ0); + if (++tail >= info->nr) + tail = 0; + + *ring_event = *event; + + dprintk("aio_write_evt: %p[%lu]: %Lx %Lx %Lx %Lx\n", + ctx, tail, event->obj, event->data, event->res, event->res2); + + /* after flagging the request as done, we + * must never even look at it again + */ + + smp_wmb(); /* make event visible before updating tail */ + + info->tail = tail; + ring->tail = tail; + + put_aio_ring_event(ring_event, KM_IRQ0); + kunmap_atomic(ring, KM_IRQ1); + + pr_debug("added to ring at [%lu]\n", tail); +} + /* aio_complete * Called when the io request on the given iocb is complete. * Returns true if this is the last user of the request. The @@ -940,11 +1032,8 @@ EXPORT_SYMBOL(kick_iocb); int fastcall aio_complete(struct kiocb *iocb, long res, long res2) { struct kioctx *ctx = iocb->ki_ctx; - struct aio_ring_info *info; - struct aio_ring *ring; - struct io_event *event; + struct io_event event; unsigned long flags; - unsigned long tail; int ret; /* Special case handling for sync iocbs: events go directly @@ -971,14 +1060,13 @@ int fastcall aio_complete(struct kiocb * return ret; } - info = &ctx->ring_info; + /* insert event in the event ring */ + + event.obj = (u64)(unsigned long)iocb->ki_obj.user; + event.data = iocb->ki_user_data; + event.res = res; + event.res2 = res2; - /* add a completion event to the ring buffer. - * must be done holding ctx->ctx_lock to prevent - * other code from messing with the tail - * pointer since we might be called from irq - * context. - */ spin_lock_irqsave(&ctx->ctx_lock, flags); if (iocb->ki_run_list.prev && !list_empty(&iocb->ki_run_list)) @@ -991,34 +1079,10 @@ int fastcall aio_complete(struct kiocb * if (kiocbIsCancelled(iocb)) goto put_rq; - ring = kmap_atomic(info->ring_pages[0], KM_IRQ1); - - tail = info->tail; - event = aio_ring_event(info, tail, KM_IRQ0); - if (++tail >= info->nr) - tail = 0; - - event->obj = (u64)(unsigned long)iocb->ki_obj.user; - event->data = iocb->ki_user_data; - event->res = res; - event->res2 = res2; - - dprintk("aio_complete: %p[%lu]: %p: %p %Lx %lx %lx\n", - ctx, tail, iocb, iocb->ki_obj.user, iocb->ki_user_data, - res, res2); - - /* after flagging the request as done, we - * must never even look at it again - */ - smp_wmb(); /* make event visible before updating tail */ - - info->tail = tail; - ring->tail = tail; + __aio_write_evt(ctx, &event); - put_aio_ring_event(event, KM_IRQ0); - kunmap_atomic(ring, KM_IRQ1); - - pr_debug("added to ring %p at [%lu]\n", iocb, tail); + if (iocb->ki_signo) + __aio_send_signal(iocb); pr_debug("%ld retries: %d of %d\n", iocb->ki_retried, iocb->ki_nbytes - iocb->ki_left, iocb->ki_nbytes); @@ -1499,6 +1563,71 @@ int aio_wake_function(wait_queue_t *wait return 0; } +static long iocb_setup_sigevent(struct kiocb *req, + struct sigevent __user *user_event) +{ + int notify = 0; + pid_t aio_pid; + int aio_signo = 0; + + req->ki_uid = current->uid; + req->ki_euid = current->euid; + + if (!access_ok(VERIFY_READ, user_event, sizeof(struct sigevent))) + return -EFAULT; + /* + * We avoid copying the whole sigevent bunch and only get the + * needed fields. + */ + if (unlikely(__get_user(aio_pid, &user_event->sigev_notify_thread_id))) + return -EFAULT; + + if (unlikely(__get_user(aio_signo, &user_event->sigev_signo))) + return -EFAULT; + + if (unlikely(__copy_from_user(&req->ki_sigev_value, + &user_event->sigev_value, + sizeof(sigval_t)))) + return -EFAULT; + + if (!aio_signo) + return 0; /* no signal number, we do nothing */ + + if (aio_pid == 0) { + /* notify itself */ + aio_pid = current->pid; + notify = IO_NOTIFY_SIGNAL; + } else { + pid_t group_id; + task_t *ptask; + /* notify given thread */ + + /* caller thread and target thread must be in same + * thread group + */ + read_lock(&tasklist_lock); + + ptask = find_task_by_pid(aio_pid); + if (ptask) + group_id = ptask->tgid; + read_unlock(&tasklist_lock); + + if (unlikely (ptask == NULL)) + return -ESRCH; + + if (group_id != current->tgid) + return -EINVAL; + + notify = IO_NOTIFY_THREAD_ID; + } + + req->ki_pid = aio_pid; + req->ki_signo = aio_signo; + req->ki_notify = notify; + + return 0; +} + int fastcall io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, struct iocb *iocb) { @@ -1507,8 +1636,7 @@ int fastcall io_submit_one(struct kioctx ssize_t ret; /* enforce forwards compatibility on users */ - if (unlikely(iocb->aio_reserved1 || iocb->aio_reserved2 || - iocb->aio_reserved3)) { + if (unlikely(iocb->aio_reserved1)) { pr_debug("EINVAL: io_submit: reserve field set\n"); return -EINVAL; } @@ -1517,6 +1645,7 @@ int fastcall io_submit_one(struct kioctx if (unlikely( (iocb->aio_buf != (unsigned long)iocb->aio_buf) || (iocb->aio_nbytes != (size_t)iocb->aio_nbytes) || + (iocb->aio_sigeventp != (unsigned long)iocb->aio_sigeventp) || ((ssize_t)iocb->aio_nbytes < 0) )) { pr_debug("EINVAL: io_submit: overflow check\n"); @@ -1552,8 +1681,16 @@ int fastcall io_submit_one(struct kioctx req->ki_run_list.next = req->ki_run_list.prev = NULL; req->ki_retried = 0; - ret = aio_setup_iocb(req); + /* handle setting up the sigevent for POSIX AIO signals */ + if (iocb->aio_sigeventp) { + ret = iocb_setup_sigevent(req, + (struct sigevent __user *)(unsigned long) + iocb->aio_sigeventp); + if (ret) + goto out_put_req; + } + ret = aio_setup_iocb(req); if (ret) goto out_put_req; diff -purN --exclude=description 816_iocb_run_deadlock/include/linux/aio.h 817_sigevent/include/linux/aio.h --- 816_iocb_run_deadlock/include/linux/aio.h 2005-08-15 15:32:51.000000000 -0400 +++ 817_sigevent/include/linux/aio.h 2005-08-15 15:48:34.000000000 -0400 @@ -6,6 +6,12 @@ #include #include +#include + +enum { + IO_NOTIFY_SIGNAL = 0, /* send signal to a processe */ + IO_NOTIFY_THREAD_ID = 1, /* send signal to a specific thread */ +}; #define AIO_MAXSEGS 4 #define AIO_KIOGRP_NR_ATOMIC 8 @@ -87,6 +93,13 @@ struct kiocb { long ki_queued; /* just for testing */ void *private; + + /* to notify a process on I/O event only valid if ki_signo != 0 */ + pid_t ki_pid; + __u16 ki_signo; + __u16 ki_notify; + uid_t ki_uid, ki_euid; + sigval_t ki_sigev_value; }; #define is_sync_kiocb(iocb) ((iocb)->ki_key == KIOCB_SYNC_KEY) diff -purN --exclude=description 816_iocb_run_deadlock/include/linux/aio_abi.h 817_sigevent/include/linux/aio_abi.h --- 816_iocb_run_deadlock/include/linux/aio_abi.h 2005-06-20 13:33:38.000000000 -0400 +++ 817_sigevent/include/linux/aio_abi.h 2005-08-15 15:48:34.000000000 -0400 @@ -80,8 +80,9 @@ struct iocb { __u64 aio_nbytes; __s64 aio_offset; + __u64 aio_sigeventp; /* pointer to struct sigevent */ + /* extra parameters */ - __u64 aio_reserved2; /* TODO: use this for a (struct sigevent *) */ __u64 aio_reserved3; }; /* 64 bytes */