Andrew Morton wrote:
> What is the testing status of this?
I've beaten on it with my multithreaded test program on a 2-way
Pentium II box.
> Have any real-life applications been converted?
Not that I'm aware of. My test program uses a real-life async i/o
thread pool framework which I've converted to Linux aio, but that
framework also depends on IOCB_CMD_NOOP (which Ben LaHaise has been
sitting on for months).
Note that aio poll is primarily for programs which either have a
thread pool architecture or which need to use the aio framework for
other operations. Single threaded straight-networking applications
would be better served by epoll.
> Does it require libaio changes to test?
Don't think so--one could create an IO_CMD_POLL iocb by hand. I just
found it easier to change libaio.
> growl. Please don't put function prototypes in .c files. It defeats
> typechecking. I've moved this to aio.h.
That's how the patch was before I got my hands on it. I would have
moved it to poll.h, but whatever.
> > + iocb->ki_users++;
>
> There is no locking around this modification of ->ki_users. Is this
> correct?
Yes. This is the submission path. No other thread can access that
iocb until after the wmb().
> Barriers always need comments explaining why they are there.
I looked through LXR to find similar wmb() calls to mimic, but didn't
find any non-trivially commented calls. I hope the comment I added is
acceptable.
> > + if (unlikely(apiocb->outofmem) && xchg(&apiocb->armed, NULL)) {
> > + async_poll_freewait(apiocb, NULL);
> > + aio_put_req(iocb);
> > + aio_put_req(iocb);
>
> Is the double-put intentional?
Yes it was intentional and yes it is a bug. One frees the reference
held by the async_poll() call (obtained before the wmb() call) as
protection against submit/complete and submit/cancel races. The other
frees the reference held by the poll operation itself (obtained in
aio_get_req()), but when the routine returns -ENOMEM that causes
io_submit_one() to call aio_complete() which again frees that
reference.
Corrected patch follows.
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./fs/Makefile ../linux-aiopoll/fs/Makefile
--- ./fs/Makefile Fri May 16 17:01:25 2003
+++ ../linux-aiopoll/fs/Makefile Fri May 16 16:57:55 2003
@@ -10,7 +10,8 @@
namei.o fcntl.o ioctl.o readdir.o select.o fifo.o locks.o \
dcache.o inode.o attr.o bad_inode.o file.o dnotify.o \
filesystems.o namespace.o seq_file.o xattr.o libfs.o \
- fs-writeback.o mpage.o direct-io.o aio.o eventpoll.o
+ fs-writeback.o mpage.o direct-io.o aio.o eventpoll.o \
+ aiopoll.o
obj-$(CONFIG_COMPAT) += compat.o
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./fs/aio.c ../linux-aiopoll/fs/aio.c
--- ./fs/aio.c Fri May 16 17:01:25 2003
+++ ../linux-aiopoll/fs/aio.c Mon May 19 11:38:32 2003
@@ -984,6 +984,19 @@
return -EINVAL;
}
+ssize_t generic_aio_poll(struct file *file, struct kiocb *req, struct iocb *iocb)
+{
+ unsigned events = iocb->aio_buf;
+
+ /* Did the user set any bits they weren't supposed to? (The
+ * above is actually a cast.
+ */
+ if (unlikely(events != iocb->aio_buf))
+ return -EINVAL;
+
+ return async_poll(req, events);
+}
+
int FASTCALL(io_submit_one(struct kioctx *ctx, struct iocb *user_iocb,
struct iocb *iocb));
int io_submit_one(struct kioctx *ctx, struct iocb *user_iocb,
@@ -1070,6 +1083,9 @@
if (file->f_op->aio_fsync)
ret = file->f_op->aio_fsync(req, 0);
break;
+ case IOCB_CMD_POLL:
+ ret = generic_aio_poll(file, req, iocb);
+ break;
default:
dprintk("EINVAL: io_submit: no operation provided\n");
ret = -EINVAL;
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./fs/aiopoll.c ../linux-aiopoll/fs/aiopoll.c
--- fs/aiopoll.c 1969-12-31 16:00:00.000000000 -0800
+++ ../linux-aiopoll/fs/aiopoll.c 2003-05-19 12:26:41.000000000 -0700
@@ -0,0 +1,178 @@
+/*
+ * This file contains the procedures for the handling of aio poll
+ */
+
+#include <linux/slab.h>
+#include <linux/poll.h>
+#include <linux/aio.h>
+#include <linux/init.h>
+
+struct async_poll_iocb;
+
+struct async_poll_entry {
+ wait_queue_t wait;
+ wait_queue_head_t *whead;
+ struct async_poll_entry *next;
+ struct async_poll_iocb *apiocb;
+};
+
+struct async_poll_iocb {
+ poll_table pt;
+ void *armed;
+ int outofmem;
+ int events; /* event mask for async poll */
+ struct async_poll_entry *ehead;
+ struct async_poll_entry entry[2]; /* space for two entries */
+};
+
+static kmem_cache_t *async_poll_entry_cache;
+
+static inline struct async_poll_iocb *kiocb_to_apiocb(struct kiocb *iocb)
+{
+ BUG_ON(sizeof(struct async_poll_iocb) > KIOCB_PRIVATE_SIZE);
+ return (struct async_poll_iocb *)iocb->private;
+}
+
+static inline struct kiocb *apiocb_to_kiocb(struct async_poll_iocb *apiocb)
+{
+ return container_of((void *)apiocb, struct kiocb, private);
+}
+
+static void async_poll_freewait(struct async_poll_iocb *apiocb, wait_queue_t *wait)
+{
+ struct async_poll_entry *entry = apiocb->ehead;
+ struct async_poll_entry *old;
+
+ while (entry) {
+ if (wait != &entry->wait) {
+ remove_wait_queue(entry->whead, &entry->wait);
+ } else {
+ __remove_wait_queue(entry->whead, &entry->wait);
+ }
+ old = entry;
+ entry = entry->next;
+ if (old != &apiocb->entry[0] && old != &apiocb->entry[1]) {
+ kmem_cache_free(async_poll_entry_cache, old);
+ }
+ }
+}
+
+static int async_poll_waiter(wait_queue_t *wait, unsigned mode, int sync)
+{
+ struct async_poll_entry *entry = (struct async_poll_entry *)wait;
+ struct async_poll_iocb *apiocb = entry->apiocb;
+ struct kiocb *iocb = apiocb_to_kiocb(apiocb);
+ unsigned int mask;
+
+ mask = iocb->ki_filp->f_op->poll(iocb->ki_filp, NULL);
+ mask &= apiocb->events | POLLERR | POLLHUP;
+ if (mask) {
+ if (xchg(&apiocb->armed, NULL)) {
+ async_poll_freewait(apiocb, wait);
+ aio_complete(iocb, mask, 0);
+ return 1;
+ }
+ }
+ return 0;
+}
+
+int async_poll_cancel(struct kiocb *iocb, struct io_event *res)
+{
+ struct async_poll_iocb *apiocb = kiocb_to_apiocb(iocb);
+ void *armed;
+
+ armed = xchg(&apiocb->armed, NULL);
+ aio_put_req(iocb);
+ if (armed) {
+ async_poll_freewait(apiocb, NULL);
+ /*
+ * Since async_poll_freewait() locks the wait queue, we
+ * know that async_poll_waiter() is either not going to
+ * be run or has finished all its work.
+ */
+ aio_put_req(iocb);
+ return 0;
+ }
+ return -EAGAIN;
+}
+
+static void async_poll_queue_proc(struct file *file, wait_queue_head_t *whead, poll_table *pt)
+{
+ struct async_poll_iocb *apiocb = (struct async_poll_iocb *)pt;
+ struct async_poll_entry *entry;
+
+ if (!apiocb->ehead) {
+ entry = &apiocb->entry[0];
+ } else if (apiocb->ehead == &apiocb->entry[0]) {
+ entry = &apiocb->entry[1];
+ } else {
+ entry = kmem_cache_alloc(async_poll_entry_cache, SLAB_KERNEL);
+ if (!entry) {
+ apiocb->outofmem = 1;
+ return;
+ }
+ }
+ init_waitqueue_func_entry(&entry->wait, async_poll_waiter);
+ entry->whead = whead;
+ entry->next = apiocb->ehead;
+ entry->apiocb = apiocb;
+ add_wait_queue(whead, &entry->wait);
+ apiocb->ehead = entry;
+}
+
+int async_poll(struct kiocb *iocb, int events)
+{
+ unsigned int mask;
+ struct async_poll_iocb *apiocb = kiocb_to_apiocb(iocb);
+
+ /* Fast path */
+ if (iocb->ki_filp->f_op && iocb->ki_filp->f_op->poll) {
+ mask = iocb->ki_filp->f_op->poll(iocb->ki_filp, NULL);
+ mask &= events | POLLERR | POLLHUP;
+ if (mask & events)
+ return events;
+ }
+
+ init_poll_funcptr(&apiocb->pt, async_poll_queue_proc);
+ apiocb->armed = &apiocb;
+ apiocb->outofmem = 0;
+ apiocb->events = events;
+ apiocb->ehead = NULL;
+ iocb->ki_users++;
+
+ /*
+ * Flush the preceeding before letting the complete
+ * or cancel paths get at this iocb.
+ */
+ wmb();
+
+ mask = DEFAULT_POLLMASK;
+ if (iocb->ki_filp->f_op && iocb->ki_filp->f_op->poll)
+ mask = iocb->ki_filp->f_op->poll(iocb->ki_filp, &apiocb->pt);
+ mask &= events | POLLERR | POLLHUP;
+ if (mask && xchg(&apiocb->armed, NULL)) {
+ async_poll_freewait(apiocb, NULL);
+ aio_complete(iocb, mask, 0);
+ }
+ if (unlikely(apiocb->outofmem) && xchg(&apiocb->armed, NULL)) {
+ async_poll_freewait(apiocb, NULL);
+ aio_put_req(iocb);
+ return -ENOMEM;
+ }
+
+ iocb->ki_cancel = async_poll_cancel;
+ aio_put_req(iocb);
+ return -EIOCBQUEUED;
+}
+
+static int __init async_poll_init(void)
+{
+ async_poll_entry_cache = kmem_cache_create("async poll entry",
+ sizeof(struct async_poll_entry), 0, 0, NULL, NULL);
+ if (!async_poll_entry_cache)
+ panic("unable to alloc poll_entry_cache");
+ return 0;
+}
+
+module_init(async_poll_init);
+
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./fs/select.c ../linux-aiopoll/fs/select.c
--- ./fs/select.c Fri May 16 17:01:26 2003
+++ ../linux-aiopoll/fs/select.c Fri May 16 16:57:55 2003
@@ -24,7 +24,6 @@
#include <asm/uaccess.h>
#define ROUND_UP(x,y) (((x)+(y)-1)/(y))
-#define DEFAULT_POLLMASK (POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM)
struct poll_table_entry {
struct file * filp;
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./include/linux/aio.h ../linux-aiopoll/include/linux/aio.h
--- ./include/linux/aio.h Fri May 16 17:02:12 2003
+++ ../linux-aiopoll/include/linux/aio.h Mon May 19 11:38:32 2003
@@ -166,4 +166,6 @@
/* for sysctl: */
extern unsigned aio_max_nr, aio_max_size, aio_max_pinned;
+extern int async_poll(struct kiocb *iocb, int events);
+
#endif /* __LINUX__AIO_H */
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./include/linux/aio_abi.h ../linux-aiopoll/include/linux/aio_abi.h
--- ./include/linux/aio_abi.h Fri May 16 17:02:12 2003
+++ ../linux-aiopoll/include/linux/aio_abi.h Fri May 16 16:58:29 2003
@@ -38,8 +38,8 @@
IOCB_CMD_FDSYNC = 3,
/* These two are experimental.
* IOCB_CMD_PREADX = 4,
- * IOCB_CMD_POLL = 5,
*/
+ IOCB_CMD_POLL = 5,
IOCB_CMD_NOOP = 6,
};
diff -Nur --exclude=SCCS --exclude=BitKeeper --exclude=ChangeSet ./include/linux/poll.h ../linux-aiopoll/include/linux/poll.h
--- ./include/linux/poll.h Fri May 16 17:02:16 2003
+++ ../linux-aiopoll/include/linux/poll.h Mon May 19 11:20:13 2003
@@ -10,6 +10,8 @@
#include <linux/mm.h>
#include <asm/uaccess.h>
+#define DEFAULT_POLLMASK (POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM)
+
struct poll_table_struct;
/*
On Llu, 2003-05-19 at 20:38, John Myers wrote:
> Andrew Morton wrote:
> > What is the testing status of this?
>
> I've beaten on it with my multithreaded test program on a 2-way
> Pentium II box.
Can someone beat this code up on an SMP PPC/PPC64 box - that would show
up far more than x86 does
Alan Cox <[email protected]> wrote:
>
> On Llu, 2003-05-19 at 20:38, John Myers wrote:
> > Andrew Morton wrote:
> > > What is the testing status of this?
> >
> > I've beaten on it with my multithreaded test program on a 2-way
> > Pentium II box.
>
> Can someone beat this code up on an SMP PPC/PPC64 box - that would show
> up far more than x86 does
>
I can do that, but would need the test app...
Andrew Morton wrote:
>I can do that, but would need the test app...
>
>
Unfortunately, the thread pool framework the test app uses is proprietary.
[email protected] (John Myers) wrote:
>
> Andrew Morton wrote:
>
> >I can do that, but would need the test app...
> >
> >
> Unfortunately, the thread pool framework the test app uses is proprietary.
>
That is a problem.
Generally, adding new code in codepaths which lots of apps use is low-risk
because problems will be found quickly.
But from a stabilising-the-kernel viewpoint, adding new code which does not
require elevated privileges, and which is not used by any current
applications and which doesn't even have a decent test suite is a big
worry.
There is a little testcase in
http://www.kvack.org/~bcrl/aio/aio-userspace-18a.tar.gz but it is singly
threaded.