Use WARN() instead of a printk+WARN_ON() pair; this way the message
becomes part of the warning section for better reporting/collection.

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

If we have multiple tasks freeing io contexts when as-iosched
is being unloaded, we could complete() ioc_gone twice. Fix that by
protecting ioc_gone complete() and clearing with a spinlock for
just that purpose. Doesn't matter from a performance perspective,
since it'll only enter that path when ioc_gone != NULL (when as-iosched
is being rmmod'ed).

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

AS scheduler alternates between issuing read and write batches. It does
the batch switch only after all requests from the previous batch are
completed.

When switching to a write batch, if there is an on-going read request,
it waits for its completion and indicates its intention of switching by
setting ad->changed_batch and the new direction but does not update the
batch_expire_time for the new write batch which it does in the case of
no previous pending requests.
On completion of the read request, it sees that we were waiting for the
switch and schedules work for kblockd right away and resets the
ad->changed_data flag.
Now when kblockd enters dispatch_request where it is expected to pick
up a write request, it in turn ends the write batch because the
batch_expire_timer was not updated and shows the expire timestamp for
the previous batch.

This results in the write starvation for all the cases where there is
the intention for switching to a write batch, but there is a previous
in-flight read request and the batch gets reverted to a read_batch
right away.

This also holds true in the reverse case (switching from a write batch
to a read batch with an in-flight write request).

I've checked that this bug exists on 2.6.11, 2.6.18, 2.6.24 and
linux-2.6-block git HEAD. I've tested the fix on x86 platforms with
SCSI drives where the driver asks for the next request while a current
request is in-flight.

This patch is based off linux-2.6-block git HEAD.

Bug reproduction:
A simple scenario which reproduces this bug is:
- dd if=/dev/hda3 of=/dev/null &
- lilo
   The lilo takes forever to complete.

This can also be reproduced fairly easily with the earlier dd and
another test
program doing msync().

The example test program below should print out a message after every
iteration
but it simply hangs forever. With this bugfix it makes forward progress.

====
Example test program using msync() (thanks to suleiman AT google DOT
com)

inline uint64_t
rdtsc(void)
{
         int64_t tsc;

         __asm __volatile("rdtsc" : "=A" (tsc));
         return (tsc);
}

int
main(int argc, char **argv)
{
         struct stat st;
         uint64_t e, s, t;
         char *p, q;
         long i;
         int fd;

         if (argc < 2) {
                 printf("Usage: %s <file>\n", argv[0]);
                 return (1);
         }

         if ((fd = open(argv[1], O_RDWR | O_NOATIME)) < 0)
                 err(1, "open");

         if (fstat(fd, &st) < 0)
                 err(1, "fstat");

         p = mmap(NULL, st.st_size, PROT_READ | PROT_WRITE,
MAP_SHARED, fd, 0);

         t = 0;
         for (i = 0; i < 1000; i++) {
                 *p = 0;
                 msync(p, 4096, MS_SYNC);
                 s = rdtsc();
                *p = 0;
                 __asm __volatile(""::: "memory");
                 e = rdtsc();
                 if (argc > 2)
                         printf("%d: %lld cycles %jd %jd\n",
                                i, e - s, (intmax_t)s, (intmax_t)e);
                 t += e - s;
         }
         printf("average time: %lld cycles\n", t / 1000);
         return (0);
}

Cc: <stable@kernel.org>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

It blindly copies everything in the io_context, including the lock.
That doesn't work so well for either lock ordering or lockdep.

There seems zero point in swapping io contexts on a request to request
merge, so the best point of action is to just remove it.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

Since it's acquired from irq context, all locking must be of the
irq safe variant. Most are already inside the queue lock (which
already disables interrupts), but the io scheduler rmmod path
always has irqs enabled and the put_io_context() path may legally
be called with irqs enabled (even if it isn't usually). So fixup
those two.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

If the two requests belong to the same io context, we will attempt
to lock the same lock twice. But swapping contexts is pointless in
that case, so just check for rioc == nioc before doing the double
lock and copy.

Tested-by: Olof Johansson <olof@lixom.net>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

changes to anticipatory io scheduler for io_context sharing

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

elv_register() always returns 0, and there isn't anything it does where
it should return an error (the only error condition is so grave that
it's handled with a BUG_ON).

Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

New write batches currently start from where the last one completed.
We have no idea where the head is after switching batches, so this
makes little sense.  Instead, start the next batch from the request
with the earliest deadline in the hope that we avoid a deadline
expiry later on.

Signed-off-by: Aaron Carroll <aaronc@gelato.unsw.edu.au>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>

Two comments refer to deadlines applying to reads only.  This is
not the case.

Signed-off-by: Aaron Carroll <aaronc@gelato.unsw.edu.au>
Acked-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>