There is no need for release memory ordering when assigning a null pointer
to an rcu pointer, since the null pointer cannot be dereferenced.  Add
a specialization of assign() that takes advantage of this fact.

Reviewed-by: Nadav Har'El <nyh@cloudius-systems.com>
Signed-off-by: Avi Kivity <avi@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The add_buf_wait will sleep if there is not enough buf in the avail
ring.

When there is more space available, it can be woke up by wakeup_waiter.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

In some cases netperf relies on SIGALRM to wake sleeping
send*/recv*() calls. alarm() API on OSv doesn't fully comply to
specification, i.e. it executes SIGALRM handler but doesn't
interrupt sleeping syscalls. Because of this there is a race
condition that makes netserver sleep infinitely in some cases.

This commit introduces patch for netperf that acts as a work-around
for this problem by setting socket timeout for network operations.

This commit also introduces simple script that builds patched netperf
in automated manner.

Signed-off-by: Dmitry Fleytman <dmitry@daynix.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The name is shorter and clearer and more consistent with the name used
in virtio-scsi.

Signed-off-by: Asias He <asias.hejun@gmail.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias.hejun@gmail.com>
[ penberg: keep comments vertically aligned ]
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

bio is already signalled in all cases. No need to do it in destructor.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

It's better not to indent after the namespace specifier.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Change from blk::make_virtio_request to blk::make_request.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

We are under the virtio namespace, it makes no sense to repeat the
virtio prefix again in the virito_blk driver.

Change the naming from virtio::virtio_blk to virtio::blk.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The check is moved to bdev_read and bdev_write.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The check is moved to bdev_read and bdev_write.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The check is moved to bdev_read and bdev_write.

Sigaed-off-by: Asias He <asias@cloudius-systems.com>

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Over the process of creating a zpool, history_str_get() will be eventually
called by zfs_ioc_pool_create(). history_str_get() will not work as expected as
it fully relies on the function copyinstr() which is currently unimplemented.

After that, spa_create() will be called with history_str storing trash, thus
making the spa_history_log filled with completely wrong entries.

Let's fix this problem simply by implementing copyinstr. By the way, copystr
has basically the same semantics as copyinstr, so let's implement it as well.
It's not being used anywhere, but who knows? It might be needed in the future
by some bsd compliant application.

As far as I know, this problem didn't manifest *visibly* anywhere as the spa
history log wasn't intentionally used by us yet. I found the problem while
auditing code.

Signed-off-by: Raphael S. Carvalho <raphaelsc@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This reverts commit 97626d36. As pointed
out Tomek, "-u" flag is already passed in mkzfs.py.

mprotect() should fails with ENOMEM if it is called on non mapped
virtual address, but this check is done by mmu::ismapped().

Signed-off-by: Gleb Natapov <gleb@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Just came across this. Trivial to fix.

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Since there's no user data on the image, we can safely cache it, improving
build performance.  Helpful duing 'git bisect'.

Signed-off-by: Avi Kivity <avi@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Commit 649654af ("Fix race between join() and thread completion") fixed
the problem so bring back the test case.

Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Currently we use implementation from
bsd/sys/cddl/compat/opensolaris/kern/opensolaris_atomic.c
which is using global mutex.

This showes up on wait profile of ZFS writes:

(...)
   \-- 469.96 us (0.00%, #250) arc_do_user_evicts
       \-- 469.96 us (0.00%, #250) dbuf_do_evict
           |-- 355.93 us (0.00%, #154) dbuf_destroy
(...)
           |   \-- 76.60 us (0.00%, #53) zrl_add
           |                             atomic_cas_32
           |                             lockfree_mutex_lock
           |                             sched::thread::wait()
           |
           |-- 66.14 us (0.00%, #56) dbuf_clear
           |                         zrl_add
           |                         atomic_cas_32
           |                         lockfree_mutex_lock
           |                         sched::thread::wait()
           |
           \-- 47.89 us (0.00%, #40) dbuf_rele_and_unlock
                                     atomic_add_64_nv
                                     lockfree_mutex_lock
                                     sched::thread::wait()

This patch switches to amd64 implementation.

Signed-off-by: Tomasz Grabiec <tgrabiec@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

af_local.h declares a couple of functions implemented in af_local.cc.
There is no reason for pipe.cc to include it.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This patch adds the basic of memory tracking, and exposes an interface to for
that data to be collected.

We basically start with all stats at zero, and as we add memory to the System,
we bump it up and recalculate the watermarks (to avoid recomputing them all the
time). When a page range comes up, it will be added as free memory.

We operate based on what is currently sitting in the page ranges. This means
that we are effectively ignoring memory that sit in pools for memory usage. I
think it is a good assumption because it allow us to focus in the big picture,
and leave the pools to be used as liquid currency.

Signed-off-by: Glauber Costa <glommer@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This adds initial virtio-scsi support. We have no scsi layer in osv, in
this implementation virtio-scsi works directly with the bio layer. It
translates BIO_READ, BIO_WRITE and BIO_FLUSH to SCSI CMD.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

bio with BIO_SCSI flag contains a SCSI command.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

This can be used by the low-level driver to store private data, e.g.
virtio-scsi driver uses it to store a virtio-scsi request bound to this
bio.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

The lock used to protect _waiting_request_thread can go away.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

With net channels, poll() needs to wait not only on poll wakeups and the
timeout, but also requests from network interfaces to flush net channels
for polled sockets.

In preparation for that, switch from bsd msleep() to native wait_until().

Signed-off-by: Avi Kivity <avi@cloudius-systems.com>

While booting with many cpus (e.g., run.py -c 20, but I sometimes saw this
with as few as 7), we crashed while trying to sleep with preemption disabled:

Assertion failed: preemptable() (/home/nyh/osv/include/sched.hh: do_wait_until: 605)
Aborted

Turns out that since commit 223b2252 (half a year ago), main_cont()
ran smp_launch() with preemption disabled. But smp_launch creates threads,
and thread's constructor may sleep (e.g., on the thread_list_mutex), and
we cannot do this with preemption disabled...

So dropped the preempt_disable()/enable() from main_cont(). The reasoning for
it as expressed in 223b2252 appears to be no longer relevant.

Fixes #130.

Signed-off-by: Nadav Har'El <nyh@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Printing asynchronously is not affected by write() stalls and hence the period
for partial amount can be more or less constant.

Signed-off-by: Tomasz Grabiec <tgrabiec@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Reviewed-by: Dor Laor <dor@cloudius-systems.com>
Signed-off-by: Raphael S. Carvalho <raphaelsc@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>

We can skip to construct a vring::sg_node.

Signed-off-by: Asias He <asias@cloudius-systems.com>
Signed-off-by: Pekka Enberg <penberg@cloudius-systems.com>