qcow2: avoid extra flushes in qcow2

The problem with excessive flushing was found by a couple of performance
tests:
  - parallel directory tree creation (from 2 processes)
  - 32 cached writes + fsync at the end in a loop

For the first one results improved from 2.6 loops/sec to 3.5 loops/sec.
Each loop creates 10^3 directories with 10 files in each.

For the second one results improved from ~600 fsync/sec to ~1100
fsync/sec. Though, it was run on SSD so it probably won't show such
performance gain on rotational media.

qcow2_cache_flush() calls bdrv_flush() unconditionally after writing
cache entries of a particular cache. This can lead to as many as
2 additional fdatasyncs inside bdrv_flush.

We can simply skip all fdatasync calls inside qcow2_co_flush_to_os
as bdrv_flush for sure will do the job. These flushes are necessary to
keep the right order of writes to the different caches. Though this is
not necessary in the current code base as this ordering is ensured through
the flush in qcow2_cache_flush_dependency().

Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Pavel Borzenkov <pborzenkov@virtuozzo.com>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

raw-posix: Fetch max sectors for host block device

This is sometimes a useful value we should count in.

Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: assert that bs->request_alignment is a power of 2

at least bdrv_co_preadv/pwritev expect this.

Signed-off-by: Peter Lieven <pl@kamp.de>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

migration/block: Convert saving to BlockBackend

This creates a new BlockBackend for copying data from an images to the
migration stream on the source host. All I/O for block migration goes
through BlockBackend now.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

migration/block: Convert load to BlockBackend

This converts the loading part of block migration to use BlockBackend
interfaces rather than accessing the BlockDriverState directly.

Note that this takes a lazy shortcut. We should really use a separate
BlockBackend that is configured for the migration rather than for the
guest (e.g. writethrough caching is unnecessary) and holds its own
reference to the BlockDriverState, but the impact isn't that big and we
didn't have a separate migration reference before either, so it must be
good enough, I guess...

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

block: Kill bdrv_co_write_zeroes()

Now that all drivers have been converted to a byte interface,
we no longer need a sector interface.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

vmdk: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

raw_bsd: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

raw-posix: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
[ kwolf: Fixed up trace_paio_submit_co() call for qiov == NULL ]
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qed: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Kill an abuse of the comma operator while at it (fortunately,
the semantics were still right).  Also, the test for requests
not aligned to clusters should be applied always, not just
when a backing file is present.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

gluster: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

blkreplay: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

iscsi: Convert to bdrv_co_pwrite_zeroes()

Another step on our continuing quest to switch to byte-based
interfaces.

As this is the first byte-based iscsi interface, convert
is_request_lun_aligned() into two versions, one for sectors
and one for bytes.  Also, change from outright -EINVAL failure
on an unaligned request, to instead failing with -ENOTSUP to
trigger a read-modify-write fallback, particularly since the
block layer should be honoring bs->request_alignment to avoid
-EINVAL on read/write requests.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: Switch bdrv_write_zeroes() to byte interface

Rename to bdrv_pwrite_zeroes() to let the compiler ensure we
cater to the updated semantics.  Do the same for bdrv_co_write_zeroes().

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: Add .bdrv_co_pwrite_zeroes()

Update bdrv_co_do_write_zeroes() to be byte-based, and select
between the new byte-based bdrv_co_pwrite_zeroes() or the old
bdrv_co_write_zeroes().  The next patches will convert drivers,
then remove the old interface.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: Track write zero limits in bytes

Another step towards removing sector-based interfaces: convert
the maximum write and minimum alignment values from sectors to
bytes.  Rename the variables to let the compiler check that all
users are converted to the new semantics.

The maximum remains an int as long as BDRV_REQUEST_MAX_SECTORS
is constrained by INT_MAX (this means that we can't even
support a 2G write_zeroes, but just under it) - changing
operation lengths to unsigned or to 64-bits is a much bigger
audit, and debatable if we even want to do it (since at the
core, a 32-bit platform will still have ssize_t as its
underlying limit on write()).

Meanwhile, alignment is changed to 'uint32_t', since it makes no
sense to have an alignment larger than the maximum write, and
less painful to use an unsigned type with well-defined behavior
in bit operations than to have to worry about what happens if
a driver mistakenly supplies a negative alignment.

Add an assert that no one was trying to use sectors to get a
write zeroes larger than 2G, and therefore that a later conversion
to bytes won't be impacted by keeping the limit at 32 bits.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

iscsi: Use block size as minimum zero/discard alignment

If hardware does not advertise a minimum zero/discard
alignment, we still want to guarantee that the block layer
will align requests to our blocks, rather than the arbitrary
512-byte BDRV sector size.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: Catch more unaligned write_zero into zero cluster

is_zero_cluster() and is_zero_cluster_top_locked() are used only
by qcow2_co_write_zeroes().  The former is too broad (we don't
care if the sectors we are about to overwrite are non-zero, only
that all other sectors in the cluster are zero), so it needs to
be called up to twice but with smaller limits - rename it along
with adding the neeeded parameter.  The latter can be inlined for
more compact code.

The testsuite change shows that we now have a sparser top file
when an unaligned write_zeroes overwrites the only portion of
the backing file with data.

Based on a patch proposal by Denis V. Lunev.

CC: Denis V. Lunev <den@openvz.org>
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Denis V. Lunev <den@openvz.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qemu-iotests: Test one more spot for optimizing write_zeroes

Add another test to 154, showing that we currently allocate a
data cluster in the top layer if any sector of the backing file
was allocated.  The next patch will optimize this case.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: add tracepoints for qcow2_co_write_zeroes

This patch follows guidelines of all other tracepoints in qcow2, like ones
in qcow2_co_writev. I think that they should dump values in the same
quantities or be changed all together.

Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Eric Blake <eblake@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
Message-Id: <1463476543-3087-4-git-send-email-den@openvz.org>
[eblake: typo fix in commit message]
Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

qcow2: simplify logic in qcow2_co_write_zeroes

Unaligned requests will occupy only one cluster. This is true since the
previous commit. Simplify the code taking this consideration into
account.

In other words, the caller is now buggy if it ever passes us an unaligned
request that crosses cluster boundaries (the only requests that can cross
boundaries will be aligned).

There are no other changes so far.

Signed-off-by: Denis V. Lunev <den@openvz.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
CC: Eric Blake <eblake@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
Message-Id: <1463476543-3087-3-git-send-email-den@openvz.org>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

block: split write_zeroes always

We should split requests even if they are less than write_zeroes_alignment.
For example we can have the following request:
  offset 62k
  size   4k
  write_zeroes_alignment 64k
The original code sent 1 request covering 2 qcow2 clusters, and resulted
in both clusters being allocated. But by splitting the request, we can
cater to the case where one of the two clusters can be zeroed as a
whole, for only 1 cluster allocated after the operation.

Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Eric Blake <eblake@redhat.com>
CC: Kevin Wolf <kwolf@redhat.com>
Message-Id: <1463476543-3087-2-git-send-email-den@openvz.org>

[eblake: Avoid exceeding nb_sectors, hoist alignment checks out of
loop, and update testsuite to show that patch works]

Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

Merge remote-tracking branch 'remotes/mjt/tags/pull-trivial-patches-2016-06-07' into staging

trivial patches for 2016-06-07

# gpg: Signature made Tue 07 Jun 2016 16:20:52 BST
# gpg:                using RSA key 0xBEE59D74A4C3D7DB
# gpg: Good signature from "Michael Tokarev <mjt@tls.msk.ru>"
# gpg:                 aka "Michael Tokarev <mjt@corpit.ru>"
# gpg:                 aka "Michael Tokarev <mjt@debian.org>"

* remotes/mjt/tags/pull-trivial-patches-2016-06-07: (51 commits)
  hbitmap: Use DIV_ROUND_UP
  qemu-timer: Use DIV_ROUND_UP
  linux-user: Use DIV_ROUND_UP
  slirp: Use DIV_ROUND_UP
  usb: Use DIV_ROUND_UP
  rocker: Use DIV_ROUND_UP
  SPICE: Use DIV_ROUND_UP
  audio: Use DIV_ROUND_UP
  xen: Use DIV_ROUND_UP
  crypto: Use DIV_ROUND_UP
  block: Use DIV_ROUND_UP
  qed: Use DIV_ROUND_UP
  qcow/qcow2: Use DIV_ROUND_UP
  parallels: Use DIV_ROUND_UP
  coccinelle: use macro DIV_ROUND_UP instead of (((n) + (d) - 1) /(d))
  thunk: Rename args and fields in host-target bitmask conversion code
  thunk: Drop unused NO_THUNK_TYPE_SIZE guards
  qemu-common.h: Drop WORDS_ALIGNED define
  host-utils: Prefer 'false' for bool type
  docs/multi-thread-compression: Fix wrong command string
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

hbitmap: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qemu-timer: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

linux-user: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Riku Voipio <riku.voipio@iki.fi>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

slirp: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Samuel Thibault <samuel.thibault@ens-lyon.org>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Acked-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

usb: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

rocker: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

SPICE: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

audio: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

xen: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Stefano Stabellini <sstabellini@kernel.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

crypto: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

block: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: qemu-block@nongnu.org
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qed: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: qemu-block@nongnu.org
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qcow/qcow2: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: qemu-block@nongnu.org
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

parallels: Use DIV_ROUND_UP

Replace (((n) + (d) - 1) /(d)) by DIV_ROUND_UP(n,d).

This patch is the result of coccinelle script
scripts/coccinelle/round.cocci

CC: qemu-block@nongnu.org
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

coccinelle: use macro DIV_ROUND_UP instead of (((n) + (d) - 1) /(d))

sample from http://coccinellery.org/

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

thunk: Rename args and fields in host-target bitmask conversion code

The target_to_host_bitmask() and host_to_target_bitmask() functions
and the associated struct bitmask_transtbl are completely generic,
but for historical reasons the target related fields and parameters
are named 'x86' and the host related fields are named 'alpha'.
Rename them to 'target' and 'host'.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

thunk: Drop unused NO_THUNK_TYPE_SIZE guards

The thunk_type_size_array() and thunk_type_align_array() functions
are only provided if NO_THUNK_TYPE_SIZE is not defined. However
nothing in the codebase defines that, and so in fact these functions
are always present. Drop the unnecessary #ifdefs.

(Over a decade ago thunk.h used to be included by some softmmu
files, which defined NO_THUNK_TYPE_SIZE, but these includes are
long gone; see for instance commit f193c7979c2f7.)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qemu-common.h: Drop WORDS_ALIGNED define

The WORDS_ALIGNED #define is not used anywhere, and hasn't been since
2013 when commit 612d590ebc6cef rewrote the various ld<type>_<endian>_p
functions to not use it. Remove the #define and the comment describing it.
Also remove the line in the comment about TARGET_WORDS_ALIGNED, since
it has never actually existed.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

host-utils: Prefer 'false' for bool type

Mixing '0' and 'bool' looks stupid.

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

docs/multi-thread-compression: Fix wrong command string

s/info_migrate_capabilities/info migrate_capabilities

Signed-off-by: Wei Jiangang <weijg.fnst@cn.fujitsu.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Liang Li <liang.z.li@intel.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

all: Remove unnecessary glib.h includes

Remove glib.h includes, as it is provided by osdep.h.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qga: Remove unnecessary glib.h includes

Remove glib.h includes, as it is provided by osdep.h.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

tests: Remove unnecessary glib.h includes

Remove glib.h includes, as it is provided by osdep.h.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

clean-includes: Add glib.h to list of unneeded includes

osdep.h pulls in glib.h via glib-compat.h, so add it to the list of
includes that we remove. (This then means we must avoid running
clean-includes on glib-compat.h or it will delete the glib.h include.)

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Tested-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

pc: cleanup unused struct PcRomPciInfo

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

e1000: Removing unnecessary if statement

Since mit_delay can never be 0 this if statement is
superfluous.

Signed-off-by: Sameeh Jubran <sameeh@daynix.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

Fix configure test for PBKDF2 in nettle

On my Debian jessie system, including nettle/pbkdf2.h does not cause
NULL to be defined, which causes the test to fail to compile.  Include
stddef.h to bring in a definition of NULL.

Cc: qemu-trivial@nongnu.org
Cc: qemu-stable@nongnu.org
Signed-off-by: Steven Luo <steven+qemu@steven676.net>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

docs: Fix a couple of typos in throttle.txt

Signed-off-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

hw: Clean up includes

Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

replay: Clean up includes

Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.

This commit was created with scripts/clean-includes.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

fw_cfg: follow CODING_STYLE

Replace tab with 4 spaces; brace the indented statement.

Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Acked-by: Corey Minyard <cminyard@mvista.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qdev: Clean up around properties

include:
1. remove unnecessary declaration of static function
2. fix inconsistency between comment and function name, and typo OOM->QOM
2. update comments of functions, use uniform format(GTK-Doc style)

Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

monitor: Typo fix

s/partinal/partial/

Signed-off-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

ICH9: fix typo

Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

scripts: Use $(..) instead of deprecated `..`

This fixes these warnings from shellcheck:

    ^-- SC2006: Use $(..) instead of deprecated `..`

Update also a comment using the same pattern.

Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

configure: Use $(..) instead of deprecated `..`

This fixes these warnings from shellcheck:

    ^-- SC2006: Use $(..) instead of deprecated `..`

Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qemu-options.hx: Specify the units for -machine kvm_shadow_mem

The -machine kvm_shadow_mem option takes a size in bytes; say
so explicitly in its documentation.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reported-by: Tobi (github.com/tobimensch)
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

Fix linking relocatable objects on Sparc

On Sparc, gcc implicitly passes --relax to the linker, but -r is
incompatible with this. Therefore, if --no-relax is supported, it should
be passed to the linker.

Signed-off-by: James Clarke <jrtc27@jrtc27.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

ppc: Remove a potential overflow in muldiv64()

The coccinelle script:
scripts/coccinelle/overflow_muldiv64.cocci
gives us a list of potential overflows in muldiv64()
(the two first parameters are 64bit values).

This patch fixes one, as the fix seems obvious:

replace muldiv64(a, b, c) by muldiv64(b, a, c)
as "a" and "b" are 64bit values but a <= NANOSECONDS_PER_SECOND.
(10^9 -> 30bit value).

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

replace muldiv64(a, b, c) by (uint64_t)a * b / c

When "a" and "b" are 32bit values, we don't have to cast
them to 128bit, 64bit is enough.

This patch is the result of coccinelle script
scripts/coccinelle/simplify_muldiv64.cocci

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
For xtensa PIC:
Acked-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

remove useless muldiv64()

muldiv64(a, 1, b) is like "a / b".

This patch is the result of coccinelle script
scripts/coccinelle/remove_muldiv64.cocci.

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

The only 64bit parameter of muldiv64() is the first one.

muldiv64() is "uint64_t muldiv64(uint64_t a, uint32_t b, uint32_t c)"

Some time it is used as muldiv64(uint32_t a, uint64_t b, uint32_t c)"

This patch is the result of coccinelle script
scripts/coccinelle/swap_muldiv64.cocci to reorder arguments.

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Richard Henderson <rth@twiddle.net>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

scripts: add muldiv64() checking coccinelle scripts

Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

gdbstub: set listen backlog to 1

Avoid possible connection drops on Linux (when tcp_syncookies is
disabled) or fallbacks to SYN cookies with the following kernel warning:

    TCP: request_sock_TCP: Possible SYN flooding on port 1234. Sending cookies.  Check SNMP counters.

Since Linux 4.4 (ef547f2ac16b "tcp: remove max_qlen_log"), a backlog of
zero is really treated as the "queue length for completely established
sockets waiting to be accepted" (listen(2)). This is apparently a valid
interpretation of an "implementation-defined minimum value" for a
backlog value of 0 (listen(3p)). Previous kernels would use 8 as
minimum value, but that is no longer the case.

Signed-off-by: Peter Wu <peter@lekensteyn.nl>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

po/Makefile: call rm -f directly

Default variables are undefined in rules.mak and this is what the rest
of the build system uses.
Fixes make clean in ./po/

Signed-off-by: Jan Vesely <jano.vesely@gmail.com>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

target-moxie: Remove unused struct elements

Signed-off-by: Stefan Weil <sw@weilnetz.de>
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

fsdev: spelling fix

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

qga: spelling fix

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

docs: "specify" spell fix

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Eric Blake <eblake@redhat.com>

hw/ipmi: fix spelling

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Acked-by: Corey Minyard <cminyard@mvista.com>

s390x/virtio-ccw: fix spelling

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Cornelia Huck <cornelia.huck@de.ibm.com>

Merge remote-tracking branch 'remotes/stefanha/tags/block-pull-request' into staging

# gpg: Signature made Tue 07 Jun 2016 15:26:09 BST
# gpg:                using RSA key 0x9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>"
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>"

* remotes/stefanha/tags/block-pull-request:
  throttle: refuse iops-size without iops-total/read/write
  block: Drop bdrv_ioctl_bh_cb
  block: Move BlockRequest type to io.c
  block/io: optimize bdrv_co_pwritev for small requests
  iostatus: fix comments for block_job_iostatus_reset
  block/io: Remove unused bdrv_aio_write_zeroes()
  virtio: drop duplicate virtio_queue_get_id() function
  virtio-scsi: Remove op blocker for dataplane
  virtio-blk: Remove op blocker for dataplane
  blockdev-backup: Don't move target AioContext if it's attached
  blockdev-backup: Use bdrv_lookup_bs on target
  tests: avoid coroutine pool test crash

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

Merge remote-tracking branch 'remotes/mst/tags/for_upstream' into staging

pc, pci, virtio: new features, cleanups, fixes

This includes some infrastructure for ipmi smbios tables.
Beginning of acpi hotplug rework by Igor for supporting >255 CPUs.
Misc cleanups and fixes.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# gpg: Signature made Tue 07 Jun 2016 13:55:22 BST
# gpg:                using RSA key 0x281F0DB8D28D5469
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>"
# gpg:                 aka "Michael S. Tsirkin <mst@redhat.com>"

* remotes/mst/tags/for_upstream: (25 commits)
  virtio: move bi-endian target support to a single location
  pc-dimm: introduce realize callback
  pc-dimm: get memory region from ->get_memory_region()
  acpi: make bios_linker_loader_add_checksum() API offset based
  acpi: make bios_linker_loader_add_pointer() API offset based
  tpm: apci: cleanup TCPA table initialization
  acpi: cleanup bios_linker_loader_cleanup()
  acpi: simplify bios_linker API by removing redundant 'table' argument
  acpi: convert linker from GArray to BIOSLinker structure
  pc: use AcpiDeviceIfClass.send_event to issue GPE events
  acpi: extend ACPI interface to provide send_event hook
  pc: Postpone SMBIOS table installation to post machine init
  ipmi: rework the fwinfo to be fetched from the interface
  tests: acpi: update tables with consolidated legacy cpu-hotplug AML
  pc: acpi: cpuhp-legacy: switch ProcessorID to possible_cpus idx
  pc: acpi: simplify build_legacy_cpu_hotplug_aml() signature
  pc: acpi: consolidate legacy CPU hotplug in one file
  pc: acpi: mark current CPU hotplug functions as legacy
  pc: acpi: cpu-hotplug: make AML CPU_foo defines local to cpu_hotplug_acpi_table.c
  pc: acpi: consolidate \GPE._E02 with the rest of CPU hotplug AML
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

docker: Don't use eval trick on Makefile

The eval trick for defining DOCKER_SRC_COPY doesn't do anything
useful, as DOCKER_SRC_COPY is immediately expanded just after it
is defined, and CUR_TIME is already defined using ":=". Simply
define it using ":=" so it is evaluated only once.

The eval trick was also triggering an weird error on Travis builds:
  qemu/tests/docker/Makefile.include:34: *** unterminated variable reference.  Stop.

The issue is not easily reproducible (maybe it's a bug in some
versions of Make), but it is avoided if removing the eval trick.

Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

throttle: refuse iops-size without iops-total/read/write

In a similar vein to commit ee2bdc33c913b7d765baa5aa338c29fb30a05c9a
("throttle: refuse bps_max/iops_max without bps/iops") it is likely that
the user made a configuration error if iops-size has been set but no
iops limit has been set.

Print an error message so the user can check their throttling
configuration.  They should either remove iops-size if they don't want
any throttling or specify one of iops-total, iops-read, or iops-write.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Message-id: 1464828031-25601-1-git-send-email-stefanha@redhat.com

block: Drop bdrv_ioctl_bh_cb

Similar to the "!drv || !drv->bdrv_aio_ioctl" case above, here it is
okay to set co.ret and return. As pointed out by Paolo, a BH will be
created as necessary by the caller (bdrv_co_maybe_schedule_bh).
Besides, as pointed out by Kevin, "data" was leaked before.

Reported-by: Kevin Wolf <kwolf@redhat.com>
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20160601015223.19277-1-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block: Move BlockRequest type to io.c

I was thrown by the fact that the public type BlockRequest had
an anonymous union, but no obvious discriminator.  Turns out
that the only client of the second branch of the union was code
internal to io.c, now that commit 91c6e4b killed public
multiwrite, so move it into io.c and improve the comments.

Signed-off-by: Eric Blake <eblake@redhat.com>
Message-Id: <1463699150-19445-1-git-send-email-eblake@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>

block/io: optimize bdrv_co_pwritev for small requests

in a read-modify-write cycle a small request might cause
head and tail to fall into the same aligned block. Currently
QEMU reads the same block twice in this case which is
not necessary.

Signed-off-by: Peter Lieven <pl@kamp.de>
Message-id: 1464607873-28206-1-git-send-email-pl@kamp.de
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

iostatus: fix comments for block_job_iostatus_reset

Signed-off-by: Changlong Xie <xiecl.fnst@cn.fujitsu.com>
Message-id: 1464600491-23340-1-git-send-email-xiecl.fnst@cn.fujitsu.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

block/io: Remove unused bdrv_aio_write_zeroes()

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-id: 1464599852-15392-1-git-send-email-kwolf@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

virtio: drop duplicate virtio_queue_get_id() function

The virtio_queue_get_id() function is the lesser used duplicate of
virtio_get_queue_index().  Use the latter instead.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1463767461-17922-1-git-send-email-stefanha@redhat.com

virtio-scsi: Remove op blocker for dataplane

The previous patch dropped all op blockers from virtio-blk data plane.
The situation of virtio-scsi is exactly the same it can drop them too.

Signed-off-by: Fam Zheng <famz@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-id: 1463969978-24970-5-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

virtio-blk: Remove op blocker for dataplane

Block layer is prepared to unspecialize dataplane, an evidence is this
almost complete list of unblocked operations. It has all types except
two (actually three if DATAPLANE itself counts but blockdev.c makes sure
attaching twice is not possible): MIRROR_TARGET and BACKUP_TARGET.

blockdev-mirror refuses to start if target is attached, so the first is
not a problem.

By removing BACKUP_TARGET, blockdev-backup will become permissive to
write to a virtio-blk dataplane disk, but that is not worse than
non-dataplane given the latter is already possible. In either case,
blockdev.c always checks the target and source are on the same
AioContext, or bring them together if possible.

Signed-off-by: Fam Zheng <famz@redhat.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Message-id: 1463969978-24970-4-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

blockdev-backup: Don't move target AioContext if it's attached

If the BDS is attached, it will want to stay on the AioContext where its
BlockBackend is. Don't call bdrv_set_aio_context in this case.

Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1463969978-24970-3-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

blockdev-backup: Use bdrv_lookup_bs on target

This allows backing up to a BDS that has not been attached to any BB.

Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1463969978-24970-2-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

tests: avoid coroutine pool test crash

Skip the test_co_queue test case if the coroutine pool is not enabled.
The test case does not work without the pool because it touches memory
belonging to a freed coroutine (on purpose).

Reported-by: Eduardo Habkost <ehabkost@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 1463767231-13379-1-git-send-email-stefanha@redhat.com

virtio: move bi-endian target support to a single location

Paolo's recent cpu.h cleanups broke legacy virtio for ppc64 LE guests (and
arm BE guests as well, even if I have not verified that). Especially, commit
"33c11879fd42 qemu-common: push cpu.h inclusion out of qemu-common.h" has
the side-effect of silently hiding the TARGET_IS_BIENDIAN macro from the
virtio memory accessors, and thus fully disabling support of endian changing
targets.

To be sure this cannot happen again, let's gather all the bi-endian bits
where they belong in include/hw/virtio/virtio-access.h.

The changes in hw/virtio/vhost.c are safe because vhost_needs_vring_endian()
is not called on a hot path and non bi-endian targets will return false
anyway.

While here, also rename TARGET_IS_BIENDIAN to be more precise: it is only for
legacy virtio and bi-endian guests.

Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Cédric Le Goater <clg@kaod.org>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

pc-dimm: introduce realize callback

nvdimm needs to  check if the backend memory is large enough to contain
label data and init its memory region when the device is realized, so
introduce realize callback which is called after common dimm has been
realize

Signed-off-by: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

pc-dimm: get memory region from ->get_memory_region()

Curretly, the memory region of backed memory is all directly
mapped to guest's address space, however, it will be not true
for nvdimm device if we introduce nvdimm label which only can
be indirectly accessed by ACPI DSM method

Also it improves the comments a bit to reflect this fact

Signed-off-by: Xiao Guangrong <guangrong.xiao@linux.intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>

acpi: make bios_linker_loader_add_checksum() API offset based

It should help to make clear that bios_linker works in terms
of offsets within a file. Also it should prevent mistakes
where user passes as arguments pointers to unrelated to file blobs.

While at it, considering that it's a ACPI checksum and
it's initial value must be 0, move checksum field zeroing
into bios_linker_loader_add_checksum() instead of doing it
at every call site manually before bios_linker_loader_add_checksum()
is called.

In addition add extra boundary checks.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: make bios_linker_loader_add_pointer() API offset based

cleanup bios_linker_loader_add_pointer() API by switching
arguments to taking offsets relative to corresponding files
instead of doing pointer arithmetic on behalf of user which
were confusing.

Also make offset inside of source file explicit in API
so that user won't have to manually set it in
destination file blob and while at it add additional
boundary checks.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

tpm: apci: cleanup TCPA table initialization

At the time build_tpm_tcpa() is called the tcpalog size is
always 0, so log_area_start_address which is actually offset
from the start of ACPI_BUILD_TPMLOG_FILE is always 0.

Also as 'TCPA' is allocated 0 filled, there is no point
in calculating always 0 log_area_start_address and set
tcpa->log_area_start_address to it since the field should
always point to start of ACPI_BUILD_TPMLOG_FILE.
Make code easier to read dropping not needed offset
calculations.
While at that move tcpalog allocation closer to the code
that defines its size.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: cleanup bios_linker_loader_cleanup()

bios_linker_loader_cleanup() is called only from one place
and returned value is immediately freed wich makes returning
pointer from bios_linker_loader_cleanup() useless.

Cleanup bios_linker_loader_cleanup() by freeing
data there so that caller won't have to free it.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: simplify bios_linker API by removing redundant 'table' argument

'table' argument in bios_linker_add_foo() commands is
a data blob of one of files also passed to the same API.
So instead of passing blob in every API call, add and keep
file name association with related blob at bios_linker_loader_alloc()
time.

And find blob by name looking up allocated file entries
inside of bios_linker_add_foo() commands.

It will:
 - make API less confusing,
 - enforce calling bios_linker_loader_alloc() before
   calling any bios_linker_add_foo()
 - make sure that blob is the correct one, i.e.
   associated with the right file name

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

acpi: convert linker from GArray to BIOSLinker structure

Patch just changes type of of linker variables to
a structure, there aren't any functional changes.

Converting linker to a structure will allow to extend
it functionality in follow up patch adding sanity blob
checks.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>

pc: use AcpiDeviceIfClass.send_event to issue GPE events

it reduces number of args passed in handlers by 1 and
a number of used proxy wrappers saving ~20LOC.
Also it allows to make cpu/mem hotplug code more
universal as it would allow ARM to reuse it without
rewrite by providing its own send_event callback
to trigger events usiong GPIO instead of GPE
as fixed hadrware ACPI model doen't have GPE at all.

Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>