OSDN Git Service

qmiga/qemu.git
3 years agovhost-user-blk: Fail gracefully on too large queue size
Kevin Wolf [Tue, 13 Apr 2021 16:56:54 +0000 (18:56 +0200)]
vhost-user-blk: Fail gracefully on too large queue size

virtio_add_queue() aborts when queue_size > VIRTQUEUE_MAX_SIZE, so
vhost_user_blk_device_realize() should check this before calling it.

Simple reproducer:

qemu-system-x86_64 \
    -chardev null,id=foo \
    -device vhost-user-blk-pci,queue-size=4096,chardev=foo

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1935014
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210413165654.50810-1-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoqemu-img convert: Unshare write permission for source
Kevin Wolf [Thu, 22 Apr 2021 16:43:44 +0000 (18:43 +0200)]
qemu-img convert: Unshare write permission for source

For a successful conversion of an image, we must make sure that its
content doesn't change during the conversion.

A special case of this is using the same image file both as the source
and as the destination. If both input and output format are raw, the
operation would just be useless work, with other formats it is a sure
way to destroy the image. This will now fail because the image file
can't be opened a second time for the output when opening it for the
input has already acquired file locks to unshare BLK_PERM_WRITE.

Nevertheless, if there is some reason in a special case why it is
actually okay to allow writes to the image while it is being converted,
-U can still be used to force sharing all permissions.

Note that for most image formats, BLK_PERM_WRITE would already be
unshared by the format driver, so this only really makes a difference
for raw source images (but any output format).

Reported-by: Xueqiang Wei <xuwei@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210422164344.283389-3-kwolf@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: Add BDRV_O_NO_SHARE for blk_new_open()
Kevin Wolf [Thu, 22 Apr 2021 16:43:43 +0000 (18:43 +0200)]
block: Add BDRV_O_NO_SHARE for blk_new_open()

Normally, blk_new_open() just shares all permissions. This was fine
originally when permissions only protected against uses in the same
process because no other part of the code would actually get to access
the block nodes opened with blk_new_open(). However, since we use it for
file locking now, unsharing permissions becomes desirable.

Add a new BDRV_O_NO_SHARE flag that is used in blk_new_open() to unshare
any permissions that can be unshared.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210422164344.283389-2-kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: refactor bdrv_node_check_perm()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:18:04 +0000 (18:18 +0300)]
block: refactor bdrv_node_check_perm()

Now, bdrv_node_check_perm() is called only with fresh cumulative
permissions, so its actually "refresh_perm".

Move permission calculation to the function. Also, drop unreachable
error message and rewrite the remaining one to be more generic (as now
we don't know which node is added and which was already here).

Add also Virtuozzo copyright, as big work is done at this point.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-37-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: rename bdrv_replace_child_safe() to bdrv_replace_child()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:18:03 +0000 (18:18 +0300)]
block: rename bdrv_replace_child_safe() to bdrv_replace_child()

We don't have bdrv_replace_child(), so it's time for
bdrv_replace_child_safe() to take its place.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-36-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: refactor bdrv_child_set_perm_safe() transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:18:02 +0000 (18:18 +0300)]
block: refactor bdrv_child_set_perm_safe() transaction action

Old interfaces dropped, nobody directly calls
bdrv_child_set_perm_abort() and bdrv_child_set_perm_commit(), so we can
use personal state structure for the action and stop exploiting
BdrvChild structure. Also, drop "_safe" suffix which is redundant now.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-35-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: inline bdrv_replace_child()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:18:01 +0000 (18:18 +0300)]
block: inline bdrv_replace_child()

bdrv_replace_child() has only one caller, the second argument is
unused. Inline it now. This triggers deletion of some more unused
interfaces.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-34-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: inline bdrv_check_perm_common()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:18:00 +0000 (18:18 +0300)]
block: inline bdrv_check_perm_common()

bdrv_check_perm_common() has only one caller, so no more sense in
"common".

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-33-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: drop unused permission update functions
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:59 +0000 (18:17 +0300)]
block: drop unused permission update functions

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-32-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: bdrv_reopen_multiple: refresh permissions on updated graph
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:58 +0000 (18:17 +0300)]
block: bdrv_reopen_multiple: refresh permissions on updated graph

Move bdrv_reopen_multiple to new paradigm of permission update:
first update graph relations, then do refresh the permissions.

We have to modify reopen process in file-posix driver: with new scheme
we don't have prepared permissions in raw_reopen_prepare(), so we
should reconfigure fd in raw_check_perm(). Still this seems more native
and simple anyway.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-31-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: bdrv_reopen_multiple(): move bdrv_flush to separate pre-prepare
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:57 +0000 (18:17 +0300)]
block: bdrv_reopen_multiple(): move bdrv_flush to separate pre-prepare

During reopen we may add backing bs from other aio context, which may
lead to changing original context of top bs.

We are going to move graph modification to prepare stage. So, it will
be possible that bdrv_flush() in bdrv_reopen_prepare called on bs in
non-original aio context, which we didn't aquire which leads to crash.

To avoid this problem move bdrv_flush() to be a separate reopen stage
before bdrv_reopen_prepare().

This doesn't seem correct to acquire only one aio context and not all
contexts participating in reopen. But it's not obvious how to do it
correctly, keeping in mind:

 1. rules of bdrv_set_aio_context_ignore() that requires new_context
    lock not being held

 2. possible deadlocks because of holding all (or several?) AioContext
    locks

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-30-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_set_backing_noperm() transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:56 +0000 (18:17 +0300)]
block: add bdrv_set_backing_noperm() transaction action

Split out no-perm part of bdrv_set_backing_hd() as a separate
transaction action. Note the in case of existing BdrvChild we reuse it,
not recreate, just to do less actions.

We don't need to create extra reference to backing_hd as we don't lose
it in bdrv_attach_child().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-29-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: make bdrv_refresh_limits() to be a transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:55 +0000 (18:17 +0300)]
block: make bdrv_refresh_limits() to be a transaction action

To be used in further commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-28-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: make bdrv_unset_inherits_from to be a transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:54 +0000 (18:17 +0300)]
block: make bdrv_unset_inherits_from to be a transaction action

To be used in the further commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-27-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: drop ignore_children for permission update functions
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:53 +0000 (18:17 +0300)]
block: drop ignore_children for permission update functions

This argument is always NULL. Drop it.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-26-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock/backup-top: drop .active
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:52 +0000 (18:17 +0300)]
block/backup-top: drop .active

We don't need this workaround anymore: bdrv_append is already smart
enough and we can use new bdrv_drop_filter().

This commit efficiently reverts also recent 705dde27c6c53b73, which
checked .active on io path. Still it said that the problem should be
theoretical. And the logic of filter removement is changed anyway.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-25-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: introduce bdrv_drop_filter()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:51 +0000 (18:17 +0300)]
block: introduce bdrv_drop_filter()

Using bdrv_replace_node() for removing filter is not good enough: it
keeps child reference of the filter, which may conflict with original
top node during permission update.

Instead let's create new interface, which will do all graph
modifications first and then update permissions.

Let's modify bdrv_replace_node_common(), allowing it additionally drop
backing chain child link pointing to new node. This is quite
appropriate for bdrv_drop_intermediate() and makes possible to add
new bdrv_drop_filter() as a simple wrapper.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-24-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_remove_filter_or_cow transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:50 +0000 (18:17 +0300)]
block: add bdrv_remove_filter_or_cow transaction action

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-23-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: adapt bdrv_append() for inserting filters
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:49 +0000 (18:17 +0300)]
block: adapt bdrv_append() for inserting filters

bdrv_append is not very good for inserting filters: it does extra
permission update as part of bdrv_set_backing_hd(). During this update
filter may conflict with other parents of top_bs.

Instead, let's first do all graph modifications and after it update
permissions.

append-greedy-filter test-case in test-bdrv-graph-mod is now works, so
move it out of debug option.

Note: bdrv_append() is still only works for backing-child based
filters. It's something to improve later.

Note2: we use the fact that bdrv_append() is used to append new nodes,
without backing child, so we don't need frozen check and inherits_from
logic from bdrv_set_backing_hd().

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-22-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: split out bdrv_replace_node_noperm()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:48 +0000 (18:17 +0300)]
block: split out bdrv_replace_node_noperm()

Split part of bdrv_replace_node_common() to be used separately.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-21-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_attach_child_noperm() transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:47 +0000 (18:17 +0300)]
block: add bdrv_attach_child_noperm() transaction action

Split no-perm part of bdrv_attach_child as separate transaction action.
It will be used in later commits.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-20-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_attach_child_common() transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:46 +0000 (18:17 +0300)]
block: add bdrv_attach_child_common() transaction action

Split out no-perm part of bdrv_root_attach_child() into separate
transaction action. bdrv_root_attach_child() now moves to new
permission update paradigm: first update graph relations then update
permissions.

qsd-jobs test output updated. Seems now permission update goes in
another order. Still, the test comment say that we only want to check
that command doesn't crash, and it's still so.

Error message is a bit misleading as it looks like job was added first.
But actually in new paradigm of graph update we can't distinguish such
things. We should update the error message, but let's not do it now.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210428151804.439460-19-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: fix bdrv_replace_node_common
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:45 +0000 (18:17 +0300)]
block: fix bdrv_replace_node_common

inore_children thing doesn't help to track all propagated permissions
of children we want to ignore. The simplest way to correctly update
permissions is update graph first and then do permission update. In
this case we just referesh permissions for the whole subgraph (in
topological-sort defined order) and everything is correctly calculated
automatically without any ignore_children.

So, refactor bdrv_replace_node_common to first do graph update and then
refresh the permissions.

Test test_parallel_exclusive_write() now pass, so move it out of
debugging "if".

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-18-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_replace_child_safe() transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:44 +0000 (18:17 +0300)]
block: add bdrv_replace_child_safe() transaction action

To be used in the following commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-17-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_list_* permission update functions
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:43 +0000 (18:17 +0300)]
block: add bdrv_list_* permission update functions

Add new interface, allowing use of existing node list. It will be used
to fix bdrv_replace_node() in the further commit.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-16-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: add bdrv_drv_set_perm transaction action
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:42 +0000 (18:17 +0300)]
block: add bdrv_drv_set_perm transaction action

Refactor calling driver callbacks to a separate transaction action to
be used later.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-15-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: use topological sort for permission update
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:41 +0000 (18:17 +0300)]
block: use topological sort for permission update

Rewrite bdrv_check_perm(), bdrv_abort_perm_update() and bdrv_set_perm()
to update nodes in topological sort order instead of simple DFS. With
topologically sorted nodes, we update a node only when all its parents
already updated. With DFS it's not so.

Consider the following example:

    A -+
    |  |
    |  v
    |  B
    |  |
    v  |
    C<-+

A is parent for B and C, B is parent for C.

Obviously, to update permissions, we should go in order A B C, so, when
we update C, all parent permissions already updated. But with current
approach (simple recursion) we can update in sequence A C B C (C is
updated twice). On first update of C, we consider old B permissions, so
doing wrong thing. If it succeed, all is OK, on second C update we will
finish with correct graph. But if the wrong thing failed, we break the
whole process for no reason (it's possible that updated B permission
will be less strict, but we will never check it).

Also new approach gives a way to simultaneously and correctly update
several nodes, we just need to run bdrv_topological_dfs() several times
to add all nodes and their subtrees into one topologically sorted list
(next patch will update bdrv_replace_node() in this manner).

Test test_parallel_perm_update() is now passing, so move it out of
debugging "if".

We also need to support ignore_children in
bdrv_parent_perms_conflict()

For test 283 order of conflicting parents check is changed.

Note also that in bdrv_check_perm() we don't check for parents conflict
at root bs, as we may be in the middle of permission update in
bdrv_reopen_multiple(). bdrv_reopen_multiple() will be updated soon.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-14-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: inline bdrv_child_*() permission functions calls
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:40 +0000 (18:17 +0300)]
block: inline bdrv_child_*() permission functions calls

Each of them has only one caller. Open-coding simplifies further
pemission-update system changes.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-13-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: rewrite bdrv_child_try_set_perm() using bdrv_refresh_perms()
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:39 +0000 (18:17 +0300)]
block: rewrite bdrv_child_try_set_perm() using bdrv_refresh_perms()

We are going to drop recursive bdrv_child_* functions, so stop use them
in bdrv_child_try_set_perm() as a first step.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-12-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: refactor bdrv_child* permission functions
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:38 +0000 (18:17 +0300)]
block: refactor bdrv_child* permission functions

Split out non-recursive parts, and refactor as block graph transaction
action.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-11-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: bdrv_refresh_perms: check for parents permissions conflict
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:37 +0000 (18:17 +0300)]
block: bdrv_refresh_perms: check for parents permissions conflict

Add additional check that node parents do not interfere with each
other. This should not hurt existing callers and allows in further
patch use bdrv_refresh_perms() to update a subtree of changed
BdrvChild (check that change is correct).

New check will substitute bdrv_check_update_perm() in following
permissions refactoring, so keep error messages the same to avoid
unit test result changes.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-10-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoutil: add transactions.c
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:36 +0000 (18:17 +0300)]
util: add transactions.c

Add simple transaction API to use in further update of block graph
operations.

Supposed usage is:

- "prepare" is main function of the action and it should make the main
  effect of the action to be visible for the following actions, keeping
  possibility of roll-back, saving necessary things in action state,
  which is prepended to the action list (to do that, prepare func
  should call tran_add()). So, driver struct doesn't include "prepare"
  field, as it is supposed to be called directly.

- commit/rollback is supposed to be called for the list of action
  states, to commit/rollback all the actions in reverse order

- When possible "commit" should not make visible effect for other
  actions, which make possible transparent logical interaction between
  actions.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-9-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: make bdrv_reopen_{prepare,commit,abort} private
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:35 +0000 (18:17 +0300)]
block: make bdrv_reopen_{prepare,commit,abort} private

These functions are called only from bdrv_reopen_multiple() in block.c.
No reason to publish them.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-8-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: drop ctx argument from bdrv_root_attach_child
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:34 +0000 (18:17 +0300)]
block: drop ctx argument from bdrv_root_attach_child

Passing parent aio context is redundant, as child_class and parent
opaque pointer are enough to retrieve it. Drop the argument and use new
bdrv_child_get_parent_aio_context() interface.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-7-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: BdrvChildClass: add .get_parent_aio_context handler
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:33 +0000 (18:17 +0300)]
block: BdrvChildClass: add .get_parent_aio_context handler

Add new handler to get aio context and implement it in all child
classes. Add corresponding public interface to be used soon.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Alberto Garcia <berto@igalia.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-6-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoblock: bdrv_append(): don't consume reference
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:32 +0000 (18:17 +0300)]
block: bdrv_append(): don't consume reference

We have too much comments for this feature. It seems better just don't
do it. Most of real users (tests don't count) have to create additional
reference.

Drop also comment in external_snapshot_prepare:
 - bdrv_append doesn't "remove" old bs in common sense, it sounds
   strange
 - the fact that bdrv_append can fail is obvious from the context
 - the fact that we must rollback all changes in transaction abort is
   known (it's the direct role of abort)

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-5-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agotests/test-bdrv-graph-mod: add test_append_greedy_filter
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:31 +0000 (18:17 +0300)]
tests/test-bdrv-graph-mod: add test_append_greedy_filter

bdrv_append() is not quite good for inserting filters: it does extra
permission update in intermediate state, where filter get it filtered
child but is not yet replace it in a backing chain.

Some filters (for example backup-top) may want permissions even when
have no parents. And described intermediate state becomes invalid.

That's (half a) reason, why we need "inactive" state for backup-top
filter.

bdrv_append() will be improved later, now let's add a unit test.

Now test fails, so it runs only with -d flag. To run do

  ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/append-greedy-filter

from <build-directory>/tests.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-4-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agotests/test-bdrv-graph-mod: add test_parallel_perm_update
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:30 +0000 (18:17 +0300)]
tests/test-bdrv-graph-mod: add test_parallel_perm_update

Add test to show that simple DFS recursion order is not correct for
permission update. Correct order is topological-sort order, which will
be introduced later.

Consider the block driver which has two filter children: one active
with exclusive write access and one inactive with no specific
permissions.

And, these two children has a common base child, like this:

┌─────┐     ┌──────┐
│ fl2 │ ◀── │ top  │
└─────┘     └──────┘
  │           │
  │           │ w
  │           ▼
  │         ┌──────┐
  │         │ fl1  │
  │         └──────┘
  │           │
  │           │ w
  │           ▼
  │         ┌──────┐
  └───────▶ │ base │
            └──────┘

So, exclusive write is propagated.

Assume, we want to make fl2 active instead of fl1.
So, we set some option for top driver and do permission update.

If permission update (remember, it's DFS) goes first through
top->fl1->base branch it will succeed: it firstly drop exclusive write
permissions and than apply them for another BdrvChildren.
But if permission update goes first through top->fl2->base branch it
will fail, as when we try to update fl2->base child, old not yet
updated fl1->base child will be in conflict.

Now test fails, so it runs only with -d flag. To run do

  ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/parallel-perm-update

from <build-directory>/tests.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-3-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agotests/test-bdrv-graph-mod: add test_parallel_exclusive_write
Vladimir Sementsov-Ogievskiy [Wed, 28 Apr 2021 15:17:29 +0000 (18:17 +0300)]
tests/test-bdrv-graph-mod: add test_parallel_exclusive_write

Add the test that shows that concept of ignore_children is incomplete.
Actually, when we want to update something, ignoring permission of some
existing BdrvChild, we should ignore also the propagated effect of this
child to the other children. But that's not done. Better approach
(update permissions on already updated graph) will be implemented
later.

Now the test fails, so it's added with -d argument to not break make
check.

Test fails with

 "Conflicts with use by fl1 as 'backing', which does not allow 'write' on base"

because when updating permissions we can ignore original top->fl1
BdrvChild. But we don't ignore exclusive write permission in fl1->base
BdrvChild, which is propagated. Correct thing to do is make graph
change first and then do permission update from the top node.

To run test do

  ./test-bdrv-graph-mod -d -p /bdrv-graph-mod/parallel-exclusive-write

from <build-directory>/tests.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20210428151804.439460-2-vsementsov@virtuozzo.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
3 years agoOpen 6.1 development tree
Peter Maydell [Fri, 30 Apr 2021 10:15:40 +0000 (11:15 +0100)]
Open 6.1 development tree

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoUpdate version for v6.0.0 release
Peter Maydell [Thu, 29 Apr 2021 17:05:29 +0000 (18:05 +0100)]
Update version for v6.0.0 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoUpdate version for v6.0.0-rc5 release
Peter Maydell [Mon, 26 Apr 2021 15:24:54 +0000 (16:24 +0100)]
Update version for v6.0.0-rc5 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agohw/block/nvme: fix invalid msix exclusive uninit
Klaus Jensen [Fri, 23 Apr 2021 05:21:26 +0000 (07:21 +0200)]
hw/block/nvme: fix invalid msix exclusive uninit

Commit 1901b4967c3f changed the nvme device from using a bar exclusive
for MSI-x to sharing it on bar0.

Unfortunately, the msix_uninit_exclusive_bar() call remains in
nvme_exit() which causes havoc when the device is removed with, say,
device_del. Fix this.

Additionally, a subregion is added but it is not removed on exit which
causes a reference to linger and the drive to never be unlocked.

Fixes: 1901b4967c3f ("hw/block/nvme: move msix table and pba to BAR 0")
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotarget/s390x: fix s390_probe_access to check PAGE_WRITE_ORG for writeability
Alex Bennée [Thu, 22 Apr 2021 15:44:27 +0000 (16:44 +0100)]
target/s390x: fix s390_probe_access to check PAGE_WRITE_ORG for writeability

We can remove PAGE_WRITE when (internally) marking a page read-only
because it contains translated code. This can get confused when we are
executing signal return code on signal stacks.

Fixes: e56552cf07 ("target/s390x: Implement the MVPG condition-code-option bit")
Found-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: David Hildenbrand <david@redhat.com>
Message-id: 20210422154427.13038-1-alex.bennee@linaro.org
Cc: Cornelia Huck <cohuck@redhat.com>
Cc: Thomas Huth <thuth@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Laurent Vivier <laurent@vivier.eu>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agonet: check the existence of peer before trying to pad
Jason Wang [Fri, 23 Apr 2021 03:18:03 +0000 (11:18 +0800)]
net: check the existence of peer before trying to pad

There could be case that peer is NULL. This can happen when during
network device hot-add where net device needs to be added first. So
the patch check the existence of peer before trying to do the pad.

Fixes: 969e50b61a285 ("net: Pad short frames to minimum size before sending from SLiRP/TAP")
Signed-off-by: Jason Wang <jasowang@redhat.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Stefan Weil <sw@weilnetz.de>
Message-id: 20210423031803.1479-1-jasowang@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoUpdate version for v6.0.0-rc4 release
Peter Maydell [Tue, 20 Apr 2021 15:30:01 +0000 (16:30 +0100)]
Update version for v6.0.0-rc4 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoqga: fix guest-get-disks regression
Marc-André Lureau [Tue, 20 Apr 2021 12:58:31 +0000 (16:58 +0400)]
qga: fix guest-get-disks regression

Commit 54aa3de72 ("qapi: Use QAPI_LIST_PREPEND() where possible")
inadvertently removed the has_dependencies from the partition disk
info, resulting in empty list being returned.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1950833

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-id: 20210420125831.233092-1-marcandre.lureau@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotarget/mips/rel6_translate: Change license to GNU LGPL v2.1 (or later)
Philippe Mathieu-Daudé [Tue, 20 Apr 2021 10:06:32 +0000 (12:06 +0200)]
target/mips/rel6_translate: Change license to GNU LGPL v2.1 (or later)

When adding this file and its new content in commit 3f7a927847a
("target/mips: LSA/DLSA R6 decodetree helpers") I did 2 mistakes:

1: Listed authors who haven't been involved in its development,
2: Used an incorrect GNU GPLv2 license text (using 'and' instead
   of 'or').

Instead of correcting the GNU GPLv2 license text, replace the license
by the 'GNU LGPL v2.1 or later' one, to be coherent with the other
translation files in the target/mips/ folder.

Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20210420100633.1752440-1-f4bug@amsat.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agomigration: Deprecate redundant query-migrate result @blocked
Markus Armbruster [Tue, 20 Apr 2021 05:19:06 +0000 (07:19 +0200)]
migration: Deprecate redundant query-migrate result @blocked

Result @blocked is true when and only when result @blocked-reasons is
present.  It's always non-empty when present.  @blocked is redundant.
It was introduced in commit 3af8554bd0 "migration: Add blocker
information", and has not been released.  This gives us a chance to
fix the interface with minimal fuss.

Unfortunately, we're already too close to the release to risk dropping
it.  Deprecate it instead.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-id: 20210420051907.891470-1-armbru@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agomptsas: Remove unused MPTSASState 'pending' field (CVE-2021-3392)
Michael Tokarev [Mon, 19 Apr 2021 13:42:47 +0000 (15:42 +0200)]
mptsas: Remove unused MPTSASState 'pending' field (CVE-2021-3392)

While processing SCSI i/o requests in mptsas_process_scsi_io_request(),
the Megaraid emulator appends new MPTSASRequest object 'req' to
the 's->pending' queue. In case of an error, this same object gets
dequeued in mptsas_free_request() only if SCSIRequest object
'req->sreq' is initialised. This may lead to a use-after-free issue.

Since s->pending is actually not used, simply remove it from
MPTSASState.

Cc: qemu-stable@nongnu.org
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reported-by: Cheolwoo Myung <cwmyung@snu.ac.kr>
Message-id: 20210419134247.1467982-1-f4bug@amsat.org
Message-Id: <20210416102243.1293871-1-mjt@msgid.tls.msk.ru>
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Reported-by: Cheolwoo Myung <cwmyung@snu.ac.kr>
BugLink: https://bugs.launchpad.net/qemu/+bug/1914236
Fixes: e351b826112 ("hw: Add support for LSI SAS1068 (mptsas) device")
[PMD: Reworded description, added more tags]
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210417' into...
Peter Maydell [Sat, 17 Apr 2021 19:47:32 +0000 (20:47 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210417' into staging

Fixes for rc4:
 * Fix compile failures of C++ files with new glib headers
 * mps3-an547: Use correct Cortex-M55 CPU and don't disable its FPU
 * accel/tcg: Fix assertion failure executing from non-RAM with -icount

# gpg: Signature made Sat 17 Apr 2021 20:39:58 BST
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20210417:
  accel/tcg: avoid re-translating one-shot instructions
  target/arm: drop CF_LAST_IO/dc->condjump check
  hw/arm/armsse: Make SSE-300 use Cortex-M55
  hw/arm/armsse: Give SSE-300 its own Property array
  include/qemu/osdep.h: Move system includes to top
  osdep: protect qemu/osdep.h with extern "C"
  osdep: include glib-compat.h before other QEMU headers

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoaccel/tcg: avoid re-translating one-shot instructions
Alex Bennée [Thu, 15 Apr 2021 16:24:53 +0000 (17:24 +0100)]
accel/tcg: avoid re-translating one-shot instructions

By definition a single instruction is capable of being an IO
instruction. This avoids a problem of triggering a cpu_io_recompile on
a non-recorded translation which then fails because it expects
tcg_tb_lookup() to succeed unconditionally. The normal use case
requires a TB to be able to resolve machine state.

The other users of tcg_tb_lookup() are able to tolerate a missing TB
if the machine state has been resolved by other means - which in the
single-shot case is always true because machine state is synced at the
start of a block.

Reported-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210415162454.22056-1-alex.bennee@linaro.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotarget/arm: drop CF_LAST_IO/dc->condjump check
Alex Bennée [Fri, 16 Apr 2021 17:02:07 +0000 (18:02 +0100)]
target/arm: drop CF_LAST_IO/dc->condjump check

This is a left over erroneous check from the days front-ends handled
io start/end themselves. Regardless just because IO could be performed
on the last instruction doesn't obligate the front end to do so.

This fixes an abort faced by the aspeed execute-in-place support which
will necessarily trigger this state (even before the one-shot
CF_LAST_IO fix). The test still seems to hang once it attempts to boot
the Linux kernel but I suspect this is an unrelated issue with icount
and the timer handling code.

The original intention of the cpu_abort (added in commit 2e70f6efa8b9
when the icount stuff was first added) seems to have been to act as
an assert() to catch an unhandled corner case where the generated code
would be something like:
    conditional branch to condlabel if its cc failed
    implementation of the insn (a conditional branch or trap)
    code emitted by gen_io_end()
 condlabel:
    gen_goto_tb or equivalent thing to go to next insn

At runtime the cc-failed case would skip over the code emitted by
gen_io_end(), leaving the can_do_io flag incorrectly set.

In commit ba3e7926691ed33 we switched to an implementation which
always clears can_do_io at the start of the following TB instead
of trying to clear it at the end of a TB that did IO. So the corner
case that this cpu_abort() was trying to flag is no longer possible,
because the gen_io_end() call has been deleted. We can therefore
safely remove the no-longer-valid assertion.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20210416170207.12504-1-alex.bennee@linaro.org
Cc: Cédric Le Goater <clg@kaod.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agohw/arm/armsse: Make SSE-300 use Cortex-M55
Peter Maydell [Fri, 16 Apr 2021 10:40:10 +0000 (11:40 +0100)]
hw/arm/armsse: Make SSE-300 use Cortex-M55

The SSE-300 has a Cortex-M55 (which was the whole reason for us
modelling it), but we forgot to actually update the code to let it
have a different CPU type from the IoTKit and SSE-200.  Add CPU type
as a field for ARMSSEInfo instead of hardcoding it to always use a
Cortex-M33.

Buglink: https://bugs.launchpad.net/qemu/+bug/1923861
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210416104010.13228-1-peter.maydell@linaro.org

3 years agohw/arm/armsse: Give SSE-300 its own Property array
Peter Maydell [Thu, 15 Apr 2021 18:23:53 +0000 (19:23 +0100)]
hw/arm/armsse: Give SSE-300 its own Property array

SSE-300 currently shares the SSE-200 Property array. This is
bad principally because the default values of the CPU0_FPU
and CPU0_DSP properties disable the FPU and DSP on the CPU.
That is correct for the SSE-200 but not the SSE-300.
Give the SSE-300 its own Property array with the correct
SSE-300 specific settings:
 * SSE-300 has only one CPU, so no CPU1* properties
 * SSE-300 CPU has FPU and DSP

Buglink: https://bugs.launchpad.net/qemu/+bug/1923861
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210415182353.8173-1-peter.maydell@linaro.org

3 years agoinclude/qemu/osdep.h: Move system includes to top
Peter Maydell [Fri, 16 Apr 2021 13:55:40 +0000 (14:55 +0100)]
include/qemu/osdep.h: Move system includes to top

Mostly osdep.h puts the system includes at the top of the file; but
there are a couple of exceptions where we include a system header
halfway through the file.  Move these up to the top with the rest
so that all the system headers we include are included before
we include os-win32.h or os-posix.h.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210416135543.20382-4-peter.maydell@linaro.org
Message-id: 20210414184343.26235-1-peter.maydell@linaro.org

3 years agoosdep: protect qemu/osdep.h with extern "C"
Paolo Bonzini [Fri, 16 Apr 2021 13:55:39 +0000 (14:55 +0100)]
osdep: protect qemu/osdep.h with extern "C"

System headers may include templates if compiled with a C++ compiler,
which cause the compiler to complain if qemu/osdep.h is included
within a C++ source file's 'extern "C"' block.  Add
an 'extern "C"' block directly to qemu/osdep.h, so that
system headers can be kept out of it.

There is a stray declaration early in qemu/osdep.h, which needs
to be special cased.  Add a definition in qemu/compiler.h to
make it look nice.

config-host.h, CONFIG_TARGET, exec/poison.h and qemu/compiler.h
are included outside the 'extern "C"' block; that is not
an issue because they consist entirely of preprocessor directives.

This allows us to move the include of osdep.h in our two C++
source files outside the extern "C" block they were previously
using for it, which in turn means that they compile successfully
against newer versions of glib which insist that glib.h is
*not* inside an extern "C" block.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210416135543.20382-3-peter.maydell@linaro.org
[PMM: Moved disas/arm-a64.cc osdep.h include out of its extern "C" block;
 explained in commit message why we're doing this]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoosdep: include glib-compat.h before other QEMU headers
Paolo Bonzini [Fri, 16 Apr 2021 13:55:38 +0000 (14:55 +0100)]
osdep: include glib-compat.h before other QEMU headers

glib-compat.h is sort of like a system header, and it needs to include
system headers (glib.h) that may dislike being included under
'extern "C"'.  Move it right after all system headers and before
all other QEMU headers.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20210416135543.20382-2-peter.maydell@linaro.org
[PMM: Added comment about why glib-compat.h is special]
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoUpdate version for v6.0.0-rc3 release
Peter Maydell [Wed, 14 Apr 2021 21:06:18 +0000 (22:06 +0100)]
Update version for v6.0.0-rc3 release

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoqapi/qom.json: Do not use CONFIG_VIRTIO_CRYPTO in common code
Thomas Huth [Mon, 12 Apr 2021 16:07:09 +0000 (18:07 +0200)]
qapi/qom.json: Do not use CONFIG_VIRTIO_CRYPTO in common code

The ObjectType enum and ObjectOptions are included from qapi-types-qom.h
into common code. We should not use target-specific config switches like
CONFIG_VIRTIO_CRYPTO here, since this is not defined in common code and
thus the enum will look differently between common and target specific
code. For this case, it's hopefully enough to check for CONFIG_VHOST_CRYPTO
only (which is a host specific config switch, i.e. it's the same on all
targets).

Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210412160710.639800-1-thuth@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/dgilbert-gitlab/tags/pull-virtiofs-20210413...
Peter Maydell [Tue, 13 Apr 2021 21:12:27 +0000 (22:12 +0100)]
Merge remote-tracking branch 'remotes/dgilbert-gitlab/tags/pull-virtiofs-20210413' into staging

virtiofs: Fix feature negotiation (for 6.0)

A 6.0 fix for feature negotiation on vhost-user.

Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
# gpg: Signature made Tue 13 Apr 2021 16:21:41 BST
# gpg:                using RSA key 45F5C71B4A0CB7FB977A9FA90516331EBC5BFDE7
# gpg: Good signature from "Dr. David Alan Gilbert (RH2) <dgilbert@redhat.com>" [full]
# Primary key fingerprint: 45F5 C71B 4A0C B7FB 977A  9FA9 0516 331E BC5B FDE7

* remotes/dgilbert-gitlab/tags/pull-virtiofs-20210413:
  vhost-user-fs: fix features handling

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/maxreitz/tags/pull-block-2021-04-13' into staging
Peter Maydell [Tue, 13 Apr 2021 20:05:17 +0000 (21:05 +0100)]
Merge remote-tracking branch 'remotes/maxreitz/tags/pull-block-2021-04-13' into staging

Block patches for 6.0-rc3:
- Use-after-free fix for block/nbd.c

# gpg: Signature made Tue 13 Apr 2021 14:35:48 BST
# gpg:                using RSA key 91BEB60A30DB3E8857D11829F407DB0061D5CF40
# gpg:                issuer "mreitz@redhat.com"
# gpg: Good signature from "Max Reitz <mreitz@redhat.com>" [full]
# Primary key fingerprint: 91BE B60A 30DB 3E88 57D1  1829 F407 DB00 61D5 CF40

* remotes/maxreitz/tags/pull-block-2021-04-13:
  block/nbd: fix possible use after free of s->connect_thread

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210413' into...
Peter Maydell [Tue, 13 Apr 2021 16:08:13 +0000 (17:08 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210413' into staging

target-arm queue:
 * Fix MPC setting for AN524 SRAM block
 * sphinx: qapidoc: Wrap "If" section body in a paragraph node

# gpg: Signature made Tue 13 Apr 2021 13:07:12 BST
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20210413:
  sphinx: qapidoc: Wrap "If" section body in a paragraph node
  hw/arm/mps2-tz: Assert if more than one RAM is attached to an MPC
  hw/arm/mps2-tz: Fix MPC setting for AN524 SRAM block

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agovhost-user-fs: fix features handling
Anton Kuchin [Thu, 8 Apr 2021 19:55:34 +0000 (22:55 +0300)]
vhost-user-fs: fix features handling

Make virtio-fs take into account server capabilities.

Just returning requested features assumes they all of then are implemented
by server and results in setting unsupported configuration if some of them
are absent.

Signed-off-by: Anton Kuchin <antonkuchin@yandex-team.ru>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
  With changes suggested by Stefan
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
3 years agoblock/nbd: fix possible use after free of s->connect_thread
Vladimir Sementsov-Ogievskiy [Tue, 6 Apr 2021 15:51:14 +0000 (18:51 +0300)]
block/nbd: fix possible use after free of s->connect_thread

If on nbd_close() we detach the thread (in
nbd_co_establish_connection_cancel() thr->state becomes
CONNECT_THREAD_RUNNING_DETACHED), after that point we should not use
s->connect_thread (which is set to NULL), as running thread may free it
at any time.

Still nbd_co_establish_connection() does exactly this: it saves
s->connect_thread to local variable (just for better code style) and
use it even after yield point, when thread may be already detached.

Fix that. Also check thr to be non-NULL on
nbd_co_establish_connection() start for safety.

After this patch "case CONNECT_THREAD_RUNNING_DETACHED" becomes
impossible in the second switch in nbd_co_establish_connection().
Still, don't add extra abort() just before the release. If it somehow
possible to reach this "case:" it won't hurt. Anyway, good refactoring
of all this reconnect mess will come soon.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Message-Id: <20210406155114.1057355-1-vsementsov@virtuozzo.com>
Reviewed-by: Roman Kagan <rvkagan@yandex-team.ru>
Signed-off-by: Max Reitz <mreitz@redhat.com>
3 years agoMerge remote-tracking branch 'remotes/philmd/tags/mips-20210413' into staging
Peter Maydell [Tue, 13 Apr 2021 13:32:44 +0000 (14:32 +0100)]
Merge remote-tracking branch 'remotes/philmd/tags/mips-20210413' into staging

MIPS patches queue

- Fix invalid Kconfig dependency
- Fix missing migrated value
- Fix TCG temporary leak

# gpg: Signature made Tue 13 Apr 2021 11:10:11 BST
# gpg:                using RSA key FAABE75E12917221DCFD6BB2E3E32C2CDEADC0DE
# gpg: Good signature from "Philippe Mathieu-Daudé (F4BUG) <f4bug@amsat.org>" [full]
# Primary key fingerprint: FAAB E75E 1291 7221 DCFD  6BB2 E3E3 2C2C DEAD C0DE

* remotes/philmd/tags/mips-20210413:
  target/mips: Fix TCG temporary leak in gen_cache_operation()
  hw/isa/piix4: Migrate Reset Control Register
  hw/isa/Kconfig: Add missing dependency VIA VT82C686 -> APM

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/dg-gitlab/tags/ppc-for-6.0-20210412' into staging
Peter Maydell [Tue, 13 Apr 2021 12:05:07 +0000 (13:05 +0100)]
Merge remote-tracking branch 'remotes/dg-gitlab/tags/ppc-for-6.0-20210412' into staging

ppc patch queue for 2021-04-21

Here's what I hope is the last ppc related pull request for qemu-6.0.

The 2 patches here revert a behavioural change that after further
discussion we concluded was a bad idea (adding a timeout for
possibly-failed hot unplug requests).  Instead it implements a
different approach to the original problem: we again let unplug
requests the guest doesn't respond to remain pending indefinitely, but
no longer allow those to block attempts to retry the same unplug
again.

The change is a bit more complex than I'd like for this late in the
freeze.  Nonetheless, I think it's important to merge this for 6.0, so
we don't allow a release which has the probably-a-bad-idea timeout
behaviour.

# gpg: Signature made Mon 12 Apr 2021 06:25:58 BST
# gpg:                using RSA key 75F46586AE61A66CC44E87DC6C38CACA20D9B392
# gpg: Good signature from "David Gibson <david@gibson.dropbear.id.au>" [full]
# gpg:                 aka "David Gibson (Red Hat) <dgibson@redhat.com>" [full]
# gpg:                 aka "David Gibson (ozlabs.org) <dgibson@ozlabs.org>" [full]
# gpg:                 aka "David Gibson (kernel.org) <dwg@kernel.org>" [unknown]
# Primary key fingerprint: 75F4 6586 AE61 A66C C44E  87DC 6C38 CACA 20D9 B392

* remotes/dg-gitlab/tags/ppc-for-6.0-20210412:
  spapr.c: always pulse guest IRQ in spapr_core_unplug_request()
  spapr: rollback 'unplug timeout' for CPU hotunplugs

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotarget/mips: Fix TCG temporary leak in gen_cache_operation()
Philippe Mathieu-Daudé [Tue, 6 Apr 2021 20:26:21 +0000 (22:26 +0200)]
target/mips: Fix TCG temporary leak in gen_cache_operation()

Fix a TCG temporary leak when translating CACHE opcode.

Fixes: 0d74a222c27 ("make ITC Configuration Tags accessible to the CPU")
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20210406202857.1440744-1-f4bug@amsat.org>

3 years agohw/isa/piix4: Migrate Reset Control Register
Philippe Mathieu-Daudé [Wed, 24 Mar 2021 13:54:43 +0000 (14:54 +0100)]
hw/isa/piix4: Migrate Reset Control Register

When adding the Reset register in commit 5790b757cfb we
forgot to migrate it.

While it is possible a VM using the PIIX4 is migrated just
after requesting a system shutdown, it is very unlikely.
However when restoring a migrated VM, we might have the
RCR bit #4 set on the stack and when the VM resume it
directly shutdowns.

Add a post_load() migration handler and set the default
RCR value to 0 for earlier versions, assuming the VM was
not going to shutdown before migration.

Fixes: 5790b757cfb ("piix4: Add the Reset Control Register")
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20210324200334.729899-1-f4bug@amsat.org>

3 years agohw/isa/Kconfig: Add missing dependency VIA VT82C686 -> APM
Philippe Mathieu-Daudé [Tue, 2 Mar 2021 08:00:42 +0000 (09:00 +0100)]
hw/isa/Kconfig: Add missing dependency VIA VT82C686 -> APM

TYPE_VIA_PM calls apm_init() in via_pm_realize(), so
requires APM to be selected.

Reported-by: BALATON Zoltan <balaton@eik.bme.hu>
Fixes: dd0ff8191ab ("isa: express SuperIO dependencies with Kconfig")
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210302080531.913802-1-f4bug@amsat.org>

3 years agoMerge remote-tracking branch 'remotes/mcayland/tags/qemu-sparc-20210412' into staging
Peter Maydell [Tue, 13 Apr 2021 10:00:34 +0000 (11:00 +0100)]
Merge remote-tracking branch 'remotes/mcayland/tags/qemu-sparc-20210412' into staging

qemu-sparc queue

# gpg: Signature made Mon 12 Apr 2021 23:13:12 BST
# gpg:                using RSA key CC621AB98E82200D915CC9C45BC2C56FAE0F321F
# gpg:                issuer "mark.cave-ayland@ilande.co.uk"
# gpg: Good signature from "Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>" [full]
# Primary key fingerprint: CC62 1AB9 8E82 200D 915C  C9C4 5BC2 C56F AE0F 321F

* remotes/mcayland/tags/qemu-sparc-20210412:
  tests/qtest: add tests for am53c974 device
  esp: ensure that do_cmd is set to zero before submitting an ESP select command
  esp: don't reset async_len directly in esp_select() if cancelling request
  esp: don't overflow cmdfifo if TC is larger than the cmdfifo size
  esp: don't overflow cmdfifo in get_cmd()
  esp: don't underflow cmdfifo in do_cmd()
  esp: ensure cmdfifo is not empty and current_dev is non-NULL
  esp: introduce esp_fifo_pop_buf() and use it instead of fifo8_pop_buf()
  esp: consolidate esp_cmdfifo_pop() into esp_fifo_pop()
  esp: consolidate esp_cmdfifo_push() into esp_fifo_push()
  esp: rework write_response() to avoid using the FIFO for DMA transactions
  esp: always check current_req is not NULL before use in DMA callbacks
  esp: fix setting of ESPState mig_version_id when launching QEMU with -S option

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agosphinx: qapidoc: Wrap "If" section body in a paragraph node
John Snow [Tue, 6 Apr 2021 14:19:09 +0000 (10:19 -0400)]
sphinx: qapidoc: Wrap "If" section body in a paragraph node

These sections need to be wrapped in a block-level element, such as
Paragraph in order for them to be rendered into Texinfo correctly.

Before (e.g.):

<section ids="qapidoc-713">
  <title>If</title>
  <literal>defined(CONFIG_REPLICATION)</literal>
</section>

became:

  .SS If
  \fBdefined(CONFIG_REPLICATION)\fP.SS \fBBlockdevOptionsReplication\fP (Object)
  ...

After:

<section ids="qapidoc-713">
  <title>If</title>
  <paragraph>
    <literal>defined(CONFIG_REPLICATION)</literal>
  </paragraph>
</section>

becomes:

  .SS If
  .sp
  \fBdefined(CONFIG_REPLICATION)\fP
  .SS \fBBlockdevOptionsReplication\fP (Object)
  ...

Reported-by: Markus Armbruster <armbru@redhat.com>
Tested-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20210406141909.1992225-2-jsnow@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotests/qtest: add tests for am53c974 device
Mark Cave-Ayland [Wed, 7 Apr 2021 19:58:01 +0000 (20:58 +0100)]
tests/qtest: add tests for am53c974 device

Use the autogenerated fuzzer test cases as the basis for a set of am53c974
regression tests.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-13-mark.cave-ayland@ilande.co.uk>

3 years agoesp: ensure that do_cmd is set to zero before submitting an ESP select command
Mark Cave-Ayland [Wed, 7 Apr 2021 19:58:00 +0000 (20:58 +0100)]
esp: ensure that do_cmd is set to zero before submitting an ESP select command

When a CDB has been received and is about to be submitted to the SCSI layer
via one of the ESP select commands, ensure that do_cmd is set to zero before
executing the command.

Otherwise a guest executing 2 valid CDBs in quick sequence can invoke the SCSI
.transfer_data callback again before do_cmd is set to zero by the callback
function triggering an assert at the start of esp_transfer_data().

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210407195801.685-12-mark.cave-ayland@ilande.co.uk>

3 years agoesp: don't reset async_len directly in esp_select() if cancelling request
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:59 +0000 (20:57 +0100)]
esp: don't reset async_len directly in esp_select() if cancelling request

Instead let the SCSI layer invoke the .cancel callback itself to cancel and
reset the request state.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210407195801.685-11-mark.cave-ayland@ilande.co.uk>

3 years agoesp: don't overflow cmdfifo if TC is larger than the cmdfifo size
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:58 +0000 (20:57 +0100)]
esp: don't overflow cmdfifo if TC is larger than the cmdfifo size

If a guest transfers the message out/command phase data using DMA with a TC
that is larger than the cmdfifo size then the cmdfifo overflows triggering
an assert. Limit the size of the transfer to the free space available in
cmdfifo.

Buglink: https://bugs.launchpad.net/qemu/+bug/1919036
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-10-mark.cave-ayland@ilande.co.uk>

3 years agoesp: don't overflow cmdfifo in get_cmd()
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:57 +0000 (20:57 +0100)]
esp: don't overflow cmdfifo in get_cmd()

If the guest tries to read a CDB using DMA and cmdfifo is not empty then it is
possible to overflow cmdfifo.

Since this can only occur by issuing deliberately incorrect instruction
sequences, ensure that the maximum length of the CDB transferred to cmdfifo is
limited to the available free space within cmdfifo.

Buglink: https://bugs.launchpad.net/qemu/+bug/1909247
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-9-mark.cave-ayland@ilande.co.uk>

3 years agoesp: don't underflow cmdfifo in do_cmd()
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:56 +0000 (20:57 +0100)]
esp: don't underflow cmdfifo in do_cmd()

If the guest tries to execute a CDB when cmdfifo is not empty before the start
of the message out phase then clearing the message out phase data will cause
cmdfifo to underflow due to cmdfifo_cdb_offset being larger than the amount of
data within.

Since this can only occur by issuing deliberately incorrect instruction
sequences, ensure that the maximum length of esp_fifo_pop_buf() is limited to
the size of the data within cmdfifo.

Buglink: https://bugs.launchpad.net/qemu/+bug/1909247
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-8-mark.cave-ayland@ilande.co.uk>

3 years agoesp: ensure cmdfifo is not empty and current_dev is non-NULL
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:55 +0000 (20:57 +0100)]
esp: ensure cmdfifo is not empty and current_dev is non-NULL

When about to execute a SCSI command, ensure that cmdfifo is not empty and
current_dev is non-NULL. This can happen if the guest tries to execute a TI
(Transfer Information) command without issuing one of the select commands
first.

Buglink: https://bugs.launchpad.net/qemu/+bug/1910723
Buglink: https://bugs.launchpad.net/qemu/+bug/1909247
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-7-mark.cave-ayland@ilande.co.uk>

3 years agoesp: introduce esp_fifo_pop_buf() and use it instead of fifo8_pop_buf()
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:54 +0000 (20:57 +0100)]
esp: introduce esp_fifo_pop_buf() and use it instead of fifo8_pop_buf()

The const pointer returned by fifo8_pop_buf() lies directly within the array used
to model the FIFO. Building with address sanitizers enabled shows that if the
caller expects a minimum number of bytes present then if the FIFO is nearly full,
the caller may unexpectedly access past the end of the array.

Introduce esp_fifo_pop_buf() which takes a destination buffer and performs a
memcpy() in it to guarantee that the caller cannot overwrite the FIFO array and
update all callers to use it. Similarly add underflow protection similar to
esp_fifo_push() and esp_fifo_pop() so that instead of triggering an assert()
the operation becomes a no-op.

Buglink: https://bugs.launchpad.net/qemu/+bug/1909247
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20210407195801.685-6-mark.cave-ayland@ilande.co.uk>

3 years agoesp: consolidate esp_cmdfifo_pop() into esp_fifo_pop()
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:53 +0000 (20:57 +0100)]
esp: consolidate esp_cmdfifo_pop() into esp_fifo_pop()

Each FIFO currently has its own pop functions with the only difference being
the capacity check. The original reason for this was that the fifo8
implementation doesn't have a formal API for retrieving the FIFO capacity,
however there are multiple examples within QEMU where the capacity field is
accessed directly.

Change esp_fifo_pop() to access the FIFO capacity directly and then consolidate
esp_cmdfifo_pop() into esp_fifo_pop().

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-5-mark.cave-ayland@ilande.co.uk>

3 years agoesp: consolidate esp_cmdfifo_push() into esp_fifo_push()
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:52 +0000 (20:57 +0100)]
esp: consolidate esp_cmdfifo_push() into esp_fifo_push()

Each FIFO currently has its own push functions with the only difference being
the capacity check. The original reason for this was that the fifo8
implementation doesn't have a formal API for retrieving the FIFO capacity,
however there are multiple examples within QEMU where the capacity field is
accessed directly.

Change esp_fifo_push() to access the FIFO capacity directly and then consolidate
esp_cmdfifo_push() into esp_fifo_push().

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-4-mark.cave-ayland@ilande.co.uk>

3 years agoesp: rework write_response() to avoid using the FIFO for DMA transactions
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:51 +0000 (20:57 +0100)]
esp: rework write_response() to avoid using the FIFO for DMA transactions

The code for write_response() has always used the FIFO to store the data for
the status/message in phases, even for DMA transactions. Switch to using a
separate buffer that can be used directly for DMA transactions and restrict
the FIFO use to the non-DMA case.

Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-Id: <20210407195801.685-3-mark.cave-ayland@ilande.co.uk>

3 years agoesp: always check current_req is not NULL before use in DMA callbacks
Mark Cave-Ayland [Wed, 7 Apr 2021 19:57:50 +0000 (20:57 +0100)]
esp: always check current_req is not NULL before use in DMA callbacks

After issuing a SCSI command the SCSI layer can call the SCSIBusInfo .cancel
callback which resets both current_req and current_dev to NULL. If any data
is left in the transfer buffer (async_len != 0) then the next TI (Transfer
Information) command will attempt to reference the NULL pointer causing a
segfault.

Buglink: https://bugs.launchpad.net/qemu/+bug/1910723
Buglink: https://bugs.launchpad.net/qemu/+bug/1909247
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Tested-by: Alexander Bulekov <alxndr@bu.edu>
Message-Id: <20210407195801.685-2-mark.cave-ayland@ilande.co.uk>

3 years agoesp: fix setting of ESPState mig_version_id when launching QEMU with -S option
Mark Cave-Ayland [Wed, 7 Apr 2021 12:48:42 +0000 (13:48 +0100)]
esp: fix setting of ESPState mig_version_id when launching QEMU with -S option

If QEMU is launched with the -S option then the ESPState mig_version_id property
is left unset due to the ordering of the VMState fields in the VMStateDescription
for sysbusespscsi and pciespscsi. If the VM is migrated and restored in this
stopped state, the version tests in the vmstate_esp VMStateDescription and
esp_post_load() become confused causing the migration to fail.

Fix the ordering problem by moving the setting of mig_version_id to a common
esp_pre_save() function which is invoked first by both sysbusespscsi and
pciespscsi rather than at the point where ESPState is itself serialised into the
migration stream.

Buglink: https://bugs.launchpad.net/qemu/+bug/1922611
Fixes: 0bd005be78 ("esp: add vmstate_esp version to embedded ESPState")
Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20210407124842.32695-1-mark.cave-ayland@ilande.co.uk>

3 years agohw/arm/mps2-tz: Assert if more than one RAM is attached to an MPC
Peter Maydell [Fri, 9 Apr 2021 15:05:27 +0000 (16:05 +0100)]
hw/arm/mps2-tz: Assert if more than one RAM is attached to an MPC

Each board in mps2-tz.c specifies a RAMInfo[] array providing
information about each RAM in the board.  The .mpc field of the
RAMInfo struct specifies which MPC, if any, the RAM is attached to.
We already assert if the array doesn't have any entry for an MPC, but
we don't diagnose the error of using the same MPC number twice (which
is quite easy to do by accident if copy-and-pasting structure
entries).

Enhance find_raminfo_for_mpc() so that it detects multiple entries
for the MPC as well as missing entries.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20210409150527.15053-3-peter.maydell@linaro.org

3 years agohw/arm/mps2-tz: Fix MPC setting for AN524 SRAM block
Peter Maydell [Fri, 9 Apr 2021 15:05:26 +0000 (16:05 +0100)]
hw/arm/mps2-tz: Fix MPC setting for AN524 SRAM block

The AN524 has three MPCs: one for the BRAM, one for the QSPI flash,
and one for the DDR.  We incorrectly set the .mpc field in the
RAMInfo struct for the SRAM block to 1, giving it the same MPC we are
using for the QSPI.  The effect of this was that the QSPI didn't get
mapped into the system address space at all, via an MPC or otherwise,
and guest programs which tried to read from the QSPI would get a bus
error.  Correct the SRAM RAMInfo to indicate that it does not have an
associated MPC.

Fixes: 25ff112a8cc ("hw/arm/mps2-tz: Add new mps3-an524 board")
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Message-id: 20210409150527.15053-2-peter.maydell@linaro.org

3 years agoMerge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210412' into...
Peter Maydell [Mon, 12 Apr 2021 11:12:09 +0000 (12:12 +0100)]
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20210412' into staging

target-arm queue:
 * hw/arm/virt-acpi-build: Fix GSIV values of the {GERR, Sync} interrupts
 * hw/arm/smmuv3: Emulate CFGI_STE_RANGE for an aligned range of StreamIDs
 * accel/tcg: Preserve PAGE_ANON when changing page permissions
 * target/arm: Check PAGE_WRITE_ORG for MTE writeability
 * exec: Fix overlap of PAGE_ANON and PAGE_TARGET_1

# gpg: Signature made Mon 12 Apr 2021 11:31:15 BST
# gpg:                using RSA key E1A5C593CD419DE28E8315CF3C2525ED14360CDE
# gpg:                issuer "peter.maydell@linaro.org"
# gpg: Good signature from "Peter Maydell <peter.maydell@linaro.org>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@gmail.com>" [ultimate]
# gpg:                 aka "Peter Maydell <pmaydell@chiark.greenend.org.uk>" [ultimate]
# Primary key fingerprint: E1A5 C593 CD41 9DE2 8E83  15CF 3C25 25ED 1436 0CDE

* remotes/pmaydell/tags/pull-target-arm-20210412:
  exec: Fix overlap of PAGE_ANON and PAGE_TARGET_1
  target/arm: Check PAGE_WRITE_ORG for MTE writeability
  accel/tcg: Preserve PAGE_ANON when changing page permissions
  hw/arm/smmuv3: Emulate CFGI_STE_RANGE for an aligned range of StreamIDs
  hw/arm/virt-acpi-build: Fix GSIV values of the {GERR, Sync} interrupts

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoexec: Fix overlap of PAGE_ANON and PAGE_TARGET_1
Richard Henderson [Tue, 6 Apr 2021 17:40:31 +0000 (10:40 -0700)]
exec: Fix overlap of PAGE_ANON and PAGE_TARGET_1

Unfortuately, the elements of PAGE_* were not in numerical
order and so PAGE_ANON was added to an "unused" bit.
As an arbitrary choice, move PAGE_TARGET_{1,2} together.

Cc: Laurent Vivier <laurent@vivier.eu>
Fixes: 26bab757d41b ("linux-user: Introduce PAGE_ANON")
Buglink: https://bugs.launchpad.net/bugs/1922617
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Laurent Vivier <laurent@vivier.eu>
Tested-by: Laurent Vivier <laurent@vivier.eu>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agotarget/arm: Check PAGE_WRITE_ORG for MTE writeability
Richard Henderson [Tue, 6 Apr 2021 17:40:21 +0000 (10:40 -0700)]
target/arm: Check PAGE_WRITE_ORG for MTE writeability

We can remove PAGE_WRITE when (internally) marking a page
read-only because it contains translated code.

This can be triggered by tests/tcg/aarch64/bti-2, after
having serviced SIGILL trampolines on the stack.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoaccel/tcg: Preserve PAGE_ANON when changing page permissions
Richard Henderson [Tue, 6 Apr 2021 17:40:20 +0000 (10:40 -0700)]
accel/tcg: Preserve PAGE_ANON when changing page permissions

Using mprotect() to change PROT_* does not change the MAP_ANON
previously set with mmap().  Our linux-user version of MTE only
works with MAP_ANON pages, so losing PAGE_ANON caused MTE to
stop working.

Reported-by: Stephen Long <steplong@quicinc.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agohw/arm/smmuv3: Emulate CFGI_STE_RANGE for an aligned range of StreamIDs
Zenghui Yu [Fri, 2 Apr 2021 10:04:49 +0000 (18:04 +0800)]
hw/arm/smmuv3: Emulate CFGI_STE_RANGE for an aligned range of StreamIDs

In emulation of the CFGI_STE_RANGE command, we now take StreamID as the
start of the invalidation range, regardless of whatever the Range is,
whilst the spec clearly states that

 - "Invalidation is performed for an *aligned* range of 2^(Range+1)
    StreamIDs."

 - "The bottom Range+1 bits of the StreamID parameter are IGNORED,
    aligning the range to its size."

Take CFGI_ALL (where Range == 31) as an example, if there are some random
bits in the StreamID field, we'll fail to perform the full invalidation but
get a strange range (e.g., SMMUSIDRange={.start=1, .end=0}) instead. Rework
the emulation a bit to get rid of the discrepancy with the spec.

Signed-off-by: Zenghui Yu <yuzenghui@huawei.com>
Acked-by: Eric Auger <eric.auger@redhat.com>
Message-id: 20210402100449.528-1-yuzenghui@huawei.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agohw/arm/virt-acpi-build: Fix GSIV values of the {GERR, Sync} interrupts
Zenghui Yu [Fri, 2 Apr 2021 08:47:31 +0000 (16:47 +0800)]
hw/arm/virt-acpi-build: Fix GSIV values of the {GERR, Sync} interrupts

The GSIV values in SMMUv3 IORT node are not correct as they don't match
the SMMUIrq enumeration, which describes the IRQ<->PIN mapping used by
our emulated vSMMU.

Fixes: a703b4f6c1ee ("hw/arm/virt-acpi-build: Add smmuv3 node in IORT table")
Signed-off-by: Zenghui Yu <yuzenghui@huawei.com>
Acked-by: Eric Auger <eric.auger@redhat.com>
Message-id: 20210402084731.93-1-yuzenghui@huawei.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agoMerge remote-tracking branch 'remotes/nvme/tags/nvme-fixes-20210412-pull-request...
Peter Maydell [Mon, 12 Apr 2021 09:41:11 +0000 (10:41 +0100)]
Merge remote-tracking branch 'remotes/nvme/tags/nvme-fixes-20210412-pull-request' into staging

emulated nvme docs and fixes for -rc3

- documentation
- fixes

# gpg: Signature made Mon 12 Apr 2021 07:56:09 BST
# gpg:                using RSA key 522833AA75E2DCE6A24766C04DE1AF316D4F0DE9
# gpg: Good signature from "Klaus Jensen <its@irrelevant.dk>" [unknown]
# gpg:                 aka "Klaus Jensen <k.jensen@samsung.com>" [unknown]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: DDCA 4D9C 9EF9 31CC 3468  4272 63D5 6FC5 E55D A838
#      Subkey fingerprint: 5228 33AA 75E2 DCE6 A247  66C0 4DE1 AF31 6D4F 0DE9

* remotes/nvme/tags/nvme-fixes-20210412-pull-request:
  hw/block/nvme: drain namespaces on sq deletion
  hw/block/nvme: store aiocb in compare
  hw/block/nvme: map prp fix if prp2 contains non-zero offset
  docs: add nvme emulation documentation

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
3 years agohw/block/nvme: drain namespaces on sq deletion
Klaus Jensen [Thu, 8 Apr 2021 10:44:05 +0000 (12:44 +0200)]
hw/block/nvme: drain namespaces on sq deletion

For most commands, when issuing an AIO, the BlockAIOCB is stored in the
NvmeRequest aiocb pointer when the AIO is issued. The main use of this
is cancelling AIOs when deleting submission queues (it is currently not
used for Abort).

However, some commands like Dataset Management Zone Management Send
(zone reset) may involve more than one AIO and here the AIOs are issued
without saving a reference to the BlockAIOCB. This is a problem since
nvme_del_sq() will attempt to cancel outstanding AIOs, potentially with
an invalid BlockAIOCB since the aiocb pointer is not NULL'ed when the
request structure is recycled.

Fix this by

  1. making sure the aiocb pointer is NULL'ed when requests are recycled
  2. only attempt to cancel the AIO if the aiocb is non-NULL
  3. if any AIOs could not be cancelled, drain all aio as a last resort.

Fixes: dc04d25e2f3f ("hw/block/nvme: add support for the format nvm command")
Fixes: c94973288cd9 ("hw/block/nvme: add broadcast nsid support flush command")
Fixes: e4e430b3d6ba ("hw/block/nvme: add simple copy command")
Fixes: 5f5dc4c6a942 ("hw/block/nvme: zero out zones on reset")
Fixes: 2605257a26b8 ("hw/block/nvme: add the dataset management command")
Cc: Gollu Appalanaidu <anaidu.gollu@samsung.com>
Cc: Minwoo Im <minwoo.im@samsung.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Minwoo Im <minwoo.im.dev@gmail.com>
3 years agohw/block/nvme: store aiocb in compare
Klaus Jensen [Thu, 8 Apr 2021 11:46:03 +0000 (13:46 +0200)]
hw/block/nvme: store aiocb in compare

nvme_compare() fails to store the aiocb from the blk_aio_preadv() call.
Fix this.

Fixes: 0a384f923f51 ("hw/block/nvme: add compare command")
Cc: Gollu Appalanaidu <anaidu.gollu@samsung.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Gollu Appalanaidu <anaidu.gollu@samsung.com>
Reviewed-by: Minwoo Im <minwoo.im.dev@gmail.com>
3 years agohw/block/nvme: map prp fix if prp2 contains non-zero offset
Padmakar Kalghatgi [Fri, 9 Apr 2021 07:25:48 +0000 (12:55 +0530)]
hw/block/nvme: map prp fix if prp2 contains non-zero offset

nvme_map_prp needs to calculate the number of list entries based on the
offset value. For the subsequent PRP2 list, need to ensure the number of
entries is within the MAX number of PRP entries for a page.

Signed-off-by: Padmakar Kalghatgi <p.kalghatgi@samsung.com>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
3 years agodocs: add nvme emulation documentation
Klaus Jensen [Thu, 8 Apr 2021 09:11:02 +0000 (11:11 +0200)]
docs: add nvme emulation documentation

Remove the docs/specs/nvme.txt and replace it with proper documentation
in docs/system/nvme.rst.

Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
3 years agospapr.c: always pulse guest IRQ in spapr_core_unplug_request()
Daniel Henrique Barboza [Thu, 1 Apr 2021 00:04:37 +0000 (21:04 -0300)]
spapr.c: always pulse guest IRQ in spapr_core_unplug_request()

Commit 47c8c915b162 fixed a problem where multiple spapr_drc_detach()
requests were breaking QEMU. The solution was to just spapr_drc_detach()
once, and use spapr_drc_unplug_requested() to filter whether we already
detached it or not. The commit also tied the hotplug request to the
guest in the same condition.

Turns out that there is a reliable way for a CPU hotunplug to fail. If a
guest with one CPU hotplugs a CPU1, then offline CPU0s via 'echo 0 >
/sys/devices/system/cpu/cpu0/online', then attempts to hotunplug CPU1,
the kernel will refuse it because it's the last online CPU of the
system. Given that we're pulsing the IRQ only in the first try, in a
failed attempt, all other CPU1 hotunplug attempts will fail, regardless
of the online state of CPU1 in the kernel, because we're simply not
letting the guest know that we want to hotunplug the device.

Let's move spapr_hotplug_req_remove_by_index() back out of the "if
(!spapr_drc_unplug_requested(drc))" conditional, allowing for multiple
'device_del' requests to the same CPU core to reach the guest, in case
the CPU core didn't fully hotunplugged previously.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210401000437.131140-3-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
3 years agospapr: rollback 'unplug timeout' for CPU hotunplugs
Daniel Henrique Barboza [Thu, 1 Apr 2021 00:04:36 +0000 (21:04 -0300)]
spapr: rollback 'unplug timeout' for CPU hotunplugs

The pseries machines introduced the concept of 'unplug timeout' for CPU
hotunplugs. The idea was to circunvent a deficiency in the pSeries
specification (PAPR), that currently does not define a proper way for
the hotunplug to fail. If the guest refuses to release the CPU (see [1]
for an example) there is no way for QEMU to detect the failure.

Further discussions about how to send a QAPI event to inform about the
hotunplug timeout [2] exposed problems that weren't predicted back when
the idea was developed. Other QEMU machines don't have any type of
hotunplug timeout mechanism for any device, e.g. ACPI based machines
have a way to make hotunplug errors visible to the hypervisor. This
would make this timeout mechanism exclusive to pSeries, which is not
ideal.

The real problem is that a QAPI event that reports hotunplug timeouts
puts the management layer (namely Libvirt) in a weird spot. We're not
telling that the hotunplug failed, because we can't be 100% sure of
that, and yet we're resetting the unplug state back, preventing any
DEVICE_DEL events to reach out in case the guest decides to release the
device. Libvirt would need to inspect the guest itself to see if the
device was released or not, otherwise the internal domain states will be
inconsistent.  Moreover, Libvirt already has an 'unplug timeout'
concept, and a QEMU side timeout would need to be juggled together with
the existing Libvirt timeout.

All this considered, this solution ended up creating more trouble than
it solved. This patch reverts the 3 commits that introduced the timeout
mechanism for CPU hotplugs in pSeries machines.

This reverts commit 4515a5f786024fabf0bef4cf3d28adf5647e6e82
"qemu_timer.c: add timer_deadline_ms() helper"

This reverts commit d1c2e3ce3d5a5424651967bce1cf1f4caa0c6d91
"spapr_drc.c: add hotunplug timeout for CPUs"

This reverts commit 51254ffb320183a4636635840c23ee0e3a1efffa
"spapr_drc.c: introduce unplug_timeout_timer"

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1911414
[2] https://lists.gnu.org/archive/html/qemu-devel/2021-03/msg04682.html

CC: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20210401000437.131140-2-danielhb413@gmail.com>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>