In case a hostname resolves to a different IP address (e.g. long
running mounts), make sure to resolve it every time prior to calling
generic_ip_connect() in reconnect.
Suggested-by: Steve French <stfrench@microsoft.com> Signed-off-by: Paulo Alcantara <palcantara@suse.de> Signed-off-by: Steve French <stfrench@microsoft.com> Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This fixes the case when md array assembly fails because of raid cache recovery
unable to allocate a stripe, despite attempts to replay stripes and increase
cache size. This happens because stripes released by r5c_recovery_replay_stripes
and raid5_set_cache_size don't become available for allocation immediately.
Released stripes first are placed on conf->released_stripes list and require
md thread to merge them on conf->inactive_list before they can be allocated.
Patch allows final allocation attempt during cache recovery to wait for
new stripes to become availabe for allocation.
Cc: linux-raid@vger.kernel.org Cc: Shaohua Li <shli@kernel.org> Cc: linux-stable <stable@vger.kernel.org> # 4.10+ Fixes: b4c625c67362 ("md/r5cache: r5cache recovery: part 1") Signed-off-by: Alexei Naberezhnov <anaberezhnov@fb.com> Signed-off-by: Song Liu <songliubraving@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When allocating a new node, add_changeset_node() was duplicating the
properties from the respective node in the overlay instead of
allocating a node with no properties.
When this patch is applied the errors reported by the devictree
unittest from patch "of: overlay: add tests to validate kfrees from
overlay removal" will no longer occur. These error messages are of
the form:
"OF: ERROR: ..."
and the unittest results will change from:
### dt-test ### end of unittest - 203 passed, 7 failed
to
### dt-test ### end of unittest - 210 passed, 0 failed
Tested-by: Alan Tull <atull@kernel.org> Signed-off-by: Frank Rowand <frank.rowand@sony.com> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The changeset entry 'update property' was used for new properties in
an overlay instead of 'add property'.
The decision of whether to use 'update property' was based on whether
the property already exists in the subtree where the node is being
spliced into. At the top level of creating a changeset describing the
overlay, the target node is in the live devicetree, so checking whether
the property exists in the target node returns the correct result.
As soon as the changeset creation algorithm recurses into a new node,
the target is no longer in the live devicetree, but is instead in the
detached overlay tree, thus all properties are incorrectly found to
already exist in the target.
This fix will expose another devicetree bug that will be fixed
in the following patch in the series.
When this patch is applied the errors reported by the devictree
unittest will change, and the unittest results will change from:
### dt-test ### end of unittest - 210 passed, 0 failed
to
### dt-test ### end of unittest - 203 passed, 7 failed
Tested-by: Alan Tull <atull@kernel.org> Signed-off-by: Frank Rowand <frank.rowand@sony.com> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a matching of_node_put() in __of_detach_node_sysfs()
Remove misleading comment from function header comment for
of_detach_node().
This patch may result in memory leaks from code that directly calls
the dynamic node add and delete functions directly instead of
using changesets.
This commit should result in powerpc systems that dynamically
allocate a node, then later deallocate the node to have a
memory leak when the node is deallocated.
The next commit will fix the leak.
Tested-by: Alan Tull <atull@kernel.org> Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc) Signed-off-by: Frank Rowand <frank.rowand@sony.com> Cc: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add checks:
- attempted kfree due to refcount reaching zero before overlay
is removed
- properties linked to an overlay node when the node is removed
- node refcount > one during node removal in a changeset destroy,
if the node was created by the changeset
After applying this patch, several validation warnings will be
reported from the devicetree unittest during boot due to
pre-existing devicetree bugs. The warnings will be similar to:
We had a race in the old balloon compaction code before b1123ea6d3b3
("mm: balloon: use general non-lru movable page feature") refactored it
that became visible after backporting 195a8c43e93d ("virtio-balloon:
deflate via a page list") without the refactoring.
The bug existed from commit d6d86c0a7f8d ("mm/balloon_compaction:
redesign ballooned pages management") till b1123ea6d3b3 ("mm: balloon:
use general non-lru movable page feature"). d6d86c0a7f8d
("mm/balloon_compaction: redesign ballooned pages management") was
backported to 3.12, so the broken kernels are stable kernels [3.12 -
4.7].
There was a subtle race between dropping the page lock of the newpage in
__unmap_and_move() and checking for __is_movable_balloon_page(newpage).
Just after dropping this page lock, virtio-balloon could go ahead and
deflate the newpage, effectively dequeueing it and clearing PageBalloon,
in turn making __is_movable_balloon_page(newpage) fail.
This resulted in dropping the reference of the newpage via
putback_lru_page(newpage) instead of put_page(newpage), leading to
page->lru getting modified and a !LRU page ending up in the LRU lists.
With 195a8c43e93d ("virtio-balloon: deflate via a page list")
backported, one would suddenly get corrupted lists in
release_pages_balloon():
- WARNING: CPU: 13 PID: 6586 at lib/list_debug.c:59 __list_del_entry+0xa1/0xd0
- list_del corruption. prev->next should be ffffe253961090a0, but was dead000000000100
Nowadays this race is no longer possible, but it is hidden behind very
ugly handling of __ClearPageMovable() and __PageMovable().
__ClearPageMovable() will not make __PageMovable() fail, only
PageMovable(). So the new check (__PageMovable(newpage)) will still
hold even after newpage was dequeued by virtio-balloon.
If anybody would ever change that special handling, the BUG would be
introduced again. So instead, make it explicit and use the information
of the original isolated page before migration.
This patch can be backported fairly easy to stable kernels (in contrast
to the refactoring).
Link: http://lkml.kernel.org/r/20190129233217.10747-1-david@redhat.com Fixes: d6d86c0a7f8d ("mm/balloon_compaction: redesign ballooned pages management") Signed-off-by: David Hildenbrand <david@redhat.com> Reported-by: Vratislav Bendel <vbendel@redhat.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Rafael Aquini <aquini@redhat.com> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: Jan Kara <jack@suse.cz> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Dominik Brodowski <linux@dominikbrodowski.net> Cc: Matthew Wilcox <willy@infradead.org> Cc: Vratislav Bendel <vbendel@redhat.com> Cc: Rafael Aquini <aquini@redhat.com> Cc: Konstantin Khlebnikov <k.khlebnikov@samsung.com> Cc: Minchan Kim <minchan@kernel.org> Cc: <stable@vger.kernel.org> [3.12 - 4.7] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently memory_failure() is racy against process's exiting, which
results in kernel crash by null pointer dereference.
The root cause is that memory_failure() uses force_sig() to forcibly
kill asynchronous (meaning not in the current context) processes. As
discussed in thread https://lkml.org/lkml/2010/6/8/236 years ago for OOM
fixes, this is not a right thing to do. OOM solves this issue by using
do_send_sig_info() as done in commit d2d393099de2 ("signal:
oom_kill_task: use SEND_SIG_FORCED instead of force_sig()"), so this
patch is suggesting to do the same for hwpoison. do_send_sig_info()
properly accesses to siglock with lock_task_sighand(), so is free from
the reported race.
I confirmed that the reported bug reproduces with inserting some delay
in kill_procs(), and it never reproduces with this patch.
Note that memory_failure() can send another type of signal using
force_sig_mceerr(), and the reported race shouldn't happen on it because
force_sig_mceerr() is called only for synchronous processes (i.e.
BUS_MCEERR_AR happens only when some process accesses to the corrupted
memory.)
Link: http://lkml.kernel.org/r/20190116093046.GA29835@hori1.linux.bs1.fc.nec.co.jp Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Reported-by: Jane Chu <jane.chu@oracle.com> Reviewed-by: Dan Williams <dan.j.williams@intel.com> Reviewed-by: William Kucharski <william.kucharski@oracle.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Syzbot instance running on upstream kernel found a use-after-free bug in
oom_kill_process. On further inspection it seems like the process
selected to be oom-killed has exited even before reaching
read_lock(&tasklist_lock) in oom_kill_process(). More specifically the
tsk->usage is 1 which is due to get_task_struct() in oom_evaluate_task()
and the put_task_struct within for_each_thread() frees the tsk and
for_each_thread() tries to access the tsk. The easiest fix is to do
get/put across the for_each_thread() on the selected task.
Now the next question is should we continue with the oom-kill as the
previously selected task has exited? However before adding more
complexity and heuristics, let's answer why we even look at the children
of oom-kill selected task? The select_bad_process() has already selected
the worst process in the system/memcg. Due to race, the selected
process might not be the worst at the kill time but does that matter?
The userspace can use the oom_score_adj interface to prefer children to
be killed before the parent. I looked at the history but it seems like
this is there before git history.
Link: http://lkml.kernel.org/r/20190121215850.221745-1-shakeelb@google.com Reported-by: syzbot+7fbbfa368521945f0e3d@syzkaller.appspotmail.com Fixes: 6b0c81b3be11 ("mm, oom: reduce dependency on tasklist_lock") Signed-off-by: Shakeel Butt <shakeelb@google.com> Reviewed-by: Roman Gushchin <guro@fb.com> Acked-by: Michal Hocko <mhocko@suse.com> Cc: David Rientjes <rientjes@google.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is the same sort of error we saw in commit 17e2e7d7e1b8 ("mm,
page_alloc: fix has_unmovable_pages for HugePages").
Gigantic hugepages cross several memblocks, so it can be that the page
we get in scan_movable_pages() is a page-tail belonging to a
1G-hugepage. If that happens, page_hstate()->size_to_hstate() will
return NULL, and we will blow up in hugepage_migration_supported().
Arkadiusz reported that enabling memcg's group oom killing causes
strange memcg statistics where there is no task in a memcg despite the
number of tasks in that memcg is not 0. It turned out that there is a
bug in wake_oom_reaper() which allows enqueuing same task twice which
makes impossible to decrease the number of tasks in that memcg due to a
refcount leak.
This bug existed since the OOM reaper became invokable from
task_will_free_mem(current) path in out_of_memory() in Linux 4.7,
T1@P1 |T2@P1 |T3@P1 |OOM reaper
----------+----------+----------+------------
# Processing an OOM victim in a different memcg domain.
try_charge()
mem_cgroup_out_of_memory()
mutex_lock(&oom_lock)
try_charge()
mem_cgroup_out_of_memory()
mutex_lock(&oom_lock)
try_charge()
mem_cgroup_out_of_memory()
mutex_lock(&oom_lock)
out_of_memory()
oom_kill_process(P1)
do_send_sig_info(SIGKILL, @P1)
mark_oom_victim(T1@P1)
wake_oom_reaper(T1@P1) # T1@P1 is enqueued.
mutex_unlock(&oom_lock)
out_of_memory()
mark_oom_victim(T2@P1)
wake_oom_reaper(T2@P1) # T2@P1 is enqueued.
mutex_unlock(&oom_lock)
out_of_memory()
mark_oom_victim(T1@P1)
wake_oom_reaper(T1@P1) # T1@P1 is enqueued again due to oom_reaper_list == T2@P1 && T1@P1->oom_reaper_list == NULL.
mutex_unlock(&oom_lock)
# Completed processing an OOM victim in a different memcg domain.
spin_lock(&oom_reaper_lock)
# T1P1 is dequeued.
spin_unlock(&oom_reaper_lock)
but memcg's group oom killing made it easier to trigger this bug by
calling wake_oom_reaper() on the same task from one out_of_memory()
request.
Fix this bug using an approach used by commit 855b018325737f76 ("oom,
oom_reaper: disable oom_reaper for oom_kill_allocating_task"). As a
side effect of this patch, this patch also avoids enqueuing multiple
threads sharing memory via task_will_free_mem(current) path.
Link: http://lkml.kernel.org/r/e865a044-2c10-9858-f4ef-254bc71d6cc2@i-love.sakura.ne.jp Link: http://lkml.kernel.org/r/5ee34fc6-1485-34f8-8790-903ddabaa809@i-love.sakura.ne.jp Fixes: af8e15cc85a25315 ("oom, oom_reaper: do not enqueue task if it is on the oom_reaper_list head") Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp> Reported-by: Arkadiusz Miskiewicz <arekm@maven.pl> Tested-by: Arkadiusz Miskiewicz <arekm@maven.pl> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Roman Gushchin <guro@fb.com> Cc: Tejun Heo <tj@kernel.org> Cc: Aleksa Sarai <asarai@suse.de> Cc: Jay Kamat <jgkamat@fb.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
hugetlb needs the same fix as faultin_nopage (which was applied in
commit 96312e61282a ("mm/gup.c: teach get_user_pages_unlocked to handle
FOLL_NOWAIT")) or KVM hangs because it thinks the mmap_sem was already
released by hugetlb_fault() if it returned VM_FAULT_RETRY, but it wasn't
in the FOLL_NOWAIT case.
Link: http://lkml.kernel.org/r/20190109020203.26669-2-aarcange@redhat.com Fixes: ce53053ce378 ("kvm: switch get_user_page_nowait() to get_user_pages_unlocked()") Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Tested-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com> Reported-by: "Dr. David Alan Gilbert" <dgilbert@redhat.com> Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com> Reviewed-by: Peter Xu <peterx@redhat.com> Cc: Mike Rapoport <rppt@linux.vnet.ibm.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, exit_ptrace() adds all ptraced tasks in a dead list, then
zap_pid_ns_processes() waits on all tasks in a current pidns, and only
then are tasks from the dead list released.
zap_pid_ns_processes() can get stuck on waiting tasks from the dead
list. In this case, we will have one unkillable process with one or
more dead children.
Thanks to Oleg for the advice to release tasks in find_child_reaper().
Link: http://lkml.kernel.org/r/20190110175200.12442-1-avagin@gmail.com Fixes: 7c8bd2322c7f ("exit: ptrace: shift "reap dead" code from exit_ptrace() to forget_original_parent()") Signed-off-by: Andrei Vagin <avagin@gmail.com> Signed-off-by: Oleg Nesterov <oleg@redhat.com> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The subvol_name is allocated in btrfs_parse_subvol_options and is
consumed and freed in mount_subvol. Add a free to the error paths that
don't call mount_subvol so that it is guaranteed that subvol_name is
freed when an error happens.
Fixes: 312c89fbca06 ("btrfs: cleanup btrfs_mount() using btrfs_mount_root()") Cc: stable@vger.kernel.org # v4.19+ Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When splitting a leaf or node from one of the trees that are modified when
flushing pending block groups (extent, chunk, device and free space trees),
we need to allocate a new tree block, which in turn can result in the need
to allocate a new block group. After allocating the new block group we may
need to flush new block groups that were previously allocated during the
course of the current transaction, which is what may cause a deadlock due
to attempts to write lock twice the same leaf or node, as when splitting
a leaf or node we are holding a write lock on it and its parent node.
The same type of deadlock can also happen when increasing the tree's
height, since we are holding a lock on the existing root while allocating
the tree block to use as the new root node.
An example trace when the deadlock happens during the leaf split path is:
Fix this by preventing the flushing of new blocks groups when splitting a
leaf/node and when inserting a new root node for one of the trees modified
by the flushing operation, similar to what is done when COWing a node/leaf
from on of these trees.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=202383 Reported-by: Eli V <eliventer@gmail.com> CC: stable@vger.kernel.org # 4.4+ Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
According to Asus firmware engineers, the meaning of these codes is only
to notify the OS that the screen brightness has been turned on/off by
the EC. This does not match the meaning of KEY_DISPLAYTOGGLE /
KEY_DISPLAY_OFF, where userspace is expected to change the display
brightness.
Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
When the OS registers to handle events from the display off hotkey the
EC will send a notification with 0x35 for every key press, independent
of the backlight state.
The behavior of this key on Windows, with the ATKACPI driver from Asus
installed, is turning off the backlight of all connected displays with a
fading effect, and any cursor input or key press turning the backlight
back on. The key press or cursor input that wakes up the display is also
passed through to the application under the cursor or under focus.
The key that matches this behavior the closest is KEY_SCREENLOCK.
Signed-off-by: João Paulo Rechi Vita <jprvita@endlessm.com> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Fix potential memory corruption and panic in loopback for IB_WR_SEND
variants.
The code blindly assumes the posted length will fit in the fetched rwqe,
which is not a valid assumption.
Fix by adding a limit test, and triggering the appropriate send completion
and putting the QP in an error state. This mimics the handling for
non-loopback QPs.
Fixes: 15703461533a ("IB/{hfi1, qib, rdmavt}: Move ruc_loopback to rdmavt") Cc: <stable@vger.kernel.org> #v4.20+ Reviewed-by: Michael J. Ruhl <michael.j.ruhl@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Applications that use the stack for execution purposes cause userspace PSM
jobs to fail during mmap().
Both Fortran (non-standard format parsing) and C (callback functions
located in the stack) applications can be written such that stack
execution is required. The linker notes this via the gnu_stack ELF flag.
This causes READ_IMPLIES_EXEC to be set which forces all PROT_READ mmaps
to have PROT_EXEC for the process.
Checking for VM_EXEC bit and failing the request with EPERM is overly
conservative and will break any PSM application using executable stacks.
Cc: <stable@vger.kernel.org> #v4.14+ Fixes: 12220267645c ("IB/hfi: Protect against writable mmap") Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Reviewed-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com> Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The vma->vm_mm can become impossible to get before rdma_umap_close() is
called, in this case we must not try to get an mm that is already
undergoing process exit. In this case there is no need to wait for
anything as the VMA will be destroyed by another thread soon and is
already effectively 'unreachable' by userspace.
The async_file might be freed before the disassociation has been ended,
causing qp shutdown to use after free on it.
Since uverbs_destroy_ufile_hw is not a fence, it returns if a
disassociation is ongoing in another thread. It has to be written this way
to avoid deadlock. However this means that the ufile FD close cannot
destroy anything that may still be used by an active kref, such as the the
async_file.
To fix that move the kref_put() to be in ib_uverbs_release_file().
Cc: <stable@vger.kernel.org> # 4.2 Fixes: 036b10635739 ("IB/uverbs: Enable device removal when there are active user space applications") Signed-off-by: Yishai Hadas <yishaih@mellanox.com> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When the trigger=off is passed for a PCM OSS stream, it sets the
start_threshold of the given substream to the boundary size, so that
it won't be automatically started. This can be problematic for a
capture stream, unfortunately, as detected by syzkaller. The scenario
is like the following:
- In __snd_pcm_lib_xfer() that is invoked from snd_pcm_oss_read()
loop, we have a check whether the stream was already started or the
stream can be auto-started.
- The function at this check returns 0 with trigger=off since we
explicitly disable the auto-start.
- The loop continues and repeats calling __snd_pcm_lib_xfer() tightly,
which may lead to an RCU stall.
This patch fixes the bug by simply allowing the wait for non-started
stream in the case of OSS capture. For native usages, it's supposed
to be done by the caller side (which is user-space), hence it returns
zero like before.
(In theory, __snd_pcm_lib_xfer() could wait even for the native API
usage cases, too; but I'd like to stay in a safer side for not
breaking the existing stuff for now.)
The hp_pin value that is referred in alc294_hp_init() is always zero
at the moment the function gets called, hence this is actually
useless as in the current code.
And, this kind of init sequence should be called from the codec init
callback, instead of the parser function. So, the first fix in this
patch to move the call call into its own init_hook.
OTOH, this function is needed to be called only once after the boot,
and it'd take too long for invoking at each resume (where the init
callback gets called). So we add a new flag and invoke this only
once as an additional fix.
The one case is still not covered, though: S4 resume. But this
change itself won't lead to any regression in that regard, so we
leave S4 issue as is for now and fix it later. -- tiwai ]
This patch adds quirk VID/PID IDs for the Opus #3 DAP (made by 'The Bit')
in order to enable Native DSD support.
[ NOTE: this could be handled in the generic way with fp->dvd_raw if
we add 0x10cb to the vendor whitelist, but since 0x10cb shows a
different vendor name (Erantech), put to the individual entry at
this time -- tiwai ]
The BCM2835 MMC host driver requests a DMA channel on probe but neglects
to release the channel in the probe error path. The channel may
therefore be leaked, in particular if devm_clk_get() causes probe
deferral. Fix it.
Fixes: 660fc733bd74 ("mmc: bcm2835: Add new driver for the sdhost controller.") Signed-off-by: Lukas Wunner <lukas@wunner.de> Cc: stable@vger.kernel.org # v4.12+ Cc: Frank Pavlic <f.pavlic@kunbus.de> Tested-by: Stefan Wahren <stefan.wahren@i2se.com> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It turns out that the fix can lead to a ~20 percent performance regression
in initial writes to the page cache according to iozone. Let's revert this
for now to have more time for a proper fix.
Cc: stable@vger.kernel.org # v3.13+ Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com> Signed-off-by: Bob Peterson <rpeterso@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When multiple instances of pcf857x chips are present, a fix up
message [1] is printed during the probe of the 2nd and later
instances.
The issue is that the driver is using the same irq_chip data
structure between multiple instances.
Fix this by allocating the irq_chip data structure per instance.
[1] fix up message addressed by this patch
[ 1.212100] gpio gpiochip9: (pcf8575): detected irqchip that is shared with multiple gpiochips: please fix the driver.
During resume hibernate restores all physical memory. Any memory
that is accessed with the MMU disabled needs to be cleaned to the
PoC.
KVMs __hyp_text was previously ommitted as it runs with the MMU
enabled, but now that the hyp-stub is located in this section,
we must clean __hyp_text too.
This ensures secondary CPUs that come online after hibernate
has finished resuming, and load KVM via the freshly written
hyp-stub see the correct instructions.
Signed-off-by: James Morse <james.morse@arm.com> Cc: stable@vger.kernel.org Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The hyp-stub is loaded by the kernel's early startup code at EL2
during boot, before KVM takes ownership later. The hyp-stub's
text is part of the regular kernel text, meaning it can be kprobed.
A breakpoint in the hyp-stub causes the CPU to spin in el2_sync_invalid.
Add it to the __hyp_text.
Signed-off-by: James Morse <james.morse@arm.com> Cc: stable@vger.kernel.org Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 3b8c9f1cdfc5 ("arm64: IPI each CPU after invalidating the I-cache
for kernel mappings") was aimed at fixing the I-cache invalidation for
kernel mappings. However, it inadvertently caused all cache maintenance
for user mappings via set_pte_at() -> __sync_icache_dcache() ->
sync_icache_aliases() to call kick_all_cpus_sync().
Reported-by: Shijith Thotton <sthotton@marvell.com> Tested-by: Shijith Thotton <sthotton@marvell.com> Reported-by: Wandun Chen <chenwandun@huawei.com> Fixes: 3b8c9f1cdfc5 ("arm64: IPI each CPU after invalidating the I-cache for kernel mappings") Cc: <stable@vger.kernel.org> # 4.19.x- Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 1598ecda7b23 ("arm64: kaslr: ensure randomized quantities are
clean to the PoC") added cache maintenance to ensure that global
variables set by the kaslr init routine are not wiped clean due to
cache invalidation occurring during the second round of page table
creation.
However, if kaslr_early_init() exits early with no randomization
being applied (either due to the lack of a seed, or because the user
has disabled kaslr explicitly), no cache maintenance is performed,
leading to the same issue we attempted to fix earlier, as far as the
module_alloc_base variable is concerned.
Note that module_alloc_base cannot be initialized statically, because
that would cause it to be subject to a R_AARCH64_RELATIVE relocation,
causing it to be overwritten by the second round of KASLR relocation
processing.
Fixes: f80fb3a3d508 ("arm64: add support for kernel ASLR") Cc: <stable@vger.kernel.org> # v4.6+ Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Originally, cns3xxx used its own functions for mapping, reading and
writing config registers.
Commit 802b7c06adc7 ("ARM: cns3xxx: Convert PCI to use generic config
accessors") removed the internal PCI config write function in favor of
the generic one:
cns3xxx_pci_write_config() expected aligned addresses, being produced by
cns3xxx_pci_map_bus() while the generic one pci_generic_config_write()
actually expects the real address as both the function and hardware are
capable of byte-aligned writes.
This currently leads to pci_generic_config_write() writing to the wrong
registers.
For instance, upon ath9k module loading:
- driver ath9k gets loaded
- The driver wants to write value 0xA8 to register PCI_LATENCY_TIMER,
located at 0x0D
- cns3xxx_pci_map_bus() aligns the address to 0x0C
- pci_generic_config_write() effectively writes 0xA8 into register 0x0C
(CACHE_LINE_SIZE)
Fix the bug by removing the alignment in the cns3xxx mapping function.
Fixes: 802b7c06adc7 ("ARM: cns3xxx: Convert PCI to use generic config accessors") Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
[lorenzo.pieralisi@arm.com: updated commit log] Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Acked-by: Krzysztof Halasa <khalasa@piap.pl> Acked-by: Tim Harvey <tharvey@gateworks.com> Acked-by: Arnd Bergmann <arnd@arndb.de> CC: stable@vger.kernel.org # v4.0+ CC: Bjorn Helgaas <bhelgaas@google.com> CC: Olof Johansson <olof@lixom.net> CC: Robin Leblon <robin.leblon@ncentric.com> CC: Rob Herring <robh@kernel.org> CC: Russell King <linux@armlinux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Passing EPERM during syscall skipping was confusing since the test wasn't
actually exercising the errno evaluation -- it was just passing a literal
"1" (EPERM). Instead, expand the tests to check both direct value returns
(positive, 45000 in this case), and errno values (negative, -ESRCH in this
case) to check both fake success and fake failure during syscall skipping.
Reported-by: Colin Ian King <colin.king@canonical.com> Fixes: a33b2d0359a0 ("selftests/seccomp: Add tests for basic ptrace actions") Cc: stable@vger.kernel.org Signed-off-by: Kees Cook <keescook@chromium.org> Signed-off-by: Shuah Khan <shuah@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 9d3a4de4cb8d ("iommu: Disambiguate MSI region types") changed
the reserved region type in intel_iommu_get_resv_regions() from
IOMMU_RESV_RESERVED to IOMMU_RESV_MSI, but it forgot to also change
the type in intel_iommu_put_resv_regions().
This leads to a memory leak, because now the check in
intel_iommu_put_resv_regions() for IOMMU_RESV_RESERVED will never
be true, and no allocated regions will be freed.
Fix this by changing the region type in intel_iommu_put_resv_regions()
to IOMMU_RESV_MSI, matching the type of the allocated regions.
The nr_dentry_unused per-cpu counter tracks dentries in both the LRU
lists and the shrink lists where the DCACHE_LRU_LIST bit is set.
The shrink_dcache_sb() function moves dentries from the LRU list to a
shrink list and subtracts the dentry count from nr_dentry_unused. This
is incorrect as the nr_dentry_unused count will also be decremented in
shrink_dentry_list() via d_shrink_del().
To fix this double decrement, the decrement in the shrink_dcache_sb()
function is taken out.
Fixes: 4e717f5c1083 ("list_lru: remove special case function list_lru_dispose_all." Cc: stable@kernel.org Signed-off-by: Waiman Long <longman@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When doing reads beyound the end of a file the server returns
error STATUS_END_OF_FILE error which is mapped to -ENODATA.
Currently we report it as a failure which confuses read stats.
Change it to not consider -ENODATA as failure for stat purposes.
Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Steve French <stfrench@microsoft.com> CC: Stable <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently we log success once we send an async IO request to
the server. Instead we need to analyse a response and then log
success or failure for a particular command. Also fix argument
list for read logging.
Cc: <stable@vger.kernel.org> # 4.18 Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Steve French <stfrench@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Allocation of a page array for non-cached IO was separated from
allocation of rdata and wdata structures and this introduced memory
leaks and a possible null pointer dereference. This patch fixes
these problems.
Cc: <stable@vger.kernel.org> Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Steve French <stfrench@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Steve French <stfrench@microsoft.com> CC: Stable <stable@vger.kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
IPv6 does not consider if the socket is bound to a device when binding
to an address. The result is that a socket can be bound to eth0 and then
bound to the address of eth1. If the device is a VRF, the result is that
a socket can only be bound to an address in the default VRF.
Resolve by considering the device if sk_bound_dev_if is set.
This problem exists from the beginning of git history.
Signed-off-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
I had to fold commit df133f3f9625 ("virtio_net: bulk free tx skbs")
into this to make it work. ]
We do not reset or free up unused buffers when enabling/disabling XDP,
so it can happen that xdp_frames are freed after disabling XDP or
sk_buffs are freed after enabling XDP on xdp tx queues.
Thus we need to handle both forms (xdp_frames and sk_buffs) regardless
of XDP setting.
One way to trigger this problem is to disable XDP when napi_tx is
enabled. In that case, virtnet_xdp_set() calls virtnet_napi_enable()
which kicks NAPI. The NAPI handler will call virtnet_poll_cleantx()
which invokes free_old_xmit_skbs() for queues which have been used by
XDP.
Note that even with this change we need to keep skipping
free_old_xmit_skbs() from NAPI handlers when XDP is enabled, because XDP
tx queues do not aquire queue locks.
- v2: Use napi_consume_skb() instead of dev_consume_skb_any()
Fixes: 4941d472bf95 ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
put_page() can work as a fallback for freeing xdp_frames, but the
appropriate way is to use xdp_return_frame().
Fixes: cac320c850ef ("virtio_net: convert to use generic xdp_frame and xdp_return_frame API") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Jesper Dangaard Brouer <brouer@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 8dcc5b0ab0ec ("virtio_net: fix ndo_xdp_xmit crash towards dev not
ready for XDP") tried to avoid access to unexpected sq while XDP is
disabled, but was not complete.
There was a small window which causes out of bounds sq access in
virtnet_xdp_xmit() while disabling XDP.
An example case of
- curr_queue_pairs = 6 (2 for SKB and 4 for XDP)
- online_cpu_num = xdp_queue_paris = 4
when XDP is enabled:
CPU 0 CPU 1
(Disabling XDP) (Processing redirected XDP frames)
virtnet_xdp_xmit()
virtnet_xdp_set()
_virtnet_set_queues()
set curr_queue_pairs (2)
check if rq->xdp_prog is not NULL
virtnet_xdp_sq(vi)
qp = curr_queue_pairs -
xdp_queue_pairs +
smp_processor_id()
= 2 - 4 + 1 = -1
sq = &vi->sq[qp] // out of bounds access
set xdp_queue_pairs (0)
rq->xdp_prog = NULL
Basically we should not change curr_queue_pairs and xdp_queue_pairs
while someone can read the values. Thus, when disabling XDP, assign NULL
to rq->xdp_prog first, and wait for RCU grace period, then change
xxx_queue_pairs.
Note that we need to keep the current order when enabling XDP though.
- v2: Make rcu_assign_pointer/synchronize_net conditional instead of
_virtnet_set_queues.
Fixes: 186b3c998c50 ("virtio-net: support XDP_REDIRECT") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When XDP is disabled, curr_queue_pairs + smp_processor_id() can be
larger than max_queue_pairs.
There is no guarantee that we have enough XDP send queues dedicated for
each cpu when XDP is disabled, so do not count drops on sq in that case.
Fixes: 5b8f3c8d30a6 ("virtio_net: Add XDP related stats") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When _virtnet_set_queues() failed we did not restore real_num_rx_queues.
Fix this by placing the change of real_num_rx_queues after
_virtnet_set_queues().
This order is also in line with virtnet_set_channels().
Fixes: 4941d472bf95 ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When napi_tx is enabled, virtnet_poll_cleantx() called
free_old_xmit_skbs() even for xdp send queue.
This is bogus since the queue has xdp_frames, not sk_buffs, thus mangled
device tx bytes counters because skb->len is meaningless value, and even
triggered oops due to general protection fault on freeing them.
Since xdp send queues do not aquire locks, old xdp_frames should be
freed only in virtnet_xdp_xmit(), so just skip free_old_xmit_skbs() for
xdp send queues.
Similarly virtnet_poll_tx() called free_old_xmit_skbs(). This NAPI
handler is called even without calling start_xmit() because cb for tx is
by default enabled. Once the handler is called, it enabled the cb again,
and then the handler would be called again. We don't need this handler
for XDP, so don't enable cb as well as not calling free_old_xmit_skbs().
Also, we need to disable tx NAPI when disabling XDP, so
virtnet_poll_tx() can safely access curr_queue_pairs and
xdp_queue_pairs, which are not atomically updated while disabling XDP.
Fixes: b92f1e6751a6 ("virtio-net: transmit napi") Fixes: 7b0411ef4aa6 ("virtio-net: clean tx descriptors from rx napi") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 4e09ff536284 ("virtio-net: disable NAPI only when enabled during
XDP set") tried to fix inappropriate NAPI enabling/disabling when
!netif_running(), but was not complete.
On error path virtio_net could enable NAPI even when !netif_running().
This can cause enabling NAPI twice on virtnet_open(), which would
trigger BUG_ON() in napi_enable().
Fixes: 4941d472bf95b ("virtio-net: do not reset during XDP set") Signed-off-by: Toshiaki Makita <makita.toshiaki@lab.ntt.co.jp> Acked-by: Jason Wang <jasowang@redhat.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
aead_request_set_crypt takes an iv pointer, and we change the iv
soon after setting it. Some async crypto algorithms don't save the iv,
so we need to save it in the tls_rec for async requests.
Found by hardcoding x64 aesni to use async crypto manager (to test the async
codepath), however I don't think this combination can happen in the wild.
Presumably other hardware offloads will need this fix, but there have been
no user reports.
Fixes: a42055e8d2c30 ("Add support for async encryption of records...") Signed-off-by: Dave Watson <davejwatson@fb.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If there are outstanding async tx requests (when crypto returns EINPROGRESS),
there is a potential deadlock: the tx work acquires the lock, while we
cancel_delayed_work_sync() while holding the lock. Drop the lock while waiting
for the work to complete.
Fixes: a42055e8d2c30 ("Add support for async encryption of records...") Signed-off-by: Dave Watson <davejwatson@fb.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Now sctp_transport_pmtu() passes transport->saddr into .get_dst() to set
flow sport from 'saddr'. However, transport->saddr is set only when
transport->dst exists in sctp_transport_route().
If sctp_transport_pmtu() is called without transport->saddr set, like
when transport->dst doesn't exists, the flow sport will be set to 0
from transport->saddr, which will cause a wrong route to be got.
Commit 6e91b578bf3f ("sctp: re-use sctp_transport_pmtu in
sctp_transport_route") made the issue be triggered more easily
since sctp_transport_pmtu() would be called in sctp_transport_route()
after that.
In gerneral, fl4->fl4_sport should always be set to
htons(asoc->base.bind_addr.port), unless transport->asoc doesn't exist
in sctp_v4/6_get_dst(), which is the case:
sctp_ootb_pkt_new() ->
sctp_transport_route()
For that, we can simply handle it by setting flow sport from saddr only
when it's 0 in sctp_v4/6_get_dst().
Fixes: 6e91b578bf3f ("sctp: re-use sctp_transport_pmtu in sctp_transport_route") Reported-by: Ying Xu <yinxu@redhat.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The new chunk 'retval' transport is set from the incoming chunk 'chunk'
transport. However, 'retval' transport belong to the new asoc, which
is a different one from 'chunk' transport's asoc.
It will cause that the 'retval' chunk gets set with a wrong transport.
Later when sending it and because of Commit b9fd683982c9 ("sctp: add
sctp_packet_singleton"), sctp_packet_singleton() will set some fields,
like vtag to 'retval' chunk from that wrong transport's asoc.
This patch is to fix it by setting 'retval' transport correctly which
belongs to the right asoc in sctp_make_init_ack() and
sctp_sf_do_5_1D_ce().
Fixes: b9fd683982c9 ("sctp: add sctp_packet_singleton") Reported-by: Ying Xu <yinxu@redhat.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
With the original commit, eswitch instance will not be initialized for
a function which is vport group manager but not eswitch manager such as
host PF on SmartNIC (BlueField) card. This will result in a kernel crash
when such a vport group manager is trying to access vports in its group.
E.g, PF vport manager (not eswitch manager) tries to configure the MAC
of its VF vport, a kernel trace will happen similar as bellow:
BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
...
RIP: 0010:mlx5_eswitch_get_vport_config+0xc/0x180 [mlx5_core]
...
Fixes: 5f5991f36dce ("net/mlx5e: E-Switch, Initialize eswitch only if eswitch manager") Signed-off-by: Bodong Wang <bodong@mellanox.com> Reported-by: Yuval Avnery <yuvalav@mellanox.com> Reviewed-by: Daniel Jurgens <danielj@mellanox.com> Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When the MC route socket is closed, mroute_clean_tables() is called to
cleanup existing routes. Mistakenly notifiers call was put on the cleanup
of the unresolved MC route entries cache.
In a case where the MC socket closes before an unresolved route expires,
the notifier call leads to a crash, caused by the driver trying to
increment a non initialized refcount_t object [1] and then when handling
is done, to decrement it [2]. This was detected by a test recently added in
commit 6d4efada3b82 ("selftests: forwarding: Add multicast routing test").
Fix that by putting notifiers call on the resolved entries traversal,
instead of on the unresolved entries traversal.
This patch is to improve sctp stream adding events in 2 places:
1. In sctp_process_strreset_addstrm_out(), move up SCTP_MAX_STREAM
and in stream allocation failure checks, as the adding has to
succeed after reconf_timer stops for the in stream adding
request retransmission.
3. In sctp_process_strreset_addstrm_in(), no event should be sent,
as no in or out stream is added here.
Fixes: 50a41591f110 ("sctp: implement receiver-side procedures for the Add Outgoing Streams Request Parameter") Fixes: c5c4ebb3ab87 ("sctp: implement receiver-side procedures for the Add Incoming Streams Request Parameter") Reported-by: Ying Xu <yinxu@redhat.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As Erspan_v4, Erspan_v6 protocol relies on o_key to configure
session id header field. However TUNNEL_KEY bit is cleared in
ip6erspan_tunnel_xmit since ERSPAN protocol does not set the key field
of the external GRE header and so the configured o_key is not reported
to userspace. The issue can be triggered with the following reproducer:
$ip link add ip6erspan1 type ip6erspan local 2000::1 remote 2000::2 \
key 1 seq erspan_ver 1
$ip link set ip6erspan1 up
ip -d link sh ip6erspan1
ip6erspan1@NONE: <BROADCAST,MULTICAST> mtu 1422 qdisc noop state DOWN mode DEFAULT
link/ether ba:ff:09:24:c3:0e brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 68 maxmtu 1500
ip6erspan remote 2000::2 local 2000::1 encaplimit 4 flowlabel 0x00000 ikey 0.0.0.1 iseq oseq
Fix the issue adding TUNNEL_KEY bit to the o_flags parameter in
ip6gre_fill_info
Fixes: 5a963eb61b7c ("ip6_gre: Add ERSPAN native tunnel support") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After batched used ring updating was introduced in commit e2b3b35eb989
("vhost_net: batch used ring update in rx"). We tend to batch heads in
vq->heads for more than one packet. But the quota passed to
get_rx_bufs() was not correctly limited, which can result a OOB write
in vq->heads.
UIO_MAXIOV was still used which is wrong since we could have batched
used in vq->heads, this will cause OOB if the next buffer needs more
than 960 (1024 (UIO_MAXIOV) - 64 (VHOST_NET_BATCH)) heads after we've
batched 64 (VHOST_NET_BATCH) heads: Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
=============================================================================
BUG kmalloc-8k (Tainted: G B ): Redzone overwritten
-----------------------------------------------------------------------------
Fixing this by allocating UIO_MAXIOV + VHOST_NET_BATCH iovs for
vhost-net. This is done through set the limitation through
vhost_dev_init(), then set_owner can allocate the number of iov in a
per device manner.
This fixes CVE-2018-16880.
Fixes: e2b3b35eb989 ("vhost_net: batch used ring update in rx") Signed-off-by: Jason Wang <jasowang@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After a timeout event caused by for example a broadcast storm, when
the MAC and PHY are reset, the BQL TX queue needs to be reset as
well. Otherwise, the device will exhibit severe performance issues
even after the storm has ended.
Co-authored-by: David Gounaris <david.gounaris@infinera.com> Signed-off-by: Mathias Thore <mathias.thore@infinera.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Call tun_set_real_num_queues() after the increment of tun->numqueues
since the former depends on it. Otherwise, the number of queues is not
correctly accounted for, which results to warnings similar to:
"vnet0 selects TX queue 11, but real number of TX queues is 11".
Fixes: 0b7959b62573 ("tun: publish tfile after it's fully initialized") Reported-and-tested-by: George Amanakis <gamanakis@gmail.com> Signed-off-by: George Amanakis <gamanakis@gmail.com> Signed-off-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch is to improve sctp stream reset events in 4 places:
1. In sctp_process_strreset_outreq(), the flag should always be set with
SCTP_STREAM_RESET_INCOMING_SSN instead of OUTGOING, as receiver's in
stream is reset here.
2. In sctp_process_strreset_outreq(), move up SCTP_STRRESET_ERR_WRONG_SSN
check, as the reset has to succeed after reconf_timer stops for the
in stream reset request retransmission.
3. In sctp_process_strreset_inreq(), no event should be sent, as no in
or out stream is reset here.
4. In sctp_process_strreset_resp(), SCTP_STREAM_RESET_INCOMING_SSN or
OUTGOING event should always be sent for stream reset requests, no
matter it fails or succeeds to process the request.
Fixes: 810544764536 ("sctp: implement receiver-side procedures for the Outgoing SSN Reset Request Parameter") Fixes: 16e1a91965b0 ("sctp: implement receiver-side procedures for the Incoming SSN Reset Request Parameter") Fixes: 11ae76e67a17 ("sctp: implement receiver-side procedures for the Reconf Response Parameter") Reported-by: Ying Xu <yinxu@redhat.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
EtherAVB may provide a checksum of packet data appended to packet data. In
order to allow this checksum to be received by the host descriptor data
needs to be enlarged by 2 bytes to accommodate the checksum.
In the case of MTU-sized packets without a VLAN tag the
checksum were already accommodated by virtue of the space reserved for the
VLAN tag. However, a packet of MTU-size with a VLAN tag consumed all
packet data space provided by a descriptor leaving no space for the
trailing checksum.
This was not detected by the driver which incorrectly used the last two
bytes of packet data as the checksum and truncate the packet by two bytes.
This resulted all such packets being dropped.
A work around is to disable RX checksum offload
# ethtool -K eth0 rx off
This patch resolves this problem by increasing the size available for
packet data in RX descriptors by two bytes.
Tested on R-Car E3 (r8a77990) ES1.0 based Ebisu-4D board
v2
* Use sizeof(__sum16) directly rather than adding a driver-local
#define for the size of the checksum provided by the hw (2 bytes).
Fixes: 4d86d3818627 ("ravb: RX checksum offload") Signed-off-by: Simon Horman <horms+renesas@verge.net.au> Reviewed-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Assign a default net namespace to netdevs created by init_dummy_netdev().
Fixes a NULL pointer dereference caused by busy-polling a socket bound to
an iwlwifi wireless device, which bumps the per-net BUSYPOLLRXPACKETS stat
if napi_poll() received packets:
BUG: unable to handle kernel NULL pointer dereference at 0000000000000190
IP: napi_busy_loop+0xd6/0x200
Call Trace:
sock_poll+0x5e/0x80
do_sys_poll+0x324/0x5a0
SyS_poll+0x6c/0xf0
do_syscall_64+0x6b/0x1f0
entry_SYSCALL_64_after_hwframe+0x3d/0xa2
Fixes: 7db6b048da3b ("net: Commonize busy polling code to focus on napi_id instead of socket") Signed-off-by: Josh Elsasser <jelsasser@appneta.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When an internally generated frame is handled by rose_xmit(),
rose_route_frame() is called:
if (!rose_route_frame(skb, NULL)) {
dev_kfree_skb(skb);
stats->tx_errors++;
return NETDEV_TX_OK;
}
We have the same code sequence in Net/Rom where an internally generated
frame is handled by nr_xmit() calling nr_route_frame(skb, NULL).
However, in this function NULL argument is tested while it is not in
rose_route_frame().
Then kernel panic occurs later on when calling ax25cmp() with a NULL
ax25_cb argument as reported many times and recently with syzbot.
We need to test if ax25 is NULL before using it.
Testing:
Built kernel with CONFIG_ROSE=y.
Signed-off-by: Bernard Pidoux <f6bvp@free.fr> Acked-by: Dmitry Vyukov <dvyukov@google.com> Reported-by: syzbot+1a2c456a1ea08fa5b5f7@syzkaller.appspotmail.com Cc: "David S. Miller" <davem@davemloft.net> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Bernard Pidoux <f6bvp@free.fr> Cc: linux-hams@vger.kernel.org Cc: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Driver reads the query HCA capabilities without the corresponding masks.
Without the correct masks, the base addresses of the queues are
unaligned. In addition some reserved bits were wrongly read. Using the
correct masks, ensures alignment of the base addresses and allows future
firmware versions safe use of the reserved bits.
Fixes: ab9c17a009ee ("mlx4_core: Modify driver initialization flow to accommodate SRIOV for Ethernet") Fixes: 0ff1fb654bec ("{NET, IB}/mlx4: Add device managed flow steering firmware API") Signed-off-by: Aya Levin <ayal@mellanox.com> Signed-off-by: Tariq Toukan <tariqt@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
in6_dump_addrs() returns a positive 1 if there was nothing to dump.
This return value can not be passed as return from inet6_dump_addr()
as is, because it will confuse rtnetlink, resulting in NLMSG_DONE
never getting set:
$ ip addr list dev lo
EOF on netlink
Dump terminated
v2: flip condition to avoid a new goto (DaveA)
Fixes: 7c1e8a3817c5 ("netlink: fixup regression in RTM_GETADDR") Reported-by: Brendan Galloway <brendan.galloway@netronome.com> Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by: David Ahern <dsahern@gmail.com> Tested-by: David Ahern <dsahern@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use ERSPAN key header field as tunnel key in gre_parse_header routine
since ERSPAN protocol sets the key field of the external GRE header to
0 resulting in a tunnel lookup fail in ip6gre_err.
In addition remove key field parsing and pskb_may_pull check in
erspan_rcv and ip6erspan_rcv
Fixes: 5a963eb61b7c ("ip6_gre: Add ERSPAN native tunnel support") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Erspan protocol (version 1 and 2) relies on o_key to configure
session id header field. However TUNNEL_KEY bit is cleared in
erspan_xmit since ERSPAN protocol does not set the key field
of the external GRE header and so the configured o_key is not reported
to userspace. The issue can be triggered with the following reproducer:
$ip link add erspan1 type erspan local 192.168.0.1 remote 192.168.0.2 \
key 1 seq erspan_ver 1
$ip link set erspan1 up
$ip -d link sh erspan1
Fix the issue adding TUNNEL_KEY bit to the o_flags parameter in
ipgre_fill_info
Fixes: 84e54fe0a5ea ("gre: introduce native tunnel support for ERSPAN") Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use pskb_may_pull() to make sure the optional fields are in skb linear
parts, so we can safely read them later.
It's easy to reproduce the issue with a net driver that supports paged
skb data. Just create a L2TPv3 over IP tunnel and then generates some
network traffic.
Once reproduced, rx err in /sys/kernel/debug/l2tp/tunnels will increase.
Changes in v4:
1. s/l2tp_v3_pull_opt/l2tp_v3_ensure_opt_in_linear/
2. s/tunnel->version != L2TP_HDR_VER_2/tunnel->version == L2TP_HDR_VER_3/
3. Add 'Fixes' in commit messages.
Changes in v3:
1. To keep consistency, move the code out of l2tp_recv_common.
2. Use "net" instead of "net-next", since this is a bug fix.
Changes in v2:
1. Only fix L2TPv3 to make code simple.
To fix both L2TPv3 and L2TPv2, we'd better refactor l2tp_recv_common.
It's complicated to do so.
2. Reloading pointers after pskb_may_pull
Fixes: f7faffa3ff8e ("l2tp: Add L2TPv3 protocol support") Fixes: 0d76751fad77 ("l2tp: Add L2TPv3 IP encapsulation (no UDP) support") Fixes: a32e0eec7042 ("l2tp: introduce L2TPv3 IP encapsulation support for IPv6") Signed-off-by: Jacob Wen <jian.w.wen@oracle.com> Acked-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The size of L2TPv2 header with all optional fields is 14 bytes.
l2tp_udp_recv_core only moves 10 bytes to the linear part of a
skb. This may lead to l2tp_recv_common read data outside of a skb.
This patch make sure that there is at least 14 bytes in the linear
part of a skb to meet the maximum need of l2tp_udp_recv_core and
l2tp_recv_common. The minimum size of both PPP HDLC-like frame and
Ethernet frame is larger than 14 bytes, so we are safe to do so.
Also remove L2TP_HDR_SIZE_NOSEQ, it is unused now.
Fixes: fd558d186df2 ("l2tp: Split pppol2tp patch into separate l2tp and ppp parts") Suggested-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: Jacob Wen <jian.w.wen@oracle.com> Acked-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
While implementing ipvlan l3 and l3s mode for kubernetes CNI plugin,
I ran into the issue that while l3 mode is working fine, l3s mode
does not have any connectivity to kube-apiserver and hence all pods
end up in Error state as well. The ipvlan master device sits on
top of a bond device and hostns traffic to kube-apiserver (also running
in hostns) is DNATed from 10.152.183.1:443 to 139.178.29.207:37573
where the latter is the address of the bond0. While in l3 mode, a
curl to https://10.152.183.1:443 or to https://139.178.29.207:37573
works fine from hostns, neither of them do in case of l3s. In the
latter only a curl to https://127.0.0.1:37573 appeared to work where
for local addresses of bond0 I saw kernel suddenly starting to emit
ARP requests to query HW address of bond0 which remained unanswered
and neighbor entries in INCOMPLETE state. These ARP requests only
happen while in l3s.
Debugging this further, I found the issue is that l3s mode is piggy-
backing on l3 master device, and in this case local routes are using
l3mdev_master_dev_rcu(dev) instead of net->loopback_dev as per commit f5a0aab84b74 ("net: ipv4: dst for local input routes should use l3mdev
if relevant") and 5f02ce24c269 ("net: l3mdev: Allow the l3mdev to be
a loopback"). I found that reverting them back into using the
net->loopback_dev fixed ipvlan l3s connectivity and got everything
working for the CNI.
Now judging from 4fbae7d83c98 ("ipvlan: Introduce l3s mode") and the
l3mdev paper in [0] the only sole reason why ipvlan l3s is relying
on l3 master device is to get the l3mdev_ip_rcv() receive hook for
setting the dst entry of the input route without adding its own
ipvlan specific hacks into the receive path, however, any l3 domain
semantics beyond just that are breaking l3s operation. Note that
ipvlan also has the ability to dynamically switch its internal
operation from l3 to l3s for all ports via ipvlan_set_port_mode()
at runtime. In any case, l3 vs l3s soley distinguishes itself by
'de-confusing' netfilter through switching skb->dev to ipvlan slave
device late in NF_INET_LOCAL_IN before handing the skb to L4.
Minimal fix taken here is to add a IFF_L3MDEV_RX_HANDLER flag which,
if set from ipvlan setup, gets us only the wanted l3mdev_l3_rcv() hook
without any additional l3mdev semantics on top. This should also have
minimal impact since dev->priv_flags is already hot in cache. With
this set, l3s mode is working fine and I also get things like
masquerading pod traffic on the ipvlan master properly working.
Fixes: f5a0aab84b74 ("net: ipv4: dst for local input routes should use l3mdev if relevant") Fixes: 5f02ce24c269 ("net: l3mdev: Allow the l3mdev to be a loopback") Fixes: 4fbae7d83c98 ("ipvlan: Introduce l3s mode") Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Mahesh Bandewar <maheshb@google.com> Cc: David Ahern <dsa@cumulusnetworks.com> Cc: Florian Westphal <fw@strlen.de> Cc: Martynas Pumputis <m@lambda.lt> Acked-by: David Ahern <dsa@cumulusnetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
skb->cb may contain data from previous layers (in an observed case
IPv4 with L3 Master Device). In the observed scenario, the data in
IPCB(skb)->frags was misinterpreted as IP6CB(skb)->frag_max_size,
eventually caused an unexpected IPv6 fragmentation in ip6_fragment()
through ip6_finish_output().
This patch clears IP6CB(skb), which potentially contains garbage data,
on the SRH ip4ip6 encapsulation.
Fixes: 32d99d0b6702 ("ipv6: sr: add support for ip4ip6 encapsulation") Signed-off-by: Yohei Kanemaru <yohei.kanemaru@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When debugfs is disabled, but coredump is turned on, the adreno driver fails to build:
drivers/gpu/drm/msm/adreno/a3xx_gpu.c:460:4: error: 'struct msm_gpu_funcs' has no member named 'show'
.show = adreno_show,
^~~~
drivers/gpu/drm/msm/adreno/a3xx_gpu.c:460:11: note: (near initialization for 'funcs.base')
drivers/gpu/drm/msm/adreno/a3xx_gpu.c:460:11: error: initialization of 'void (*)(struct msm_gpu *, struct msm_gem_submit *, struct msm_file_private *)' from incompatible pointer type 'void (*)(struct msm_gpu *, struct msm_gpu_state *, struct drm_printer *)' [-Werror=incompatible-pointer-types]
drivers/gpu/drm/msm/adreno/a3xx_gpu.c:460:11: note: (near initialization for 'funcs.base.submit')
drivers/gpu/drm/msm/adreno/a4xx_gpu.c:546:4: error: 'struct msm_gpu_funcs' has no member named 'show'
drivers/gpu/drm/msm/adreno/a5xx_gpu.c:1460:4: error: 'struct msm_gpu_funcs' has no member named 'show'
drivers/gpu/drm/msm/adreno/a6xx_gpu.c:769:4: error: 'struct msm_gpu_funcs' has no member named 'show'
drivers/gpu/drm/msm/msm_gpu.c: In function 'msm_gpu_devcoredump_read':
drivers/gpu/drm/msm/msm_gpu.c:289:12: error: 'const struct msm_gpu_funcs' has no member named 'show'
Adjust the #ifdef to make it build again.
Fixes: c0fec7f562ec ("drm/msm/gpu: Capture the GPU state on a GPU hang") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Update mt76x0_phy_ant_select() to conform vendor driver, most notably
add dual antenna mode support, read configuration from EEPROM and
move ant select out of channel config to init phase. Plus small MT7630E
quirk for MT_CMB_CTRL register which vendor driver dedicated to this
chip do.
This make MT7630E workable with mt76x0e driver and do not cause any
problems on MT7610U for me.
Acked-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Driver works better for MT7630 without MCU calibration, which
looks like it can hangs the firmware. Vendor driver do not
perform it for MT7630 as well.
Assure that after we initialize dev->cal.low_gain to -1 this
will cause update gain calibration. Otherwise this might or
might not happen depending on value of second bit of low_gain
and values read from registers in mt76x02_phy_adjust_vga_gain().
If we are associated and scanning is performed, sw_scan_complete callback
is done after we get back to operating channel, so we do not perform
queue cal work. Fix this queue cal work from sw_scan_complete().
On mt76x0 we have to restore gain in MT_BBP(AGC, 8) register after
scanning, as it was multiple times modified by channel switch code.
So queue cal work without any delay to set AGC gain value.
Similar like in mt76x2 init AGC gain only when set operating channel
and just check before queuing cal work in sw_scan_complete() if
initialization was already done.
Fixes: bbd10586f0df ("mt76x0: phy: do not run calibration during channel switch") Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
MT_BBP(AGC, 8) register has values depend on band in
mt76x0_bbp_switch_tab, so we should not overwrite other fields
than MT_BBP_AGC_GAIN when setting gain.
This can fix performance issues when connecting to 2.4GHz AP.
Fixes: 4636a2544c3b ("mt76x0: phy: align channel gain logic to mt76x2 one") Acked-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com> Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by: Felix Fietkau <nbd@nbd.name> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
OUT endpoint requests may somtimes have this flag set when
preparing to be submitted to HW indicating that there is an
additional TRB chained to the request for alignment purposes.
If that request is removed before the controller can execute the
transfer (e.g. ep_dequeue/ep_disable), the request will not go
through the dwc3_gadget_ep_cleanup_completed_request() handler
and will not have its needs_extra_trb flag cleared when
dwc3_gadget_giveback() is called. This same request could be
later requeued for a new transfer that does not require an
extra TRB and if it is successfully completed, the cleanup
and TRB reclamation will incorrectly process the additional TRB
which belongs to the next request, and incorrectly advances the
TRB dequeue pointer, thereby messing up calculation of the next
requeust's actual/remaining count when it completes.
The right thing to do here is to ensure that the flag is cleared
before it is given back to the function driver. A good place
to do that is in dwc3_gadget_del_and_unmap_request().
Fixes: c6267a51639b ("usb: dwc3: gadget: align transfers to wMaxPacketSize") Cc: stable@vger.kernel.org Signed-off-by: Jack Pham <jackp@codeaurora.org> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
[jackp: backport to <= 4.20: replaced 'needs_extra_trb' with 'unaligned'
and 'zero' members in patch and reworded commit text] Signed-off-by: Jack Pham <jackp@codeaurora.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The underlying assumption that one sparse section belongs into a single
numa node doesn't hold really. Robert Shteynfeld has reported a boot
failure. The boot log was not captured but his memory layout is as
follows:
This means that node0 starts in the middle of a memory section which is
also in node1. memmap_init_zone tries to initialize padding of a
section even when it is outside of the given pfn range because there are
code paths (e.g. memory hotplug) which assume that the full worth of
memory section is always initialized.
In this particular case, though, such a range is already intialized and
most likely already managed by the page allocator. Scribbling over
those pages corrupts the internal state and likely blows up when any of
those pages gets used.
Reported-by: Robert Shteynfeld <robert.shteynfeld@gmail.com> Fixes: 2830bf6f05fb ("mm, memory_hotplug: initialize struct pages for the full memory section") Cc: stable@kernel.org Signed-off-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The changes to split ring allocation from open/close, broke
the cleanup of subchannels. This resulted in problems using
uio on network devices because the subchannel was left behind
when the network device was unbound.
The cause was in the disconnect logic which used list splice
to move the subchannel list into a local variable. This won't
work because the subchannel list is needed later during the
process of the rescind messages (relid2channel).
The fix is to just leave the subchannel list in place
which is what the original code did. The list is cleaned
up later when the host rescind is processed.
Without the fix, we have a lot of "hang" issues in netvsc when we
try to change the NIC's MTU, set the number of channels, etc.
Fixes: ae6935ed7d42 ("vmbus: split ring buffer allocation from open") Cc: stable@vger.kernel.org Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com> Signed-off-by: Dexuan Cui <decui@microsoft.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Commit d86adf482b84 ("scsi: storvsc: Enable multi-queue support") removed
the usage of the API in Jan 2017, and the API is not used since then.
netvsc and storvsc have their own algorithms to determine the outgoing
channel, so this API is useless.
And the API is potentially unsafe, because it reads primary->num_sc without
any lock held. This can be risky considering the RESCIND-OFFER message.
Let's remove the API.
Cc: Long Li <longli@microsoft.com> Cc: Stephen Hemminger <sthemmin@microsoft.com> Cc: K. Y. Srinivasan <kys@microsoft.com> Cc: Haiyang Zhang <haiyangz@microsoft.com> Signed-off-by: Dexuan Cui <decui@microsoft.com> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Dexuan Cui <decui@microsoft.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
During review I noticed that inner meta map setup for map in
map is buggy in that it does not propagate all needed data
from the reference map which the verifier is later accessing.
In particular one such case is index masking to prevent out of
bounds access under speculative execution due to missing the
map's unpriv_array/index_mask field propagation. Fix this such
that the verifier is generating the correct code for inlined
lookups in case of unpriviledged use.
Before patch (test_verifier's 'map in map access' dump):
While 979d63d50c0c ("bpf: prevent out of bounds speculation on pointer
arithmetic") took care of rejecting alu op on pointer when e.g. pointer
came from two different map values with different map properties such as
value size, Jann reported that a case was not covered yet when a given
alu op is used in both "ptr_reg += reg" and "numeric_reg += reg" from
different branches where we would incorrectly try to sanitize based
on the pointer's limit. Catch this corner case and reject the program
instead.
Fixes: 979d63d50c0c ("bpf: prevent out of bounds speculation on pointer arithmetic") Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Jann reported that the original commit back in b2157399cc98
("bpf: prevent out-of-bounds speculation") was not sufficient
to stop CPU from speculating out of bounds memory access:
While b2157399cc98 only focussed on masking array map access
for unprivileged users for tail calls and data access such
that the user provided index gets sanitized from BPF program
and syscall side, there is still a more generic form affected
from BPF programs that applies to most maps that hold user
data in relation to dynamic map access when dealing with
unknown scalars or "slow" known scalars as access offset, for
example:
- Load a map value pointer into R6
- Load an index into R7
- Do a slow computation (e.g. with a memory dependency) that
loads a limit into R8 (e.g. load the limit from a map for
high latency, then mask it to make the verifier happy)
- Exit if R7 >= R8 (mispredicted branch)
- Load R0 = R6[R7]
- Load R0 = R6[R0]
For unknown scalars there are two options in the BPF verifier
where we could derive knowledge from in order to guarantee
safe access to the memory: i) While </>/<=/>= variants won't
allow to derive any lower or upper bounds from the unknown
scalar where it would be safe to add it to the map value
pointer, it is possible through ==/!= test however. ii) another
option is to transform the unknown scalar into a known scalar,
for example, through ALU ops combination such as R &= <imm>
followed by R |= <imm> or any similar combination where the
original information from the unknown scalar would be destroyed
entirely leaving R with a constant. The initial slow load still
precedes the latter ALU ops on that register, so the CPU
executes speculatively from that point. Once we have the known
scalar, any compare operation would work then. A third option
only involving registers with known scalars could be crafted
as described in [0] where a CPU port (e.g. Slow Int unit)
would be filled with many dependent computations such that
the subsequent condition depending on its outcome has to wait
for evaluation on its execution port and thereby executing
speculatively if the speculated code can be scheduled on a
different execution port, or any other form of mistraining
as described in [1], for example. Given this is not limited
to only unknown scalars, not only map but also stack access
is affected since both is accessible for unprivileged users
and could potentially be used for out of bounds access under
speculation.
In order to prevent any of these cases, the verifier is now
sanitizing pointer arithmetic on the offset such that any
out of bounds speculation would be masked in a way where the
pointer arithmetic result in the destination register will
stay unchanged, meaning offset masked into zero similar as
in array_index_nospec() case. With regards to implementation,
there are three options that were considered: i) new insn
for sanitation, ii) push/pop insn and sanitation as inlined
BPF, iii) reuse of ax register and sanitation as inlined BPF.
Option i) has the downside that we end up using from reserved
bits in the opcode space, but also that we would require
each JIT to emit masking as native arch opcodes meaning
mitigation would have slow adoption till everyone implements
it eventually which is counter-productive. Option ii) and iii)
have both in common that a temporary register is needed in
order to implement the sanitation as inlined BPF since we
are not allowed to modify the source register. While a push /
pop insn in ii) would be useful to have in any case, it
requires once again that every JIT needs to implement it
first. While possible, amount of changes needed would also
be unsuitable for a -stable patch. Therefore, the path which
has fewer changes, less BPF instructions for the mitigation
and does not require anything to be changed in the JITs is
option iii) which this work is pursuing. The ax register is
already mapped to a register in all JITs (modulo arm32 where
it's mapped to stack as various other BPF registers there)
and used in constant blinding for JITs-only so far. It can
be reused for verifier rewrites under certain constraints.
The interpreter's tmp "register" has therefore been remapped
into extending the register set with hidden ax register and
reusing that for a number of instructions that needed the
prior temporary variable internally (e.g. div, mod). This
allows for zero increase in stack space usage in the interpreter,
and enables (restricted) generic use in rewrites otherwise as
long as such a patchlet does not make use of these instructions.
The sanitation mask is dynamic and relative to the offset the
map value or stack pointer currently holds.
There are various cases that need to be taken under consideration
for the masking, e.g. such operation could look as follows:
ptr += val or val += ptr or ptr -= val. Thus, the value to be
sanitized could reside either in source or in destination
register, and the limit is different depending on whether
the ALU op is addition or subtraction and depending on the
current known and bounded offset. The limit is derived as
follows: limit := max_value_size - (smin_value + off). For
subtraction: limit := umax_value + off. This holds because
we do not allow any pointer arithmetic that would
temporarily go out of bounds or would have an unknown
value with mixed signed bounds where it is unclear at
verification time whether the actual runtime value would
be either negative or positive. For example, we have a
derived map pointer value with constant offset and bounded
one, so limit based on smin_value works because the verifier
requires that statically analyzed arithmetic on the pointer
must be in bounds, and thus it checks if resulting
smin_value + off and umax_value + off is still within map
value bounds at time of arithmetic in addition to time of
access. Similarly, for the case of stack access we derive
the limit as follows: MAX_BPF_STACK + off for subtraction
and -off for the case of addition where off := ptr_reg->off +
ptr_reg->var_off.value. Subtraction is a special case for
the masking which can be in form of ptr += -val, ptr -= -val,
or ptr -= val. In the first two cases where we know that
the value is negative, we need to temporarily negate the
value in order to do the sanitation on a positive value
where we later swap the ALU op, and restore original source
register if the value was in source.
The sanitation of pointer arithmetic alone is still not fully
sufficient as is, since a scenario like the following could
happen ...
... and therefore still access out of bounds. To prevent such
case, the verifier is also analyzing safety for potential out
of bounds access under speculative execution. Meaning, it is
also simulating pointer access under truncation. We therefore
"branch off" and push the current verification state after the
ALU operation with known 0 to the verification stack for later
analysis. Given the current path analysis succeeded it is
likely that the one under speculation can be pruned. In any
case, it is also subject to existing complexity limits and
therefore anything beyond this point will be rejected. In
terms of pruning, it needs to be ensured that the verification
state from speculative execution simulation must never prune
a non-speculative execution path, therefore, we mark verifier
state accordingly at the time of push_stack(). If verifier
detects out of bounds access under speculative execution from
one of the possible paths that includes a truncation, it will
reject such program.
Given we mask every reg-based pointer arithmetic for
unprivileged programs, we've been looking into how it could
affect real-world programs in terms of size increase. As the
majority of programs are targeted for privileged-only use
case, we've unconditionally enabled masking (with its alu
restrictions on top of it) for privileged programs for the
sake of testing in order to check i) whether they get rejected
in its current form, and ii) by how much the number of
instructions and size will increase. We've tested this by
using Katran, Cilium and test_l4lb from the kernel selftests.
For Katran we've evaluated balancer_kern.o, Cilium bpf_lxc.o
and an older test object bpf_lxc_opt_-DUNKNOWN.o and l4lb
we've used test_l4lb.o as well as test_l4lb_noinline.o. We
found that none of the programs got rejected by the verifier
with this change, and that impact is rather minimal to none.
balancer_kern.o had 13,904 bytes (1,738 insns) xlated and
7,797 bytes JITed before and after the change. Most complex
program in bpf_lxc.o had 30,544 bytes (3,817 insns) xlated
and 18,538 bytes JITed before and after and none of the other
tail call programs in bpf_lxc.o had any changes either. For
the older bpf_lxc_opt_-DUNKNOWN.o object we found a small
increase from 20,616 bytes (2,576 insns) and 12,536 bytes JITed
before to 20,664 bytes (2,582 insns) and 12,558 bytes JITed
after the change. Other programs from that object file had
similar small increase. Both test_l4lb.o had no change and
remained at 6,544 bytes (817 insns) xlated and 3,401 bytes
JITed and for test_l4lb_noinline.o constant at 5,080 bytes
(634 insns) xlated and 3,313 bytes JITed. This can be explained
in that LLVM typically optimizes stack based pointer arithmetic
by using K-based operations and that use of dynamic map access
is not overly frequent. However, in future we may decide to
optimize the algorithm further under known guarantees from
branch and value speculation. Latter seems also unclear in
terms of prediction heuristics that today's CPUs apply as well
as whether there could be collisions in e.g. the predictor's
Value History/Pattern Table for triggering out of bounds access,
thus masking is performed unconditionally at this point but could
be subject to relaxation later on. We were generally also
brainstorming various other approaches for mitigation, but the
blocker was always lack of available registers at runtime and/or
overhead for runtime tracking of limits belonging to a specific
pointer. Thus, we found this to be minimally intrusive under
given constraints.
With that in place, a simple example with sanitized access on
unprivileged load at post-verification time looks as follows:
JIT blinding example with non-conflicting use of r10:
[...]
d5: je 0x0000000000000106 _
d7: mov 0x0(%rax),%edi |
da: mov $0xf153246,%r10d | Index load from map value and
e0: xor $0xf153259,%r10 | (const blinded) mask with 0x1f.
e7: and %r10,%rdi |_
ea: mov $0x2f,%r10d |
f0: sub %rdi,%r10 | Sanitized addition. Both use r10
f3: or %rdi,%r10 | but do not interfere with each
f6: neg %r10 | other. (Neither do these instructions
f9: sar $0x3f,%r10 | interfere with the use of ax as temp
fd: and %r10,%rdi | in interpreter.)
100: add %rax,%rdi |_
103: mov 0x0(%rdi),%eax
[...]
Tested that it fixes Jann's reproducer, and also checked that test_verifier
and test_progs suite with interpreter, JIT and JIT with hardening enabled
on x86-64 and arm64 runs successfully.
[0] Speculose: Analyzing the Security Implications of Speculative
Execution in CPUs, Giorgi Maisuradze and Christian Rossow,
https://arxiv.org/pdf/1801.04084.pdf
[1] A Systematic Evaluation of Transient Execution Attacks and
Defenses, Claudio Canella, Jo Van Bulck, Michael Schwarz,
Moritz Lipp, Benjamin von Berg, Philipp Ortner, Frank Piessens,
Dmitry Evtyushkin, Daniel Gruss,
https://arxiv.org/pdf/1811.05441.pdf
In check_map_access() we probe actual bounds through __check_map_access()
with offset of reg->smin_value + off for lower bound and offset of
reg->umax_value + off for the upper bound. However, even though the
reg->smin_value could have a negative value, the final result of the
sum with off could be positive when pointer arithmetic with known and
unknown scalars is combined. In this case we reject the program with
an error such as "R<x> min value is negative, either use unsigned index
or do a if (index >=0) check." even though the access itself would be
fine. Therefore extend the check to probe whether the actual resulting
reg->smin_value + off is less than zero.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
For unknown scalars of mixed signed bounds, meaning their smin_value is
negative and their smax_value is positive, we need to reject arithmetic
with pointer to map value. For unprivileged the goal is to mask every
map pointer arithmetic and this cannot reliably be done when it is
unknown at verification time whether the scalar value is negative or
positive. Given this is a corner case, the likelihood of breaking should
be very small.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Signed-off-by: Sasha Levin <sashal@kernel.org>