Vlastimil Babka [Fri, 2 Aug 2019 16:06:14 +0000 (18:06 +0200)]
x86, mm, gup: prevent get_page() race with munmap in paravirt guest
The x86 version of get_user_pages_fast() relies on disabled interrupts to
synchronize gup_pte_range() between gup_get_pte(ptep); and get_page() against
a parallel munmap. The munmap side nulls the pte, then flushes TLBs, then
releases the page. As TLB flush is done synchronously via IPI disabling
interrupts blocks the page release, and get_page(), which assumes existing
reference on page, is thus safe.
However when TLB flush is done by a hypercall, e.g. in a Xen PV guest, there is
no blocking thanks to disabled interrupts, and get_page() can succeed on a page
that was already freed or even reused.
We have recently seen this happen with our 4.4 and 4.12 based kernels, with
userspace (java) that exits a thread, where mm_release() performs a futex_wake()
on tsk->clear_child_tid, and another thread in parallel unmaps the page where
tsk->clear_child_tid points to. The spurious get_page() succeeds, but futex code
immediately releases the page again, while it's already on a freelist. Symptoms
include a bad page state warning, general protection faults acessing a poisoned
list prev/next pointer in the freelist, or free page pcplists of two cpus joined
together in a single list. Oscar has also reproduced this scenario, with a
patch inserting delays before the get_page() to make the race window larger.
Fix this by removing the dependency on TLB flush interrupts the same way as the
generic get_user_pages_fast() code by using page_cache_add_speculative() and
revalidating the PTE contents after pinning the page. Mainline is safe since
4.13 where the x86 gup code was removed in favor of the common code. Accessing
the page table itself safely also relies on disabled interrupts and TLB flush
IPIs that don't happen with hypercalls, which was acknowledged in commit 9e52fc2b50de ("x86/mm: Enable RCU based page table freeing
(CONFIG_HAVE_RCU_TABLE_FREE=y)"). That commit with follups should also be
backported for full safety, although our reproducer didn't hit a problem
without that backport.
Reproduced-by: Oscar Salvador <osalvador@suse.de> Signed-off-by: Vlastimil Babka <vbabka@suse.cz> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Juergen Gross <jgross@suse.com> Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Vitaly Kuznetsov <vkuznets@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Borislav Petkov <bp@alien8.de> Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The upcoming GCC 9 release extends the -Wmissing-attributes warnings
(enabled by -Wall) to C and aliases: it warns when particular function
attributes are missing in the aliases but not in their target.
In particular, it triggers for all the init/cleanup_module
aliases in the kernel (defined by the module_init/exit macros),
ending up being very noisy.
These aliases point to the __init/__exit functions of a module,
which are defined as __cold (among other attributes). However,
the aliases themselves do not have the __cold attribute.
Since the compiler behaves differently when compiling a __cold
function as well as when compiling paths leading to calls
to __cold functions, the warning is trying to point out
the possibly-forgotten attribute in the alias.
In order to keep the warning enabled, we decided to silence
this case. Ideally, we would mark the aliases directly
as __init/__exit. However, there are currently around 132 modules
in the kernel which are missing __init/__exit in their init/cleanup
functions (either because they are missing, or for other reasons,
e.g. the functions being called from somewhere else); and
a section mismatch is a hard error.
A conservative alternative was to mark the aliases as __cold only.
However, since we would like to eventually enforce __init/__exit
to be always marked, we chose to use the new __copy function
attribute (introduced by GCC 9 as well to deal with this).
With it, we copy the attributes used by the target functions
into the aliases. This way, functions that were not marked
as __init/__exit won't have their aliases marked either,
and therefore there won't be a section mismatch.
Note that the warning would go away marking either the extern
declaration, the definition, or both. However, we only mark
the definition of the alias, since we do not want callers
(which only see the declaration) to be compiled as if the function
was __cold (and therefore the paths leading to those calls
would be assumed to be unlikely).
When fixing the race conditions between the coredump and the mmap_sem
holders outside the context of the process, we focused on
mmget_not_zero()/get_task_mm() callers in 04f5866e41fb70 ("coredump: fix
race condition between mmget_not_zero()/get_task_mm() and core
dumping"), but those aren't the only cases where the mmap_sem can be
taken outside of the context of the process as Michal Hocko noticed
while backporting that commit to older -stable kernels.
If mmgrab() is called in the context of the process, but then the
mm_count reference is transferred outside the context of the process,
that can also be a problem if the mmap_sem has to be taken for writing
through that mm_count reference.
khugepaged registration calls mmgrab() in the context of the process,
but the mmap_sem for writing is taken later in the context of the
khugepaged kernel thread.
collapse_huge_page() after taking the mmap_sem for writing doesn't
modify any vma, so it's not obvious that it could cause a problem to the
coredump, but it happens to modify the pmd in a way that breaks an
invariant that pmd_trans_huge_lock() relies upon. collapse_huge_page()
needs the mmap_sem for writing just to block concurrent page faults that
call pmd_trans_huge_lock().
Specifically the invariant that "!pmd_trans_huge()" cannot become a
"pmd_trans_huge()" doesn't hold while collapse_huge_page() runs.
The coredump will call __get_user_pages() without mmap_sem for reading,
which eventually can invoke a lockless page fault which will need a
functional pmd_trans_huge_lock().
So collapse_huge_page() needs to use mmget_still_valid() to check it's
not running concurrently with the coredump... as long as the coredump
can invoke page faults without holding the mmap_sem for reading.
This has "Fixes: khugepaged" to facilitate backporting, but in my view
it's more a bug in the coredump code that will eventually have to be
rewritten to stop invoking page faults without the mmap_sem for reading.
So the long term plan is still to drop all mmget_still_valid().
Link: http://lkml.kernel.org/r/20190607161558.32104-1-aarcange@redhat.com Fixes: ba76149f47d8 ("thp: khugepaged") Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Reported-by: Michal Hocko <mhocko@suse.com> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Jann Horn <jannh@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Mike Rapoport <rppt@linux.vnet.ibm.com> Cc: Mike Kravetz <mike.kravetz@oracle.com> Cc: Peter Xu <peterx@redhat.com> Cc: Jason Gunthorpe <jgg@mellanox.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[Ajay: Just adjusted to apply on v4.9] Signed-off-by: Ajay Kaher <akaher@vmware.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Ajay Kaher [Sun, 4 Aug 2019 03:59:26 +0000 (09:29 +0530)]
infiniband: fix race condition between infiniband mlx4, mlx5 driver and core dumping
This patch is the extension of following upstream commit to fix
the race condition between get_task_mm() and core dumping
for IB->mlx4 and IB->mlx5 drivers:
commit 04f5866e41fb ("coredump: fix race condition between
mmget_not_zero()/get_task_mm() and core dumping")'
The core dumping code has always run without holding the mmap_sem for
writing, despite that is the only way to ensure that the entire vma
layout will not change from under it. Only using some signal
serialization on the processes belonging to the mm is not nearly enough.
This was pointed out earlier. For example in Hugh's post from Jul 2017:
"Not strictly relevant here, but a related note: I was very surprised
to discover, only quite recently, how handle_mm_fault() may be called
without down_read(mmap_sem) - when core dumping. That seems a
misguided optimization to me, which would also be nice to correct"
In particular because the growsdown and growsup can move the
vm_start/vm_end the various loops the core dump does around the vma will
not be consistent if page faults can happen concurrently.
Pretty much all users calling mmget_not_zero()/get_task_mm() and then
taking the mmap_sem had the potential to introduce unexpected side
effects in the core dumping code.
Adding mmap_sem for writing around the ->core_dump invocation is a
viable long term fix, but it requires removing all copy user and page
faults and to replace them with get_dump_page() for all binary formats
which is not suitable as a short term fix.
For the time being this solution manually covers the places that can
confuse the core dump either by altering the vma layout or the vma flags
while it runs. Once ->core_dump runs under mmap_sem for writing the
function mmget_still_valid() can be dropped.
Allowing mmap_sem protected sections to run in parallel with the
coredump provides some minor parallelism advantage to the swapoff code
(which seems to be safe enough by never mangling any vma field and can
keep doing swapins in parallel to the core dumping) and to some other
corner case.
In order to facilitate the backporting I added "Fixes: 86039bd3b4e6"
however the side effect of this same race condition in /proc/pid/mem
should be reproducible since before 2.6.12-rc2 so I couldn't add any
other "Fixes:" because there's no hash beyond the git genesis commit.
Because find_extend_vma() is the only location outside of the process
context that could modify the "mm" structures under mmap_sem for
reading, by adding the mmget_still_valid() check to it, all other cases
that take the mmap_sem for reading don't need the new check after
mmget_not_zero()/get_task_mm(). The expand_stack() in page fault
context also doesn't need the new check, because all tasks under core
dumping are frozen.
Link: http://lkml.kernel.org/r/20190325224949.11068-1-aarcange@redhat.com Fixes: 86039bd3b4e6 ("userfaultfd: add new syscall to provide memory externalization") Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Reported-by: Jann Horn <jannh@google.com> Suggested-by: Oleg Nesterov <oleg@redhat.com> Acked-by: Peter Xu <peterx@redhat.com> Reviewed-by: Mike Rapoport <rppt@linux.ibm.com> Reviewed-by: Oleg Nesterov <oleg@redhat.com> Reviewed-by: Jann Horn <jannh@google.com> Acked-by: Jason Gunthorpe <jgg@mellanox.com> Acked-by: Michal Hocko <mhocko@suse.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[akaher@vmware.com: stable 4.9 backport
- handle binder_update_page_range - mhocko@suse.com] Signed-off-by: Ajay Kaher <akaher@vmware.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The specification for the Toeplitz function doesn't require to set the key
explicitly to be symmetric. In case a symmetric functionality is required
a symmetric key can be simply used.
Wrongly forcing the algorithm to symmetric causes the wrong packet
distribution and a performance degradation.
The condition in xen_swiotlb_free_coherent() for deciding whether to
call xen_destroy_contiguous_region() is wrong: in case the region to
be freed is not contiguous calling xen_destroy_contiguous_region() is
the wrong thing to do: it would result in inconsistent mappings of
multiple PFNs to the same MFN. This will lead to various strange
crashes or data corruption.
Instead of calling xen_destroy_contiguous_region() in that case a
warning should be issued as that situation should never occur.
Cc: stable@vger.kernel.org Signed-off-by: Juergen Gross <jgross@suse.com> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Reviewed-by: Jan Beulich <jbeulich@suse.com> Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by: Juergen Gross <jgross@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Handling of the CPU_PM_ENTER_FAILED transition in the Arm PMU PM
notifier code incorrectly skips restoration of the counters. Fix the
logic so that CPU_PM_ENTER_FAILED follows the same path as CPU_PM_EXIT.
Cc: <stable@vger.kernel.org> Fixes: da4e4f18afe0f372 ("drivers/perf: arm_pmu: implement CPU_PM notifier") Reported-by: Anders Roxell <anders.roxell@linaro.org> Acked-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After getting a storage server event that causes the DASD device driver
to update its unit address configuration during a device shutdown there is
the possibility of an endless loop in the device driver.
In the system log there will be ongoing DASD error messages with RC: -19.
The reason is that the loop starting the ruac request only terminates when
the retry counter is decreased to 0. But in the sleep_on function there are
early exit paths that do not decrease the retry counter.
Prevent an endless loop by handling those cases separately.
Remove the unnecessary do..while loop since the sleep_on function takes
care of retries by itself.
Fixes: 8e09f21574ea ("[S390] dasd: add hyper PAV support to DASD device driver, part 1") Cc: stable@vger.kernel.org # 2.6.25+ Signed-off-by: Stefan Haberland <sth@linux.ibm.com> Reviewed-by: Jan Hoeppner <hoeppner@linux.ibm.com> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since roles_init() adds some entries to the role hash table, we need to
destroy also its keys/values on error, otherwise we get a memory leak in
the error path.
Cc: <stable@vger.kernel.org> Reported-by: syzbot+fee3a14d4cdf92646287@syzkaller.appspotmail.com Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Ondrej Mosnacek <omosnace@redhat.com> Signed-off-by: Paul Moore <paul@paul-moore.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a pin is active-low, logical trigger edge should be inverted to match
the same interrupt opportunity.
For example, a button pushed triggers falling edge in ACTIVE_HIGH case; in
ACTIVE_LOW case, the button pushed triggers rising edge. For user space the
IRQ requesting doesn't need to do any modification except to configuring
GPIOHANDLE_REQUEST_ACTIVE_LOW.
For example, we want to catch the event when the button is pushed. The
button on the original board drives level to be low when it is pushed, and
drives level to be high when it is released.
while (1) {
read(fd, &dat, sizeof(dat));
if (dat.id == GPIOEVENT_EVENT_FALLING_EDGE)
printf("button pushed\n");
}
Run the same logic on another board which the polarity of the button is
inverted; it drives level to be high when pushed, and level to be low when
released. For this inversion we add flag GPIOHANDLE_REQUEST_ACTIVE_LOW:
At the result, there are no any events caught when the button is pushed.
By the way, button releasing will emit a "falling" event. The timing of
"falling" catching is not expected.
In commit 46d179525a1f ("mmc: dw_mmc: Wait for data transfer after
response errors.") we fixed a tuning-induced hang that I saw when
stress testing tuning on certain SD cards. I won't re-hash that whole
commit, but the summary is that as a normal part of tuning you need to
deal with transfer errors and there were cases where these transfer
errors was putting my system into a bad state causing all future
transfers to fail. That commit fixed handling of the transfer errors
for me.
In downstream Chrome OS my fix landed and had the same behavior for
all SD/MMC commands. However, it looks like when the commit landed
upstream we limited it to only SD tuning commands. Presumably this
was to try to get around problems that Alim Akhtar reported on exynos
[1].
Unfortunately while stress testing reboots (and suspend/resume) on
some rk3288-based Chromebooks I found the same problem on the eMMC on
some of my Chromebooks (the ones with Hynix eMMC). Since the eMMC
tuning command is different (MMC_SEND_TUNING_BLOCK_HS200
vs. MMC_SEND_TUNING_BLOCK) we were basically getting back into the
same situation.
I'm hoping that whatever problems exynos was having in the past are
somehow magically fixed now and we can make the behavior the same for
all commands.
Fixes: 46d179525a1f ("mmc: dw_mmc: Wait for data transfer after response errors.") Signed-off-by: Douglas Anderson <dianders@chromium.org> Cc: Marek Szyprowski <m.szyprowski@samsung.com> Cc: Alim Akhtar <alim.akhtar@gmail.com> Cc: Enric Balletbo i Serra <enric.balletbo@collabora.com> Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When doing an incremental send operation we can fail if we previously did
deduplication operations against a file that exists in both snapshots. In
that case we will fail the send operation with -EIO and print a message
to dmesg/syslog like the following:
BTRFS error (device sdc): Send: inconsistent snapshot, found updated \
extent for inode 257 without updated inode item, send root is 258, \
parent root is 257
This requires that we deduplicate to the same file in both snapshots for
the same amount of times on each snapshot. The issue happens because a
deduplication only updates the iversion of an inode and does not update
any other field of the inode, therefore if we deduplicate the file on
each snapshot for the same amount of time, the inode will have the same
iversion value (stored as the "sequence" field on the inode item) on both
snapshots, therefore it will be seen as unchanged between in the send
snapshot while there are new/updated/deleted extent items when comparing
to the parent snapshot. This makes the send operation return -EIO and
print an error message.
Example reproducer:
$ mkfs.btrfs -f /dev/sdb
$ mount /dev/sdb /mnt
# Create our first file. The first half of the file has several 64Kb
# extents while the second half as a single 512Kb extent.
$ xfs_io -f -s -c "pwrite -S 0xb8 -b 64K 0 512K" /mnt/foo
$ xfs_io -c "pwrite -S 0xb8 512K 512K" /mnt/foo
# Create the base snapshot and the parent send stream from it.
$ btrfs subvolume snapshot -r /mnt /mnt/mysnap1
$ btrfs send -f /tmp/1.snap /mnt/mysnap1
# Create our second file, that has exactly the same data as the first
# file.
$ xfs_io -f -c "pwrite -S 0xb8 0 1M" /mnt/bar
# Create the second snapshot, used for the incremental send, before
# doing the file deduplication.
$ btrfs subvolume snapshot -r /mnt /mnt/mysnap2
# Now before creating the incremental send stream:
#
# 1) Deduplicate into a subrange of file foo in snapshot mysnap1. This
# will drop several extent items and add a new one, also updating
# the inode's iversion (sequence field in inode item) by 1, but not
# any other field of the inode;
#
# 2) Deduplicate into a different subrange of file foo in snapshot
# mysnap2. This will replace an extent item with a new one, also
# updating the inode's iversion by 1 but not any other field of the
# inode.
#
# After these two deduplication operations, the inode items, for file
# foo, are identical in both snapshots, but we have different extent
# items for this inode in both snapshots. We want to check this doesn't
# cause send to fail with an error or produce an incorrect stream.
This issue started happening back in 2015 when deduplication was updated
to not update the inode's ctime and mtime and update only the iversion.
Back then we would hit a BUG_ON() in send, but later in 2016 send was
updated to return -EIO and print the error message instead of doing the
BUG_ON().
A test case for fstests follows soon.
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203933 Fixes: 1c919a5e13702c ("btrfs: don't update mtime/ctime on deduped inodes") 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>
After making a change to improve objtool's sibling call detection, it
started showing the following warning:
arch/x86/kvm/vmx/nested.o: warning: objtool: .fixup+0x15: sibling call from callable instruction with modified stack frame
The problem is the ____kvm_handle_fault_on_reboot() macro. It does a
fake call by pushing a fake RIP and doing a jump. That tricks the
unwinder into printing the function which triggered the exception,
rather than the .fixup code.
Instead of the hack to make it look like the original function made the
call, just change the macro so that the original function actually does
make the call. This allows removal of the hack, and also makes objtool
happy.
I triggered a vmx instruction exception and verified that the stack
trace is still sane:
if (mq_open("/testing", 0x40, 3, &attr) == (mqd_t) -1)
perror("mq_open");
mqueue_get_inode() was correctly rejecting the giant mq_msgsize, and
preparing to return -EINVAL. During the cleanup, it calls
mqueue_evict_inode() which performed resource usage tracking math for
updating "user", before checking if there was a valid "user" at all
(which would indicate that the calculations would be sane). Instead,
delay this check to after seeing a valid "user".
The overflow was real, but the results went unused, so while the flaw is
harmless, it's noisy for kernel fuzzers, so just fix it by moving the
calculation under the non-NULL "user" where it actually gets used.
The dev_info.name[] array has space for RIO_MAX_DEVNAME_SZ + 1
characters. But the problem here is that we don't ensure that the user
put a NUL terminator on the end of the string. It could lead to an out
of bounds read.
Only users of upc_req in kernel side fs/coda/psdev.c and
fs/coda/upcall.c already include linux/coda_psdev.h.
Suggested by Jan Harkes <jaharkes@cs.cmu.edu> in
https://lore.kernel.org/lkml/20150531111913.GA23377@cs.cmu.edu/
Fixes these include/uapi/linux/coda_psdev.h compilation errors in userspace:
linux/coda_psdev.h:12:19: error: field `uc_chain' has incomplete type
struct list_head uc_chain;
^
linux/coda_psdev.h:13:2: error: unknown type name `caddr_t'
caddr_t uc_data;
^
linux/coda_psdev.h:14:2: error: unknown type name `u_short'
u_short uc_flags;
^
linux/coda_psdev.h:15:2: error: unknown type name `u_short'
u_short uc_inSize; /* Size is at most 5000 bytes */
^
linux/coda_psdev.h:16:2: error: unknown type name `u_short'
u_short uc_outSize;
^
linux/coda_psdev.h:17:2: error: unknown type name `u_short'
u_short uc_opcode; /* copied from data to save lookup */
^
linux/coda_psdev.h:19:2: error: unknown type name `wait_queue_head_t'
wait_queue_head_t uc_sleep; /* process' wait queue */
^
Link: http://lkml.kernel.org/r/9f99f5ce6a0563d5266e6cf7aa9585aac2cae971.1558117389.git.jaharkes@cs.cmu.edu Signed-off-by: Mikko Rapeli <mikko.rapeli@iki.fi> Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Colin Ian King <colin.king@canonical.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: David Howells <dhowells@redhat.com> Cc: Fabian Frederick <fabf@skynet.be> Cc: Sam Protsenko <semen.protsenko@linaro.org> Cc: Yann Droneaud <ydroneaud@opteya.com> Cc: Zhouyang Jia <jiazhouyang09@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The kernel is self-contained project and can be built with bare-metal
toolchain. But bare-metal toolchain doesn't define __linux__. Because
of this u_quad_t type is not defined when using bare-metal toolchain and
codafs build fails. This patch fixes it by defining u_quad_t type
unconditionally.
Link: http://lkml.kernel.org/r/3cbb40b0a57b6f9923a9d67b53473c0b691a3eaa.1558117389.git.jaharkes@cs.cmu.edu Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Colin Ian King <colin.king@canonical.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: David Howells <dhowells@redhat.com> Cc: Fabian Frederick <fabf@skynet.be> Cc: Mikko Rapeli <mikko.rapeli@iki.fi> Cc: Yann Droneaud <ydroneaud@opteya.com> Cc: Zhouyang Jia <jiazhouyang09@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The description of cma_declare_contiguous() indicates that if the
'fixed' argument is true the reserved contiguous area must be exactly at
the address of the 'base' argument.
However, the function currently allows the 'base', 'size', and 'limit'
arguments to be silently adjusted to meet alignment constraints. This
commit enforces the documented behavior through explicit checks that
return an error if the region does not fit within a specified region.
Link: http://lkml.kernel.org/r/1561422051-16142-1-git-send-email-opendmb@gmail.com Fixes: 5ea3b1b2f8ad ("cma: add placement specifier for "cma=" kernel parameter") Signed-off-by: Doug Berger <opendmb@gmail.com> Acked-by: Michal Nazarewicz <mina86@mina86.com> Cc: Yue Hu <huyue2@yulong.com> Cc: Mike Rapoport <rppt@linux.ibm.com> Cc: Laura Abbott <labbott@redhat.com> Cc: Peng Fan <peng.fan@nxp.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Marek Szyprowski <m.szyprowski@samsung.com> Cc: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
In file included from ./arch/x86/include/asm/smp.h:13,
from ./arch/x86/include/asm/mmzone_64.h:11,
from ./arch/x86/include/asm/mmzone.h:5,
from ./include/linux/mmzone.h:969,
from ./include/linux/gfp.h:6,
from ./include/linux/mm.h:10,
from arch/x86/kernel/apic/io_apic.c:34:
arch/x86/kernel/apic/io_apic.c: In function 'check_timer':
./arch/x86/include/asm/apic.h:37:11: warning: comparison of unsigned
expression >= 0 is always true [-Wtype-limits]
if ((v) <= apic_verbosity) \
^~
arch/x86/kernel/apic/io_apic.c:2160:2: note: in expansion of macro
'apic_printk'
apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X "
^~~~~~~~~~~
./arch/x86/include/asm/apic.h:37:11: warning: comparison of unsigned
expression >= 0 is always true [-Wtype-limits]
if ((v) <= apic_verbosity) \
^~
arch/x86/kernel/apic/io_apic.c:2207:4: note: in expansion of macro
'apic_printk'
apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: "
^~~~~~~~~~~
APIC_QUIET is 0, so silence them by making apic_verbosity type int.
While changing the number of interrupt channels, be2net stops adapter
operation (including netif_tx_disable()) but it doesn't signal that it
cannot transmit. This may lead dev_watchdog() to falsely trigger during
that time.
Add the missing call to netif_carrier_off(), following the pattern used in
many other drivers. netif_carrier_on() is already taken care of in
be_open().
Signed-off-by: Benjamin Poirier <bpoirier@suse.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
clang gets confused by an uninitialized variable in what looks
to it like a never executed code path:
arch/x86/kernel/acpi/boot.c:618:13: error: variable 'polarity' is uninitialized when used here [-Werror,-Wuninitialized]
polarity = polarity ? ACPI_ACTIVE_LOW : ACPI_ACTIVE_HIGH;
^~~~~~~~
arch/x86/kernel/acpi/boot.c:606:32: note: initialize the variable 'polarity' to silence this warning
int rc, irq, trigger, polarity;
^
= 0
arch/x86/kernel/acpi/boot.c:617:12: error: variable 'trigger' is uninitialized when used here [-Werror,-Wuninitialized]
trigger = trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE;
^~~~~~~
arch/x86/kernel/acpi/boot.c:606:22: note: initialize the variable 'trigger' to silence this warning
int rc, irq, trigger, polarity;
^
= 0
This is unfortunately a design decision in clang and won't be fixed.
Changing the acpi_get_override_irq() macro to an inline function
reliably avoids the issue.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
GCC v9 emits this warning:
CC drivers/s390/scsi/zfcp_erp.o
drivers/s390/scsi/zfcp_erp.c: In function 'zfcp_erp_action_enqueue':
drivers/s390/scsi/zfcp_erp.c:217:26: warning: 'erp_action' may be used uninitialized in this function [-Wmaybe-uninitialized]
217 | struct zfcp_erp_action *erp_action;
| ^~~~~~~~~~
This is a possible false positive case, as also documented in the GCC
documentations:
https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html#index-Wmaybe-uninitialized
The actual code-sequence is like this:
Various callers can invoke the function below with the argument "want"
being one of:
ZFCP_ERP_ACTION_REOPEN_ADAPTER,
ZFCP_ERP_ACTION_REOPEN_PORT_FORCED,
ZFCP_ERP_ACTION_REOPEN_PORT, or
ZFCP_ERP_ACTION_REOPEN_LUN.
zfcp_erp_action_enqueue(want, ...)
...
need = zfcp_erp_required_act(want, ...)
need = want
...
maybe: need = ZFCP_ERP_ACTION_REOPEN_PORT
maybe: need = ZFCP_ERP_ACTION_REOPEN_ADAPTER
...
return need
...
zfcp_erp_setup_act(need, ...)
struct zfcp_erp_action *erp_action; // <== line 217
...
switch(need) {
case ZFCP_ERP_ACTION_REOPEN_LUN:
...
erp_action = &zfcp_sdev->erp_action;
WARN_ON_ONCE(erp_action->port != port); // <== access
...
break;
case ZFCP_ERP_ACTION_REOPEN_PORT:
case ZFCP_ERP_ACTION_REOPEN_PORT_FORCED:
...
erp_action = &port->erp_action;
WARN_ON_ONCE(erp_action->port != port); // <== access
...
break;
case ZFCP_ERP_ACTION_REOPEN_ADAPTER:
...
erp_action = &adapter->erp_action;
WARN_ON_ONCE(erp_action->port != NULL); // <== access
...
break;
}
...
WARN_ON_ONCE(erp_action->adapter != adapter); // <== access
When zfcp_erp_setup_act() is called, 'need' will never be anything else
than one of the 4 possible enumeration-names that are used in the
switch-case, and 'erp_action' is initialized for every one of them, before
it is used. Thus the warning is a false positive, as documented.
We introduce the extra if{} in the beginning to create an extra code-flow,
so the compiler can be convinced that the switch-case will never see any
other value.
BUG_ON()/BUG() is intentionally not used to not crash anything, should
this ever happen anyway - right now it's impossible, as argued above; and
it doesn't introduce a 'default:' switch-case to retain warnings should
'enum zfcp_erp_act_type' ever be extended and no explicit case be
introduced. See also v5.0 commit 399b6c8bc9f7 ("scsi: zfcp: drop old
default switch case which might paper over missing case").
Signed-off-by: Benjamin Block <bblock@linux.ibm.com> Reviewed-by: Jens Remus <jremus@linux.ibm.com> Reviewed-by: Steffen Maier <maier@linux.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The getxattr manpage states that we should return ERANGE if the
destination buffer size is too small to hold the value.
ceph_vxattrcb_layout does this internally, but we should be doing
this for all vxattrs.
Fix the only caller of getxattr_cb to check the returned size
against the buffer length and return -ERANGE if it doesn't fit.
Drop the same check in ceph_vxattrcb_layout and just rely on the
caller to handle it.
The list of profiles in btrfs_chunk_max_errors lists DUP as a profile
DUP able to tolerate 1 device missing. Though this profile is special
with 2 copies, it still needs the device, unlike the others.
Looking at the history of changes, thre's no clear reason why DUP is
there, functions were refactored and blocks of code merged to one
helper.
d20983b40e828 Btrfs: fix writing data into the seed filesystem
- factor code to a helper
de11cc12df173 Btrfs: don't pre-allocate btrfs bio
- unrelated change, DUP still in the list with max errors 1
a236aed14ccb0 Btrfs: Deal with failed writes in mirrored configurations
- introduced the max errors, leaves DUP and RAID1 in the same group
Fix a use-after-free bug during filesystem initialisation, where we
access the disc record (which is stored in a buffer) after we have
released the buffer.
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Sasha Levin <sashal@kernel.org>
While the .device_prep_slave_sg() callback rejects empty scatterlists,
it still accepts single-entry scatterlists with a zero-length segment.
These may happen if a driver calls dmaengine_prep_slave_single() with a
zero len parameter. The corresponding DMA request will never complete,
leading to messages like:
Although requesting a zero-length DMA request is a driver bug, rejecting
it early eases debugging. Note that the .device_prep_dma_memcpy()
callback already rejects requests to copy zero bytes.
The modification of EXIN register doesn't clean the bitfield before
the writing of a new value. After a few modifications the bitfield would
accumulate only '1's.
Microsoft HyperV disables the X86_FEATURE_SMCA bit on AMD systems, and
linux guests boot with repeated errors:
amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)
amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)
The warnings occur because the module code erroneously returns -EEXIST
for modules that have failed to load and are in the process of being
removed from the module list.
module amd64_edac_mod has a dependency on module edac_mce_amd. Using
modules.dep, systemd will load edac_mce_amd for every request of
amd64_edac_mod. When the edac_mce_amd module loads, the module has
state MODULE_STATE_UNFORMED and once the module load fails and the state
becomes MODULE_STATE_GOING. Another request for edac_mce_amd module
executes and add_unformed_module() will erroneously return -EEXIST even
though the previous instance of edac_mce_amd has MODULE_STATE_GOING.
Upon receiving -EEXIST, systemd attempts to load amd64_edac_mod, which
fails because of unknown symbols from edac_mce_amd.
add_unformed_module() must wait to return for any case other than
MODULE_STATE_LIVE to prevent a race between multiple loads of
dependent modules.
Custom trampolines can only be enabled if there is only a single ops
attached to it. If there's only a single callback registered to a function,
and the ops has a trampoline registered for it, then we can call the
trampoline directly. This is very useful for improving the performance of
ftrace and livepatch.
If more than one callback is registered to a function, the general
trampoline is used, and the custom trampoline is not restored back to the
direct call even if all the other callbacks were unregistered and we are
back to one callback for the function.
To fix this, set FTRACE_FL_TRAMP flag if rec count is decremented
to one, and the ops that left has a trampoline.
This is similar to commit e6186820a745 ("arm64: dts: rockchip: Arch
counter doesn't tick in system suspend"). Specifically on the rk3288
it can be seen that the timer stops ticking in suspend if we end up
running through the "osc_disable" path in rk3288_slp_mode_set(). In
that path the 24 MHz clock will turn off and the timer stops.
To test this, I ran this on a Chrome OS filesystem:
before=$(date); \
suspend_stress_test -c1 --suspend_min=30 --suspend_max=31; \
echo ${before}; date
...and I found that unless I plug in a device that requests USB wakeup
to be active that the two calls to "date" would show that fewer than
30 seconds passed.
NOTE: deep suspend (where the 24 MHz clock gets disabled) isn't
supported yet on upstream Linux so this was tested on a downstream
kernel.
Signed-off-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
When I try to boot rk3288-veyron-mickey I totally fail to make the
eMMC work. Specifically my logs (on Chrome OS 4.19):
mmc_host mmc1: card is non-removable.
mmc_host mmc1: Bus speed (slot 0) = 400000Hz (slot req 400000Hz, actual 400000HZ div = 0)
mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 52000000Hz, actual 50000000HZ div = 0)
mmc1: switch to bus width 8 failed
mmc1: switch to bus width 4 failed
mmc1: new high speed MMC card at address 0001
mmcblk1: mmc1:0001 HAG2e 14.7 GiB
mmcblk1boot0: mmc1:0001 HAG2e partition 1 4.00 MiB
mmcblk1boot1: mmc1:0001 HAG2e partition 2 4.00 MiB
mmcblk1rpmb: mmc1:0001 HAG2e partition 3 4.00 MiB, chardev (243:0)
mmc_host mmc1: Bus speed (slot 0) = 400000Hz (slot req 400000Hz, actual 400000HZ div = 0)
mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 52000000Hz, actual 50000000HZ div = 0)
mmc1: switch to bus width 8 failed
mmc1: switch to bus width 4 failed
mmc1: tried to HW reset card, got error -110
mmcblk1: error -110 requesting status
mmcblk1: recovery failed!
print_req_error: I/O error, dev mmcblk1, sector 0
...
When I remove the '/delete-property/mmc-hs200-1_8v' then everything is
hunky dory.
That line comes from the original submission of the mickey dts
upstream, so presumably at the time the HS200 was failing and just
enumerating things as a high speed device was fine. ...or maybe it's
just that some mickey devices work when enumerating at "high speed",
just not mine?
In any case, hs200 seems good now. Let's turn it on.
Signed-off-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
As some point hs200 was failing on rk3288-veyron-minnie. See commit 984926781122 ("ARM: dts: rockchip: temporarily remove emmc hs200 speed
from rk3288 minnie"). Although I didn't track down exactly when it
started working, it seems to work OK now, so let's turn it back on.
To test this, I booted from SD card and then used this script to
stress the enumeration process after fixing a memory leak [1]:
cd /sys/bus/platform/drivers/dwmmc_rockchip
for i in $(seq 1 3000); do
echo "========================" $i
echo ff0f0000.dwmmc > unbind
sleep .5
echo ff0f0000.dwmmc > bind
while true; do
if [ -e /dev/mmcblk2 ]; then
break;
fi
sleep .1
done
done
DMA got broken a while back in two different ways:
1) a change in the behaviour of disable_irq() to wait for the interrupt
to finish executing causes us to deadlock at the end of DMA.
2) a change to avoid modifying the scatterlist left the first transfer
uninitialised.
DMA is only used with expansion cards, so has gone unnoticed.
Fixes: fa4e99899932 ("[ARM] dma: RiscPC: don't modify DMA SG entries") Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: Sasha Levin <sashal@kernel.org>
The PPS assert/clear offset corrections are set by the PPS_SETPARAMS
ioctl in the pps_ktime structs, which also contain flags. The flags are
not initialized by applications (using the timepps.h header) and they
are not used by the kernel for anything except returning them back in
the PPS_GETPARAMS ioctl.
Set the flags to zero to make it clear they are unused and avoid leaking
uninitialized data of the PPS_SETPARAMS caller to other applications
that have a read access to the PPS device.
Link: http://lkml.kernel.org/r/20190702092251.24303-1-mlichvar@redhat.com Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com> Reviewed-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Rodolfo Giometti <giometti@enneenne.com> Cc: Greg KH <greg@kroah.com> Cc: Dan Carpenter <dan.carpenter@oracle.com> 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>
When going through execve(), zero out the NUMA fault statistics instead of
freeing them.
During execve, the task is reachable through procfs and the scheduler. A
concurrent /proc/*/sched reader can read data from a freed ->numa_faults
allocation (confirmed by KASAN) and write it back to userspace.
I believe that it would also be possible for a use-after-free read to occur
through a race between a NUMA fault and execve(): task_numa_fault() can
lead to task_numa_compare(), which invokes task_weight() on the currently
running task of a different CPU.
Another way to fix this would be to make ->numa_faults RCU-managed or add
extra locking, but it seems easier to wipe the NUMA fault statistics on
execve.
Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Petr Mladek <pmladek@suse.com> Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Will Deacon <will@kernel.org> Fixes: 82727018b0d3 ("sched/numa: Call task_numa_free() from do_execve()") Link: https://lkml.kernel.org/r/20190716152047.14424-1-jannh@google.com Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Certain ttys operations (pty_unix98_ops) lack tiocmget() and tiocmset()
functions which are called by the certain HCI UART protocols (hci_ath,
hci_bcm, hci_intel, hci_mrvl, hci_qca) via hci_uart_set_flow_control()
or directly. This leads to an execution at NULL and can be triggered by
an unprivileged user. Fix this by adding a helper function and a check
for the missing tty operations in the protocols code.
This fixes CVE-2019-10207. The Fixes: lines list commits where calls to
tiocm[gs]et() or hci_uart_set_flow_control() were added to the HCI UART
protocols.
Link: https://syzkaller.appspot.com/bug?id=1b42faa2848963564a5b1b7f8c837ea7b55ffa50 Reported-by: syzbot+79337b501d6aa974d0f6@syzkaller.appspotmail.com Cc: stable@vger.kernel.org # v2.6.36+ Fixes: b3190df62861 ("Bluetooth: Support for Atheros AR300x serial chip") Fixes: 118612fb9165 ("Bluetooth: hci_bcm: Add suspend/resume PM functions") Fixes: ff2895592f0f ("Bluetooth: hci_intel: Add Intel baudrate configuration support") Fixes: 162f812f23ba ("Bluetooth: hci_uart: Add Marvell support") Fixes: fa9ad876b8e0 ("Bluetooth: hci_qca: Add support for Qualcomm Bluetooth chip wcn3990") Signed-off-by: Vladis Dronov <vdronov@redhat.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org> Reviewed-by: Yu-Chen, Cho <acho@suse.com> Tested-by: Yu-Chen, Cho <acho@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change devm_k*alloc to k*alloc to manually allocate memory
The manual allocation and freeing of memory is necessary because when
the USB radio is disconnected, the memory associated with devm_k*alloc
is freed. Meaning if we still have unresolved references to the radio
device, then we get use-after-free errors.
This patch fixes this by manually allocating memory, and freeing it in
the v4l2.release callback that gets called when the last radio device
exits.
Reported-and-tested-by: syzbot+a4387f5b6b799f6becbf@syzkaller.appspotmail.com Signed-off-by: Luke Nowakowski-Krijger <lnowakow@eng.ucsd.edu> Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
[hverkuil-cisco@xs4all.nl: cleaned up two small checkpatch.pl warnings]
[hverkuil-cisco@xs4all.nl: prefix subject with driver name] Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Kasan reported a use after free in cpia2_usb_disconnect()
It first freed everything and then woke up those waiting.
The reverse order is correct.
Fixes: 6c493f8b28c67 ("[media] cpia2: major overhaul to get it in a working state again") Signed-off-by: Oliver Neukum <oneukum@suse.com> Reported-by: syzbot+0c90fc937c84f97d0aa6@syzkaller.appspotmail.com Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Phong Tran <tranmanphong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If any I2C error comes during BAM DMA transfer, then the QUP I2C
interrupt will be generated and the flush operation will be
carried out to make I2C consume all scheduled DMA transfer.
Currently, the same completion structure is being used for BAM
transfer which has already completed without reinit. It will make
flush operation wait_for_completion_timeout completed immediately
and will proceed for freeing the DMA resources where the
descriptors are still in process.
Signed-off-by: Abhishek Sahu <absahu@codeaurora.org> Acked-by: Sricharan R <sricharan@codeaurora.org> Reviewed-by: Austin Christ <austinwc@codeaurora.org> Reviewed-by: Andy Gross <andy.gross@linaro.org> Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Amit Pundir <amit.pundir@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
tcp_write_queue_purge clears all the SKBs in the write queue
but does not reset the sk_send_head. As a result, we can have
a NULL pointer dereference anywhere that we use tcp_send_head
instead of the tcp_write_queue_tail.
For example, after a27fd7a8ed38 (tcp: purge write queue upon RST),
we can purge the write queue on RST. Prior to 75c119afe14f (tcp: implement rb-tree based retransmit queue),
tcp_push will only check tcp_send_head and then accesses
tcp_write_queue_tail to send the actual SKB. As a result, it will
dereference a NULL pointer.
This has been reported twice for 4.14 where we don't have 75c119afe14f:
stable 4.4 and stable 4.9 don't have the commit abb4a8b870b5 ("tcp: purge write queue upon RST")
which is referred in dbbf2d1e4077,
in tcp_connect_init, it calls tcp_write_queue_purge, and does not reset sk_send_head, then UAF.
stable 4.14 have the commit abb4a8b870b5 ("tcp: purge write queue upon RST"),
in tcp_reset, it calls tcp_write_queue_purge(sk), and does not reset sk_send_head, then UAF.
So this patch can be used to fix stable 4.4 and 4.9.
Fixes: a27fd7a8ed38 (tcp: purge write queue upon RST) Reported-by: Timofey Titovets <nefelim4ag@gmail.com> Reported-by: Yongjian Xu <yongjianchn@gmail.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com> Tested-by: Yongjian Xu <yongjianchn@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Mao Wenan <maowenan@huawei.com> Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 5e1859fbcc3c ("ipv4: ipmr: various fixes and cleanups") fixed
the issue for ipv4 ipmr:
ip_mroute_setsockopt() & ip_mroute_getsockopt() should not
access/set raw_sk(sk)->ipmr_table before making sure the socket
is a raw socket, and protocol is IGMP
The same fix should be done for ipv6 ipmr as well.
This patch can fix the panic caused by overwriting the same offset
as ipmr_table as in raw_sk(sk) when accessing other type's socket
by ip_mroute_setsockopt().
Signed-off-by: Xin Long <lucien.xin@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It turns out that 'access()' (and 'faccessat()') can cause a lot of RCU
work because it installs a temporary credential that gets allocated and
freed for each system call.
The allocation and freeing overhead is mostly benign, but because
credentials can be accessed under the RCU read lock, the freeing
involves a RCU grace period.
Which is not a huge deal normally, but if you have a lot of access()
calls, this causes a fair amount of seconday damage: instead of having a
nice alloc/free patterns that hits in hot per-CPU slab caches, you have
all those delayed free's, and on big machines with hundreds of cores,
the RCU overhead can end up being enormous.
But it turns out that all of this is entirely unnecessary. Exactly
because access() only installs the credential as the thread-local
subjective credential, the temporary cred pointer doesn't actually need
to be RCU free'd at all. Once we're done using it, we can just free it
synchronously and avoid all the RCU overhead.
So add a 'non_rcu' flag to 'struct cred', which can be set by users that
know they only use it in non-RCU context (there are other potential
users for this). We can make it a union with the rcu freeing list head
that we need for the RCU case, so this doesn't need any extra storage.
Note that this also makes 'get_current_cred()' clear the new non_rcu
flag, in case we have filesystems that take a long-term reference to the
cred and then expect the RCU delayed freeing afterwards. It's not
entirely clear that this is required, but it makes for clear semantics:
the subjective cred remains non-RCU as long as you only access it
synchronously using the thread-local accessors, but you _can_ use it as
a generic cred if you want to.
It is possible that we should just remove the whole RCU markings for
->cred entirely. Only ->real_cred is really supposed to be accessed
through RCU, and the long-term cred copies that nfs uses might want to
explicitly re-enable RCU freeing if required, rather than have
get_current_cred() do it implicitly.
But this is a "minimal semantic changes" change for the immediate
problem.
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Eric Dumazet <edumazet@google.com> Acked-by: Paul E. McKenney <paulmck@linux.ibm.com> Cc: Oleg Nesterov <oleg@redhat.com> Cc: Jan Glauber <jglauber@marvell.com> Cc: Jiri Kosina <jikos@kernel.org> Cc: Jayachandran Chandrasekharan Nair <jnair@marvell.com> Cc: Greg KH <greg@kroah.com> Cc: Kees Cook <keescook@chromium.org> Cc: David Howells <dhowells@redhat.com> Cc: Miklos Szeredi <miklos@szeredi.hu> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On systems like P9 powernv where we have no TM (or P8 booted with
ppc_tm=off), userspace can construct a signal context which still has
the MSR TS bits set. The kernel tries to restore this context which
results in the following crash:
The problem is the signal code assumes TM is enabled when
CONFIG_PPC_TRANSACTIONAL_MEM is enabled. This may not be the case as
with P9 powernv or if `ppc_tm=off` is used on P8.
This means any local user can crash the system.
Fix the problem by returning a bad stack frame to the user if they try
to set the MSR TS bits with sigreturn() on systems where TM is not
supported.
Found with sigfuz kernel selftest on P9.
This fixes CVE-2019-13648.
Fixes: 2b0a576d15e0 ("powerpc: Add new transactional memory state to the signal context") Cc: stable@vger.kernel.org # v3.9 Reported-by: Praveen Pandey <Praveen.Pandey@in.ibm.com> Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20190719050502.405-1-mikey@neuling.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This conexant codec isn't in the supported codec list yet, the hda
generic driver can drive this codec well, but on a Lenovo machine
with mute/mic-mute leds, we need to apply CXT_FIXUP_THINKPAD_ACPI
to make the leds work. After adding this codec to the list, the
driver patch_conexant.c will apply THINKPAD_ACPI to this machine.
X86_HYPER_NATIVE isn't accurate for checking if running on native platform,
e.g. CONFIG_HYPERVISOR_GUEST isn't set or "nopv" is enabled.
Checking the CPU feature bit X86_FEATURE_HYPERVISOR to determine if it's
running on native platform is more accurate.
This still doesn't cover the platforms on which X86_FEATURE_HYPERVISOR is
unsupported, e.g. VMware, but there is nothing which can be done about this
scenario.
Some Lenovo 2-in-1s with a detachable keyboard have a portrait screen but
advertise a landscape resolution and pitch, resulting in a messed up
display if the kernel tries to show anything on the efifb (because of the
wrong pitch).
Fix this by adding a new DMI match table for devices which need to have
their width and height swapped.
At first it was tried to use the existing table for overriding some of the
efifb parameters, but some of the affected devices have variants with
different LCD resolutions which will not work with hardcoded override
values.
Reference: https://bugzilla.redhat.com/show_bug.cgi?id=1730783 Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20190721152418.11644-1-hdegoede@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The AMD PLL USB quirk is incorrectly enabled on newer Ryzen
chipsets. The logic in usb_amd_find_chipset_info currently checks
for unaffected chipsets rather than affected ones. This broke
once a new chipset was added in e788787ef. It makes more sense
to reverse the logic so it won't need to be updated as new
chipsets are added. Note that the core of the workaround in
usb_amd_quirk_pll does correctly check the chipset.
Signed-off-by: Ryan Kennedy <ryan5544@gmail.com> Fixes: e788787ef4f9 ("usb:xhci:Add quirk for Certain failing HP keyboard on reset after resume") Cc: stable <stable@vger.kernel.org> Acked-by: Alan Stern <stern@rowland.harvard.edu> Link: https://lore.kernel.org/r/20190704153529.9429-2-ryan5544@gmail.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
syzboot reported that
https://syzkaller.appspot.com/bug?extid=fd2bd7df88c606eea4ef
There is not consitency parameter in cluste_id_get/put calling.
In case of getting the id with result is failure, the wusbhc->cluster_id
will not be updated and this can not be used for wusb_cluster_id_put().
The stats variable nr_unused_locks is incremented every time a new lock
class is register and decremented when the lock is first used in
__lock_acquire(). And after all, it is shown and checked in lockdep_stats.
However, under configurations that either CONFIG_TRACE_IRQFLAGS or
CONFIG_PROVE_LOCKING is not defined:
The commit:
091806515124b20 ("locking/lockdep: Consolidate lock usage bit initialization")
missed marking the LOCK_USED flag at IRQ usage initialization because
as mark_usage() is not called. And the commit:
Make mmu_notifier_register() safer by issuing a memory barrier before
registering a new notifier. This fixes a theoretical bug on weakly
ordered CPUs. For example, take this simplified use of notifiers by a
driver:
Once mmu_notifier_register() releases the mm locks, another thread can
invalidate a range:
mmu_notifier_invalidate_range()
...
hlist_for_each_entry_rcu(mn, &mm->mmu_notifiers, hlist) {
if (mn->ops->invalidate_range)
The read side relies on the data dependency between mn and ops to ensure
that the pointer is properly initialized. But the write side doesn't have
any dependency between (1) and (2), so they could be reordered and the
readers could dereference an invalid mn->ops. mmu_notifier_register()
does take all the mm locks before adding to the hlist, but those have
acquire semantics which isn't sufficient.
By calling hlist_add_head_rcu() instead of hlist_add_head() we update the
hlist using a store-release, ensuring that readers see prior
initialization of my_struct. This situation is better illustated by
litmus test MP+onceassign+derefonce.
Fix the callback 9p passes to read_cache_page to actually have the
proper type expected. Casting around function pointers can easily
hide typing bugs, and defeats control flow protection.
Link: http://lkml.kernel.org/r/20190520055731.24538-5-hch@lst.de Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Kees Cook <keescook@chromium.org> Cc: Sami Tolvanen <samitolvanen@google.com> Cc: Nick Desaulniers <ndesaulniers@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
in_softirq() is a wrong predicate to check if we are in a softirq
context. It also returns true if we have BH disabled, so objects are
falsely stamped with "softirq" comm. The correct predicate is
in_serving_softirq().
If user does cat from /sys/kernel/debug/kmemleak previously they would
see this, which is clearly wrong, this is system call context (see the
comm):
In commit 4a7b06c157a2 ("powerpc/eeh: Handle hugepages in ioremap
space") support for using hugepages in the vmalloc and ioremap areas was
enabled for radix. Unfortunately this broke EEH MMIO error checking.
Detection works by inserting a hook which checks the results of the
ioreadXX() set of functions. When a read returns a 0xFFs response we
need to check for an error which we do by mapping the (virtual) MMIO
address back to a physical address, then mapping physical address to a
PCI device via an interval tree.
When translating virt -> phys we currently assume the ioremap space is
only populated by PAGE_SIZE mappings. If a hugepage mapping is found we
emit a WARN_ON(), but otherwise handles the check as though a normal
page was found. In pathalogical cases such as copying a buffer
containing a lot of 0xFFs from BAR memory this can result in the system
not booting because it's too busy printing WARN_ON()s.
There's no real reason to assume huge pages can't be present and we're
prefectly capable of handling them, so do that.
Fixes: 4a7b06c157a2 ("powerpc/eeh: Handle hugepages in ioremap space") Reported-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by: Oliver O'Halloran <oohall@gmail.com> Tested-by: Sachin Sant <sachinp@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20190710150517.27114-1-oohall@gmail.com Signed-off-by: Sasha Levin <sashal@kernel.org>
Previously, if mbox_request_channel_byname was used with a name
which did not exist in the "mbox-names" property of a mailbox
client, the mailbox corresponding to the last entry in the
"mbox-names" list would be incorrectly selected.
With this patch, -EINVAL is returned if the named mailbox is
not found.
The next commit will make the way of passing CONFIG options more robust.
Unfortunately, it would uncover another hidden issue; without this
commit, skiroot_defconfig would be broken like this:
| WRAP arch/powerpc/boot/zImage.pseries
| arch/powerpc/boot/wrapper.a(decompress.o): In function `bcj_powerpc.isra.10':
| decompress.c:(.text+0x720): undefined reference to `get_unaligned_be32'
| decompress.c:(.text+0x7a8): undefined reference to `put_unaligned_be32'
| make[1]: *** [arch/powerpc/boot/Makefile;383: arch/powerpc/boot/zImage.pseries] Error 1
| make: *** [arch/powerpc/Makefile;295: zImage] Error 2
skiroot_defconfig is the only defconfig that enables CONFIG_KERNEL_XZ
for ppc, which has never been correctly built before.
I figured out the root cause in lib/decompress_unxz.c:
Calculate the correct byte_len on the receiving side when a work
completion is generated with IB_WC_RECV_RDMA_WITH_IMM opcode.
According to the IBA byte_len must indicate the number of written bytes,
whereas it was always equal to zero for the IB_WC_RECV_RDMA_WITH_IMM
opcode, even though data was transferred.
Fixes: 8700e3e7c485 ("Soft RoCE driver") Signed-off-by: Konstantin Taranov <konstantin.taranov@inf.ethz.ch> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Running the 'perf test' command after building perf with a memory
sanitizer causes a warning that says:
WARNING: MemorySanitizer: use-of-uninitialized-value... in mmap-thread-lookup.c
Initializing the go variable to 0 silences this harmless warning.
Committer warning:
This was harmless, just a simple test writing whatever was at that
sizeof(int) memory area just to signal another thread blocked reading
that file created with pipe(). Initialize it tho so that we don't get
this warning.
Signed-off-by: Numfor Mbiziwo-Tiapo <nums@google.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Drayton <mbd@fb.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Song Liu <songliubraving@fb.com> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/20190702173716.181223-1-nums@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
gcc asan instrumentation emits the following sequence to store frame pc
when the kernel is built with CONFIG_RELOCATABLE:
debug/vsprintf.s:
.section .data.rel.ro.local,"aw"
.align 8
.LC3:
.quad .LASANPC4826@GOTOFF
.text
.align 8
.type number, @function
number:
.LASANPC4826:
and in case reloc is issued for LASANPC label it also gets into .symtab
with the same address as actual function symbol:
$ nm -n vmlinux | grep 0000000001397150 0000000001397150 t .LASANPC4826 0000000001397150 t number
When uart_flush_buffer() is called, the .flush_buffer() callback zeroes
the tx_dma_len field. This may race with the work queue function
handling transmit DMA requests:
1. If the buffer is flushed before the first DMA API call,
dmaengine_prep_slave_single() may be called with a zero length,
causing the DMA request to never complete, leading to messages
like:
2. If the buffer is flushed after the first DMA API call, but before
the second, dma_sync_single_for_device() may be called with a zero
length, causing the transmit data not to be flushed to RAM, and
leading to stale data being output.
Fix this by:
1. Letting sci_dma_tx_work_fn() return immediately if the transmit
buffer is empty,
2. Extending the critical section to cover all DMA preparational work,
so tx_dma_len stays consistent for all of it,
3. Using local copies of circ_buf.head and circ_buf.tail, to make sure
they match the actual operation above.
While the .flush_buffer() callback clears sci_port.tx_dma_len since
commit 1cf4a7efdc71cab8 ("serial: sh-sci: Fix race condition causing
garbage during shutdown"), it does not terminate a transmit DMA
operation that may be in progress.
Fix this by terminating any pending DMA operations, and resetting the
corresponding cookie.
When testing out gpio-keys with a button, a spurious
interrupt (and therefore a key press or release event)
gets triggered as soon as the driver enables the irq
line for the first time.
This patch clears any potential bogus generated interrupt
that was caused by the switching of the associated irq's
type and polarity.
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Sasha Levin <sashal@kernel.org>
down_read(&old_mm->mmap_sem);
...
task_lock(tsk);
...
activate_mm(active_mm, mm);
(which does down_write(&mm->mmap_sem))
I'm not really sure why lockdep throws in the whole knowledge
about the task lock, but it seems that old_mm and mm shouldn't
ever be the same (and it doesn't deadlock) so tell lockdep that
they're different.
Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Richard Weinberger <richard@nod.at> Signed-off-by: Sasha Levin <sashal@kernel.org>
Since devm_regmap_init_mmio_clk can fail, add return value checking.
Signed-off-by: Axel Lin <axel.lin@ingics.com> Acked-by: Chen Feng <puck.chen@hisilicon.com> Signed-off-by: Lee Jones <lee.jones@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When the driver is used with a subdevice that is disabled in the
kernel configuration, clang gets a little confused about the
control flow and fails to notice that n_subdevs is only
uninitialized when subdevs is NULL, and we check for that,
leading to a false-positive warning:
drivers/mfd/arizona-core.c:1423:19: error: variable 'n_subdevs' is uninitialized when used here
[-Werror,-Wuninitialized]
subdevs, n_subdevs, NULL, 0, NULL);
^~~~~~~~~
drivers/mfd/arizona-core.c:999:15: note: initialize the variable 'n_subdevs' to silence this warning
int n_subdevs, ret, i;
^
= 0
Ideally, we would rearrange the code to avoid all those early
initializations and have an explicit exit in each disabled case,
but it's much easier to chicken out and add one more initialization
here to shut up the warning.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Nathan Chancellor <natechancellor@gmail.com> Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Lee Jones <lee.jones@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
The logic for setting the of_node on devices created by mfd did not set
the fwnode pointer to match, which caused fwnode-based APIs to
malfunction on these devices since the fwnode pointer was null. Fix
this.
Signed-off-by: Robert Hancock <hancock@sedsystems.ca> Signed-off-by: Lee Jones <lee.jones@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Looking at the relocation records in __mcount_loc shows a few spurious
entries:
RELOCATION RECORDS FOR [__mcount_loc]:
OFFSET TYPE VALUE 0000000000000000 R_PPC64_ADDR64 .text.unlikely+0x0000000000000008 0000000000000008 R_PPC64_ADDR64 .text.unlikely+0x0000000000000014 0000000000000010 R_PPC64_ADDR64 .text.unlikely+0x0000000000000060 0000000000000018 R_PPC64_ADDR64 .text.unlikely+0x00000000000000b4 0000000000000020 R_PPC64_ADDR64 .init.text+0x0000000000000008 0000000000000028 R_PPC64_ADDR64 .init.text+0x0000000000000014
The first entry in each section is incorrect. Looking at the
relocation records, the spurious entries correspond to the
R_PPC64_ENTRY records:
RELOCATION RECORDS FOR [.text.unlikely]:
OFFSET TYPE VALUE 0000000000000000 R_PPC64_REL64 .TOC.-0x0000000000000008 0000000000000008 R_PPC64_ENTRY *ABS* 0000000000000014 R_PPC64_REL24 _mcount
<snip>
The problem is that we are not validating the return value from
get_mcountsym() in sift_rel_mcount(). With this entry, mcountsym is 0,
but Elf_r_sym(relp) also ends up being 0. Fix this by ensuring
mcountsym is valid before processing the entry.
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org> Tested-by: Satheesh Rajendran <sathnaga@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Sasha Levin <sashal@kernel.org>
On some machines, iio-sensor-proxy was returning all 0's for IIO sensor
values. It turns out that the bits_used for this sensor is 32, which makes
the mask calculation:
*mask = (1 << 32) - 1;
If the compiler interprets the 1 literals as 32-bit ints, it generates
undefined behavior depending on compiler version and optimization level.
On my system, it optimizes out the shift, so the mask value becomes
*mask = (1) - 1;
With a mask value of 0, iio-sensor-proxy will always return 0 for every axis.
Avoid incorrect 0 values caused by compiler optimization.
See original fix by Brett Dutro <brett.dutro@gmail.com> in
iio-sensor-proxy:
https://github.com/hadess/iio-sensor-proxy/commit/9615ceac7c134d838660e209726cd86aa2064fd3
According to the PCI Local Bus specification Revision 3.0,
section 6.8.1.3 (Message Control for MSI), endpoints that
are Multiple Message Capable as defined by bits [3:1] in
the Message Control for MSI can request a number of vectors
that is power of two aligned.
As specified in section 6.8.1.6 "Message data for MSI", the Multiple
Message Enable field (bits [6:4] of the Message Control register)
defines the number of low order message data bits the function is
permitted to modify to generate its system software allocated
vectors.
The MSI controller in the Xilinx NWL PCIe controller supports a number
of MSI vectors specified through a bitmap and the hwirq number for an
MSI, that is the value written in the MSI data TLP is determined by
the bitmap allocation.
For instance, in a situation where two endpoints sitting on
the PCI bus request the following MSI configuration, with
the current PCI Xilinx bitmap allocation code (that does not
align MSI vector allocation on a power of two boundary):
The bitmap value(s) corresponds to the hwirq number that is programmed
into the Message Data for MSI field in the endpoint MSI capability
and is detected by the root complex to fire the corresponding
MSI irqs. The value written in Message Data for MSI field corresponds
to the first bit allocated in the bitmap for Multi MSI vectors.
The current Xilinx NWL MSI allocation code allows a bitmap allocation
that is not a power of two boundaries, so endpoint #2, is allowed to
toggle Message Data bit[0] to differentiate between its two vectors
(meaning that the MSI data will be respectively 0x0 and 0x1 for the two
vectors allocated to endpoint #2).
This clearly aliases with the Endpoint #1 vector allocation, resulting
in a broken Multi MSI implementation.
Update the code to allocate MSI bitmap ranges with a power of two
alignment, fixing the bug.
Fixes: ab597d35ef11 ("PCI: xilinx-nwl: Add support for Xilinx NWL PCIe Host Controller") Suggested-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Bharat Kumar Gogada <bharat.kumar.gogada@xilinx.com>
[lorenzo.pieralisi@arm.com: updated commit log] Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Acked-by: Marc Zyngier <marc.zyngier@arm.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In commit ebcc5928c5d9 ("arm64: Silence gcc warnings about arch ABI
drift"), the arm64 Makefile added -Wno-psabi to KBUILD_CFLAGS, which is
a GCC only option so clang rightfully complains:
However, by default, this is merely a warning so the build happily goes
on with a slew of these warnings in the process.
Commit c3f0d0bc5b01 ("kbuild, LLVMLinux: Add -Werror to cc-option to
support clang") worked around this behavior in cc-option by adding
-Werror so that unknown flags cause an error. However, this all happens
silently and when an unknown flag is added to the build unconditionally
like -Wno-psabi, cc-option will always fail because there is always an
unknown flag in the list of flags. This manifested as link time failures
in the arm64 libstub because -fno-stack-protector didn't get added to
KBUILD_CFLAGS.
To avoid these weird cryptic failures in the future, make clang behave
like gcc and immediately error when it encounters an unknown flag by
adding -Werror=unknown-warning-option to CLANG_FLAGS. This can be added
unconditionally for clang because it is supported by at least 3.0.0,
according to godbolt [1] and 4.0.0, according to its documentation [2],
which is far earlier than we typically support.
With CONFIG_PROVE_LOCKING=y, using sysfs to remove a bridge with a device
below it causes a lockdep warning, e.g.,
# echo 1 > /sys/class/pci_bus/0000:00/device/0000:00:00.0/remove
============================================
WARNING: possible recursive locking detected
...
pci_bus 0000:01: busn_res: [bus 01] is released
The remove recursively removes the subtree below the bridge. Each call
uses a different lock so there's no deadlock, but the locks were all
created with the same lockdep key so the lockdep checker can't tell them
apart.
Mark the "remove" sysfs attribute with __ATTR_IGNORE_LOCKDEP() as it is
safe to ignore the lockdep check between different "remove" kernfs
instances.
There's discussion about a similar issue in USB at [1], which resulted in 356c05d58af0 ("sysfs: get rid of some lockdep false positives") and e9b526fe7048 ("i2c: suppress lockdep warning on delete_device"), which do
basically the same thing for USB "remove" and i2c "delete_device" files.
[1] https://lore.kernel.org/r/Pine.LNX.4.44L0.1204251436140.1206-100000@iolanthe.rowland.org Link: https://lore.kernel.org/r/20190526225151.3865-1-marek.vasut@gmail.com Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
[bhelgaas: trim commit log, details at above links] Signed-off-by: Bjorn Helgaas <bhelgaas@google.com> Cc: Geert Uytterhoeven <geert+renesas@glider.be> Cc: Phil Edworthy <phil.edworthy@renesas.com> Cc: Simon Horman <horms+renesas@verge.net.au> Cc: Tejun Heo <tj@kernel.org> Cc: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Sasha Levin <sashal@kernel.org>
When the firmware does PCI BAR resource allocation, it passes the assigned
addresses and flags (prefetch/64bit/...) via the "reg" property of
a PCI device device tree node so the kernel does not need to do
resource allocation.
The flags are stored in resource::flags - the lower byte stores
PCI_BASE_ADDRESS_SPACE/etc bits and the other bytes are IORESOURCE_IO/etc.
Some flags from PCI_BASE_ADDRESS_xxx and IORESOURCE_xxx are duplicated,
such as PCI_BASE_ADDRESS_MEM_PREFETCH/PCI_BASE_ADDRESS_MEM_TYPE_64/etc.
When parsing the "reg" property, we copy the prefetch flag but we skip
on PCI_BASE_ADDRESS_MEM_TYPE_64 which leaves the flags out of sync.
The missing IORESOURCE_MEM_64 flag comes into play under 2 conditions:
1. we remove PCI_PROBE_ONLY for pseries (by hacking pSeries_setup_arch()
or by passing "/chosen/linux,pci-probe-only");
2. we request resource alignment (by passing pci=resource_alignment=
via the kernel cmd line to request PAGE_SIZE alignment or defining
ppc_md.pcibios_default_alignment which returns anything but 0). Note that
the alignment requests are ignored if PCI_PROBE_ONLY is enabled.
With 1) and 2), the generic PCI code in the kernel unconditionally
decides to:
- reassign the BARs in pci_specified_resource_alignment() (works fine)
- write new BARs to the device - this fails for 64bit BARs as the generic
code looks at IORESOURCE_MEM_64 (not set) and writes only lower 32bits
of the BAR and leaves the upper 32bit unmodified which breaks BAR mapping
in the hypervisor.
This fixes the issue by copying the flag. This is useful if we want to
enforce certain BAR alignment per platform as handling subpage sized BARs
is proven to cause problems with hotplug (SLOF already aligns BARs to 64k).
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Reviewed-by: Sam Bobroff <sbobroff@linux.ibm.com> Reviewed-by: Oliver O'Halloran <oohall@gmail.com> Reviewed-by: Shawn Anastasio <shawn@anastas.io> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Sasha Levin <sashal@kernel.org>
In some cases the "Allocate & copy" block in ffs_epfile_io() is not
executed. Consequently, in such a case ffs_alloc_buffer() is never called
and struct ffs_io_data is not initialized properly. This in turn leads to
problems when ffs_free_buffer() is called at the end of ffs_epfile_io().
This patch uses kzalloc() instead of kmalloc() in the aio case and memset()
in non-aio case to properly initialize struct ffs_io_data.
Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@collabora.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
A bug was introduced by commit b3b576461864 ("tty: serial_core: convert
uart_open to use tty_port_open"). It caused a constant warning printed
into the system log regarding the tty and port counter mismatch:
in case if session hangup was detected so the warning is printed starting
from the second open-close iteration.
Particularly the problem was discovered in situation when there is a
serial tty device without hardware back-end being setup. It is considered
by the tty-serial subsystems as a hardware problem with session hang up.
In this case uart_startup() will return a positive value with TTY_IO_ERROR
flag set in corresponding tty_struct instance. The same value will get
passed to be returned from the activate() callback and then being returned
from tty_port_open(). But since in this case tty_port_block_til_ready()
isn't called the TTY_PORT_ACTIVE flag isn't set (while the method had been
called before tty_port_open conversion was introduced and the rest of the
subsystem code expected the bit being set in this case), which prevents the
uart_hangup() method to perform any cleanups including the tty port
counter setting to zero. So the next attempt to open/close the tty device
will discover the counters mismatch.
In order to fix the problem we need to manually set the TTY_PORT_ACTIVE
flag in case if uart_startup() returned a positive value. In this case
the hang up procedure will perform a full set of cleanup actions including
the port ref-counter resetting.
When fixing up the clock in vop_crtc_mode_fixup() we're not doing it
quite correctly. Specifically if we've got the true clock 266666667 Hz,
we'll perform this calculation: 266666667 / 1000 => 266666
Later when we try to set the clock we'll do clk_set_rate(266666 *
1000). The common clock framework won't actually pick the proper clock
in this case since it always wants clocks <= the specified one.
Let's solve this by using DIV_ROUND_UP.
Fixes: b59b8de31497 ("drm/rockchip: return a true clock rate to adjusted_mode") Signed-off-by: Douglas Anderson <dianders@chromium.org> Signed-off-by: Sean Paul <seanpaul@chromium.org> Reviewed-by: Yakir Yang <ykk@rock-chips.com> Signed-off-by: Heiko Stuebner <heiko@sntech.de> Link: https://patchwork.freedesktop.org/patch/msgid/20190614224730.98622-1-dianders@chromium.org Signed-off-by: Sasha Levin <sashal@kernel.org>
This patch fixes memory leak at error paths of the probe function.
In for_each_child_of_node, if the loop returns, the driver should
call of_put_node() before returns.
Reported-by: Julia Lawall <julia.lawall@lip6.fr> Fixes: 1233f59f745b237 ("phy: Renesas R-Car Gen2 PHY driver") Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
After data is copied to the cache entry, atomic_set is used indicate
that the data is the entry is valid without appropriate memory barriers.
Similarly the read side was missing the corresponding memory barriers.
Interrupt handler checked THRE bit (transmitter holding register
empty) in LSR to detect if TX fifo is empty.
In case when there is only receive interrupts the TX handling
got called because THRE bit in LSR is set when there is no
transmission (FIFO empty). TX handling caused TX stop, which in
RS-485 half-duplex mode actually resets receiver FIFO. This is not
desired during reception because of possible data loss.
The fix is to check if THRI is set in IER in addition of the TX
fifo status. THRI in IER is set when TX is started and cleared
when TX is stopped.
This ensures that TX handling is only called when there is really
transmission on going and an interrupt for THRE and not when there
are only RX interrupts.
The function msm_wait_for_xmitr can be taken with interrupts
disabled. In order to avoid a potential system lockup - demonstrated
under stress testing conditions on SoC QCS404/5 - make sure we wait
for a bounded amount of time.
projects / linux-stable / log