If there is a possibility that a VM may migrate to a Skylake host,
then the hypervisor should report IA32_ARCH_CAPABILITIES.RSBA[bit 2]
as being set (future work, of course). This implies that
CPUID.(EAX=7,ECX=0):EDX.ARCH_CAPABILITIES[bit 29] should be
set. Therefore, kvm should report this CPUID bit as being supported
whether or not the host supports it. Userspace is still free to clear
the bit if it chooses.
For more information on RSBA, see Intel's white paper, "Retpoline: A
Branch Target Injection Mitigation" (Document Number 337131-001),
currently available at https://bugzilla.kernel.org/show_bug.cgi?id=199511.
Since the IA32_ARCH_CAPABILITIES MSR is emulated in kvm, there is no
dependency on hardware support for this feature.
The CPUID bits of OSXSAVE (function=0x1) and OSPKE (func=0x7, leaf=0x0)
allows user apps to detect if OS has set CR4.OSXSAVE or CR4.PKE. KVM is
supposed to update these CPUID bits when CR4 is updated. Current KVM
code doesn't handle some special cases when updates come from emulator.
Here is one example:
Step 1: guest boots
Step 2: guest OS enables XSAVE ==> CR4.OSXSAVE=1 and CPUID.OSXSAVE=1
Step 3: guest hot reboot ==> QEMU reset CR4 to 0, but CPUID.OSXAVE==1
Step 4: guest os checks CPUID.OSXAVE, detects 1, then executes xgetbv
Step 4 above will cause an #UD and guest crash because guest OS hasn't
turned on OSXAVE yet. This patch solves the problem by comparing the the
old_cr4 with cr4. If the related bits have been changed,
kvm_update_cpuid() needs to be called.
The X86_FEATURE_SSBD is an synthetic CPU feature - that is
it bit location has no relevance to the real CPUID 0x7.EBX[31]
bit position. For that we need the new CPU feature name.
Fixes: 52817587e706 ("x86/cpufeatures: Disentangle SSBD enumeration") Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: kvm@vger.kernel.org Cc: "Radim Krčmář" <rkrcmar@redhat.com> Cc: stable@vger.kernel.org Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Link: https://lkml.kernel.org/r/20180521215449.26423-2-konrad.wilk@oracle.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix this by sanitizing *resource* before using it to index
current->signal->rlim
Notice that given that speculation windows are large, the policy is to
kill the speculation on the first load and not worry if it can be
completed with a dependent load/store [1].
KASAN uses different routines to map shadow for hot added memory and
memory obtained in boot process. Attempt to offline memory onlined by
normal boot process leads to this:
Trying to vfree() nonexistent vm area (000000005d3b34b9)
WARNING: CPU: 2 PID: 13215 at mm/vmalloc.c:1525 __vunmap+0x147/0x190
Obviously we can't call vfree() to free memory that wasn't allocated via
vmalloc(). Use find_vm_area() to see if we can call vfree().
Unfortunately it's a bit tricky to properly unmap and free shadow
allocated during boot, so we'll have to keep it. If memory will come
online again that shadow will be reused.
Matthew asked: how can you call vfree() on something that isn't a
vmalloc address?
vfree() is able to free any address returned by
__vmalloc_node_range(). And __vmalloc_node_range() gives you any
address you ask. It doesn't have to be an address in [VMALLOC_START,
VMALLOC_END] range.
That's also how the module_alloc()/module_memfree() works on
architectures that have designated area for modules.
[aryabinin@virtuozzo.com: improve comments] Link: http://lkml.kernel.org/r/dabee6ab-3a7a-51cd-3b86-5468718e0390@virtuozzo.com
[akpm@linux-foundation.org: fix typos, reflow comment] Link: http://lkml.kernel.org/r/20180201163349.8700-1-aryabinin@virtuozzo.com Fixes: fa69b5989bb0 ("mm/kasan: add support for memory hotplug") Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com> Reported-by: Paul Menzel <pmenzel+linux-kasan-dev@molgen.mpg.de> Cc: Alexander Potapenko <glider@google.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Matthew Wilcox <willy@infradead.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>
shmat()'s SHM_REMAP option forbids passing a nil address for; this is in
fact the very first thing we check for. Andrea reported that for
SHM_RND|SHM_REMAP cases we can end up bypassing the initial addr check,
but we need to check again if the address was rounded down to nil. As
of this patch, such cases will return -EINVAL.
Link: http://lkml.kernel.org/r/20180503204934.kk63josdu6u53fbd@linux-n805 Signed-off-by: Davidlohr Bueso <dbueso@suse.de> Reported-by: Andrea Arcangeli <aarcange@redhat.com> Cc: Joe Lawrence <joe.lawrence@redhat.com> Cc: Manfred Spraul <manfred@colorfullife.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>
Patch series "ipc/shm: shmat() fixes around nil-page".
These patches fix two issues reported[1] a while back by Joe and Andrea
around how shmat(2) behaves with nil-page.
The first reverts a commit that it was incorrectly thought that mapping
nil-page (address=0) was a no no with MAP_FIXED. This is not the case,
with the exception of SHM_REMAP; which is address in the second patch.
I chose two patches because it is easier to backport and it explicitly
reverts bogus behaviour. Both patches ought to be in -stable and ltp
testcases need updated (the added testcase around the cve can be
modified to just test for SHM_RND|SHM_REMAP).
Commit 95e91b831f87 ("ipc/shm: Fix shmat mmap nil-page protection")
worked on the idea that we should not be mapping as root addr=0 and
MAP_FIXED. However, it was reported that this scenario is in fact
valid, thus making the patch both bogus and breaks userspace as well.
For example X11's libint10.so relies on shmat(1, SHM_RND) for lowmem
initialization[1].
If the radix tree underlying the IDR happens to be full and we attempt
to remove an id which is larger than any id in the IDR, we will call
__radix_tree_delete() with an uninitialised 'slot' pointer, at which
point anything could happen. This was easiest to hit with a single
entry at id 0 and attempting to remove a non-0 id, but it could have
happened with 64 entries and attempting to remove an id >= 64.
Roman said:
The syzcaller test boils down to opening /dev/kvm, creating an
eventfd, and calling a couple of KVM ioctls. None of this requires
superuser. And the result is dereferencing an uninitialized pointer
which is likely a crash. The specific path caught by syzbot is via
KVM_HYPERV_EVENTD ioctl which is new in 4.17. But I guess there are
other user-triggerable paths, so cc:stable is probably justified.
Matthew added:
We have around 250 calls to idr_remove() in the kernel today. Many of
them pass an ID which is embedded in the object they're removing, so
they're safe. Picking a few likely candidates:
drivers/firewire/core-cdev.c looks unsafe; the ID comes from an ioctl.
drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c is similar
drivers/atm/nicstar.c could be taken down by a handcrafted packet
Link: http://lkml.kernel.org/r/20180518175025.GD6361@bombadil.infradead.org Fixes: 0a835c4f090a ("Reimplement IDR and IDA using the radix tree") Reported-by: <syzbot+35666cba7f0a337e2e79@syzkaller.appspotmail.com> Debugged-by: Roman Kagan <rkagan@virtuozzo.com> Signed-off-by: Matthew Wilcox <mawilcox@microsoft.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>
We're casting the CDROM layer request_sense to the SCSI sense
buffer, but the former is 64 bytes and the latter is 96 bytes.
As we generally allocate these on the stack, we end up blowing
up the stack.
Fix this by wrapping the scsi_execute() call with a properly
sized sense buffer, and copying back the bits for the CDROM
layer.
Cc: stable@vger.kernel.org Reported-by: Piotr Gabriel Kosinski <pg.kosinski@gmail.com> Reported-by: Daniel Shapira <daniel@twistlock.com> Tested-by: Kees Cook <keescook@chromium.org> Fixes: 82ed4db499b8 ("block: split scsi_request out of struct request") Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
User-space may invoke ibv_reg_mr and ibv_dereg_mr in different threads.
If ibv_dereg_mr is called after the thread which invoked ibv_reg_mr has
exited, get_pid_task will return NULL and ib_umem_release will not
decrease mm->pinned_vm.
Instead of using threads to locate the mm, use the overall tgid from the
ib_ucontext struct instead. This matches the behavior of ODP and
disassociate in handling the mm of the process that called ibv_reg_mr.
Cc: <stable@vger.kernel.org> Fixes: 87773dd56d54 ("IB: ib_umem_release() should decrement mm->pinned_vm from ib_umem_get") Signed-off-by: Lidong Chen <lidongchen@tencent.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A pio send egress error can occur when the PSM library attempts to
to send a bad packet. That issue is still being investigated.
The pio error interrupt handler then attempts to progress the recovery
of the errored pio send context.
Code inspection reveals that the handling lacks the necessary locking
if that recovery interleaves with a PSM close of the "context" object
contains the pio send context.
The lack of the locking can cause the recovery to access the already
freed pio send context object and incorrectly deduce that the pio
send context is actually a kernel pio send context as shown by the
NULL deref stack below:
Clear the PCR (Processor Compatibility Register) on boot to ensure we
are not running in a compatibility mode.
We've seen this cause problems when a crash (and kdump) occurs while
running compat mode guests. The kdump kernel then runs with the PCR
set and causes problems. The symptom in the kdump kernel (also seen in
petitboot after fast-reboot) is early userspace programs taking
sigills on newer instructions (seen in libc).
Signed-off-by: Michael Neuling <mikey@neuling.org> Cc: stable@vger.kernel.org Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For LSE atomics that read and write a register operand, we need to
ensure that these operands are annotated as "early clobber" if the
register is written before all of the input operands have been consumed.
Failure to do so can result in the compiler allocating the same register
to both operands, leading to splats such as:
Unable to handle kernel paging request at virtual address 11111122222221
[...]
x1 : 1111111122222222 x0 : 1111111122222221
Process swapper/0 (pid: 1, stack limit = 0x000000008209f908)
Call trace:
test_atomic64+0x1360/0x155c
where x0 has been allocated as both the value to be stored and also the
atomic_t pointer.
This patch adds the missing clobbers.
Cc: <stable@vger.kernel.org> Cc: Dave Martin <dave.martin@arm.com> Cc: Robin Murphy <robin.murphy@arm.com> Reported-by: Mark Salter <msalter@redhat.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Depending on whether the kernel is compiled with frame-pointer or not,
the temporary memory location used for the bp parameter in these macros
is referenced relative to the stack pointer or the frame pointer.
Hence we can never reference that parameter when we've modified either
the stack pointer or the frame pointer, because then the compiler would
generate an incorrect stack reference.
Fix this by pushing the temporary memory parameter on a known location on
the stack before modifying the stack- and frame pointers.
Cc: <stable@vger.kernel.org> Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Reviewed-by: Brian Paul <brianp@vmware.com> Reviewed-by: Sinclair Yeh <syeh@vmware.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When run raidconfig from Dom0 we found that the Xen DMA heap is reduced,
but Dom Heap is increased by the same size. Tracing raidconfig we found
that the related ioctl() in megaraid_sas will call dma_alloc_coherent()
to apply memory. If the memory allocated by Dom0 is not in the DMA area,
it will exchange memory with Xen to meet the requiment. Later drivers
call dma_free_coherent() to free the memory, on xen_swiotlb_free_coherent()
the check condition (dev_addr + size - 1 <= dma_mask) is always false,
it prevents calling xen_destroy_contiguous_region() to return the memory
to the Xen DMA heap.
This issue introduced by commit 6810df88dcfc2 "xen-swiotlb: When doing
coherent alloc/dealloc check before swizzling the MFNs.".
Signed-off-by: Joe Jin <joe.jin@oracle.com> Tested-by: John Sobecki <john.sobecki@oracle.com> Reviewed-by: Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: stable@vger.kernel.org Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
While whitelisting Micron M500DC drives, the tweaked blacklist entry
enabled queued TRIM from M500IT variants also. But these do not support
queued TRIM. And while using those SSDs with the latest kernel we have
seen errors and even the partition table getting corrupted.
Some part from the dmesg:
[ 6.727384] ata1.00: ATA-9: Micron_M500IT_MTFDDAK060MBD, MU01, max UDMA/133
[ 6.727390] ata1.00: 117231408 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 6.741026] ata1.00: supports DRM functions and may not be fully accessible
[ 6.759887] ata1.00: configured for UDMA/133
[ 6.762256] scsi 0:0:0:0: Direct-Access ATA Micron_M500IT_MT MU01 PQ: 0 ANSI: 5
and then for the error:
[ 120.860334] ata1.00: exception Emask 0x1 SAct 0x7ffc0007 SErr 0x0 action 0x6 frozen
[ 120.860338] ata1.00: irq_stat 0x40000008
[ 120.860342] ata1.00: failed command: SEND FPDMA QUEUED
[ 120.860351] ata1.00: cmd 64/01:00:00:00:00/00:00:00:00:00/a0 tag 0 ncq dma 512 out
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x5 (timeout)
[ 120.860353] ata1.00: status: { DRDY }
[ 120.860543] ata1: hard resetting link
[ 121.166128] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 121.166376] ata1.00: supports DRM functions and may not be fully accessible
[ 121.186238] ata1.00: supports DRM functions and may not be fully accessible
[ 121.204445] ata1.00: configured for UDMA/133
[ 121.204454] ata1.00: device reported invalid CHS sector 0
[ 121.204541] sd 0:0:0:0: [sda] tag#18 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
[ 121.204546] sd 0:0:0:0: [sda] tag#18 Sense Key : 0x5 [current]
[ 121.204550] sd 0:0:0:0: [sda] tag#18 ASC=0x21 ASCQ=0x4
[ 121.204555] sd 0:0:0:0: [sda] tag#18 CDB: opcode=0x93 93 08 00 00 00 00 00 04 28 80 00 00 00 30 00 00
[ 121.204559] print_req_error: I/O error, dev sda, sector 272512
After few reboots with these errors, and the SSD is corrupted.
After blacklisting it, the errors are not seen and the SSD does not get
corrupted any more.
Fixes: 243918be6393 ("libata: Do not blacklist Micron M500DC") Cc: Martin K. Petersen <martin.petersen@oracle.com> Cc: stable@vger.kernel.org Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com> Signed-off-by: Tejun Heo <tj@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The SDHCI_QUIRK2_HOST_OFF_CARD_ON is needed for the driver to
properly reset the host controller (reset all) on initialization
after exiting deep sleep.
Signed-off-by: Corneliu Doban <corneliu.doban@broadcom.com> Signed-off-by: Scott Branden <scott.branden@broadcom.com> Reviewed-by: Ray Jui <ray.jui@broadcom.com> Reviewed-by: Srinath Mannam <srinath.mannam@broadcom.com> Fixes: c833e92bbb60 ("mmc: sdhci-iproc: support standard byte register accesses") Cc: stable@vger.kernel.org # v4.10+ Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When the host controller accepts only 32bit writes, the value of the
16bit TRANSFER_MODE register, that has the same 32bit address as the
16bit COMMAND register, needs to be saved and it will be written
in a 32bit write together with the command as this will trigger the
host to send the command on the SD interface.
When sending the tuning command, TRANSFER_MODE is written and then
sdhci_set_transfer_mode reads it back to clear AUTO_CMD12 bit and
write it again resulting in wrong value to be written because the
initial write value was saved in a shadow and the read-back returned
a wrong value, from the register.
Fix sdhci_iproc_readw to return the saved value of TRANSFER_MODE
when a saved value exist.
Same fix for read of BLOCK_SIZE and BLOCK_COUNT registers, that are
saved for a different reason, although a scenario that will cause the
mentioned problem on this registers is not probable.
Remove hard coded mmc cap 1.8v from platform data as it is board specific.
The 1.8v DDR mmc caps can be enabled using DTS property for those
boards that support it.
For anything NFS-exported we do _not_ want to unlock new inode
before it has grown an alias; original set of fixes got the
ordering right, but missed the nasty complication in case of
lockdep being enabled - unlock_new_inode() does
lockdep_annotate_inode_mutex_key(inode)
which can only be done before anyone gets a chance to touch
->i_mutex. Unfortunately, flipping the order and doing
unlock_new_inode() before d_instantiate() opens a window when
mkdir can race with open-by-fhandle on a guessed fhandle, leading
to multiple aliases for a directory inode and all the breakage
that follows from that.
Correct solution: a new primitive (d_instantiate_new())
combining these two in the right order - lockdep annotate, then
d_instantiate(), then the rest of unlock_new_inode(). All
combinations of d_instantiate() with unlock_new_inode() should
be converted to that.
Cc: stable@kernel.org # 2.6.29 and later Tested-by: Mike Marshall <hubcap@omnibond.com> Reviewed-by: Andreas Dilger <adilger@dilger.ca> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit f65e0d299807 ("ALSA: timer: Call notifier in the same spinlock")
combined the start/continue and stop/pause functions, and in doing so
changed the event code for the pause case to SNDRV_TIMER_EVENT_CONTINUE.
Change it back to SNDRV_TIMER_EVENT_PAUSE.
Fixes: f65e0d299807 ("ALSA: timer: Call notifier in the same spinlock") Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk> Cc: stable@vger.kernel.org Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
kill_ioctx() used to have an explicit RCU delay between removing the
reference from ->ioctx_table and percpu_ref_kill() dropping the refcount.
At some point that delay had been removed, on the theory that
percpu_ref_kill() itself contained an RCU delay. Unfortunately, that was
the wrong kind of RCU delay and it didn't care about rcu_read_lock() used
by lookup_ioctx(). As the result, we could get ctx freed right under
lookup_ioctx(). Tejun has fixed that in a6d7cff472e ("fs/aio: Add explicit
RCU grace period when freeing kioctx"); however, that fix is not enough.
Suppose io_destroy() from one thread races with e.g. io_setup() from another;
CPU1 removes the reference from current->mm->ioctx_table[...] just as CPU2
has picked it (under rcu_read_lock()). Then CPU1 proceeds to drop the
refcount, getting it to 0 and triggering a call of free_ioctx_users(),
which proceeds to drop the secondary refcount and once that reaches zero
calls free_ioctx_reqs(). That does
INIT_RCU_WORK(&ctx->free_rwork, free_ioctx);
queue_rcu_work(system_wq, &ctx->free_rwork);
and schedules freeing the whole thing after RCU delay.
In the meanwhile CPU2 has gotten around to percpu_ref_get(), bumping the
refcount from 0 to 1 and returned the reference to io_setup().
Tejun's fix (that queue_rcu_work() in there) guarantees that ctx won't get
freed until after percpu_ref_get(). Sure, we'd increment the counter before
ctx can be freed. Now we are out of rcu_read_lock() and there's nothing to
stop freeing of the whole thing. Unfortunately, CPU2 assumes that since it
has grabbed the reference, ctx is *NOT* going away until it gets around to
dropping that reference.
The fix is obvious - use percpu_ref_tryget_live() and treat failure as miss.
It's not costlier than what we currently do in normal case, it's safe to
call since freeing *is* delayed and it closes the race window - either
lookup_ioctx() comes before percpu_ref_kill() (in which case ctx->users
won't reach 0 until the caller of lookup_ioctx() drops it) or lookup_ioctx()
fails, ctx->users is unaffected and caller of lookup_ioctx() doesn't see
the object in question at all.
Cc: stable@kernel.org Fixes: a6d7cff472e "fs/aio: Add explicit RCU grace period when freeing kioctx" Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We recently had an oops reported on a 4.14 kernel in
xfs_reclaim_inodes_count() where sb->s_fs_info pointed to garbage
and so the m_perag_tree lookup walked into lala land. It produces
an oops down this path during the failed mount:
The problem is that the superblock shrinker is running before the
filesystem structures it depends on have been fully set up. i.e.
the shrinker is registered in sget(), before ->fill_super() has been
called, and the shrinker can call into the filesystem before
fill_super() does it's setup work. Essentially we are exposed to
both use-after-free and use-before-initialisation bugs here.
To fix this, add a check for the SB_BORN flag in super_cache_count.
In general, this flag is not set until ->fs_mount() completes
successfully, so we know that it is set after the filesystem
setup has completed. This matches the trylock_super() behaviour
which will not let super_cache_scan() run if SB_BORN is not set, and
hence will not allow the superblock shrinker from entering the
filesystem while it is being set up or after it has failed setup
and is being torn down.
Cc: stable@kernel.org Signed-Off-By: Dave Chinner <dchinner@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
we unlock the directory hash too early - if we are looking at secondary
link and primary (in another directory) gets removed just as we unlock,
we could have the old primary moved in place of the secondary, leaving
us to look into freed entry (and leaving our dentry with ->d_fsdata
pointing to a freed entry).
Cc: stable@vger.kernel.org # 2.4.4+ Acked-by: David Sterba <dsterba@suse.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Check the TIF_32BIT_FPREGS task setting of the tracee rather than the
tracer in determining the layout of floating-point general registers in
the floating-point context, correcting access to odd-numbered registers
for o32 tracees where the setting disagrees between the two processes.
Fixes: 597ce1723e0f ("MIPS: Support for 64-bit FP with O32 binaries") Signed-off-by: Maciej W. Rozycki <macro@mips.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: linux-mips@linux-mips.org Cc: <stable@vger.kernel.org> # 3.14+ Signed-off-by: James Hogan <jhogan@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Correct commit 7aeb753b5353 ("MIPS: Implement task_user_regset_view.")
and expose the FIR register using the unused 4 bytes at the end of the
NT_PRFPREG regset. Without that register included clients cannot use
the PTRACE_GETREGSET request to retrieve the complete FPU register set
and have to resort to one of the older interfaces, either PTRACE_PEEKUSR
or PTRACE_GETFPREGS, to retrieve the missing piece of data. Also the
register is irreversibly missing from core dumps.
This register is architecturally hardwired and read-only so the write
path does not matter. Ignore data supplied on writes then.
Fixes: 7aeb753b5353 ("MIPS: Implement task_user_regset_view.") Signed-off-by: James Hogan <jhogan@kernel.org> Signed-off-by: Maciej W. Rozycki <macro@mips.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: linux-mips@linux-mips.org Cc: <stable@vger.kernel.org> # 3.13+
Patchwork: https://patchwork.linux-mips.org/patch/19273/ Signed-off-by: James Hogan <jhogan@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When DMA will be performed to a MIPS32 1004K CPS, the L1-cache for the
range needs to be flushed and invalidated first.
The code currently takes one of two approaches.
1/ If the range is less than the size of the dcache, then HIT type
requests flush/invalidate cache lines for the particular addresses.
HIT-type requests a globalised by the CPS so this is safe on SMP.
2/ If the range is larger than the size of dcache, then INDEX type
requests flush/invalidate the whole cache. INDEX type requests affect
the local cache only. CPS does not propagate them in any way. So this
invalidation is not safe on SMP CPS systems.
Data corruption due to '2' can quite easily be demonstrated by
repeatedly "echo 3 > /proc/sys/vm/drop_caches" and then sha1sum a file
that is several times the size of available memory. Dropping caches
means that large contiguous extents (large than dcache) are more likely.
This was not a problem before Linux-4.8 because option 2 was never used
if CONFIG_MIPS_CPS was defined. The commit which removed that apparently
didn't appreciate the full consequence of the change.
We could, in theory, globalize the INDEX based flush by sending an IPI
to other cores. These cache invalidation routines can be called with
interrupts disabled and synchronous IPI require interrupts to be
enabled. Asynchronous IPI may not trigger writeback soon enough. So we
cannot use IPI in practice.
We can already test if IPI would be needed for an INDEX operation with
r4k_op_needs_ipi(R4K_INDEX). If this is true then we mustn't try the
INDEX approach as we cannot use IPI. If this is false (e.g. when there
is only one core and hence one L1 cache) then it is safe to use the
INDEX approach without IPI.
This patch avoids options 2 if r4k_op_needs_ipi(R4K_INDEX), and so
eliminates the corruption.
Fix the following warning in MIPS allmodconfig by adding a
MODULE_LICENSE() at the end of rtc-goldfish.c, based on the file header
comment which says GNU General Public License version 2:
WARNING: modpost: missing MODULE_LICENSE() in drivers/rtc/rtc-goldfish.o
The shifting of buf[5] by 24 bits to the left will be promoted to
a 32 bit signed int and then sign-extended to an unsigned long. If
the top bit of buf[5] is set then all then all the upper bits sec
end up as also being set because of the sign-extension. Fix this by
casting buf[5] to an unsigned long before the shift.
Detected by CoverityScan, CID#1465292 ("Unintended sign extension")
Fixes: 0e1492330cd2 ("rtc: add rtc-tx4939 driver") Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Also, there is another possible race condition. The probe function is not
allowed to fail after the RTC is registered because the following may
happen:
On 32bit platforms, time_t is still a signed 32bit long. If it is
overflowed, userspace and the kernel cant agree on the current system time.
This causes multiple issues, in particular with systemd:
https://github.com/systemd/systemd/issues/1143
A good workaround is to simply avoid using hctosys which is something I
greatly encourage as the time is better set by userspace.
However, many distribution enable it and use systemd which is rendering the
system unusable in case the RTC holds a date after 2038 (and more so after
2106). Many drivers have workaround for this case and they should be
eliminated so there is only one place left to fix when userspace is able to
cope with dates after the 31bit overflow.
commit 179a502f8c46 ("rtc: snvs: add Freescale rtc-snvs driver") introduces
the SNVS RTC driver with a function snvs_rtc_enable().
snvs_rtc_enable() can return an error on the enable path however this
driver does not currently trap that failure on the probe() path and
consequently if enabling the RTC fails we encounter a later error spinning
forever in rtc_write_sync_lp().
[ 36.093481] [<c010d630>] (__irq_svc) from [<c0c2e9ec>] (_raw_spin_unlock_irqrestore+0x34/0x44)
[ 36.102122] [<c0c2e9ec>] (_raw_spin_unlock_irqrestore) from [<c072e32c>] (regmap_read+0x4c/0x5c)
[ 36.110938] [<c072e32c>] (regmap_read) from [<c085d0f4>] (rtc_write_sync_lp+0x6c/0x98)
[ 36.118881] [<c085d0f4>] (rtc_write_sync_lp) from [<c085d160>] (snvs_rtc_alarm_irq_enable+0x40/0x4c)
[ 36.128041] [<c085d160>] (snvs_rtc_alarm_irq_enable) from [<c08567b4>] (rtc_timer_do_work+0xd8/0x1a8)
[ 36.137291] [<c08567b4>] (rtc_timer_do_work) from [<c01441b8>] (process_one_work+0x28c/0x76c)
[ 36.145840] [<c01441b8>] (process_one_work) from [<c01446cc>] (worker_thread+0x34/0x58c)
[ 36.153961] [<c01446cc>] (worker_thread) from [<c014aee4>] (kthread+0x138/0x150)
[ 36.161388] [<c014aee4>] (kthread) from [<c0107e14>] (ret_from_fork+0x14/0x20)
[ 36.168635] rcu_sched kthread starved for 2602 jiffies! g496 c495 f0x2 RCU_GP_WAIT_FQS(3) ->state=0x0 ->cpu=0
[ 36.178564] rcu_sched R running task 0 8 2 0x00000000
[ 36.185664] [<c0c288b0>] (__schedule) from [<c0c29134>] (schedule+0x3c/0xa0)
[ 36.192739] [<c0c29134>] (schedule) from [<c0c2db80>] (schedule_timeout+0x78/0x4e0)
[ 36.200422] [<c0c2db80>] (schedule_timeout) from [<c01a7ab0>] (rcu_gp_kthread+0x648/0x1864)
[ 36.208800] [<c01a7ab0>] (rcu_gp_kthread) from [<c014aee4>] (kthread+0x138/0x150)
[ 36.216309] [<c014aee4>] (kthread) from [<c0107e14>] (ret_from_fork+0x14/0x20)
This patch fixes by parsing the result of rtc_write_sync_lp() and
propagating both in the probe and elsewhere. If the RTC doesn't start we
don't proceed loading the driver and don't get into this loop mess later
on.
Most register accesses in the altera driver honor port->regshift by
using altera_uart_writel(). There are a few accesses however that were
missed when the driver was converted to use port->regshift and some
others were added later in commit 4d9d7d896d77 ("serial: altera_uart:
add earlycon support").
Fixes: 2780ad42f5fe ("tty: serial: altera_uart: Use port->regshift to store bus shift") Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Acked-by: Tobias Klauser <tklauser@distanz.ch> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, data in RX FIFO is read based on UART_LSR register state even
if RDI and RLSI interrupts are disabled in UART_IER register.
This is because when IRQ handler is called due to TX FIFO empty event,
RX FIFO is serviced based on UART_LSR register status instead of
UART_IIR status. This defeats the purpose of disabling UART RX
FIFO interrupts during throttling(see, omap_8250_throttle()) as IRQ
handler continues to drain UART RX FIFO resulting in overflow of buffer
at tty layer.
Fix this by making sure that driver drains UART RX FIFO only when
UART_IIR_RDI is set along with UART_LSR_BI or UART_LSR_DR bits.
The imx_ports[] array is indexed using a value derived from the
"serialN" alias in DT, or from platform data, which may lead to an
out-of-bounds access.
The auart_port[] array is indexed using a value derived from the
"serialN" alias in DT, or from platform data, which may lead to an
out-of-bounds access.
The s3c24xx_serial_ports[] array is indexed using a value derived from
the "serialN" alias in DT, or from an incrementing probe index, which
may lead to an out-of-bounds access.
Fix this by adding a range check.
Note that the array size is defined by a Kconfig symbol
(CONFIG_SERIAL_SAMSUNG_UARTS), so this can even be triggered using
a legitimate DTB or legitimate board code.
Currently an out of range dev->nr is detected by just reporting the
issue and later on an out-of-bounds read on array card occurs because
of this. Fix this by checking the upper range of dev->nr with the size
of array card (removes the hard coded size), move this check earlier
and also exit with the error -ENOSYS to avoid the later out-of-bounds
array read.
Detected by CoverityScan, CID#711191 ("Out-of-bounds-read")
Fixes: commit 02b20b0b4cde ("V4L/DVB (12730): Add conexant cx25821 driver") Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
[hans.verkuil@cisco.com: %ld -> %zd] Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The vivid driver has two custom controls that change the behavior of RDS.
Depending on the control setting the V4L2_CAP_READWRITE capability is toggled.
However, after an earlier commit the capability was no longer set correctly.
This is now fixed.
Fixes: 9765a32cd8 ("vivid: set device_caps in video_device") Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixes vb2_vmalloc_get_userptr() to ioremap correct area.
Since the current code does ioremap the page address, if the offset > 0,
it does not do ioremap the last page and results in kernel panic.
This fixes to pass the size + offset to ioremap so that ioremap
can map correct area. Also, this uses __pfn_to_phys() to get the physical
address of given PFN.
The ADV748x handles interlaced media using V4L2_FIELD_ALTERNATE field
types. The correct specification for the height on the mbus is the
image height, in this instance, the field height.
The AFE component already correctly adjusts the height on the mbus, but
the HDMI component got left behind.
Adjust the mbus height to correctly describe the image height of the
fields when processing interlaced video for HDMI pipelines.
Make sure we don't accept more inputs than the hardware can handle. This
is a temporary fix to avoid display stall, we need to instead allocate
the BRU or BRS to display pipelines dynamically based on the number of
planes they each use.
Both lgdt33606a_release and lgdt3306a_remove kfree state, but _release is
called first, then _remove operates on states members before kfree'ing it.
This can lead to random oops/GPF/etc on USB disconnect.
While experimenting with older compiler versions, I ran
into a warning that no longer shows up on gcc-4.8 or newer:
drivers/media/platform/s3c-camif/camif-capture.c: In function '__camif_subdev_try_format':
drivers/media/platform/s3c-camif/camif-capture.c:1265:25: error: array subscript is below array bounds
This is an off-by-one bug, leading to an access before the start of the
array, while newer compilers silently assume this undefined behavior
cannot happen and leave the loop at index 0 if no other entry matches.
As Sylvester explains, we actually need to ensure that the
value is within the range, so this reworks the loop to be
easier to parse correctly, and an additional check to fall
back on the first format value for any unexpected input.
I found an existing gcc bug for it and added a reduced version
of the function there.
Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69249#c3 Fixes: babde1c243b2 ("[media] V4L: Add driver for S3C24XX/S3C64XX SoC series camera interface") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently clk_freq is ignored entirely, because the cx235840 driver
configures the xtal at the chip defaults. This is an issue if a
board is produced with a non-default frequency crystal. If clk_freq
is not zero the cx25840 will attempt to use the setting provided,
or fall back to defaults otherwise.
Hauppauge produced a revision of ImpactVCBe using an 888,
with a 25MHz crystal, instead of using the default third
overtone 50Mhz crystal. This overrides that frequency so
that the cx25840 is properly configured. Without the proper
crystal setup the cx25840 cannot load the firmware or
decode video.
The device node obtained with of_graph_get_next_endpoint() should be
released by calling of_node_put(). But it was not released when
v4l2_fwnode_endpoint_parse() failed.
This change moves the of_node_put() call before the error check and
fixes the issue.
As pointed by Dan, possible values for bits[3:0] of te Line Mode Registers
can range from 0x0 to 0xf, but the check logic allow values ranging
from 0x0 to 0xe.
As static arrays are initialized with zero, using a value without
an explicit initializer at the array won't cause any harm.
Returning -EINVAL when an ioctl is not implemented is a very
bad idea, as it is hard to distinguish from other error
contitions that an ioctl could lead. Replace it by its
right error code: -ENOTTY.
Rates declared in PLL rate tables should match exactly rates calculated
from PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
In this patch an erroneous P value for 74176002 output frequency is also
corrected.
Rates declared in PLL rate tables should match exactly rates calculated
from PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
Rates declared in PLL rate tables should match exactly rates calculated from
the PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
Rates declared in PLL rate tables should match exactly rates calculated from
the PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
Rates declared in PLL rate tables should match exactly rates calculated from
the PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
Rates declared in PLL rate tables should match exactly rates calculated from
the PLL coefficients. If that is not the case, rate of the PLL's child clock
might be set not as expected. For instance, if in the PLL rates table we have
a 393216000 Hz entry and the real value as returned by the PLL's recalc_rate
callback is 393216003, after setting PLL's clk rate to 393216000 clk_get_rate
will return 393216003. If we now attempt to set rate of a PLL's child divider
clock to 393216000/2 its rate will be 131072001, rather than 196608000.
That is, the divider will be set to 3 instead of 2, because 393216003/2 is
greater than 196608000.
To fix this issue declared rates are changed to exactly match rates generated
by the PLL, as calculated from the P, M, S, K coefficients.
The MMC sample and drv clock for rockchip platforms are derived from
the bus clock output to the MMC/SDIO card. So it should never happens
that the clk rate is zero given it should inherits the clock rate from
its parent. If something goes wrong and makes the clock rate to be zero,
the calculation would be wrong but may still make the mmc tuning process
work luckily. However it makes people harder to debug when the following
data transfer is unstable.
Of course this won't quite work leading to the following messages:
[ 6.559593] usb 2-1: new full-speed USB device number 2 using tegra-
ehci
[ 11.759173] usb 2-1: device descriptor read/64, error -110
[ 27.119453] usb 2-1: device descriptor read/64, error -110
[ 27.389217] usb 2-1: new full-speed USB device number 3 using tegra-
ehci
[ 32.559454] usb 2-1: device descriptor read/64, error -110
[ 47.929777] usb 2-1: device descriptor read/64, error -110
[ 48.049658] usb usb2-port1: attempt power cycle
[ 48.759475] usb 2-1: new full-speed USB device number 4 using tegra-
ehci
[ 59.349457] usb 2-1: device not accepting address 4, error -110
[ 59.509449] usb 2-1: new full-speed USB device number 5 using tegra-
ehci
[ 70.069457] usb 2-1: device not accepting address 5, error -110
[ 70.079721] usb usb2-port1: unable to enumerate USB device
Fix this by actually allowing the rate also being set from within
the Linux kernel.
It exposes an issue that clk core, clk_core_get_phase, always
returns the cached core->phase which should be either updated
by calling clk_set_phase or directly from the first place the
clk was registered.
When registering the clk, the core->phase geting from ->get_phase()
may return negative value indicating error. This is quite common
since the clk's phase may be highly related to its parent chain,
but it was temporarily orphan when registered, since its parent
chains hadn't be ready at that time, so the clk drivers decide to
return error in this case. However, if no clk_set_phase is called or
maybe the ->set_phase() isn't even implemented, the core->phase would
never be updated. This is wrong, and we should try to update it when
all its parent chains are settled down, like the way of updating clock
rate for that. But it's not deserved to complicate the code now and
just update it anyway when calling clk_core_get_phase, which would be
much simple and enough.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com> Acked-by: Jerome Brunet <jbrunet@baylibre.com> Signed-off-by: Stephen Boyd <sboyd@kernel.org> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c420c1e4db22 ("clk: rockchip: Prevent calculating mmc phase
if clock rate is zero") catches one gremlin again for clk-rk3228.c
that the parent of SDMMC phase clock should be sclk_sdmmc0, but not
sclk_sdmmc. However, the naming of the sdmmc clocks varies in the
manual with the card clock having the 0 while the hclk is named
without appended 0. So standardize one one format to prevent
confusion, as there also is only one (non-sdio) mmc controller on
the soc.
If the RCLK mux clock configuration is specified in DT and no set_sysclk()
callback is used in the sound card driver the sclk_srcrate field will remain
set to 0, leading to an incorrect PSR divider setting.
To fix this the frequency value is retrieved from the CLK_I2S_RCLK_SRC clock,
so the actual RCLK mux selection is taken into account.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Acked-by: Krzysztof Kozlowski <krzk@kernel.org> Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds the change required to create the TLV data
for dapm widget kcontrols from topology. This also fixes the following
TLV read error shown in amixer while showing the card control contents.
"amixer: Control hw:1 element TLV read error: No such device or address"
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case of sample rates lower than 44100 currently there is too low MCLK
frequency set for the CODEC. Playback fails with following errors:
$ speaker-test -c2 -t sine -f 1500 -l2 -r 32000
Sine wave rate is 1500.0000Hz
Rate set to 32000Hz (requested 32000Hz)
Buffer size range from 128 to 131072
Period size range from 64 to 65536
Using max buffer size 131072
Periods = 4
Unable to set hw params for playback: Invalid argument
Setting of hwparams failed: Invalid argument
[ 497.883700] max98090 1-0010: Invalid master clock frequency
To fix this the I2S root clock's frequency is increased, depending
on sampling rate.
Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Acked-by: Krzysztof Kozlowski <krzk@kernel.org> Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The hcp->chmap_info must not be freed up in the hdmi_codec_remove()
function as it leads to kernel crash due ALSA core's
pcm_chmap_ctl_private_free() is trying to free it up again when the card
destroyed via snd_card_free.
Commit cd6111b26280a ("ASoC: hdmi-codec: add channel mapping control")
should not have added the kfree(hcp->chmap_info); to the hdmi_codec_remove
function.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com> Reviewed-by: Jyri Sarha <jsarha@ti.com> Tested-by: Jyri Sarha <jsarha@ti.com> Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver controls when the hardware sends completions that communicate
consumption of elements from the WQ. This is done by setting a WQEC bit
on a WQE.
The current driver sets it on every Nth WQE posting. However, the driver
isn't clearing the bit if the WQE is reused. Thus, if the queue depth
isn't evenly divisible by N, with enough time, it can be set on every
element, creating a lot of overhead and risking CQ full conditions.
Correct by clearing the bit when not setting it on an Nth element.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver was simultaneously setting a combination of flags that caused
lpfc_do_work()to effectively spin between slow path work and new event
data, causing the lockup.
Ensure in the typical wq completions, that new event data flags are set
if the slow path flag is running. The slow path will eventually
reschedule the wq handling.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver ignored checks on whether the link should be kept
administratively down after a link bounce. Correct the checks.
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com> Signed-off-by: James Smart <james.smart@broadcom.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The SCSI PRE-FETCH (10 or 16) command is present both on hard disks
and some SSDs. It is useful when the address of the next block(s) to
be read is known but it is not following the LBA of the current READ
(so read-ahead won't help). It returns two "good" SCSI Status values.
If the requested blocks have fitted (or will most likely fit (when
the IMMED bit is set)) into the disk's cache, it returns CONDITION
MET. If it didn't (or will not) fit then it returns GOOD status.
The goal of this patch is to stop the SCSI subsystem treating the
CONDITION MET SCSI status as an error. The current state makes the
PRE-FETCH command effectively unusable via pass-throughs.
Signed-off-by: Douglas Gilbert <dgilbert@interlog.com> Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If a recursive IOP_RESET is invoked, usually due to the eh_thread
handling errors after the first reset, be sure we flag that the command
thread has been stopped to avoid an Oops of the form;
iscsi tcp will first send out data, then calculate and send data
digest. If we don't have BDI_CAP_STABLE_WRITES, the page cache will be
written in spite of the on going writeback. Consequently, wrong digest
will be got and sent to target.
To fix this, set BDI_CAP_STABLE_WRITES when data digest is enabled
in iscsi_tcp .slave_configure callback.
Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com> Acked-by: Chris Leech <cleech@redhat.com> Acked-by: Lee Duncan <lduncan@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit reads the disk's old state and combines it with the device
disk-reported state rather than unconditionally marking it as RW.
Reported-by: Li Ning <lining916740672@icloud.com> Signed-off-by: Jeremy Cline <jeremy@jcline.org> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The firmware event workqueue should not be marked as WQ_MEM_RECLAIM
as it's doesn't need to make forward progress under memory pressure.
In the current state it will result in a deadlock if the device had been
forcefully removed.
Cc: Sreekanth Reddy <sreekanth.reddy@broadcom.com> Cc: Suganath Prabu Subramani <suganath-prabu.subramani@broadcom.com> Acked-by: Sreekanth Reddy <Sreekanth.Reddy@broadcom.com> Signed-off-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
and that assumes shost has been assigned with pci_set_drvdata().
However, pci_set_drvdata(pdev, shost) is done in aac_probe_one() far
after bailing out with error from calling the init function
((*aac_drivers[index].init)(aac)), and when the init function fails, no
error is returned from aac_probe_one() so PCI layer assumes there is
driver attached, and tries to shut it down later.
Fix it by returning error from aac_probe_one() when card-specific init
function fails.
This fixes reboot on my HP NetRAID-4M with dead battery.
Signed-off-by: Meelis Roos <mroos@linux.ee> Reviewed-by: Dave Carroll <david.carroll@microsemi.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The data in NVRAM is not guaranteed to be NUL terminated. Since
snprintf expects byte-stream to accommodate null byte, the CHAP secret
is truncated. Use sprintf instead of snprintf to fix the truncation of
CHAP name and secret.
Signed-off-by: Andrew Vasquez <andrew.vasquez@cavium.com> Signed-off-by: Nilesh Javali <nilesh.javali@cavium.com> Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com> Acked-by: Chris Leech <cleech@redhat.com> Acked-by: Lee Duncan <lduncan@suse.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Increase cmd_per_lun to allow more I/Os in progress per device,
particularly for NVMe's. The Hyper-V host side can handle the higher
count with no issues.
Signed-off-by: Michael Kelley <mikelley@microsoft.com> Reviewed-by: K. Y. Srinivasan <kys@microsoft.com> Acked-by: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We wanted to exit the loop with "div" set to zero, but instead, if we
don't hit the break then "div" is -1 when we finish the loop. It leads
to an array underflow a few lines later.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de> Acked-by: Matthew Wilcox <mawilcox@microsoft.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a request times out we set the io_req flag BNX2FC_FLAG_IO_COMPL so
that if a subsequent completion comes in on that task ID we will ignore
it. The issue is that in the check for this flag there is a missing
return so we will continue to process a request which may have already
been returned to the ownership of the SCSI layer. This can cause
unpredictable results.
Solution is to add in the missing return.
[mkp: typo plus title shortening]
Signed-off-by: Chad Dupuis <chad.dupuis@cavium.com> Reviewed-by: Laurence Oberman <loberman@redhat.com> Tested-by: Laurence Oberman <loberman@redhat.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cpu_msix_table is allocated to store online cpus, but pci_irq_get_affinity
may return cpu_possible_mask which is then used to access cpu_msix_table.
That causes bad user experience. Fix limits access to only online cpus,
I've also added an additional test to protect from an unlikely change in
cpu_online_mask.
[mkp: checkpatch]
Fixes: 1d55abc0e98a ("scsi: mpt3sas: switch to pci_alloc_irq_vectors") Signed-off-by: Tomas Henzl <thenzl@redhat.com> Acked-by: Suganath Prabu Subramani <suganath-prabu.subramani@broadcom.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When exiting a transformation, the cra_exit() helper is called in each
driver providing one. The Inside Secure SafeXcel driver has one, which
is responsible of freeing some areas and of sending one invalidation
request to the crypto engine, to invalidate the context that was used
during the transformation.
We could see in some setups (when lots of transformations were being
used with a short lifetime, and hence lots of cra_exit() calls) NULL
pointer dereferences and other weird issues. All these issues were
coming from accessing the tfm context.
The issue is the invalidation request completion is checked using a
wait_for_completion_interruptible() call in both the cipher and hash
cra_exit() helpers. In some cases this was interrupted while the
invalidation request wasn't processed yet. And then cra_exit() returned,
and its caller was freeing the tfm instance. Only then the request was
being handled by the SafeXcel driver, which lead to the said issues.
This patch fixes this by using wait_for_completion() calls in these
specific cases.
projects / users / jedix / linux-maple.git / log