platform/kernel/linux-rpi.git
4 years agonfsd: fsnotify on rmdir under nfsd/clients/
J. Bruce Fields [Thu, 19 Mar 2020 15:30:38 +0000 (11:30 -0400)]
nfsd: fsnotify on rmdir under nfsd/clients/

commit 69afd267982e733a555fede4e85fe30329ed0588 upstream.

Userspace should be able to monitor nfsd/clients/ to see when clients
come and go, but we're failing to send fsnotify events.

Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agopowerpc/pseries: Avoid NULL pointer dereference when drmem is unavailable
Libor Pechacek [Fri, 31 Jan 2020 13:28:29 +0000 (14:28 +0100)]
powerpc/pseries: Avoid NULL pointer dereference when drmem is unavailable

commit a83836dbc53e96f13fec248ecc201d18e1e3111d upstream.

In guests without hotplugagble memory drmem structure is only zero
initialized. Trying to manipulate DLPAR parameters results in a crash.

  $ echo "memory add count 1" > /sys/kernel/dlpar
  Oops: Kernel access of bad area, sig: 11 [#1]
  LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA pSeries
  ...
  NIP:  c0000000000ff294 LR: c0000000000ff248 CTR: 0000000000000000
  REGS: c0000000fb9d3880 TRAP: 0300   Tainted: G            E      (5.5.0-rc6-2-default)
  MSR:  8000000000009033 <SF,EE,ME,IR,DR,RI,LE>  CR: 28242428  XER: 20000000
  CFAR: c0000000009a6c10 DAR: 0000000000000010 DSISR: 40000000 IRQMASK: 0
  ...
  NIP dlpar_memory+0x6e4/0xd00
  LR  dlpar_memory+0x698/0xd00
  Call Trace:
    dlpar_memory+0x698/0xd00 (unreliable)
    handle_dlpar_errorlog+0xc0/0x190
    dlpar_store+0x198/0x4a0
    kobj_attr_store+0x30/0x50
    sysfs_kf_write+0x64/0x90
    kernfs_fop_write+0x1b0/0x290
    __vfs_write+0x3c/0x70
    vfs_write+0xd0/0x260
    ksys_write+0xdc/0x130
    system_call+0x5c/0x68

Taking closer look at the code, I can see that for_each_drmem_lmb is a
macro expanding into `for (lmb = &drmem_info->lmbs[0]; lmb <=
&drmem_info->lmbs[drmem_info->n_lmbs - 1]; lmb++)`. When drmem_info->lmbs
is NULL, the loop would iterate through the whole address range if it
weren't stopped by the NULL pointer dereference on the next line.

This patch aligns for_each_drmem_lmb and for_each_drmem_lmb_in_range
macro behavior with the common C semantics, where the end marker does
not belong to the scanned range, and alters get_lmb_range() semantics.
As a side effect, the wraparound observed in the crash is prevented.

Fixes: 6c6ea53725b3 ("powerpc/mm: Separate ibm, dynamic-memory data from DT format")
Cc: stable@vger.kernel.org # v4.16+
Signed-off-by: Libor Pechacek <lpechacek@suse.cz>
Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200131132829.10281-1-msuchanek@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodrm/amdgpu: unify fw_write_wait for new gfx9 asics
Aaron Liu [Tue, 7 Apr 2020 09:46:04 +0000 (17:46 +0800)]
drm/amdgpu: unify fw_write_wait for new gfx9 asics

commit 2960758cce2310774de60bbbd8d6841d436c54d9 upstream.

Make the fw_write_wait default case true since presumably all new
gfx9 asics will have updated firmware. That is using unique WAIT_REG_MEM
packet with opration=1.

Signed-off-by: Aaron Liu <aaron.liu@amd.com>
Tested-by: Aaron Liu <aaron.liu@amd.com>
Tested-by: Yuxian Dai <Yuxian.Dai@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodrm/amdgpu/powerplay: using the FCLK DPM table to set the MCLK
Yuxian Dai [Wed, 1 Apr 2020 11:26:26 +0000 (19:26 +0800)]
drm/amdgpu/powerplay: using the FCLK DPM table to set the MCLK

commit 022ac4c9c55be35a2d1f71019a931324c51b0dab upstream.

1.Using the FCLK DPM table to set the MCLK for DPM states consist of
three entities:
 FCLK
 UCLK
 MEMCLK
All these three clk change together, MEMCLK from FCLK, so use the fclk
frequency.
2.we should show the current working clock freqency from clock table metric

Signed-off-by: Yuxian Dai <Yuxian.Dai@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Kevin Wang <Kevin1.Wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodrm: Remove PageReserved manipulation from drm_pci_alloc
Chris Wilson [Sun, 2 Feb 2020 17:16:31 +0000 (17:16 +0000)]
drm: Remove PageReserved manipulation from drm_pci_alloc

commit ea36ec8623f56791c6ff6738d0509b7920f85220 upstream.

drm_pci_alloc/drm_pci_free are very thin wrappers around the core dma
facilities, and we have no special reason within the drm layer to behave
differently. In particular, since

commit de09d31dd38a50fdce106c15abd68432eebbd014
Author: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Date:   Fri Jan 15 16:51:42 2016 -0800

    page-flags: define PG_reserved behavior on compound pages

    As far as I can see there's no users of PG_reserved on compound pages.
    Let's use PF_NO_COMPOUND here.

it has been illegal to combine GFP_COMP with SetPageReserved, so lets
stop doing both and leave the dma layer to its own devices.

Reported-by: Taketo Kabe
Bug: https://gitlab.freedesktop.org/drm/intel/issues/1027
Fixes: de09d31dd38a ("page-flags: define PG_reserved behavior on compound pages")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: <stable@vger.kernel.org> # v4.5+
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200202171635.4039044-1-chris@chris-wilson.co.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodrm/etnaviv: rework perfmon query infrastructure
Christian Gmeiner [Fri, 28 Feb 2020 10:37:49 +0000 (11:37 +0100)]
drm/etnaviv: rework perfmon query infrastructure

commit ed1dd899baa32d47d9a93d98336472da50564346 upstream.

Report the correct perfmon domains and signals depending
on the supported feature flags.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: 9e2c2e273012 ("drm/etnaviv: add infrastructure to query perf counter")
Cc: stable@vger.kernel.org
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodrm/i915/gem: Flush all the reloc_gpu batch
Chris Wilson [Mon, 6 Apr 2020 11:48:21 +0000 (12:48 +0100)]
drm/i915/gem: Flush all the reloc_gpu batch

commit 1aaea8476d9f014667d2cb24819f9bcaf3ebb7a4 upstream.

__i915_gem_object_flush_map() takes a byte range, so feed it the written
bytes and do not mistake the u32 index as bytes!

Fixes: a679f58d0510 ("drm/i915: Flush pages on acquisition")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Cc: <stable@vger.kernel.org> # v5.2+
Reviewed-by: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200406114821.10949-1-chris@chris-wilson.co.uk
(cherry picked from commit 30c88a47f1abd5744908d3681f54dcf823fe2a12)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agovfio: platform: Switch to platform_get_irq_optional()
Eric Auger [Tue, 24 Mar 2020 15:26:30 +0000 (09:26 -0600)]
vfio: platform: Switch to platform_get_irq_optional()

commit 723fe298ad85ad1278bd2312469ad14738953cc6 upstream.

Since commit 7723f4c5ecdb ("driver core: platform: Add an error
message to platform_get_irq*()"), platform_get_irq() calls dev_err()
on an error. As we enumerate all interrupts until platform_get_irq()
fails, we now systematically get a message such as:
"vfio-platform fff51000.ethernet: IRQ index 3 not found" which is
a false positive.

Let's use platform_get_irq_optional() instead.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
Cc: stable@vger.kernel.org # v5.3+
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Tested-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoselftests/powerpc: Add tlbie_test in .gitignore
Christophe Leroy [Fri, 28 Feb 2020 00:00:09 +0000 (00:00 +0000)]
selftests/powerpc: Add tlbie_test in .gitignore

commit 47bf235f324c696395c30541fe4fcf99fcd24188 upstream.

The commit identified below added tlbie_test but forgot to add it in
.gitignore.

Fixes: 93cad5f78995 ("selftests/powerpc: Add test case for tlbie vs mtpidr ordering issue")
Cc: stable@vger.kernel.org # v5.4+
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/259f9c06ed4563c4fa4fa8ffa652347278d769e7.1582847784.git.christophe.leroy@c-s.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoselftests/vm: fix map_hugetlb length used for testing read and write
Christophe Leroy [Thu, 2 Apr 2020 04:11:51 +0000 (21:11 -0700)]
selftests/vm: fix map_hugetlb length used for testing read and write

commit cabc30da10e677c67ab9a136b1478175734715c5 upstream.

Commit fa7b9a805c79 ("tools/selftest/vm: allow choosing mem size and page
size in map_hugetlb") added the possibility to change the size of memory
mapped for the test, but left the read and write test using the default
value.  This is unnoticed when mapping a length greater than the default
one, but segfaults otherwise.

Fix read_bytes() and write_bytes() by giving them the real length.

Also fix the call to munmap().

Fixes: fa7b9a805c79 ("tools/selftest/vm: allow choosing mem size and page size in map_hugetlb")
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Leonardo Bras <leonardo@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Shuah Khan <shuah@kernel.org>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/9a404a13c871c4bd0ba9ede68f69a1225180dd7e.1580978385.git.christophe.leroy@c-s.fr
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoselftests: vm: drop dependencies on page flags from mlock2 tests
Michal Hocko [Thu, 2 Apr 2020 04:10:25 +0000 (21:10 -0700)]
selftests: vm: drop dependencies on page flags from mlock2 tests

commit eea274d64e6ea8aff2224d33d0851133a84cc7b5 upstream.

It was noticed that mlock2 tests are failing after 9c4e6b1a7027f ("mm,
mlock, vmscan: no more skipping pagevecs") because the patch has changed
the timing on when the page is added to the unevictable LRU list and thus
gains the unevictable page flag.

The test was just too dependent on the implementation details which were
true at the time when it was introduced.  Page flags and the timing when
they are set is something no userspace should ever depend on.  The test
should be testing only for the user observable contract of the tested
syscalls.  Those are defined pretty well for the mlock and there are other
means for testing them.  In fact this is already done and testing for page
flags can be safely dropped to achieve the aimed purpose.  Present bits
can be checked by /proc/<pid>/smaps RSS field and the locking state by
VmFlags although I would argue that Locked: field would be more
appropriate.

Drop all the page flag machinery and considerably simplify the test.  This
should be more robust for future kernel changes while checking the
promised contract is still valid.

Fixes: 9c4e6b1a7027f ("mm, mlock, vmscan: no more skipping pagevecs")
Reported-by: Rafael Aquini <aquini@redhat.com>
Signed-off-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Rafael Aquini <aquini@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Eric B Munson <emunson@akamai.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/20200324154218.GS19542@dhcp22.suse.cz
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoarm64: armv8_deprecated: Fix undef_hook mask for thumb setend
Fredrik Strupe [Wed, 8 Apr 2020 11:29:41 +0000 (13:29 +0200)]
arm64: armv8_deprecated: Fix undef_hook mask for thumb setend

commit fc2266011accd5aeb8ebc335c381991f20e26e33 upstream.

For thumb instructions, call_undef_hook() in traps.c first reads a u16,
and if the u16 indicates a T32 instruction (u16 >= 0xe800), a second
u16 is read, which then makes up the the lower half-word of a T32
instruction. For T16 instructions, the second u16 is not read,
which makes the resulting u32 opcode always have the upper half set to
0.

However, having the upper half of instr_mask in the undef_hook set to 0
masks out the upper half of all thumb instructions - both T16 and T32.
This results in trapped T32 instructions with the lower half-word equal
to the T16 encoding of setend (b650) being matched, even though the upper
half-word is not 0000 and thus indicates a T32 opcode.

An example of such a T32 instruction is eaa0b650, which should raise a
SIGILL since T32 instructions with an eaa prefix are unallocated as per
Arm ARM, but instead works as a SETEND because the second half-word is set
to b650.

This patch fixes the issue by extending instr_mask to include the
upper u32 half, which will still match T16 instructions where the upper
half is 0, but not T32 instructions.

Fixes: 2d888f48e056 ("arm64: Emulate SETEND for AArch32 tasks")
Cc: <stable@vger.kernel.org> # 4.0.x-
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Fredrik Strupe <fredrik@strupe.net>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoarm64: dts: ti: k3-am65: Add clocks to dwc3 nodes
Dave Gerlach [Wed, 11 Mar 2020 14:41:11 +0000 (16:41 +0200)]
arm64: dts: ti: k3-am65: Add clocks to dwc3 nodes

commit a81e5442d796ccfa2cc97d205a5477053264d978 upstream.

The TI sci-clk driver can scan the DT for all clocks provided by system
firmware and does this by checking the clocks property of all nodes, so
we must add this to the dwc3 nodes so USB clocks are available.

Without this USB does not work with latest system firmware i.e.
[    1.714662] clk: couldn't get parent clock 0 for /interconnect@100000/dwc3@4020000

Fixes: cc54a99464ccd ("arm64: dts: ti: k3-am6: add USB suppor")
Signed-off-by: Dave Gerlach <d-gerlach@ti.com>
Signed-off-by: Roger Quadros <rogerq@ti.com>
Cc: stable@kernel.org
Signed-off-by: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoARM: dts: exynos: Fix polarity of the LCD SPI bus on UniversalC210 board
Marek Szyprowski [Wed, 4 Mar 2020 14:37:26 +0000 (15:37 +0100)]
ARM: dts: exynos: Fix polarity of the LCD SPI bus on UniversalC210 board

commit 32a1671ff8e84f0dfff3a50d4b2091d25e91f5e2 upstream.

Recent changes in the SPI core and the SPI-GPIO driver revealed that the
GPIO lines for the LD9040 LCD controller on the UniversalC210 board are
defined incorrectly. Fix the polarity for those lines to match the old
behavior and hardware requirements to fix LCD panel operation with
recent kernels.

Cc: <stable@vger.kernel.org> # 5.0.x
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Andrzej Hajda <a.hajda@samsung.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoscsi: lpfc: Fix lpfc_io_buf resource leak in lpfc_get_scsi_buf_s4 error path
James Smart [Tue, 28 Jan 2020 00:23:02 +0000 (16:23 -0800)]
scsi: lpfc: Fix lpfc_io_buf resource leak in lpfc_get_scsi_buf_s4 error path

commit 0ab384a49c548baf132ccef249f78d9c6c506380 upstream.

If a call to lpfc_get_cmd_rsp_buf_per_hdwq returns NULL (memory allocation
failure), a previously allocated lpfc_io_buf resource is leaked.

Fix by releasing the lpfc_io_buf resource in the failure path.

Fixes: d79c9e9d4b3d ("scsi: lpfc: Support dynamic unbounded SGL lists on G7 hardware.")
Cc: <stable@vger.kernel.org> # v5.4+
Link: https://lore.kernel.org/r/20200128002312.16346-3-jsmart2021@gmail.com
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoscsi: ufs: fix Auto-Hibern8 error detection
Stanley Chu [Wed, 29 Jan 2020 10:52:50 +0000 (18:52 +0800)]
scsi: ufs: fix Auto-Hibern8 error detection

commit 5a244e0ea67b293abb1d26c825db2ddde5f2862f upstream.

Auto-Hibern8 may be disabled by some vendors or sysfs in runtime even if
Auto-Hibern8 capability is supported by host. If Auto-Hibern8 capability is
supported by host but not actually enabled, Auto-Hibern8 error shall not
happen.

To fix this, provide a way to detect if Auto-Hibern8 is actually enabled
first, and bypass Auto-Hibern8 disabling case in
ufshcd_is_auto_hibern8_error().

Fixes: 821744403913 ("scsi: ufs: Add error-handling of Auto-Hibernate")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200129105251.12466-4-stanley.chu@mediatek.com
Reviewed-by: Bean Huo <beanhuo@micron.com>
Reviewed-by: Alim Akhtar <alim.akhtar@samsung.com>
Reviewed-by: Asutosh Das <asutoshd@codeaurora.org>
Reviewed-by: Can Guo <cang@codeaurora.org>
Signed-off-by: Stanley Chu <stanley.chu@mediatek.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoscsi: zfcp: fix missing erp_lock in port recovery trigger for point-to-point
Steffen Maier [Thu, 12 Mar 2020 17:44:56 +0000 (18:44 +0100)]
scsi: zfcp: fix missing erp_lock in port recovery trigger for point-to-point

commit 819732be9fea728623e1ed84eba28def7384ad1f upstream.

v2.6.27 commit cc8c282963bd ("[SCSI] zfcp: Automatically attach remote
ports") introduced zfcp automatic port scan.

Before that, the user had to use the sysfs attribute "port_add" of an FCP
device (adapter) to add and open remote (target) ports, even for the remote
peer port in point-to-point topology. That code path did a proper port open
recovery trigger taking the erp_lock.

Since above commit, a new helper function zfcp_erp_open_ptp_port()
performed an UNlocked port open recovery trigger. This can race with other
parallel recovery triggers. In zfcp_erp_action_enqueue() this could corrupt
e.g. adapter->erp_total_count or adapter->erp_ready_head.

As already found for fabric topology in v4.17 commit fa89adba1941 ("scsi:
zfcp: fix infinite iteration on ERP ready list"), there was an endless loop
during tracing of rport (un)block.  A subsequent v4.18 commit 9e156c54ace3
("scsi: zfcp: assert that the ERP lock is held when tracing a recovery
trigger") introduced a lockdep assertion for that case.

As a side effect, that lockdep assertion now uncovered the unlocked code
path for PtP. It is from within an adapter ERP action:

zfcp_erp_strategy[1479]  intentionally DROPs erp lock around
                         zfcp_erp_strategy_do_action()
zfcp_erp_strategy_do_action[1441]      NO erp lock
zfcp_erp_adapter_strategy[876]         NO erp lock
zfcp_erp_adapter_strategy_open[855]    NO erp lock
zfcp_erp_adapter_strategy_open_fsf[806]NO erp lock
zfcp_erp_adapter_strat_fsf_xconf[772]  erp lock only around
                                       zfcp_erp_action_to_running(),
                                       BUT *_not_* around
                                       zfcp_erp_enqueue_ptp_port()
zfcp_erp_enqueue_ptp_port[728]         BUG: *_not_* taking erp lock
_zfcp_erp_port_reopen[432]             assumes to be called with erp lock
zfcp_erp_action_enqueue[314]           assumes to be called with erp lock
zfcp_dbf_rec_trig[288]                 _checks_ to be called with erp lock:
lockdep_assert_held(&adapter->erp_lock);

It causes the following lockdep warning:

WARNING: CPU: 2 PID: 775 at drivers/s390/scsi/zfcp_dbf.c:288
                            zfcp_dbf_rec_trig+0x16a/0x188
no locks held by zfcperp0.0.17c0/775.

Fix this by using the proper locked recovery trigger helper function.

Link: https://lore.kernel.org/r/20200312174505.51294-2-maier@linux.ibm.com
Fixes: cc8c282963bd ("[SCSI] zfcp: Automatically attach remote ports")
Cc: <stable@vger.kernel.org> #v2.6.27+
Reviewed-by: Jens Remus <jremus@linux.ibm.com>
Reviewed-by: Benjamin Block <bblock@linux.ibm.com>
Signed-off-by: Steffen Maier <maier@linux.ibm.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: ccree - dec auth tag size from cryptlen map
Gilad Ben-Yossef [Sun, 2 Feb 2020 16:19:14 +0000 (18:19 +0200)]
crypto: ccree - dec auth tag size from cryptlen map

commit 8962c6d2c2b8ca51b0f188109015b15fc5f4da44 upstream.

Remove the auth tag size from cryptlen before mapping the destination
in out-of-place AEAD decryption thus resolving a crash with
extended testmgr tests.

Signed-off-by: Gilad Ben-Yossef <gilad@benyossef.com>
Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: stable@vger.kernel.org # v4.19+
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: ccree - only try to map auth tag if needed
Gilad Ben-Yossef [Wed, 29 Jan 2020 14:37:55 +0000 (16:37 +0200)]
crypto: ccree - only try to map auth tag if needed

commit 504e84abec7a635b861afd8d7f92ecd13eaa2b09 upstream.

Make sure to only add the size of the auth tag to the source mapping
for encryption if it is an in-place operation. Failing to do this
previously caused us to try and map auth size len bytes from a NULL
mapping and crashing if both the cryptlen and assoclen are zero.

Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Gilad Ben-Yossef <gilad@benyossef.com>
Cc: stable@vger.kernel.org # v4.19+
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: ccree - protect against empty or NULL scatterlists
Gilad Ben-Yossef [Wed, 29 Jan 2020 14:37:54 +0000 (16:37 +0200)]
crypto: ccree - protect against empty or NULL scatterlists

commit ce0fc6db38decf0d2919bfe783de6d6b76e421a9 upstream.

Deal gracefully with a NULL or empty scatterlist which can happen
if both cryptlen and assoclen are zero and we're doing in-place
AEAD encryption.

This fixes a crash when this causes us to try and map a NULL page,
at least with some platforms / DMA mapping configs.

Cc: stable@vger.kernel.org # v4.19+
Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Gilad Ben-Yossef <gilad@benyossef.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: caam - update xts sector size for large input length
Andrei Botila [Fri, 28 Feb 2020 10:46:48 +0000 (12:46 +0200)]
crypto: caam - update xts sector size for large input length

commit 3f142b6a7b573bde6cff926f246da05652c61eb4 upstream.

Since in the software implementation of XTS-AES there is
no notion of sector every input length is processed the same way.
CAAM implementation has the notion of sector which causes different
results between the software implementation and the one in CAAM
for input lengths bigger than 512 bytes.
Increase sector size to maximum value on 16 bits.

Fixes: c6415a6016bf ("crypto: caam - add support for acipher xts(aes)")
Cc: <stable@vger.kernel.org> # v4.12+
Signed-off-by: Andrei Botila <andrei.botila@nxp.com>
Reviewed-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: caam/qi2 - fix chacha20 data size error
Horia Geantă [Fri, 28 Feb 2020 06:51:23 +0000 (08:51 +0200)]
crypto: caam/qi2 - fix chacha20 data size error

commit 3a5a9e1ef37b030b836d92df8264f840988f4a38 upstream.

HW generates a Data Size error for chacha20 requests that are not
a multiple of 64B, since algorithm state (AS) does not have
the FINAL bit set.

Since updating req->iv (for chaining) is not required,
modify skcipher descriptors to set the FINAL bit for chacha20.

[Note that for skcipher decryption we know that ctx1_iv_off is 0,
which allows for an optimization by not checking algorithm type,
since append_dec_op1() sets FINAL bit for all algorithms except AES.]

Also drop the descriptor operations that save the IV.
However, in order to keep code logic simple, things like
S/G tables generation etc. are not touched.

Cc: <stable@vger.kernel.org> # v5.3+
Fixes: 334d37c9e263 ("crypto: caam - update IV using HW support")
Signed-off-by: Horia Geantă <horia.geanta@nxp.com>
Tested-by: Valentin Ciocoi Radulescu <valentin.ciocoi@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoxarray: Fix early termination of xas_for_each_marked
Matthew Wilcox (Oracle) [Thu, 12 Mar 2020 21:29:11 +0000 (17:29 -0400)]
xarray: Fix early termination of xas_for_each_marked

commit 7e934cf5ace1dceeb804f7493fa28bb697ed3c52 upstream.

xas_for_each_marked() is using entry == NULL as a termination condition
of the iteration. When xas_for_each_marked() is used protected only by
RCU, this can however race with xas_store(xas, NULL) in the following
way:

TASK1                                   TASK2
page_cache_delete()                  find_get_pages_range_tag()
                                          xas_for_each_marked()
                                            xas_find_marked()
                                              off = xas_find_chunk()

  xas_store(&xas, NULL)
    xas_init_marks(&xas);
    ...
    rcu_assign_pointer(*slot, NULL);
                                              entry = xa_entry(off);

And thus xas_for_each_marked() terminates prematurely possibly leading
to missed entries in the iteration (translating to missing writeback of
some pages or a similar problem).

If we find a NULL entry that has been marked, skip it (unless we're trying
to allocate an entry).

Reported-by: Jan Kara <jack@suse.cz>
CC: stable@vger.kernel.org
Fixes: ef8e5717db01 ("page cache: Convert delete_batch to XArray")
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoXArray: Fix xas_pause for large multi-index entries
Matthew Wilcox (Oracle) [Fri, 31 Jan 2020 11:17:09 +0000 (06:17 -0500)]
XArray: Fix xas_pause for large multi-index entries

commit c36d451ad386b34f452fc3c8621ff14b9eaa31a6 upstream.

Inspired by the recent Coverity report, I looked for other places where
the offset wasn't being converted to an unsigned long before being
shifted, and I found one in xas_pause() when the entry being paused is
of order >32.

Fixes: b803b42823d0 ("xarray: Add XArray iterators")
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodm clone metadata: Fix return type of dm_clone_nr_of_hydrated_regions()
Nikos Tsironis [Fri, 27 Mar 2020 14:01:11 +0000 (16:01 +0200)]
dm clone metadata: Fix return type of dm_clone_nr_of_hydrated_regions()

commit 81d5553d1288c2ec0390f02f84d71ca0f0f9f137 upstream.

dm_clone_nr_of_hydrated_regions() returns the number of regions that
have been hydrated so far. In order to do so it employs bitmap_weight().

Until now, the return type of dm_clone_nr_of_hydrated_regions() was
unsigned long.

Because bitmap_weight() returns an int, in case BITS_PER_LONG == 64 and
the return value of bitmap_weight() is 2^31 (the maximum allowed number
of regions for a device), the result is sign extended from 32 bits to 64
bits and an incorrect value is displayed, in the status output of
dm-clone, as the number of hydrated regions.

Fix this by having dm_clone_nr_of_hydrated_regions() return an unsigned
int.

Fixes: 7431b7835f55 ("dm: add clone target")
Cc: stable@vger.kernel.org # v5.4+
Signed-off-by: Nikos Tsironis <ntsironis@arrikto.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodm clone: Add overflow check for number of regions
Nikos Tsironis [Fri, 27 Mar 2020 14:01:09 +0000 (16:01 +0200)]
dm clone: Add overflow check for number of regions

commit cd481c12269b4d276f1a52eda0ebd419079bfe3a upstream.

Add overflow check for clone->nr_regions variable, which holds the
number of regions of the target.

The overflow can occur with sufficiently large devices, if BITS_PER_LONG
== 32. E.g., if the region size is 8 sectors (4K), the overflow would
occur for device sizes > 34359738360 sectors (~16TB).

This could result in multiple device sectors wrongly mapping to the same
region number, due to the truncation from 64 bits to 32 bits, which
would lead to data corruption.

Fixes: 7431b7835f55 ("dm: add clone target")
Cc: stable@vger.kernel.org # v5.4+
Signed-off-by: Nikos Tsironis <ntsironis@arrikto.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodm verity fec: fix memory leak in verity_fec_dtr
Shetty, Harshini X (EXT-Sony Mobile) [Tue, 17 Mar 2020 09:15:45 +0000 (09:15 +0000)]
dm verity fec: fix memory leak in verity_fec_dtr

commit 75fa601934fda23d2f15bf44b09c2401942d8e15 upstream.

Fix below kmemleak detected in verity_fec_ctr. output_pool is
allocated for each dm-verity-fec device. But it is not freed when
dm-table for the verity target is removed. Hence free the output
mempool in destructor function verity_fec_dtr.

unreferenced object 0xffffffffa574d000 (size 4096):
  comm "init", pid 1667, jiffies 4294894890 (age 307.168s)
  hex dump (first 32 bytes):
    8e 36 00 98 66 a8 0b 9b 00 00 00 00 00 00 00 00  .6..f...........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<0000000060e82407>] __kmalloc+0x2b4/0x340
    [<00000000dd99488f>] mempool_kmalloc+0x18/0x20
    [<000000002560172b>] mempool_init_node+0x98/0x118
    [<000000006c3574d2>] mempool_init+0x14/0x20
    [<0000000008cb266e>] verity_fec_ctr+0x388/0x3b0
    [<000000000887261b>] verity_ctr+0x87c/0x8d0
    [<000000002b1e1c62>] dm_table_add_target+0x174/0x348
    [<000000002ad89eda>] table_load+0xe4/0x328
    [<000000001f06f5e9>] dm_ctl_ioctl+0x3b4/0x5a0
    [<00000000bee5fbb7>] do_vfs_ioctl+0x5dc/0x928
    [<00000000b475b8f5>] __arm64_sys_ioctl+0x70/0x98
    [<000000005361e2e8>] el0_svc_common+0xa0/0x158
    [<000000001374818f>] el0_svc_handler+0x6c/0x88
    [<000000003364e9f4>] el0_svc+0x8/0xc
    [<000000009d84cec9>] 0xffffffffffffffff

Fixes: a739ff3f543af ("dm verity: add support for forward error correction")
Depends-on: 6f1c819c219f7 ("dm: convert to bioset_init()/mempool_init()")
Cc: stable@vger.kernel.org
Signed-off-by: Harshini Shetty <harshini.x.shetty@sony.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodm integrity: fix a crash with unusually large tag size
Mikulas Patocka [Sun, 22 Mar 2020 19:42:21 +0000 (20:42 +0100)]
dm integrity: fix a crash with unusually large tag size

commit b93b6643e9b5a7f260b931e97f56ffa3fa65e26d upstream.

If the user specifies tag size larger than HASH_MAX_DIGESTSIZE,
there's a crash in integrity_metadata().

Cc: stable@vger.kernel.org
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agodm writecache: add cond_resched to avoid CPU hangs
Mikulas Patocka [Fri, 27 Mar 2020 11:22:36 +0000 (07:22 -0400)]
dm writecache: add cond_resched to avoid CPU hangs

commit 1edaa447d958bec24c6a79685a5790d98976fd16 upstream.

Initializing a dm-writecache device can take a long time when the
persistent memory device is large.  Add cond_resched() to a few loops
to avoid warnings that the CPU is stuck.

Cc: stable@vger.kernel.org # v4.18+
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agomm, memcg: do not high throttle allocators based on wraparound
Jakub Kicinski [Fri, 10 Apr 2020 21:32:19 +0000 (14:32 -0700)]
mm, memcg: do not high throttle allocators based on wraparound

commit 9b8b17541f13809d06f6f873325305ddbb760e3e upstream.

If a cgroup violates its memory.high constraints, we may end up unduly
penalising it.  For example, for the following hierarchy:

  A:   max high, 20 usage
  A/B: 9 high, 10 usage
  A/C: max high, 10 usage

We would end up doing the following calculation below when calculating
high delay for A/B:

  A/B: 10 - 9 = 1...
  A:   20 - PAGE_COUNTER_MAX = 21, so set max_overage to 21.

This gets worse with higher disparities in usage in the parent.

I have no idea how this disappeared from the final version of the patch,
but it is certainly Not Good(tm).  This wasn't obvious in testing because,
for a simple cgroup hierarchy with only one child, the result is usually
roughly the same.  It's only in more complex hierarchies that things go
really awry (although still, the effects are limited to a maximum of 2
seconds in schedule_timeout_killable at a maximum).

[chris@chrisdown.name: changelog]
Fixes: e26733e0d0ec ("mm, memcg: throttle allocators based on ancestral memory.high")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Chris Down <chris@chrisdown.name>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: <stable@vger.kernel.org> [5.4.x]
Link: http://lkml.kernel.org/r/20200331152424.GA1019937@chrisdown.name
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoarm64: dts: allwinner: h5: Fix PMU compatible
Maxime Ripard [Mon, 10 Feb 2020 09:55:59 +0000 (10:55 +0100)]
arm64: dts: allwinner: h5: Fix PMU compatible

commit 4ae7a3c3d7d31260f690d8d658f0365f3eca67a2 upstream.

The commit c35a516a4618 ("arm64: dts: allwinner: H5: Add PMU node")
introduced support for the PMU found on the Allwinner H5. However, the
binding only allows for a single compatible, while the patch was adding
two.

Make sure we follow the binding.

Fixes: c35a516a4618 ("arm64: dts: allwinner: H5: Add PMU node")
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agosched/core: Remove duplicate assignment in sched_tick_remote()
Scott Wood [Tue, 4 Feb 2020 00:35:58 +0000 (19:35 -0500)]
sched/core: Remove duplicate assignment in sched_tick_remote()

commit 82e0516ce3a147365a5dd2a9bedd5ba43a18663d upstream.

A redundant "curr = rq->curr" was added; remove it.

Fixes: ebc0f83c78a2 ("timers/nohz: Update NOHZ load in remote tick")
Signed-off-by: Scott Wood <swood@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/1580776558-12882-1-git-send-email-swood@redhat.com
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoarm64: dts: allwinner: h6: Fix PMU compatible
Maxime Ripard [Mon, 10 Feb 2020 09:56:00 +0000 (10:56 +0100)]
arm64: dts: allwinner: h6: Fix PMU compatible

commit 4c7eeb9af3e41ae7d840977119c58f3bbb3f4f59 upstream.

The commit 7aa9b9eb7d6a ("arm64: dts: allwinner: H6: Add PMU mode")
introduced support for the PMU found on the Allwinner H6. However, the
binding only allows for a single compatible, while the patch was adding
two.

Make sure we follow the binding.

Fixes: 7aa9b9eb7d6a ("arm64: dts: allwinner: H6: Add PMU mode")
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agonet: qualcomm: rmnet: Allow configuration updates to existing devices
Subash Abhinov Kasiviswanathan [Wed, 1 Apr 2020 21:23:55 +0000 (15:23 -0600)]
net: qualcomm: rmnet: Allow configuration updates to existing devices

commit 2abb5792387eb188b12051337d5dcd2cba615cb0 upstream.

This allows the changelink operation to succeed if the mux_id was
specified as an argument. Note that the mux_id must match the
existing mux_id of the rmnet device or should be an unused mux_id.

Fixes: 1dc49e9d164c ("net: rmnet: do not allow to change mux id if mux id is duplicated")
Reported-and-tested-by: Alex Elder <elder@linaro.org>
Signed-off-by: Sean Tranchetti <stranche@codeaurora.org>
Signed-off-by: Subash Abhinov Kasiviswanathan <subashab@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agotools: gpio: Fix out-of-tree build regression
Anssi Hannula [Wed, 25 Mar 2020 10:31:54 +0000 (12:31 +0200)]
tools: gpio: Fix out-of-tree build regression

commit 82f04bfe2aff428b063eefd234679b2d693228ed upstream.

Commit 0161a94e2d1c7 ("tools: gpio: Correctly add make dependencies for
gpio_utils") added a make rule for gpio-utils-in.o but used $(output)
instead of the correct $(OUTPUT) for the output directory, breaking
out-of-tree build (O=xx) with the following error:

  No rule to make target 'out/tools/gpio/gpio-utils-in.o', needed by 'out/tools/gpio/lsgpio-in.o'.  Stop.

Fix that.

Fixes: 0161a94e2d1c ("tools: gpio: Correctly add make dependencies for gpio_utils")
Cc: <stable@vger.kernel.org>
Cc: Laura Abbott <labbott@redhat.com>
Signed-off-by: Anssi Hannula <anssi.hannula@bitwise.fi>
Link: https://lore.kernel.org/r/20200325103154.32235-1-anssi.hannula@bitwise.fi
Reviewed-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agopowerpc/pseries: Drop pointless static qualifier in vpa_debugfs_init()
YueHaibing [Mon, 18 Feb 2019 12:56:44 +0000 (12:56 +0000)]
powerpc/pseries: Drop pointless static qualifier in vpa_debugfs_init()

commit 11dd34f3eae5a468013bb161a1dcf1fecd2ca321 upstream.

There is no need to have the 'struct dentry *vpa_dir' variable static
since new value always be assigned before use it.

Fixes: c6c26fb55e8e ("powerpc/pseries: Export raw per-CPU VPA data via debugfs")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190218125644.87448-1-yuehaibing@huawei.com
Cc: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agommc: sdhci-of-esdhc: fix esdhc_reset() for different controller versions
Yangbo Lu [Wed, 8 Jan 2020 04:07:12 +0000 (12:07 +0800)]
mmc: sdhci-of-esdhc: fix esdhc_reset() for different controller versions

commit 2aa3d826adb578b26629a79b775a552cfe3fedf7 upstream.

This patch is to fix operating in esdhc_reset() for different
controller versions, and to add bus-width restoring after data
reset for eSDHC (verdor version <= 2.2).

Also add annotation for understanding.

Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20200108040713.38888-1-yangbo.lu@nxp.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoio_uring: honor original task RLIMIT_FSIZE
Jens Axboe [Fri, 20 Mar 2020 17:23:41 +0000 (11:23 -0600)]
io_uring: honor original task RLIMIT_FSIZE

commit 4ed734b0d0913e566a9d871e15d24eb240f269f7 upstream.

With the previous fixes for number of files open checking, I added some
debug code to see if we had other spots where we're checking rlimit()
against the async io-wq workers. The only one I found was file size
checking, which we should also honor.

During write and fallocate prep, store the max file size and override
that for the current ask if we're in io-wq worker context.

Cc: stable@vger.kernel.org # 5.1+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoerofs: correct the remaining shrink objects
Gao Xiang [Wed, 26 Feb 2020 08:10:06 +0000 (16:10 +0800)]
erofs: correct the remaining shrink objects

commit 9d5a09c6f3b5fb85af20e3a34827b5d27d152b34 upstream.

The remaining count should not include successful
shrink attempts.

Fixes: e7e9a307be9d ("staging: erofs: introduce workstation for decompression")
Cc: <stable@vger.kernel.org> # 4.19+
Link: https://lore.kernel.org/r/20200226081008.86348-1-gaoxiang25@huawei.com
Reviewed-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Gao Xiang <gaoxiang25@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: mxs-dcp - fix scatterlist linearization for hash
Rosioru Dragos [Tue, 25 Feb 2020 15:05:52 +0000 (17:05 +0200)]
crypto: mxs-dcp - fix scatterlist linearization for hash

commit fa03481b6e2e82355c46644147b614f18c7a8161 upstream.

The incorrect traversal of the scatterlist, during the linearization phase
lead to computing the hash value of the wrong input buffer.
New implementation uses scatterwalk_map_and_copy()
to address this issue.

Cc: <stable@vger.kernel.org>
Fixes: 15b59e7c3733 ("crypto: mxs - Add Freescale MXS DCP driver")
Signed-off-by: Rosioru Dragos <dragos.rosioru@nxp.com>
Reviewed-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocrypto: rng - Fix a refcounting bug in crypto_rng_reset()
Dan Carpenter [Mon, 20 Jan 2020 14:38:04 +0000 (17:38 +0300)]
crypto: rng - Fix a refcounting bug in crypto_rng_reset()

commit eed74b3eba9eda36d155c11a12b2b4b50c67c1d8 upstream.

We need to decrement this refcounter on these error paths.

Fixes: f7d76e05d058 ("crypto: user - fix use_after_free of struct xxx_request")
Cc: <stable@vger.kernel.org>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoremoteproc: Fix NULL pointer dereference in rproc_virtio_notify
Nikita Shubin [Fri, 6 Mar 2020 07:24:53 +0000 (10:24 +0300)]
remoteproc: Fix NULL pointer dereference in rproc_virtio_notify

commit 791c13b709dd51eb37330f2a5837434e90c87c27 upstream.

Undefined rproc_ops .kick method in remoteproc driver will result in
"Unable to handle kernel NULL pointer dereference" in rproc_virtio_notify,
after firmware loading if:

 1) .kick method wasn't defined in driver
 2) resource_table exists in firmware and has "Virtio device entry" defined

Let's refuse to register an rproc-induced virtio device if no kick method was
defined for rproc.

[   13.180049][  T415] 8<--- cut here ---
[   13.190558][  T415] Unable to handle kernel NULL pointer dereference at virtual address 00000000
[   13.212544][  T415] pgd = (ptrval)
[   13.217052][  T415] [00000000] *pgd=00000000
[   13.224692][  T415] Internal error: Oops: 80000005 [#1] PREEMPT SMP ARM
[   13.231318][  T415] Modules linked in: rpmsg_char imx_rproc virtio_rpmsg_bus rpmsg_core [last unloaded: imx_rproc]
[   13.241687][  T415] CPU: 0 PID: 415 Comm: unload-load.sh Not tainted 5.5.2-00002-g707df13bbbdd #6
[   13.250561][  T415] Hardware name: Freescale i.MX7 Dual (Device Tree)
[   13.257009][  T415] PC is at 0x0
[   13.260249][  T415] LR is at rproc_virtio_notify+0x2c/0x54
[   13.265738][  T415] pc : [<00000000>]    lr : [<8050f6b0>]    psr: 60010113
[   13.272702][  T415] sp : b8d47c48  ip : 00000001  fp : bc04de00
[   13.278625][  T415] r10: bc04c000  r9 : 00000cc0  r8 : b8d46000
[   13.284548][  T415] r7 : 00000000  r6 : b898f200  r5 : 00000000  r4 : b8a29800
[   13.291773][  T415] r3 : 00000000  r2 : 990a3ad4  r1 : 00000000  r0 : b8a29800
[   13.299000][  T415] Flags: nZCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[   13.306833][  T415] Control: 10c5387d  Table: b8b4806a  DAC: 00000051
[   13.313278][  T415] Process unload-load.sh (pid: 415, stack limit = 0x(ptrval))
[   13.320591][  T415] Stack: (0xb8d47c48 to 0xb8d48000)
[   13.325651][  T415] 7c40:                   b895b680 00000001 b898f200 803c6430 b895bc80 7f00ae18
[   13.334531][  T415] 7c60: 00000035 00000000 00000000 b9393200 80b3ed80 00004000 b9393268 bbf5a9a2
[   13.343410][  T415] 7c80: 00000e00 00000200 00000000 7f00aff0 7f00a014 b895b680 b895b800 990a3ad4
[   13.352290][  T415] 7ca0: 00000001 b898f210 b898f200 00000000 00000000 7f00e000 00000001 00000000
[   13.361170][  T415] 7cc0: 00000000 803c62e0 80b2169c 802a0924 b898f210 00000000 00000000 b898f210
[   13.370049][  T415] 7ce0: 80b9ba44 00000000 80b9ba48 00000000 7f00e000 00000008 80b2169c 80400114
[   13.378929][  T415] 7d00: 80b2169c 8061fd64 b898f210 7f00e000 80400744 b8d46000 80b21634 80b21634
[   13.387809][  T415] 7d20: 80b2169c 80400614 80b21634 80400718 7f00e000 00000000 b8d47d7c 80400744
[   13.396689][  T415] 7d40: b8d46000 80b21634 80b21634 803fe338 b898f254 b80fe76c b8d32e38 990a3ad4
[   13.405569][  T415] 7d60: fffffff3 b898f210 b8d46000 00000001 b898f254 803ffe7c 80857a90 b898f210
[   13.414449][  T415] 7d80: 00000001 990a3ad4 b8d46000 b898f210 b898f210 80b17aec b8a29c20 803ff0a4
[   13.423328][  T415] 7da0: b898f210 00000000 b8d46000 803fb8e0 b898f200 00000000 80b17aec b898f210
[   13.432209][  T415] 7dc0: b8a29c20 990a3ad4 b895b900 b898f200 8050fb7c 80b17aec b898f210 b8a29c20
[   13.441088][  T415] 7de0: b8a29800 b895b900 b8a29a04 803c5ec0 b8a29c00 b898f200 b8a29a20 00000007
[   13.449968][  T415] 7e00: b8a29c20 8050fd78 b8a29800 00000000 b8a29a20 b8a29c04 b8a29820 b8a299d0
[   13.458848][  T415] 7e20: b895b900 8050e5a4 b8a29800 b8a299d8 b8d46000 b8a299e0 b8a29820 b8a299d0
[   13.467728][  T415] 7e40: b895b900 8050e008 000041ed 00000000 b8b8c440 b8a299d8 b8a299e0 b8a299d8
[   13.476608][  T415] 7e60: b8b8c440 990a3ad4 00000000 b8a29820 b8b8c400 00000006 b8a29800 b895b880
[   13.485487][  T415] 7e80: b8d47f78 00000000 00000000 8050f4b4 00000006 b895b890 b8b8c400 008fbea0
[   13.494367][  T415] 7ea0: b895b880 8029f530 00000000 00000000 b8d46000 00000006 b8d46000 008fbea0
[   13.503246][  T415] 7ec0: 8029f434 00000000 b8d46000 00000000 00000000 8021e2e4 0000000a 8061fd0c
[   13.512125][  T415] 7ee0: 0000000a b8af0c00 0000000a b8af0c40 00000001 b8af0c40 00000000 8061f910
[   13.521005][  T415] 7f00: 0000000a 80240af4 00000002 b8d46000 00000000 8061fd0c 00000002 80232d7c
[   13.529884][  T415] 7f20: 00000000 b8d46000 00000000 990a3ad4 00000000 00000006 b8a62d80 008fbea0
[   13.538764][  T415] 7f40: b8d47f78 00000000 b8d46000 00000000 00000000 802210c0 b88f2900 00000000
[   13.547644][  T415] 7f60: b8a62d80 b8a62d80 b8d46000 00000006 008fbea0 80221320 00000000 00000000
[   13.556524][  T415] 7f80: b8af0c00 990a3ad4 0000006c 008fbea0 76f1cda0 00000004 80101204 00000004
[   13.565403][  T415] 7fa0: 00000000 80101000 0000006c 008fbea0 00000001 008fbea0 00000006 00000000
[   13.574283][  T415] 7fc0: 0000006c 008fbea0 76f1cda0 00000004 00000006 00000006 00000000 00000000
[   13.583162][  T415] 7fe0: 00000004 7ebaf7d0 76eb4c0b 76e3f206 600d0030 00000001 00000000 00000000
[   13.592056][  T415] [<8050f6b0>] (rproc_virtio_notify) from [<803c6430>] (virtqueue_notify+0x1c/0x34)
[   13.601298][  T415] [<803c6430>] (virtqueue_notify) from [<7f00ae18>] (rpmsg_probe+0x280/0x380 [virtio_rpmsg_bus])
[   13.611663][  T415] [<7f00ae18>] (rpmsg_probe [virtio_rpmsg_bus]) from [<803c62e0>] (virtio_dev_probe+0x1f8/0x2c4)
[   13.622022][  T415] [<803c62e0>] (virtio_dev_probe) from [<80400114>] (really_probe+0x200/0x450)
[   13.630817][  T415] [<80400114>] (really_probe) from [<80400614>] (driver_probe_device+0x16c/0x1ac)
[   13.639873][  T415] [<80400614>] (driver_probe_device) from [<803fe338>] (bus_for_each_drv+0x84/0xc8)
[   13.649102][  T415] [<803fe338>] (bus_for_each_drv) from [<803ffe7c>] (__device_attach+0xd4/0x164)
[   13.658069][  T415] [<803ffe7c>] (__device_attach) from [<803ff0a4>] (bus_probe_device+0x84/0x8c)
[   13.666950][  T415] [<803ff0a4>] (bus_probe_device) from [<803fb8e0>] (device_add+0x444/0x768)
[   13.675572][  T415] [<803fb8e0>] (device_add) from [<803c5ec0>] (register_virtio_device+0xa4/0xfc)
[   13.684541][  T415] [<803c5ec0>] (register_virtio_device) from [<8050fd78>] (rproc_add_virtio_dev+0xcc/0x1b8)
[   13.694466][  T415] [<8050fd78>] (rproc_add_virtio_dev) from [<8050e5a4>] (rproc_start+0x148/0x200)
[   13.703521][  T415] [<8050e5a4>] (rproc_start) from [<8050e008>] (rproc_boot+0x384/0x5c0)
[   13.711708][  T415] [<8050e008>] (rproc_boot) from [<8050f4b4>] (state_store+0x3c/0xc8)
[   13.719723][  T415] [<8050f4b4>] (state_store) from [<8029f530>] (kernfs_fop_write+0xfc/0x214)
[   13.728348][  T415] [<8029f530>] (kernfs_fop_write) from [<8021e2e4>] (__vfs_write+0x30/0x1cc)
[   13.736971][  T415] [<8021e2e4>] (__vfs_write) from [<802210c0>] (vfs_write+0xac/0x17c)
[   13.744985][  T415] [<802210c0>] (vfs_write) from [<80221320>] (ksys_write+0x64/0xe4)
[   13.752825][  T415] [<80221320>] (ksys_write) from [<80101000>] (ret_fast_syscall+0x0/0x54)
[   13.761178][  T415] Exception stack(0xb8d47fa8 to 0xb8d47ff0)
[   13.766932][  T415] 7fa0:                   0000006c 008fbea0 00000001 008fbea0 00000006 00000000
[   13.775811][  T415] 7fc0: 0000006c 008fbea0 76f1cda0 00000004 00000006 00000006 00000000 00000000
[   13.784687][  T415] 7fe0: 00000004 7ebaf7d0 76eb4c0b 76e3f206
[   13.790442][  T415] Code: bad PC value
[   13.839214][  T415] ---[ end trace 1fe21ecfc9f28852 ]---

Reviewed-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Nikita Shubin <NShubin@topcon.com>
Fixes: 7a186941626d ("remoteproc: remove the single rpmsg vdev limitation")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200306072452.24743-1-NShubin@topcon.com
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoremoteproc: qcom_q6v5_mss: Reload the mba region on coredump
Sibi Sankar [Wed, 4 Mar 2020 19:47:29 +0000 (01:17 +0530)]
remoteproc: qcom_q6v5_mss: Reload the mba region on coredump

commit d96f2571dc84d128cacf1944f4ecc87834c779a6 upstream.

On secure devices after a wdog/fatal interrupt, the mba region has to be
refreshed in order to prevent the following errors during mba load.

Err Logs:
remoteproc remoteproc2: stopped remote processor 4080000.remoteproc
qcom-q6v5-mss 4080000.remoteproc: PBL returned unexpected status -284031232
qcom-q6v5-mss 4080000.remoteproc: PBL returned unexpected status -284031232
....
qcom-q6v5-mss 4080000.remoteproc: PBL returned unexpected status -284031232
qcom-q6v5-mss 4080000.remoteproc: MBA booted, loading mpss

Fixes: 7dd8ade24dc2a ("remoteproc: qcom: q6v5-mss: Add custom dump function for modem")
Cc: stable@vger.kernel.org
Signed-off-by: Sibi Sankar <sibis@codeaurora.org>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Link: https://lore.kernel.org/r/20200304194729.27979-4-sibis@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoremoteproc: qcom_q6v5_mss: Don't reassign mpss region on shutdown
Bjorn Andersson [Wed, 4 Mar 2020 19:47:27 +0000 (01:17 +0530)]
remoteproc: qcom_q6v5_mss: Don't reassign mpss region on shutdown

commit 900fc60df22748dbc28e4970838e8f7b8f1013ce upstream.

Trying to reclaim mpss memory while the mba is not running causes the
system to crash on devices with security fuses blown, so leave it
assigned to the remote on shutdown and recover it on a subsequent boot.

Fixes: 6c5a9dc2481b ("remoteproc: qcom: Make secure world call for mem ownership switch")
Cc: stable@vger.kernel.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Sibi Sankar <sibis@codeaurora.org>
Tested-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Link: https://lore.kernel.org/r/20200304194729.27979-2-sibis@codeaurora.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: use nofs allocations for running delayed items
Josef Bacik [Thu, 19 Mar 2020 14:11:32 +0000 (10:11 -0400)]
btrfs: use nofs allocations for running delayed items

commit 351cbf6e4410e7ece05e35d0a07320538f2418b4 upstream.

Zygo reported the following lockdep splat while testing the balance
patches

======================================================
WARNING: possible circular locking dependency detected
5.6.0-c6f0579d496a+ #53 Not tainted
------------------------------------------------------
kswapd0/1133 is trying to acquire lock:
ffff888092f622c0 (&delayed_node->mutex){+.+.}, at: __btrfs_release_delayed_node+0x7c/0x5b0

but task is already holding lock:
ffffffff8fc5f860 (fs_reclaim){+.+.}, at: __fs_reclaim_acquire+0x5/0x30

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #1 (fs_reclaim){+.+.}:
       fs_reclaim_acquire.part.91+0x29/0x30
       fs_reclaim_acquire+0x19/0x20
       kmem_cache_alloc_trace+0x32/0x740
       add_block_entry+0x45/0x260
       btrfs_ref_tree_mod+0x6e2/0x8b0
       btrfs_alloc_tree_block+0x789/0x880
       alloc_tree_block_no_bg_flush+0xc6/0xf0
       __btrfs_cow_block+0x270/0x940
       btrfs_cow_block+0x1ba/0x3a0
       btrfs_search_slot+0x999/0x1030
       btrfs_insert_empty_items+0x81/0xe0
       btrfs_insert_delayed_items+0x128/0x7d0
       __btrfs_run_delayed_items+0xf4/0x2a0
       btrfs_run_delayed_items+0x13/0x20
       btrfs_commit_transaction+0x5cc/0x1390
       insert_balance_item.isra.39+0x6b2/0x6e0
       btrfs_balance+0x72d/0x18d0
       btrfs_ioctl_balance+0x3de/0x4c0
       btrfs_ioctl+0x30ab/0x44a0
       ksys_ioctl+0xa1/0xe0
       __x64_sys_ioctl+0x43/0x50
       do_syscall_64+0x77/0x2c0
       entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&delayed_node->mutex){+.+.}:
       __lock_acquire+0x197e/0x2550
       lock_acquire+0x103/0x220
       __mutex_lock+0x13d/0xce0
       mutex_lock_nested+0x1b/0x20
       __btrfs_release_delayed_node+0x7c/0x5b0
       btrfs_remove_delayed_node+0x49/0x50
       btrfs_evict_inode+0x6fc/0x900
       evict+0x19a/0x2c0
       dispose_list+0xa0/0xe0
       prune_icache_sb+0xbd/0xf0
       super_cache_scan+0x1b5/0x250
       do_shrink_slab+0x1f6/0x530
       shrink_slab+0x32e/0x410
       shrink_node+0x2a5/0xba0
       balance_pgdat+0x4bd/0x8a0
       kswapd+0x35a/0x800
       kthread+0x1e9/0x210
       ret_from_fork+0x3a/0x50

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(fs_reclaim);
                               lock(&delayed_node->mutex);
                               lock(fs_reclaim);
  lock(&delayed_node->mutex);

 *** DEADLOCK ***

3 locks held by kswapd0/1133:
 #0: ffffffff8fc5f860 (fs_reclaim){+.+.}, at: __fs_reclaim_acquire+0x5/0x30
 #1: ffffffff8fc380d8 (shrinker_rwsem){++++}, at: shrink_slab+0x1e8/0x410
 #2: ffff8881e0e6c0e8 (&type->s_umount_key#42){++++}, at: trylock_super+0x1b/0x70

stack backtrace:
CPU: 2 PID: 1133 Comm: kswapd0 Not tainted 5.6.0-c6f0579d496a+ #53
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
Call Trace:
 dump_stack+0xc1/0x11a
 print_circular_bug.isra.38.cold.57+0x145/0x14a
 check_noncircular+0x2a9/0x2f0
 ? print_circular_bug.isra.38+0x130/0x130
 ? stack_trace_consume_entry+0x90/0x90
 ? save_trace+0x3cc/0x420
 __lock_acquire+0x197e/0x2550
 ? btrfs_inode_clear_file_extent_range+0x9b/0xb0
 ? register_lock_class+0x960/0x960
 lock_acquire+0x103/0x220
 ? __btrfs_release_delayed_node+0x7c/0x5b0
 __mutex_lock+0x13d/0xce0
 ? __btrfs_release_delayed_node+0x7c/0x5b0
 ? __asan_loadN+0xf/0x20
 ? pvclock_clocksource_read+0xeb/0x190
 ? __btrfs_release_delayed_node+0x7c/0x5b0
 ? mutex_lock_io_nested+0xc20/0xc20
 ? __kasan_check_read+0x11/0x20
 ? check_chain_key+0x1e6/0x2e0
 mutex_lock_nested+0x1b/0x20
 ? mutex_lock_nested+0x1b/0x20
 __btrfs_release_delayed_node+0x7c/0x5b0
 btrfs_remove_delayed_node+0x49/0x50
 btrfs_evict_inode+0x6fc/0x900
 ? btrfs_setattr+0x840/0x840
 ? do_raw_spin_unlock+0xa8/0x140
 evict+0x19a/0x2c0
 dispose_list+0xa0/0xe0
 prune_icache_sb+0xbd/0xf0
 ? invalidate_inodes+0x310/0x310
 super_cache_scan+0x1b5/0x250
 do_shrink_slab+0x1f6/0x530
 shrink_slab+0x32e/0x410
 ? do_shrink_slab+0x530/0x530
 ? do_shrink_slab+0x530/0x530
 ? __kasan_check_read+0x11/0x20
 ? mem_cgroup_protected+0x13d/0x260
 shrink_node+0x2a5/0xba0
 balance_pgdat+0x4bd/0x8a0
 ? mem_cgroup_shrink_node+0x490/0x490
 ? _raw_spin_unlock_irq+0x27/0x40
 ? finish_task_switch+0xce/0x390
 ? rcu_read_lock_bh_held+0xb0/0xb0
 kswapd+0x35a/0x800
 ? _raw_spin_unlock_irqrestore+0x4c/0x60
 ? balance_pgdat+0x8a0/0x8a0
 ? finish_wait+0x110/0x110
 ? __kasan_check_read+0x11/0x20
 ? __kthread_parkme+0xc6/0xe0
 ? balance_pgdat+0x8a0/0x8a0
 kthread+0x1e9/0x210
 ? kthread_create_worker_on_cpu+0xc0/0xc0
 ret_from_fork+0x3a/0x50

This is because we hold that delayed node's mutex while doing tree
operations.  Fix this by just wrapping the searches in nofs.

CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: fix missing semaphore unlock in btrfs_sync_file
Robbie Ko [Tue, 17 Mar 2020 06:31:02 +0000 (14:31 +0800)]
btrfs: fix missing semaphore unlock in btrfs_sync_file

commit 6ff06729c22ec0b7498d900d79cc88cfb8aceaeb upstream.

Ordered ops are started twice in sync file, once outside of inode mutex
and once inside, taking the dio semaphore. There was one error path
missing the semaphore unlock.

Fixes: aab15e8ec2576 ("Btrfs: fix rare chances for data loss when doing a fast fsync")
CC: stable@vger.kernel.org # 4.19+
Signed-off-by: Robbie Ko <robbieko@synology.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
[ add changelog ]
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: unset reloc control if we fail to recover
Josef Bacik [Wed, 4 Mar 2020 16:18:25 +0000 (11:18 -0500)]
btrfs: unset reloc control if we fail to recover

commit fb2d83eefef4e1c717205bac71cb1941edf8ae11 upstream.

If we fail to load an fs root, or fail to start a transaction we can
bail without unsetting the reloc control, which leads to problems later
when we free the reloc control but still have it attached to the file
system.

In the normal path we'll end up calling unset_reloc_control() twice, but
all it does is set fs_info->reloc_control = NULL, and we can only have
one balance at a time so it's not racey.

CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: fix missing file extent item for hole after ranged fsync
Filipe Manana [Mon, 9 Mar 2020 12:41:05 +0000 (12:41 +0000)]
btrfs: fix missing file extent item for hole after ranged fsync

commit 95418ed1d10774cd9a49af6f39e216c1256f1eeb upstream.

When doing a fast fsync for a range that starts at an offset greater than
zero, we can end up with a log that when replayed causes the respective
inode miss a file extent item representing a hole if we are not using the
NO_HOLES feature. This is because for fast fsyncs we don't log any extents
that cover a range different from the one requested in the fsync.

Example scenario to trigger it:

  $ mkfs.btrfs -O ^no-holes -f /dev/sdd
  $ mount /dev/sdd /mnt

  # Create a file with a single 256K and fsync it to clear to full sync
  # bit in the inode - we want the msync below to trigger a fast fsync.
  $ xfs_io -f -c "pwrite -S 0xab 0 256K" -c "fsync" /mnt/foo

  # Force a transaction commit and wipe out the log tree.
  $ sync

  # Dirty 768K of data, increasing the file size to 1Mb, and flush only
  # the range from 256K to 512K without updating the log tree
  # (sync_file_range() does not trigger fsync, it only starts writeback
  # and waits for it to finish).

  $ xfs_io -c "pwrite -S 0xcd 256K 768K" /mnt/foo
  $ xfs_io -c "sync_range -abw 256K 256K" /mnt/foo

  # Now dirty the range from 768K to 1M again and sync that range.
  $ xfs_io -c "mmap -w 768K 256K"        \
           -c "mwrite -S 0xef 768K 256K" \
           -c "msync -s 768K 256K"       \
           -c "munmap"                   \
           /mnt/foo

  <power fail>

  # Mount to replay the log.
  $ mount /dev/sdd /mnt
  $ umount /mnt

  $ btrfs check /dev/sdd
  Opening filesystem to check...
  Checking filesystem on /dev/sdd
  UUID: 482fb574-b288-478e-a190-a9c44a78fca6
  [1/7] checking root items
  [2/7] checking extents
  [3/7] checking free space cache
  [4/7] checking fs roots
  root 5 inode 257 errors 100, file extent discount
  Found file extent holes:
       start: 262144, len: 524288
  ERROR: errors found in fs roots
  found 720896 bytes used, error(s) found
  total csum bytes: 512
  total tree bytes: 131072
  total fs tree bytes: 32768
  total extent tree bytes: 16384
  btree space waste bytes: 123514
  file data blocks allocated: 589824
    referenced 589824

Fix this issue by setting the range to full (0 to LLONG_MAX) when the
NO_HOLES feature is not enabled. This results in extra work being done
but it gives the guarantee we don't end up with missing holes after
replaying the log.

CC: stable@vger.kernel.org # 4.19+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
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>
4 years agobtrfs: drop block from cache on error in relocation
Josef Bacik [Wed, 4 Mar 2020 16:18:23 +0000 (11:18 -0500)]
btrfs: drop block from cache on error in relocation

commit 8e19c9732ad1d127b5575a10f4fbcacf740500ff upstream.

If we have an error while building the backref tree in relocation we'll
process all the pending edges and then free the node.  However if we
integrated some edges into the cache we'll lose our link to those edges
by simply freeing this node, which means we'll leak memory and
references to any roots that we've found.

Instead we need to use remove_backref_node(), which walks through all of
the edges that are still linked to this node and free's them up and
drops any root references we may be holding.

CC: stable@vger.kernel.org # 4.9+
Reviewed-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: set update the uuid generation as soon as possible
Josef Bacik [Fri, 14 Feb 2020 20:22:06 +0000 (15:22 -0500)]
btrfs: set update the uuid generation as soon as possible

commit 75ec1db8717a8f0a9d9c8d033e542fdaa7b73898 upstream.

In my EIO stress testing I noticed I was getting forced to rescan the
uuid tree pretty often, which was weird.  This is because my error
injection stuff would sometimes inject an error after log replay but
before we loaded the UUID tree.  If log replay committed the transaction
it wouldn't have updated the uuid tree generation, but the tree was
valid and didn't change, so there's no reason to not update the
generation here.

Fix this by setting the BTRFS_FS_UPDATE_UUID_TREE_GEN bit immediately
after reading all the fs roots if the uuid tree generation matches the
fs generation.  Then any transaction commits that happen during mount
won't screw up our uuid tree state, forcing us to do needless uuid
rescans.

Fixes: 70f801754728 ("Btrfs: check UUID tree during mount if required")
CC: stable@vger.kernel.org # 4.19+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: reloc: clean dirty subvols if we fail to start a transaction
Josef Bacik [Wed, 4 Mar 2020 16:18:27 +0000 (11:18 -0500)]
btrfs: reloc: clean dirty subvols if we fail to start a transaction

commit 6217b0fadd4473a16fabc6aecd7527a9f71af534 upstream.

If we do merge_reloc_roots() we could insert a few roots onto the dirty
subvol roots list, where we hold a ref on them.  If we fail to start the
transaction we need to run clean_dirty_subvols() in order to cleanup the
refs.

CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoBtrfs: fix crash during unmount due to race with delayed inode workers
Filipe Manana [Fri, 28 Feb 2020 13:04:36 +0000 (13:04 +0000)]
Btrfs: fix crash during unmount due to race with delayed inode workers

commit f0cc2cd70164efe8f75c5d99560f0f69969c72e4 upstream.

During unmount we can have a job from the delayed inode items work queue
still running, that can lead to at least two bad things:

1) A crash, because the worker can try to create a transaction just
   after the fs roots were freed;

2) A transaction leak, because the worker can create a transaction
   before the fs roots are freed and just after we committed the last
   transaction and after we stopped the transaction kthread.

A stack trace example of the crash:

 [79011.691214] kernel BUG at lib/radix-tree.c:982!
 [79011.692056] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC PTI
 [79011.693180] CPU: 3 PID: 1394 Comm: kworker/u8:2 Tainted: G        W         5.6.0-rc2-btrfs-next-54 #2
 (...)
 [79011.696789] Workqueue: btrfs-delayed-meta btrfs_work_helper [btrfs]
 [79011.697904] RIP: 0010:radix_tree_tag_set+0xe7/0x170
 (...)
 [79011.702014] RSP: 0018:ffffb3c84a317ca0 EFLAGS: 00010293
 [79011.702949] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
 [79011.704202] RDX: ffffb3c84a317cb0 RSI: ffffb3c84a317ca8 RDI: ffff8db3931340a0
 [79011.705463] RBP: 0000000000000005 R08: 0000000000000005 R09: ffffffff974629d0
 [79011.706756] R10: ffffb3c84a317bc0 R11: 0000000000000001 R12: ffff8db393134000
 [79011.708010] R13: ffff8db3931340a0 R14: ffff8db393134068 R15: 0000000000000001
 [79011.709270] FS:  0000000000000000(0000) GS:ffff8db3b6a00000(0000) knlGS:0000000000000000
 [79011.710699] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 [79011.711710] CR2: 00007f22c2a0a000 CR3: 0000000232ad4005 CR4: 00000000003606e0
 [79011.712958] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 [79011.714205] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
 [79011.715448] Call Trace:
 [79011.715925]  record_root_in_trans+0x72/0xf0 [btrfs]
 [79011.716819]  btrfs_record_root_in_trans+0x4b/0x70 [btrfs]
 [79011.717925]  start_transaction+0xdd/0x5c0 [btrfs]
 [79011.718829]  btrfs_async_run_delayed_root+0x17e/0x2b0 [btrfs]
 [79011.719915]  btrfs_work_helper+0xaa/0x720 [btrfs]
 [79011.720773]  process_one_work+0x26d/0x6a0
 [79011.721497]  worker_thread+0x4f/0x3e0
 [79011.722153]  ? process_one_work+0x6a0/0x6a0
 [79011.722901]  kthread+0x103/0x140
 [79011.723481]  ? kthread_create_worker_on_cpu+0x70/0x70
 [79011.724379]  ret_from_fork+0x3a/0x50
 (...)

The following diagram shows a sequence of steps that lead to the crash
during ummount of the filesystem:

        CPU 1                                             CPU 2                                CPU 3

 btrfs_punch_hole()
   btrfs_btree_balance_dirty()
     btrfs_balance_delayed_items()
       --> sees
           fs_info->delayed_root->items
           with value 200, which is greater
           than
           BTRFS_DELAYED_BACKGROUND (128)
           and smaller than
           BTRFS_DELAYED_WRITEBACK (512)
       btrfs_wq_run_delayed_node()
         --> queues a job for
             fs_info->delayed_workers to run
             btrfs_async_run_delayed_root()

                                                                                            btrfs_async_run_delayed_root()
                                                                                              --> job queued by CPU 1

                                                                                              --> starts picking and running
                                                                                                  delayed nodes from the
                                                                                                  prepare_list list

                                                 close_ctree()

                                                   btrfs_delete_unused_bgs()

                                                   btrfs_commit_super()

                                                     btrfs_join_transaction()
                                                       --> gets transaction N

                                                     btrfs_commit_transaction(N)
                                                       --> set transaction state
                                                        to TRANTS_STATE_COMMIT_START

                                                                                             btrfs_first_prepared_delayed_node()
                                                                                               --> picks delayed node X through
                                                                                                   the prepared_list list

                                                       btrfs_run_delayed_items()

                                                         btrfs_first_delayed_node()
                                                           --> also picks delayed node X
                                                               but through the node_list
                                                               list

                                                         __btrfs_commit_inode_delayed_items()
                                                            --> runs all delayed items from
                                                                this node and drops the
                                                                node's item count to 0
                                                                through call to
                                                                btrfs_release_delayed_inode()

                                                         --> finishes running any remaining
                                                             delayed nodes

                                                       --> finishes transaction commit

                                                   --> stops cleaner and transaction threads

                                                   btrfs_free_fs_roots()
                                                     --> frees all roots and removes them
                                                         from the radix tree
                                                         fs_info->fs_roots_radix

                                                                                             btrfs_join_transaction()
                                                                                               start_transaction()
                                                                                                 btrfs_record_root_in_trans()
                                                                                                   record_root_in_trans()
                                                                                                     radix_tree_tag_set()
                                                                                                       --> crashes because
                                                                                                           the root is not in
                                                                                                           the radix tree
                                                                                                           anymore

If the worker is able to call btrfs_join_transaction() before the unmount
task frees the fs roots, we end up leaking a transaction and all its
resources, since after the call to btrfs_commit_super() and stopping the
transaction kthread, we don't expect to have any transaction open anymore.

When this situation happens the worker has a delayed node that has no
more items to run, since the task calling btrfs_run_delayed_items(),
which is doing a transaction commit, picks the same node and runs all
its items first.

We can not wait for the worker to complete when running delayed items
through btrfs_run_delayed_items(), because we call that function in
several phases of a transaction commit, and that could cause a deadlock
because the worker calls btrfs_join_transaction() and the task doing the
transaction commit may have already set the transaction state to
TRANS_STATE_COMMIT_DOING.

Also it's not possible to get into a situation where only some of the
items of a delayed node are added to the fs/subvolume tree in the current
transaction and the remaining ones in the next transaction, because when
running the items of a delayed inode we lock its mutex, effectively
waiting for the worker if the worker is running the items of the delayed
node already.

Since this can only cause issues when unmounting a filesystem, fix it in
a simple way by waiting for any jobs on the delayed workers queue before
calling btrfs_commit_supper() at close_ctree(). This works because at this
point no one can call btrfs_btree_balance_dirty() or
btrfs_balance_delayed_items(), and if we end up waiting for any worker to
complete, btrfs_commit_super() will commit the transaction created by the
worker.

CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Filipe Manana <fdmanana@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agobtrfs: Don't submit any btree write bio if the fs has errors
Qu Wenruo [Wed, 12 Feb 2020 06:12:44 +0000 (14:12 +0800)]
btrfs: Don't submit any btree write bio if the fs has errors

commit b3ff8f1d380e65dddd772542aa9bff6c86bf715a upstream.

[BUG]
There is a fuzzed image which could cause KASAN report at unmount time.

  BUG: KASAN: use-after-free in btrfs_queue_work+0x2c1/0x390
  Read of size 8 at addr ffff888067cf6848 by task umount/1922

  CPU: 0 PID: 1922 Comm: umount Tainted: G        W         5.0.21 #1
  Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014
  Call Trace:
   dump_stack+0x5b/0x8b
   print_address_description+0x70/0x280
   kasan_report+0x13a/0x19b
   btrfs_queue_work+0x2c1/0x390
   btrfs_wq_submit_bio+0x1cd/0x240
   btree_submit_bio_hook+0x18c/0x2a0
   submit_one_bio+0x1be/0x320
   flush_write_bio.isra.41+0x2c/0x70
   btree_write_cache_pages+0x3bb/0x7f0
   do_writepages+0x5c/0x130
   __writeback_single_inode+0xa3/0x9a0
   writeback_single_inode+0x23d/0x390
   write_inode_now+0x1b5/0x280
   iput+0x2ef/0x600
   close_ctree+0x341/0x750
   generic_shutdown_super+0x126/0x370
   kill_anon_super+0x31/0x50
   btrfs_kill_super+0x36/0x2b0
   deactivate_locked_super+0x80/0xc0
   deactivate_super+0x13c/0x150
   cleanup_mnt+0x9a/0x130
   task_work_run+0x11a/0x1b0
   exit_to_usermode_loop+0x107/0x130
   do_syscall_64+0x1e5/0x280
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

[CAUSE]
The fuzzed image has a completely screwd up extent tree:

  leaf 29421568 gen 8 total ptrs 6 free space 3587 owner EXTENT_TREE
  refs 2 lock (w:0 r:0 bw:0 br:0 sw:0 sr:0) lock_owner 0 current 5938
          item 0 key (12587008 168 4096) itemoff 3942 itemsize 53
                  extent refs 1 gen 9 flags 1
                  ref#0: extent data backref root 5 objectid 259 offset 0 count 1
          item 1 key (12591104 168 8192) itemoff 3889 itemsize 53
                  extent refs 1 gen 9 flags 1
                  ref#0: extent data backref root 5 objectid 271 offset 0 count 1
          item 2 key (12599296 168 4096) itemoff 3836 itemsize 53
                  extent refs 1 gen 9 flags 1
                  ref#0: extent data backref root 5 objectid 259 offset 4096 count 1
          item 3 key (29360128 169 0) itemoff 3803 itemsize 33
                  extent refs 1 gen 9 flags 2
                  ref#0: tree block backref root 5
          item 4 key (29368320 169 1) itemoff 3770 itemsize 33
                  extent refs 1 gen 9 flags 2
                  ref#0: tree block backref root 5
          item 5 key (29372416 169 0) itemoff 3737 itemsize 33
                  extent refs 1 gen 9 flags 2
                  ref#0: tree block backref root 5

Note that leaf 29421568 doesn't have its backref in the extent tree.
Thus extent allocator can re-allocate leaf 29421568 for other trees.

In short, the bug is caused by:

- Existing tree block gets allocated to log tree
  This got its generation bumped.

- Log tree balance cleaned dirty bit of offending tree block
  It will not be written back to disk, thus no WRITTEN flag.

- Original owner of the tree block gets COWed
  Since the tree block has higher transid, no WRITTEN flag, it's reused,
  and not traced by transaction::dirty_pages.

- Transaction aborted
  Tree blocks get cleaned according to transaction::dirty_pages. But the
  offending tree block is not recorded at all.

- Filesystem unmount
  All pages are assumed to be are clean, destroying all workqueue, then
  call iput(btree_inode).
  But offending tree block is still dirty, which triggers writeback, and
  causes use-after-free bug.

The detailed sequence looks like this:

- Initial status
  eb: 29421568, header=WRITTEN bflags_dirty=0, page_dirty=0, gen=8,
      not traced by any dirty extent_iot_tree.

- New tree block is allocated
  Since there is no backref for 29421568, it's re-allocated as new tree
  block.
  Keep in mind that tree block 29421568 is still referred by extent
  tree.

- Tree block 29421568 is filled for log tree
  eb: 29421568, header=0 bflags_dirty=1, page_dirty=1, gen=9 << (gen bumped)
      traced by btrfs_root::dirty_log_pages

- Some log tree operations
  Since the fs is using node size 4096, the log tree can easily go a
  level higher.

- Log tree needs balance
  Tree block 29421568 gets all its content pushed to right, thus now
  it is empty, and we don't need it.
  btrfs_clean_tree_block() from __push_leaf_right() get called.

  eb: 29421568, header=0 bflags_dirty=0, page_dirty=0, gen=9
      traced by btrfs_root::dirty_log_pages

- Log tree write back
  btree_write_cache_pages() goes through dirty pages ranges, but since
  page of tree block 29421568 gets cleaned already, it's not written
  back to disk. Thus it doesn't have WRITTEN bit set.
  But ranges in dirty_log_pages are cleared.

  eb: 29421568, header=0 bflags_dirty=0, page_dirty=0, gen=9
      not traced by any dirty extent_iot_tree.

- Extent tree update when committing transaction
  Since tree block 29421568 has transid equal to running trans, and has
  no WRITTEN bit, should_cow_block() will use it directly without adding
  it to btrfs_transaction::dirty_pages.

  eb: 29421568, header=0 bflags_dirty=1, page_dirty=1, gen=9
      not traced by any dirty extent_iot_tree.

  At this stage, we're doomed. We have a dirty eb not tracked by any
  extent io tree.

- Transaction gets aborted due to corrupted extent tree
  Btrfs cleans up dirty pages according to transaction::dirty_pages and
  btrfs_root::dirty_log_pages.
  But since tree block 29421568 is not tracked by neither of them, it's
  still dirty.

  eb: 29421568, header=0 bflags_dirty=1, page_dirty=1, gen=9
      not traced by any dirty extent_iot_tree.

- Filesystem unmount
  Since all cleanup is assumed to be done, all workqueus are destroyed.
  Then iput(btree_inode) is called, expecting no dirty pages.
  But tree 29421568 is still dirty, thus triggering writeback.
  Since all workqueues are already freed, we cause use-after-free.

This shows us that, log tree blocks + bad extent tree can cause wild
dirty pages.

[FIX]
To fix the problem, don't submit any btree write bio if the filesytem
has any error.  This is the last safe net, just in case other cleanup
haven't caught catch it.

Link: https://github.com/bobfuzzer/CVE/tree/master/CVE-2019-19377
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Signed-off-by: Qu Wenruo <wqu@suse.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agomtd: spinand: Do not erase the block before writing a bad block marker
Frieder Schrempf [Tue, 18 Feb 2020 10:05:35 +0000 (10:05 +0000)]
mtd: spinand: Do not erase the block before writing a bad block marker

commit b645ad39d56846618704e463b24bb994c9585c7f upstream.

Currently when marking a block, we use spinand_erase_op() to erase
the block before writing the marker to the OOB area. Doing so without
waiting for the operation to finish can lead to the marking failing
silently and no bad block marker being written to the flash.

In fact we don't need to do an erase at all before writing the BBM.
The ECC is disabled for raw accesses to the OOB data and we don't
need to work around any issues with chips reporting ECC errors as it
is known to be the case for raw NAND.

Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs")
Cc: stable@vger.kernel.org
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200218100432.32433-4-frieder.schrempf@kontron.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agomtd: spinand: Stop using spinand->oobbuf for buffering bad block markers
Frieder Schrempf [Tue, 18 Feb 2020 10:05:14 +0000 (10:05 +0000)]
mtd: spinand: Stop using spinand->oobbuf for buffering bad block markers

commit 2148937501ee3d663e0010e519a553fea67ad103 upstream.

For reading and writing the bad block markers, spinand->oobbuf is
currently used as a buffer for the marker bytes. During the
underlying read and write operations to actually get/set the content
of the OOB area, the content of spinand->oobbuf is reused and changed
by accessing it through spinand->oobbuf and/or spinand->databuf.

This is a flaw in the original design of the SPI NAND core and at the
latest from 13c15e07eedf ("mtd: spinand: Handle the case where
PROGRAM LOAD does not reset the cache") on, it results in not having
the bad block marker written at all, as the spinand->oobbuf is
cleared to 0xff after setting the marker bytes to zero.

To fix it, we now just store the two bytes for the marker on the
stack and let the read/write operations copy it from/to the page
buffer later.

Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs")
Cc: stable@vger.kernel.org
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200218100432.32433-2-frieder.schrempf@kontron.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoCIFS: Fix bug which the return value by asynchronous read is error
Yilu Lin [Wed, 18 Mar 2020 03:59:19 +0000 (11:59 +0800)]
CIFS: Fix bug which the return value by asynchronous read is error

commit 97adda8b3ab703de8e4c8d27646ddd54fe22879c upstream.

This patch is used to fix the bug in collect_uncached_read_data()
that rc is automatically converted from a signed number to an
unsigned number when the CIFS asynchronous read fails.
It will cause ctx->rc is error.

Example:
Share a directory and create a file on the Windows OS.
Mount the directory to the Linux OS using CIFS.
On the CIFS client of the Linux OS, invoke the pread interface to
deliver the read request.

The size of the read length plus offset of the read request is greater
than the maximum file size.

In this case, the CIFS server on the Windows OS returns a failure
message (for example, the return value of
smb2.nt_status is STATUS_INVALID_PARAMETER).

After receiving the response message, the CIFS client parses
smb2.nt_status to STATUS_INVALID_PARAMETER
and converts it to the Linux error code (rdata->result=-22).

Then the CIFS client invokes the collect_uncached_read_data function to
assign the value of rdata->result to rc, that is, rc=rdata->result=-22.

The type of the ctx->total_len variable is unsigned integer,
the type of the rc variable is integer, and the type of
the ctx->rc variable is ssize_t.

Therefore, during the ternary operation, the value of rc is
automatically converted to an unsigned number. The final result is
ctx->rc=4294967274. However, the expected result is ctx->rc=-22.

Signed-off-by: Yilu Lin <linyilu@huawei.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
CC: Stable <stable@vger.kernel.org>
Acked-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agosmb3: fix performance regression with setting mtime
Steve French [Mon, 24 Feb 2020 23:37:39 +0000 (17:37 -0600)]
smb3: fix performance regression with setting mtime

commit cf5371ae460eb8e484e4884747af270c86c3c469 upstream.

There are cases when we don't want to send the SMB2 flush operation
(e.g. when user specifies mount parm "nostrictsync") and it can be
a very expensive operation on the server.  In most cases in order
to set mtime, we simply need to flush (write) the dirtry pages from
the client and send the writes to the server not also send a flush
protocol operation to the server.

Fixes: aa081859b10c ("cifs: flush before set-info if we have writeable handles")
CC: Stable <stable@vger.kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: VMX: fix crash cleanup when KVM wasn't used
Vitaly Kuznetsov [Wed, 1 Apr 2020 08:13:48 +0000 (10:13 +0200)]
KVM: VMX: fix crash cleanup when KVM wasn't used

commit dbef2808af6c594922fe32833b30f55f35e9da6d upstream.

If KVM wasn't used at all before we crash the cleanup procedure fails with
 BUG: unable to handle page fault for address: ffffffffffffffc8
 #PF: supervisor read access in kernel mode
 #PF: error_code(0x0000) - not-present page
 PGD 23215067 P4D 23215067 PUD 23217067 PMD 0
 Oops: 0000 [#8] SMP PTI
 CPU: 0 PID: 3542 Comm: bash Kdump: loaded Tainted: G      D           5.6.0-rc2+ #823
 RIP: 0010:crash_vmclear_local_loaded_vmcss.cold+0x19/0x51 [kvm_intel]

The root cause is that loaded_vmcss_on_cpu list is not yet initialized,
we initialize it in hardware_enable() but this only happens when we start
a VM.

Previously, we used to have a bitmap with enabled CPUs and that was
preventing [masking] the issue.

Initialized loaded_vmcss_on_cpu list earlier, right before we assign
crash_vmclear_loaded_vmcss pointer. blocked_vcpu_on_cpu list and
blocked_vcpu_on_cpu_lock are moved altogether for consistency.

Fixes: 31603d4fc2bb ("KVM: VMX: Always VMCLEAR in-use VMCSes during crash with kexec support")
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Message-Id: <20200401081348.1345307-1-vkuznets@redhat.com>
Reviewed-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: VMX: Add a trampoline to fix VMREAD error handling
Sean Christopherson [Thu, 26 Mar 2020 16:07:12 +0000 (09:07 -0700)]
KVM: VMX: Add a trampoline to fix VMREAD error handling

commit 842f4be95899df22b5843ba1a7c8cf37e831a6e8 upstream.

Add a hand coded assembly trampoline to preserve volatile registers
across vmread_error(), and to handle the calling convention differences
between 64-bit and 32-bit due to asmlinkage on vmread_error().  Pass
@field and @fault on the stack when invoking the trampoline to avoid
clobbering volatile registers in the context of the inline assembly.

Calling vmread_error() directly from inline assembly is partially broken
on 64-bit, and completely broken on 32-bit.  On 64-bit, it will clobber
%rdi and %rsi (used to pass @field and @fault) and any volatile regs
written by vmread_error().  On 32-bit, asmlinkage means vmread_error()
expects the parameters to be passed on the stack, not via regs.

Opportunistically zero out the result in the trampoline to save a few
bytes of code for every VMREAD.  A happy side effect of the trampoline
is that the inline code footprint is reduced by three bytes on 64-bit
due to PUSH/POP being more efficent (in terms of opcode bytes) than MOV.

Fixes: 6e2020977e3e6 ("KVM: VMX: Add error handling to VMREAD helper")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Message-Id: <20200326160712.28803-1-sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: x86: Gracefully handle __vmalloc() failure during VM allocation
Sean Christopherson [Mon, 27 Jan 2020 00:41:11 +0000 (16:41 -0800)]
KVM: x86: Gracefully handle __vmalloc() failure during VM allocation

commit d18b2f43b9147c8005ae0844fb445d8cc6a87e31 upstream.

Check the result of __vmalloc() to avoid dereferencing a NULL pointer in
the event that allocation failres.

Fixes: d1e5b0e98ea27 ("kvm: Make VM ioctl do valloc for some archs")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: VMX: Always VMCLEAR in-use VMCSes during crash with kexec support
Sean Christopherson [Sat, 21 Mar 2020 19:37:49 +0000 (12:37 -0700)]
KVM: VMX: Always VMCLEAR in-use VMCSes during crash with kexec support

commit 31603d4fc2bb4f0815245d496cb970b27b4f636a upstream.

VMCLEAR all in-use VMCSes during a crash, even if kdump's NMI shootdown
interrupted a KVM update of the percpu in-use VMCS list.

Because NMIs are not blocked by disabling IRQs, it's possible that
crash_vmclear_local_loaded_vmcss() could be called while the percpu list
of VMCSes is being modified, e.g. in the middle of list_add() in
vmx_vcpu_load_vmcs().  This potential corner case was called out in the
original commit[*], but the analysis of its impact was wrong.

Skipping the VMCLEARs is wrong because it all but guarantees that a
loaded, and therefore cached, VMCS will live across kexec and corrupt
memory in the new kernel.  Corruption will occur because the CPU's VMCS
cache is non-coherent, i.e. not snooped, and so the writeback of VMCS
memory on its eviction will overwrite random memory in the new kernel.
The VMCS will live because the NMI shootdown also disables VMX, i.e. the
in-progress VMCLEAR will #UD, and existing Intel CPUs do not flush the
VMCS cache on VMXOFF.

Furthermore, interrupting list_add() and list_del() is safe due to
crash_vmclear_local_loaded_vmcss() using forward iteration.  list_add()
ensures the new entry is not visible to forward iteration unless the
entire add completes, via WRITE_ONCE(prev->next, new).  A bad "prev"
pointer could be observed if the NMI shootdown interrupted list_del() or
list_add(), but list_for_each_entry() does not consume ->prev.

In addition to removing the temporary disabling of VMCLEAR, open code
loaded_vmcs_init() in __loaded_vmcs_clear() and reorder VMCLEAR so that
the VMCS is deleted from the list only after it's been VMCLEAR'd.
Deleting the VMCS before VMCLEAR would allow a race where the NMI
shootdown could arrive between list_del() and vmcs_clear() and thus
neither flow would execute a successful VMCLEAR.  Alternatively, more
code could be moved into loaded_vmcs_init(), but that gets rather silly
as the only other user, alloc_loaded_vmcs(), doesn't need the smp_wmb()
and would need to work around the list_del().

Update the smp_*() comments related to the list manipulation, and
opportunistically reword them to improve clarity.

[*] https://patchwork.kernel.org/patch/1675731/#3720461

Fixes: 8f536b7697a0 ("KVM: VMX: provide the vmclear function and a bitmap to support VMCLEAR in kdump")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Message-Id: <20200321193751.24985-2-sean.j.christopherson@intel.com>
Reviewed-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: x86: Allocate new rmap and large page tracking when moving memslot
Sean Christopherson [Tue, 18 Feb 2020 21:07:15 +0000 (13:07 -0800)]
KVM: x86: Allocate new rmap and large page tracking when moving memslot

commit edd4fa37baa6ee8e44dc65523b27bd6fe44c94de upstream.

Reallocate a rmap array and recalcuate large page compatibility when
moving an existing memslot to correctly handle the alignment properties
of the new memslot.  The number of rmap entries required at each level
is dependent on the alignment of the memslot's base gfn with respect to
that level, e.g. moving a large-page aligned memslot so that it becomes
unaligned will increase the number of rmap entries needed at the now
unaligned level.

Not updating the rmap array is the most obvious bug, as KVM accesses
garbage data beyond the end of the rmap.  KVM interprets the bad data as
pointers, leading to non-canonical #GPs, unexpected #PFs, etc...

  general protection fault: 0000 [#1] SMP
  CPU: 0 PID: 1909 Comm: move_memory_reg Not tainted 5.4.0-rc7+ #139
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:rmap_get_first+0x37/0x50 [kvm]
  Code: <48> 8b 3b 48 85 ff 74 ec e8 6c f4 ff ff 85 c0 74 e3 48 89 d8 5b c3
  RSP: 0018:ffffc9000021bbc8 EFLAGS: 00010246
  RAX: ffff00617461642e RBX: ffff00617461642e RCX: 0000000000000012
  RDX: ffff88827400f568 RSI: ffffc9000021bbe0 RDI: ffff88827400f570
  RBP: 0010000000000000 R08: ffffc9000021bd00 R09: ffffc9000021bda8
  R10: ffffc9000021bc48 R11: 0000000000000000 R12: 0030000000000000
  R13: 0000000000000000 R14: ffff88827427d700 R15: ffffc9000021bce8
  FS:  00007f7eda014700(0000) GS:ffff888277a00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f7ed9216ff8 CR3: 0000000274391003 CR4: 0000000000162eb0
  Call Trace:
   kvm_mmu_slot_set_dirty+0xa1/0x150 [kvm]
   __kvm_set_memory_region.part.64+0x559/0x960 [kvm]
   kvm_set_memory_region+0x45/0x60 [kvm]
   kvm_vm_ioctl+0x30f/0x920 [kvm]
   do_vfs_ioctl+0xa1/0x620
   ksys_ioctl+0x66/0x70
   __x64_sys_ioctl+0x16/0x20
   do_syscall_64+0x4c/0x170
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x7f7ed9911f47
  Code: <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 21 6f 2c 00 f7 d8 64 89 01 48
  RSP: 002b:00007ffc00937498 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
  RAX: ffffffffffffffda RBX: 0000000001ab0010 RCX: 00007f7ed9911f47
  RDX: 0000000001ab1350 RSI: 000000004020ae46 RDI: 0000000000000004
  RBP: 000000000000000a R08: 0000000000000000 R09: 00007f7ed9214700
  R10: 00007f7ed92149d0 R11: 0000000000000246 R12: 00000000bffff000
  R13: 0000000000000003 R14: 00007f7ed9215000 R15: 0000000000000000
  Modules linked in: kvm_intel kvm irqbypass
  ---[ end trace 0c5f570b3358ca89 ]---

The disallow_lpage tracking is more subtle.  Failure to update results
in KVM creating large pages when it shouldn't, either due to stale data
or again due to indexing beyond the end of the metadata arrays, which
can lead to memory corruption and/or leaking data to guest/userspace.

Note, the arrays for the old memslot are freed by the unconditional call
to kvm_free_memslot() in __kvm_set_memory_region().

Fixes: 05da45583de9b ("KVM: MMU: large page support")
Cc: stable@vger.kernel.org
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: s390: vsie: Fix delivery of addressing exceptions
David Hildenbrand [Fri, 3 Apr 2020 15:30:47 +0000 (17:30 +0200)]
KVM: s390: vsie: Fix delivery of addressing exceptions

commit 4d4cee96fb7a3cc53702a9be8299bf525be4ee98 upstream.

Whenever we get an -EFAULT, we failed to read in guest 2 physical
address space. Such addressing exceptions are reported via a program
intercept to the nested hypervisor.

We faked the intercept, we have to return to guest 2. Instead, right
now we would be returning -EFAULT from the intercept handler, eventually
crashing the VM.
the correct thing to do is to return 1 as rc == 1 is the internal
representation of "we have to go back into g2".

Addressing exceptions can only happen if the g2->g3 page tables
reference invalid g2 addresses (say, either a table or the final page is
not accessible - so something that basically never happens in sane
environments.

Identified by manual code inspection.

Fixes: a3508fbe9dc6 ("KVM: s390: vsie: initial support for nested virtualization")
Cc: <stable@vger.kernel.org> # v4.8+
Signed-off-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/20200403153050.20569-3-david@redhat.com
Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
[borntraeger@de.ibm.com: fix patch description]
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: s390: vsie: Fix region 1 ASCE sanity shadow address checks
David Hildenbrand [Fri, 3 Apr 2020 15:30:46 +0000 (17:30 +0200)]
KVM: s390: vsie: Fix region 1 ASCE sanity shadow address checks

commit a1d032a49522cb5368e5dfb945a85899b4c74f65 upstream.

In case we have a region 1 the following calculation
(31 + ((gmap->asce & _ASCE_TYPE_MASK) >> 2)*11)
results in 64. As shifts beyond the size are undefined the compiler is
free to use instructions like sllg. sllg will only use 6 bits of the
shift value (here 64) resulting in no shift at all. That means that ALL
addresses will be rejected.

The can result in endless loops, e.g. when prefix cannot get mapped.

Fixes: 4be130a08420 ("s390/mm: add shadow gmap support")
Tested-by: Janosch Frank <frankja@linux.ibm.com>
Reported-by: Janosch Frank <frankja@linux.ibm.com>
Cc: <stable@vger.kernel.org> # v4.8+
Signed-off-by: David Hildenbrand <david@redhat.com>
Link: https://lore.kernel.org/r/20200403153050.20569-2-david@redhat.com
Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
[borntraeger@de.ibm.com: fix patch description, remove WARN_ON_ONCE]
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKVM: nVMX: Properly handle userspace interrupt window request
Sean Christopherson [Tue, 3 Mar 2020 06:27:35 +0000 (22:27 -0800)]
KVM: nVMX: Properly handle userspace interrupt window request

commit a1c77abb8d93381e25a8d2df3a917388244ba776 upstream.

Return true for vmx_interrupt_allowed() if the vCPU is in L2 and L1 has
external interrupt exiting enabled.  IRQs are never blocked in hardware
if the CPU is in the guest (L2 from L1's perspective) when IRQs trigger
VM-Exit.

The new check percolates up to kvm_vcpu_ready_for_interrupt_injection()
and thus vcpu_run(), and so KVM will exit to userspace if userspace has
requested an interrupt window (to inject an IRQ into L1).

Remove the @external_intr param from vmx_check_nested_events(), which is
actually an indicator that userspace wants an interrupt window, e.g.
it's named @req_int_win further up the stack.  Injecting a VM-Exit into
L1 to try and bounce out to L0 userspace is all kinds of broken and is
no longer necessary.

Remove the hack in nested_vmx_vmexit() that attempted to workaround the
breakage in vmx_check_nested_events() by only filling interrupt info if
there's an actual interrupt pending.  The hack actually made things
worse because it caused KVM to _never_ fill interrupt info when the
LAPIC resides in userspace (kvm_cpu_has_interrupt() queries
interrupt.injected, which is always cleared by prepare_vmcs12() before
reaching the hack in nested_vmx_vmexit()).

Fixes: 6550c4df7e50 ("KVM: nVMX: Fix interrupt window request with "Acknowledge interrupt on exit"")
Cc: stable@vger.kernel.org
Cc: Liran Alon <liran.alon@oracle.com>
Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoplatform/x86: asus-wmi: Support laptops where the first battery is named BATT
Kristian Klausen [Tue, 3 Mar 2020 18:02:15 +0000 (19:02 +0100)]
platform/x86: asus-wmi: Support laptops where the first battery is named BATT

commit 6b3586d45bba14f6912f37488090c37a3710e7b4 upstream.

The WMI method to set the charge threshold does not provide a
way to specific a battery, so we assume it is the first/primary
battery (by checking if the name is BAT0).
On some newer ASUS laptops (Zenbook UM431DA) though, the
primary/first battery isn't named BAT0 but BATT, so we need
to support that case.

Fixes: 7973353e92ee ("platform/x86: asus-wmi: Refactor charge threshold to use the battery hooking API")
Cc: stable@vger.kernel.org
Signed-off-by: Kristian Klausen <kristian@klausen.dk>
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agox86/entry/32: Add missing ASM_CLAC to general_protection entry
Thomas Gleixner [Tue, 25 Feb 2020 21:36:37 +0000 (22:36 +0100)]
x86/entry/32: Add missing ASM_CLAC to general_protection entry

commit 3d51507f29f2153a658df4a0674ec5b592b62085 upstream.

All exception entry points must have ASM_CLAC right at the
beginning. The general_protection entry is missing one.

Fixes: e59d1b0a2419 ("x86-32, smap: Add STAC/CLAC instructions to 32-bit kernel entry")
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Reviewed-by: Alexandre Chartre <alexandre.chartre@oracle.com>
Reviewed-by: Andy Lutomirski <luto@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20200225220216.219537887@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agox86/tsc_msr: Make MSR derived TSC frequency more accurate
Hans de Goede [Sun, 23 Feb 2020 14:06:10 +0000 (15:06 +0100)]
x86/tsc_msr: Make MSR derived TSC frequency more accurate

commit fac01d11722c92a186b27ee26cd429a8066adfb5 upstream.

The "Intel 64 and IA-32 Architectures Software Developer’s Manual Volume 4:
Model-Specific Registers" has the following table for the values from
freq_desc_byt:

   000B: 083.3 MHz
   001B: 100.0 MHz
   010B: 133.3 MHz
   011B: 116.7 MHz
   100B: 080.0 MHz

Notice how for e.g the 83.3 MHz value there are 3 significant digits, which
translates to an accuracy of a 1000 ppm, where as a typical crystal
oscillator is 20 - 100 ppm, so the accuracy of the frequency format used in
the Software Developer’s Manual is not really helpful.

As far as we know Bay Trail SoCs use a 25 MHz crystal and Cherry Trail
uses a 19.2 MHz crystal, the crystal is the source clock for a root PLL
which outputs 1600 and 100 MHz. It is unclear if the root PLL outputs are
used directly by the CPU clock PLL or if there is another PLL in between.

This does not matter though, we can model the chain of PLLs as a single PLL
with a quotient equal to the quotients of all PLLs in the chain multiplied.

So we can create a simplified model of the CPU clock setup using a
reference clock of 100 MHz plus a quotient which gets us as close to the
frequency from the SDM as possible.

For the 83.3 MHz example from above this would give 100 MHz * 5 / 6 = 83
and 1/3 MHz, which matches exactly what has been measured on actual
hardware.

Use a simplified PLL model with a reference clock of 100 MHz for all Bay
and Cherry Trail models.

This has been tested on the following models:

              CPU freq before:        CPU freq after:
Intel N2840   2165.800 MHz            2166.667 MHz
Intel Z3736   1332.800 MHz            1333.333 MHz
Intel Z3775   1466.300 MHz            1466.667 MHz
Intel Z8350   1440.000 MHz            1440.000 MHz
Intel Z8750   1600.000 MHz            1600.000 MHz

This fixes the time drifting by about 1 second per hour (20 - 30 seconds
per day) on (some) devices which rely on the tsc_msr.c code to determine
the TSC frequency.

Reported-by: Vipul Kumar <vipulk0511@gmail.com>
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
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/20200223140610.59612-3-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agox86/tsc_msr: Fix MSR_FSB_FREQ mask for Cherry Trail devices
Hans de Goede [Sun, 23 Feb 2020 14:06:09 +0000 (15:06 +0100)]
x86/tsc_msr: Fix MSR_FSB_FREQ mask for Cherry Trail devices

commit c8810e2ffc30c7e1577f9c057c4b85d984bbc35a upstream.

According to the "Intel 64 and IA-32 Architectures Software Developer's
Manual Volume 4: Model-Specific Registers" on Cherry Trail (Airmont)
devices the 4 lowest bits of the MSR_FSB_FREQ mask indicate the bus freq
unlike on e.g. Bay Trail where only the lowest 3 bits are used.

This is also the reason why MAX_NUM_FREQS is defined as 9, since Cherry
Trail SoCs have 9 possible frequencies, so the lo value from the MSR needs
to be masked with 0x0f, not with 0x07 otherwise the 9th frequency will get
interpreted as the 1st.

Bump MAX_NUM_FREQS to 16 to avoid any possibility of addressing the array
out of bounds and makes the mask part of the cpufreq struct so it can be
set it per model.

While at it also log an error when the index points to an uninitialized
part of the freqs lookup-table.

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/20200223140610.59612-2-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agox86/tsc_msr: Use named struct initializers
Hans de Goede [Sun, 23 Feb 2020 14:06:08 +0000 (15:06 +0100)]
x86/tsc_msr: Use named struct initializers

commit 812c2d7506fde7cdf83cb2532810a65782b51741 upstream.

Use named struct initializers for the freq_desc struct-s initialization
and change the "u8 msr_plat" to a "bool use_msr_plat" to make its meaning
more clear instead of relying on a comment to explain it.

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/20200223140610.59612-1-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agosignal: Extend exec_id to 64bits
Eric W. Biederman [Tue, 31 Mar 2020 00:01:04 +0000 (19:01 -0500)]
signal: Extend exec_id to 64bits

commit d1e7fd6462ca9fc76650fbe6ca800e35b24267da upstream.

Replace the 32bit exec_id with a 64bit exec_id to make it impossible
to wrap the exec_id counter.  With care an attacker can cause exec_id
wrap and send arbitrary signals to a newly exec'd parent.  This
bypasses the signal sending checks if the parent changes their
credentials during exec.

The severity of this problem can been seen that in my limited testing
of a 32bit exec_id it can take as little as 19s to exec 65536 times.
Which means that it can take as little as 14 days to wrap a 32bit
exec_id.  Adam Zabrocki has succeeded wrapping the self_exe_id in 7
days.  Even my slower timing is in the uptime of a typical server.
Which means self_exec_id is simply a speed bump today, and if exec
gets noticably faster self_exec_id won't even be a speed bump.

Extending self_exec_id to 64bits introduces a problem on 32bit
architectures where reading self_exec_id is no longer atomic and can
take two read instructions.  Which means that is is possible to hit
a window where the read value of exec_id does not match the written
value.  So with very lucky timing after this change this still
remains expoiltable.

I have updated the update of exec_id on exec to use WRITE_ONCE
and the read of exec_id in do_notify_parent to use READ_ONCE
to make it clear that there is no locking between these two
locations.

Link: https://lore.kernel.org/kernel-hardening/20200324215049.GA3710@pi3.com.pl
Fixes: 2.3.23pre2
Cc: stable@vger.kernel.org
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoath9k: Handle txpower changes even when TPC is disabled
Remi Pommarel [Sat, 29 Feb 2020 16:13:47 +0000 (17:13 +0100)]
ath9k: Handle txpower changes even when TPC is disabled

commit 968ae2caad0782db5dbbabb560d3cdefd2945d38 upstream.

When TPC is disabled IEEE80211_CONF_CHANGE_POWER event can be handled to
reconfigure HW's maximum txpower.

This fixes 0dBm txpower setting when user attaches to an interface for
the first time with the following scenario:

ieee80211_do_open()
    ath9k_add_interface()
        ath9k_set_txpower() /* Set TX power with not yet initialized
                               sc->hw->conf.power_level */

    ieee80211_hw_config() /* Iniatilize sc->hw->conf.power_level and
                             raise IEEE80211_CONF_CHANGE_POWER */

    ath9k_config() /* IEEE80211_CONF_CHANGE_POWER is ignored */

This issue can be reproduced with the following:

  $ modprobe -r ath9k
  $ modprobe ath9k
  $ wpa_supplicant -i wlan0 -c /tmp/wpa.conf &
  $ iw dev /* Here TX power is either 0 or 3 depending on RF chain */
  $ killall wpa_supplicant
  $ iw dev /* TX power goes back to calibrated value and subsequent
              calls will be fine */

Fixes: 283dd11994cde ("ath9k: add per-vif TX power capability")
Cc: stable@vger.kernel.org
Signed-off-by: Remi Pommarel <repk@triplefau.lt>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPM: sleep: wakeup: Skip wakeup_source_sysfs_remove() if device is not there
Neeraj Upadhyay [Mon, 23 Mar 2020 05:08:51 +0000 (10:38 +0530)]
PM: sleep: wakeup: Skip wakeup_source_sysfs_remove() if device is not there

commit 87de6594dc45dbf6819f3e0ef92f9331c5a9444c upstream.

Skip wakeup_source_sysfs_remove() to fix a NULL pinter dereference via
ws->dev, if the wakeup source is unregistered before registering the
wakeup class from device_add().

Fixes: 2ca3d1ecb8c4 ("PM / wakeup: Register wakeup class kobj after device is added")
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>
Cc: 5.4+ <stable@vger.kernel.org> # 5.4+
[ rjw: Subject & changelog, white space ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPM / Domains: Allow no domain-idle-states DT property in genpd when parsing
Ulf Hansson [Tue, 10 Mar 2020 10:40:23 +0000 (11:40 +0100)]
PM / Domains: Allow no domain-idle-states DT property in genpd when parsing

commit 56cb26891ea4180121265dc6b596015772c4a4b8 upstream.

Commit 2c361684803e ("PM / Domains: Don't treat zero found compatible idle
states as an error"), moved of_genpd_parse_idle_states() towards allowing
none compatible idle state to be found for the device node, rather than
returning an error code.

However, it didn't consider that the "domain-idle-states" DT property may
be missing as it's optional, which makes of_count_phandle_with_args() to
return -ENOENT. Let's fix this to make the behaviour consistent.

Fixes: 2c361684803e ("PM / Domains: Don't treat zero found compatible idle states as an error")
Reported-by: Benjamin Gaignard <benjamin.gaignard@st.com>
Cc: 4.20+ <stable@vger.kernel.org> # 4.20+
Reviewed-by: Sudeep Holla <sudeep.holla@arm.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoMIPS: OCTEON: irq: Fix potential NULL pointer dereference
Gustavo A. R. Silva [Tue, 22 Jan 2019 20:18:42 +0000 (14:18 -0600)]
MIPS: OCTEON: irq: Fix potential NULL pointer dereference

commit 792a402c2840054533ef56279c212ef6da87d811 upstream.

There is a potential NULL pointer dereference in case kzalloc()
fails and returns NULL.

Fix this by adding a NULL check on *cd*

This bug was detected with the help of Coccinelle.

Fixes: 64b139f97c01 ("MIPS: OCTEON: irq: add CIB and other fixes")
Cc: stable@vger.kernel.org
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoMIPS/tlbex: Fix LDDIR usage in setup_pw() for Loongson-3
Huacai Chen [Wed, 25 Mar 2020 03:44:54 +0000 (11:44 +0800)]
MIPS/tlbex: Fix LDDIR usage in setup_pw() for Loongson-3

commit d191aaffe3687d1e73e644c185f5f0550ec242b5 upstream.

LDDIR/LDPTE is Loongson-3's acceleration for Page Table Walking. If BD
(Base Directory, the 4th page directory) is not enabled, then GDOffset
is biased by BadVAddr[63:62]. So, if GDOffset (aka. BadVAddr[47:36] for
Loongson-3) is big enough, "0b11(BadVAddr[63:62])|BadVAddr[47:36]|...."
can far beyond pg_swapper_dir. This means the pg_swapper_dir may NOT be
accessed by LDDIR correctly, so fix it by set PWDirExt in CP0_PWCtl.

Cc: <stable@vger.kernel.org>
Signed-off-by: Pei Huang <huangpei@loongson.cn>
Signed-off-by: Huacai Chen <chenhc@lemote.com>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agopstore: pstore_ftrace_seq_next should increase position index
Vasily Averin [Tue, 25 Feb 2020 08:11:20 +0000 (11:11 +0300)]
pstore: pstore_ftrace_seq_next should increase position index

commit 6c871b7314dde9ab64f20de8f5aa3d01be4518e8 upstream.

In Aug 2018 NeilBrown noticed
commit 1f4aace60b0e ("fs/seq_file.c: simplify seq_file iteration code and interface")
"Some ->next functions do not increment *pos when they return NULL...
Note that such ->next functions are buggy and should be fixed.
A simple demonstration is

 dd if=/proc/swaps bs=1000 skip=1

Choose any block size larger than the size of /proc/swaps. This will
always show the whole last line of /proc/swaps"

/proc/swaps output was fixed recently, however there are lot of other
affected files, and one of them is related to pstore subsystem.

If .next function does not change position index, following .show function
will repeat output related to current position index.

There are at least 2 related problems:
- read after lseek beyond end of file, described above by NeilBrown
  "dd if=<AFFECTED_FILE> bs=1000 skip=1" will generate whole last list
- read after lseek on in middle of last line will output expected rest of
  last line but then repeat whole last line once again.

If .show() function generates multy-line output (like
pstore_ftrace_seq_show() does ?) following bash script cycles endlessly

 $ q=;while read -r r;do echo "$((++q)) $r";done < AFFECTED_FILE

Unfortunately I'm not familiar enough to pstore subsystem and was unable
to find affected pstore-related file on my test node.

If .next function does not change position index, following .show function
will repeat output related to current position index.

Cc: stable@vger.kernel.org
Fixes: 1f4aace60b0e ("fs/seq_file.c: simplify seq_file iteration code ...")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=206283
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Link: https://lore.kernel.org/r/4e49830d-4c88-0171-ee24-1ee540028dad@virtuozzo.com
[kees: with robustness tweak from Joel Fernandes <joelaf@google.com>]
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoio_uring: remove bogus RLIMIT_NOFILE check in file registration
Jens Axboe [Fri, 3 Apr 2020 19:54:26 +0000 (13:54 -0600)]
io_uring: remove bogus RLIMIT_NOFILE check in file registration

commit c336e992cb1cb1db9ee608dfb30342ae781057ab upstream.

We already checked this limit when the file was opened, and we keep it
open in the file table. Hence when we added unit_inflight to the count
we want to register, we're doubly accounting these files. This results
in -EMFILE for file registration, if we're at half the limit.

Cc: stable@vger.kernel.org # v5.1+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoirqchip/versatile-fpga: Apply clear-mask earlier
Sungbo Eo [Sat, 21 Mar 2020 13:38:42 +0000 (22:38 +0900)]
irqchip/versatile-fpga: Apply clear-mask earlier

commit 6a214a28132f19ace3d835a6d8f6422ec80ad200 upstream.

Clear its own IRQs before the parent IRQ get enabled, so that the
remaining IRQs do not accidentally interrupt the parent IRQ controller.

This patch also fixes a reboot bug on OX820 SoC, where the remaining
rps-timer IRQ raises a GIC interrupt that is left pending. After that,
the rps-timer IRQ is cleared during driver initialization, and there's
no IRQ left in rps-irq when local_irq_enable() is called, which evokes
an error message "unexpected IRQ trap".

Fixes: bdd272cbb97a ("irqchip: versatile FPGA: support cascaded interrupts from DT")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200321133842.2408823-1-mans0n@gorani.run
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agogenirq/debugfs: Add missing sanity checks to interrupt injection
Thomas Gleixner [Fri, 6 Mar 2020 13:03:42 +0000 (14:03 +0100)]
genirq/debugfs: Add missing sanity checks to interrupt injection

commit a740a423c36932695b01a3e920f697bc55b05fec upstream.

Interrupts cannot be injected when the interrupt is not activated and when
a replay is already in progress.

Fixes: 536e2e34bd00 ("genirq/debugfs: Triggering of interrupts from userspace")
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Marc Zyngier <maz@kernel.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20200306130623.500019114@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agocpu/hotplug: Ignore pm_wakeup_pending() for disable_nonboot_cpus()
Thomas Gleixner [Fri, 27 Mar 2020 11:06:44 +0000 (12:06 +0100)]
cpu/hotplug: Ignore pm_wakeup_pending() for disable_nonboot_cpus()

commit e98eac6ff1b45e4e73f2e6031b37c256ccb5d36b upstream.

A recent change to freeze_secondary_cpus() which added an early abort if a
wakeup is pending missed the fact that the function is also invoked for
shutdown, reboot and kexec via disable_nonboot_cpus().

In case of disable_nonboot_cpus() the wakeup event needs to be ignored as
the purpose is to terminate the currently running kernel.

Add a 'suspend' argument which is only set when the freeze is in context of
a suspend operation. If not set then an eventually pending wakeup event is
ignored.

Fixes: a66d955e910a ("cpu/hotplug: Abort disabling secondary CPUs if wakeup is pending")
Reported-by: Boqun Feng <boqun.feng@gmail.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Pavankumar Kondeti <pkondeti@codeaurora.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/874kuaxdiz.fsf@nanos.tec.linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoKEYS: reaching the keys quotas correctly
Yang Xu [Fri, 28 Feb 2020 04:41:51 +0000 (12:41 +0800)]
KEYS: reaching the keys quotas correctly

commit 2e356101e72ab1361821b3af024d64877d9a798d upstream.

Currently, when we add a new user key, the calltrace as below:

add_key()
  key_create_or_update()
    key_alloc()
    __key_instantiate_and_link
      generic_key_instantiate
        key_payload_reserve
          ......

Since commit a08bf91ce28e ("KEYS: allow reaching the keys quotas exactly"),
we can reach max bytes/keys in key_alloc, but we forget to remove this
limit when we reserver space for payload in key_payload_reserve. So we
can only reach max keys but not max bytes when having delta between plen
and type->def_datalen. Remove this limit when instantiating the key, so we
can keep consistent with key_alloc.

Also, fix the similar problem in keyctl_chown_key().

Fixes: 0b77f5bfb45c ("keys: make the keyring quotas controllable through /proc/sys")
Fixes: a08bf91ce28e ("KEYS: allow reaching the keys quotas exactly")
Cc: stable@vger.kernel.org # 5.0.x
Cc: Eric Biggers <ebiggers@google.com>
Signed-off-by: Yang Xu <xuyang2018.jy@cn.fujitsu.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agotpm: tpm2_bios_measurements_next should increase position index
Vasily Averin [Tue, 25 Feb 2020 06:26:22 +0000 (09:26 +0300)]
tpm: tpm2_bios_measurements_next should increase position index

commit f9bf8adb55cd5a357b247a16aafddf8c97b276e0 upstream.

If .next function does not change position index,
following .show function will repeat output related
to current position index.

For /sys/kernel/security/tpm0/binary_bios_measurements:
1) read after lseek beyound end of file generates whole last line.
2) read after lseek to middle of last line generates
expected end of last line and unexpected whole last line once again.

Cc: stable@vger.kernel.org # 4.19.x
Fixes: 1f4aace60b0e ("fs/seq_file.c: simplify seq_file iteration code ...")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=206283
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agotpm: tpm1_bios_measurements_next should increase position index
Vasily Averin [Tue, 25 Feb 2020 06:26:08 +0000 (09:26 +0300)]
tpm: tpm1_bios_measurements_next should increase position index

commit d7a47b96ed1102551eb7325f97937e276fb91045 upstream.

If .next function does not change position index,
following .show function will repeat output related
to current position index.

In case of /sys/kernel/security/tpm0/ascii_bios_measurements
and binary_bios_measurements:
1) read after lseek beyound end of file generates whole last line.
2) read after lseek to middle of last line generates
expected end of last line and unexpected whole last line once again.

Cc: stable@vger.kernel.org # 4.19.x
Fixes: 1f4aace60b0e ("fs/seq_file.c: simplify seq_file iteration code ...")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=206283
Signed-off-by: Vasily Averin <vvs@virtuozzo.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agotpm: Don't make log failures fatal
Matthew Garrett [Thu, 2 Jan 2020 21:55:18 +0000 (13:55 -0800)]
tpm: Don't make log failures fatal

commit 805fa88e0780b7ce1cc9b649dd91a0a7164c6eb4 upstream.

If a TPM is in disabled state, it's reasonable for it to have an empty
log. Bailing out of probe in this case means that the PPI interface
isn't available, so there's no way to then enable the TPM from the OS.
In general it seems reasonable to ignore log errors - they shouldn't
interfere with any other TPM functionality.

Signed-off-by: Matthew Garrett <matthewgarrett@google.com>
Cc: stable@vger.kernel.org # 4.19.x
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>
Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agosched/fair: Fix enqueue_task_fair warning
Vincent Guittot [Fri, 6 Mar 2020 13:52:57 +0000 (14:52 +0100)]
sched/fair: Fix enqueue_task_fair warning

commit fe61468b2cbc2b7ce5f8d3bf32ae5001d4c434e9 upstream.

When a cfs rq is throttled, the latter and its child are removed from the
leaf list but their nr_running is not changed which includes staying higher
than 1. When a task is enqueued in this throttled branch, the cfs rqs must
be added back in order to ensure correct ordering in the list but this can
only happens if nr_running == 1.
When cfs bandwidth is used, we call unconditionnaly list_add_leaf_cfs_rq()
when enqueuing an entity to make sure that the complete branch will be
added.

Similarly unthrottle_cfs_rq() can stop adding cfs in the list when a parent
is throttled. Iterate the remaining entity to ensure that the complete
branch will be added in the list.

Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: stable@vger.kernel.org
Cc: stable@vger.kernel.org #v5.1+
Link: https://lkml.kernel.org/r/20200306135257.25044-1-vincent.guittot@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPCI: endpoint: Fix for concurrent memory allocation in OB address region
Kishon Vijay Abraham I [Mon, 24 Feb 2020 09:53:36 +0000 (15:23 +0530)]
PCI: endpoint: Fix for concurrent memory allocation in OB address region

commit 04e046ca57ebed3943422dee10eec9e73aec081e upstream.

pci-epc-mem uses a bitmap to manage the Endpoint outbound (OB) address
region. This address region will be shared by multiple endpoint
functions (in the case of multi function endpoint) and it has to be
protected from concurrent access to avoid updating an inconsistent state.

Use a mutex to protect bitmap updates to prevent the memory
allocation API from returning incorrect addresses.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: stable@vger.kernel.org # v4.14+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPCI: qcom: Fix the fixup of PCI_VENDOR_ID_QCOM
Bjorn Andersson [Fri, 27 Dec 2019 01:27:17 +0000 (17:27 -0800)]
PCI: qcom: Fix the fixup of PCI_VENDOR_ID_QCOM

commit 604f3956524a6a53c1e3dd27b4b685b664d181ec upstream.

There exists non-bridge PCIe devices with PCI_VENDOR_ID_QCOM, so limit
the fixup to only affect the relevant PCIe bridges.

Fixes: 322f03436692 ("PCI: qcom: Use default config space read function")
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Acked-by: Stanimir Varbanov <svarbanov@mm-sol.com>
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPCI: Add boot interrupt quirk mechanism for Xeon chipsets
Sean V Kelley [Thu, 20 Feb 2020 19:29:29 +0000 (11:29 -0800)]
PCI: Add boot interrupt quirk mechanism for Xeon chipsets

commit b88bf6c3b6ff77948c153cac4e564642b0b90632 upstream.

The following was observed by Kar Hin Ong with RT patchset:

  Backtrace:
  irq 19: nobody cared (try booting with the "irqpoll" option)
  CPU: 0 PID: 3329 Comm: irq/34-nipalk Tainted:4.14.87-rt49 #1
  Hardware name: National Instruments NI PXIe-8880/NI PXIe-8880,
           BIOS 2.1.5f1 01/09/2020
  Call Trace:
  <IRQ>
    ? dump_stack+0x46/0x5e
    ? __report_bad_irq+0x2e/0xb0
    ? note_interrupt+0x242/0x290
    ? nNIKAL100_memoryRead16+0x8/0x10 [nikal]
    ? handle_irq_event_percpu+0x55/0x70
    ? handle_irq_event+0x4f/0x80
    ? handle_fasteoi_irq+0x81/0x180
    ? handle_irq+0x1c/0x30
    ? do_IRQ+0x41/0xd0
    ? common_interrupt+0x84/0x84
  </IRQ>
  ...
  handlers:
  [<ffffffffb3297200>] irq_default_primary_handler threaded
  [<ffffffffb3669180>] usb_hcd_irq
  Disabling IRQ #19

The problem being that this device is triggering boot interrupts
due to threaded interrupt handling and masking of the IO-APIC. These
boot interrupts are then forwarded on to the legacy PCH's PIRQ lines
where there is no handler present for the device.

Whenever a PCI device fires interrupt (INTx) to Pin 20 of IOAPIC 2
(GSI 44), the kernel receives two interrupts:

   1. Interrupt from Pin 20 of IOAPIC 2  -> Expected
   2. Interrupt from Pin 19 of IOAPIC 1  -> UNEXPECTED

Quirks for disabling boot interrupts (preferred) or rerouting the
handler exist but do not address these Xeon chipsets' mechanism:
https://lore.kernel.org/lkml/12131949181903-git-send-email-sassmann@suse.de/

Add a new mechanism via PCI CFG for those chipsets supporting CIPINTRC
register's dis_intx_rout2ich bit.

Link: https://lore.kernel.org/r/20200220192930.64820-2-sean.v.kelley@linux.intel.com
Reported-by: Kar Hin Ong <kar.hin.ong@ni.com>
Tested-by: Kar Hin Ong <kar.hin.ong@ni.com>
Signed-off-by: Sean V Kelley <sean.v.kelley@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPCI/ASPM: Clear the correct bits when enabling L1 substates
Yicong Yang [Fri, 13 Mar 2020 09:53:47 +0000 (17:53 +0800)]
PCI/ASPM: Clear the correct bits when enabling L1 substates

commit 58a3862a10a317a81097ab0c78aecebabb1704f5 upstream.

In pcie_config_aspm_l1ss(), we cleared the wrong bits when enabling ASPM L1
Substates.  Instead of the L1.x enable bits (PCI_L1SS_CTL1_L1SS_MASK, 0xf), we
cleared the Link Activation Interrupt Enable bit (PCI_L1SS_CAP_L1_PM_SS,
0x10).

Clear the L1.x enable bits before writing the new L1.x configuration.

[bhelgaas: changelog]
Fixes: aeda9adebab8 ("PCI/ASPM: Configure L1 substate settings")
Link: https://lore.kernel.org/r/1584093227-1292-1-git-send-email-yangyicong@hisilicon.com
Signed-off-by: Yicong Yang <yangyicong@hisilicon.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
CC: stable@vger.kernel.org # v4.11+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoPCI: pciehp: Fix indefinite wait on sysfs requests
Lukas Wunner [Wed, 18 Mar 2020 11:33:12 +0000 (12:33 +0100)]
PCI: pciehp: Fix indefinite wait on sysfs requests

commit 3e487d2e4aa466decd226353755c9d423e8fbacc upstream.

David Hoyer reports that powering pciehp slots up or down via sysfs may
hang:  The call to wait_event() in pciehp_sysfs_enable_slot() and
_disable_slot() does not return because ctrl->ist_running remains true.

This flag, which was introduced by commit 157c1062fcd8 ("PCI: pciehp: Avoid
returning prematurely from sysfs requests"), signifies that the IRQ thread
pciehp_ist() is running.  It is set to true at the top of pciehp_ist() and
reset to false at the end.  However there are two additional return
statements in pciehp_ist() before which the commit neglected to reset the
flag to false and wake up waiters for the flag.

That omission opens up the following race when powering up the slot:

* pciehp_ist() runs because a PCI_EXP_SLTSTA_PDC event was requested
  by pciehp_sysfs_enable_slot()

* pciehp_ist() turns on slot power via the following call stack:
  pciehp_handle_presence_or_link_change() -> pciehp_enable_slot() ->
  __pciehp_enable_slot() -> board_added() -> pciehp_power_on_slot()

* after slot power is turned on, the link comes up, resulting in a
  PCI_EXP_SLTSTA_DLLSC event

* the IRQ handler pciehp_isr() stores the event in ctrl->pending_events
  and returns IRQ_WAKE_THREAD

* the IRQ thread is already woken (it's bringing up the slot), but the
  genirq code remembers to re-run the IRQ thread after it has finished
  (such that it can deal with the new event) by setting IRQTF_RUNTHREAD
  via __handle_irq_event_percpu() -> __irq_wake_thread()

* the IRQ thread removes PCI_EXP_SLTSTA_DLLSC from ctrl->pending_events
  via board_added() -> pciehp_check_link_status() in order to deal with
  presence and link flaps per commit 6c35a1ac3da6 ("PCI: pciehp:
  Tolerate initially unstable link")

* after pciehp_ist() has successfully brought up the slot, it resets
  ctrl->ist_running to false and wakes up the sysfs requester

* the genirq code re-runs pciehp_ist(), which sets ctrl->ist_running
  to true but then returns with IRQ_NONE because ctrl->pending_events
  is empty

* pciehp_sysfs_enable_slot() is finally woken but notices that
  ctrl->ist_running is true, hence continues waiting

The only way to get the hung task going again is to trigger a hotplug
event which brings down the slot, e.g. by yanking out the card.

The same race exists when powering down the slot because remove_board()
likewise clears link or presence changes in ctrl->pending_events per commit
3943af9d01e9 ("PCI: pciehp: Ignore Link State Changes after powering off a
slot") and thereby may cause a re-run of pciehp_ist() which returns with
IRQ_NONE without resetting ctrl->ist_running to false.

Fix by adding a goto label before the teardown steps at the end of
pciehp_ist() and jumping to that label from the two return statements which
currently neglect to reset the ctrl->ist_running flag.

Fixes: 157c1062fcd8 ("PCI: pciehp: Avoid returning prematurely from sysfs requests")
Link: https://lore.kernel.org/r/cca1effa488065cb055120aa01b65719094bdcb5.1584530321.git.lukas@wunner.de
Reported-by: David Hoyer <David.Hoyer@netapp.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Cc: stable@vger.kernel.org # v4.19+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoefi/x86: Add TPM related EFI tables to unencrypted mapping checks
Tom Lendacky [Fri, 28 Feb 2020 12:14:03 +0000 (13:14 +0100)]
efi/x86: Add TPM related EFI tables to unencrypted mapping checks

commit f10e80a19b07b58fc2adad7945f8313b01503bae upstream.

When booting with SME active, EFI tables must be mapped unencrypted since
they were built by UEFI in unencrypted memory. Update the list of tables
to be checked during early_memremap() processing to account for the EFI
TPM tables.

This fixes a bug where an EFI TPM log table has been created by UEFI, but
it lives in memory that has been marked as usable rather than reserved.

Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: linux-efi@vger.kernel.org
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: David Hildenbrand <david@redhat.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: <stable@vger.kernel.org> # v5.4+
Link: https://lore.kernel.org/r/4144cd813f113c20cdfa511cf59500a64e6015be.1582662842.git.thomas.lendacky@amd.com
Link: https://lore.kernel.org/r/20200228121408.9075-2-ardb@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agonvme-fc: Revert "add module to ops template to allow module references"
James Smart [Fri, 3 Apr 2020 14:33:20 +0000 (07:33 -0700)]
nvme-fc: Revert "add module to ops template to allow module references"

commit 8c5c660529209a0e324c1c1a35ce3f83d67a2aa5 upstream.

The original patch was to resolve the lldd being able to be unloaded
while being used to talk to the boot device of the system. However, the
end result of the original patch is that any driver unload while a nvme
controller is live via the lldd is now being prohibited. Given the module
reference, the module teardown routine can't be called, thus there's no
way, other than manual actions to terminate the controllers.

Fixes: 863fbae929c7 ("nvme_fc: add module to ops template to allow module references")
Cc: <stable@vger.kernel.org> # v5.4+
Signed-off-by: James Smart <jsmart2021@gmail.com>
Reviewed-by: Himanshu Madhani <himanshu.madhani@oracle.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agonvmet-tcp: fix maxh2cdata icresp parameter
Sagi Grimberg [Wed, 26 Feb 2020 00:42:27 +0000 (16:42 -0800)]
nvmet-tcp: fix maxh2cdata icresp parameter

commit 9cda34e37489244a8c8628617e24b2dbc8a8edad upstream.

MAXH2CDATA is not zero based. Also no reason to limit ourselves to
1M transfers as we can do more easily. Make this an arbitrary limit
of 16M.

Reported-by: Wenhua Liu <liuw@vmware.com>
Cc: stable@vger.kernel.org # v5.0+
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agothermal: devfreq_cooling: inline all stubs for CONFIG_DEVFREQ_THERMAL=n
Martin Blumenstingl [Fri, 3 Apr 2020 20:51:33 +0000 (22:51 +0200)]
thermal: devfreq_cooling: inline all stubs for CONFIG_DEVFREQ_THERMAL=n

commit 3f5b9959041e0db6dacbea80bb833bff5900999f upstream.

When CONFIG_DEVFREQ_THERMAL is disabled all functions except
of_devfreq_cooling_register_power() were already inlined. Also inline
the last function to avoid compile errors when multiple drivers call
of_devfreq_cooling_register_power() when CONFIG_DEVFREQ_THERMAL is not
set. Compilation failed with the following message:
  multiple definition of `of_devfreq_cooling_register_power'
(which then lists all usages of of_devfreq_cooling_register_power())

Thomas Zimmermann reported this problem [0] on a kernel config with
CONFIG_DRM_LIMA={m,y}, CONFIG_DRM_PANFROST={m,y} and
CONFIG_DEVFREQ_THERMAL=n after both, the lima and panfrost drivers
gained devfreq cooling support.

[0] https://www.spinics.net/lists/dri-devel/msg252825.html

Fixes: a76caf55e5b356 ("thermal: Add devfreq cooling")
Cc: stable@vger.kernel.org
Reported-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Link: https://lore.kernel.org/r/20200403205133.1101808-1-martin.blumenstingl@googlemail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoACPI: PM: s2idle: Refine active GPEs check
Rafael J. Wysocki [Wed, 25 Mar 2020 10:55:48 +0000 (11:55 +0100)]
ACPI: PM: s2idle: Refine active GPEs check

commit d5406284ff803a578ca503373624312770319054 upstream.

The check for any active GPEs added by commit fdde0ff8590b ("ACPI:
PM: s2idle: Prevent spurious SCIs from waking up the system") turns
out to be insufficiently precise to prevent some systems from
resuming prematurely due to a spurious EC wakeup, so refine it
by first checking if any GPEs other than the EC GPE are active
and skipping all of the SCIs coming from the EC that do not produce
any genuine wakeup events after processing.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=206629
Fixes: fdde0ff8590b ("ACPI: PM: s2idle: Prevent spurious SCIs from waking up the system")
Reported-by: Ondřej Caletka <ondrej@caletka.cz>
Tested-by: Ondřej Caletka <ondrej@caletka.cz>
Cc: 5.4+ <stable@vger.kernel.org> # 5.4+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoACPICA: Allow acpi_any_gpe_status_set() to skip one GPE
Rafael J. Wysocki [Wed, 25 Mar 2020 10:54:29 +0000 (11:54 +0100)]
ACPICA: Allow acpi_any_gpe_status_set() to skip one GPE

commit 0ce792d660bda990c675eaf14ce09594a9b85cbf upstream.

The check carried out by acpi_any_gpe_status_set() is not precise enough
for the suspend-to-idle implementation in Linux and in some cases it is
necessary make it skip one GPE (specifically, the EC GPE) from the check
to prevent a race condition leading to a premature system resume from
occurring.

For this reason, redefine acpi_any_gpe_status_set() to take the number
of a GPE to skip as an argument.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=206629
Tested-by: Ondřej Caletka <ondrej@caletka.cz>
Cc: 5.4+ <stable@vger.kernel.org> # 5.4+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoacpi/x86: ignore unspecified bit positions in the ACPI global lock field
Jan Engelhardt [Thu, 5 Mar 2020 12:24:25 +0000 (13:24 +0100)]
acpi/x86: ignore unspecified bit positions in the ACPI global lock field

commit ecb9c790999fd6c5af0f44783bd0217f0b89ec2b upstream.

The value in "new" is constructed from "old" such that all bits defined
as reserved by the ACPI spec[1] are left untouched. But if those bits
do not happen to be all zero, "new < 3" will not evaluate to true.

The firmware of the laptop(s) Medion MD63490 / Akoya P15648 comes with
garbage inside the "FACS" ACPI table. The starting value is
old=0x4944454d, therefore new=0x4944454e, which is >= 3. Mask off
the reserved bits.

[1] https://uefi.org/sites/default/files/resources/ACPI_6_2.pdf

Link: https://bugzilla.kernel.org/show_bug.cgi?id=206553
Cc: All applicable <stable@vger.kernel.org>
Signed-off-by: Jan Engelhardt <jengelh@inai.de>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agoseccomp: Add missing compat_ioctl for notify
Sven Schnelle [Tue, 10 Mar 2020 12:33:32 +0000 (13:33 +0100)]
seccomp: Add missing compat_ioctl for notify

commit 3db81afd99494a33f1c3839103f0429c8f30cb9d upstream.

Executing the seccomp_bpf testsuite under a 64-bit kernel with 32-bit
userland (both s390 and x86) doesn't work because there's no compat_ioctl
handler defined. Add the handler.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Fixes: 6a21cc50f0c7 ("seccomp: add a return code to trap to userspace")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20200310123332.42255-1-svens@linux.ibm.com
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
4 years agomedia: ti-vpe: cal: fix a kernel oops when unloading module
Benoit Parrot [Fri, 6 Mar 2020 13:08:39 +0000 (14:08 +0100)]
media: ti-vpe: cal: fix a kernel oops when unloading module

commit 80264809ea0a3fd2ee8251f31a9eb85d2c3fc77e upstream.

After the switch to use v4l2_async_notifier_add_subdev() and
v4l2_async_notifier_cleanup(), unloading the ti_cal module would cause a
kernel oops.

This was root cause to the fact that v4l2_async_notifier_cleanup() tries
to kfree the asd pointer passed into v4l2_async_notifier_add_subdev().

In our case the asd reference was from a statically allocated struct.
So in effect v4l2_async_notifier_cleanup() was trying to free a pointer
that was not kalloc.

So here we switch to using a kzalloc struct instead of a static one.
To achieve this we re-order some of the calls to prevent asd allocation
from leaking.

Fixes: d079f94c9046 ("media: platform: Switch to v4l2_async_notifier_add_subdev")
Cc: stable@vger.kernel.org
Signed-off-by: Benoit Parrot <bparrot@ti.com>
Reviewed-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>