Jens Axboe [Sun, 23 Aug 2020 17:00:37 +0000 (11:00 -0600)]
io_uring: don't recurse on tsk->sighand->siglock with signalfd
If an application is doing reads on signalfd, and we arm the poll handler
because there's no data available, then the wakeup can recurse on the
tasks sighand->siglock as the signal delivery from task_work_add() will
use TWA_SIGNAL and that attempts to lock it again.
We can detect the signalfd case pretty easily by comparing the poll->head
wait_queue_head_t with the target task signalfd wait queue. Just use
normal task wakeup for this case.
Kai-Heng Feng [Fri, 21 Aug 2020 09:15:48 +0000 (12:15 +0300)]
xhci: Do warm-reset when both CAS and XDEV_RESUME are set
Sometimes re-plugging a USB device during system sleep renders the device
useless:
[ 173.418345] xhci_hcd 0000:00:14.0: Get port status 2-4 read: 0x14203e2, return 0x10262
...
[ 176.496485] usb 2-4: Waited 2000ms for CONNECT
[ 176.496781] usb usb2-port4: status 0000.0262 after resume, -19
[ 176.497103] usb 2-4: can't resume, status -19
[ 176.497438] usb usb2-port4: logical disconnect
Because PLS equals to XDEV_RESUME, xHCI driver reports U3 to usbcore,
despite of CAS bit is flagged.
So proritize CAS over XDEV_RESUME to let usbcore handle warm-reset for
the port.
Merge tag 'misc-habanalabs-fixes-2020-08-22' of git://people.freedesktop.org/~gabbayo/linux into char-misc-linus
Oded writes:
This tag contains the following bug fixes for 5.9-rc2/3:
- Correct cleanup of PCI bar mapping in case of failure during
initialization.
- Several security fixes:
- Validating user addresses before mapping them
- Validating packet id (from user) before using it as index for array.
- Validating F/W file size before coping it.
- Prevent possible overflow when validating address from user in
profiler.
- Validate queue index (from user) before using it as index for array.
- Check for correct vmalloc return code
- Fix memory corruption in debugfs entry
- Fix a loop in gaudi_extract_ecc_info()
- Fix the set clock gating function in gaudi code
- Set maximum power to F/W according to the card type
- Cix incorrect check on failed workqueue create
- Correctly report error when configuring the PCI controller
* tag 'misc-habanalabs-fixes-2020-08-22' of git://people.freedesktop.org/~gabbayo/linux:
habanalabs: correctly report inbound pci region cfg error
habanalabs: check correct vmalloc return code
habanalabs: validate FW file size
habanalabs: fix incorrect check on failed workqueue create
habanalabs: set max power according to card type
habanalabs: proper handling of alloc size in coresight
habanalabs: set clock gating according to mask
habanalabs: verify user input in cs_ioctl_signal_wait
habanalabs: Fix a loop in gaudi_extract_ecc_info()
habanalabs: Fix memory corruption in debugfs
habanalabs: validate packet id during CB parse
habanalabs: Validate user address before mapping
habanalabs: unmap PCI bars upon iATU failure
Linus Torvalds [Sun, 23 Aug 2020 00:11:38 +0000 (17:11 -0700)]
Merge branch 'work.epoll' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull epoll fixes from Al Viro:
"Fix reference counting and clean up exit paths"
* 'work.epoll' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
do_epoll_ctl(): clean the failure exits up a bit
epoll: Keep a reference on files added to the check list
Marc Zyngier [Wed, 19 Aug 2020 16:12:17 +0000 (17:12 +0100)]
epoll: Keep a reference on files added to the check list
When adding a new fd to an epoll, and that this new fd is an
epoll fd itself, we recursively scan the fds attached to it
to detect cycles, and add non-epool files to a "check list"
that gets subsequently parsed.
However, this check list isn't completely safe when deletions
can happen concurrently. To sidestep the issue, make sure that
a struct file placed on the check list sees its f_count increased,
ensuring that a concurrent deletion won't result in the file
disapearing from under our feet.
Currently the nexthop code will use an empty NHA_GROUP attribute, but it
requires at least 1 entry in order to function properly. Otherwise we
end up derefencing null or random pointers all over the place due to not
having any nh_grp_entry members allocated, nexthop code relies on having at
least the first member present. Empty NHA_GROUP doesn't make any sense so
just disallow it.
Also add a WARN_ON for any future users of nexthop_create_group().
Jonathan Marek [Thu, 13 Aug 2020 18:54:48 +0000 (14:54 -0400)]
drm/msm/a6xx: fix frequency not always being restored on GMU resume
The patch reorganizing the set_freq function made it so the gmu resume
doesn't always set the frequency, because a6xx_gmu_set_freq() exits early
when the frequency hasn't been changed. Note this always happens when
resuming GMU after recovering from a hang.
Use a simple workaround to prevent this from happening.
Fixes: 1f60d11423db ("drm: msm: a6xx: send opp instead of a frequency") Signed-off-by: Jonathan Marek <[email protected]> Signed-off-by: Rob Clark <[email protected]>
Rob Clark [Tue, 11 Aug 2020 23:36:57 +0000 (16:36 -0700)]
drm/msm/a6xx: add module param to enable debugbus snapshot
For production devices, the debugbus sections will typically be fused
off and empty in the gpu device coredump. But since this may contain
data like cache contents, don't capture it by default.
Linus Torvalds [Sat, 22 Aug 2020 17:22:44 +0000 (10:22 -0700)]
Merge tag 'kbuild-fixes-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- move -Wsign-compare warning from W=2 to W=3
- fix the keyword _restrict to __restrict in genksyms
- fix more bugs in qconf
* tag 'kbuild-fixes-v5.9' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: qconf: replace deprecated QString::sprintf() with QTextStream
kconfig: qconf: remove redundant help in the info view
kconfig: qconf: remove qInfo() to get back Qt4 support
kconfig: qconf: remove unused colNr
kconfig: qconf: fix the popup menu in the ConfigInfoView window
kconfig: qconf: fix signal connection to invalid slots
genksyms: keywords: Use __restrict not _restrict
kbuild: remove redundant patterns in filter/filter-out
extract-cert: add static to local data
Makefile.extrawarn: Move sign-compare from W=2 to W=3
Linus Torvalds [Sat, 22 Aug 2020 17:17:36 +0000 (10:17 -0700)]
Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Catalin Marinas:
- Allow booting of late secondary CPUs affected by erratum 1418040
(currently they are parked if none of the early CPUs are affected by
this erratum).
- Add the 32-bit vdso Makefile to the vdso_install rule so that 'make
vdso_install' installs the 32-bit compat vdso when it is compiled.
- Print a warning that untrusted guests without a CPU erratum
workaround (Cortex-A57 832075) may deadlock the affected system.
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
ARM64: vdso32: Install vdso32 from vdso_install
KVM: arm64: Print warning when cpu erratum can cause guests to deadlock
arm64: Allow booting of late CPUs affected by erratum 1418040
arm64: Move handling of erratum 1418040 into C code
Linus Torvalds [Sat, 22 Aug 2020 17:03:05 +0000 (10:03 -0700)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
- PAE and PKU bugfixes for x86
- selftests fix for new binutils
- MMU notifier fix for arm64
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: arm64: Only reschedule if MMU_NOTIFIER_RANGE_BLOCKABLE is not set
KVM: Pass MMU notifier range flags to kvm_unmap_hva_range()
kvm: x86: Toggling CR4.PKE does not load PDPTEs in PAE mode
kvm: x86: Toggling CR4.SMAP does not load PDPTEs in PAE mode
KVM: x86: fix access code passed to gva_to_gpa
selftests: kvm: Use a shorter encoding to clear RAX
Linus Torvalds [Sat, 22 Aug 2020 16:56:42 +0000 (09:56 -0700)]
Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"23 fixes in 5 drivers (qla2xxx, ufs, scsi_debug, fcoe, zfcp). The bulk
of the changes are in qla2xxx and ufs and all are mostly small and
definitely don't impact the core"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (23 commits)
Revert "scsi: qla2xxx: Disable T10-DIF feature with FC-NVMe during probe"
Revert "scsi: qla2xxx: Fix crash on qla2x00_mailbox_command"
scsi: qla2xxx: Fix null pointer access during disconnect from subsystem
scsi: qla2xxx: Check if FW supports MQ before enabling
scsi: qla2xxx: Fix WARN_ON in qla_nvme_register_hba
scsi: qla2xxx: Allow ql2xextended_error_logging special value 1 to be set anytime
scsi: qla2xxx: Reduce noisy debug message
scsi: qla2xxx: Fix login timeout
scsi: qla2xxx: Indicate correct supported speeds for Mezz card
scsi: qla2xxx: Flush I/O on zone disable
scsi: qla2xxx: Flush all sessions on zone disable
scsi: qla2xxx: Use MBX_TOV_SECONDS for mailbox command timeout values
scsi: scsi_debug: Fix scp is NULL errors
scsi: zfcp: Fix use-after-free in request timeout handlers
scsi: ufs: No need to send Abort Task if the task in DB was cleared
scsi: ufs: Clean up completed request without interrupt notification
scsi: ufs: Improve interrupt handling for shared interrupts
scsi: ufs: Fix interrupt error message for shared interrupts
scsi: ufs-pci: Add quirk for broken auto-hibernate for Intel EHL
scsi: ufs-mediatek: Fix incorrect time to wait link status
...
Linus Torvalds [Sat, 22 Aug 2020 16:31:11 +0000 (09:31 -0700)]
Merge tag 'devicetree-fixes-for-5.9-2' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux
Pull devicetree fixes from Rob Herring:
"Another set of DT fixes:
- restore range parsing error check
- workaround PCI range parsing with missing 'device_type' now
required
- correct description of 'phy-connection-type'
- fix erroneous matching on 'snps,dw-pcie' by 'intel,lgm-pcie' schema
- a couple of grammar and whitespace fixes
- update Shawn Guo's email"
* tag 'devicetree-fixes-for-5.9-2' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux:
dt-bindings: vendor-prefixes: Remove trailing whitespace
dt-bindings: net: correct description of phy-connection-type
dt-bindings: PCI: intel,lgm-pcie: Fix matching on all snps,dw-pcie instances
of: address: Work around missing device_type property in pcie nodes
dt: writing-schema: Miscellaneous grammar fixes
dt-bindings: Use Shawn Guo's preferred e-mail for i.MX bindings
of/address: check for invalid range.cpu_addr
Ofir Bitton [Tue, 11 Aug 2020 05:11:49 +0000 (08:11 +0300)]
habanalabs: correctly report inbound pci region cfg error
During inbound iATU configuration we can get errors while
configuring PCI registers, there is a certain scenario in which these
errors are not reflected and driver is loaded with wrong configuration.
Colin Ian King [Thu, 30 Jul 2020 08:20:22 +0000 (09:20 +0100)]
habanalabs: fix incorrect check on failed workqueue create
The null check on a failed workqueue create is currently null checking
hdev->cq_wq rather than the pointer hdev->cq_wq[i] and so the test
will never be true on a failed workqueue create. Fix this by checking
hdev->cq_wq[i].
Addresses-Coverity: ("Dereference before null check") Fixes: 5574cb2194b1 ("habanalabs: Assign each CQ with its own work queue") Signed-off-by: Colin Ian King <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
Oded Gabbay [Sat, 8 Aug 2020 20:34:47 +0000 (23:34 +0300)]
habanalabs: set max power according to card type
In Gaudi, the default max power setting is different between PCI and PMC
cards. Therefore, the driver need to set the default after knowing what is
the card type.
The current code has a bug where it limits the maximum power of the PMC
card to 200W after a reset occurs.
During command buffer parsing, driver extracts packet id
from user buffer. Driver must validate this packet id, since it is
being used in order to extract information from internal structures.
In case the driver fails to configure the PCI controller iATU, it needs to
unmap the PCI bars before exiting so if the driver is removed, the bars
won't be left mapped.
Rob Clark [Tue, 18 Aug 2020 16:31:19 +0000 (09:31 -0700)]
drm/msm: enable vblank during atomic commits
This has roughly the same effect as drm_atomic_helper_wait_for_vblanks(),
basically just ensuring that vblank accounting is enabled so that we get
valid timestamp/seqn on pageflip events.
nvme: just check the status code type in nvme_is_path_error
Check the SCT sub-field for a path related status instead of enumerating
invididual status code. As of NVMe 1.4 this adds "Internal Path Error"
and "Controller Pathing Error" to the list, but it also future proofs for
additional status codes added to the category.
Lift all the code to decide the dispostition of a completed command
from nvme_complete_rq and nvme_failover_req into a new helper, which
returns an emum of the potential actions. nvme_complete_rq then
just switches on those and calls the proper helper for the action.
nvme_end_request is a bit misnamed, as it wraps around the
blk_mq_complete_* API. It's semantics also are non-trivial, so give it
a more descriptive name and add a comment explaining the semantics.
Keith Busch [Fri, 7 Aug 2020 16:32:35 +0000 (09:32 -0700)]
nvme: skip noiob for zoned devices
Zoned block devices reuse the chunk_sectors queue limit to define zone
boundaries. If a such a device happens to also report an optimal
boundary, do not use that to define the chunk_sectors as that may
intermittently interfere with io splitting and zone size queries.
All operations are based on the controller, not the host page size.
Switch the dma pool to use the controller page size as well to avoid
massive overallocations on large page size systems.
John Garry [Fri, 14 Aug 2020 15:34:25 +0000 (23:34 +0800)]
nvme-pci: Use u32 for nvme_dev.q_depth and nvme_queue.q_depth
Recently nvme_dev.q_depth was changed from an int to u16 type.
This falls over for the queue depth calculation in nvme_pci_enable(),
where NVME_CAP_MQES(dev->ctrl.cap) + 1 may overflow as a u16, as
NVME_CAP_MQES() is a 16b number also. That happens for me, and this is the
result:
Also use u32 for nvme_dev.q_depth, as we assign this value from
nvme_dev.q_depth, and nvme_dev.q_depth will possibly hold 65536 - this
avoids the same crash as above.
Logan Gunthorpe [Wed, 12 Aug 2020 23:24:44 +0000 (17:24 -0600)]
nvme: Use spin_lock_irq() when taking the ctrl->lock
When locking the ctrl->lock spinlock IRQs need to be disabled to avoid a
dead lock. The new spin_lock() calls recently added produce the
following lockdep warning when running the blktest nvme/003:
================================
WARNING: inconsistent lock state
--------------------------------
inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
ksoftirqd/2/22 [HC0[0]:SC1[1]:HE0:SE0] takes: ffff888276a8c4c0 (&ctrl->lock){+.?.}-{2:2}, at: nvme_keep_alive_end_io+0x50/0xc0
{SOFTIRQ-ON-W} state was registered at:
lock_acquire+0x164/0x500
_raw_spin_lock+0x28/0x40
nvme_get_effects_log+0x37/0x1c0
nvme_init_identify+0x9e4/0x14f0
nvme_reset_work+0xadd/0x2360
process_one_work+0x66b/0xb70
worker_thread+0x6e/0x6c0
kthread+0x1e7/0x210
ret_from_fork+0x22/0x30
irq event stamp: 1449221
hardirqs last enabled at (1449220): [<ffffffff81c58e69>] ktime_get+0xf9/0x140
hardirqs last disabled at (1449221): [<ffffffff83129665>] _raw_spin_lock_irqsave+0x25/0x60
softirqs last enabled at (1449210): [<ffffffff83400447>] __do_softirq+0x447/0x595
softirqs last disabled at (1449215): [<ffffffff81b489b5>] run_ksoftirqd+0x35/0x50
other info that might help us debug this:
Possible unsafe locking scenario:
Instead of calling blk_put_request() which calls blk_mq_free_request(),
call blk_mq_free_request() directly for NVMeOF passthru. This is to
mainly avoid an extra function call in the completion path
nvmet_passthru_req_done().
In the existing NVMeOF Passthru core command handling on failure of
nvme_alloc_request() it errors out with rq value set to NULL. In the
error handling path it calls blk_put_request() without checking if
rq is set to NULL or not which produces following Oops:-
We fix this Oops by adding a new goto label out_put_req and reordering
the blk_put_request call to avoid calling blk_put_request() with rq
value is set to NULL. Here we also update the rest of the code
accordingly.
In the current implementation before submitting the passthru cmd we
may come across error e.g. getting ns from passthru controller,
allocating a request from passthru controller, etc. For all the failure
cases it only uses single goto label fail_out.
In the target code, we follow the pattern to have a separate label for
each error out the case when setting up multiple things before the actual
action.
This patch follows the same pattern and renames generic fail_out label
to out_put_ns and updates the error out cases in the
nvmet_passthru_execute_cmd() where it is needed.
Martin Wilck [Thu, 6 Aug 2020 13:19:31 +0000 (15:19 +0200)]
nvme: multipath: round-robin: fix single non-optimized path case
If there's only one usable, non-optimized path, nvme_round_robin_path()
returns NULL, which is wrong. Fix it by falling back to "old", like in
the single optimized path case. Also, if the active path isn't changed,
there's no need to re-assign the pointer.
Yufen Yu [Tue, 11 Aug 2020 02:21:16 +0000 (22:21 -0400)]
blkcg: fix memleak for iolatency
Normally, blkcg_iolatency_exit() will free related memory in iolatency
when cleanup queue. But if blk_throtl_init() return error and queue init
fail, blkcg_iolatency_exit() will not do that for us. Then it cause
memory leak.
MAINTAINERS: Add missing header files to BLOCK LAYER section
The various <linux/blk*.h> header files are part of the Block Layer.
Add them to the corresponding section in the MAINTAINERS file, so
scripts/get_maintainer.pl will pick them up.
Keith Busch [Thu, 6 Aug 2020 21:58:37 +0000 (14:58 -0700)]
block: fix get_max_io_size()
A previous commit aligning splits to physical block sizes inadvertently
modified one return case such that that it now returns 0 length splits
when the number of sectors doesn't exceed the physical offset. This
later hits a BUG in bio_split(). Restore the previous working behavior.
Ming Lei [Tue, 18 Aug 2020 09:07:28 +0000 (17:07 +0800)]
blk-mq: insert request not through ->queue_rq into sw/scheduler queue
c616cbee97ae ("blk-mq: punt failed direct issue to dispatch list") supposed
to add request which has been through ->queue_rq() to the hw queue dispatch
list, however it adds request running out of budget or driver tag to hw queue
too. This way basically bypasses request merge, and causes too many request
dispatched to LLD, and system% is unnecessary increased.
Fixes this issue by adding request not through ->queue_rq into sw/scheduler
queue, and this way is safe because no ->queue_rq is called on this request
yet.
High %system can be observed on Azure storvsc device, and even soft lock
is observed. This patch reduces %system during heavy sequential IO,
meantime decreases soft lockup risk.
block/rnbd: Ensure err is always initialized in process_rdma
Clang warns:
drivers/block/rnbd/rnbd-srv.c:150:6: warning: variable 'err' is used
uninitialized whenever 'if' condition is true
[-Wsometimes-uninitialized]
if (IS_ERR(bio)) {
^~~~~~~~~~~
drivers/block/rnbd/rnbd-srv.c:177:9: note: uninitialized use occurs here
return err;
^~~
drivers/block/rnbd/rnbd-srv.c:150:2: note: remove the 'if' if its
condition is always false
if (IS_ERR(bio)) {
^~~~~~~~~~~~~~~~~~
drivers/block/rnbd/rnbd-srv.c:126:9: note: initialize the variable 'err'
to silence this warning
int err;
^
= 0
1 warning generated.
err is indeed uninitialized when this statement is taken. Ensure that it
is assigned the error value of bio before jumping to the error handling
label.
Will Deacon [Tue, 11 Aug 2020 10:27:25 +0000 (11:27 +0100)]
KVM: arm64: Only reschedule if MMU_NOTIFIER_RANGE_BLOCKABLE is not set
When an MMU notifier call results in unmapping a range that spans multiple
PGDs, we end up calling into cond_resched_lock() when crossing a PGD boundary,
since this avoids running into RCU stalls during VM teardown. Unfortunately,
if the VM is destroyed as a result of OOM, then blocking is not permitted
and the call to the scheduler triggers the following BUG():
Will Deacon [Tue, 11 Aug 2020 10:27:24 +0000 (11:27 +0100)]
KVM: Pass MMU notifier range flags to kvm_unmap_hva_range()
The 'flags' field of 'struct mmu_notifier_range' is used to indicate
whether invalidate_range_{start,end}() are permitted to block. In the
case of kvm_mmu_notifier_invalidate_range_start(), this field is not
forwarded on to the architecture-specific implementation of
kvm_unmap_hva_range() and therefore the backend cannot sensibly decide
whether or not to block.
Add an extra 'flags' parameter to kvm_unmap_hva_range() so that
architectures are aware as to whether or not they are permitted to block.
Linus Torvalds [Fri, 21 Aug 2020 21:59:16 +0000 (14:59 -0700)]
Merge tag 'io_uring-5.9-2020-08-21' of git://git.kernel.dk/linux-block
Pull io_uring fixes from Jens Axboe:
- Make sure the head link cancelation includes async work
- Get rid of kiocb_wait_page_queue_init(), makes no sense to have it as
a separate function since you moved it into io_uring itself
- io_import_iovec cleanups (Pavel, me)
- Use system_unbound_wq for ring exit work, to avoid spawning tons of
these if we have tons of rings exiting at the same time
- Fix req->flags overflow flag manipulation (Pavel)
* tag 'io_uring-5.9-2020-08-21' of git://git.kernel.dk/linux-block:
io_uring: kill extra iovec=NULL in import_iovec()
io_uring: comment on kfree(iovec) checks
io_uring: fix racy req->flags modification
io_uring: use system_unbound_wq for ring exit work
io_uring: cleanup io_import_iovec() of pre-mapped request
io_uring: get rid of kiocb_wait_page_queue_init()
io_uring: find and cancel head link async work on files exit
* emailed patches from Andrew Morton <[email protected]>:
mm, page_alloc: fix core hung in free_pcppages_bulk()
mm: include CMA pages in lowmem_reserve at boot
squashfs: avoid bio_alloc() failure with 1Mbyte blocks
uprobes: __replace_page() avoid BUG in munlock_vma_page()
kernel/relay.c: fix memleak on destroy relay channel
romfs: fix uninitialized memory leak in romfs_dev_read()
mm/rodata_test.c: fix missing function declaration
mm/vunmap: add cond_resched() in vunmap_pmd_range
khugepaged: adjust VM_BUG_ON_MM() in __khugepaged_enter()
hugetlb_cgroup: convert comma to semicolon
mailmap: add Andi Kleen
Linus Torvalds [Fri, 21 Aug 2020 19:32:42 +0000 (12:32 -0700)]
Merge tag 'riscv-for-linus-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux
Pull RISC-V fixes from Palmer Dabbelt:
- The CLINT driver has been split in two: one to handle the M-mode
CLINT (memory mapped and used on NOMMU systems) and one to handle the
S-mode CLINT (via SBI).
- The addition of SiFive's drivers to rv32_defconfig
* tag 'riscv-for-linus-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
riscv: Add SiFive drivers to rv32_defconfig
dt-bindings: timer: Add CLINT bindings
RISC-V: Remove CLINT related code from timer and arch
clocksource/drivers: Add CLINT timer driver
RISC-V: Add mechanism to provide custom IPI operations
Linus Torvalds [Fri, 21 Aug 2020 19:28:33 +0000 (12:28 -0700)]
Merge tag 'for-linus-5.9-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fixes from Juergen Gross:
"One build fix and a minor fix for suppressing a useless warning when
booting a Xen dom0 via UEFI"
* tag 'for-linus-5.9-rc2-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
Fix build error when CONFIG_ACPI is not set/enabled:
efi: avoid error message when booting under Xen
Linus Torvalds [Fri, 21 Aug 2020 19:26:58 +0000 (12:26 -0700)]
Merge tag 'pm-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
"These fix a few issues in the operating performance points (OPP)
framework.
Specifics:
- Fix re-enabling of resources in dev_pm_opp_set_rate() (Rajendra
Nayak)
- Fix OPP table reference counting in error paths (Stephen Boyd)"
* tag 'pm-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
opp: Enable resources again if they were disabled earlier
opp: Put opp table in dev_pm_opp_set_rate() if _set_opp_bw() fails
opp: Put opp table in dev_pm_opp_set_rate() for empty tables
Tom Rix [Fri, 21 Aug 2020 13:56:00 +0000 (06:56 -0700)]
net: dsa: b53: check for timeout
clang static analysis reports this problem
b53_common.c:1583:13: warning: The left expression of the compound
assignment is an uninitialized value. The computed value will
also be garbage
ent.port &= ~BIT(port);
~~~~~~~~ ^
ent is set by a successful call to b53_arl_read(). Unsuccessful
calls are caught by an switch statement handling specific returns.
b32_arl_read() calls b53_arl_op_wait() which fails with the
unhandled -ETIMEDOUT.
So add -ETIMEDOUT to the switch statement. Because
b53_arl_op_wait() already prints out a message, do not add another
one.
Fixes: 1da6df85c6fb ("net: dsa: b53: Implement ARL add/del/dump operations") Signed-off-by: Tom Rix <[email protected]> Acked-by: Florian Fainelli <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Tom Rix [Thu, 20 Aug 2020 13:19:32 +0000 (06:19 -0700)]
hwmon: (applesmc) check status earlier.
clang static analysis reports this representative problem
applesmc.c:758:10: warning: 1st function call argument is an
uninitialized value
left = be16_to_cpu(*(__be16 *)(buffer + 6)) >> 2;
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
buffer is filled by the earlier call
ret = applesmc_read_key(LIGHT_SENSOR_LEFT_KEY, ...
This problem is reported because a goto skips the status check.
Other similar problems use data from applesmc_read_key before checking
the status. So move the checks to before the use.
Jason Baron [Fri, 21 Aug 2020 16:20:14 +0000 (12:20 -0400)]
hwmon: (nct7904) Correct divide by 0
We hit a kernel panic due to a divide by 0 in nct7904_read_fan() for
the hwmon_fan_min case. Extend the check to hwmon_fan_input case as well
for safety.
Heikki Krogerus [Fri, 21 Aug 2020 10:53:42 +0000 (13:53 +0300)]
device property: Fix the secondary firmware node handling in set_primary_fwnode()
When the primary firmware node pointer is removed from a
device (set to NULL) the secondary firmware node pointer,
when it exists, is made the primary node for the device.
However, the secondary firmware node pointer of the original
primary firmware node is never cleared (set to NULL).
To avoid situation where the secondary firmware node pointer
is pointing to a non-existing object, clearing it properly
when the primary node is removed from a device in
set_primary_fwnode().
Fixes: 97badf873ab6 ("device property: Make it possible to use secondary firmware nodes") Cc: All applicable <[email protected]> Signed-off-by: Heikki Krogerus <[email protected]> Signed-off-by: Rafael J. Wysocki <[email protected]>
Ard Biesheuvel [Tue, 18 Aug 2020 09:13:53 +0000 (11:13 +0200)]
ACPI: ioremap: avoid redundant rounding to OS page size
The arm64 implementation of acpi_os_ioremap() was recently updated to
tighten the checks around which parts of memory are permitted to be
mapped by ACPI code, which generally only needs access to memory regions
that are statically described by firmware, and any attempts to access
memory that is in active use by the OS is generally a bug or a hacking
attempt. This tightening is based on the EFI memory map, which describes
all memory in the system.
The AArch64 architecture permits page sizes of 16k and 64k in addition
to the EFI default, which is 4k, which means that the EFI memory map may
describe regions that cannot be mapped seamlessly if the OS page size is
greater than 4k. This is usually not a problem, given that the EFI spec
does not permit memory regions requiring different memory attributes to
share a 64k page frame, and so the usual rounding to page size performed
by ioremap() is sufficient to deal with this. However, this rounding does
complicate our EFI memory map permission check, due to the loss of
information that occurs when several small regions share a single 64k
page frame (where rounding each of them will result in the same 64k
single page region).
However, due to the fact that the region check occurs *before* the call
to ioremap() where the necessary rounding is performed, we can deal
with this issue simply by removing the redundant rounding performed by
acpi_os_map_iomem(), as it appears to be the only place where the
arguments to a call to acpi_os_ioremap() are rounded up. So omit the
rounding in the call, and instead, apply the necessary masking when
assigning the map->virt member.
Fixes: 1583052d111f ("arm64/acpi: disallow AML memory opregions to access kernel memory") Signed-off-by: Ard Biesheuvel <[email protected]> Signed-off-by: Rafael J. Wysocki <[email protected]>
Furquan Shaikh [Wed, 19 Aug 2020 17:16:56 +0000 (10:16 -0700)]
ACPI: SoC: APD: Check return value of acpi_dev_get_property()
`fch_misc_setup()` uses `acpi_dev_get_property()` to read the value of
"is-rv" passed in by BIOS in ACPI tables. However, not all BIOSes
might pass in this property and hence it is important to first check
the return value of `acpi_dev_get_property()` before referencing the
object filled by it.
Sumit Gupta [Tue, 11 Aug 2020 19:43:17 +0000 (01:13 +0530)]
cpufreq: replace cpu_logical_map() with read_cpuid_mpir()
Commit eaecca9e7710 ("arm64: Fix __cpu_logical_map undefined issue")
fixes the issue with building tegra194 cpufreq driver as module. But
the fix might cause problem while supporting physical CPU hotplug[1].
This patch fixes the original problem by avoiding use of cpu_logical_map().
Instead calling read_cpuid_mpidr() to get MPIDR on target CPU.
Linus Torvalds [Fri, 21 Aug 2020 18:03:38 +0000 (11:03 -0700)]
Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
Pull ext4 updates from Ted Ts'o:
"Improvements to ext4's block allocator performance for very large file
systems, especially when the file system or files which are highly
fragmented. There is a new mount option, prefetch_block_bitmaps which
will pull in the block bitmaps and set up the in-memory buddy bitmaps
when the file system is initially mounted.
Beyond that, a lot of bug fixes and cleanups. In particular, a number
of changes to make ext4 more robust in the face of write errors or
file system corruptions"
* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (46 commits)
ext4: limit the length of per-inode prealloc list
ext4: reorganize if statement of ext4_mb_release_context()
ext4: add mb_debug logging when there are lost chunks
ext4: Fix comment typo "the the".
jbd2: clean up checksum verification in do_one_pass()
ext4: change to use fallthrough macro
ext4: remove unused parameter of ext4_generic_delete_entry function
mballoc: replace seq_printf with seq_puts
ext4: optimize the implementation of ext4_mb_good_group()
ext4: delete invalid comments near ext4_mb_check_limits()
ext4: fix typos in ext4_mb_regular_allocator() comment
ext4: fix checking of directory entry validity for inline directories
fs: prevent BUG_ON in submit_bh_wbc()
ext4: correctly restore system zone info when remount fails
ext4: handle add_system_zone() failure in ext4_setup_system_zone()
ext4: fold ext4_data_block_valid_rcu() into the caller
ext4: check journal inode extents more carefully
ext4: don't allow overlapping system zones
ext4: handle error of ext4_setup_system_zone() on remount
ext4: delete the invalid BUGON in ext4_mb_load_buddy_gfp()
...
David Howells [Fri, 21 Aug 2020 09:15:12 +0000 (10:15 +0100)]
afs: Fix NULL deref in afs_dynroot_depopulate()
If an error occurs during the construction of an afs superblock, it's
possible that an error occurs after a superblock is created, but before
we've created the root dentry. If the superblock has a dynamic root
(ie. what's normally mounted on /afs), the afs_kill_super() will call
afs_dynroot_depopulate() to unpin any created dentries - but this will
oops if the root hasn't been created yet.
Fix this by skipping that bit of code if there is no root dentry.
Linus Torvalds [Fri, 21 Aug 2020 17:14:16 +0000 (10:14 -0700)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull rdma fixes from Jason Gunthorpe:
"One regression from 5.8 and a few bugs from earlier kernels:
- Various spelling corrections in kernel prints
- Bug fixes in hfi1 and bntx_re
- Revert a 5.8 patch in hns
- Batch update for Mellanox and Cumulus maintainers emails"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
MAINTAINERS: Update Mellanox and Cumulus Network addresses to new domain
Revert "RDMA/hns: Reserve one sge in order to avoid local length error"
RDMA/hfi1: Correct an interlock issue for TID RDMA WRITE request
RDMA/bnxt_re: Do not add user qps to flushlist
RDMA/core: Fix spelling mistake "Could't" -> "Couldn't"
RDMA/usnic: Fix spelling mistake "transistion" -> "transition"
RDMA/hns: Fix spelling mistake "epmty" -> "empty"
Linus Torvalds [Fri, 21 Aug 2020 17:07:54 +0000 (10:07 -0700)]
Merge tag 'sound-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"A collection of small fixes over several drivers, but all are driver-
specific and nothing looks scary.
Slightly large changes are seen in ASoC qcom driver for the bugs that
were revealed by the recent ASoC core change to report the invalid
register access errors. Also ASoC fsl got a slight intensive change
for the distortion fix.
Others are only trivial fixes or device-specific quirks"
* tag 'sound-5.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound: (25 commits)
ALSA: hda: avoid reset of sdo_limit
ALSA: hda/realtek: Add quirk for Samsung Galaxy Book Ion
ALSA: usb-audio: ignore broken processing/extension unit
ASoC: intel: Fix memleak in sst_media_open
ASoC: wm8994: Avoid attempts to read unreadable registers
ASoC: msm8916-wcd-analog: fix register Interrupt offset
ASoC: wm8994: Prevent access to invalid VU register bits on WM1811
ALSA: hda/realtek: Add model alc298-samsung-headphone
ALSA: usb-audio: Update documentation comment for MS2109 quirk
ALSA: isa: fix spelling mistakes in the comments
ALSA: usb-audio: Add capture support for Saffire 6 (USB 1.1)
ALSA: hda/realtek: Add quirk for Samsung Galaxy Flex Book
ASoC: q6routing: add dummy register read/write function
ASoC: q6afe-dai: mark all widgets registers as SND_SOC_NOPM
ASoC: Make soc_component_read() returning an error code again
ASoC: amd: Replacing component->name with codec_dai->name.
ASoC: fsl: Fix unused variable warning
ASoC: tegra: tegra210_i2s: Fix compile warning with CONFIG_PM=n
ASoC: tegra: tegra210_dmic: Fix compile warning with CONFIG_PM=n
ASoC: tegra: tegra210_ahub: Fix compile warning with CONFIG_PM=n
...
Linus Torvalds [Fri, 21 Aug 2020 17:02:44 +0000 (10:02 -0700)]
Merge tag 'drm-fixes-2020-08-21' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"Regular fixes pull for rc2. Usual rc2 doesn't seem too busy, mainly
i915 and amdgpu. I'd expect the usual uptick for rc3.
amdgpu:
- Fix allocation size
- SR-IOV fixes
- Vega20 SMU feature state caching fix
- Fix custom pptable handling
- Arcturus golden settings update
- Several display fixes
- Fixes for Navy Flounder
- Misc display fixes
- RAS fix
amdkfd:
- SDMA fix for renoir
i915:
- Fix device parameter usage for selftest mock i915 device
- Fix LPSP capability debugfs NULL dereference
- Fix buddy register pagemask table
- Fix intel_atomic_check() non-negative return value
- Fix selftests passing a random 0 into ilog2()
- Fix TGL power well enable/disable ordering
- Switch to PMU module refcounting
- GVT fixes
virtio:
- Add missing dma_fence_put() in virtio_gpu_execbuffer_ioctl()
- Fix memory leak in virtio_gpu_cleanup_object()"
* tag 'drm-fixes-2020-08-21' of git://anongit.freedesktop.org/drm/drm: (34 commits)
Revert "drm/amdgpu: disable gfxoff for navy_flounder"
drm/i915/tgl: Make sure TC-cold is blocked before enabling TC AUX power wells
drm/i915/selftests: Avoid passing a random 0 into ilog2
drm/i915: Fix wrong return value in intel_atomic_check()
drm/i915: Update bw_buddy pagemask table
drm/i915/display: Check for an LPSP encoder before dereferencing
drm/i915: Copy default modparams to mock i915_device
drm/i915: Provide the perf pmu.module
drm/amd/display: fix pow() crashing when given base 0
drm/amd/display: Reset scrambling on Test Pattern
drm/amd/display: fix dcn3 wide timing dsc validation
drm/amd/display: Fix DFPstate hang due to view port changed
drm/amd/display: Assign correct left shift
drm/amd/display: Call DMUB for eDP power control
drm/amdkfd: fix the wrong sdma instance query for renoir
drm/amdgpu: parse ta firmware for navy_flounder
drm/amdgpu: fix NULL pointer access issue when unloading driver
drm/amdgpu: fix uninit-value in arcturus_log_thermal_throttling_event()
drm/amdgpu: disable gfxoff for navy_flounder
drm/amdgpu/display: use GFP_ATOMIC in dcn20_validate_bandwidth_internal
...
mm, page_alloc: fix core hung in free_pcppages_bulk()
The following race is observed with the repeated online, offline and a
delay between two successive online of memory blocks of movable zone.
P1 P2
Online the first memory block in
the movable zone. The pcp struct
values are initialized to default
values,i.e., pcp->high = 0 &
pcp->batch = 1.
Allocate the pages from the
movable zone.
Try to Online the second memory
block in the movable zone thus it
entered the online_pages() but yet
to call zone_pcp_update().
This process is entered into
the exit path thus it tries
to release the order-0 pages
to pcp lists through
free_unref_page_commit().
As pcp->high = 0, pcp->count = 1
proceed to call the function
free_pcppages_bulk().
Update the pcp values thus the
new pcp values are like, say,
pcp->high = 378, pcp->batch = 63.
Read the pcp's batch value using
READ_ONCE() and pass the same to
free_pcppages_bulk(), pcp values
passed here are, batch = 63,
count = 1.
Since num of pages in the pcp
lists are less than ->batch,
then it will stuck in
while(list_empty(list)) loop
with interrupts disabled thus
a core hung.
Avoid this by ensuring free_pcppages_bulk() is called with proper count of
pcp list pages.
The mentioned race is some what easily reproducible without [1] because
pcp's are not updated for the first memory block online and thus there is
a enough race window for P2 between alloc+free and pcp struct values
update through onlining of second memory block.
With [1], the race still exists but it is very narrow as we update the pcp
struct values for the first memory block online itself.
This is not limited to the movable zone, it could also happen in cases
with the normal zone (e.g., hotplug to a node that only has DMA memory, or
no other memory yet).
Doug Berger [Fri, 21 Aug 2020 00:42:24 +0000 (17:42 -0700)]
mm: include CMA pages in lowmem_reserve at boot
The lowmem_reserve arrays provide a means of applying pressure against
allocations from lower zones that were targeted at higher zones. Its
values are a function of the number of pages managed by higher zones and
are assigned by a call to the setup_per_zone_lowmem_reserve() function.
The function is initially called at boot time by the function
init_per_zone_wmark_min() and may be called later by accesses of the
/proc/sys/vm/lowmem_reserve_ratio sysctl file.
The function init_per_zone_wmark_min() was moved up from a module_init to
a core_initcall to resolve a sequencing issue with khugepaged.
Unfortunately this created a sequencing issue with CMA page accounting.
The CMA pages are added to the managed page count of a zone when
cma_init_reserved_areas() is called at boot also as a core_initcall. This
makes it uncertain whether the CMA pages will be added to the managed page
counts of their zones before or after the call to
init_per_zone_wmark_min() as it becomes dependent on link order. With the
current link order the pages are added to the managed count after the
lowmem_reserve arrays are initialized at boot.
This means the lowmem_reserve values at boot may be lower than the values
used later if /proc/sys/vm/lowmem_reserve_ratio is accessed even if the
ratio values are unchanged.
In many cases the difference is not significant, but for example
an ARM platform with 1GB of memory and the following memory layout
cma: Reserved 256 MiB at 0x0000000030000000
Zone ranges:
DMA [mem 0x0000000000000000-0x000000002fffffff]
Normal empty
HighMem [mem 0x0000000030000000-0x000000003fffffff]
would result in 0 lowmem_reserve for the DMA zone. This would allow
userspace to deplete the DMA zone easily.
Funnily enough
$ cat /proc/sys/vm/lowmem_reserve_ratio
would fix up the situation because as a side effect it forces
setup_per_zone_lowmem_reserve.
This commit breaks the link order dependency by invoking
init_per_zone_wmark_min() as a postcore_initcall so that the CMA pages
have the chance to be properly accounted in their zone(s) and allowing
the lowmem_reserve arrays to receive consistent values.
Phillip Lougher [Fri, 21 Aug 2020 00:42:21 +0000 (17:42 -0700)]
squashfs: avoid bio_alloc() failure with 1Mbyte blocks
This is a regression introduced by the patch "migrate from ll_rw_block
usage to BIO".
Bio_alloc() is limited to 256 pages (1 Mbyte). This can cause a failure
when reading 1 Mbyte block filesystems. The problem is a datablock can be
fully (or almost uncompressed), requiring 256 pages, but, because blocks
are not aligned to page boundaries, it may require 257 pages to read.
Bio_kmalloc() can handle 1024 pages, and so use this for the edge
condition.
Hugh Dickins [Fri, 21 Aug 2020 00:42:17 +0000 (17:42 -0700)]
uprobes: __replace_page() avoid BUG in munlock_vma_page()
syzbot crashed on the VM_BUG_ON_PAGE(PageTail) in munlock_vma_page(), when
called from uprobes __replace_page(). Which of many ways to fix it?
Settled on not calling when PageCompound (since Head and Tail are equals
in this context, PageCompound the usual check in uprobes.c, and the prior
use of FOLL_SPLIT_PMD will have cleared PageMlocked already).
'chan->buf' is malloced in relay_open() by alloc_percpu() but not free
while destroy the relay channel. Fix it by adding free_percpu() before
return from relay_destroy_channel().
Jann Horn [Fri, 21 Aug 2020 00:42:11 +0000 (17:42 -0700)]
romfs: fix uninitialized memory leak in romfs_dev_read()
romfs has a superblock field that limits the size of the filesystem; data
beyond that limit is never accessed.
romfs_dev_read() fetches a caller-supplied number of bytes from the
backing device. It returns 0 on success or an error code on failure;
therefore, its API can't represent short reads, it's all-or-nothing.
However, when romfs_dev_read() detects that the requested operation would
cross the filesystem size limit, it currently silently truncates the
requested number of bytes. This e.g. means that when the content of a
file with size 0x1000 starts one byte before the filesystem size limit,
->readpage() will only fill a single byte of the supplied page while
leaving the rest uninitialized, leaking that uninitialized memory to
userspace.
Fix it by returning an error code instead of truncating the read when the
requested read operation would go beyond the end of the filesystem.
Like zap_pte_range add cond_resched so that we can avoid softlockups as
reported below. On non-preemptible kernel with large I/O map region (like
the one we get when using persistent memory with sector mode), an unmap of
the namespace can report below softlockups.
Hugh Dickins [Fri, 21 Aug 2020 00:42:02 +0000 (17:42 -0700)]
khugepaged: adjust VM_BUG_ON_MM() in __khugepaged_enter()
syzbot crashes on the VM_BUG_ON_MM(khugepaged_test_exit(mm), mm) in
__khugepaged_enter(): yes, when one thread is about to dump core, has set
core_state, and is waiting for others, another might do something calling
__khugepaged_enter(), which now crashes because I lumped the core_state
test (known as "mmget_still_valid") into khugepaged_test_exit(). I still
think it's best to lump them together, so just in this exceptional case,
check mm->mm_users directly instead of khugepaged_test_exit().
Thomas Gleixner [Wed, 19 Aug 2020 19:44:39 +0000 (21:44 +0200)]
core/entry: Respect syscall number rewrites
The transcript of the x86 entry code to the generic version failed to
reload the syscall number from ptregs after ptrace and seccomp have run,
which both can modify the syscall number in ptregs. It returns the original
syscall number instead which is obviously not the right thing to do.
x86/entry/64: Do not use RDPID in paranoid entry to accomodate KVM
KVM has an optmization to avoid expensive MRS read/writes on
VMENTER/EXIT. It caches the MSR values and restores them either when
leaving the run loop, on preemption or when going out to user space.
The affected MSRs are not required for kernel context operations. This
changed with the recently introduced mechanism to handle FSGSBASE in the
paranoid entry code which has to retrieve the kernel GSBASE value by
accessing per CPU memory. The mechanism needs to retrieve the CPU number
and uses either LSL or RDPID if the processor supports it.
Unfortunately RDPID uses MSR_TSC_AUX which is in the list of cached and
lazily restored MSRs, which means between the point where the guest value
is written and the point of restore, MSR_TSC_AUX contains a random number.
If an NMI or any other exception which uses the paranoid entry path happens
in such a context, then RDPID returns the random guest MSR_TSC_AUX value.
As a consequence this reads from the wrong memory location to retrieve the
kernel GSBASE value. Kernel GS is used to for all regular this_cpu_*()
operations. If the GSBASE in the exception handler points to the per CPU
memory of a different CPU then this has the obvious consequences of data
corruption and crashes.
As the paranoid entry path is the only place which accesses MSR_TSX_AUX
(via RDPID) and the fallback via LSL is not significantly slower, remove
the RDPID alternative from the entry path and always use LSL.
The alternative would be to write MSR_TSC_AUX on every VMENTER and VMEXIT
which would be inflicting massive overhead on that code path.
Kajol Jain [Fri, 21 Aug 2020 08:06:10 +0000 (13:36 +0530)]
powerpc/perf/hv-24x7: Move cpumask file to top folder of hv-24x7 driver
Commit 792f73f747b8 ("powerpc/hv-24x7: Add sysfs files inside hv-24x7
device to show cpumask") added cpumask file as part of hv-24x7 driver
inside the interface folder. The cpumask file is supposed to be in the
top folder of the PMU driver in order to make hotplug work.
This patch fixes that issue and creates new group 'cpumask_attr_group'
to add cpumask file and make sure it added in top folder.
Rob Herring [Mon, 3 Aug 2020 19:31:25 +0000 (13:31 -0600)]
KVM: arm64: Print warning when cpu erratum can cause guests to deadlock
If guests don't have certain CPU erratum workarounds implemented, then
there is a possibility a guest can deadlock the system. IOW, only trusted
guests should be used on systems with the erratum.
Marc Zyngier [Fri, 31 Jul 2020 17:38:24 +0000 (18:38 +0100)]
arm64: Allow booting of late CPUs affected by erratum 1418040
As we can now switch from a system that isn't affected by 1418040
to a system that globally is affected, let's allow affected CPUs
to come in at a later time.
projects / linux.git / log