The kernel needs to explicitly enable FSGSBASE. So, the application needs
to know if it can safely use these instructions. Just looking at the CPUID
bit is not enough because it may be running in a kernel that does not
enable the instructions.
One way for the application would be to just try and catch the SIGILL.
But that is difficult to do in libraries which may not want to overwrite
the signal handlers of the main application.
Enumerate the enabled FSGSBASE capability in bit 1 of AT_HWCAP2 in the ELF
aux vector. AT_HWCAP2 is already used by PPC for similar purposes.
The application can access it open coded or by using the getauxval()
function in newer versions of glibc.
[ tglx: Massaged changelog ]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Link: https://lkml.kernel.org/r/1557309753-24073-18-git-send-email-chang.seok.bae@intel.com
Without FSGSBASE, user space cannot change GSBASE other than through a
PRCTL. The kernel enforces that the user space GSBASE value is postive as
negative values are used for detecting the kernel space GSBASE value in the
paranoid entry code.
If FSGSBASE is enabled, user space can set arbitrary GSBASE values without
kernel intervention, including negative ones, which breaks the paranoid
entry assumptions.
To avoid this, paranoid entry needs to unconditionally save the current
GSBASE value independent of the interrupted context, retrieve and write the
kernel GSBASE and unconditionally restore the saved value on exit. The
restore happens either in paranoid_exit or in the special exit path of the
NMI low level code.
All other entry code pathes which use unconditional SWAPGS are not affected
as they do not depend on the actual content.
[ tglx: Massaged changelogs and comments ]
Suggested-by: H. Peter Anvin <hpa@zytor.com>
Suggested-by: Andy Lutomirski <luto@kernel.org>
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lkml.kernel.org/r/1557309753-24073-13-git-send-email-chang.seok.bae@intel.com
With the new FSGSBASE instructions, FS and GSABSE can be efficiently read
and writen in __switch_to(). Use that capability to preserve the full
state.
This will enable user code to do whatever it wants with the new
instructions without any kernel-induced gotchas. (There can still be
architectural gotchas: movl %gs,%eax; movl %eax,%gs may change GSBASE if
WRGSBASE was used, but users are expected to read the CPU manual before
doing things like that.)
This is a considerable speedup. It seems to save about 100 cycles
per context switch compared to the baseline 4.6-rc1 behavior on a
Skylake laptop.
[ chang: 5~10% performance improvements were seen with a context switch
benchmark that ran threads with different FS/GSBASE values (to the
baseline 4.16). Minor edit on the changelog. ]
[ tglx: Masaage changelog ]
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Link: https://lkml.kernel.org/r/1557309753-24073-8-git-send-email-chang.seok.bae@intel.com
Add cpu feature conditional FSGSBASE access to the relevant helper
functions. That allows to accelerate certain FS/GS base operations in
subsequent changes.
Note, that while possible, the user space entry/exit GSBASE operations are
not going to use the new FSGSBASE instructions. The reason is that it would
require additional storage for the user space value which adds more
complexity to the low level code and experiments have shown marginal
benefit. This may be revisited later but for now the SWAPGS based handling
in the entry code is preserved except for the paranoid entry/exit code.
To preserve the SWAPGS entry mechanism introduce __[rd|wr]gsbase_inactive()
helpers. Note, for Xen PV, paravirt hooks can be added later as they might
allow a very efficient but different implementation.
[ tglx: Massaged changelog ]
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ravi Shankar <ravi.v.shankar@intel.com>
Cc: Andrew Cooper <andrew.cooper3@citrix.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Link: https://lkml.kernel.org/r/1557309753-24073-7-git-send-email-chang.seok.bae@intel.com
When a ptracer writes a ptracee's FS/GSBASE with a different value, the
selector is also cleared. This behavior is not correct as the selector
should be preserved.
Update only the base value and leave the selector intact. To simplify the
code further remove the conditional checking for the same value as this
code is not performance critical.
The only recognizable downside of this change is when the selector is
already nonzero on write. The base will be reloaded according to the
selector. But the case is highly unexpected in real usages.
[ tglx: Massage changelog ]
Suggested-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Chang S. Bae <chang.seok.bae@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Link: https://lkml.kernel.org/r/9040CFCD-74BD-4C17-9A01-B9B713CF6B10@intel.com
... into a separate function for better readability. Split out from a
patch from Fenghua Yu <fenghua.yu@intel.com> to keep the mechanical,
sole code movement separate for easy review.
No functional changes.
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: x86@kernel.org
Pull ceph fixes from Ilya Dryomov:
"A change to call iput() asynchronously to avoid a possible deadlock
when iput_final() needs to wait for in-flight I/O (e.g. readahead) and
a fixup for a cleanup that went into -rc1"
* tag 'ceph-for-5.2-rc4' of git://github.com/ceph/ceph-client:
ceph: fix error handling in ceph_get_caps()
ceph: avoid iput_final() while holding mutex or in dispatch thread
ceph: single workqueue for inode related works
Pull xen fix from Juergen Gross:
"Just one fix for the Xen block frontend driver avoiding allocations
with order > 0"
* tag 'for-linus-5.2b-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen-blkfront: switch kcalloc to kvcalloc for large array allocation
Pull s390 fixes from Heiko Carstens:
- fix stack unwinder: the stack unwinder rework has on off-by-one bug
which prevents following stack backchains over more than one context
(e.g. irq -> process).
- fix address space detection in exception handler: if user space
switches to access register mode, which is not supported anymore, the
exception handler may resolve to the wrong address space.
* tag 's390-5.2-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390/unwind: correct stack switching during unwind
s390/mm: fix address space detection in exception handling
Pull MIPS fixes from Paul Burton:
- Declare ginvt() __always_inline due to its use of an argument as an
inline asm immediate.
- A VDSO build fix following Kbuild changes made this cycle.
- A fix for boot failures on txx9 systems following memory
initialization changes made this cycle.
- Bounds check virt_addr_valid() to prevent it spuriously indicating
that bogus addresses are valid, in turn fixing hardened usercopy
failures that have been present since v4.12.
- Build uImage.gz for pistachio systems by default, since this is the
image we need in order to actually boot on a board.
- Remove an unused variable in our uprobes code.
* tag 'mips_fixes_5.2_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
MIPS: uprobes: remove set but not used variable 'epc'
MIPS: pistachio: Build uImage.gz by default
MIPS: Make virt_addr_valid() return bool
MIPS: Bounds check virt_addr_valid
MIPS: TXx9: Fix boot crash in free_initmem()
MIPS: remove a space after -I to cope with header search paths for VDSO
MIPS: mark ginvt() as __always_inline
Pull yet more SPDX updates from Greg KH:
"Another round of SPDX header file fixes for 5.2-rc4
These are all more "GPL-2.0-or-later" or "GPL-2.0-only" tags being
added, based on the text in the files. We are slowly chipping away at
the 700+ different ways people tried to write the license text. All of
these were reviewed on the spdx mailing list by a number of different
people.
We now have over 60% of the kernel files covered with SPDX tags:
$ ./scripts/spdxcheck.py -v 2>&1 | grep Files
Files checked: 64533
Files with SPDX: 40392
Files with errors: 0
I think the majority of the "easy" fixups are now done, it's now the
start of the longer-tail of crazy variants to wade through"
* tag 'spdx-5.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (159 commits)
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 450
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 449
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 448
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 446
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 445
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 444
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 443
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 442
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 441
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 440
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 438
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 437
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 436
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 435
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 434
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 433
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 432
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 431
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 430
treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 429
...
Pull char/misc driver fixes from Greg KH:
"Here are some small char and misc driver fixes for 5.2-rc4 to resolve
a number of reported issues.
The most "notable" one here is the kernel headers in proc^Wsysfs
fixes. Those changes move the header file info into sysfs and fixes
the build issues that you reported.
Other than that, a bunch of small habanalabs driver fixes, some fpga
driver fixes, and a few other tiny driver fixes.
All of these have been in linux-next for a while with no reported
issues"
* tag 'char-misc-5.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
habanalabs: Read upper bits of trace buffer from RWPHI
habanalabs: Fix virtual address access via debugfs for 2MB pages
fpga: zynqmp-fpga: Correctly handle error pointer
habanalabs: fix bug in checking huge page optimization
habanalabs: Avoid using a non-initialized MMU cache mutex
habanalabs: fix debugfs code
uapi/habanalabs: add opcode for enable/disable device debug mode
habanalabs: halt debug engines on user process close
test_firmware: Use correct snprintf() limit
genwqe: Prevent an integer overflow in the ioctl
parport: Fix mem leak in parport_register_dev_model
fpga: dfl: expand minor range when registering chrdev region
fpga: dfl: Add lockdep classes for pdata->lock
fpga: dfl: afu: Pass the correct device to dma_mapping_error()
fpga: stratix10-soc: fix use-after-free on s10_init()
w1: ds2408: Fix typo after 49695ac468 (reset on output_write retry with readback)
kheaders: Do not regenerate archive if config is not changed
kheaders: Move from proc to sysfs
lkdtm/bugs: Adjust recursion test to avoid elision
lkdtm/usercopy: Moves the KERNEL_DS test to non-canonical
Pull i2c fixes from Wolfram Sang:
"I2C has a driver bugfix and a MAINTAINERS fix"
* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
MAINTAINERS: Karthikeyan Ramasubramanian is MIA
i2c: xiic: Add max_read_len quirk
Pull dmaengine fixes from Vinod Koul:
- jz4780 transfer fix for acking descriptors early
- fsl-qdma: clean registers on error
- dw-axi-dmac: null pointer dereference fix
- mediatek-cqdma: fix sleeping in atomic context
- tegra210-adma: fix bunch os issues like crashing in driver probe,
channel FIFO configuration etc.
- sprd: Fixes for possible crash on descriptor status, block length
overflow. For 2-stage transfer fix incorrect start, configuration and
interrupt handling.
* tag 'dmaengine-fix-5.2-rc4' of git://git.infradead.org/users/vkoul/slave-dma:
dmaengine: sprd: Add interrupt support for 2-stage transfer
dmaengine: sprd: Fix the right place to configure 2-stage transfer
dmaengine: sprd: Fix block length overflow
dmaengine: sprd: Fix the incorrect start for 2-stage destination channels
dmaengine: sprd: Add validation of current descriptor in irq handler
dmaengine: sprd: Fix the possible crash when getting descriptor status
dmaengine: tegra210-adma: Fix spelling
dmaengine: tegra210-adma: Fix channel FIFO configuration
dmaengine: tegra210-adma: Fix crash during probe
dmaengine: mediatek-cqdma: sleeping in atomic context
dmaengine: dw-axi-dmac: fix null dereference when pointer first is null
dmaengine: fsl-qdma: Add improvement
dmaengine: jz4780: Fix transfers being ACKed too soon
Pull block fixes from Jens Axboe:
- Allow symlink from the bfq.weight cgroup parameter to the general
weight (Angelo)
- Damien is new skd maintainer (Bart)
- NVMe pull request from Sagi, with a few small fixes.
- Ensure we set DMA segment size properly, dma-debug is now tripping on
these (Christoph)
- Remove useless debugfs_create() return check (Greg)
- Remove redundant unlikely() check on IS_ERR() (Kefeng)
- Fixup request freeing on exit (Ming)
* tag 'for-linus-20190608' of git://git.kernel.dk/linux-block:
block, bfq: add weight symlink to the bfq.weight cgroup parameter
cgroup: let a symlink too be created with a cftype file
block: free sched's request pool in blk_cleanup_queue
nvme-rdma: use dynamic dma mapping per command
nvme: Fix u32 overflow in the number of namespace list calculation
mmc: also set max_segment_size in the device
mtip32xx: also set max_segment_size in the device
rsxx: don't call dma_set_max_seg_size
nvme-pci: don't limit DMA segement size
block: Drop unlikely before IS_ERR(_OR_NULL)
block: aoe: no need to check return value of debugfs_create functions
nvmet: fix data_len to 0 for bdev-backed write_zeroes
MAINTAINERS: Hand over skd maintainership
nvme-tcp: fix queue mapping when queue count is limited
nvme-rdma: fix queue mapping when queue count is limited
Pull SCSI fixes from James Bottomley:
"Two bug fixes, both for fairly serious problems; the UFS one looks
like it could be used to exfiltrate data from the kernel, although
probably only a privileged user has access to the command management
interface and the missing unlock in smartpqi is long standing and
probably a little used error path"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: smartpqi: unlock on error in pqi_submit_raid_request_synchronous()
scsi: ufs: Check that space was properly alloced in copy_query_response
Pull Kselftest fix from Shuah Khan:
"This consists of a single fix for a vm test build failure regression
when it is built by itself"
* tag 'linux-kselftest-5.2-rc4-2' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
selftests: vm: Fix test build failure when built by itself
Pull drm fixes from Dave Airlie:
"A small bit more lively this week but not majorly so. I'm away in
Japan next week for family holiday, so I'll be pretty disconnected,
I've asked Daniel to do fixes for the week while I'm out.
The nouveau firmware changes are a bit large, but they address a big
problem where a whole set of boards don't load with the driver, and
the new firmware fixes that, so I think it's worth trying to land it
now.
core:
- Allow fb changes in async commits (drivers as well)
udmabuf:
- Unmap scatterlist when unmapping udmabuf
nouveau:
- firmware loading fixes for secboot firmware on new GPU revision.
komeda:
- oops, dma mapping and warning fixes
arm-hdlcd:
- clock fixes
- mode validation fix
i915:
- Add a missing Icelake workaround
- GVT - DMA map fault fix and enforcement fixes
amdgpu:
- DCE resume fix
- New raven variation updates"
* tag 'drm-fixes-2019-06-07-1' of git://anongit.freedesktop.org/drm/drm: (33 commits)
drm/nouveau/secboot/gp10[2467]: support newer FW to fix SEC2 failures on some boards
drm/nouveau/secboot: enable loading of versioned LS PMU/SEC2 ACR msgqueue FW
drm/nouveau/secboot: split out FW version-specific LS function pointers
drm/nouveau/secboot: pass max supported FW version to LS load funcs
drm/nouveau/core: support versioned firmware loading
drm/nouveau/core: pass subdev into nvkm_firmware_get, rather than device
drm/komeda: Potential error pointer dereference
drm/komeda: remove set but not used variable 'kcrtc'
drm/amd/amdgpu: add RLC firmware to support raven1 refresh
drm/amd/powerplay: add set_power_profile_mode for raven1_refresh
drm/amdgpu: fix ring test failure issue during s3 in vce 3.0 (V2)
udmabuf: actually unmap the scatterlist
drm/arm/hdlcd: Allow a bit of clock tolerance
drm/arm/hdlcd: Actually validate CRTC modes
drm/arm/mali-dp: Add a loop around the second set CVAL and try 5 times
drm/komeda: fixing of DMA mapping sg segment warning
drm: don't block fb changes for async plane updates
drm/vc4: fix fb references in async update
drm/msm: fix fb references in async update
drm/amd: fix fb references in async update
...
A mail just bounced back with "user unknown":
550 5.1.1 <kramasub@codeaurora.org> User doesn't exist
I also couldn't find a more recent address in git history. So, remove
this stale entry.
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
This driver does not support reading more than 255 bytes at once because
the register for storing the number of bytes to read is only 8 bits. Add
a max_read_len quirk to enforce this.
This was found when using this driver with the SFP driver, which was
previously reading all 256 bytes in the SFP EEPROM in one transaction.
This caused a bunch of hard-to-debug errors in the xiic driver since the
driver/logic was treating the number of bytes to read as zero.
Rejecting transactions that aren't supported at least allows the problem
to be diagnosed more easily.
Signed-off-by: Robert Hancock <hancock@sedsystems.ca>
Reviewed-by: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Cc: stable@kernel.org
Pull hwmon fixes from Guenter Roeck:
- Fix a couple of inconsistencies and locking problems in pmbus driver
- Register with thermal subsystem only on systems supporting devicetree
* tag 'hwmon-for-v5.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:
hwmon: (pmbus/core) Treat parameters as paged if on multiple pages
hwmon: (pmbus/core) mutex_lock write in pmbus_set_samples
hwmon: (core) add thermal sensors only if dev->of_node is present
The lockref cmpxchg loop is unbound as long as the spinlock is not
taken. Depending on the hardware implementation of compare-and-swap
a high number of loop retries might happen.
Add an upper bound to the loop to force the fallback to spinlocks
after some time. A retry value of 100 should not impact any hardware
that does not have this issue.
With the retry limit the performance of an open-close testcase
improved between 60-70% on ThunderX2.
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Jan Glauber <jglauber@marvell.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Architectures that support memory tagging have a need to perform untagging
(stripping the tag) in various parts of the kernel. This patch adds an
untagged_addr() macro, which is defined as noop for architectures that do
not support memory tagging. The oncoming patch series will define it at
least for sparc64 and arm64.
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: Khalid Aziz <khalid.aziz@oracle.com>
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull xtensa fix from Max Filippov:
"Fix a section mismatch between memblock_reserve and mem_reserve.
This fixes tinyconfig xtensa builds"
* tag 'xtensa-20190607' of git://github.com/jcmvbkbc/linux-xtensa:
xtensa: Fix section mismatch between memblock_reserve and mem_reserve
Pull NVMe fixes from Sagi.
* 'nvme-5.2-rc-next' of git://git.infradead.org/nvme:
nvme-rdma: use dynamic dma mapping per command
nvme: Fix u32 overflow in the number of namespace list calculation
nvmet: fix data_len to 0 for bdev-backed write_zeroes
nvme-tcp: fix queue mapping when queue count is limited
nvme-rdma: fix queue mapping when queue count is limited
Pull more Kbuild fixes from Masahiro Yamada:
- fix kselftest-merge to find config fragments in deeper directories
- fix kconfig unit test, which was broken by SPDX tag addition
- add + prefix to buildtar to suppress jobserver unavailable warning
- fix checkstack.pl to recognize arch=arm64
- suppress noisy warning from cc-cross-prefix
* tag 'kbuild-fixes-v5.2-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: use more portable 'command -v' for cc-cross-prefix
scripts/checkstack.pl: Fix arm64 wrong or unknown architecture
kbuild: tar-pkg: enable communication with jobserver
kconfig: tests: fix recursive inclusion unit test
kbuild: teach kselftest-merge to find nested config files
Pull power management fixes from Rafael Wysocki:
"These fix a crash during resume from hibernation introduced during the
4.19 cycle, cause the new Performance and Energy Bias Hint (EPB) code
to be built only if CONFIG_PM is set and add a few missing kerneldoc
comments.
Specifics:
- Fix a crash that occurs when a kernel with 'nosmt' in the command
line is used to resume the system from hibernation (as the
"restore" kernel), because memory mapping differences between the
restore and image kernels cause SMT siblings to be woken up from
idle states and subsequently they try to fetch instructions from
incorrect memory locations (Jiri Kosina).
- Cause the new Performance and Energy Bias Hint (EPB) code to be
built only if CONFIG_PM is set, because that code is not really
necessary otherwise (Rafael Wysocki).
- Add kerneldoc comments to documents some helper functions related
to system-wide suspend to avoid possible confusion regarding their
purpose (Rafael Wysocki)"
* tag 'pm-5.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
x86/power: Fix 'nosmt' vs hibernation triple fault during resume
PM: sleep: Add kerneldoc comments to some functions
x86: intel_epb: Do not build when CONFIG_PM is unset
