This driver often takes on the order of 10ms to start, but in some cases
as much as 190ms. It shouldn't have many cross-device dependencies to
race with, nor racy access to shared state with other drivers, so this
should be a relatively low risk change. We've done similarly with a
variety of other MMC host drivers already.
This driver was pinpointed as part of a survey of top slowest initcalls
(i.e., are built in, and probing synchronously) on a lab of ChromeOS
systems.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221028155633.1.I6c4bfb31e88fad934e7360242cb662e01612c1bb@changeid
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
For standard tuning method on usdhc, the previous tuning result can
impact current tuning result, let current tuning can't set the correct
delay cell. And from the logic, this is also reasonable for manual
tuning method. So reset the tuning logic before execute tuning.
To avoid compile issue, this patch also move the esdhc_reset_tuning()
upper.
Find this issue when support SDIO WiFi in band wakeup feature. After
system resume back, will do re-tuning, but then meet data CRC error.
Do not meet this issue on SD/eMMC, because we already call
esdhc_reset_tuning() when config the legency ios, and SD/eMMC need
to re-init when system resume back, but SDIO device don't do re-init
if it has MMC_PM_KEEP_POWER pm_flags.
Signed-off-by: Haibo Chen <haibo.chen@nxp.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/1666947869-7904-1-git-send-email-haibo.chen@nxp.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Add dynamic configuration support for Xilinx ZynqMP which takes care of
configuring the SD secure space configuration registers using EEMI APIs,
performing SD reset assert and deassert.
High level sequence:
- Check for the PM dynamic configuration support, if no error proceed with
SD dynamic configurations(next steps) otherwise skip the dynamic
configuration.
- Put the SD Controller in reset.
- Configure SD Fixed configurations.
- Configure the SD Slot Type.
- Configure the BASE_CLOCK.
- Configure the 8-bit support.
- Bring the SD Controller out of reset.
- Wait for 1msec delay.
Signed-off-by: Sai Krishna Potthuri <sai.krishna.potthuri@amd.com>
Acked-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/20221019054857.8286-1-sai.krishna.potthuri@amd.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The binding was describing a single clock list for all platforms, but
that's not really suitable:
Most platforms using at least 2 clocks (source, hclk), some of them
a third "source_cg". Mt2712 requires an extra 'bus_clk' on some of
its controllers, while mt8192 requires 8 clocks.
Move the clock definitions inside if blocks that match on the
compatibles.
I used Patch from Nícolas F. R. A. Prado and modified it to not using
"not" statement.
Fixes: 59a23395d8 ("dt-bindings: mmc: Add support for MT8192 SoC")
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com>
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20221025132953.81286-2-linux@fw-web.de
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-15-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-14-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Link: https://lore.kernel.org/r/20221005101951.3165-13-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
When the second argument to kmap_atomic was removed by commit 482fce997e
("mmc: remove the second argument of k[un]map_atomic()"),
local_irq_{save,restore}() should have been removed also.
Remove it now.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Link: https://lore.kernel.org/r/20221005101951.3165-12-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-11-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
tifm_sd_bounce_block() calls functions that ultimate use kmap_atomic() to
map pages.
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
When the second argument to kmap_atomic was removed by commit 482fce997e
("mmc: remove the second argument of k[un]map_atomic()"),
local_irq_{save,restore}() should have been removed also.
Remove it now.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-10-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
tifm_sd_transfer_data() calls functions that ultimate use kmap_atomic() to
map pages.
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
When the second argument to kmap_atomic was removed by commit 482fce997e
("mmc: remove the second argument of k[un]map_atomic()"),
local_irq_{save,restore}() should have been removed also.
Remove it now.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-9-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
sg_copy_{from,to}_buffer() call sg_copy_buffer() which uses an
sg_mapping_iter with flag SG_MITER_ATOMIC, so then sg_miter_next() uses
kmap_atomic() to map pages.
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
Remove local_irq_{save,restore}() around sg_copy_{from,to}_buffer().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-8-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-7-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
sg_miter_next() using an sg_mapping_iter with flag SG_MITER_ATOMIC uses
kmap_atomic() to map pages.
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
Remove local_irq_{save,restore}() around sg_miter_{next,stop}().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-6-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
Unfortunately, that unnecessary pattern of code has been copied since
and persists in bcm2385.c.
Remove it.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-5-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
kmap_local_page() is equivalent to kmap_atomic() except that it does not
disable page faults or preemption. Where possible kmap_local_page() is
preferred to kmap_atomic() - refer kernel highmem documentation.
In this case, there is no need to disable page faults or preemption, so
replace kmap_atomic() with kmap_local_page(), and, correspondingly,
kunmap_atomic() with kunmap_local().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-4-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
sg_miter_next() using an sg_mapping_iter with flag SG_MITER_ATOMIC uses
kmap_atomic() to map pages.
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
Remove local_irq_{save,restore}() around sg_miter_{next,stop}().
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-3-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
A long time ago the kmap_atomic API required a slot to be provided which
risked the possibility that other code might use the same slot at the
same time. Disabling interrupts prevented the possibility of an interrupt
handler doing that. However, that went away with
commit 3e4d3af501 ("mm: stack based kmap_atomic()").
When the second argument to kmap_atomic was removed by commit 482fce997e
("mmc: remove the second argument of k[un]map_atomic()"),
local_irq_{save,restore}() should have been removed also.
Remove it now.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20221005101951.3165-2-adrian.hunter@intel.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Currently tpm transactions are executed unconditionally in
tpm_pm_suspend() function, which may lead to races with other tpm
accessors in the system.
Specifically, the hw_random tpm driver makes use of tpm_get_random(),
and this function is called in a loop from a kthread, which means it's
not frozen alongside userspace, and so can race with the work done
during system suspend:
tpm tpm0: tpm_transmit: tpm_recv: error -52
tpm tpm0: invalid TPM_STS.x 0xff, dumping stack for forensics
CPU: 0 PID: 1 Comm: init Not tainted 6.1.0-rc5+ #135
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.0-20220807_005459-localhost 04/01/2014
Call Trace:
tpm_tis_status.cold+0x19/0x20
tpm_transmit+0x13b/0x390
tpm_transmit_cmd+0x20/0x80
tpm1_pm_suspend+0xa6/0x110
tpm_pm_suspend+0x53/0x80
__pnp_bus_suspend+0x35/0xe0
__device_suspend+0x10f/0x350
Fix this by calling tpm_try_get_ops(), which itself is a wrapper around
tpm_chip_start(), but takes the appropriate mutex.
Signed-off-by: Jan Dabros <jsd@semihalf.com>
Reported-by: Vlastimil Babka <vbabka@suse.cz>
Tested-by: Jason A. Donenfeld <Jason@zx2c4.com>
Tested-by: Vlastimil Babka <vbabka@suse.cz>
Link: https://lore.kernel.org/all/c5ba47ef-393f-1fba-30bd-1230d1b4b592@suse.cz/
Cc: stable@vger.kernel.org
Fixes: e891db1a18 ("tpm: turn on TPM on suspend for TPM 1.x")
[Jason: reworked commit message, added metadata]
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Reviewed-by: Jarkko Sakkinen <jarkko@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull perf fix from Borislav Petkov:
- Fix a use-after-free case where the perf pending task callback would
see an already freed event
* tag 'perf_urgent_for_v6.1_rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf: Fix perf_pending_task() UaF
Pull timer fix from Borislav Petkov:
- Revert a fix to RISC-V timers supposed to address an uncertainty
whether clock events are received during S3 or not which locks up
other RISC-V platforms. The issue will be fixed differently later.
* tag 'timers_urgent_for_v6.1_rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Revert "clocksource/drivers/riscv: Events are stopped during CPU suspend"
Pull powerpc fixes from Michael Ellerman:
- Fix oops in 32-bit BPF tail call tests
- Add missing declaration for machine_check_early_boot()
Thanks to Christophe Leroy and Naveen N. Rao.
* tag 'powerpc-6.1-6' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s: Add missing declaration for machine_check_early_boot()
powerpc/bpf/32: Fix Oops on tail call tests
Pull input fix from Dmitry Torokhov:
- a fix for Raydium touchscreen driver to stop leaking memory when
sending commands to the chip
* tag 'input-for-v6.1-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
Input: raydium_ts_i2c - fix memory leak in raydium_i2c_send()
Pull i2c fixes from Wolfram Sang:
"A power state fix in the core for ACPI devices, a regression fix
regarding bus recovery for the cadence driver, a DMA handling fix for
the imx driver, and two error path fixes (npcm7xx and qcom-geni)"
* tag 'i2c-for-6.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
i2c: imx: Only DMA messages with I2C_M_DMA_SAFE flag set
i2c: qcom-geni: fix error return code in geni_i2c_gpi_xfer
i2c: cadence: Fix regression with bus recovery
i2c: Restore initial power state if probe fails
i2c: npcm7xx: Fix error handling in npcm_i2c_init()
Pull dax fixes from Dan Williams:
"A few bug fixes around the handling of "Soft Reserved" memory and
memory tiering information.
Linux is starting to enounter more real world systems that deploy an
ACPI HMAT to describe different performance classes of memory, as well
the "special purpose" (Linux "Soft Reserved") designation from EFI.
These fixes result from that testing.
It has all appeared in -next for a while with no known issues.
- Fix duplicate overlapping device-dax instances for HMAT described
"Soft Reserved" Memory
- Fix missing node targets in the sysfs representation of memory
tiers
- Remove a confusing variable initialization"
* tag 'dax-fixes-6.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:
device-dax: Fix duplicate 'hmem' device registration
ACPI: HMAT: Fix initiator registration for single-initiator systems
ACPI: HMAT: remove unnecessary variable initialization
Pull block fixes from Jens Axboe:
"Just a small NVMe merge for this week, fixing protection of the name
space list, and a missing clear of a reserved field when unused"
* tag 'block-6.1-2022-12-02' of git://git.kernel.dk/linux:
nvme: fix SRCU protection of nvme_ns_head list
nvme-pci: clear the prp2 field when not used
Pull pin control fixes from Linus Walleij:
"Three driver fixes. The Intel fix looks like the most important.
- Fix a potential divide by zero in pinctrl-singe (OMAP and
HiSilicon)
- Disable IRQs on startup in the Mediatek driver. This is a classic,
we should be looking out for this more.
- Save and restore pins in 'direct IRQ' mode in the Intel driver,
this works around firmware bugs"
* tag 'pinctrl-v6.1-5' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl:
pinctrl: intel: Save and restore pins in "direct IRQ" mode
pinctrl: meditatek: Startup with the IRQs disabled
pinctrl: single: Fix potential division by zero
Pull RISC-V fixes from Palmer Dabbelt:
- build fix for the NR_CPUS Kconfig SBI version dependency
- fixes to early memory initialization, to fix page permissions in EFI
and post-initmem-free
- build fix for the VDSO, to avoid trying to profile the VDSO functions
- fixes for kexec crash handling, to fix multi-core and interrupt
related initialization inside the crash kernel
- fix for a race condition when handling multiple concurrect kernel
stack overflows
* tag 'riscv-for-linus-6.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
riscv: kexec: Fixup crash_smp_send_stop without multi cores
riscv: kexec: Fixup irq controller broken in kexec crash path
riscv: mm: Proper page permissions after initmem free
riscv: vdso: fix section overlapping under some conditions
riscv: fix race when vmap stack overflow
riscv: Sync efi page table's kernel mappings before switching
riscv: Fix NR_CPUS range conditions
Pull MMC fixes from Ulf Hansson:
"MMC core:
- Fix ambiguous TRIM and DISCARD args
- Fix removal of debugfs file for mmc_test
MMC host:
- mtk-sd: Add missing clk_disable_unprepare() in an error path
- sdhci: Fix I/O voltage switch delay for UHS-I SD cards
- sdhci-esdhc-imx: Fix CQHCI exit halt state check
- sdhci-sprd: Fix voltage switch"
* tag 'mmc-v6.1-rc5-2' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
mmc: sdhci-sprd: Fix no reset data and command after voltage switch
mmc: sdhci: Fix voltage switch delay
mmc: mtk-sd: Fix missing clk_disable_unprepare in msdc_of_clock_parse()
mmc: mmc_test: Fix removal of debugfs file
mmc: sdhci-esdhc-imx: correct CQHCI exit halt state check
mmc: core: Fix ambiguous TRIM and DISCARD arg
The "force" argument to write_spec_ctrl_current() is currently ambiguous
as it does not guarantee the MSR write. This is due to the optimization
that writes to the MSR happen only when the new value differs from the
cached value.
This is fine in most cases, but breaks for S3 resume when the cached MSR
value gets out of sync with the hardware MSR value due to S3 resetting
it.
When x86_spec_ctrl_current is same as x86_spec_ctrl_base, the MSR write
is skipped. Which results in SPEC_CTRL mitigations not getting restored.
Move the MSR write from write_spec_ctrl_current() to a new function that
unconditionally writes to the MSR. Update the callers accordingly and
rename functions.
[ bp: Rework a bit. ]
Fixes: caa0ff24d5 ("x86/bugs: Keep a per-CPU IA32_SPEC_CTRL value")
Suggested-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Pawan Gupta <pawan.kumar.gupta@linux.intel.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: <stable@kernel.org>
Link: https://lore.kernel.org/r/806d39b0bfec2fe8f50dc5446dff20f5bb24a959.1669821572.git.pawan.kumar.gupta@linux.intel.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull sound fixes from Takashi Iwai:
"Likely the last piece for 6.1; the only significant fixes are ASoC
core ops fixes, while others are device-specific (rather minor) fixes
in ASoC and FireWire drivers.
All appear safe enough to take as a late stage material"
* tag 'sound-6.1-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: dice: fix regression for Lexicon I-ONIX FW810S
ASoC: cs42l51: Correct PGA Volume minimum value
ASoC: ops: Correct bounds check for second channel on SX controls
ASoC: tlv320adc3xxx: Fix build error for implicit function declaration
ASoC: ops: Check bounds for second channel in snd_soc_put_volsw_sx()
ASoC: ops: Fix bounds check for _sx controls
ASoC: fsl_micfil: explicitly clear CHnF flags
ASoC: fsl_micfil: explicitly clear software reset bit
Pull drm fixes from Dave Airlie:
"Things do seem to have finally settled down, just four i915 and one
amdgpu this week. Probably won't have much for next week if you do
push rc8 out.
i915:
- Fix dram info readout
- Remove non-existent pipes from bigjoiner pipe mask
- Fix negative value passed as remaining time
- Never return 0 if not all requests retired
amdgpu:
- VCN fix for vangogh"
* tag 'drm-fixes-2022-12-02' of git://anongit.freedesktop.org/drm/drm:
drm/amdgpu: enable Vangogh VCN indirect sram mode
drm/i915: Never return 0 if not all requests retired
drm/i915: Fix negative value passed as remaining time
drm/i915: Remove non-existent pipes from bigjoiner pipe mask
drm/i915/mtl: Fix dram info readout
Pull misc hotfixes from Andrew Morton:
"15 hotfixes, 11 marked cc:stable.
Only three or four of the latter address post-6.0 issues, which is
hopefully a sign that things are converging"
* tag 'mm-hotfixes-stable-2022-12-02' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
revert "kbuild: fix -Wimplicit-function-declaration in license_is_gpl_compatible"
Kconfig.debug: provide a little extra FRAME_WARN leeway when KASAN is enabled
drm/amdgpu: temporarily disable broken Clang builds due to blown stack-frame
mm/khugepaged: invoke MMU notifiers in shmem/file collapse paths
mm/khugepaged: fix GUP-fast interaction by sending IPI
mm/khugepaged: take the right locks for page table retraction
mm: migrate: fix THP's mapcount on isolation
mm: introduce arch_has_hw_nonleaf_pmd_young()
mm: add dummy pmd_young() for architectures not having it
mm/damon/sysfs: fix wrong empty schemes assumption under online tuning in damon_sysfs_set_schemes()
tools/vm/slabinfo-gnuplot: use "grep -E" instead of "egrep"
nilfs2: fix NULL pointer dereference in nilfs_palloc_commit_free_entry()
hugetlb: don't delete vma_lock in hugetlb MADV_DONTNEED processing
madvise: use zap_page_range_single for madvise dontneed
mm: replace VM_WARN_ON to pr_warn if the node is offline with __GFP_THISNODE
The V4L2_MEMORY_USERPTR interface is long deprecated and shouldn't be
used (and is discouraged for any modern v4l drivers). And Seth Jenkins
points out that the fallback to VM_PFNMAP/VM_IO is fundamentally racy
and dangerous.
Note that it's not even a case that should trigger, since any normal
user pointer logic ends up just using the pin_user_pages_fast() call
that does the proper page reference counting. That's not the problem
case, only if you try to use special device mappings do you have any
issues.
Normally I'd just remove this during the merge window, but since Seth
pointed out the problem cases, we really want to know as soon as
possible if there are actually any users of this odd special case of a
legacy interface. Neither Hans nor Mauro seem to think that such
mis-uses of the old legacy interface should exist. As Mauro says:
"See, V4L2 has actually 4 streaming APIs:
- Kernel-allocated mmap (usually referred simply as just mmap);
- USERPTR mmap;
- read();
- dmabuf;
The USERPTR is one of the oldest way to use it, coming from V4L
version 1 times, and by far the least used one"
And Hans chimed in on the USERPTR interface:
"To be honest, I wouldn't mind if it goes away completely, but that's a
bit of a pipe dream right now"
but while removing this legacy interface entirely may be a pipe dream we
can at least try to remove the unlikely (and actively broken) case of
using special device mappings for USERPTR accesses.
This replaces it with a WARN_ONCE() that we can remove once we've
hopefully confirmed that no actual users exist.
NOTE! Longer term, this means that a 'struct frame_vector' only ever
contains proper page pointers, and all the games we have with converting
them to pages can go away (grep for 'frame_vector_to_pages()' and the
uses of 'vec->is_pfns'). But this is just the first step, to verify
that this code really is all dead, and do so as quickly as possible.
Reported-by: Seth Jenkins <sethjenkins@google.com>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
Acked-by: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull NVMe fixes from Christoph:
"nvme fixes for Linux 6.1
- fix SRCU protection of nvme_ns_head list (Caleb Sander)
- clear the prp2 field when not used (Lei Rao)"
* tag 'nvme-6.1-2022-01-02' of git://git.infradead.org/nvme:
nvme: fix SRCU protection of nvme_ns_head list
nvme-pci: clear the prp2 field when not used
for_each_pci_dev() is implemented by pci_get_device(). The comment of
pci_get_device() says that it will increase the reference count for the
returned pci_dev and also decrease the reference count for the input
pci_dev @from if it is not NULL.
If we break for_each_pci_dev() loop with pdev not NULL, we need to call
pci_dev_put() to decrease the reference count. Add the missing
pci_dev_put() for the error path to avoid reference count leak.
Fixes: 2e45528930 ("iommu/vt-d: Unify the way to process DMAR device scope array")
Signed-off-by: Xiongfeng Wang <wangxiongfeng2@huawei.com>
Link: https://lore.kernel.org/r/20221121113649.190393-3-wangxiongfeng2@huawei.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
