Running cpufreq-hw driver on Lenovo Yoga C630 laptop, the following
warning messages will be seen.
[ 3.415340] cpu cpu4: Voltage update failed freq=2841600
[ 3.418755] cpu cpu4: failed to update OPP for freq=2841600
[ 3.422949] cpu cpu4: Voltage update failed freq=2956800
[ 3.427086] cpu cpu4: failed to update OPP for freq=2956800
This is because the cpufreq-hw lookup table of SDM850 provides these two
set-points, but they are missing from OPP table in DT. Let's create
sdm850.dtsi to add the OPP for them, so that the warning will be gone.
Signed-off-by: Steev Klimaszewski <steev@kali.org>
Signed-off-by: Shawn Guo <shawn.guo@linaro.org>
Link: https://lore.kernel.org/r/20210112090640.20062-1-shawn.guo@linaro.org
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
This commit adds support for deep idling of the entire unified DynamIQ
CPU cluster on sm8150. In this idle state, the LLCC (Last-Level Cache
Controller) is powered off and the AOP (Always-On Processor) enters a
low-power sleep state.
I'm not sure what the per-CPU 0x400000f4 idle state previously
contributed by Qualcomm as the "cluster sleep" state is, but the
downstream kernel has no such state. The real deep cluster idle state
is 0x41000c244, composed of:
Cluster idle state: (0xc24) << 4 = 0xc240
Is reset state: 1 << 30 = 0x40000000
Affinity level: 1 << 24 = 0x1000000
CPU idle state: 0x4 (power collapse)
This setup can be replicated with the PSCI power domain cpuidle driver,
which utilizes OSI to enter cluster idle when the last active CPU
enters idle.
The cluster idle state cannot be used as a plain cpuidle state because
it requires that all CPUs in the cluster are idling.
Reviewed-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Danny Lin <danny@kdrag0n.dev>
Link: https://lore.kernel.org/r/20210105201000.913183-1-danny@kdrag0n.dev
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
For a bunch of rails we really don't do anything with them in Linux.
These are things like modem voltage rails that the modem manages these
itself and core rails (like IO rails) that are setup to just
automagically do the right thing by the firmware.
Let's stop even listing those rails in our device tree.
The net result of this is that some of these rails might be able to go
down to a lower voltage or perhaps transition to LPM (low power mode)
sometimes.
Here's a list of what we're doing and why:
* L1A - only goes to SoC and doesn't seem associated with any
particular peripheral. Kernel isn't doing anything with
this. Removing from dts. NET IMPACT: rail might drop from 1.2V to
1.178V and switch to LPM in some cases depending on firmware.
* L2A - only goes to SoC and doesn't seem associated with any
particular peripheral. Kernel isn't doing anything with
this. Removing from dts. NET IMPACT: rail might switch to LPM in
some cases depending on firmware.
* L3A - only goes to SoC and doesn't seem associated with any
particular peripheral. Kernel isn't doing anything with
this. Removing from dts. NET IMPACT: rail might switch to LPM in
some cases depending on firmware.
* L5A - seems to be totally unused as far as I can tell and doesn't
even come off QSIP. Removing from dts.
* L6A - only goes to SoC and doesn't seem associated with any
particular peripheral (I think?). Kernel isn't doing anything with
this. Removing from dts. NET IMPACT: rail might switch to LPM in
some cases depending on firmware.
* L16A - Looks like this is only used for internal RF stuff. Removing
from dts. NET IMPACT: rail might switch to LPM in some cases
depending on firmware.
* L1C - Just goes to WiFi / Bluetooth. Trust how IDP has this set and
put this back at 1.616V min.
* L4C - This goes out to the eSIM among other places. This looks like
it's intended to be for SIM card and modem manages. NET IMPACT:
rail might switch to LPM in some cases depending on firmware.
* L5C - This goes to the physical SIM. This looks like it's intended
to be for SIM card and modem manages. NET IMPACT: rail might drop
from 1.8V to 1.648V and switch to LPM in some cases depending on
firmware.
NOTE: in general for anything which is supposed to be managed by Linux
I still left it all forced to HPM since I'm not 100% sure that all the
needed calls to regulator_set_load() are in place and HPM is safer.
Switching more things to LPM can happen in a future patch.
ALSO NOTE: Power measurements showed no measurable difference after
applying this patch, so perhaps it should be viewed more as a cleanup
than any power savings.
Reviewed-by: Alexandru M Stan <amstan@google.com>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20201207143255.1.Ib92ec35163682dec4b2fbb4bde0785cb6e6dde27@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Unlike most MSM8916 boards, samsung-a5u uses WCN3660B instead of
WCN3620 to support the 5 GHz band additionally.
WCN3660B has similar requirements as WCN3620, but it needs the XO
clock to run at 48 MHz instead of 19.2 MHz. So far it was possible
to describe that configuration using the qcom,wcn3680 compatible.
However, as of commit 8490987bdb ("wcn36xx: Hook and identify RF_IRIS_WCN3680"),
the wcn36xx driver will now use the qcom,wcn3680 compatible
to enable functionality specific to WCN3680. In particular,
WCN3680 supports 802.11ac, which is not available in WCN3660B.
Use the new qcom,wcn3660b compatible to describe the chip properly.
Fixes: 0d70519991 ("arm64: dts: msm8916-samsung-a5u: Override iris compatible")
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Link: https://lore.kernel.org/r/20210106102134.59801-4-stephan@gerhold.net
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Like other Qualcomm SoCs, sm8150 exposes CPU and cluster idle states
through PSCI. Define the idle states to save power when the CPU is not
in active use.
These idle states, latency, and residency values match the downstream
4.14 kernel from Qualcomm as of LA.UM.8.1.r1-15600-sm8150.0.
It's worth noting that the CPU has an additional C3 power collapse idle
state between WFI and rail power collapse (with PSCI mode 0x40000003),
but it is not officially used in downstream kernels due to "thermal
throttling issues."
Signed-off-by: Danny Lin <danny@kdrag0n.dev>
Link: https://lore.kernel.org/r/20201221002907.2870059-3-danny@kdrag0n.dev
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
It never makes sense to set the IO voltage of the SD card (vqmmc) to a
voltage that's higher than the voltage of the card's main power supply
(vmmc). The card's main voltage is 2.952V on trogdor, so let's set
the max for the IO voltage to the same.
NOTE: On Linux, this is pretty much a no-op currently. Linux already
makes an effort to match vqmmc with vmmc when running at "3.3" signal
voltage, so both before and after this change we end up running vqmmc
at 2.904V when talking to non-UHS cards. It still seems cleaner to
make it a little more correct, though.
Also note: as per above, on Linux right now we end up running vqmmc as
2.904V even though vmmc is 2.952V. This isn't super ideal but
shouldn't really hurt.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20201204104900.1.I0a4ac2c7f4d405431cf95eb7b7c36800660516ec@changeid
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Since commit 36e2c7421f ("fs: don't allow splice read/write without
explicit ops") we've required that file operation structures explicitly
enable splice support, rather than falling back to the default handlers.
Most /proc files use the indirect 'struct proc_ops' to describe their
file operations, and were fixed up to support splice earlier in commits
40be821d627c..b24c30c67863, but the mountinfo files interact with the
VFS directly using their own 'struct file_operations' and got missed as
a result.
This adds the necessary support for splice to work for /proc/*/mountinfo
and friends.
Reported-by: Joan Bruguera Micó <joanbrugueram@gmail.com>
Reported-by: Jussi Kivilinna <jussi.kivilinna@iki.fi>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=209971
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull NTB fixes from Jon Mason:
"Bug fix for IDT NTB and Intel NTB LTR management support"
* tag 'ntb-5.11' of git://github.com/jonmason/ntb:
ntb: intel: add Intel NTB LTR vendor support for gen4 NTB
ntb: idt: fix error check in ntb_hw_idt.c
Pull crypto fixes from Herbert Xu:
"Fix a number of autobuild failures due to missing Kconfig
dependencies"
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
crypto: qat - add CRYPTO_AES to Kconfig dependencies
crypto: keembay - Add dependency on HAS_IOMEM
crypto: keembay - CRYPTO_DEV_KEEMBAY_OCS_AES_SM4 should depend on ARCH_KEEMBAY
Pull objtool fix from Ingo Molnar:
"Fix a segfault that occurs when built with Clang"
* tag 'objtool-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
objtool: Fix seg fault with Clang non-section symbols
Pull timer fixes from Ingo Molnar:
"Update/fix two CPU sanity checks in the hotplug and the boot code, and
fix a typo in the Kconfig help text.
[ Context: the first two commits are the result of an ongoing
annotation+review work of (intentional) tick_do_timer_cpu() data
races reported by KCSAN, but the annotations aren't fully cooked
yet ]"
* tag 'timers-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timekeeping: Fix spelling mistake in Kconfig "fullfill" -> "fulfill"
tick/sched: Remove bogus boot "safety" check
tick: Remove pointless cpu valid check in hotplug code
Pull scheduler fix from Ingo Molnar:
"Fix a context switch performance regression"
* tag 'sched-urgent-2020-12-27' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched: Optimize finish_lock_switch()
Commit c9a3c4e637 ("mfd: ab8500-debugfs: Remove extraneous curly
brace") removed a left-over curly brace that caused build failures, but
Joe Perches points out that the subsequent 'seq_putc()' should also be
removed, because the commit that caused all these problems already added
the final '\n' to the seq_printf() above it.
Reported-by: Joe Perches <joe@perches.com>
Fixes: 886c812165 ("mfd: ab8500-debugfs: Remove the racy fiddling with irq_desc")
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull PCI fixes from Bjorn Helgaas:
- Fix a tegra enumeration regression (Rob Herring)
- Fix a designware-host check that warned on *success*, not failure
(Alexander Lobakin)
* tag 'pci-v5.11-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:
PCI: dwc: Fix inverted condition of DMA mask setup warning
PCI: tegra: Fix host link initialization
Clang errors:
drivers/mfd/ab8500-debugfs.c:1526:2: error: non-void function does not return a value [-Werror,-Wreturn-type]
}
^
drivers/mfd/ab8500-debugfs.c:1528:2: error: expected identifier or '('
return 0;
^
drivers/mfd/ab8500-debugfs.c:1529:1: error: extraneous closing brace ('}')
}
^
3 errors generated.
The cleanup in ab8500_interrupts_show left a curly brace around, remove
it to fix the error.
Fixes: 886c812165 ("mfd: ab8500-debugfs: Remove the racy fiddling with irq_desc")
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit 660c486590 ("PCI: dwc: Set 32-bit DMA mask for MSI target address
allocation") added dma_mask_set() call to explicitly set 32-bit DMA mask
for MSI message mapping, but for now it throws a warning on ret == 0, while
dma_set_mask() returns 0 in case of success.
Fix this by inverting the condition.
[bhelgaas: join string to make it greppable]
Fixes: 660c486590 ("PCI: dwc: Set 32-bit DMA mask for MSI target address allocation")
Link: https://lore.kernel.org/r/20201222150708.67983-1-alobakin@pm.me
Signed-off-by: Alexander Lobakin <alobakin@pm.me>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Commit b9ac0f9dc8 ("PCI: dwc: Move dw_pcie_setup_rc() to DWC common
code") broke enumeration of downstream devices on Tegra:
In non-working case (next-20201211):
0001:00:00.0 PCI bridge: NVIDIA Corporation Device 1ad2 (rev a1)
0001:01:00.0 SATA controller: Marvell Technology Group Ltd. Device 9171 (rev 13)
0005:00:00.0 PCI bridge: NVIDIA Corporation Device 1ad0 (rev a1)
In working case (v5.10-rc7):
0001:00:00.0 PCI bridge: Molex Incorporated Device 1ad2 (rev a1)
0001:01:00.0 SATA controller: Marvell Technology Group Ltd. Device 9171 (rev 13)
0005:00:00.0 PCI bridge: Molex Incorporated Device 1ad0 (rev a1)
0005:01:00.0 PCI bridge: PLX Technology, Inc. Device 3380 (rev ab)
0005:02:02.0 PCI bridge: PLX Technology, Inc. Device 3380 (rev ab)
0005:03:00.0 USB controller: PLX Technology, Inc. Device 3380 (rev ab)
The problem seems to be dw_pcie_setup_rc() is now called twice before and
after the link up handling. The fix is to move Tegra's link up handling to
.start_link() function like other DWC drivers. Tegra is a bit more
complicated than others as it re-inits the whole DWC controller to retry
the link. With this, the initialization ordering is restored to match the
prior sequence.
Fixes: b9ac0f9dc8 ("PCI: dwc: Move dw_pcie_setup_rc() to DWC common code")
Link: https://lore.kernel.org/r/20201218143905.1614098-1-robh@kernel.org
Reported-by: Mian Yousaf Kaukab <ykaukab@suse.de>
Tested-by: Mian Yousaf Kaukab <ykaukab@suse.de>
Signed-off-by: Rob Herring <robh@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Jonathan Hunter <jonathanh@nvidia.com>
Cc: Vidya Sagar <vidyas@nvidia.com>
clang (quite rightly) complains fairly loudly about the newly added
mpc1_get_mpc_out_mux() function returning an uninitialized value if the
'opp_id' checks don't pass.
This may not happen in practice, but the code really shouldn't return
garbage if the sanity checks don't pass.
So just initialize 'val' to zero to avoid the issue.
Fixes: 110b055b28 ("drm/amd/display: add getter routine to retrieve mpcc mux")
Cc: Josip Pavic <Josip.Pavic@amd.com>
Cc: Bindu Ramamurthy <bindu.r@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull more perf tools updates from Arnaldo Carvalho de Melo:
- Refactor 'perf stat' per CPU/socket/die/thread aggregation fixing use
cases in ARM machines.
- Fix memory leak when synthesizing SDT probes in 'perf probe'.
- Update kernel header copies related to KVM, epol_pwait. msr-index and
powerpc and s390 syscall tables.
* tag 'perf-tools-2020-12-24' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux: (24 commits)
perf probe: Fix memory leak when synthesizing SDT probes
perf stat aggregation: Add separate thread member
perf stat aggregation: Add separate core member
perf stat aggregation: Add separate die member
perf stat aggregation: Add separate socket member
perf stat aggregation: Add separate node member
perf stat aggregation: Start using cpu_aggr_id in map
perf cpumap: Drop in cpu_aggr_map struct
perf cpumap: Add new map type for aggregation
perf stat: Replace aggregation ID with a struct
perf cpumap: Add new struct for cpu aggregation
perf cpumap: Use existing allocator to avoid using malloc
perf tests: Improve topology test to check all aggregation types
perf tools: Update s390's syscall.tbl copy from the kernel sources
perf tools: Update powerpc's syscall.tbl copy from the kernel sources
perf s390: Move syscall.tbl check into check-headers.sh
perf powerpc: Move syscall.tbl check to check-headers.sh
tools headers UAPI: Synch KVM's svm.h header with the kernel
tools kvm headers: Update KVM headers from the kernel sources
tools headers UAPI: Sync KVM's vmx.h header with the kernel sources
...
Pull coccinelle updates from Julia Lawall.
* 'for-5.11' of git://git.kernel.org/pub/scm/linux/kernel/git/jlawall/linux:
scripts: coccicheck: Correct usage of make coccicheck
coccinelle: update expiring email addresses
coccinnelle: Remove ptr_ret script
kbuild: do not use scripts/ld-version.sh for checking spatch version
remove boolinit.cocci
Commit 64a1b95bb9 ("genirq: Restrict export of irq_to_desc()") removed
the export of irq_to_desc() unless powerpc KVM is being built, because
there is still a use of irq_to_desc() in modular code there.
However it used:
#ifdef CONFIG_KVM_BOOK3S_64_HV
Which doesn't work when that symbol is =m, leading to a build failure:
ERROR: modpost: "irq_to_desc" [arch/powerpc/kvm/kvm-hv.ko] undefined!
Fix it by checking for the definedness of the correct symbol which is
CONFIG_KVM_BOOK3S_64_HV_MODULE.
Fixes: 64a1b95bb9 ("genirq: Restrict export of irq_to_desc()")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull misc vfs updates from Al Viro:
"Assorted patches from previous cycle(s)..."
* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fix hostfs_open() use of ->f_path.dentry
Make sure that make_create_in_sticky() never sees uninitialized value of dir_mode
fs: Kill DCACHE_DONTCACHE dentry even if DCACHE_REFERENCED is set
fs: Handle I_DONTCACHE in iput_final() instead of generic_drop_inode()
fs/namespace.c: WARN if mnt_count has become negative
Pull documentation fixes from Jonathan Corbet:
"A small set of late-arriving, small documentation fixes"
* tag 'docs-5.11-2' of git://git.lwn.net/linux:
docs: admin-guide: Fix default value of max_map_count in sysctl/vm.rst
Documentation/submitting-patches: Document the SoB chain
Documentation: process: Correct numbering
docs: submitting-patches: Trivial - fix grammatical error
