aboutsummaryrefslogtreecommitdiffstats
path: root/arch
AgeCommit message (Collapse)AuthorFilesLines
2020-09-15KVM: arm64: Choose hyp symbol based on contextAndrew Scull1-6/+19
Make CHOOSE_HYP_SYM select the symbol of the active hypervisor for the host, the nVHE symbol for nVHE and the VHE symbol for VHE. The nVHE and VHE hypervisors see their own symbols without prefixes and trigger a link error when trying to use a symbol of the other hypervisor. Signed-off-by: Andrew Scull <ascull@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Cc: David Brazdil <dbrazdil@google.com> Link: https://lore.kernel.org/r/20200915104643.2543892-5-ascull@google.com
2020-09-15KVM: arm64: Remove kvm_host_data_t typedefAndrew Scull2-5/+3
The kvm_host_data_t typedef is used inconsistently and goes against the kernel's coding style. Remove it in favour of the full struct specifier. Signed-off-by: Andrew Scull <ascull@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20200915104643.2543892-4-ascull@google.com
2020-09-15KVM: arm64: Remove hyp_panic argumentsAndrew Scull5-16/+15
hyp_panic is able to find all the context it needs from within itself so remove the argument. The __hyp_panic wrapper becomes redundant so is also removed. Signed-off-by: Andrew Scull <ascull@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20200915104643.2543892-3-ascull@google.com
2020-09-15KVM: arm64: Remove __activate_vm wrapperAndrew Scull3-14/+9
The __activate_vm wrapper serves no useful function and has a misleading name as it simply calls __load_guest_stage2 and does not touch HCR_EL2.VM so remove it. Also rename __deactivate_vm to __load_host_stage2 to match naming pattern. Signed-off-by: Andrew Scull <ascull@google.com> Signed-off-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/20200915104643.2543892-2-ascull@google.com
2020-09-15arm64: dts: qcom: Makefile: Sort linesStephan Gerhold1-5/+5
The Makefile is in a bit of a weird order at the moment. It's almost sorted alphabetically, but not entirely. Also, one element uses a space before the += instead of a tab. Fix this up and sort the lines alphabetically so we have a consistent order in the Makefile. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-15-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: pm8916: Sort nodesStephan Gerhold1-30/+30
Sort nodes by unit address so we have a consistent order in pm8916.dtsi. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-14-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Sort nodesStephan Gerhold1-1107/+1107
Just like in commit 50aa72ccb30b ("arm64: dts: qcom: msm8996: Sort all nodes in msm8996.dtsi"), sort all the nodes by unit address, then alphabetically by their name. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-13-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Pad addressesStephan Gerhold1-69/+69
Just like in commit 86f6d6225e5e ("arm64: dts: qcom: msm8996: Pad addresses"), pad all addresses to 8 digits to make it easier to see the correct order of the nodes. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-12-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Rename "x-smp2p" to "smp2p-x"Stephan Gerhold1-2/+2
This allows grouping them together when sorting nodes alphabetically. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-11-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Use more generic node namesStephan Gerhold3-17/+17
Now that all MSM8916 boards are referencing nodes by label instead of name, we can easily make some more nodes use more generic names (as recommended in the device tree specification or the binding documentation). Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-10-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Add MSM8916-specific compatibles to SCM/MSSStephan Gerhold1-2/+2
Over the time, the SCM and MSS driver were refactored to use SoC-specific compatibles. While the generic compatibles still work correctly, add the MSM8916-specific compatibles so they are actually used somewhere. For SCM this will ensure that we actually manage to obtain all three of the specified clocks, since those are required on MSM8916. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-9-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Minor style fixesStephan Gerhold1-13/+13
Fix usages of spaces for indentation, break a long line and remove duplicate new lines. Add some spaces where appropriate. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-8-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Drop qcom,tcsr-mutex sysconStephan Gerhold1-8/+3
The hwlock device node does not (directly) use memory resources of the SoC, so we should move it outside the "soc" node. However, as of commit 7a1e6fb1c606 ("hwspinlock: qcom: Allow mmio usage in addition to syscon") we can now assign the memory region directly to the hwlock device node. This works because the register space used by it is actually separate and not used by any other components. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-7-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Use IRQ defines, add IRQ typesStephan Gerhold1-18/+20
dt-bindings/interrupt-controller/arm-gic.h has a GIC_SPI define that allows specifying interrupts more clearly, but right now only some device nodes in msm8916.dtsi make use of it. Convert all others to use it. The same applies to the IRQ_TYPE_* defines in dt-bindings/interrupt-controller/irq.h. Some interrupts were defined with raw numbers, or even with IRQ_TYPE_NONE (0). Convert all these to use appropriate IRQ types. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-6-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Fix MDP/DSI interruptsStephan Gerhold1-2/+2
The mdss node sets #interrupt-cells = <1>, so its interrupts should be referenced using a single cell (in this case: only the interrupt number). However, right now the mdp/dsi node both have two interrupt cells set, e.g. interrupts = <4 0>. The 0 is probably meant to say IRQ_TYPE_NONE (= 0), but with #interrupt-cells = <1> this is actually interpreted as a second interrupt line. Remove the IRQ flags from both interrupts to fix this. Fixes: 305410ffd1b2 ("arm64: dts: msm8916: Add display support") Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-5-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: pm8916: Remove invalid reg size from wcd_codecStephan Gerhold1-1/+1
Tha parent node of "wcd_codec" specifies #address-cells = <1> and #size-cells = <0>, which means that each resource should be described by one cell for the address and size omitted. However, wcd_codec currently lists 0x200 as second cell (probably the size of the resource). When parsing this would be treated like another memory resource - which is entirely wrong. To quote the device tree specification [1]: "If the parent node specifies a value of 0 for #size-cells, the length field in the value of reg shall be omitted." [1]: https://www.devicetree.org/specifications/ Fixes: 5582fcb3829f ("arm64: dts: apq8016-sbc: add analog audio support with multicodec") Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-4-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Remove one more thermal trip point unit nameStephan Gerhold1-3/+3
Commit fe2aff0c574d2 ("arm64: dts: qcom: msm8916: remove unit name for thermal trip points") removed the unit names for most of the thermal trip points defined in msm8916.dtsi, but missed to update the one for cpu0_1-thermal. So why wasn't this spotted by "make dtbs_check"? Apparently, the name of the thermal zone is already invalid: thermal-zones.yaml specifies a regex of ^[a-zA-Z][a-zA-Z0-9\\-]{1,12}-thermal$, so it is not allowed to contain underscores. Therefore the thermal zone was never verified using the DTB schema. After replacing the underscore in the thermal zone name, the warning shows up: apq8016-sbc.dt.yaml: thermal-zones: cpu0-1-thermal:trips: 'trip-point@0' does not match any of the regexes: '^[a-zA-Z][a-zA-Z0-9\\-_]{0,63}$', 'pinctrl-[0-9]+' Fix up the thermal zone names and remove the unit name for the trip point. Cc: Amit Kucheria <amit.kucheria@linaro.org> Fixes: fe2aff0c574d2 ("arm64: dts: qcom: msm8916: remove unit name for thermal trip points") Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-3-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8916: Configure DSI port with labelsStephan Gerhold2-17/+7
&dsi0 -> ports -> port@1 -> endpoint already has the "dsi0_out" label, so we can use it for configuring instead of replicating the entire node hierarchy. Looks like I missed that when converting the boards. Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Link: https://lore.kernel.org/r/20200915071221.72895-2-stephan@gerhold.net Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: sm8250: Add OPP table for all qup devicesDmitry Baryshkov1-0/+69
qup has a requirement to vote on the performance state of the CX domain in sm8250 devices. Add OPP tables for these and also add power-domains property for all qup instances for uart and spi. i2c does not support scaling and uses a fixed clock. Acked-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Link: https://lore.kernel.org/r/20200915120203.290295-1-dmitry.baryshkov@linaro.org Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15arm64: dts: qcom: msm8996: Add VFE1_GDSC power domain to camss nodeRobert Foss1-1/+2
As the MSM8996 has two VFE IP-blocks, and each has a power domain, both of them have to be enabled. Previously only the power domain of VFE0 was enabled, but not the domain for VFE1. This patch adds the VFE1_GDSC power domain to the camss device tree node of the MSM8996 soc. Signed-off-by: Robert Foss <robert.foss@linaro.org> Link: https://lore.kernel.org/r/20200915142316.147208-1-robert.foss@linaro.org Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
2020-09-15ARM: add malloc size to decompressor kexec size structureRussell King3-4/+8
Add the required malloc size to the decompressor kexec size structure. Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: add TEXT_OFFSET to decompressor kexec image structureRussell King3-1/+6
Add the TEXT_OFFSET to the decompressor's kexec image structure to kexec knows what offset to use. Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9007/1: l2c: fix prefetch bits init in L2X0_AUX_CTRL using DT valuesGuillaume Tucker1-4/+12
The L310_PREFETCH_CTRL register bits 28 and 29 to enable data and instruction prefetch respectively can also be accessed via the L2X0_AUX_CTRL register. They appear to be actually wired together in hardware between the registers. Changing them in the prefetch register only will get undone when restoring the aux control register later on. For this reason, set these bits in both registers during initialisation according to the devicetree property values. Link: https://lore.kernel.org/lkml/76f2f3ad5e77e356e0a5b99ceee1e774a2842c25.1597061474.git.guillaume.tucker@collabora.com/ Fixes: ec3bd0e68a67 ("ARM: 8391/1: l2c: add options to overwrite prefetching behavior") Signed-off-by: Guillaume Tucker <guillaume.tucker@collabora.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9010/1: uncompress: Print the location of appended DTBLinus Walleij1-0/+23
When using the kernel with an appended DTB it is useful to know where this will end up in the physical memory at the time the kernel boots. We add a debug print macro that will help out with this. Here is a sample debug print after passing -DDEBUG to head.S during compilation: DTB:0x40CEBA70 (0x000051B5) C:0x402080C0-0x40CF0CE0->0x41801D00-0x422EA920 DTB:0x422E56B0 (0x00005262) This means that the appended DTB is first found after the compressed kernel at 0x40CEBA70 of size 0x51B5 and then after the compressed kernel is moved to 0x41801D00 it is found again at 0x422E56B0 and is there size 0x5262. The growth in size of the FDT is due to the call to atags_to_fdt() that augments the DTB with ATAG information. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9009/1: uncompress: Enable debug in head.SLinus Walleij1-0/+1
The assembly file head.S includes some debug code that does not get enabled when we select CONFIG_DEBUG_UNCOMPRESS. The debug in head.S relies on the user tagging on -DDEBUG on the compilation command line. To simplify debugging, tag on -DDEBUG so that we also get these debug messages when selecting CONFIG_DEBUG_UNCOMPRESS. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9008/1: uncompress: Drop excess whitespace printLinus Walleij1-2/+0
This drops some whitespace from the debug message about where we move the compressed kernel: r after the message is completely surplus since the putc routine will anyway add r after n, and the initial linefeed just assumes that this will always be the first message on the console, which is not certain to be true. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9006/1: uncompress: Wait for ready and busy in debug printsLinus Walleij1-5/+10
For some platforms such as Qualcomm we need to wait for the UART to be ready before writing characters to the UART in the same manner as the macro in debug.S used with the main "Uncompressing Linux ..." text. Pass an extra temporary variable to writeb and make it call waituarttxrdy and busyuart just like the other decomression messages. Optionally it will also call waituartcts if and only if CONFIG_DEBUG_UART_FLOW_CONTROL is selected. After this the decompression debug messages work fine on Qualcomm platforms if you compile head.S with -DDEBUG. Cc: Nicolas Pitre <nico@fluxnic.net> Cc: Fabrizio Castro <fabrizio.castro@bp.renesas.com> Cc: Ard Biesheuvel <ardb@kernel.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9005/1: debug: Select flow control for all debug UARTsLinus Walleij5-10/+19
Instead of a flow control selection mechanism specifically for 8250, make this available for all debug UARTs. If the debug UART supports waiting for CTS to be asserted, then this code can be activated for terminals that need it. We keep the defaults for EBSA110, Footbridge, Gemini and RPC so that this still works as expected for these older platforms: they assume that flow control shall be enabled for debug prints. I switch the location of the check for ifdef CONFIG_DEBUG_UART_FLOW_CONTROL from the actual debug UART drivers: the code would get compiled-out for 8250 and Tegra unless their custom config (or passing -DFLOW_CONTROL in the Tegra case) was not set. Instead this is conditional at the three places where we print debug messages. The idea is that debug UARTs can be implemented without this ifdef boilerplate so they look cleaner, alas the ifdef has to be somewhere. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9004/1: debug: Split waituart to CTS and TXRDYLinus Walleij27-30/+114
This patch was triggered by a remark from Russell that introducing a call to the waituart (needed to fix debug prints on the Qualcomm platforms) was dangerous because in some cases this will involve waiting for a modem CTS (clear to send) signal, and debug messages would maybe not work on platforms with no modem connected to the UART port: they will just hang waiting for the modem to assert CTS and this might never happen. Looking through all UART debug drivers implementing the waituart macro I discovered that all users except two actually use this macro to check if the UART is ready for TX, let's call this TXRDY. Only two debug UART drivers actually check for CTS: - arch/arm/include/debug/8250.S - arch/arm/include/debug/tegra.S The former is very significant since the 8250 is possibly the most common UART on the planet. We have the following problem: the semantics of waituart are ambiguous making it dangerous to introduce the macro to debug code fixing debug prints for Qualcomm. To start to pry this problem apart, this patch does the following: - Convert all debug UART drivers to define two macros: - waituartcts with the clear semantic to wait for CTS to be asserted - waituarttxrdy with the clear semantic to wait for the TX capability of the UART to be ready - When doing this take care to assign the right function to each drivers macro, so they now do exactly the above. - Update the three sites in the kernel invoking the waituart macro to call waituartcts/waituarttxrdy in sequence, so that the functional impact on the kernel should be zero. After this we can start to change the code sites using this code to do the right thing. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 9003/1: uncompress: Delete unused debug macrosLinus Walleij1-30/+0
The debug macros debug_reloc_start and debug_reloc_end were rendered unused in commit 6d7d0ae51574943bf571d269da3243257a2d15db "ARM: 6750/1: improvements to compressed/head.S". Later on a different debug macro named dbgkc was introduced in commit f3c899927e19d1be39818145efc39ea27b8efc69 "ARM: 8786/1: Debug kernel copy by printing". Delete the dead debug code. Cc: Nicolas Pitre <nico@fluxnic.net> Cc: Fabrizio Castro <fabrizio.castro@bp.renesas.com> Cc: Ard Biesheuvel <ardb@kernel.org> Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15ARM: 8997/2: hw_breakpoint: Handle inexact watchpoint addressesDouglas Anderson1-28/+72
This is commit fdfeff0f9e3d ("arm64: hw_breakpoint: Handle inexact watchpoint addresses") but ported to arm32, which has the same problem. This problem was found by Android CTS tests, notably the "watchpoint_imprecise" test [1]. I tested locally against a copycat (simplified) version of the test though. [1] https://android.googlesource.com/platform/bionic/+/master/tests/sys_ptrace_test.cpp Link: https://lkml.kernel.org/r/20191019111216.1.I82eae759ca6dc28a245b043f485ca490e3015321@changeid Signed-off-by: Douglas Anderson <dianders@chromium.org> Reviewed-by: Matthias Kaehlcke <mka@chromium.org> Acked-by: Will Deacon <will@kernel.org> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
2020-09-15powerpc/pci: unmap legacy INTx interrupts when a PHB is removedCédric Le Goater2-0/+120
When a passthrough IO adapter is removed from a pseries machine using hash MMU and the XIVE interrupt mode, the POWER hypervisor expects the guest OS to clear all page table entries related to the adapter. If some are still present, the RTAS call which isolates the PCI slot returns error 9001 "valid outstanding translations" and the removal of the IO adapter fails. This is because when the PHBs are scanned, Linux maps automatically the INTx interrupts in the Linux interrupt number space but these are never removed. To solve this problem, we introduce a PPC platform specific pcibios_remove_bus() routine which clears all interrupt mappings when the bus is removed. This also clears the associated page table entries of the ESB pages when using XIVE. For this purpose, we record the logical interrupt numbers of the mapped interrupt under the PHB structure and let pcibios_remove_bus() do the clean up. Since some PCI adapters, like GPUs, use the "interrupt-map" property to describe interrupt mappings other than the legacy INTx interrupts, we can not restrict the size of the mapping array to PCI_NUM_INTX. The number of interrupt mappings is computed from the "interrupt-map" property and the mapping array is allocated accordingly. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20200807101854.844619-1-clg@kaod.org
2020-09-15powerpc/powernv/idle: add a basic stop 0-3 driver for POWER10Nicholas Piggin4-96/+211
This driver does not restore stop > 3 state, so it limits itself to states which do not lose full state or TB. The POWER10 SPRs are sufficiently different from P9 that it seems easier to split out the P10 code. The POWER10 deep sleep code (e.g., the BHRB restore) has been taken out, but it can be re-added when stop > 3 support is added. Signed-off-by: Nicholas Piggin <npiggin@gmail.com> Tested-by: Pratik Rajesh Sampat<psampat@linux.ibm.com> Tested-by: Vaidyanathan Srinivasan <svaidy@linux.ibm.com> Reviewed-by: Pratik Rajesh Sampat<psampat@linux.ibm.com> Reviewed-by: Gautham R. Shenoy <ego@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20200819094700.493399-1-npiggin@gmail.com
2020-09-15powerepc/book3s64/hash: Align start/end address correctly with bolt mappingAneesh Kumar K.V2-3/+10
This ensures we don't do a partial mapping of memory. With nvdimm, when creating namespaces with size not aligned to 16MB, the kernel ends up partially mapping the pages. This can result in kernel adding multiple hash page table entries for the same range. A new namespace will result in create_section_mapping() with start and end overlapping an already existing bolted hash page table entry. commit: 6acd7d5ef264 ("libnvdimm/namespace: Enforce memremap_compat_align()") made sure that we always create namespaces aligned to 16MB. But we can do better by avoiding mapping pages that are not aligned. This helps to catch access to these partially mapped pages early. Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20200907072539.67310-1-aneesh.kumar@linux.ibm.com
2020-09-15powerpc/ps3: make two symbols staticJason Yan1-2/+2
This addresses the following sparse warning: arch/powerpc/platforms/ps3/spu.c:451:33: warning: symbol 'spu_management_ps3_ops' was not declared. Should it be static? arch/powerpc/platforms/ps3/spu.c:592:28: warning: symbol 'spu_priv1_ps3_ops' was not declared. Should it be static? Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: Jason Yan <yanaijie@huawei.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/20200911020121.1464585-1-yanaijie@huawei.com
2020-09-15powerpc/kasan: Fix CONFIG_KASAN_VMALLOC for 8xxChristophe Leroy1-3/+9
Before the commit identified below, pages tables allocation was performed after the allocation of final shadow area for linear memory. But that commit switched the order, leading to page tables being already allocated at the time 8xx kasan_init_shadow_8M() is called. Due to this, kasan_init_shadow_8M() doesn't map the needed shadow entries because there are already page tables. kasan_init_shadow_8M() installs huge PMD entries instead of page tables. We could at that time free the page tables, but there is no point in creating page tables that get freed before being used. Only book3s/32 hash needs early allocation of page tables. For other variants, we can keep the initial order and create remaining page tables after the allocation of final shadow memory for linear mem. Move back the allocation of shadow page tables for CONFIG_KASAN_VMALLOC into kasan_init() after the loop which creates final shadow memory for linear mem. Fixes: 41ea93cf7ba4 ("powerpc/kasan: Fix shadow pages allocation failure") Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/8ae4554357da4882612644a74387ae05525b2aaa.1599800716.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/powermac: Fix low_sleep_handler with KUAP and KUEPChristophe Leroy2-9/+2
low_sleep_handler() has an hardcoded restore of segment registers that doesn't take KUAP and KUEP into account. Use head_32's load_segment_registers() routine instead. Fixes: a68c31fc01ef ("powerpc/32s: Implement Kernel Userspace Access Protection") Fixes: 31ed2b13c48d ("powerpc/32s: Implement Kernel Userspace Execution Prevention.") Cc: stable@vger.kernel.org Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/21b05f7298c1b18f73e6e5b4cd5005aafa24b6da.1599820109.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Remove useless #ifdef CONFIG_PPC_FPUChristophe Leroy1-5/+4
Add a stub for __giveup_fpu() when CONFIG_PPC_FPU is not selected, as done for CONFIG_SPE and CONFIG_ALTIVEC. This allows to remove some #ifdef CONFIG_PPC_FPU. Also change one to IS_ENABLED(). Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/69c8b7954ceeccc6b849e52e1fa41b3a0f10f6c1.1597643221.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Remove useless #ifdef CONFIG_SPEChristophe Leroy2-14/+8
cpu_has_feature(CPU_FTR_SPE) returns false when CONFIG_SPE is not set. There is no need to enclose the test in an #ifdef CONFIG_SPE. Remove it. CPU_FTR_SPE only exists on 32 bits. Define it as 0 on 64 bits. We have a couple of places like: #ifdef CONFIG_SPE if (cpu_has_feature(CPU_FTR_SPE)) { do_something_that_requires_CONFIG_SPE } else { return -EINVAL; } #else return -EINVAL; #endif Replace them by a cleaner version: if (cpu_has_feature(CPU_FTR_SPE)) { #ifdef CONFIG_SPE do_something_that_requires_CONFIG_SPE #endif } else { return -EINVAL; } When CONFIG_SPE is not set, this resolves to an unconditional return of -EINVAL Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/698df8387555765b70ea42e4a7fa48141c309c1f.1597643221.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Remove useless #ifdef CONFIG_ALTIVECChristophe Leroy1-4/+0
cpu_has_feature(CPU_FTR_ALTIVEC) returns false when CONFIG_ALTIVEC is not set. There is no need to enclose the test in an #ifdef CONFIG_ALTIVEC. Remove it. Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/03ba6b52344ca7c336df2bc6e3d31d736c804ae2.1597643221.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Remove useless #ifdef CONFIG_VSXChristophe Leroy1-12/+1
cpu_has_feature(CPU_FTR_VSX) returns false when CONFIG_VSX is not set. There is no need to enclose the test in an #ifdef CONFIG_VSX. Remove it. Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/0eb61cf0dc66d781d47deb2228498cd61d03a754.1597643221.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Tag an #endif to help locate the matching #ifdef.Christophe Leroy1-1/+1
That #endif is more than 100 lines after the matching #ifdef, and there are several #ifdef/#else/#endif inbetween. Tag it as /* CONFIG_PPC_BOOK3S_64 */ to help locate the matching #ifdef. Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/3612a8f8aaca16de3fc414a7e66293319d6e213c.1597643147.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Replace #ifdef CONFIG_KALLSYMS by IS_ENABLED()Christophe Leroy1-4/+4
The #ifdef CONFIG_KALLSYMS encloses some printk which can compile in all cases. Replace by IS_ENABLED(). Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/2d89732a9062b2cf2651728804e4b8f6c9b9358e.1597643164.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Replace an #if defined(CONFIG_4xx) || defined(CONFIG_BOOKE) by IS_ENABLED()Christophe Leroy1-6/+7
The #if defined(CONFIG_4xx) || defined(CONFIG_BOOKE) encloses some printk which can be compiled in all cases. Replace by IS_ENABLED(). Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/a1b6ef3d657c8f249193442f56868fc358ea5b6c.1597643160.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Replace an #ifdef CONFIG_PPC_BOOK3S_64 by IS_ENABLED()Christophe Leroy1-3/+1
This #ifdef CONFIG_PPC_BOOK3S_64 calls preload_new_slb_context() when radix is not enabled. radix_enabled() is always defined, and the prototype for preload_new_slb_context() is always present, so the #ifdef is unneeded. Replace it by IS_ENABLED(). Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/d31506ca9bac9def68cf7424eded63fdc4fb6660.1597643167.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/process: Replace an #ifdef CONFIG_PPC_47x by IS_ENABLED()Christophe Leroy1-3/+2
isync() is always defined, no need for an #ifdef. Replace it by IS_ENABLED(CONFIG_PPC_47x). Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/ac8da0e3baa91dda805e1e492fd65aecd90c1fb5.1597643156.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/32: Fix vmap stack - Properly set r1 before activating MMUChristophe Leroy1-14/+29
We need r1 to be properly set before activating MMU, otherwise any new exception taken while saving registers into the stack in exception prologs will use the user stack, which is wrong and will even lockup or crash when KUAP is selected. Do that by switching the meaning of r11 and r1 until we have saved r1 to the stack: copy r1 into r11 and setup the new stack pointer in r1. To avoid complicating and impacting all generic and specific prolog code (and more), copy back r1 into r11 once r11 is save onto the stack. We could get rid of copying r1 back and forth at the cost of rewriting everything to use r1 instead of r11 all the way when CONFIG_VMAP_STACK is set, but the effort is probably not worth it. Fixes: 028474876f47 ("powerpc/32: prepare for CONFIG_VMAP_STACK") Cc: stable@vger.kernel.org Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/8f85e8752ac5af602db7237ef53d634f4f3d3892.1599486108.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/32: Fix vmap stack - Do not activate MMU before reading task structChristophe Leroy2-31/+6
We need r1 to be properly set before activating MMU, so reading task_struct->stack must be done with MMU off. This means we need an additional register to play with MSR bits while r11 now points to the stack. For that, move r10 back to CR (As is already done for hash MMU) and use r10. We still don't have r1 correct yet when we activate MMU. It is done in following patch. Fixes: 028474876f47 ("powerpc/32: prepare for CONFIG_VMAP_STACK") Cc: stable@vger.kernel.org Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/a027d447022a006c9c4958ac734128e577a3c5c1.1599486108.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/uaccess: Remove __put_user_asm() and __put_user_asm2()Christophe Leroy1-36/+5
__put_user_asm() and __put_user_asm2() are not used anymore. Remove them. Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/d66c4a372738d2fbd81f433ca86e4295871ace6a.1599216721.git.christophe.leroy@csgroup.eu
2020-09-15powerpc/uaccess: Switch __patch_instruction() to __put_user_asm_goto()Christophe Leroy1-10/+7
__patch_instruction() is the only user of __put_user_asm() outside of asm/uaccess.h Switch to the new __put_user_asm_goto() to enable retirement of __put_user_asm() in a later patch. Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/b9745b122f4a9ae72cef445c61320022ab8b77b7.1599216721.git.christophe.leroy@csgroup.eu
Copyright © 1996 – 2026 Jason A. Donenfeld. All Rights Reverse Engineered.