- 09 Sep, 2022 8 commits
-
-
Krzysztof Kozlowski authored
Child node of SMD must be either smd-edge or rpm, so the pattern can be narrowed. Signed-off-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20220909125403.803158-7-krzysztof.kozlowski@linaro.orgSigned-off-by:
Rob Herring <robh@kernel.org>
-
Krzysztof Kozlowski authored
GLINK edge contains subnodes representing devices related to the remote processor. Each of such device (e.g. APR or FastRPC) must have a "qcom,glink-channels" property. However the APR or FastRPC device can be also part of other communication - SMD - thus their schema does not require "qcom,glink-channels". To make the GLINK edge bindings strict, require "qcom,glink-channels" for its children. Signed-off-by:
-
Krzysztof Kozlowski authored
The parent of FastRPC node is GLINK edge, so the properties should be matching GLINK, not SMD. Correct the example and extend it based on SM8350. Signed-off-by:
-
Krzysztof Kozlowski authored
The Qualcomm FastRPC is expected to in GLINK or SMD edge node and its channel name is always the same. Signed-off-by:
-
Krzysztof Kozlowski authored
The children of FastRPC have either one or two IOMMUs in existing DTS (SM8150, SM8450 and others). Signed-off-by:
-
Krzysztof Kozlowski authored
The property in old TXT bindings, in Linux driver and in DTS is qcom,nsessions. Fixes: ed18a9b1 ("dt-bindings: misc: fastrpc convert bindings to yaml") Signed-off-by:
-
Luca Weiss authored
Convert the PDC interrupt controller bindings to YAML. Signed-off-by:
Luca Weiss <luca.weiss@fairphone.com> Reviewed-by:
-
Marek Bykowski authored
If the properties 'linux,initrd-start' and 'linux,initrd-end' of the chosen node populated from the bootloader, eg. U-Boot, are so that start > end, then the phys_initrd_size calculated from end - start is negative that subsequently gets converted to a high positive value for being unsigned long long. Then, the memory region with the (invalid) size is added to the bootmem and attempted being paged in paging_init() that results in the kernel fault. For example, on the FVP ARM64 system I'm running, the U-Boot populates the 'linux,initrd-start' with 8800_0000 and 'linux,initrd-end' with 0. The phys_initrd_size calculated is then ffff_ffff_7800_0000 (= 0 - 8800_0000 = -8800_0000 + ULLONG_MAX + 1). paging_init() then attempts to map the address 8800_0000 + ffff_ffff_7800_0000 and oops'es as below. It should be stressed, it is generally a fault of the bootloader's with the kernel relying on it, however we should not allow the bootloader's misconfiguration to lead to the kernel oops. Not only the kernel should be bullet proof against it but also finding the root cause of the paging fault spanning over the bootloader, DT, and kernel may happen is not so easy. Unable to handle kernel paging request at virtual address fffffffefe43c000 Mem abort info: ESR = 0x96000007 EC = 0x25: DABT (current EL), IL = 32 bits SET = 0, FnV = 0 EA = 0, S1PTW = 0 Data abort info: ISV = 0, ISS = 0x00000007 CM = 0, WnR = 0 swapper pgtable: 4k pages, 39-bit VAs, pgdp=0000000080e3d000 [fffffffefe43c000] pgd=0000000080de9003, pud=0000000080de9003 Unable to handle kernel paging request at virtual address ffffff8000de9f90 Mem abort info: ESR = 0x96000005 EC = 0x25: DABT (current EL), IL = 32 bits SET = 0, FnV = 0 EA = 0, S1PTW = 0 Data abort info: ISV = 0, ISS = 0x00000005 CM = 0, WnR = 0 swapper pgtable: 4k pages, 39-bit VAs, pgdp=0000000080e3d000 [ffffff8000de9f90] pgd=0000000000000000, pud=0000000000000000 Internal error: Oops: 96000005 [#1] PREEMPT SMP Modules linked in: CPU: 0 PID: 0 Comm: swapper Not tainted 5.4.51-yocto-standard #1 Hardware name: FVP Base (DT) pstate: 60000085 (nZCv daIf -PAN -UAO) pc : show_pte+0x12c/0x1b4 lr : show_pte+0x100/0x1b4 sp : ffffffc010ce3b30 x29: ffffffc010ce3b30 x28: ffffffc010ceed80 x27: fffffffefe43c000 x26: fffffffefe43a028 x25: 0000000080bf0000 x24: 0000000000000025 x23: ffffffc010b8d000 x22: ffffffc010e3d000 x23: ffffffc010b8d000 x22: ffffffc010e3d000 x21: 0000000080de9000 x20: ffffff7f80000f90 x19: fffffffefe43c000 x18: 0000000000000030 x17: 0000000000001400 x16: 0000000000001c00 x15: ffffffc010cef1b8 x14: ffffffffffffffff x13: ffffffc010df1f40 x12: ffffffc010df1b70 x11: ffffffc010ce3b30 x10: ffffffc010ce3b30 x9 : 00000000ffffffc8 x8 : 0000000000000000 x7 : 000000000000000f x6 : ffffffc010df16e8 x5 : 0000000000000000 x4 : 0000000000000000 x3 : 00000000ffffffff x2 : 0000000000000000 x1 : 0000008080000000 x0 : ffffffc010af1d68 Call trace: show_pte+0x12c/0x1b4 die_kernel_fault+0x54/0x78 __do_kernel_fault+0x11c/0x128 do_translation_fault+0x58/0xac do_mem_abort+0x50/0xb0 el1_da+0x1c/0x90 __create_pgd_mapping+0x348/0x598 paging_init+0x3f0/0x70d0 setup_arch+0x2c0/0x5d4 start_kernel+0x94/0x49c Code: 92748eb5 900052a0 9135a000 cb010294 (f8756a96) Signed-off-by:Marek Bykowski <marek.bykowski@gmail.com> Link: https://lore.kernel.org/r/20220909023358.76881-1-marek.bykowski@gmail.comSigned-off-by:
-
- 08 Sep, 2022 1 commit
-
-
Olliver Schinagl authored
In commit 853a78a7 (dt-bindings: leds: Add LED_COLOR_ID definitions, Sun Jun 9 20:19:04 2019 +0200) the most basic color definitions where added. However, there's a little more very common LED colors. While the documentation states 'add what is missing', engineers tend to be lazy and will just use what currently exists. So this patch will take (a) list from online retailers [0], [1], [2] and use the common LED colors from there, this being reasonable as this is what is currently available to purchase. Note, that LIME seems to be the modern take to 'Yellow-green' or 'Yellowish-green' from some older datasheets. [0]: https://www.digikey.com/en/products/filter/led-lighting-color/125 [1]: https://eu.mouser.com/c/optoelectronics/led-lighting/led-emitters/standard-leds-smd [2]: https://nl.farnell.com/en-NL/c/optoelectronics-displays/led-products/standard-single-colour-leds-under-75maSigned-off-by:
Olliver Schinagl <oliver@schinagl.nl> Acked-by:
Alexander Dahl <ada@thorsis.com> Acked-by:
Jacek Anaszewski <jacek.anaszewski@gmail.com> Link: https://lore.kernel.org/r/20220830134613.1564059-1-oliver@schinagl.nlSigned-off-by:
-
- 07 Sep, 2022 3 commits
-
-
Andrew Davis authored
For K3 devices the clock fed into the RNG module is shared with the rest of the Crypto module, it is not dedicated to the RNG module and cannot be controlled by the RNG driver. The driver does not require this clock to always be defined and the DT binding should not force it either. Remove this so we can start dropping out the clock properties as needed. Signed-off-by:
Andrew Davis <afd@ti.com> Link: https://lore.kernel.org/r/20220901171041.32056-1-afd@ti.comSigned-off-by:
-
Alexander Sverdlin authored
New pr_err(), a copy of preceeding pr_debug(), faciliates debugging. This change was inspired by a long lasting debugging of the octeon_irq_init_ciu() which fails completely silently and leaves the interrupt controller half-way configured which in turn had very non-obvious effects. Signed-off-by:
Alexander Sverdlin <alexander.sverdlin@nokia.com> Link: https://lore.kernel.org/r/20220907121629.54330-1-alexander.sverdlin@nokia.comSigned-off-by:
-
Abel Vesa authored
Convert Qualcomm FastRPC bindings to yaml format, so that we could validate dt-entries correctly and any future additions can go into yaml format. Use compute-cb@ subnodes instead of just cb@. Add qcom,glink-channels and qcom,smd-channels missing properties to make sure dtbs_check doesn't fail right off the bat. Correct the name of the parent node in the example from smd-edge to glink-edge. Since now the qcom,fastrpc bindings document is yaml, update the reference to it in qcom,glink-edge and also use $ref. Also update the MAINTAINERS file to point to the yaml version. Co-developed-by:
Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Signed-off-by:
David Heidelberg <david@ixit.cz> Signed-off-by:
Abel Vesa <abel.vesa@linaro.org> Reviewed-by:
-
- 31 Aug, 2022 2 commits
-
-
Rob Herring authored
In order to ensure only documented properties are present, node schemas must have unevaluatedProperties or additionalProperties set to false (typically). Signed-off-by:
-
Rob Herring authored
In order to ensure only documented properties are present, node schemas must have unevaluatedProperties or additionalProperties set to false (typically). Signed-off-by:
Ulf Hansson <ulf.hansson@linaro.org> Link: https://lore.kernel.org/r/20220823145649.3118479-15-robh@kernel.org
-
- 29 Aug, 2022 10 commits
-
-
Jayesh Choudhary authored
crypto driver itself is not dma-coherent. It is the dmaengine that moves data and the buffers are to be mapped to the dmaengine provider. So this property should be dropped. Fixes: 2ce9a729 ('dt-bindings: crypto: Add TI SA2UL crypto accelerator documentation') Signed-off-by:
Jayesh Choudhary <j-choudhary@ti.com> Link: https://lore.kernel.org/r/20220826091142.262325-1-j-choudhary@ti.comSigned-off-by:
-
Krzysztof Kozlowski authored
On PM660L, PMI8994 and PMI8998, the WLED has two address spaces. This also fixes dtbs_check warnings like: arch/arm64/boot/dts/qcom/sm7225-fairphone-fp4.dtb: leds@d800: reg: [[55296], [55552]] is too long Signed-off-by:
-
Krzysztof Kozlowski authored
Document compatibles for QFPROM used on IPQ8064 and SDM630. They are compatible with generic QFPROM fallback. Signed-off-by:
-
Krzysztof Kozlowski authored
minItems, if missing, are implicitly equal to maxItems, so drop redundant piece to reduce size of code. Signed-off-by:
Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Reviewed-by:
Alim Akhtar <alim.akhtar@samsung.com> Link: https://lore.kernel.org/r/20220825113334.196908-4-krzysztof.kozlowski@linaro.orgSigned-off-by:
-
Krzysztof Kozlowski authored
minItems, if missing, are implicitly equal to maxItems, so drop redundant piece to reduce size of code. Signed-off-by:
Damien Le Moal <damien.lemoal@opensource.wdc.com> Link: https://lore.kernel.org/r/20220825113334.196908-2-krzysztof.kozlowski@linaro.orgSigned-off-by:
-
Krzysztof Kozlowski authored
minItems, if missing, are implicitly equal to maxItems, so drop redundant piece to reduce size of code. Signed-off-by:
-
Rob Herring authored
In order to ensure only documented properties are present, node schemas must have unevaluatedProperties or additionalProperties set to false (typically). Signed-off-by:
-
Krzysztof Kozlowski authored
minItems, if missing, are implicitly equal to maxItems, so drop redundant piece to reduce size of code. Signed-off-by:
Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Reviewed-by:
-
Krzysztof Kozlowski authored
reg-io-width is a standard property, so no need for defining its type with $ref. Signed-off-by:
-
Rob Herring authored
In order to ensure only documented properties are present, node schemas must have unevaluatedProperties or additionalProperties set to false (typically). Signed-off-by:
-
- 25 Aug, 2022 3 commits
-
-
Rob Herring authored
DT bindings using the graph binding must have references to the graph binding schema. These are missing from the adi,adv7511 and adi,adv7533 bindings, so add them. Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Link: https://lore.kernel.org/r/20220823145649.3118479-12-robh@kernel.org
-
Andrew Lunn authored
As indicated in link: https://lore.kernel.org/all/20220822204945.GA808626-robh@kernel.org/ DT schema files should not have 'Device Tree Binding' as part of there title: line. Remove this in most .yaml files, so hopefully preventing developers copying it into new .yaml files, and being asked to remove it. Signed-off-by:
Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/20220825020427.3460650-1-andrew@lunn.chSigned-off-by:
-
Frank Rowand authored
Make OF unittest trigger the new TAINT_TEST taint when OF unittest runs. Due to OF unittest not being intended to run on production systems, and potentially causing problems (or security issues like leaking kernel addresses), the kernel's state should not be considered safe for production use after OF unittest runs. Signed-off-by:
Frank Rowand <frank.rowand@sony.com> Link: https://lore.kernel.org/r/20220823200152.3465751-1-frowand.list@gmail.comSigned-off-by:
-
- 23 Aug, 2022 2 commits
-
-
Jean-Philippe Brucker authored
It should be valid for a GICv2m node, child of a GICv2 node, to use two cells per reg address and size. The QEMU virt device-tree currently fails validation because the schema imposes a single address and size cell. Amend the rule. Signed-off-by:
Jean-Philippe Brucker <jean-philippe@linaro.org> Reviewed-by:
Marc Zyngier <maz@kernel.org> Signed-off-by:
-
Jean-Philippe Brucker authored
A GICv3 implementation without virtualization, such as the base QEMU virt machine (without -M virtualization=on), does not issue maintenance interrupts. Therefore its device-tree node does not need an 'interrupts' property. Currently, validating the QEMU virt device-tree throws a warning that 'interrupts' is missing. Make it optional. Signed-off-by:
-
- 22 Aug, 2022 6 commits
-
-
Wolfram Sang authored
Follow the advice of the below link and prefer 'strscpy' in this subsystem. Conversion is 1:1 because the return value is not used. Generated by a coccinelle script. Link: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/Signed-off-by:
Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by:
-
Rob Herring authored
Convert the arm,versatile-sysreg binding to DT schema format. The original binding was missing 'simple-mfd' and a 'panel' sub node which the only user (versatile-ab.dts) of this binding has. Signed-off-by:
-
Rob Herring authored
The 'arm,versatile-sysreg' node in the example should not have '#address-cells' and '#size-cells' properties as the child node doesn't have 'reg'. Signed-off-by:
-
Krzysztof Kozlowski authored
Simple 'opp-table:true' accepts a boolean property as opp-table, so restrict it to object to properly enforce real OPP table nodes. Signed-off-by:
-
Krzysztof Kozlowski authored
Simple 'opp-table:true' accepts a boolean property as opp-table, so restrict it to object to properly enforce real OPP table nodes. Signed-off-by:
-
Krzysztof Kozlowski authored
Simple 'opp-table:true' accepts a boolean property as opp-table, so restrict it to object to properly enferce real OPP table nodes. Signed-off-by:
Peng Fan <peng.fan@nxp.com> Signed-off-by:
-
- 16 Aug, 2022 1 commit
-
-
Rob Herring authored
Even PSCI v1.0 compliant implementations may support v0.1 clients (i.e. "arm,psci"). Relax the compatible schema such that an implementation can claim 1.0, 0.2, and 0.1 compatibility. In the process, the schema can be simplified a bit by using 'minItems' instead of separate 'oneOf' entries. Signed-off-by:
-
- 14 Aug, 2022 4 commits
-
-
Linus Torvalds authored
-
Yury Norov authored
Radix tree header includes gfp.h for __GFP_BITS_SHIFT only. Now we have gfp_types.h for this. Fixes powerpc allmodconfig build: In file included from include/linux/nodemask.h:97, from include/linux/mmzone.h:17, from include/linux/gfp.h:7, from include/linux/radix-tree.h:12, from include/linux/idr.h:15, from include/linux/kernfs.h:12, from include/linux/sysfs.h:16, from include/linux/kobject.h:20, from include/linux/pci.h:35, from arch/powerpc/kernel/prom_init.c:24: include/linux/random.h: In function 'add_latent_entropy': >> include/linux/random.h:25:46: error: 'latent_entropy' undeclared (first use in this function); did you mean 'add_latent_entropy'? 25 | add_device_randomness((const void *)&latent_entropy, sizeof(latent_entropy)); | ^~~~~~~~~~~~~~ | add_latent_entropy include/linux/random.h:25:46: note: each undeclared identifier is reported only once for each function it appears in Reported-by:kernel test robot <lkp@intel.com> CC: Andy Shevchenko <andriy.shevchenko@linux.intel.com> CC: Andrew Morton <akpm@linux-foundation.org> CC: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by:
Yury Norov <yury.norov@gmail.com> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfsLinus Torvalds authored
Pull vfs lseek fix from Al Viro: "Fix proc_reg_llseek() breakage. Always had been possible if somebody left NULL ->proc_lseek, became a practical issue now" * tag 'pull-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: take care to handle NULL ->proc_lseek()
-
Al Viro authored
Easily done now, just by clearing FMODE_LSEEK in ->f_mode during proc_reg_open() for such entries. Fixes: 868941b1 "fs: remove no_llseek" Signed-off-by:
Al Viro <viro@zeniv.linux.org.uk>
-
