linux/drivers/of, branch v5.14

Merge tag 'devicetree-for-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux

2021-07-03T17:54:08Z

Pull devicetree updates from Rob Herring: - Refine reserved memory nomap handling - Merge some PCI and non-PCI address handling implementations - Simplify of_address.h header ifdefs - Improve printk handling of some 64-bit types - Convert adi,adv7511, Arm ccree, Arm SCMI, Arm SCU, Arm TWD timer, Arm VIC, arm,sbsa-gwdt, Arm/Amlogic SCPI, Aspeed I2C, Broadcom iProc PWM, linaro,optee-tz, MDIO GPIO, Mediatek RNG, MTD physmap, NXP pcf8563/pcf85263/pcf85363, Renesas TPU, renesas,emev2-smu, renesas,r9a06g032-sysctrl, sysc-rmobile, Tegra20 EMC, TI AM56 PCI, TI OMAP mailbox, TI SCI bindings, virtio-mmio, Zynq FPGA, and ZynqMP RTC to DT schema - Convert mux and mux controller bindings to schema. This includes MDIO IIO, and I2C muxes. - Add Arm PL031 RTC binding schema - Add vendor prefixes for StarFive Technology Co. Ltd. and Insignal Ltd - Fix some stale doc references - Remove stale property-units.txt. Superseded by schema in dt-schema repo. - Fixes for 'unevaluatedProperties' handling (enabled with experimental json-schema support) - Drop redundant usage of minItems and maxItems across the tree - Update some examples to use bindings with a schema * tag 'devicetree-for-5.14' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux: (83 commits) dt-bindings: Fix 'unevaluatedProperties' errors in DT graph users dt-bindings: display: renesas,du: Fix 'ports' reference dt-bindings: media: adv7180: Add missing video-interfaces.yaml reference dt-bindings: crypto: ccree: Convert to json-schema dt-bindings: fpga: zynq: convert bindings to YAML dt-bindings: rtc: zynqmp: convert bindings to YAML dt-bindings: interrupt-controller: Convert ARM VIC to json-schema of: of_reserved_mem: mark nomap memory instead of removing of: of_reserved_mem: only call memblock_free for normal reserved memory dt-bindings: Drop redundant minItems/maxItems dt-bindings: spmi: Correct 'reg' schema of: reserved-memory: Add stub for RESERVEDMEM_OF_DECLARE() dt-bindings: clk: vc5: Fix example dt-bindings: timer: renesas,tmu: add r8a779a0 TMU support dt-bindings: drm: bridge: adi,adv7511.txt: convert to yaml dt-bindings: PCI: ti,am65: Convert PCIe host/endpoint mode dt-bindings to YAML of: Remove superfluous casts when printing u64 values of: Fix truncation of memory sizes on 32-bit platforms dt-bindings: rtc: nxp,pcf8563: Absorb pcf85263/pcf85363 bindings dt-bindings: pwm: Use examples with documented/matching schema ...

of: of_reserved_mem: mark nomap memory instead of removing

2021-06-24T19:47:26Z

Since commit 86588296acbf ("fdt: Properly handle "no-map" field in the memory region"), nomap memory is changed to call memblock_mark_nomap() instead of memblock_remove(). But it only changed the reserved memory with fixed addr and size case in early_init_dt_reserve_memory_arch(), not including the dynamical allocation by size case in early_init_dt_alloc_reserved_memory_arch(). Cc: Rob Herring Cc: devicetree@vger.kernel.org Signed-off-by: Dong Aisheng Reviewed-by: Quentin Perret Link: https://lore.kernel.org/r/20210611131153.3731147-2-aisheng.dong@nxp.com Signed-off-by: Rob Herring

of: of_reserved_mem: only call memblock_free for normal reserved memory

2021-06-24T19:47:26Z

For nomap case, the memory block will be removed by memblock_remove() in early_init_dt_alloc_reserved_memory_arch(). So it's meaningless to call memblock_free() on error path. Cc: Rob Herring Cc: devicetree@vger.kernel.org Signed-off-by: Dong Aisheng Link: https://lore.kernel.org/r/20210611131153.3731147-1-aisheng.dong@nxp.com Signed-off-by: Rob Herring

of: Remove superfluous casts when printing u64 values

2021-06-16T19:55:05Z

"u64" is "unsigned long long" on all architectures now. Hence there is no longer a need to use casts when formatting using the "ll" length modifier. Signed-off-by: Geert Uytterhoeven Link: https://lore.kernel.org/r/ef3f4f78385b43230695ba0855d078290c958192.1623835273.git.geert+renesas@glider.be Signed-off-by: Rob Herring

of: Fix truncation of memory sizes on 32-bit platforms

2021-06-16T19:55:05Z

Variable "size" has type "phys_addr_t", which can be either 32-bit or 64-bit on 32-bit systems, while "unsigned long" is always 32-bit on 32-bit systems. Hence the cast in (unsigned long)size / SZ_1M may truncate a 64-bit size to 32-bit, as casts have a higher operator precedence than divisions. Fix this by inverting the order of the cast and division, which should be safe for memory blocks smaller than 4 PiB. Note that the division is actually a shift, as SZ_1M is a power-of-two constant, hence there is no need to use div_u64(). While at it, use "%lu" to format "unsigned long". Fixes: e8d9d1f5485b52ec ("drivers: of: add initialization code for static reserved memory") Fixes: 3f0c8206644836e4 ("drivers: of: add initialization code for dynamic reserved memory") Signed-off-by: Geert Uytterhoeven Acked-by: Marek Szyprowski Link: https://lore.kernel.org/r/4a1117e72d13d26126f57be034c20dac02f1e915.1623835273.git.geert+renesas@glider.be Signed-off-by: Rob Herring

iommu: Drop unnecessary of_iommu.h includes

2021-06-08T12:15:46Z

The only place of_iommu.h is needed is in drivers/of/device.c. Remove it from everywhere else. Cc: Will Deacon Cc: Robin Murphy Cc: Joerg Roedel Cc: Rob Clark Cc: Marek Szyprowski Cc: Krzysztof Kozlowski Cc: Bjorn Andersson Cc: Yong Wu Cc: Matthias Brugger Cc: Heiko Stuebner Cc: Jean-Philippe Brucker Cc: Frank Rowand Cc: linux-arm-kernel@lists.infradead.org Cc: iommu@lists.linux-foundation.org Signed-off-by: Rob Herring Reviewed-by: Robin Murphy Acked-by: Heiko Stuebner Link: https://lore.kernel.org/r/20210527193710.1281746-2-robh@kernel.org Signed-off-by: Joerg Roedel

of: unittest: Use DEFINE_RES_MEM() to simplify code

2021-06-04T22:29:18Z

No functional change. Signed-off-by: Zhen Lei Link: https://lore.kernel.org/r/20210601062223.9724-1-thunder.leizhen@huawei.com Signed-off-by: Rob Herring

of: Merge of_address_to_resource() and of_pci_address_to_resource() implementations

2021-06-03T01:59:29Z

of_address_to_resource() and of_pci_address_to_resource() are almost the same except the former takes an index and the latter takes a BAR number. Now that __of_get_address() can take either one, refactor the functions to use a common implementation. Cc: Frank Rowand Signed-off-by: Rob Herring Link: https://lore.kernel.org/r/20210527194547.1287934-5-robh@kernel.org

of: address: Use IS_ENABLED() for !CONFIG_PCI

2021-06-03T01:59:29Z

Convert address.c to use IS_ENABLED() instead of ifdefs for the public PCI functions. This simplifies the ifdefs in of_address.h. Cc: Frank Rowand Signed-off-by: Rob Herring Link: https://lore.kernel.org/r/20210527194547.1287934-4-robh@kernel.org

of: Merge of_get_address() and of_get_pci_address() implementations

2021-06-03T01:59:28Z

of_get_address() and of_get_pci_address() are the same implementation except of_get_pci_address() takes the PCI BAR number rather than an index. Modify the of_get_address() implementation to work on either index or BAR and provide wrapper functions for the existing functions. Cc: Frank Rowand Signed-off-by: Rob Herring Link: https://lore.kernel.org/r/20210527194547.1287934-3-robh@kernel.org