kernel.git/drivers/soc/imx/gpc.c, branch linux-4.18.y

soc: imx: gpc: restrict register range for regmap access

2018-07-19T02:24:00Z

GPC registers are NOT continuous, some registers are reserved and accessing them from userspace will trigger external abort, add regmap register access table to avoid below abort: root@imx6slevk:~# cat /sys/kernel/debug/regmap/20dc000.gpc/registers [ 108.480477] Unhandled fault: imprecise external abort (0x1406) at 0xb6db5004 [ 108.487985] pgd = 42b54bfd [ 108.490741] [b6db5004] *pgd=ba1b7831 [ 108.494386] Internal error: : 1406 [#1] SMP ARM [ 108.498943] Modules linked in: [ 108.502043] CPU: 0 PID: 389 Comm: cat Not tainted 4.18.0-rc1-00074-gc9f1f60-dirty #482 [ 108.509982] Hardware name: Freescale i.MX6 SoloLite (Device Tree) [ 108.516123] PC is at regmap_mmio_read32le+0x20/0x24 [ 108.521031] LR is at regmap_mmio_read+0x40/0x60 [ 108.525586] pc : [] lr : [] psr: 20060093 [ 108.531875] sp : eccf1d98 ip : eccf1da8 fp : eccf1da4 [ 108.537122] r10: ec2d3800 r9 : eccf1f60 r8 : ecfc0000 [ 108.542370] r7 : eccf1e2c r6 : eccf1e2c r5 : 00000028 r4 : ec338e00 [ 108.548920] r3 : 00000000 r2 : eccf1e2c r1 : f0980028 r0 : 00000000 [ 108.555474] Flags: nzCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment none [ 108.562720] Control: 10c5387d Table: acf4004a DAC: 00000051 [ 108.568491] Process cat (pid: 389, stack limit = 0xd4318a65) [ 108.574174] Stack: (0xeccf1d98 to 0xeccf2000) Fixes: 721cabf6c660 ("soc: imx: move PGC handling to a new GPC driver") Signed-off-by: Anson Huang Reviewed-by: Fabio Estevam Signed-off-by: Shawn Guo

soc: imx: gpc: Do not pass static memory as platform data

2018-04-23T09:03:15Z

Platform device core assumes the ownership of dev.platform_data as well as that it is dynamically allocated and it will try to kfree it as a part of platform_device_release(). Change the code to use platform_device_add_data() instead of a pointer to a static memory to avoid causing a BUG() when calling platform_device_put(). The problem can be reproduced by artificially enabling the error path of platform_device_add() call (around line 452). Cc: Stefan Agner Cc: Lucas Stach Cc: linux-arm-kernel@lists.infradead.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Andrey Smirnov Signed-off-by: Shawn Guo

Merge tag 'imx-drivers-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into next/drivers

2018-03-15T15:23:33Z

Pull "i.MX drivers update for 4.17" from Shawn Guo: - Set GENPD_FLAG_ALWAYS_ON flag for ARM power domain to avoid incorrect power state in sysfs pm_genpd_summary output. * tag 'imx-drivers-4.17' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux: soc: imx: gpc: ARM power domain should be always-on

Merge tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc

2018-03-01T00:11:04Z

Pull ARM SoC fixes from Arnd Bergmann: "This is the first set of bugfixes for ARM SoCs, fixing a couple of stability problems, mostly on TI OMAP and Rockchips platforms: - OMAP2 hwmod clocks must be enabled in the correct order - OMAP3 Wakeup from resume through PRM IRQ was unreliable - one regression on OMAP5 caused by a kexec fix - Rockchip ethernet needs some settings for stable operation on Rock64 - Rockchip based Chrombook Plus needs another clock setting for stable display suspend/resume - Rockchip based phyCORE-RK3288 was able to run at an invalid CPU clock frequency - Rockchip MMC link was sometimes unreliable - multiple fixes to avoid crashes in the Broadcom STB DPFE driver Other minor changes include: - Devicetree fixes for incorrect hardware description (rockchip, omap, Gemini, amlogic) - some MAINTAINER file updates to correct email and git addresses - some fixes addressing 'make W=1' dtc warnings (broadcom, amlogic, cavium, qualcomm, hisilicon, zx) - fixes for LTO-compilation (orion, davinci, clps711x) - one fix for an incorrect Kconfig errata selection - a memory leak in the OMAP timer driver - a kernel data leak in OMAP1 debugfs files" * tag 'armsoc-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc: (38 commits) MAINTAINERS: update entries for ARM/STM32 ARM: dts: bcm283x: Move arm-pmu out of soc node ARM: dts: bcm283x: Fix unit address of local_intc ARM: dts: NSP: Fix amount of RAM on BCM958625HR ARM: dts: Set D-Link DNS-313 SATA to muxmode 0 ARM: omap2: set CONFIG_LIRC=y in defconfig ARM: dts: imx6dl: Include correct dtsi file for Engicam i.CoreM6 DualLite/Solo RQS memory: brcmstb: dpfe: support new way of passing data from the DCPU memory: brcmstb: dpfe: fix type declaration of variable "ret" memory: brcmstb: dpfe: properly mask vendor error bits ARM: BCM: dts: Remove leading 0x and 0s from bindings notation ARM: orion: fix orion_ge00_switch_board_info initialization ARM: davinci: mark spi_board_info arrays as const ARM: clps711x: mark clps711x_compat as const arm: zx: dts: Remove leading 0x and 0s from bindings notation arm64: dts: Remove leading 0x and 0s from bindings notation arm64: dts: cavium: fix PCI bus dtc warnings MAINTAINERS: ARM: at91: update my email address soc: imx: gpc: de-register power domains only if initialized ARM: dts: rockchip: Fix DWMMC clocks ...

treewide/trivial: Remove ';;$' typo noise

2018-02-22T09:59:33Z

On lkml suggestions were made to split up such trivial typo fixes into per subsystem patches: --- a/arch/x86/boot/compressed/eboot.c +++ b/arch/x86/boot/compressed/eboot.c @@ -439,7 +439,7 @@ setup_uga32(void **uga_handle, unsigned long size, u32 *width, u32 *height) struct efi_uga_draw_protocol *uga = NULL, *first_uga; efi_guid_t uga_proto = EFI_UGA_PROTOCOL_GUID; unsigned long nr_ugas; - u32 *handles = (u32 *)uga_handle;; + u32 *handles = (u32 *)uga_handle; efi_status_t status = EFI_INVALID_PARAMETER; int i; This patch is the result of the following script: $ sed -i 's/;;$/;/g' $(git grep -E ';;$' | grep "\.[ch]:" | grep -vwE 'for|ia64' | cut -d: -f1 | sort | uniq) ... followed by manual review to make sure it's all good. Splitting this up is just crazy talk, let's get over with this and just do it. Reported-by: Pavel Machek Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar

soc: imx: gpc: de-register power domains only if initialized

2018-02-22T03:22:12Z

If power domain information are missing in the device tree, no power domains get initialized. However, imx_gpc_remove tries to remove power domains always in the old DT binding case. Only remove power domains when imx_gpc_probe initialized them in first place. Fixes: 721cabf6c660 ("soc: imx: move PGC handling to a new GPC driver") Signed-off-by: Stefan Agner Reviewed-by: Lucas Stach Reviewed-by: Fabio Estevam Signed-off-by: Shawn Guo

soc: imx: gpc: ARM power domain should be always-on

2018-02-12T05:07:48Z

ARM power domain does NOT support runtime off, always-on flag should be set to avoid incorrect power state in pm_genpd_summary: Before: root@imx6qpdlsolox:~# cat /sys/kernel/debug/pm_genpd/pm_genpd_summary domain status slaves /device runtime status ---------------------------------------------------------------------- ARM off-0 After: root@imx6qpdlsolox:~# cat /sys/kernel/debug/pm_genpd/pm_genpd_summary domain status slaves /device runtime status ---------------------------------------------------------------------- ARM on Signed-off-by: Anson Huang Reviewed-by: Lucas Stach Acked-by: Dong Aisheng Signed-off-by: Shawn Guo

soc: imx: gpc: Add i.MX6SX PCI power domain

2017-12-26T08:26:46Z

i.MX6SX has a PCI power domain in PGC. Add support for it. Signed-off-by: Fabio Estevam Acked-by: Lucas Stach Signed-off-by: Shawn Guo

soc: imx: gpc: add workaround for i.MX6QP to the GPC PD driver

2017-04-07T12:47:34Z

On i.MX6QP, due to hardware erratum ERR009619, the PRE clocks may be stalled during the power up sequencing of the PU power domain. As this may lead to a complete loss of display output, the recommended workaround is to keep the PU domain enabled during normal system operation. Implement this by rejecting the domain power down request on the affected SoC. Signed-off-by: Lucas Stach Signed-off-by: Shawn Guo

soc: imx: gpc: add defines for domain index

2017-04-07T12:47:24Z

Makes referencing a specfic domain in the driver code less error prone. Signed-off-by: Lucas Stach Signed-off-by: Shawn Guo