From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932991AbdBPQtq (ORCPT ); Thu, 16 Feb 2017 11:49:46 -0500 Received: from mga04.intel.com ([192.55.52.120]:16621 "EHLO mga04.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932254AbdBPQto (ORCPT ); Thu, 16 Feb 2017 11:49:44 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.35,169,1484035200"; d="scan'208";a="66615890" Reply-To: thor.thayer@linux.intel.com Subject: Re: [PATCH 3/5] reset: Add Altera Arria10 System Resource Reset Controller References: <1487195416-23827-1-git-send-email-thor.thayer@linux.intel.com> <1487195416-23827-4-git-send-email-thor.thayer@linux.intel.com> <1487241047.2377.23.camel@pengutronix.de> To: Philipp Zabel Cc: lee.jones@linaro.org, robh+dt@kernel.org, mark.rutland@arm.com, dinguyen@kernel.org, linux@armlinux.org.uk, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org From: Thor Thayer Message-ID: <05662fdf-ac7b-94f0-0a32-7a2c7a2e3270@linux.intel.com> Date: Thu, 16 Feb 2017 10:51:46 -0600 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.7.0 MIME-Version: 1.0 In-Reply-To: <1487241047.2377.23.camel@pengutronix.de> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Philipp, On 02/16/2017 04:30 AM, Philipp Zabel wrote: > Hi Thor, > > thank you for the patch. A few comments below: > > On Wed, 2017-02-15 at 15:50 -0600, thor.thayer@linux.intel.com wrote: >> From: Thor Thayer >> >> This patch adds the reset controller functionality to the Arria10 >> System Resource Manager. >> >> Signed-off-by: Thor Thayer >> --- >> MAINTAINERS | 1 + >> drivers/reset/Kconfig | 7 ++ >> drivers/reset/Makefile | 1 + >> drivers/reset/reset-a10sr.c | 176 ++++++++++++++++++++++++++++++++++++++++++++ >> 4 files changed, 185 insertions(+) >> create mode 100644 drivers/reset/reset-a10sr.c >> >> diff --git a/MAINTAINERS b/MAINTAINERS >> index a2c74db..3265cb2 100644 >> --- a/MAINTAINERS >> +++ b/MAINTAINERS >> @@ -655,6 +655,7 @@ F: drivers/gpio/gpio-altera-a10sr.c >> F: drivers/mfd/altera-a10sr.c >> F: include/linux/mfd/altera-a10sr.h >> F: include/dt-bindings/reset/altr,rst-mgr-a10sr.h >> +F: drivers/reset/reset-a10sr.c >> >> ALTERA TRIPLE SPEED ETHERNET DRIVER >> M: Vince Bridgers >> diff --git a/drivers/reset/Kconfig b/drivers/reset/Kconfig >> index f4cdfe9..b821d1b 100644 >> --- a/drivers/reset/Kconfig >> +++ b/drivers/reset/Kconfig >> @@ -39,6 +39,13 @@ config RESET_MESON >> help >> This enables the reset driver for Amlogic Meson SoCs. >> >> +config ALTERA_A10SR_RESET > > Please follow the RESET_ convention and sort alphabetically. > Whoops. Thanks. >> + tristate "Altera Arria10 System Resource Reset" >> + depends on MFD_ALTERA_A10SR >> + help >> + This option enables support for the external reset functions for >> + peripheral PHYs on the Altera Arria10 System Resource Chip. >> + >> config RESET_OXNAS >> bool >> >> diff --git a/drivers/reset/Makefile b/drivers/reset/Makefile >> index 2cd3f6c..d6a2a87 100644 >> --- a/drivers/reset/Makefile >> +++ b/drivers/reset/Makefile >> @@ -15,3 +15,4 @@ obj-$(CONFIG_TI_SYSCON_RESET) += reset-ti-syscon.o >> obj-$(CONFIG_RESET_UNIPHIER) += reset-uniphier.o >> obj-$(CONFIG_RESET_ZX2967) += reset-zx2967.o >> obj-$(CONFIG_RESET_ZYNQ) += reset-zynq.o >> +obj-$(CONFIG_ALTERA_A10SR_RESET) += reset-a10sr.o >> diff --git a/drivers/reset/reset-a10sr.c b/drivers/reset/reset-a10sr.c >> new file mode 100644 >> index 0000000..ed058f3 >> --- /dev/null >> +++ b/drivers/reset/reset-a10sr.c >> @@ -0,0 +1,176 @@ >> +/* >> + * Copyright Intel Corporation (C) 2017. All Rights Reserved >> + * >> + * This program is free software; you can redistribute it and/or modify it >> + * under the terms and conditions of the GNU General Public License, >> + * version 2, as published by the Free Software Foundation. >> + * >> + * This program is distributed in the hope it will be useful, but WITHOUT >> + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or >> + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for >> + * more details. >> + * >> + * You should have received a copy of the GNU General Public License along with >> + * this program. If not, see . >> + * >> + * Reset driver for Altera Arria10 MAX5 System Resource Chip >> + * >> + * Adapted from reset-socfpga.c >> + */ >> + >> +#include >> +#include > > This is not used. > OK, thanks. >> +#include >> +#include >> +#include >> +#include >> +#include >> +#include >> + >> +#include >> + >> +/* Number of A10 System Controller Resets */ >> +#define A10SR_RESETS 16 > > So 16 resets... > >> +struct a10sr_reset { >> + struct reset_controller_dev rcdev; >> + struct regmap *regmap; >> +}; >> + >> +static inline struct a10sr_reset *to_a10sr_rst(struct reset_controller_dev *rc) >> +{ >> + return container_of(rc, struct a10sr_reset, rcdev); >> +} >> + >> +static inline int a10sr_reset_shift(unsigned long id) >> +{ >> + switch (id) { >> + case A10SR_RESET_ENET_HPS: >> + return 1; >> + case A10SR_RESET_PCIE: >> + case A10SR_RESET_FILE: >> + case A10SR_RESET_BQSPI: >> + case A10SR_RESET_USB: >> + return id + 11; > > ... but only 5 are handled. What about the other 11? Could you point me > to documentation for this reset controller? > Whoops. Good catch, I will fix this. Remnants of my first implementation before realizing a switch statement was cleaner. I looked and apparently we don't publish the System Resource documentation yet. This is for a CPLD programmed as a system manager so there is some sharing of reset/enable and status bits. There were two 8-bit registers but only 5 bits are writable while others only show status. Bit 2 of the first register and bits 12-15 of the 2nd register are writable resets. >> + default: >> + return -EINVAL; >> + } >> +} >> + >> +static int a10sr_reset_update(struct reset_controller_dev *rcdev, >> + unsigned long id, bool assert) >> +{ >> + struct a10sr_reset *a10r = to_a10sr_rst(rcdev); >> + int offset = a10sr_reset_shift(id); >> + u8 mask = ALTR_A10SR_REG_BIT_MASK(offset); >> + int index = ALTR_A10SR_HPS_RST_REG + ALTR_A10SR_REG_OFFSET(offset); >> + >> + if (id >= rcdev->nr_resets) >> + return -EINVAL; > > Checking the id is not necessary. The ops are called with rstc->id, > which was checked in of_reset_simple_xlate. > OK. I didn't notice that check. Thanks. >> + return regmap_update_bits(a10r->regmap, index, mask, assert ? 0 : mask); >> +} >> + >> +static int a10sr_reset_assert(struct reset_controller_dev *rcdev, >> + unsigned long id) >> +{ >> + return a10sr_reset_update(rcdev, id, true); >> +} >> + >> +static int a10sr_reset_deassert(struct reset_controller_dev *rcdev, >> + unsigned long id) >> +{ >> + return a10sr_reset_update(rcdev, id, false); >> +} >> + >> +static int a10sr_reset_status(struct reset_controller_dev *rcdev, >> + unsigned long id) >> +{ >> + int error; >> + struct a10sr_reset *a10r = to_a10sr_rst(rcdev); >> + int offset = a10sr_reset_shift(id); >> + u8 mask = ALTR_A10SR_REG_BIT_MASK(offset); >> + int index = ALTR_A10SR_HPS_RST_REG + ALTR_A10SR_REG_OFFSET(offset); >> + unsigned int value; >> + >> + if (id >= rcdev->nr_resets) >> + return -EINVAL; > > Same here. reset_status should never be called with an invalid id. > >> + error = regmap_read(a10r->regmap, index, &value); >> + if (error < 0) >> + return -ENOMEM; > > Why ENOMEM? Just propagate the error. > Yes, that would be better. >> + >> + return !!(value & mask); >> +} >> + >> +static const struct reset_control_ops a10sr_reset_ops = { >> + .assert = a10sr_reset_assert, >> + .deassert = a10sr_reset_deassert, >> + .status = a10sr_reset_status, >> +}; >> + >> +static const struct of_device_id a10sr_reset_of_match[]; > > Avoid forward declarations. Just move a10sr_reset_of_match up here if > needed. Although I don't think this is necessary, see below. > OK. Yes, I agree that this isn't necessary with the below changes. Thanks. >> +static int a10sr_reset_probe(struct platform_device *pdev) >> +{ >> + struct altr_a10sr *a10sr = dev_get_drvdata(pdev->dev.parent); >> + struct a10sr_reset *a10r; >> + struct device *dev = &pdev->dev; >> + struct device_node *np = dev->of_node; > > Here np is assigned to pdev->dev.of_node. > >> + /* Ensure we have a valid DT entry. */ >> + np = of_find_matching_node(NULL, a10sr_reset_of_match); > > And here it is overwritten again. Shouldn't those be the same anyway? I > think this assignment and the DT entry check could be removed. > Yes, Thanks. >> + if (!np) { >> + dev_err(&pdev->dev, "A10 Reset DT Entry not found\n"); >> + return -EINVAL; >> + } >> + >> + if (!of_find_property(np, "#reset-cells", NULL)) { >> + dev_err(&pdev->dev, "%s missing #reset-cells property\n", >> + np->full_name); >> + return -EINVAL; >> + } > > This is unnecessary. If #reset-cells is missing, the reset lookups will > fail. > Nice, Thanks. >> + a10r = devm_kzalloc(&pdev->dev, sizeof(struct a10sr_reset), >> + GFP_KERNEL); >> + if (!a10r) >> + return -ENOMEM; >> + >> + a10r->rcdev.owner = THIS_MODULE; >> + a10r->rcdev.nr_resets = A10SR_RESETS; >> + a10r->rcdev.ops = &a10sr_reset_ops; >> + a10r->rcdev.of_node = np; >> + a10r->regmap = a10sr->regmap; >> + >> + platform_set_drvdata(pdev, a10r); >> + >> + return devm_reset_controller_register(dev, &a10r->rcdev); >> +} >> + >> +static int a10sr_reset_remove(struct platform_device *pdev) >> +{ >> + struct a10sr_reset *a10r = platform_get_drvdata(pdev); >> + >> + reset_controller_unregister(&a10r->rcdev); > > Since you use the devm_ variant of reset_controller_register above, this > happens automatically. You can drop a10sr_reset_remove completely. > OK, thanks. >> + >> + return 0; >> +} >> + >> +static const struct of_device_id a10sr_reset_of_match[] = { >> + { .compatible = "altr,a10sr-reset" }, >> + { }, >> +}; >> +MODULE_DEVICE_TABLE(of, a10sr_reset_of_match); >> + >> +static struct platform_driver a10sr_reset_driver = { >> + .probe = a10sr_reset_probe, >> + .remove = a10sr_reset_remove, >> + .driver = { >> + .name = "altr_a10sr_reset", >> + .owner = THIS_MODULE, > > Assigning .owner is not necessary anymore, __platform_driver_register > does this. > Got it. Thanks for reviewing! I will fix this and resubmit. >> + }, >> +}; >> +module_platform_driver(a10sr_reset_driver); >> + >> +MODULE_AUTHOR("Thor Thayer "); >> +MODULE_DESCRIPTION("Altera Arria10 System Resource Reset Controller Driver"); >> +MODULE_LICENSE("GPL v2"); > > regards > Philipp >