Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753176Ab3I0Irs (ORCPT ); Fri, 27 Sep 2013 04:47:48 -0400 Received: from inca-roads.misterjones.org ([213.251.177.50]:57079 "EHLO inca-roads.misterjones.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751905Ab3I0Iro (ORCPT ); Fri, 27 Sep 2013 04:47:44 -0400 To: Feng Kan Subject: Re: [PATCH 1/1] arm64: X-Gene AHBC platform bus driver X-PHP-Originating-Script: 0:main.inc MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Date: Fri, 27 Sep 2013 09:47:42 +0100 From: Marc Zyngier Cc: , Organization: ARM Ltd In-Reply-To: <1380244761-15143-1-git-send-email-fkan@apm.com> References: <1380244761-15143-1-git-send-email-fkan@apm.com> Message-ID: <02e353fa5db73dd8c344389e58dd2678@www.loen.fr> User-Agent: Roundcube Webmail/0.7.2 X-SA-Exim-Connect-IP: X-SA-Exim-Rcpt-To: fkan@apm.com, linux-kernel@vger.kernel.org, linux-arm-kernel@lists.infradead.org X-SA-Exim-Mail-From: marc.zyngier@arm.com X-SA-Exim-Scanned: No (on cheepnis.misterjones.org); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 8730 Lines: 295 Hi Feng, On 2013-09-27 02:19, Feng Kan wrote: > This driver setup the AHBC for SPI and SD drivers to use. That's a bit thin for a description. What is AHBC? How does it relate to SPI and SD? How is it used? > Signed-off-by: Feng Kan > > --- > arch/arm64/boot/dts/apm-storm.dtsi | 6 + > drivers/bus/Kconfig | 9 ++ > drivers/bus/Makefile | 2 + > drivers/bus/xgene_ahbc.c | 193 > ++++++++++++++++++++++++++++++++++++ > 4 files changed, 210 insertions(+), 0 deletions(-) > create mode 100644 drivers/bus/xgene_ahbc.c > > diff --git a/arch/arm64/boot/dts/apm-storm.dtsi > b/arch/arm64/boot/dts/apm-storm.dtsi > index d994806..b9477ec 100644 > --- a/arch/arm64/boot/dts/apm-storm.dtsi > +++ b/arch/arm64/boot/dts/apm-storm.dtsi > @@ -178,6 +178,12 @@ > }; > }; > > + ahbc: ahbc@1f2a0000 { > + device_type = "ahbc"; > + compatible = "apm,xgene-ahbc"; > + reg = <0x0 0x1f2a0000 0x0 0x80000>; > + }; Where is the binding documentation for this? > serial0: serial@1c020000 { > device_type = "serial"; > compatible = "ns16550"; > diff --git a/drivers/bus/Kconfig b/drivers/bus/Kconfig > index 1f70e84..6cc361c68 100644 > --- a/drivers/bus/Kconfig > +++ b/drivers/bus/Kconfig > @@ -42,4 +42,13 @@ config ARM_CCI > help > Driver supporting the CCI cache coherent interconnect for ARM > platforms. > + > +config XGENE_AHBC > + tristate "X-Gene SoC AHB Bus Driver" > + depends on ARM64 && ARCH_XGENE > + default y if ARCH_XGENE > + help > + X-Gene SoC AHB bus driver is required for SPI and > + SDIO to function properly. > + > endmenu > diff --git a/drivers/bus/Makefile b/drivers/bus/Makefile > index 8947bdd..0c7a4ca 100644 > --- a/drivers/bus/Makefile > +++ b/drivers/bus/Makefile > @@ -10,3 +10,5 @@ obj-$(CONFIG_OMAP_OCP2SCP) += omap-ocp2scp.o > obj-$(CONFIG_OMAP_INTERCONNECT) += omap_l3_smx.o omap_l3_noc.o > # CCI cache coherent interconnect for ARM platforms > obj-$(CONFIG_ARM_CCI) += arm-cci.o > + > +obj-$(CONFIG_XGENE_AHBC) += xgene_ahbc.o > diff --git a/drivers/bus/xgene_ahbc.c b/drivers/bus/xgene_ahbc.c > new file mode 100644 > index 0000000..9554bb4 > --- /dev/null > +++ b/drivers/bus/xgene_ahbc.c > @@ -0,0 +1,193 @@ > +/** > + * AppliedMicro X-Gene AHBC Driver > + * > + * Copyright (c) 2013, Applied Micro Circuits Corporation > + * Author: Loc Ho > + * Author: Feng Kan > + * > + * This program is free software; you can redistribute it and/or > + * modify it under the terms of the GNU General Public License as > + * published by the Free Software Foundation; either version 2 of > + * the License, or (at your option) any later version. > + * > + * This program is distributed in the hope that 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, write to the Free Software > + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, > + * MA 02111-1307 USA > + * > + */ > +#include > +#include > +#include > +#include > +#include > +#include > +#include > + > +#define AHBC_DRIVER_VER "0.1" > + > +#define AIM_8_SIZE_CTL_ADDR 0x00000104 > +#define AIM_8_AXI_LO_ADDR 0x00000108 > +#define AIM_8_AXI_HI_ADDR 0x0000010c > + > +#define ARSB_F8_WR(src) (((u32)(src)<<23) & 0x03800000) > +#define AWSB_F8_WR(src) (((u32)(src)<<20) & 0x00700000) > +#define AIM_AXI_ADDRESS_LO_8_WR(src) (((u32)(src)) & 0x000fffff) > + > +/* Global Base Address */ > +#define REGSPEC_AHBC_GLBL_DIAG_CSR_I_BASE_ADDR 0x01f2ad000ULL Global base address? Why isn't this part of your DT binding? Or is it always included in the reg property? In any case, nuke this #define. > +#define REGSPEC_CFG_MEM_RAM_SHUTDOWN_ADDR 0x00000070 > +#define REGSPEC_BLOCK_MEM_RDY_ADDR 0x00000074 > + > +static void xgene_ahb_write(void *base, u32 offset, u32 val) > +{ > + if (base == NULL) > + return; > + > + writel_relaxed(val, base + offset); > +} > + > +static void xgene_ahbc_init_mem(struct platform_device *pdev) > +{ > + void *ahbc_base = platform_get_drvdata(pdev); > + u32 diag_offset = REGSPEC_AHBC_GLBL_DIAG_CSR_I_BASE_ADDR & 0xFFFF; So diag_offset is always 0xd000? > + void *diagcsr_base = ahbc_base + diag_offset; > + int timeout; > + u32 val; > + > + if (ahbc_base == NULL) > + return; > + > + if (readl(diagcsr_base + REGSPEC_CFG_MEM_RAM_SHUTDOWN_ADDR) == 0) > + return; > + > + dev_dbg(&pdev->dev, "AHBC clear memory shutdown\n"); > + writel(0x00, diagcsr_base + REGSPEC_CFG_MEM_RAM_SHUTDOWN_ADDR); > + readl(diagcsr_base + REGSPEC_CFG_MEM_RAM_SHUTDOWN_ADDR); > + > + val = readl(diagcsr_base + REGSPEC_BLOCK_MEM_RDY_ADDR); > + timeout = 5000; > + while (val != 0xFFFFFFFF && timeout-- > 0) { > + val = readl(diagcsr_base + REGSPEC_BLOCK_MEM_RDY_ADDR); > + if (val != 0xFFFFFFFF) > + udelay(1); > + } > + if (timeout <= 0) > + dev_err(&pdev->dev, "AHBC failed to remove RAM out of reset\n"); In that case, why not report the error to the caller? > +} > + > +static int xgene_ahbc_hw_init(struct platform_device *pdev) > +{ > + void *ahbc_base = platform_get_drvdata(pdev); > + phys_addr_t ddr_phy; > + > + /** > + * Set AHBC AIM windows for 4GB regardless of DDR size. This > + * must map to Kernel memory space. > + */ > + ddr_phy = memblock_phys_mem_size(); > + dev_dbg(&pdev->dev, "Setup AHBC AIM windows DDR 0x%llX\n", > ddr_phy); > + xgene_ahb_write(ahbc_base, AIM_8_SIZE_CTL_ADDR, 0x00000000); > + xgene_ahb_write(ahbc_base, AIM_8_AXI_LO_ADDR, > + ARSB_F8_WR(1) | AWSB_F8_WR(1) | > + AIM_AXI_ADDRESS_LO_8_WR((u32) ddr_phy)); > + xgene_ahb_write(ahbc_base, AIM_8_AXI_HI_ADDR, > + (u32) ((ddr_phy >> 32) << 20)); > + > + return 0; > +} > + > +static int __init xgene_ahbc_probe(struct platform_device *pdev) > +{ > + struct resource res; > + int rc; > + void *reg; > + > + rc = of_address_to_resource(pdev->dev.of_node, 0, &res); > + if (rc != 0) { > + dev_err(&pdev->dev, "invalid resource address\n"); > + return -ENODEV; > + } > + reg = ioremap(res.start, res.end - res.start + 1); > + if (reg == NULL) { > + dev_err(&pdev->dev, "can not map resource\n"); > + return -ENODEV; > + } Use of_iomap instead of dealing with resources yourself. > + > + platform_set_drvdata(pdev, reg); > + > + /* Initialize the hardware */ > + xgene_ahbc_init_mem(pdev); > + xgene_ahbc_hw_init(pdev); > + > + dev_info(&pdev->dev, "AHBC driver v%s\n", AHBC_DRIVER_VER); > + return 0; > +} > + > +static int xgene_ahbc_remove(struct platform_device *pdev) > +{ > + void *ahbc_base = platform_get_drvdata(pdev); > + > + iounmap(ahbc_base); > + platform_set_drvdata(pdev, NULL); > + return 0; > +} > + > +#if defined(CONFIG_PM) > +static int xgene_ahbc_suspend(struct platform_device *dev, > pm_message_t state) > +{ > + /* Nothing to do here */ > + return 0; > +} > + > +static int xgene_ahbc_resume(struct platform_device *pdev) > +{ > + /* Initialize the hardware */ > + xgene_ahbc_init_mem(pdev); > + xgene_ahbc_hw_init(pdev); > + return 0; > +} > +#endif > + > +static const struct of_device_id xgene_ahbc_match[] = { > + {.compatible = "apm,xgene-ahbc",}, > + {}, > +}; > + > +MODULE_DEVICE_TABLE(of, xgene_ahbc_match); > + > +static struct platform_driver ahbc_driver = { > + .probe = xgene_ahbc_probe, > + .remove = xgene_ahbc_remove, > +#if defined(CONFIG_PM) > + .suspend = xgene_ahbc_suspend, > + .resume = xgene_ahbc_resume, > +#endif It would be nicer to define xgene_ahbc_{suspend,ersume} to NULL if !CONFIG_PM. > + .driver = { > + .name = "xgene-ahbc", > + .owner = THIS_MODULE, > + .of_match_table = xgene_ahbc_match, > + }, > +}; > + > +static int __init xgene_ahbc_init(void) > +{ > + return platform_driver_register(&ahbc_driver); > +} > +subsys_initcall(xgene_ahbc_init); > + > +static void __exit xgene_ahbc_exit(void) > +{ > + platform_driver_unregister(&ahbc_driver); > +} > +module_exit(xgene_ahbc_exit); > + > +MODULE_AUTHOR("Loc Ho "); > +MODULE_DESCRIPTION("X-Gene AHBC driver"); > +MODULE_LICENSE("GPL"); Cheers, M. -- Fast, cheap, reliable. Pick two. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/