Received: by 2002:ac0:a5a7:0:0:0:0:0 with SMTP id m36-v6csp3824516imm; Sat, 21 Jul 2018 04:09:31 -0700 (PDT) X-Google-Smtp-Source: AAOMgpfS9HtxmXYRJqS0bJ9DaCvPjztiCBAzli+kSBzxGw7/AuJquZGP3bM36SWd+UxroKHHkqVq X-Received: by 2002:a63:4283:: with SMTP id p125-v6mr5425902pga.142.1532171371657; Sat, 21 Jul 2018 04:09:31 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1532171371; cv=none; d=google.com; s=arc-20160816; b=N7FTKySv+IwbPxD1zOJmkU5Y7vk6asIdLjpZm2JE9I9B/OgAoOXGKGjwv5V/emvzmq GCMzUH6aT0mG1Hiynmr+0/JFQo/v7aptKxAcGiEP1Lj0b8flwPP5lAFDUNSnrHaJrBIq Y1kkzresiD6Y3+/XBodHKsslNq2x6Acg1eSoQSWpK+tiNu+Z6GFj6ymhIRmHV2DhpDXp v1vwzNy099nB15AIJwZF5xtkihLg7UwekgDVeS5riMMhcWOAnhMbKDhdaH18leupQVd5 akktgtED9PHkL9jn3slFh627VqkSQ2wKtlaJSHDY8dA1pgFVmaGZyRuBLkApYRsD5nWO LeJg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:dkim-signature:references:in-reply-to :message-id:date:subject:cc:to:from:arc-authentication-results; bh=s1Z+jigC7AvzchpeDXRe23MuxCPWic1motvZezwZz7o=; b=AwUpxXEyefTE2ddKhwZFAbY4lk/YgFHCSpOK3E7AYFEW5NmPJT9ztD69gLkbIsPTbC NYnSuNeya7YPIDu/7wr7JB7Jq5R0NO4mO7ilCCXPx1ytxa9b+DGV7611Xxnj4ZwewpRe QhVm0pcU+AVMx4dEogh+0h7q/gWKfp9hlRU0iSg6csw2gCzb41UUfO0rKFxAEPbGYKp5 xro7dIgRelQHhHtdBiPVeQJnkXeMRdWdBnn/JIE5nSp9VYAHjWQ4drjN+m5LPinHXe33 P4GTAfHZw8kzDQt6FaocDC4xZ9SUjmsdTLaouJqWNGCCJVCgzzNwQahQf/+sXMxH7I20 lJsw== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@crapouillou.net header.s=mail header.b=gKbf9Wa6; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=crapouillou.net Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id u7-v6si4203033pfi.96.2018.07.21.04.09.04; Sat, 21 Jul 2018 04:09:31 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@crapouillou.net header.s=mail header.b=gKbf9Wa6; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=crapouillou.net Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728099AbeGUMAF (ORCPT + 99 others); Sat, 21 Jul 2018 08:00:05 -0400 Received: from outils.crapouillou.net ([89.234.176.41]:42788 "EHLO crapouillou.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727500AbeGUMAE (ORCPT ); Sat, 21 Jul 2018 08:00:04 -0400 From: Paul Cercueil To: Vinod Koul , Rob Herring , Mark Rutland , Ralf Baechle , Paul Burton , James Hogan , Zubair Lutfullah Kakakhel Cc: Mathieu Malaterre , Daniel Silsby , Paul Cercueil , dmaengine@vger.kernel.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mips@linux-mips.org Subject: [PATCH v3 04/18] dmaengine: dma-jz4780: Separate chan/ctrl registers Date: Sat, 21 Jul 2018 13:06:29 +0200 Message-Id: <20180721110643.19624-5-paul@crapouillou.net> In-Reply-To: <20180721110643.19624-1-paul@crapouillou.net> References: <20180721110643.19624-1-paul@crapouillou.net> DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=crapouillou.net; s=mail; t=1532171261; bh=s1Z+jigC7AvzchpeDXRe23MuxCPWic1motvZezwZz7o=; h=From:To:Cc:Subject:Date:Message-Id:In-Reply-To:References; b=gKbf9Wa6osB3G5Y2Igpg1yzsqfs6jpddnQGl6qKxIUp2Y25HOGvRs4sAARnIjPCFkNKPZuzeLHYUNYcXcw6+TnimM1ypcTeTDPazJLLGG+UINFLT/dzfEzOQ61LJ/O6HchLtht4bM4YrWijlGdz+Ki9JzbOZh3JygIqg++Ji/ao= Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org The register area of the JZ4780 DMA core can be split into different sections for different purposes: * one set of registers is used to perform actions at the DMA core level, that will generally affect all channels; * one set of registers per DMA channel, to perform actions at the DMA channel level, that will only affect the channel in question. The problem rises when trying to support new versions of the JZ47xx Ingenic SoC. For instance, the JZ4770 has two DMA cores, each one with six DMA channels, and the register sets are interleaved: By using one memory resource for the channel-specific registers and one memory resource for the core-specific registers, we can support the JZ4770, by initializing the driver once per DMA core with different addresses. Signed-off-by: Paul Cercueil Tested-by: Mathieu Malaterre --- drivers/dma/dma-jz4780.c | 115 ++++++++++++++++++++++++++++++----------------- 1 file changed, 74 insertions(+), 41 deletions(-) v2: - Add a fallback mechanism for JZ4780 if the second memory resource was not supplied in the devicetree. - The documentation update was moved to patch 01/17 v3: No change diff --git a/drivers/dma/dma-jz4780.c b/drivers/dma/dma-jz4780.c index a26107c85ee7..2f17a0fb1e5c 100644 --- a/drivers/dma/dma-jz4780.c +++ b/drivers/dma/dma-jz4780.c @@ -25,26 +25,26 @@ #include "virt-dma.h" /* Global registers. */ -#define JZ_DMA_REG_DMAC 0x1000 -#define JZ_DMA_REG_DIRQP 0x1004 -#define JZ_DMA_REG_DDR 0x1008 -#define JZ_DMA_REG_DDRS 0x100c -#define JZ_DMA_REG_DMACP 0x101c -#define JZ_DMA_REG_DSIRQP 0x1020 -#define JZ_DMA_REG_DSIRQM 0x1024 -#define JZ_DMA_REG_DCIRQP 0x1028 -#define JZ_DMA_REG_DCIRQM 0x102c +#define JZ_DMA_REG_DMAC 0x00 +#define JZ_DMA_REG_DIRQP 0x04 +#define JZ_DMA_REG_DDR 0x08 +#define JZ_DMA_REG_DDRS 0x0c +#define JZ_DMA_REG_DMACP 0x1c +#define JZ_DMA_REG_DSIRQP 0x20 +#define JZ_DMA_REG_DSIRQM 0x24 +#define JZ_DMA_REG_DCIRQP 0x28 +#define JZ_DMA_REG_DCIRQM 0x2c /* Per-channel registers. */ #define JZ_DMA_REG_CHAN(n) (n * 0x20) -#define JZ_DMA_REG_DSA(n) (0x00 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DTA(n) (0x04 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DTC(n) (0x08 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DRT(n) (0x0c + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DCS(n) (0x10 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DCM(n) (0x14 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DDA(n) (0x18 + JZ_DMA_REG_CHAN(n)) -#define JZ_DMA_REG_DSD(n) (0x1c + JZ_DMA_REG_CHAN(n)) +#define JZ_DMA_REG_DSA 0x00 +#define JZ_DMA_REG_DTA 0x04 +#define JZ_DMA_REG_DTC 0x08 +#define JZ_DMA_REG_DRT 0x0c +#define JZ_DMA_REG_DCS 0x10 +#define JZ_DMA_REG_DCM 0x14 +#define JZ_DMA_REG_DDA 0x18 +#define JZ_DMA_REG_DSD 0x1c #define JZ_DMA_DMAC_DMAE BIT(0) #define JZ_DMA_DMAC_AR BIT(2) @@ -85,6 +85,8 @@ BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \ BIT(DMA_SLAVE_BUSWIDTH_4_BYTES)) +#define JZ4780_DMA_CTRL_OFFSET 0x1000 + /** * struct jz4780_dma_hwdesc - descriptor structure read by the DMA controller. * @dcm: value for the DCM (channel command) register @@ -144,7 +146,8 @@ struct jz4780_dma_soc_data { struct jz4780_dma_dev { struct dma_device dma_device; - void __iomem *base; + void __iomem *chn_base; + void __iomem *ctrl_base; struct clk *clk; unsigned int irq; enum jz_version version; @@ -178,16 +181,28 @@ static inline struct jz4780_dma_dev *jz4780_dma_chan_parent( dma_device); } -static inline uint32_t jz4780_dma_readl(struct jz4780_dma_dev *jzdma, +static inline uint32_t jz4780_dma_chn_readl(struct jz4780_dma_dev *jzdma, + unsigned int chn, unsigned int reg) +{ + return readl(jzdma->chn_base + reg + JZ_DMA_REG_CHAN(chn)); +} + +static inline void jz4780_dma_chn_writel(struct jz4780_dma_dev *jzdma, + unsigned int chn, unsigned int reg, uint32_t val) +{ + writel(val, jzdma->chn_base + reg + JZ_DMA_REG_CHAN(chn)); +} + +static inline uint32_t jz4780_dma_ctrl_readl(struct jz4780_dma_dev *jzdma, unsigned int reg) { - return readl(jzdma->base + reg); + return readl(jzdma->ctrl_base + reg); } -static inline void jz4780_dma_writel(struct jz4780_dma_dev *jzdma, +static inline void jz4780_dma_ctrl_writel(struct jz4780_dma_dev *jzdma, unsigned int reg, uint32_t val) { - writel(val, jzdma->base + reg); + writel(val, jzdma->ctrl_base + reg); } static struct jz4780_dma_desc *jz4780_dma_desc_alloc( @@ -482,17 +497,18 @@ static void jz4780_dma_begin(struct jz4780_dma_chan *jzchan) } /* Use 8-word descriptors. */ - jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), JZ_DMA_DCS_DES8); + jz4780_dma_chn_writel(jzdma, jzchan->id, + JZ_DMA_REG_DCS, JZ_DMA_DCS_DES8); /* Write descriptor address and initiate descriptor fetch. */ desc_phys = jzchan->desc->desc_phys + (jzchan->curr_hwdesc * sizeof(*jzchan->desc->desc)); - jz4780_dma_writel(jzdma, JZ_DMA_REG_DDA(jzchan->id), desc_phys); - jz4780_dma_writel(jzdma, JZ_DMA_REG_DDRS, BIT(jzchan->id)); + jz4780_dma_chn_writel(jzdma, jzchan->id, JZ_DMA_REG_DDA, desc_phys); + jz4780_dma_ctrl_writel(jzdma, JZ_DMA_REG_DDRS, BIT(jzchan->id)); /* Enable the channel. */ - jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), - JZ_DMA_DCS_DES8 | JZ_DMA_DCS_CTE); + jz4780_dma_chn_writel(jzdma, jzchan->id, JZ_DMA_REG_DCS, + JZ_DMA_DCS_DES8 | JZ_DMA_DCS_CTE); } static void jz4780_dma_issue_pending(struct dma_chan *chan) @@ -518,7 +534,7 @@ static int jz4780_dma_terminate_all(struct dma_chan *chan) spin_lock_irqsave(&jzchan->vchan.lock, flags); /* Clear the DMA status and stop the transfer. */ - jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), 0); + jz4780_dma_chn_writel(jzdma, jzchan->id, JZ_DMA_REG_DCS, 0); if (jzchan->desc) { vchan_terminate_vdesc(&jzchan->desc->vdesc); jzchan->desc = NULL; @@ -567,8 +583,8 @@ static size_t jz4780_dma_desc_residue(struct jz4780_dma_chan *jzchan, residue += desc->desc[i].dtc << jzchan->transfer_shift; if (next_sg != 0) { - count = jz4780_dma_readl(jzdma, - JZ_DMA_REG_DTC(jzchan->id)); + count = jz4780_dma_chn_readl(jzdma, jzchan->id, + JZ_DMA_REG_DTC); residue += count << jzchan->transfer_shift; } @@ -615,8 +631,8 @@ static void jz4780_dma_chan_irq(struct jz4780_dma_dev *jzdma, spin_lock(&jzchan->vchan.lock); - dcs = jz4780_dma_readl(jzdma, JZ_DMA_REG_DCS(jzchan->id)); - jz4780_dma_writel(jzdma, JZ_DMA_REG_DCS(jzchan->id), 0); + dcs = jz4780_dma_chn_readl(jzdma, jzchan->id, JZ_DMA_REG_DCS); + jz4780_dma_chn_writel(jzdma, jzchan->id, JZ_DMA_REG_DCS, 0); if (dcs & JZ_DMA_DCS_AR) { dev_warn(&jzchan->vchan.chan.dev->device, @@ -655,7 +671,7 @@ static irqreturn_t jz4780_dma_irq_handler(int irq, void *data) uint32_t pending, dmac; int i; - pending = jz4780_dma_readl(jzdma, JZ_DMA_REG_DIRQP); + pending = jz4780_dma_ctrl_readl(jzdma, JZ_DMA_REG_DIRQP); for (i = 0; i < jzdma->soc_data->nb_channels; i++) { if (!(pending & (1<base = devm_ioremap_resource(dev, res); - if (IS_ERR(jzdma->base)) - return PTR_ERR(jzdma->base); + jzdma->chn_base = devm_ioremap_resource(dev, res); + if (IS_ERR(jzdma->chn_base)) + return PTR_ERR(jzdma->chn_base); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 1); + if (res) { + jzdma->ctrl_base = devm_ioremap_resource(dev, res); + if (IS_ERR(jzdma->ctrl_base)) + return PTR_ERR(jzdma->ctrl_base); + } else if (version == ID_JZ4780) { + /* + * On JZ4780, if the second memory resource was not supplied, + * assume we're using an old devicetree, and calculate the + * offset to the control registers. + */ + jzdma->ctrl_base = jzdma->chn_base + JZ4780_DMA_CTRL_OFFSET; + } else { + dev_err(dev, "failed to get I/O memory\n"); + return -EINVAL; + } ret = platform_get_irq(pdev, 0); if (ret < 0) { @@ -869,9 +902,9 @@ static int jz4780_dma_probe(struct platform_device *pdev) * Also set the FMSC bit - it increases MSC performance, so it makes * little sense not to enable it. */ - jz4780_dma_writel(jzdma, JZ_DMA_REG_DMAC, + jz4780_dma_ctrl_writel(jzdma, JZ_DMA_REG_DMAC, JZ_DMA_DMAC_DMAE | JZ_DMA_DMAC_FMSC); - jz4780_dma_writel(jzdma, JZ_DMA_REG_DMACP, 0); + jz4780_dma_ctrl_writel(jzdma, JZ_DMA_REG_DMACP, 0); INIT_LIST_HEAD(&dd->channels); -- 2.11.0