From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-13.8 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 46C5BC433EF for ; Fri, 24 Sep 2021 08:24:37 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 2CA6161090 for ; Fri, 24 Sep 2021 08:24:37 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S244698AbhIXI0I (ORCPT ); Fri, 24 Sep 2021 04:26:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38398 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S244650AbhIXI0G (ORCPT ); Fri, 24 Sep 2021 04:26:06 -0400 Received: from metis.ext.pengutronix.de (metis.ext.pengutronix.de [IPv6:2001:67c:670:201:290:27ff:fe1d:cc33]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C3FEC061574 for ; Fri, 24 Sep 2021 01:24:33 -0700 (PDT) Received: from gallifrey.ext.pengutronix.de ([2001:67c:670:201:5054:ff:fe8d:eefb] helo=[IPv6:::1]) by metis.ext.pengutronix.de with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1mTgV9-0001IK-F1; Fri, 24 Sep 2021 10:24:11 +0200 Message-ID: <64913eca4ded0803a7e839902ff6d70c924c71c8.camel@pengutronix.de> Subject: Re: [RFC] PCI: imx6: add support for internal oscillator on i.MX7D From: Lucas Stach To: Matthias Schiffer , Richard Zhu , Shawn Guo , Sascha Hauer Cc: Lorenzo Pieralisi , Rob Herring , Krzysztof =?UTF-8?Q?Wilczy=C5=84ski?= , Bjorn Helgaas , Pengutronix Kernel Team , Fabio Estevam , NXP Linux Team , linux-pci@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Markus Niebel Date: Fri, 24 Sep 2021 10:24:09 +0200 In-Reply-To: <81c77a29362433fc5629ada442f0489046ce1051.1632319151.git.matthias.schiffer@ew.tq-group.com> References: <81c77a29362433fc5629ada442f0489046ce1051.1632319151.git.matthias.schiffer@ew.tq-group.com> Content-Type: text/plain; charset="UTF-8" User-Agent: Evolution 3.40.4 (3.40.4-1.fc34) MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-SA-Exim-Connect-IP: 2001:67c:670:201:5054:ff:fe8d:eefb X-SA-Exim-Mail-From: l.stach@pengutronix.de X-SA-Exim-Scanned: No (on metis.ext.pengutronix.de); SAEximRunCond expanded to false X-PTX-Original-Recipient: linux-kernel@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Matthias, Am Freitag, dem 24.09.2021 um 10:05 +0200 schrieb Matthias Schiffer: > Adds support for a DT property fsl,internal-osc to select the internal > oscillator for the PCIe PHY. > > Signed-off-by: Matthias Schiffer > Cc: Markus Niebel > --- > > Okay, so while this patch is nice and short, I'm note sure if it's a good > solution, hence I submit it as an RFC. It is roughly based on code from > older linux-imx versions [1] - although it seems this feature was never > supported on i.MX7D even by linux-imx (possibly because of compliance > issues with the internal clock, however I haven't found a definitive > erratum backing this), but only on other SoC like i.MX6QP. > > The device tree binding docs of the driver are somewhat lacking, but > looking at [1] it seems that an external reference clock takes the place of > the "pcie_bus" clock - various pieces of the driver skip enabling/disabling > this clock when an external clock is configured. > > From this I've come to the conclusion that the clock settings in > imx7d.dtsi do not really make sense: The pcie_bus clock is configured to > PLL_ENET_MAIN_100M_CLK, but this seems wrong for both internal and > external reference clocks: > > - For the internal clock, the correct clock should be PCIE_PHY_ROOT_CLK > according to the reference manual The pcie_bus clock should be the reference clock for the peripherals and depends on the board design. I don't think it would typically be the PCIE_PHY_ROOT_CLK, but a clock derived from the same parent PLL, if a SoC clock is used for that purpose. > - The external clocks, this should refer to an actual external clock, or > possibly a fixed-clock node > That's correct and the i.MX8MQ board DTs all point to a external clock node, as we currently default to external clocking there. > I would be great if someone with more insight into this could chime in > and tell me if my reasoning here is correct or not. > > Unfortunately I only have our MBa7x at my disposal for further > experimentation. This board does not have an external reference clock for > the PCIe PHY, so I cannot test the behaviour for settings that use an > external clock. Without this patch (and adding the new flag to the MBa7x > DTS), the boot will hang while waiting for the PCIe link to come up. > > So, for the actual question (given that my thoughts above make any sense): > How do we want to implement this? > > 1. A simple boolean flag, like this patch provides > 2. Allow Device Trees not to specify a "pcie_bus" clock at all, meaning > it should use the internal clock The internal clock is a bus clock that needs to be enabled as all other clocks, so this is not an option. > 3. Special handling when the "pcie_bus" clock is configured to > PCIE_PHY_ROOT_CLK - is such a thing even possible, or is this > breaking the clock driver's abstraction too much? > 4. Something more involved, with a proper clock sel as the source for > "pcie_bus" > > Solution 4. seems difficult to implement nicely, as the PCIe driver > also fiddles with IMX7D_GPR12_PCIE_PHY_REFCLK_SEL for power management: > the clock selection is switched back to the internal clock in > imx6_pcie_clk_disable(), which also disables its source PCIE_PHY_ROOT_CLK, > effectively gating the clock. > There is currently work under way to support this case properly. The first step is to actually abstract the PCIe PHY in the right way, Richard has already sent some patches for this. Then we can add support for the different possibilities of using the refclock pad, as this isn't a simple choice between using the internal clock or a externally supplied one. The options are: 1. Use internal clock to drive the PHY, if there isn't some other path to output this clock to the board this is effectively a split clocking setup. 2. Use externally supplied clock provided to the PHY via the refclock pad. 3. Use internal clock to driver the PHY, but output this clock on the refclock pad, which is known to be possible with the i.MX8MM PCIe PHY. Regards, Lucas > Regards, > Matthias > > > [1] https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/pci/host/pci-imx6.c?h=imx_4.1.15_2.0.0_ga > > drivers/pci/controller/dwc/pci-imx6.c | 8 +++++++- > 1 file changed, 7 insertions(+), 1 deletion(-) > > diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c > index 80fc98acf097..021499b9ee7c 100644 > --- a/drivers/pci/controller/dwc/pci-imx6.c > +++ b/drivers/pci/controller/dwc/pci-imx6.c > @@ -83,6 +83,7 @@ struct imx6_pcie { > struct regulator *vpcie; > struct regulator *vph; > void __iomem *phy_base; > + bool internal_osc; > > /* power domain for pcie */ > struct device *pd_pcie; > @@ -637,7 +638,9 @@ static void imx6_pcie_init_phy(struct imx6_pcie *imx6_pcie) > break; > case IMX7D: > regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, > - IMX7D_GPR12_PCIE_PHY_REFCLK_SEL, 0); > + IMX7D_GPR12_PCIE_PHY_REFCLK_SEL, > + imx6_pcie->internal_osc ? > + IMX7D_GPR12_PCIE_PHY_REFCLK_SEL : 0); > break; > case IMX6SX: > regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, > @@ -1130,6 +1133,9 @@ static int imx6_pcie_probe(struct platform_device *pdev) > &imx6_pcie->tx_swing_low)) > imx6_pcie->tx_swing_low = 127; > > + if (of_property_read_bool(node, "fsl,internal-osc")) > + imx6_pcie->internal_osc = true; > + > /* Limit link speed */ > pci->link_gen = 1; > ret = of_property_read_u32(node, "fsl,max-link-speed", &pci->link_gen); From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-15.2 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id B8FCCC433F5 for ; Fri, 24 Sep 2021 08:26:30 +0000 (UTC) Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 8AED861090 for ; Fri, 24 Sep 2021 08:26:30 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.1 mail.kernel.org 8AED861090 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=pengutronix.de Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=lists.infradead.org DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:MIME-Version:References:In-Reply-To: Date:Cc:To:From:Subject:Message-ID:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=ga0gzga57OLHbJVNW8ZKg/zSVH8nk+eXi0X1vGqaql0=; b=nuZguSdbaW2bh/ u20kSMQsDzy47kv2ZaQ3zjiDb2pJcc5qXkkL85SJP5WpOkDimKPs1UGPouHcv1YdE2qLFqRvBkjW3 SQxSqalyc2ifQzLmGcGEGsyv3NtxG3IKnA0hHjrwgcqHvP2xAEd1pRmHsnWmNtFOlwormknX7+KPt +JEAWDPWXtjdIC72jdZ7dPuxgZ1hBY89J2Lpst3qZLgbqpbw3dLuJ8mILHqJDJ+lxSTqz9vpcCROn LHUIab3TieCFv0HQakKTOub8jQ08fo18A2M5ZQOUUBti3ZVjHCHli7FAd5Eh0AevQO90EMjpaz9mu FdtGhq3j1uGduUJE5QHQ==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1mTgVY-00DWTx-GD; Fri, 24 Sep 2021 08:24:36 +0000 Received: from metis.ext.pengutronix.de ([2001:67c:670:201:290:27ff:fe1d:cc33]) by bombadil.infradead.org with esmtps (Exim 4.94.2 #2 (Red Hat Linux)) id 1mTgVS-00DWSD-3H for linux-arm-kernel@lists.infradead.org; Fri, 24 Sep 2021 08:24:31 +0000 Received: from gallifrey.ext.pengutronix.de ([2001:67c:670:201:5054:ff:fe8d:eefb] helo=[IPv6:::1]) by metis.ext.pengutronix.de with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1mTgV9-0001IK-F1; Fri, 24 Sep 2021 10:24:11 +0200 Message-ID: <64913eca4ded0803a7e839902ff6d70c924c71c8.camel@pengutronix.de> Subject: Re: [RFC] PCI: imx6: add support for internal oscillator on i.MX7D From: Lucas Stach To: Matthias Schiffer , Richard Zhu , Shawn Guo , Sascha Hauer Cc: Lorenzo Pieralisi , Rob Herring , Krzysztof =?UTF-8?Q?Wilczy=C5=84ski?= , Bjorn Helgaas , Pengutronix Kernel Team , Fabio Estevam , NXP Linux Team , linux-pci@vger.kernel.org, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, Markus Niebel Date: Fri, 24 Sep 2021 10:24:09 +0200 In-Reply-To: <81c77a29362433fc5629ada442f0489046ce1051.1632319151.git.matthias.schiffer@ew.tq-group.com> References: <81c77a29362433fc5629ada442f0489046ce1051.1632319151.git.matthias.schiffer@ew.tq-group.com> User-Agent: Evolution 3.40.4 (3.40.4-1.fc34) MIME-Version: 1.0 X-SA-Exim-Connect-IP: 2001:67c:670:201:5054:ff:fe8d:eefb X-SA-Exim-Mail-From: l.stach@pengutronix.de X-SA-Exim-Scanned: No (on metis.ext.pengutronix.de); SAEximRunCond expanded to false X-PTX-Original-Recipient: linux-arm-kernel@lists.infradead.org X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20210924_012430_178329_91B1BD22 X-CRM114-Status: GOOD ( 46.18 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org Hi Matthias, Am Freitag, dem 24.09.2021 um 10:05 +0200 schrieb Matthias Schiffer: > Adds support for a DT property fsl,internal-osc to select the internal > oscillator for the PCIe PHY. > > Signed-off-by: Matthias Schiffer > Cc: Markus Niebel > --- > > Okay, so while this patch is nice and short, I'm note sure if it's a good > solution, hence I submit it as an RFC. It is roughly based on code from > older linux-imx versions [1] - although it seems this feature was never > supported on i.MX7D even by linux-imx (possibly because of compliance > issues with the internal clock, however I haven't found a definitive > erratum backing this), but only on other SoC like i.MX6QP. > > The device tree binding docs of the driver are somewhat lacking, but > looking at [1] it seems that an external reference clock takes the place of > the "pcie_bus" clock - various pieces of the driver skip enabling/disabling > this clock when an external clock is configured. > > From this I've come to the conclusion that the clock settings in > imx7d.dtsi do not really make sense: The pcie_bus clock is configured to > PLL_ENET_MAIN_100M_CLK, but this seems wrong for both internal and > external reference clocks: > > - For the internal clock, the correct clock should be PCIE_PHY_ROOT_CLK > according to the reference manual The pcie_bus clock should be the reference clock for the peripherals and depends on the board design. I don't think it would typically be the PCIE_PHY_ROOT_CLK, but a clock derived from the same parent PLL, if a SoC clock is used for that purpose. > - The external clocks, this should refer to an actual external clock, or > possibly a fixed-clock node > That's correct and the i.MX8MQ board DTs all point to a external clock node, as we currently default to external clocking there. > I would be great if someone with more insight into this could chime in > and tell me if my reasoning here is correct or not. > > Unfortunately I only have our MBa7x at my disposal for further > experimentation. This board does not have an external reference clock for > the PCIe PHY, so I cannot test the behaviour for settings that use an > external clock. Without this patch (and adding the new flag to the MBa7x > DTS), the boot will hang while waiting for the PCIe link to come up. > > So, for the actual question (given that my thoughts above make any sense): > How do we want to implement this? > > 1. A simple boolean flag, like this patch provides > 2. Allow Device Trees not to specify a "pcie_bus" clock at all, meaning > it should use the internal clock The internal clock is a bus clock that needs to be enabled as all other clocks, so this is not an option. > 3. Special handling when the "pcie_bus" clock is configured to > PCIE_PHY_ROOT_CLK - is such a thing even possible, or is this > breaking the clock driver's abstraction too much? > 4. Something more involved, with a proper clock sel as the source for > "pcie_bus" > > Solution 4. seems difficult to implement nicely, as the PCIe driver > also fiddles with IMX7D_GPR12_PCIE_PHY_REFCLK_SEL for power management: > the clock selection is switched back to the internal clock in > imx6_pcie_clk_disable(), which also disables its source PCIE_PHY_ROOT_CLK, > effectively gating the clock. > There is currently work under way to support this case properly. The first step is to actually abstract the PCIe PHY in the right way, Richard has already sent some patches for this. Then we can add support for the different possibilities of using the refclock pad, as this isn't a simple choice between using the internal clock or a externally supplied one. The options are: 1. Use internal clock to drive the PHY, if there isn't some other path to output this clock to the board this is effectively a split clocking setup. 2. Use externally supplied clock provided to the PHY via the refclock pad. 3. Use internal clock to driver the PHY, but output this clock on the refclock pad, which is known to be possible with the i.MX8MM PCIe PHY. Regards, Lucas > Regards, > Matthias > > > [1] https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/pci/host/pci-imx6.c?h=imx_4.1.15_2.0.0_ga > > drivers/pci/controller/dwc/pci-imx6.c | 8 +++++++- > 1 file changed, 7 insertions(+), 1 deletion(-) > > diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c > index 80fc98acf097..021499b9ee7c 100644 > --- a/drivers/pci/controller/dwc/pci-imx6.c > +++ b/drivers/pci/controller/dwc/pci-imx6.c > @@ -83,6 +83,7 @@ struct imx6_pcie { > struct regulator *vpcie; > struct regulator *vph; > void __iomem *phy_base; > + bool internal_osc; > > /* power domain for pcie */ > struct device *pd_pcie; > @@ -637,7 +638,9 @@ static void imx6_pcie_init_phy(struct imx6_pcie *imx6_pcie) > break; > case IMX7D: > regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, > - IMX7D_GPR12_PCIE_PHY_REFCLK_SEL, 0); > + IMX7D_GPR12_PCIE_PHY_REFCLK_SEL, > + imx6_pcie->internal_osc ? > + IMX7D_GPR12_PCIE_PHY_REFCLK_SEL : 0); > break; > case IMX6SX: > regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, > @@ -1130,6 +1133,9 @@ static int imx6_pcie_probe(struct platform_device *pdev) > &imx6_pcie->tx_swing_low)) > imx6_pcie->tx_swing_low = 127; > > + if (of_property_read_bool(node, "fsl,internal-osc")) > + imx6_pcie->internal_osc = true; > + > /* Limit link speed */ > pci->link_gen = 1; > ret = of_property_read_u32(node, "fsl,max-link-speed", &pci->link_gen); _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel