From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 832351FECB1; Sat, 28 Jun 2025 14:36:07 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1751121367; cv=none; b=WUT8DhG68aDO6xWjEtjzkynT9HYS2ITKqvHjMGOaUzOGJR+tra7Eq46zqVU69UkPnXmMS3Z0ps4W+08T9/lx6P15mT/Ny3e4wJ7mpyeOe19tvqlaLel7vHVMe5StYpteko9jKFLPGLUY3w7lfVrYp5H8olmReyhwfXwDItqTKIs= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1751121367; c=relaxed/simple; bh=LMBKCBRq57KuHMKbf/M16fyDVipmK1nxt7NY+25su0o=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=fhrMQpfeHq0zjr45VCzHPElQApFcLKeAI5D6H496+D+4A4GYFwWdpCSfVGScZBvCKY4JHIZj3CarEZscOsbWevpO8qope32p52/8Bl/8vtvEWU30dIK0CHcdk3vFuccXxpi8QKnyUqe+ho+iXLESCXd4vm+6Ktae5ETtxyQq870= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b=kDkfsgaG; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b="kDkfsgaG" Received: by smtp.kernel.org (Postfix) with ESMTPSA id B8B16C4CEF0; Sat, 28 Jun 2025 14:36:06 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1751121367; bh=LMBKCBRq57KuHMKbf/M16fyDVipmK1nxt7NY+25su0o=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=kDkfsgaGrEC3bt5bPghmzke7/i2X+/wxx8ygfjXq9/2l6p6v4OgUEVjrP2j6G/Nk3 0MNlof4YugA9iNXeLS72JH3OfmRsURY2xH7a1xzGEVYeaXcKDB0AvKGT9wmHtbRGJw j24c7esq5wjkKNtuUo7g9DH5QEq+SwSJX1n0dJCs= Date: Sat, 28 Jun 2025 16:36:04 +0200 From: Greg Kroah-Hartman To: Komal Bajaj Cc: linux-arm-msm@vger.kernel.org, linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org, Souradeep Chowdhury , Konrad Dybcio Subject: Re: [PATCH v2] usb: misc: qcom_eud: Access EUD_MODE_MANAGER2 through secure calls Message-ID: <2025062812-passive-untracked-1231@gregkh> References: <20250627125131.27606-1-komal.bajaj@oss.qualcomm.com> Precedence: bulk X-Mailing-List: linux-usb@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20250627125131.27606-1-komal.bajaj@oss.qualcomm.com> On Fri, Jun 27, 2025 at 06:21:31PM +0530, Komal Bajaj wrote: > EUD_MODE_MANAGER2 register is mapped to a memory region that is marked > as read-only for HLOS, enforcing access restrictions that prohibit > direct memory-mapped writes via writel(). > > Attempts to write to this region from HLOS can result in silent failures > or memory access violations, particularly when toggling EUD (Embedded > USB Debugger) state. To ensure secure register access, modify the driver > to use qcom_scm_io_writel(), which routes the write operation to Qualcomm > Secure Channel Monitor (SCM). SCM has the necessary permissions to access > protected memory regions, enabling reliable control over EUD state. > > SC7280, the only user of EUD is also affected, indicating that this could > never have worked on a properly fused device. > > Fixes: 9a1bf58ccd44 ("usb: misc: eud: Add driver support for Embedded USB Debugger(EUD)") > Signed-off-by: Melody Olvera > Signed-off-by: Komal Bajaj > Reviewed-by: Konrad Dybcio > --- > Changes in v2: > * Drop separate compatible to be added for secure eud > * Use secure call to access EUD mode manager register > * Link to v1: https://lore.kernel.org/all/20240807183205.803847-1-quic_molvera@quicinc.com/ > > drivers/usb/misc/qcom_eud.c | 20 ++++++++++++++------ > 1 file changed, 14 insertions(+), 6 deletions(-) > > diff --git a/drivers/usb/misc/qcom_eud.c b/drivers/usb/misc/qcom_eud.c > index 83079c414b4f..30c999c49eb0 100644 > --- a/drivers/usb/misc/qcom_eud.c > +++ b/drivers/usb/misc/qcom_eud.c > @@ -16,6 +16,8 @@ > #include > #include > > +#include Why the blank line before this #include line? > + > #define EUD_REG_INT1_EN_MASK 0x0024 > #define EUD_REG_INT_STATUS_1 0x0044 > #define EUD_REG_CTL_OUT_1 0x0074 > @@ -34,7 +36,7 @@ struct eud_chip { > struct device *dev; > struct usb_role_switch *role_sw; > void __iomem *base; > - void __iomem *mode_mgr; > + phys_addr_t mode_mgr; > unsigned int int_status; > int irq; > bool enabled; > @@ -43,10 +45,14 @@ struct eud_chip { > > static int enable_eud(struct eud_chip *priv) > { > + int ret; > + > writel(EUD_ENABLE, priv->base + EUD_REG_CSR_EUD_EN); > writel(EUD_INT_VBUS | EUD_INT_SAFE_MODE, > priv->base + EUD_REG_INT1_EN_MASK); > - writel(1, priv->mode_mgr + EUD_REG_EUD_EN2); > + ret = qcom_scm_io_writel(priv->mode_mgr + EUD_REG_EUD_EN2, 1); > + if (ret) > + return ret; So the previous writes are ok, but this one could fail? And if it does fail, what did the previous writes cause to happen to the chip / system? > return usb_role_switch_set_role(priv->role_sw, USB_ROLE_DEVICE); > } > @@ -54,7 +60,7 @@ static int enable_eud(struct eud_chip *priv) > static void disable_eud(struct eud_chip *priv) > { > writel(0, priv->base + EUD_REG_CSR_EUD_EN); > - writel(0, priv->mode_mgr + EUD_REG_EUD_EN2); > + qcom_scm_io_writel(priv->mode_mgr + EUD_REG_EUD_EN2, 0); No error checking needed? > } > > static ssize_t enable_show(struct device *dev, > @@ -178,6 +184,7 @@ static void eud_role_switch_release(void *data) > static int eud_probe(struct platform_device *pdev) > { > struct eud_chip *chip; > + struct resource *res; > int ret; > > chip = devm_kzalloc(&pdev->dev, sizeof(*chip), GFP_KERNEL); > @@ -200,9 +207,10 @@ static int eud_probe(struct platform_device *pdev) > if (IS_ERR(chip->base)) > return PTR_ERR(chip->base); > > - chip->mode_mgr = devm_platform_ioremap_resource(pdev, 1); > - if (IS_ERR(chip->mode_mgr)) > - return PTR_ERR(chip->mode_mgr); > + res = platform_get_resource(pdev, IORESOURCE_MEM, 1); > + if (!res) > + return -ENODEV; -ENOMEM perhaps? thanks, greg k-h