From patchwork Tue Nov 24 00:01:51 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Doug Anderson X-Patchwork-Id: 1405185 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=vger.kernel.org (client-ip=23.128.96.18; helo=vger.kernel.org; envelope-from=linux-gpio-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=chromium.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256 header.s=google header.b=oMy+quRS; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by ozlabs.org (Postfix) with ESMTP id 4Cg4346pTpz9sRK for ; Tue, 24 Nov 2020 11:02:40 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728578AbgKXACU (ORCPT ); Mon, 23 Nov 2020 19:02:20 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:50252 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728561AbgKXACT (ORCPT ); Mon, 23 Nov 2020 19:02:19 -0500 Received: from mail-pf1-x444.google.com (mail-pf1-x444.google.com [IPv6:2607:f8b0:4864:20::444]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 25482C0613CF for ; Mon, 23 Nov 2020 16:02:19 -0800 (PST) Received: by mail-pf1-x444.google.com with SMTP id q10so16606768pfn.0 for ; Mon, 23 Nov 2020 16:02:19 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=dVZAZuhnHNDw7eBlpqaNTuLy1qB5NlqAynArFyBMzdU=; b=oMy+quRSFEMEHuIZhuSvFxsJnohoLaGB1uP5Yc2aYTkic1+FuPSrNECJ/ysK1f6gWT ers6BO0+t5vhO1N1mWWI/Weilyqpvb2RNRGU8ocpD7IWHTijus14RXUthJJlRX5ZiOt8 0wb18gTlJlXBwbt2oaLqZ9kgPv/QZWZlenQfM= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=dVZAZuhnHNDw7eBlpqaNTuLy1qB5NlqAynArFyBMzdU=; b=EDT9p65a5CmlO6APmPwXZaAB3QrMs+kRgY6qXqBMPwRAHfDsBD7DC7frDcNJpvKDxf LZA8BtkEs/tXTlXqaofH+ja/ODLe1RTmjuZAfMKBlc9rW3R2uO4449o98bpuncp9dMyX E2NiN+oa6UW+/Wwl8RPMF5O3KQ1LOD80mCR1Y/Cz7IvxE++/TRgDqRnWbVVpUnQBgknO rjN9goZtA48S3qO3z/gOcJ9Z8XPFt6Y75jVmvA+vlNVp8NPlWVGJlvtZHTo6X0ZTZCmI BfFIvq6n4hQI1gCR6MRt3E84w/U4tGNXE9bsmukmfsD+e9c6OGkoA+vln209YoniRnEe GWzw== X-Gm-Message-State: AOAM531czTev9SrZRPqDjwf2E+b5MrxmtzD8Yj/sy72b/VZPmNJXflJv UYDDzmfozKfIxcxEaXRmTDFZew== X-Google-Smtp-Source: ABdhPJyPk8xbY5YVwigff7KJFU/YiJcXqrjKKsxb1ubCLKovQVsaw+FkeVdl3VD2BcK+j4hLt4IA0Q== X-Received: by 2002:a63:215f:: with SMTP id s31mr1567453pgm.258.1606176138623; Mon, 23 Nov 2020 16:02:18 -0800 (PST) Received: from tictac2.mtv.corp.google.com ([2620:15c:202:1:42b0:34ff:fe3d:58e6]) by smtp.gmail.com with ESMTPSA id l133sm13091945pfd.112.2020.11.23.16.02.17 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 23 Nov 2020 16:02:18 -0800 (PST) From: Douglas Anderson To: Marc Zyngier , Thomas Gleixner , Jason Cooper , Linus Walleij Cc: Maulik Shah , Srinivas Ramana , Neeraj Upadhyay , Rajendra Nayak , linux-gpio@vger.kernel.org, linux-arm-msm@vger.kernel.org, Bjorn Andersson , Stephen Boyd , Douglas Anderson , Andy Gross , linux-kernel@vger.kernel.org Subject: [PATCH 1/3] irqchip: qcom-pdc: Fix phantom irq when changing between rising/falling Date: Mon, 23 Nov 2020 16:01:51 -0800 Message-Id: <20201123160139.1.I2702919afc253e2a451bebc3b701b462b2d22344@changeid> X-Mailer: git-send-email 2.29.2.454.gaff20da3a2-goog MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-gpio@vger.kernel.org We have a problem if we use gpio-keys and configure wakeups such that we only want one edge to wake us up. AKA: wakeup-event-action = ; wakeup-source; Specifically we end up with a phantom interrupt that blocks suspend if the line was already high and we want wakeups on rising edges (AKA we want the GPIO to go low and then high again before we wake up). The opposite is also problematic. Specifically, here's what's happening today: 1. Normally, gpio-keys configures to look for both edges. Due to the current workaround introduced in commit c3c0c2e18d94 ("pinctrl: qcom: Handle broken/missing PDC dual edge IRQs on sc7180"), if the line was high we'd configure for falling edges. 2. At suspend time, we change to look for rising edges. 3. After qcom_pdc_gic_set_type() runs, we get a phantom interrupt. We can solve this by just clearing the phantom interrupt. NOTE: it is possible that this could cause problems for a client with very specific needs, but there's not much we can do with this hardware. As an example, let's say the interrupt signal is currently high and the client is looking for falling edges. The client now changes to look for rising edges. The client could possibly expect that if the line has a short pulse low (and back high) that it would always be detected. Specifically no matter when the pulse happened, it should either have tripped the (old) falling edge trigger or the (new) rising edge trigger. We will simply not trip it. We could narrow down the race a bit by polling our parent before changing types, but no matter what we do there will still be a period of time where we can't tell the difference between a real transition (or more than one transition) and the phantom. Signed-off-by: Douglas Anderson Reviewed-by: Maulik Shah Tested-by: Maulik Shah --- drivers/irqchip/qcom-pdc.c | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) diff --git a/drivers/irqchip/qcom-pdc.c b/drivers/irqchip/qcom-pdc.c index bd39e9de6ecf..7d097164aadc 100644 --- a/drivers/irqchip/qcom-pdc.c +++ b/drivers/irqchip/qcom-pdc.c @@ -159,6 +159,8 @@ static int qcom_pdc_gic_set_type(struct irq_data *d, unsigned int type) { int pin_out = d->hwirq; enum pdc_irq_config_bits pdc_type; + enum pdc_irq_config_bits old_pdc_type; + int ret; if (pin_out == GPIO_NO_WAKE_IRQ) return 0; @@ -187,9 +189,24 @@ static int qcom_pdc_gic_set_type(struct irq_data *d, unsigned int type) return -EINVAL; } + old_pdc_type = pdc_reg_read(IRQ_i_CFG, pin_out); pdc_reg_write(IRQ_i_CFG, pin_out, pdc_type); - return irq_chip_set_type_parent(d, type); + ret = irq_chip_set_type_parent(d, type); + + /* + * When we change types the PDC can give a phantom interrupt. + * Clear it. Specifically the phantom shows up if a line is already + * high and we change to rising or if a line is already low and we + * change to falling but let's be consistent and clear it always. + * + * Doing this works because we have IRQCHIP_SET_TYPE_MASKED so the + * interrupt will be cleared before the rest of the system sees it. + */ + if (old_pdc_type != pdc_type) + irq_chip_set_parent_state(d, IRQCHIP_STATE_PENDING, 0); + + return ret; } static struct irq_chip qcom_pdc_gic_chip = { From patchwork Tue Nov 24 00:01:52 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Doug Anderson X-Patchwork-Id: 1405186 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=vger.kernel.org (client-ip=23.128.96.18; helo=vger.kernel.org; envelope-from=linux-gpio-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=chromium.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256 header.s=google header.b=RQis3WvQ; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by ozlabs.org (Postfix) with ESMTP id 4Cg4352Jgrz9sSs for ; Tue, 24 Nov 2020 11:02:41 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728658AbgKXAC1 (ORCPT ); Mon, 23 Nov 2020 19:02:27 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:50264 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728606AbgKXACW (ORCPT ); Mon, 23 Nov 2020 19:02:22 -0500 Received: from mail-pl1-x642.google.com (mail-pl1-x642.google.com [IPv6:2607:f8b0:4864:20::642]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id C9C45C0613CF for ; Mon, 23 Nov 2020 16:02:20 -0800 (PST) Received: by mail-pl1-x642.google.com with SMTP id x15so9713253pll.2 for ; Mon, 23 Nov 2020 16:02:20 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=fw5pTrGt8cRDmQt546J+zjyV3tc7NtMZfxIm6B0Y+tw=; b=RQis3WvQ1SNRz7zQmrKK719gPBOlh25OBcvT9jAzxTFjvXsWrciBEkYt6v5Zbg91NT O9/zfn97owjAoxgtAYYMfa1/PCOYfbBzGoggSWHS4D7n/T/cXYhayNTDcE9VD8+APent D1L8TRD43pBJ6GAXCz1pVD+1PiT04N1XwsGx4= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=fw5pTrGt8cRDmQt546J+zjyV3tc7NtMZfxIm6B0Y+tw=; b=nkuKOxyX4TcDbmb0wfel39TKYyxwQTqlyDP78hHQLuHvr7DHrAHxBuxIic0vxqC5tb WYuiZh19vk7xTu1N+MpBJMH6hVTNrXWnKikULPoIubjR49Bp+hK/LEZRH5uLfBe0z11b LpVjgEgtfZ9BBc/qx7vW5qan+GxiZCg4yuVyqJGHLaSMTAVS5FljzrX5VV2uwgHBt0uC b5Ar/qgaCIi7DVhZZj+LfUulm+AEKtqhDbZuUADrOn5sCE6YdqZmSU4cvy+dkcfUhJ6T 8AypCUH7qns2c8SpBzP3qhww+LJ494iPXmLkoG52WdSVSMfpwtuBA2Jjj5YXi7w692lQ MhLA== X-Gm-Message-State: AOAM533HM3T23lpgXTHj2+cObzJpkGZfBCGEZN1eYjxZe9fmI2Dvbkzl JUVDuQ3XtI17Fsb/xCwJUQqiBg== X-Google-Smtp-Source: ABdhPJymEA16wber3riwA2RZRoOOVY05U+0/J44V0CVG/BRoWQRqmko0ta0Gf78Zof85Gs/h/IR8ZA== X-Received: by 2002:a17:902:8b8c:b029:d8:de6f:ed35 with SMTP id ay12-20020a1709028b8cb02900d8de6fed35mr1751936plb.36.1606176140364; Mon, 23 Nov 2020 16:02:20 -0800 (PST) Received: from tictac2.mtv.corp.google.com ([2620:15c:202:1:42b0:34ff:fe3d:58e6]) by smtp.gmail.com with ESMTPSA id l133sm13091945pfd.112.2020.11.23.16.02.18 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 23 Nov 2020 16:02:19 -0800 (PST) From: Douglas Anderson To: Marc Zyngier , Thomas Gleixner , Jason Cooper , Linus Walleij Cc: Maulik Shah , Srinivas Ramana , Neeraj Upadhyay , Rajendra Nayak , linux-gpio@vger.kernel.org, linux-arm-msm@vger.kernel.org, Bjorn Andersson , Stephen Boyd , Douglas Anderson , Andy Gross , linux-kernel@vger.kernel.org Subject: [PATCH 2/3] pinctrl: qcom: Allow SoCs to specify a GPIO function that's not 0 Date: Mon, 23 Nov 2020 16:01:52 -0800 Message-Id: <20201123160139.2.I3ad184e3423d8e479bc3e86f5b393abb1704a1d1@changeid> X-Mailer: git-send-email 2.29.2.454.gaff20da3a2-goog In-Reply-To: <20201123160139.1.I2702919afc253e2a451bebc3b701b462b2d22344@changeid> References: <20201123160139.1.I2702919afc253e2a451bebc3b701b462b2d22344@changeid> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-gpio@vger.kernel.org There's currently a comment in the code saying function 0 is GPIO. Instead of hardcoding it, let's add a member where an SoC can specify it. No known SoCs use a number other than 0, but this just makes the code clearer. NOTE: no SoC code needs to be updated since we can rely on zero-initialization. Signed-off-by: Douglas Anderson --- drivers/pinctrl/qcom/pinctrl-msm.c | 4 ++-- drivers/pinctrl/qcom/pinctrl-msm.h | 2 ++ 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c index 77a25bdf0da7..588df91274e2 100644 --- a/drivers/pinctrl/qcom/pinctrl-msm.c +++ b/drivers/pinctrl/qcom/pinctrl-msm.c @@ -210,8 +210,8 @@ static int msm_pinmux_request_gpio(struct pinctrl_dev *pctldev, if (!g->nfuncs) return 0; - /* For now assume function 0 is GPIO because it always is */ - return msm_pinmux_set_mux(pctldev, g->funcs[0], offset); + return msm_pinmux_set_mux(pctldev, + g->funcs[pctrl->soc->gpio_func], offset); } static const struct pinmux_ops msm_pinmux_ops = { diff --git a/drivers/pinctrl/qcom/pinctrl-msm.h b/drivers/pinctrl/qcom/pinctrl-msm.h index 333f99243c43..e31a5167c91e 100644 --- a/drivers/pinctrl/qcom/pinctrl-msm.h +++ b/drivers/pinctrl/qcom/pinctrl-msm.h @@ -118,6 +118,7 @@ struct msm_gpio_wakeirq_map { * @wakeirq_dual_edge_errata: If true then GPIOs using the wakeirq_map need * to be aware that their parent can't handle dual * edge interrupts. + * @gpio_func: Which function number is GPIO (usually 0). */ struct msm_pinctrl_soc_data { const struct pinctrl_pin_desc *pins; @@ -134,6 +135,7 @@ struct msm_pinctrl_soc_data { const struct msm_gpio_wakeirq_map *wakeirq_map; unsigned int nwakeirq_map; bool wakeirq_dual_edge_errata; + unsigned int gpio_func; }; extern const struct dev_pm_ops msm_pinctrl_dev_pm_ops; From patchwork Tue Nov 24 00:01:53 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Doug Anderson X-Patchwork-Id: 1405187 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=vger.kernel.org (client-ip=23.128.96.18; helo=vger.kernel.org; envelope-from=linux-gpio-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=pass (p=none dis=none) header.from=chromium.org Authentication-Results: ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=chromium.org header.i=@chromium.org header.a=rsa-sha256 header.s=google header.b=GEjSWduN; dkim-atps=neutral Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by ozlabs.org (Postfix) with ESMTP id 4Cg4354tq8z9sT6 for ; Tue, 24 Nov 2020 11:02:41 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728610AbgKXAC0 (ORCPT ); Mon, 23 Nov 2020 19:02:26 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:50278 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728614AbgKXACW (ORCPT ); Mon, 23 Nov 2020 19:02:22 -0500 Received: from mail-pl1-x632.google.com (mail-pl1-x632.google.com [IPv6:2607:f8b0:4864:20::632]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id ADE4FC061A51 for ; Mon, 23 Nov 2020 16:02:22 -0800 (PST) Received: by mail-pl1-x632.google.com with SMTP id l1so1457390pld.5 for ; Mon, 23 Nov 2020 16:02:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=L89oXQzcinW0XVkoPkY/8VBs+sljP+/OT5KD3oBT6VA=; b=GEjSWduNUiOKDZ327Z36o6TRnl57q/Pg1X0VBsUprtQZqfRdsapmA6Lk8twZOVkLfc SEkPZNNKosLv40o4esKxnKIchpxzAi0eiOdOMccM15oFCyZTgbnGmUyvzGJMZd0KbC/N EvIInLSEhOQ6e0aor7r09MI5VVzzi9rzjjIZk= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=L89oXQzcinW0XVkoPkY/8VBs+sljP+/OT5KD3oBT6VA=; b=EMVVTLro4x/hj9NkHSkevlXFMMBOAgJC8HaU0ARt2hI/6vb9X648JpCuyrAQALFEBW /ibIvW91prcYYZXBB55Frc98BYMyopgbTsPjmekmhuHRWTVadSC3Ba/0sGMJwqDM66rr Luv5zd6rONvFdsnkfpmJJdYpLXs7N1/pkb+RNXbyVBWH7rPcnPr9CwYZ8VUdXB4tGAXv 70eAead6iNf3Bi0j/BiRTNBVITkaLis9Iu331kvYI+3lqrICtPnhzxQbAq0l5+zvc6rk jE98TZloX/XdEA7XQxe+o41YUGAkvpgF+3tQ5w8yFa4Jb82zMPu9DCC4nx2t8iFU12yI dClw== X-Gm-Message-State: AOAM533F2CTpEfBEwPUbVc5tz/mODHmSr7dXssLlwnlJwoIFVDjucUE4 W/y8QdCKTc/zHwveUfaYa2bc3PVrEMJFkQ== X-Google-Smtp-Source: ABdhPJzIt+OXu8ATXhk6OvopLSMTU2ebqBFoV4Yj2uiZ6Tqox41miFCS+XKNLgw9fFBY2CNgMRiecQ== X-Received: by 2002:a17:902:c393:b029:d9:6e7:6787 with SMTP id g19-20020a170902c393b02900d906e76787mr1531403plg.78.1606176142049; Mon, 23 Nov 2020 16:02:22 -0800 (PST) Received: from tictac2.mtv.corp.google.com ([2620:15c:202:1:42b0:34ff:fe3d:58e6]) by smtp.gmail.com with ESMTPSA id l133sm13091945pfd.112.2020.11.23.16.02.20 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 23 Nov 2020 16:02:21 -0800 (PST) From: Douglas Anderson To: Marc Zyngier , Thomas Gleixner , Jason Cooper , Linus Walleij Cc: Maulik Shah , Srinivas Ramana , Neeraj Upadhyay , Rajendra Nayak , linux-gpio@vger.kernel.org, linux-arm-msm@vger.kernel.org, Bjorn Andersson , Stephen Boyd , Douglas Anderson , Andy Gross , linux-kernel@vger.kernel.org Subject: [PATCH 3/3] pinctrl: qcom: Clear possible pending irq when remuxing GPIOs Date: Mon, 23 Nov 2020 16:01:53 -0800 Message-Id: <20201123160139.3.I771b6594b2a4d5b7fe7e12a991a6640f46386e8d@changeid> X-Mailer: git-send-email 2.29.2.454.gaff20da3a2-goog In-Reply-To: <20201123160139.1.I2702919afc253e2a451bebc3b701b462b2d22344@changeid> References: <20201123160139.1.I2702919afc253e2a451bebc3b701b462b2d22344@changeid> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-gpio@vger.kernel.org Conceptually, we can envision the input on Qualcomm SoCs to pass through a bunch of blocks between coming into the chip and becoming a GPIO interrupt. From guessing and running a handful of tests, I believe that we can represent the state of the world with a drawing that looks something like this: +-----------------+ +-----------------+ +-----------------+ | INPUT | --> | PINMUX | | IS_INPUT | +-----------------+ | | --> | | | output bogus (?)| | output bogus (?)| | if not muxed | | if input disab. | +-----------------+ +-----------------+ | +---------------------------------------------------+--> to PDC | V +-----------------+ +-----------------+ +-----------------+ | INTR RAW ENABLE | | DETECTION LOGIC | | STATUS REGISTER | | | | | | | | output bogus (?)| --> | maybe handles | | latches inputs | | if disabled | | polarity diffs | --> | that are high | +-----------------+ | | | | | maybe debounces | | write 1 to clr | | level irqs | +-----------------+ +-----------------+ | | +---------------------------------------------------+ | V +-----------------+ | ENABLE | | | +-----------------+ | nothing passes | --> | SUMMARY IRQ | | through if | +-----------------+ | disabled | +-----------------+ The above might not be 100% exact, but for the purpose of this discussion, the point is that there are a whole bunch of gates and transformations on the input before it gets to the circuitry that generates interrupts. As you might guess, if you reconfigure one of the gates in the above diagram while the system is configured to detect interrupts things get a little wacky. Specifically it appears that if you gate the input at any step it can cause various glitches in the later steps because they are still paying attention to their input but their input isn't really sane anymore. I did some poking and I found that I could generate bogus interrupts in the system both when muxing away from GPIO mode and also when muxing back to GPIO mode. When configured to use the PDC path for generating interrupts I found that if the external input on the GPIO was low that I'd get what looked like a rising edge when unmuxing and a falling edge when muxing. When configured away from the PDC path I got slightly different glitch interrupts when changing muxing. These glitches when remuxing matter in reality, not just in theory. To be concrete, let's take the special "wakeup_irq" in qcom_geni_serial.c as an example. In sc7180-trogdor.dtsi we configure the uart3 to have two pinctrl states, sleep and default, and mux between the two during runtime PM and system suspend (see geni_se_resources_{on,off}() for more details). The difference between the sleep and default state is that the RX pin is muxed to a GPIO during sleep and muxed to the UART otherwise. When we switch between these two states we can generate the glitches talked about above because we're configured to look for edges but the transition from the gated input (which is bogus) to the real input can look like an edge. Historically the UART case above was handled by the fact that the "enable" function in the MSM GPIO driver did an "unmask and clear". This relied on the fact that the system happens to have the interrupt disabled until suspend time and that it would enable it after the pinmux change happened, thus clearing the interrupt. The historical solution, however, had a few problems. The first problem (that nobody seemed to have tripped) is that we can still get bogus interrupts if we remux when the interrupt isn't disabled during the muxing and re-enabled after. The second problem is that it violates how I believe that the interrupt enable path is supposed to work. In Linux, if a driver does disable_irq() and later does enable_irq() on its interrupt, I believe it's expecting these properties: * If an interrupt was pending when the driver disabled then it will still be pending after the driver re-enables. * If an edge-triggered interrupt comes in while an interrupt is disabled it should assert when the interrupt is re-enabled. If you think that the above sounds a lot like the disable_irq() and enable_irq() are supposed to be masking/unmasking the interrupt instead of disabling/enabling it then you've made an astute observation. Specifically when talking about interrupts, "mask" usually means to stop posting interrupts but keep tracking them and "disable" means to fully shut off interrupt detection. It's unfortunate that this is so confusing, but presumably this is all the way it is for historical reasons. Perhaps more confusing than the above is that, even though clients of IRQs themselves don't have a way to request mask/unmask vs. disable/enable calls, IRQ chips themselves can implement both. ...and yet more confusing is that if an IRQ chip implements disable/enable then they will be called when a client driver calls disable_irq() / enable_irq(). It does feel like some of the above could be cleared up. However, without any other core interrupt changes it should be clear that when an IRQ chip gets a request to "disable" an IRQ that it has to treat it like a mask of that IRQ. In any case, after that long interlude you can see that the "unmask and clear" can break things. Maulik tried to fix it so that we no longer did "unmask and clear" in commit 71266d9d3936 ("pinctrl: qcom: Move clearing pending IRQ to .irq_request_resources callback"), but it didn't work for two reasons: * It only tried to address the problem for interrupts that had parents (like the PDC). * It regressed the problem that the original clearing was trying to solve. I think we can safely assume that if someone muxes a pin to be something other than a GPIO and then muxes it back that we can clear any interrupts that were pending on it without violating any assumptions that client drivers are making. Presumably the client drivers are intentionally remuxing the pin away from a dedicated purpose to be a plain GPIO so they don't care what the pin state was before the mux switch and they don't expect to see the pin change level during this switch. Let's move the clearing of the IRQ to the pin muxing routine so that we'll clear a pending IRQ if we're muxing from some non-GPIO mode to a GPIO mode. Fixes: 71266d9d3936 ("pinctrl: qcom: Move clearing pending IRQ to .irq_request_resources callback") Signed-off-by: Douglas Anderson --- This is a pretty hairy little patch and presumably needs a good amount of testing / discussion before landing. If this patch is totally broken / wrong feel free to consider it as an RFC and suggest how it should be better. Also note: I wanted to put this in the same series as patch #1, but IMO that patch can stand on its own. If it looks Ok but we want to have lots of debate about this one, please land patch #1 on its own and we can split the series. I have done most of this patch testing on the Chrome OS 5.4 kernel tree (with many backports) but have sanity checked it on mainline. drivers/pinctrl/qcom/pinctrl-msm.c | 104 ++++++++++++++++++----------- 1 file changed, 64 insertions(+), 40 deletions(-) diff --git a/drivers/pinctrl/qcom/pinctrl-msm.c b/drivers/pinctrl/qcom/pinctrl-msm.c index 588df91274e2..e7c3927c7d54 100644 --- a/drivers/pinctrl/qcom/pinctrl-msm.c +++ b/drivers/pinctrl/qcom/pinctrl-msm.c @@ -166,14 +166,42 @@ static int msm_get_function_groups(struct pinctrl_dev *pctldev, return 0; } +static void msm_pinctrl_clear_pending_irq(struct msm_pinctrl *pctrl, + unsigned int group, + unsigned int irq) +{ + struct irq_data *d = irq_get_irq_data(irq); + const struct msm_pingroup *g; + unsigned long flags; + u32 val; + + if (!d) + return; + + if (d->parent_data && test_bit(d->hwirq, pctrl->skip_wake_irqs)) + irq_chip_set_parent_state(d, IRQCHIP_STATE_PENDING, 0); + + g = &pctrl->soc->groups[group]; + + raw_spin_lock_irqsave(&pctrl->lock, flags); + val = msm_readl_intr_status(pctrl, g); + val &= ~BIT(g->intr_status_bit); + msm_writel_intr_status(val, pctrl, g); + raw_spin_unlock_irqrestore(&pctrl->lock, flags); +} + static int msm_pinmux_set_mux(struct pinctrl_dev *pctldev, unsigned function, unsigned group) { struct msm_pinctrl *pctrl = pinctrl_dev_get_drvdata(pctldev); + struct gpio_chip *gc = &pctrl->chip; + unsigned int irq = irq_find_mapping(gc->irq.domain, group); const struct msm_pingroup *g; unsigned long flags; u32 val, mask; + u32 oldval; + u32 old_i; int i; g = &pctrl->soc->groups[group]; @@ -187,15 +215,26 @@ static int msm_pinmux_set_mux(struct pinctrl_dev *pctldev, if (WARN_ON(i == g->nfuncs)) return -EINVAL; - raw_spin_lock_irqsave(&pctrl->lock, flags); + disable_irq(irq); - val = msm_readl_ctl(pctrl, g); + raw_spin_lock_irqsave(&pctrl->lock, flags); + oldval = val = msm_readl_ctl(pctrl, g); val &= ~mask; val |= i << g->mux_bit; msm_writel_ctl(val, pctrl, g); - raw_spin_unlock_irqrestore(&pctrl->lock, flags); + /* + * Clear IRQs if switching to/from GPIO mode since muxing to/from + * the GPIO path can cause phantom edges. + */ + old_i = (oldval & mask) >> g->mux_bit; + if (old_i != i && + (i == pctrl->soc->gpio_func || old_i == pctrl->soc->gpio_func)) + msm_pinctrl_clear_pending_irq(pctrl, group, irq); + + enable_irq(irq); + return 0; } @@ -456,32 +495,45 @@ static const struct pinconf_ops msm_pinconf_ops = { static int msm_gpio_direction_input(struct gpio_chip *chip, unsigned offset) { const struct msm_pingroup *g; + unsigned int irq = irq_find_mapping(chip->irq.domain, offset); struct msm_pinctrl *pctrl = gpiochip_get_data(chip); unsigned long flags; + u32 oldval; u32 val; g = &pctrl->soc->groups[offset]; + disable_irq(irq); + raw_spin_lock_irqsave(&pctrl->lock, flags); - val = msm_readl_ctl(pctrl, g); + oldval = val = msm_readl_ctl(pctrl, g); val &= ~BIT(g->oe_bit); msm_writel_ctl(val, pctrl, g); raw_spin_unlock_irqrestore(&pctrl->lock, flags); + if (oldval != val) + msm_pinctrl_clear_pending_irq(pctrl, offset, irq); + + enable_irq(irq); + return 0; } static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, int value) { const struct msm_pingroup *g; + unsigned int irq = irq_find_mapping(chip->irq.domain, offset); struct msm_pinctrl *pctrl = gpiochip_get_data(chip); unsigned long flags; + u32 oldval; u32 val; g = &pctrl->soc->groups[offset]; + disable_irq(irq); + raw_spin_lock_irqsave(&pctrl->lock, flags); val = msm_readl_io(pctrl, g); @@ -491,12 +543,17 @@ static int msm_gpio_direction_output(struct gpio_chip *chip, unsigned offset, in val &= ~BIT(g->out_bit); msm_writel_io(val, pctrl, g); - val = msm_readl_ctl(pctrl, g); + oldval = msm_readl_ctl(pctrl, g); val |= BIT(g->oe_bit); msm_writel_ctl(val, pctrl, g); raw_spin_unlock_irqrestore(&pctrl->lock, flags); + if (oldval != val) + msm_pinctrl_clear_pending_irq(pctrl, offset, irq); + + enable_irq(irq); + return 0; } @@ -774,7 +831,7 @@ static void msm_gpio_irq_mask(struct irq_data *d) raw_spin_unlock_irqrestore(&pctrl->lock, flags); } -static void msm_gpio_irq_clear_unmask(struct irq_data *d, bool status_clear) +static void msm_gpio_irq_unmask(struct irq_data *d) { struct gpio_chip *gc = irq_data_get_irq_chip_data(d); struct msm_pinctrl *pctrl = gpiochip_get_data(gc); @@ -792,17 +849,6 @@ static void msm_gpio_irq_clear_unmask(struct irq_data *d, bool status_clear) raw_spin_lock_irqsave(&pctrl->lock, flags); - if (status_clear) { - /* - * clear the interrupt status bit before unmask to avoid - * any erroneous interrupts that would have got latched - * when the interrupt is not in use. - */ - val = msm_readl_intr_status(pctrl, g); - val &= ~BIT(g->intr_status_bit); - msm_writel_intr_status(val, pctrl, g); - } - val = msm_readl_intr_cfg(pctrl, g); val |= BIT(g->intr_raw_status_bit); val |= BIT(g->intr_enable_bit); @@ -815,14 +861,10 @@ static void msm_gpio_irq_clear_unmask(struct irq_data *d, bool status_clear) static void msm_gpio_irq_enable(struct irq_data *d) { - struct gpio_chip *gc = irq_data_get_irq_chip_data(d); - struct msm_pinctrl *pctrl = gpiochip_get_data(gc); - if (d->parent_data) irq_chip_enable_parent(d); - if (!test_bit(d->hwirq, pctrl->skip_wake_irqs)) - msm_gpio_irq_clear_unmask(d, true); + msm_gpio_irq_unmask(d); } static void msm_gpio_irq_disable(struct irq_data *d) @@ -837,11 +879,6 @@ static void msm_gpio_irq_disable(struct irq_data *d) msm_gpio_irq_mask(d); } -static void msm_gpio_irq_unmask(struct irq_data *d) -{ - msm_gpio_irq_clear_unmask(d, false); -} - /** * msm_gpio_update_dual_edge_parent() - Prime next edge for IRQs handled by parent. * @d: The irq dta. @@ -1097,19 +1134,6 @@ static int msm_gpio_irq_reqres(struct irq_data *d) ret = -EINVAL; goto out; } - - /* - * Clear the interrupt that may be pending before we enable - * the line. - * This is especially a problem with the GPIOs routed to the - * PDC. These GPIOs are direct-connect interrupts to the GIC. - * Disabling the interrupt line at the PDC does not prevent - * the interrupt from being latched at the GIC. The state at - * GIC needs to be cleared before enabling. - */ - if (d->parent_data && test_bit(d->hwirq, pctrl->skip_wake_irqs)) - irq_chip_set_parent_state(d, IRQCHIP_STATE_PENDING, 0); - return 0; out: module_put(gc->owner);