Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932196AbbFHRu5 (ORCPT ); Mon, 8 Jun 2015 13:50:57 -0400 Received: from mail-pd0-f176.google.com ([209.85.192.176]:35148 "EHLO mail-pd0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752335AbbFHRuu (ORCPT ); Mon, 8 Jun 2015 13:50:50 -0400 From: lucas.de.marchi@gmail.com To: linux-i2c@vger.kernel.org Cc: linux-kernel@vger.kernel.org, Wolfram Sang , Jarkko Nikula , Mika Westerberg , Fabio Mello , Lucas De Marchi Subject: [PATCH] i2c: designware: use enable on resume instead initialization Date: Mon, 8 Jun 2015 14:50:28 -0300 Message-Id: <1433785828-4100-1-git-send-email-lucas.de.marchi@gmail.com> X-Mailer: git-send-email 2.4.2 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4155 Lines: 113 From: Fabio Mello According to documentation and tests, initialization is not necessary on module resume, since the controller keeps its state between disable/enable. Change the target address is also allowed. So, this patch replaces the initialization on module resume with a simple enable, and removes the (non required anymore) enables and disables. Signed-off-by: Fabio Mello Signed-off-by: Lucas De Marchi --- These pictures explain a little more the consequence of letting the enable+disable in the code: http://pub.politreco.com/paste/TEK0011-before.jpg http://pub.politreco.com/paste/TEK0007-after.jpg The yellow line is a GPIO toggle in userspace to mark when we start and finish the i2c transactions. The blue line is the SCL in that i2c bus. Take a look on the huge pauses we have between any 2 transactions. These pauses are removed with this patch and we are able to read our sensor's values in 950usec rather than 5.24msec we had before. We are testing this using a Minnowboard Max that has a designware i2c controller. There's this comment in the code suggesting the designware controller might have problems if we don't disable it after each transfer. We left a stress code running for 3 days to check if anything bad would happen. This is the test code using a PCA9685 (just because it was the easiest device we had available to check read+write commands): http://pub.politreco.com/paste/test-i2c.c drivers/i2c/busses/i2c-designware-core.c | 19 ++++--------------- drivers/i2c/busses/i2c-designware-platdrv.c | 2 +- 2 files changed, 5 insertions(+), 16 deletions(-) diff --git a/drivers/i2c/busses/i2c-designware-core.c b/drivers/i2c/busses/i2c-designware-core.c index 6e25c01..b386c10 100644 --- a/drivers/i2c/busses/i2c-designware-core.c +++ b/drivers/i2c/busses/i2c-designware-core.c @@ -375,8 +375,12 @@ int i2c_dw_init(struct dw_i2c_dev *dev) /* configure the i2c master */ dw_writel(dev, dev->master_cfg , DW_IC_CON); + /* Enable the adapter */ + __i2c_dw_enable(dev, true); + if (dev->release_lock) dev->release_lock(dev); + return 0; } EXPORT_SYMBOL_GPL(i2c_dw_init); @@ -405,9 +409,6 @@ static void i2c_dw_xfer_init(struct dw_i2c_dev *dev) struct i2c_msg *msgs = dev->msgs; u32 ic_con, ic_tar = 0; - /* Disable the adapter */ - __i2c_dw_enable(dev, false); - /* if the slave address is ten bit address, enable 10BITADDR */ ic_con = dw_readl(dev, DW_IC_CON); if (msgs[dev->msg_write_idx].flags & I2C_M_TEN) { @@ -434,9 +435,6 @@ static void i2c_dw_xfer_init(struct dw_i2c_dev *dev) /* enforce disabled interrupts (due to HW issues) */ i2c_dw_disable_int(dev); - /* Enable the adapter */ - __i2c_dw_enable(dev, true); - /* Clear and enable interrupts */ i2c_dw_clear_int(dev); dw_writel(dev, DW_IC_INTR_DEFAULT_MASK, DW_IC_INTR_MASK); @@ -665,15 +663,6 @@ i2c_dw_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num) goto done; } - /* - * We must disable the adapter before unlocking the &dev->lock mutex - * below. Otherwise the hardware might continue generating interrupts - * which in turn causes a race condition with the following transfer. - * Needs some more investigation if the additional interrupts are - * a hardware bug or this driver doesn't handle them correctly yet. - */ - __i2c_dw_enable(dev, false); - if (dev->msg_err) { ret = dev->msg_err; goto done; diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c index c270f5f..0598695 100644 --- a/drivers/i2c/busses/i2c-designware-platdrv.c +++ b/drivers/i2c/busses/i2c-designware-platdrv.c @@ -320,7 +320,7 @@ static int dw_i2c_resume(struct device *dev) clk_prepare_enable(i_dev->clk); if (!i_dev->pm_runtime_disabled) - i2c_dw_init(i_dev); + i2c_dw_enable(i_dev); return 0; } -- 2.4.2 -- 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/