public inbox for linux-i2c@vger.kernel.org
 help / color / mirror / Atom feed
From: <Codrin.Ciubotariu@microchip.com>
To: linux-i2c@vger.kernel.org, devicetree@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org, kamel.bouhara@bootlin.com,
	wsa@the-dreams.de, Nicolas.Ferre@microchip.com,
	alexandre.belloni@bootlin.com, robh@kernel.org
Subject: Re: [PATCH v2 3/6] i2c: at91: Send bus clear command if SCL is down
Date: Thu, 9 Jan 2020 10:58:22 +0000	[thread overview]
Message-ID: <d2fb2414-5c7f-3bc0-caf5-8927e6a2ccee@microchip.com> (raw)
In-Reply-To: <20200109074729.zare4va5ugbq62f3@M43218.corp.atmel.com>

On 09.01.2020 09:47, Ludovic Desroches wrote:
> On Fri, Jan 03, 2020 at 09:49:07AM +0000, Codrin Ciubotariu - M19940 wrote:
>> After a transfer timeout, some faulty I2C slave devices might hold down
>> the SCL pin. We can generate a bus clear command, hoping that the slave
>> might release the pins.
>>
>> Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
> 
> This patch may conflict with another one removing the explicit set to
> false.

Yes, Eugen's patches got accepted in the meantime. I will rebase and 
adapt to the new style.

> 
>> ---
>>
>> Changes in v2:
>>   - use CLEAR command only if SDA is down; update patch subject to
>>     reflect this;
>>   - CLEAR command is no longer used for sama5d2, only sam9x60;
>>
>>   drivers/i2c/busses/i2c-at91-core.c   |  9 ++++++++
>>   drivers/i2c/busses/i2c-at91-master.c | 32 +++++++++++++++++++++++-----
>>   drivers/i2c/busses/i2c-at91.h        |  7 +++++-
>>   3 files changed, 42 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/i2c/busses/i2c-at91-core.c b/drivers/i2c/busses/i2c-at91-core.c
>> index e13af4874976..128eafc05adb 100644
>> --- a/drivers/i2c/busses/i2c-at91-core.c
>> +++ b/drivers/i2c/busses/i2c-at91-core.c
>> @@ -71,6 +71,7 @@ static struct at91_twi_pdata at91rm9200_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata at91sam9261_config = {
>> @@ -82,6 +83,7 @@ static struct at91_twi_pdata at91sam9261_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata at91sam9260_config = {
>> @@ -93,6 +95,7 @@ static struct at91_twi_pdata at91sam9260_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata at91sam9g20_config = {
>> @@ -104,6 +107,7 @@ static struct at91_twi_pdata at91sam9g20_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata at91sam9g10_config = {
>> @@ -115,6 +119,7 @@ static struct at91_twi_pdata at91sam9g10_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static const struct platform_device_id at91_twi_devtypes[] = {
>> @@ -148,6 +153,7 @@ static struct at91_twi_pdata at91sam9x5_config = {
>>   	.has_dig_filtr = false,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata sama5d4_config = {
>> @@ -159,6 +165,7 @@ static struct at91_twi_pdata sama5d4_config = {
>>   	.has_dig_filtr = true,
>>   	.has_adv_dig_filtr = false,
>>   	.has_ana_filtr = false,
>> +	.has_clear_cmd = false,
>>   };
>>   
>>   static struct at91_twi_pdata sama5d2_config = {
>> @@ -170,6 +177,7 @@ static struct at91_twi_pdata sama5d2_config = {
>>   	.has_dig_filtr = true,
>>   	.has_adv_dig_filtr = true,
>>   	.has_ana_filtr = true,
>> +	.has_clear_cmd = false,	/* due to errata, CLEAR cmd is not working */
> 
> In this case it's probably worth keeping it.

Yes, I also think it's useful to let this one.

Thanks and best regards,
Codrin

> 
> Ludovic
> 
>>   };
>>   
>>   static struct at91_twi_pdata sam9x60_config = {
>> @@ -181,6 +189,7 @@ static struct at91_twi_pdata sam9x60_config = {
>>   	.has_dig_filtr = true,
>>   	.has_adv_dig_filtr = true,
>>   	.has_ana_filtr = true,
>> +	.has_clear_cmd = true,
>>   };
>>   
>>   static const struct of_device_id atmel_twi_dt_ids[] = {
>> diff --git a/drivers/i2c/busses/i2c-at91-master.c b/drivers/i2c/busses/i2c-at91-master.c
>> index c55e8ff35201..8c4f20317847 100644
>> --- a/drivers/i2c/busses/i2c-at91-master.c
>> +++ b/drivers/i2c/busses/i2c-at91-master.c
>> @@ -480,7 +480,7 @@ static int at91_do_twi_transfer(struct at91_twi_dev *dev)
>>   	unsigned long time_left;
>>   	bool has_unre_flag = dev->pdata->has_unre_flag;
>>   	bool has_alt_cmd = dev->pdata->has_alt_cmd;
>> -	struct i2c_bus_recovery_info *rinfo = &dev->rinfo;
>> +	bool has_clear_cmd = dev->pdata->has_clear_cmd;
>>   
>>   	/*
>>   	 * WARNING: the TXCOMP bit in the Status Register is NOT a clear on
>> @@ -641,10 +641,32 @@ static int at91_do_twi_transfer(struct at91_twi_dev *dev)
>>   			       AT91_TWI_THRCLR | AT91_TWI_LOCKCLR);
>>   	}
>>   
>> -	if (rinfo->get_sda && !(rinfo->get_sda(&dev->adapter))) {
>> -		dev_dbg(dev->dev,
>> -			"SDA is down; clear bus using gpio\n");
>> -		i2c_recover_bus(&dev->adapter);
>> +	/*
>> +	 * some faulty I2C slave devices might hold SDA down;
>> +	 * we can send a bus clear command, hoping that the pins will be
>> +	 * released
>> +	 */
>> +	if (has_clear_cmd) {
>> +		if (!(dev->transfer_status & AT91_TWI_SDA)) {
>> +			dev_dbg(dev->dev,
>> +				"SDA is down; sending bus clear command\n");
>> +			if (dev->use_alt_cmd) {
>> +				unsigned int acr;
>> +
>> +				acr = at91_twi_read(dev, AT91_TWI_ACR);
>> +				acr &= ~AT91_TWI_ACR_DATAL_MASK;
>> +				at91_twi_write(dev, AT91_TWI_ACR, acr);
>> +			}
>> +			at91_twi_write(dev, AT91_TWI_CR, AT91_TWI_CLEAR);
>> +		}
>> +	} else {
>> +		struct i2c_bus_recovery_info *rinfo = &dev->rinfo;
>> +
>> +		if (rinfo->get_sda && !(rinfo->get_sda(&dev->adapter))) {
>> +			dev_dbg(dev->dev,
>> +				"SDA is down; clear bus using gpio\n");
>> +			i2c_recover_bus(&dev->adapter);
>> +		}
>>   	}
>>   
>>   	return ret;
>> diff --git a/drivers/i2c/busses/i2c-at91.h b/drivers/i2c/busses/i2c-at91.h
>> index eb335b71e775..80f77d523346 100644
>> --- a/drivers/i2c/busses/i2c-at91.h
>> +++ b/drivers/i2c/busses/i2c-at91.h
>> @@ -36,6 +36,7 @@
>>   #define	AT91_TWI_SVDIS		BIT(5)	/* Slave Transfer Disable */
>>   #define	AT91_TWI_QUICK		BIT(6)	/* SMBus quick command */
>>   #define	AT91_TWI_SWRST		BIT(7)	/* Software Reset */
>> +#define	AT91_TWI_CLEAR		BIT(15) /* Bus clear command */
>>   #define	AT91_TWI_ACMEN		BIT(16) /* Alternative Command Mode Enable */
>>   #define	AT91_TWI_ACMDIS		BIT(17) /* Alternative Command Mode Disable */
>>   #define	AT91_TWI_THRCLR		BIT(24) /* Transmit Holding Register Clear */
>> @@ -69,6 +70,8 @@
>>   #define	AT91_TWI_NACK		BIT(8)	/* Not Acknowledged */
>>   #define	AT91_TWI_EOSACC		BIT(11)	/* End Of Slave Access */
>>   #define	AT91_TWI_LOCK		BIT(23) /* TWI Lock due to Frame Errors */
>> +#define	AT91_TWI_SCL		BIT(24) /* TWI SCL status */
>> +#define	AT91_TWI_SDA		BIT(25) /* TWI SDA status */
>>   
>>   #define	AT91_TWI_INT_MASK \
>>   	(AT91_TWI_TXCOMP | AT91_TWI_RXRDY | AT91_TWI_TXRDY | AT91_TWI_NACK \
>> @@ -81,7 +84,8 @@
>>   #define	AT91_TWI_THR		0x0034	/* Transmit Holding Register */
>>   
>>   #define	AT91_TWI_ACR		0x0040	/* Alternative Command Register */
>> -#define	AT91_TWI_ACR_DATAL(len)	((len) & 0xff)
>> +#define	AT91_TWI_ACR_DATAL_MASK	GENMASK(15, 0)
>> +#define	AT91_TWI_ACR_DATAL(len)	((len) & AT91_TWI_ACR_DATAL_MASK)
>>   #define	AT91_TWI_ACR_DIR	BIT(8)
>>   
>>   #define AT91_TWI_FILTR		0x0044
>> @@ -118,6 +122,7 @@ struct at91_twi_pdata {
>>   	bool has_dig_filtr;
>>   	bool has_adv_dig_filtr;
>>   	bool has_ana_filtr;
>> +	bool has_clear_cmd;
>>   	struct at_dma_slave dma_slave;
>>   };
>>   
>> -- 
>> 2.20.1

  reply	other threads:[~2020-01-09 10:58 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-01-03  9:49 [PATCH v2 0/6] i2c bus recovery for Microchip SoCs Codrin.Ciubotariu
2020-01-03  9:49 ` [PATCH v2 1/6] dt-bindings: i2c: at91: document optional bus recovery properties Codrin.Ciubotariu
2020-01-03 22:29   ` Rob Herring
2020-01-06 16:34     ` Codrin.Ciubotariu
2020-01-03  9:49 ` [PATCH v2 2/6] i2c: at91: implement i2c bus recovery Codrin.Ciubotariu
2020-01-09  7:48   ` Ludovic Desroches
     [not found]     ` <20200109074819.rhlaxg3sgwlng5xm-cg6aFISJjbMT2m6MMfBamRL4W9x8LtSr@public.gmane.org>
2020-01-09 10:54       ` Codrin.Ciubotariu-UWL1GkI3JZL3oGB3hsPCZA
2020-01-03  9:49 ` [PATCH v2 3/6] i2c: at91: Send bus clear command if SCL is down Codrin.Ciubotariu
2020-01-09  7:47   ` Ludovic Desroches
2020-01-09 10:58     ` Codrin.Ciubotariu [this message]
2020-01-09 11:05   ` Russell King - ARM Linux admin
     [not found]     ` <20200109110557.GO25745-QEMnZ+CodIVURfEZ8mYm6t73F7V6hmMc@public.gmane.org>
2020-01-09 13:29       ` Codrin.Ciubotariu-UWL1GkI3JZL3oGB3hsPCZA
2020-01-03  9:49 ` [PATCH v2 5/6] ARM: at91/dt: sama5d4: add i2c gpio pinctrl Codrin.Ciubotariu
2020-01-03  9:49 ` [PATCH v2 4/6] ARM: at91/dt: sama5d3: " Codrin.Ciubotariu
2020-01-04 22:39   ` Peter Rosin
2020-01-06 16:58     ` Codrin.Ciubotariu
2020-01-03  9:49 ` [PATCH v2 6/6] ARM: at91/dt: sama5d2: " Codrin.Ciubotariu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=d2fb2414-5c7f-3bc0-caf5-8927e6a2ccee@microchip.com \
    --to=codrin.ciubotariu@microchip.com \
    --cc=Nicolas.Ferre@microchip.com \
    --cc=alexandre.belloni@bootlin.com \
    --cc=devicetree@vger.kernel.org \
    --cc=kamel.bouhara@bootlin.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-i2c@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=robh@kernel.org \
    --cc=wsa@the-dreams.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox