Patchwork i2c: mxs: Better error handling

login
register
mail settings
Submitter Alexandre Belloni
Date July 3, 2013, 8:59 a.m.
Message ID <1372841962-26310-1-git-send-email-alexandre.belloni@free-electrons.com>
Download mbox | patch
Permalink /patch/256565/
State Deferred
Headers show

Comments

Alexandre Belloni - July 3, 2013, 8:59 a.m.
This patch fixes two error handling cases in mxs_i2c_xfer_msg():

 - in PIO mode, mxs_i2c_pio_setup_xfer() may return an error but i2c->cmd_err
 will not be set so mxs_i2c_xfer_msg() fails silently.

 - in DMA mode, mxs_i2c_dma_setup_xfer() may return an error and have
 i2c->cmd_err set to -ENXIO. In that case, we want to go through the
 MXS_I2C_CTRL1_CLR_GOT_A_NAK before returning.

Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
---
 drivers/i2c/busses/i2c-mxs.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
Wolfram Sang - Aug. 15, 2013, 10:03 a.m.
On Wed, Jul 03, 2013 at 10:59:22AM +0200, Alexandre Belloni wrote:
> This patch fixes two error handling cases in mxs_i2c_xfer_msg():
> 
>  - in PIO mode, mxs_i2c_pio_setup_xfer() may return an error but i2c->cmd_err
>  will not be set so mxs_i2c_xfer_msg() fails silently.
> 
>  - in DMA mode, mxs_i2c_dma_setup_xfer() may return an error and have
>  i2c->cmd_err set to -ENXIO. In that case, we want to go through the
>  MXS_I2C_CTRL1_CLR_GOT_A_NAK before returning.

A short explanation why we want that would be perfect.

> 
> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>

I'd like to postpone this until I applied Marek's PIO rework. Could you
respin after that?
Alexandre Belloni - Aug. 15, 2013, 11:09 a.m.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi Wolfram,

On 15/08/2013 12:03, Wolfram Sang wrote:
> On Wed, Jul 03, 2013 at 10:59:22AM +0200, Alexandre Belloni wrote:
>> This patch fixes two error handling cases in mxs_i2c_xfer_msg():
>>
>>  - in PIO mode, mxs_i2c_pio_setup_xfer() may return an error but
i2c->cmd_err
>>  will not be set so mxs_i2c_xfer_msg() fails silently.
>>
>>  - in DMA mode, mxs_i2c_dma_setup_xfer() may return an error and have
>>  i2c->cmd_err set to -ENXIO. In that case, we want to go through the
>>  MXS_I2C_CTRL1_CLR_GOT_A_NAK before returning.
>
> A short explanation why we want that would be perfect.
>
>>
>> Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
>
> I'd like to postpone this until I applied Marek's PIO rework. Could you
> respin after that?
>

Actually, this patch won't be needed after Marek's PIO rework. You can
drop it now.

- -- 
Alexandre Belloni, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
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=bjCC
-----END PGP SIGNATURE-----

--
To unsubscribe from this list: send the line "unsubscribe linux-i2c" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/drivers/i2c/busses/i2c-mxs.c b/drivers/i2c/busses/i2c-mxs.c
index 2039f23..92f22b0 100644
--- a/drivers/i2c/busses/i2c-mxs.c
+++ b/drivers/i2c/busses/i2c-mxs.c
@@ -502,7 +502,7 @@  static int mxs_i2c_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg,
 		INIT_COMPLETION(i2c->cmd_complete);
 		ret = mxs_i2c_dma_setup_xfer(adap, msg, flags);
 		if (ret)
-			return ret;
+			goto exit;
 
 		ret = wait_for_completion_timeout(&i2c->cmd_complete,
 						msecs_to_jiffies(1000));
@@ -510,6 +510,7 @@  static int mxs_i2c_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg,
 			goto timeout;
 	}
 
+exit:
 	if (i2c->cmd_err == -ENXIO) {
 		/*
 		 * If the transfer fails with a NAK from the slave the
@@ -519,7 +520,8 @@  static int mxs_i2c_xfer_msg(struct i2c_adapter *adap, struct i2c_msg *msg,
 		       i2c->regs + MXS_I2C_CTRL1_SET);
 	}
 
-	ret = i2c->cmd_err;
+	if (i2c->cmd_err)
+		ret = i2c->cmd_err;
 
 	dev_dbg(i2c->dev, "Done with err=%d\n", ret);