spi: fix spi_sync
RFC PATCH 1/2 spi: fix spi_sync
Ease spi_sync calling convention. There is no more need to check the return value AND the message->status. As a side effect, this will correct the wrong behaviour of spi_read and spi_write.
---
At this time (2.6.23.1), spi_read and spi_write functions only check the return value of spi_sync. This is wrong since they should check for message->status too.
After a short inspection of all files using spi_sync in the kernel, it seems that we should change the spi_sync function (rather than all users).
Inspection details:
files using spi_sync without reading message->status:drivers/mtd/devices/m25p80.c drivers/mtd/devices/mtd_dataflash.c drivers/spi/tle62x0.cdrivers/spi/spidev.cdrivers/spi/at25.cinclude/linux/spi.h
files using spi_sync, but executing some unnecessary code when the return value or message->status report an error:drivers/rtc/rtc-max6902.cdrivers/spi/spi.cdrivers/input/touchscreen/ads7846.c
files using spi_sync the Right Way(TM):
[no files]
Regards
Marc
spi/spi.c renamed from spi/spi.c.orig +5 -8
@@ -542,10 +542,7 @@ static void spi_complete(void *arg)* Also, the caller is guaranteeing that the memory associated with the * message will not be freed before this call returns. * -* The return value is a negative error code if the message could not be -* submitted, else zero. When the value is zero, then message->status is -* also defined; it's the completion code for the transfer, either zero -* or a negative error code from the controller driver. +* It returns zero on success, else a negative error code. */ int spi_sync (struct spi_device *spi,struct spi_message *message) {
@@ -555,8 +552,10 @@ int spi_sync(struct spi_device *spi, str message->complete = spi_complete; message->context = &done; status =spi_async (spi, message); -if (status ==0 ) +if (status ==0 ) {wait_for_completion (&done); + status = m->status; + } message->context =NULL ;return status; }
@@ -629,10 +628,8 @@ int spi_write_then_read(struct spi_devic/* do the i/o */ status =spi_sync (spi, &message); -if (status ==0 ) { +if (status ==0 )memcpy (rxbuf, x[1 ].rx_buf, n_rx); - status = message.status; - }if (x[0 ].tx_buf == buf)up (&lock);
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/