Hi, Jean Delvare wrote: >>video:/soft/src/lm_sensors2/kernel/chips # cat \ >> /sys/bus/i2c/devices/0-0073/fan_div1 >>2 >>video:/soft/src/lm_sensors2/kernel/chips # echo 1 > \ >> /sys/bus/i2c/devices/0-0073/fan_div1 >>video:/soft/src/lm_sensors2/kernel/chips # cat \ >> /sys/bus/i2c/devices/0-0073/fan_div1 >>2 >>video:/soft/src/lm_sensors2/kernel/chips # >> >>Nothing special. The invalid value is simply ignored. Although >>set_fan_div() returns -1 in that case, echo doesn't show any error. > > Could you please check the value returned by echo? (i.e. "echo 1 > > fan_div1 ; echo $?"). I don't really think it can show the error, but > who knows... But why would we be allowed to return errors in these > functions if we cannot catch them later? Surprize! The driver's return value has an influence on echo's return value :-)) In the case, where my driver returns -1, echo returns 1, otherwise 0. >>How about adding a printk() for that case? > > Sound like a good idea (although I think you have to use dev_info() or > something similar, not printk(), in Linux 2.6). I see, dev_info() will always print a message, while dev_dbg() print them only only when debug messages are requested. So, although an application, that is writing the sysfs fan_div file, has a chance to detect, whether it worked or not, it might still be a good idea in this special case to show a message with acceptable values. The attached version now outputs a message like below: i2c_adapter i2c-0: fan_div value 3 not supported. Choose one of 2, 4 or 8! Bye. -- Dipl.-Inform. (FH) Reinhard Nissl mailto:rnissl at gmx.de -------------- next part -------------- A non-text attachment was scrubbed... Name: fscher.c.gz Type: application/x-gzip Size: 5037 bytes Desc: not available Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20040128/fef3c67e/attachment.gz
