]> asedeno.scripts.mit.edu Git - linux.git/commitdiff
staging: comedi: ni_mio_common: use insn->n in ni_eeprom_insn_read()
authorIan Abbott <abbotti@mev.co.uk>
Tue, 19 Mar 2019 16:54:43 +0000 (16:54 +0000)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 20 Mar 2019 07:25:17 +0000 (08:25 +0100)
The `insn_read` handler for the EEPROM subdevice on E-series boards
(`ni_eeprom_insn_read()`) currently ignores `insn->n` (the number of
samples to read) and assumes a single sample is to be read into
`data[0]`.  Fortunately, the Comedi core ensures that `data[]` has a
length of at least 16 so there is no problem with array bounds.

The usual Comedi convention for `insn_read` handlers is to read the same
channel `insn->n` times into successive elements of `data[]` so let's do
that.  (Each channel number corresponds to a single EEPROM address.)
Since we do not expect the EEPROM data at a particular address to change
between readings, let's just read it once and copy the value `insn->n`
times.

Also, follow the usual Comedi convention and return `insn->n` from the
handler to indicate success (although any non-negative value will do).

Signed-off-by: Ian Abbott <abbotti@mev.co.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/staging/comedi/drivers/ni_mio_common.c

index e008095436d7ecf0d8d5e37eb2b2e240053d8456..5b397ad2a6043bc50af1c41027fc2948eb398042 100644 (file)
@@ -4507,9 +4507,15 @@ static int ni_eeprom_insn_read(struct comedi_device *dev,
                               struct comedi_insn *insn,
                               unsigned int *data)
 {
-       data[0] = ni_read_eeprom(dev, CR_CHAN(insn->chanspec));
+       unsigned int val;
+       unsigned int i;
 
-       return 1;
+       if (insn->n) {
+               val = ni_read_eeprom(dev, CR_CHAN(insn->chanspec));
+               for (i = 0; i < insn->n; i++)
+                       data[i] = val;
+       }
+       return insn->n;
 }
 
 static int ni_m_series_eeprom_insn_read(struct comedi_device *dev,