On 29.04.2013 20:10, Eldad Zack wrote:

Hi Daniel,

I think I found a endian usage issue in the caiaq driver. I have a patch to fix it (below), but it assumes that the driver works currently on little endian CPUs - I don't have a device to test it with.

I don't currently have any either, but your findings make sense. I just know that the driver currently works on LE CPUs.

Cheers, Eldad

-- >8 -- Subject: [PATCH] ALSA: usb-audio: caiaq: fix endianness bug

Current code does this:

be16_to_cpu(buf[i * 2] << 8 | buf[(i * 2) + 1])

Which is effectively:

be16_to_cpu(be16_to_cpu(*((u16 *) buf)))

This means the int16 in the buffer is not converted at all. Assuming the device-side structure is actually little endian, change the code to use le16_to_cpu().

Signed-off-by: Eldad Zack eldad@fogrefinery.com

Acked-by: Daniel Mack zonque@gmail.com

---

sound/usb/caiaq/input.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/sound/usb/caiaq/input.c b/sound/usb/caiaq/input.c index efc70ae..2f4ae4d 100644 --- a/sound/usb/caiaq/input.c +++ b/sound/usb/caiaq/input.c @@ -488,13 +488,12 @@ static void snd_usb_caiaq_maschine_dispatch(struct snd_usb_caiaqdev *cdev, unsigned int len) { unsigned int i, pad_id;

• uint16_t pressure;

• __le16 *pressure = (__le16 *) buf;

  for (i = 0; i < MASCHINE_PADS; i++) {

• pressure = be16_to_cpu(buf[i * 2] << 8 | buf[(i * 2) + 1]);
  

• pad_id = pressure >> 12;
  

• input_report_abs(cdev->input_dev, MASCHINE_PAD(pad_id), pressure & 0xfff);
  

• pad_id = le16_to_cpu(*pressure) >> 12;
  

• input_report_abs(cdev->input_dev, MASCHINE_PAD(pad_id), le16_to_cpu(*pressure) & 0xfff);
  

• pressure++;
  

  }

  input_sync(cdev->input_dev);