On Aug 30 2016 14:32, Takashi Iwai wrote:
On Tue, 30 Aug 2016 02:21:45 +0200, Takashi Sakamoto wrote:
I misunderstand TLV packet structure (struct snd_ctl_tlv) nests in its payload. But actually, the payload stores different type of data.
This commit corrects a comment in a test program of user control element set including my misunderstanding.
Signed-off-by: Takashi Sakamoto o-takashi@sakamocchi.jp
test/user-ctl-element-set.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/test/user-ctl-element-set.c b/test/user-ctl-element-set.c index 9b9dc59..f6f050a 100644 --- a/test/user-ctl-element-set.c +++ b/test/user-ctl-element-set.c @@ -418,11 +418,11 @@ static int check_tlv(struct elem_set_trial *trial) int err;
/*
* See a layout of 'struct snd_ctl_tlv'. I don't know the reason to* construct this buffer with the same layout. It should be abstracted* inner userspace library...
* When transferring threshold level information via TLV feature of* ALSA control interface, the first two fields of packet payload* consist of data type and data length.
- orig[0] = snd_ctl_elem_id_get_numid(trial->id);
- orig[0] = SNDRV_CTL_TLVT_CONTAINER; orig[1] = 6 * sizeof(orig[0]); orig[2] = 'a'; orig[3] = 'b';
The container TLV type expects other TLVs as its content. So the above looks buggy to me...
This is a test program just to serve to myself. So I hope this degree of roughness is allowed...
Or could you please construct valid TLV array here? I don't exactly know the way to do it, because I've never touched better documentation about whole design of TLV payload data.
Regards
Takashi Sakamoto