[linux-dvb] Re: Nova card drops packets, but tuner reports 0BER & 0 Uncorr blocks

[Steve Brown <sbrown@cortland.com>]

Ralph Metzler wrote:

>Steve Brown writes:
> > For testing, I eliminated xine and my code and am using dvbtune to set 
> > the tuner and report tuner status and a hacked test_dvr to audit the pid 
> > cc's. I have an 8k array for the cc's so I can audit the whole feed.
> > 
>
>Could you please always read blocks of 512*188 bytes from PID 0x2000 
>and check with your program if these errors (in any of the PIDs) 
>occur at or near or at similar offsets of multiples of these 
>512*188 byte blocks?
>
>
>Ralph
>

Ralph,

Attached is an abbreviated snip from the test you recommended above. I 
will email you the entire gzip'd log.

The tuner claimed that the feed was error free during this capture.

After running this for a while and seeing continous errors, I killed it 
and restarted it. The errors disappeared.

In the dump, blk is the block (512 packets) count,
                         p is the packet offset,
                         r is length_read%packet_size
                         len is the length_read
                         pid is the pid in dec and hex,
                         cc_new is the cc in the current packet,
                         cc_exp(ected) is the cc +1 in the last packet 
for this pid
                         dump is the first 4 bytes of the packet with 
the cc mismatch

Looks like lots of errors around the beginning and end of the buffer. 
The error print, unfortunately, occurs at the packet that fails the 
sequence check, not the one(s) that got dropped. If you think it would 
help, I can modify the program to also show the blk/offset of the packet 
prior to the sequence break.

Steve.

BTW: Marcus, rtuxzap doesn't want to work with my nova card. I don't 
have the ir libraries. Symptom is: no response to keyboard commands. Is 
it known to work with the Nova and no ir libs? I followed it as far as 
verifying that the config file got read and the initial (default) 
setting of the tuner succeeded.

=======================================================================
                         
filter for pid:8192 (2000)
cc: blk:70 p:511 r:0 len96256 pid:512 (0x0200) cc_new:0 cc_exp:6 dump:47 
02 00 10
cc: blk:71 p:1 r:0 len96256 pid:512 (0x0200) cc_new:6 cc_exp:1 dump:47 
02 00 16
cc: blk:71 p:8 r:0 len96256 pid:1260 (0x04ec) cc_new:3 cc_exp:2 dump:47 
04 ec 93
cc: blk:103 p:6 r:0 len96256 pid:1760 (0x06e0) cc_new:3 cc_exp:2 dump:47 
06 e0 13
cc: blk:133 p:511 r:0 len96256 pid:1620 (0x0654) cc_new:8 cc_exp:d 
dump:47 46 54 18
cc: blk:134 p:9 r:0 len96256 pid:1160 (0x0488) cc_new:a cc_exp:9 dump:47 
04 88 da
cc: blk:134 p:109 r:0 len96256 pid:1620 (0x0654) cc_new:d cc_exp:9 
dump:47 06 54 3d
cc: blk:134 p:511 r:0 len96256 pid:1160 (0x0488) cc_new:5 cc_exp:d 
dump:47 04 88 d5
cc: blk:135 p:2 r:0 len96256 pid:1160 (0x0488) cc_new:d cc_exp:6 dump:47 
04 88 dd
cc: blk:135 p:9 r:0 len96256 pid:1460 (0x05b4) cc_new:c cc_exp:b dump:47 
05 b4 1c
cc: blk:165 p:511 r:0 len96256 pid:1160 (0x0488) cc_new:d cc_exp:4 
dump:47 04 88 dd
cc: blk:166 p:1 r:0 len96256 pid:1160 (0x0488) cc_new:4 cc_exp:e dump:47 
04 88 d4
cc: blk:166 p:7 r:0 len96256 pid:1260 (0x04ec) cc_new:c cc_exp:b dump:47 
04 ec dc
cc: blk:166 p:511 r:0 len96256 pid:512 (0x0200) cc_new:c cc_exp:7 
dump:47 02 00 1c
cc: blk:167 p:5 r:0 len96256 pid:512 (0x0200) cc_new:7 cc_exp:d dump:47 
02 00 17
cc: blk:167 p:9 r:0 len96256 pid:1660 (0x067c) cc_new:6 cc_exp:5 dump:47 
06 7c 16
cc: blk:197 p:511 r:0 len96256 pid:1660 (0x067c) cc_new:0 cc_exp:9 
dump:47 06 7c 10
cc: blk:198 p:5 r:0 len96256 pid:1660 (0x067c) cc_new:9 cc_exp:1 dump:47 
06 7c 19
cc: blk:198 p:7 r:0 len96256 pid:1760 (0x06e0) cc_new:5 cc_exp:4 dump:47 
06 e0 15
cc: blk:198 p:511 r:0 len96256 pid:1660 (0x067c) cc_new:c cc_exp:5 
dump:47 06 7c 1c






-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.