Beiträge von Yuri666

    if the CAM needs additional or special pids then this will never happen because VDR does not care about CI+.

    I don't use vdr now, but other software transfer uri successfully and fast in dvb card with embedded CI and very slow and partially in ddci hardware. So i think needs to care about additional pids in this case. (I don't know about what of pids...)


    BR,

    Yuri

    Hi Helmut,

    current implementation ci+ in vdr ignore uri value. But for compliance ci+ specs uri should be pass from CAM to host to every channel swithing. And host should to answer. Every new uri pass to CAM from air by emm or ecm (not speceifed in ci+ specs). Maybe in your case you didn't pass to CAM necessary pids to transfer uri to CAM and so on to host.

    BR,

    Yuri

    Hi Helmut,

    now my CAM doesn't freeze. Great! Thank you!

    But now i have CRC erors in stream and breaking picture and sound.

    I don't know it's related driver or software, I'll do more tests with other my CAMs.

    Code
    [28/04 20:34:00.996 AD4]: PID Continuity error (adapter 4): pid: 3297, Expected CC: 2, Actual CC: 3, CC Before 1
    [28/04 20:34:00.996 AD4]: PID Continuity error (adapter 4): pid: 3297, Expected CC: E, Actual CC: F, CC Before D
    [28/04 20:34:00.996 AD4]: PID Continuity error (adapter 4): pid: 3297, Expected CC: B, Actual CC: C, CC Before A
    [28/04 20:34:00.996 AD4]: PID Continuity error (adapter 4): pid: 3297, Expected CC: C, Actual CC: D, CC Before B

    BR,

    Yuri

    Hello Helmut,

    I have troubles with new CAM. In authentication process from CAM receive big tpdu (more 2kbyte). With old CAM it works fine.

    Code
    [36462.589347] [CAM ] rb_read_tpdu (0)(1)[2683] *** (7) 9F9003 ***

    Full log:

    old CAM


    But with new CAM i have error

    new CAM

    New CAM with ordinary dvb ci and kernel drivers works fine too.

    Can you look please, where might be bug?


    BR,

    Yuri

    Hi Helmut,

    sometimes device stalled with such errors in dmesg:


    Is possible recovery from this situation except disconnect and reload firmware? And where may be source this errors?


    BR,

    Yuri

    However, with modified drivers i have this strace output: (after about one minute):

    Not so much...


    BR,

    Yuri

    Hi Helmut,


    great job! :) As I understand, minisatip use kernel functions for read/write, and after your changes number submissions reduse several times.

    Code
    [136368.283358]  +++ TS-BITRATE : 18.78 Mbit/s, 48.78 CAM-subs/s, URB-load:  6.66%, TS-miss: 0
    [136378.286315]  +++ TS-BITRATE : 18.78 Mbit/s, 48.80 CAM-subs/s, URB-load:  6.66%, TS-miss: 0
    [136388.282923]  +++ TS-BITRATE : 18.82 Mbit/s, 48.90 CAM-subs/s, URB-load:  6.66%, TS-miss: 0
    [136398.295045]  +++ TS-BITRATE : 18.76 Mbit/s, 48.74 CAM-subs/s, URB-load:  6.66%, TS-miss: 0
    [136408.294683]  +++ TS-BITRATE : 18.78 Mbit/s, 48.80 CAM-subs/s, URB-load:  6.66%, TS-miss: 0

    All works fine, minisatip logs clear.

    Code
    [12/02 13:34:13.238 signal]: BW 2324KB/s, Total BW: 1115 MB, ns/read 3539, r: 66, w: 335 fw: 0, tt: 233 ms
    [12/02 13:34:14.258 signal]: BW 2391KB/s, Total BW: 1117 MB, ns/read 3733, r: 68, w: 344 fw: 0, tt: 253 ms
    [12/02 13:34:15.283 signal]: BW 2390KB/s, Total BW: 1119 MB, ns/read 3715, r: 68, w: 343 fw: 0, tt: 252 ms
    [12/02 13:34:16.302 signal]: BW 2393KB/s, Total BW: 1122 MB, ns/read 3843, r: 68, w: 344 fw: 0, tt: 261 ms
    [12/02 13:34:17.320 signal]: BW 2394KB/s, Total BW: 1124 MB, ns/read 3708, r: 68, w: 344 fw: 0, tt: 252 ms


    BR,

    Yuri

    Hi,

    no difference

    Code
    [11/02 23:26:23.085 signal]: BW 2042KB/s, Total BW: 79 MB, ns/read 5300, r: 199, w: 296 fw: 0, tt: 1054 ms
    [11/02 23:26:24.104 signal]: BW 2042KB/s, Total BW: 81 MB, ns/read 5303, r: 199, w: 295 fw: 0, tt: 1055 ms
    [11/02 23:26:24.237 AD2]: select_and_execute[7]: EOVERFLOW on socket 9 (sid:2) from 0.0.0.0:0 - type dvr errno 75
    [11/02 23:26:25.122 signal]: BW 1971KB/s, Total BW: 83 MB, ns/read 5145, r: 194, w: 285 fw: 0, tt: 998 ms
    [11/02 23:26:26.150 signal]: BW 2041KB/s, Total BW: 85 MB, ns/read 5276, r: 200, w: 293 fw: 0, tt: 1055 ms

    BR,

    Yuri

    Hi,

    yes, my speedtest showed equal results.

    But in minisatip if input stream more than 20 Mbps, i have distortions and errors in log:

    Code
    [10/02 00:19:54.876 signal]: BW 2056KB/s, Total BW: 36 MB, ns/read 15847, r: 67, w: 1600 fw: 0, tt: 1061 ms
    [10/02 00:19:55.842 AD2]: select_and_execute[7]: EOVERFLOW on socket 9 (sid:2) from 0.0.0.0:0 - type dvr errno 75

    EOVERFLOW - means that the data from DVR was not read fast enough. In case minisatip data from AD2 put to WINTV adapter using the function writev()

    WINTV driver report not much load:

    Code
    [80333.877045]  +++ TS-BITRATE : 15.79 Mbit/s, 328.20 CAM-subs/s, URB-load:  0.83%, TS-miss: 12
    [80343.876721]  +++ TS-BITRATE : 15.74 Mbit/s, 327.10 CAM-subs/s, URB-load:  0.83%, TS-miss: 12
    [80353.874308]  +++ TS-BITRATE : 15.71 Mbit/s, 326.60 CAM-subs/s, URB-load:  0.83%, TS-miss: 12
    [80363.875110]  +++ TS-BITRATE : 15.81 Mbit/s, 328.50 CAM-subs/s, URB-load:  0.83%, TS-miss: 12
    [80373.874625]  +++ TS-BITRATE : 15.73 Mbit/s, 326.90 CAM-subs/s, URB-load:  0.83%, TS-miss: 12

    Maybe needs to use other write/read method or use low latency kernel - I don't know.

    With real DDCI minisatip have not any problem. Maybe it should be - pcie 1x about four times faster than usb 2.0

    Code
     [10/02 12:22:40.977 signal]: BW 2178KB/s, Total BW: 89 MB, ns/read 1568, r: 31, w: 31 fw: 0, tt: 48 ms

    Can you test much stream load (more than 20 Mbps) to one WINTV adapter at other applications, e.g. VDR?


    BR,

    Yuri