Got it!
Line 1360 in dst_get_tuna checks for !(state->dst_type ==
DST_TYPE_IS_CABLE); if I drop that condition the tuning problem seems fixed.
Zoilo Gomez wrote:
Manu Abraham wrote:
Zoilo Gomez wrote:
Hi Manu,
Manu Abraham wrote:
Zoilo Gomez wrote:
I decided to do some tests using kernel 2.6.13.3, with following
results:
1) locking problems occur using "dvbscan -v" (nut also ("vlc -v")
in the
following situations:
=> 2.6.13.3 kernel with downloaded snapshot 20051202: locks fine,
dvb-scan returns 0x1f and dumps PMT quickly
=> 2.6.13.3 kernel with downloaded snapshot 20060115: locks fine,
dvb-scan returns 0x1f and dumps PMT quickly
=> 2.6.13.3 kernel with downloaded snapshot 20060715: no lock,
dvb-scan
returns 0x00 + timeout!
=> 2.6.13.3 kernel with downloaded snapshot 20060717: no lock,
dvb-scan
returns 0x00 + timeout!
=> 2.6.13.3 kernel with downloaded hg snapshot: no lock, dvb-scan
returns 0x00 + timeout!
So it seems that something has gone wrong between snapshots from
January
and July, with these locking problems as a result.
The strange thing is that the locking problems do not prevent xine or
vlc from coming up with streaming data (picture), however in between
"vlc -v" prints 'no lock' every 15 seconds or so, but keeps going
anyway
....
Of course I would prefer to use the latest dvb-version, but I
cannot get
it to work for reasons mentioned above.
What could possibly be wrong?
Can someone tell me who has a more recent working combination (recent
kernel + recent dvb-snapshot)?
2) in all dvb-snapshot versions that have I tested, I always need to
remove the DST_TYPE_HAS_TS204 flag from the "DCT-CI" record in
dst_list
(dst.c), or I will not get a picture; only a long list of TS
discontinuities. Aren't the necessary patches (Sept 2005) present in
recent snapshots? I don't understand why I keep running into this
problem ....
In the latest v4l-dvb HEAD if you have the TS204 flag, it fixes your
problems ?
To check once more, I have just setup once again:
=> linux-2.6.17.8
=> hg v4l-dvb (fresh clone of today)
At a fresh 'make' on v4l-dvb, I do get several warnings about missing
kernel configuration flags, but it seems that I can ignore those, as
none of those seems to affect my configuration.
The output of dvbscan -v is:
... and no PMTs get reported.
Also vlc complains that 'dvb access warning: no lock, tuning again'
every 20 seconds or so.
Unfortunately I cannot test 20060115's snapshot against linux-2.6.17.8
because the MODULE_PARM macro is no longer supported in 2.6.17 and I
get
a compile error.
But if I compile with linux-2.6.13.3:
=> vlc-dvb-20060115 works (provided I take out TS204 from DCT-CI
record)
=> vlc-dvb hg clone does not work.
So something seems to be broken...
Is this a card with a CI or just a FTA card ?
Is it the same card ..? http://thadathil.net:8000/gallery/album12
loading dst with verbose=4 and then tuning what messages do you get ?
Oh sorry, you meant during tuning of course:
On 2.6.17.8 + v4l-dvb (catch of the day):
dst(0) dst_set_freq: set Frequency 306750000
dst(0) dst_set_frontend: Set Frequency=[306750000]
dst(0) dst_set_symbolrate: set symrate 6900000
dst(0) dst_set_symbolrate: DCT-CI
dst(0) dst_write_tuna: type_flags 0x1219
dst(0) dst_comm_init: Initializing DST.
dst(0) dst_gpio_outb: mask=[ffffffff], enbb=[0001], outhigh=[0000]
dst(0) rdc_reset_state: Resetting state machine
dst(0) dst_gpio_outb: mask=[0002], enbb=[0002], outhigh=[0000]
dst(0) dst_gpio_outb: mask=[0002], enbb=[0002], outhigh=[0002]
writing [ 09 00 04 ae 3e 00 1a f4 40 b9 ]
dst(0) dst_gpio_outb: mask=[ffffffff], enbb=[0000], outhigh=[0000]
dst(0) read_dst: reply is 0xff
dst(0) dst_wait_dst_ready: dst wait ready after 42
dst(0) read_dst: reply is 0x9
0x0 0x4 0xae 0x3e 0x0 0x1a 0xf4 0x40 0xb9
dst(0) dst_get_tuna: checksum failure?
And on 2.6.13.3 + vlc-dvb=20060115:
dst_set_freq: set Frequency 306750000
dst_set_frontend: Set Frequency=[306750000]
dst_set_symbolrate: set symrate 6900000
dst_write_tuna: type_flags 0x19
dst_comm_init: Initializing DST.
dst_gpio_outb: mask=[ffffffff], enbb=[0001], outhigh=[0000]
rdc_reset_state: Resetting state machine
dst_gpio_outb: mask=[0002], enbb=[0002], outhigh=[0000]
dst_gpio_outb: mask=[0002], enbb=[0002], outhigh=[0002]
writing [ 09 00 04 ae 3e 00 1a f4 40 b9 ]
dst_gpio_outb: mask=[ffffffff], enbb=[0000], outhigh=[0000]
read_dst: reply is 0xff
dst_wait_dst_ready: dst wait ready after 2
read_dst: reply is 0x9
0x0 0x4 0xae 0x3e 0x0 0x1a 0xf4 0x40 0xb9
Mmmmh, that checksum error in todays hg looks suspicious ....
Z.
_______________________________________________
linux-dvb mailing list
linux-dvb@xxxxxxxxxxx
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
_______________________________________________
linux-dvb mailing list
linux-dvb@xxxxxxxxxxx
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
