On Saturday 07 January 2006 17:18, Walter Haidinger wrote:
> On Sat, 7 Jan 2006, Hans Verkuil wrote:
> > You are using the 0.4.0 ivtvctl with a 0.4.1 or later ivtv driver.
> > Make sure you are using matching ivtvctl and ivtv driver! (ivtvctl
> > --version reports the ivtv version for which it is compiled).
>
> Right! However:
>
> # /opt/ivtv/sbin/ivtvctl --version
> ivtvctl version 0.4.2 (development revision 3101)
> # /opt/ivtv/sbin/ivtvctl -I
> check SAA7115 input signal
> ioctl: Input detect failed (cannot read SAA regs)
> ioctl: Input detect failed (cannot read SAA regs)
> ioctl: VIDEO_STATUS = Bad
>
> SAA7115 reg[STATUS#1 - 0x1E] = 177ff8e
> Color Standard (D0-D1): PAL
> White peak loop (D2): Activated
> Gain value for act.lum. (min) (D3): Limited
> Gain value for act.lum. (max) (D4): Not limited
> WIPA-mode slow-time constant (D5): Not activated
> Horizontal frequency (D6): Locked
> Field length (D7): Standard
>
> SAA7115 reg[STATUS#2 - 0x1F] = 177ff8e
> Ready for capture (D0): No
> Macrovision 7.01 (D1): Yes
> MV Colorstripe burst (D2): Yes
> MV Colorstripe burst type 3 (D3): Yes
> Input signal timebase (D4): Non-stable (VCR)
> Field frequency (D5): 50Hz
> H & V loops (D6): Locked
> Interlace (D7): Detected
>
> The reg status look ok now, why is VIDEO_STATUS still bad?
Because the handling of errors in --input-detect is bad. The problem
here is why it cannot read the SAA7115 registers (anything after that
is bogus, I've just fixed that in 0.4.2). Are you running the saa7115
from the 2.6.15 kernel? Or from 0.4.2? In the first case the only way
to get it to work is by adding '#define CONFIG_VIDEO_ADV_DEBUG' to the
top of the drivers/media/video/saa7115.c file. Then recompile and
install the kernel. My patch adding this config option was never added
to the 2.6.15 kernel unfortunately (it sort of got lost and wasn't
found until it was too late).
But much more reliable is to use the --log-status option of ivtvctl and
look in the kernel log (or upgrade to the very latest ivtv-0.4.2 to see
the output directly on the standard output).
> Please note that I use PREFIX=/opt/ivtv and BINDIR=$(PREFIX)/sbin
> compiling the utils. To change channels, set quality etc., I use my
> own scripts located in /opt/ivtv/bin. All scripts have the line
> export PATH=/opt/ivtv/sbin:/opt/ivtv/bin:${PATH}
> to use the ivtv utils installed in /opt/ivtv/sbin.
>
> I usually never run ivtvctl or any other utility manually,
> so I forgot to specify the path and used the SuSE rpm
> versions (ivtv-0.3.7c-3) from /usr/bin. Sorry about that!
> To eliminate any further path conflicts, I've just removed the
> rpm. I don't need it anyways.
>
> Now, despite the path problems, the pvr350 is still broken.
> Is there anything I can do/run/test for further diagnosis?
1) upgrade to the very latest svn (I made some commits a few minutes ago
that might come in handy).
2) which modules are you using? The saa7115, tuner, tveeprom, etc.
modules from the 2.6.15 kernel or the modules from ivtv-0.4 svn?
3) please post the full IVTV INIT sequence from the kernel log.
4) tune to what should be a correct channel and run ivtvctl --log-status
and post the output from that too.
Hans
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
