[linux-dvb] Re: DVB driver for B2C2 based budget cards (such as Technisat SkyStar 2)

2003-02-09 Thread Roberto Deza
Hi Miernik and all, At 01:30 07/02/2003 +0100, Miernik wrote: >Wow, what a detailed analisis! Wasn't too much :-). >> >I have a Skymedia 200D and it has two big chips: >>... IOW, some good pictures are need :-). >Here are images (300 dpi scan, about 250 kB each) of two cards I have: > >http://w

[linux-dvb] Re: Crashes in dvb_demux.c

2003-02-09 Thread Emard
On Sun, Feb 09, 2003 at 07:54:43PM +0100, Florian Schirmer wrote: > Hi, > > >What if sec->seclen is already negative (-1) ? It can be -1 < count and > >also -1 < 4096, but would lead to infinite memcopy > > sec->seclen is set as "3+((buf[1]&0x0f)<<8)+buf[2]". This can not result in > a negative

[linux-dvb] drivers into the kernel

2003-02-09 Thread Alessio Sangalli
Hi. How is the quality of drivers when integrated into the kernel? Reading documentation and files inside DVB/driver directory, I noted the makepatch script. It's not that smart but after a couple of tests, I got the new menu under multimedia devices. As I can't actually use dvb cards by now (t

[linux-dvb] Re: Bug in ring buffer implementation

2003-02-09 Thread Oliver Endriss
On Saturday 08 February 2003 04:10, Oliver Endriss wrote: > while reading through the driver sources, I found that the free byte > count of the ring buffers is not calculated correctly. The following patch (against dvb head 2003-02-08) fixes the following ring buffer issues in av7110.c: - ring_b

[linux-dvb] AW: Re: Crashes in dvb_demux.c

2003-02-09 Thread Florian Schirmer
Hi, >What if sec->seclen is already negative (-1) ? It can be -1 < count and >also -1 < 4096, but would lead to infinite memcopy sec->seclen is set as "3+((buf[1]&0x0f)<<8)+buf[2]". This can not result in a negative length. Bye, Florian -- Info: To unsubscribe send a mail to [EMAIL PROT

[linux-dvb] Re: Crashes in dvb_demux.c

2003-02-09 Thread Emard
On Sun, Feb 09, 2003 at 07:29:31PM +0100, Florian Schirmer wrote: > Hi, > > > if ((count>2) && // enough data to determine sec > length? > > ((sec->seclen = section_length(buf+p)) <= > count)) { > > if (sec->seclen>4096) > >

[linux-dvb] AW: Re: Crashes in dvb_demux.c

2003-02-09 Thread Florian Schirmer
Hi, > if ((count>2) && // enough data to determine sec length? > ((sec->seclen = section_length(buf+p)) <= count)) { > if (sec->seclen>4096) >buf+p contains 188 bytes or less (188-p), assuming p is positive. > >sec->se

[linux-dvb] Re: Crashes in dvb_demux.c

2003-02-09 Thread Emard
> Did you test printing out the value when it is >0x1fff? > Does it really happen? No, I haven't. That was my mistake and I fixed it but rest of the crashes remain. I don't know much about PES data structure but I know what unix class stability code looks like. > Your ksymoops dump in your last

[linux-dvb] Re: AW: AW: mplayer divx rendering

2003-02-09 Thread Alessio Sangalli
Stephan Bieker wrote: I searched three days to find the solution...:-) I arrived to that conclusion befause I searched on the internet and I realized that it was the driver of the frontend (fe!) so it had to do with the tuner. As my card had no cable attached, I understood it could have be

[linux-dvb] Re: Crashes in dvb_demux.c

2003-02-09 Thread Ralph Metzler
Emard writes: > [1 ] > > I don't know exactly why it crashes for you (your dump rather > > points to an out of bounds *buf I think) but ts_pid() can only return values > > <=0x1fff. There should be no check needed. > > Neither do I, but it might be gcc optimizer bug. > ts_pid returns u16

[linux-dvb] Re: AW: mplayer divx rendering

2003-02-09 Thread Alessio Sangalli
Stephan Bieker wrote: Check the sat-cable connection to all of your installed dvb cards. If one of them is broken, the kdvb-fe process is going crazy and uses much cpu-time (zig-zag). Thank you. I just solved this, before reading this email. They're working on the roof so they misplaced my

[linux-dvb] Cinch cable for DVB-S rev2.1

2003-02-09 Thread Richard Robson
Hello, I have recently obtained a second hand Hauppage WinTv DVB-S rev2.1 DVB card. Unfortunately it didn't come with the Cinch cable, does anyone know where I can get the cable from or know the pin outs for it? Cheers, Richard. -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "

[linux-dvb] Re: Crashes in dvb_demux.c

2003-02-09 Thread Emard
> I don't know exactly why it crashes for you (your dump rather > points to an out of bounds *buf I think) but ts_pid() can only return values > <=0x1fff. There should be no check needed. Neither do I, but it might be gcc optimizer bug. ts_pid returns u16 value, and gcc during optimization might

[linux-dvb] Crashes in dvb_demux.c

2003-02-09 Thread Ralph Metzler
Emard writes: > void dvb_dmx_swfilter_packet(struct dvb_demux *demux, const u8 *buf) > { > struct dvb_demux_feed *feed = demux->pid2feed[ts_pid(buf)]; > > if (!feed) > return; > > dvb_dmx_swfilter_packet_type (feed, buf); > } > > Simple, but cra

[linux-dvb] GPIO0 irq oops, DEBI irq oops

2003-02-09 Thread Emil Naepflein
Hi, since I have installed a new disk controller I sometimes see messages like the following: Feb 9 13:11:20 vdrpc kernel: GPIO0 irq oops Feb 9 13:11:20 vdrpc kernel: DEBI irq oops Sometimes even a arm crash follows. I have tried with different drivers and played around with interrupt assignme

[linux-dvb] AW: can't make insmod with the metzlers 1.1.5 and kernel 2.4.20 (nexus)

2003-02-09 Thread Kai Thost
i have to specify: a restart of my computer just left the old dsp-module-problem. seems that a simple 'make rmmod' of the older drivers isn't enough ... kai -Ursprüngliche Nachricht- Von: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]Im Auftrag von Kai Thost Gesendet: Sonntag, 9. Februar 20

[linux-dvb] can't make insmod with the metzlers 1.1.5 and kernel 2.4.20 (nexus)

2003-02-09 Thread Kai Thost
make goes fine. insmod not: dvb-saa7146.o: unresolved symbol dvb_unregister_dsp dvb-saa7146.o: unresolved symbol DmxDevInit dvb-saa7146.o: unresolved symbol init_ipack dvb-saa7146.o: unresolved symbol reset_ipack dvb-saa7146.o: unresolved symbol instant_repack dvb-saa7146.o: unresolved symbol send

[linux-dvb] mplayer divx rendering

2003-02-09 Thread Alessio Sangalli
Hi. I'm trying the mplayer support to see divx through the dvb card. I use an Athlon 2000+ with kt400 motherboard, and I tried two different DVB boards. 1GB ram, ultradma hard disk, etc. I've got problems: I start mplayer directly from command line, just to be sure it's not a vdr or mplayer p

[linux-dvb] Re: channel.conf

2003-02-09 Thread Oliver Friedrich
You can go to www.vdrportal.de There are two channels.conf files (for >= 1.1.16) in the download area. But i think, that the normal one is a little bit buggy, because some channels are not working, that work fine with the channel list of my vdr 1.0.3 Is there another resource for a working channel

[linux-dvb] AW: videolans dvbinput and libdvb by metzler bros won't fit

2003-02-09 Thread Marcus Metzler
Kai Thost writes: > hi marcus, > > here we go: > > Compiling dvbinput.cpp... > g++ -I/usr/src/libdvbpsi/include -Wall -D_REENTRANT -D_FILE_OFFSET_BITS=64 - > DCONFIG_PATH='"/etc/videolan/vls"' -DBUGGY_VLC -DYY_NO_UNPUT -DSYSLOG -I/usr > /src/DVB/include -I/usr/src/DVB/libdvb -D__PLUGIN__

[linux-dvb] Crashes in dvb_demux.c

2003-02-09 Thread Emard
HI Research on vpeirq() field revealed further bugs and code dirtyness which causes crashes. Source of this comes from software demux, the stuff in dvb_demux.c in dvb-kernel. I intentionally screwed dmapos by shifting it an couple of bytes ahead (I used +441, but anything not divisible by 188 wo

[linux-dvb] AW: videolans dvbinput and libdvb by metzler bros won't fit

2003-02-09 Thread Kai Thost
hi marcus, here we go: Compiling dvbinput.cpp... g++ -I/usr/src/libdvbpsi/include -Wall -D_REENTRANT -D_FILE_OFFSET_BITS=64 - DCONFIG_PATH='"/etc/videolan/vls"' -DBUGGY_VLC -DYY_NO_UNPUT -DSYSLOG -I/usr /src/DVB/include -I/usr/src/DVB/libdvb -D__PLUGIN__ -fPIC -o obj/dvbinput.o -c dvbinput.cpp d