vdr

vdr@linuxtv.org

2 participants
6731 discussions

[PATCH] prevent vdr from using system caches for video data
by Artur Skawina 16 Nov '05

16 Nov '05

well, vdr w/ the recent cUnbufferedFile changes was flushing the data buffers in huge burst; this was even worse than slowly filling up the caches -- the large (IIRC ~10M) bursts caused latency problems (apps visibly freezing etc). This patch makes vdr use a much more aggressive disk access strategy. Writes are flushed out almost immediately and the IO is more evenly distributed. While recording and/or replaying the caches do not grow and when vdr is done accessing a video file all cached data from that file is dropped. I've tested this w/ both local disks and NFS mounted ones, and it seems to do the right thing. Writes get flushed every 1..2s at a rate of .5..1M/s instead of the >10M bursts. For async mounted NFS servers the writes get collected by the NFS server and normally written out. Local disks get an extra feature -- you can use the HD activity LED as a "recording" indicator :^) As posix_advice requires kernel v2.5.60 and glibc v2.2, you'll need at least those versions to see any difference. (w/o posix_advice you will not get the fdatasyncs every 10M - if somebody really wants them they should be controlled by a config option) Possible further improvements could be: switch from POSIX_FADV_SEQUENTIAL to POSIX_FADV_RANDOM, since we're doing manual readahead anyway (or just leave this as is, and drop the readahead) (not using POSIX_FADV_RANDOM is probably one of the causes of the "leaks", so some of the workarounds could then go too) artur --- vdr-1.3.36.org/tools.h 2005-11-05 11:54:39.000000000 +0100 +++ vdr-1.3.36/tools.h 2005-11-16 15:51:27.000000000 +0100 @@ -209,6 +209,9 @@ private: off_t end; off_t ahead; ssize_t written; + ssize_t totwritten; + size_t readahead; + size_t pendingreadahead; public: cUnbufferedFile(void); ~cUnbufferedFile(); --- vdr-1.3.36.org/tools.c 2005-11-04 17:33:18.000000000 +0100 +++ vdr-1.3.36/tools.c 2005-11-16 16:06:26.000000000 +0100 @@ -851,8 +851,7 @@ bool cSafeFile::Close(void) // --- cUnbufferedFile ------------------------------------------------------- -#define READ_AHEAD MEGABYTE(2) -#define WRITE_BUFFER MEGABYTE(10) +#define WRITE_BUFFER KILOBYTE(800) cUnbufferedFile::cUnbufferedFile(void) { @@ -869,7 +868,13 @@ int cUnbufferedFile::Open(const char *Fi Close(); fd = open(FileName, Flags, Mode); begin = end = ahead = -1; + readahead = 16*1024; + pendingreadahead = 0; written = 0; + totwritten = 0; + if (fd >= 0) { + posix_fadvise(fd, 0, 0, POSIX_FADV_SEQUENTIAL); + } return fd; } @@ -880,10 +885,10 @@ int cUnbufferedFile::Close(void) end = ahead; if (begin >= 0 && end > begin) { //dsyslog("close buffer: %d (flush: %d bytes, %ld-%ld)", fd, written, begin, end); - if (written) + if (0 && written) fdatasync(fd); - posix_fadvise(fd, begin, end - begin, POSIX_FADV_DONTNEED); } + posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED); begin = end = ahead = -1; written = 0; } @@ -899,35 +904,92 @@ off_t cUnbufferedFile::Seek(off_t Offset return -1; } +// when replaying and going eg FF->PLAY the position jumps back 2..8M +// hence we might not want to drop that data at once. +// Ignoring for now to avoid making this even more complex, but we could +// at least try to handle the common cases +// (PLAY->FF->PLAY, small jumps, moving editing marks etc) + ssize_t cUnbufferedFile::Read(void *Data, size_t Size) { if (fd >= 0) { off_t pos = lseek(fd, 0, SEEK_CUR); - // jump forward - adjust end position - if (pos > end) - end = pos; - // after adjusting end - don't clear more than previously requested - if (end > ahead) - end = ahead; - // jump backward - drop read ahead of previous run - if (pos < begin) - end = ahead; + off_t jumped = end - pos; // nonzero means we're not at the last offset - some kind of jump happened. + if (jumped) { + pendingreadahead += ahead-end+KILOBYTE(64); + // jumped forward? - treat as if we did read all the way to current pos. + if (pos > end) { + end = pos; + // but clamp at ahead so we don't clear more than previously requested. + // (would be mostly harmless anyway, unless we got more than one reader of this file) + // add a little extra readahead, JIC the kernel prefethed more than we requested. + if (end > (ahead+KILOBYTE(512))) + end = ahead+KILOBYTE(512); + } + // jumped backward? - drop both last read _and_ read-ahead + if (pos < begin) + end = ahead+KILOBYTE(512); + // jumped backward, but still inside prev read window? - pretend we read less. + if ((pos >= begin) && (pos < end) ) + end = pos; + } + + ssize_t bytesRead = safe_read(fd, Data, Size); + + // now drop all data accesed during _previous_ Read(). if (begin >= 0 && end > begin) - posix_fadvise(fd, begin - KILOBYTE(200), end - begin + KILOBYTE(200), POSIX_FADV_DONTNEED);//XXX macros/parameters??? + posix_fadvise(fd, begin, end-begin, POSIX_FADV_DONTNEED); + begin = pos; - ssize_t bytesRead = safe_read(fd, Data, Size); if (bytesRead > 0) { pos += bytesRead; - end = pos; // this seems to trigger a non blocking read - this // may or may not have been finished when we will be called next time. // If it is not finished we can't release the not yet filled buffers. // So this is commented out till we find a better solution. - //posix_fadvise(fd, pos, READ_AHEAD, POSIX_FADV_WILLNEED); - ahead = pos + READ_AHEAD; + + // Hmm, it's obviously harmless if we're actually going to read the data + // -- the whole point of read-ahead is to start the IO early... + // The comment above applies only when we jump somewhere else _before_ the + // IO started here finishes. How common would that be? Could be handled eg + // by posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED) called some time after + // we detect a jump. Ignoring this for now. /AS + + // Ugh, it seems to cause some "leaks" at every jump... Either the + // brute force approach mentioned above should work (it's not like this is + // much different than O_DIRECT) or keeping notes about the ahead reads and + // flushing them after some time. the latter seems overkill though, trying + // the former... + + if ( !jumped ) { + if ( readahead <= Size ) // automagically tune readahead size. + readahead = Size*2; + posix_fadvise(fd, pos, readahead, POSIX_FADV_WILLNEED); + ahead = pos + readahead; + } + else { + // flush it all; mostly to get rid of nonflushed readahead + // coming from _previous_ jumps. possibly ratelimit this in + // some way (bytes prefetched? nr of jumps? time?) + + //// first-jump-after-every-60s seemed as good as any :^) + //time_t t = time(NULL); + //if (t >= nextflush) { + // posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED); + // nextflush = t+60; + // } + + // the accounting is _very_ unaccurate, i've seen ~50M get flushed + // when the limit was set to 4M. As long as this triggers after + // _some_ jumps we should be ok though. + if (pendingreadahead > MEGABYTE(1)) { + posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED); + pendingreadahead = 0; + } + ahead = pos; + } } - else - end = pos; + end = pos; return bytesRead; } return -1; @@ -950,12 +1012,18 @@ ssize_t cUnbufferedFile::Write(const voi end = pos + bytesWritten; if (written > WRITE_BUFFER) { //dsyslog("flush buffer: %d (%d bytes, %ld-%ld)", fd, written, begin, end); - fdatasync(fd); - if (begin >= 0 && end > begin) - posix_fadvise(fd, begin, end - begin, POSIX_FADV_DONTNEED); - begin = end = -1; + if (begin >= 0 && end > begin) { + off_t headdrop = max((long)begin,(long)WRITE_BUFFER*2); + posix_fadvise(fd, (begin&~4095)-headdrop, end - begin + headdrop, POSIX_FADV_DONTNEED); + } + begin = end = -1; + totwritten += written; written = 0; } + if (totwritten > MEGABYTE(10)) { + posix_fadvise(fd, 0, 0, POSIX_FADV_DONTNEED); + totwritten = 0; + } } return bytesWritten; }

1 0

Re: problems with radio recordings
by Andreas Regel 16 Nov '05

16 Nov '05

Andreas Regel schrieb: > 2. When replaying a radio recording and doing a fast forward, vdr exits > with a segfault. The progress bar always shows a value between 10 and 11 > minutes. Here is the backtrace for this error: > > [..] > > I'm using vanilla vdr 1.3.19 with a single DVB card. > > Andreas This is now fixed in 1.3.24. I guess it is related to the fix in cPesAssembler. Andreas

2 1

Simple menu for DVD ?
by Gregoire Favre 16 Nov '05

16 Nov '05

Hello, I would like to have a script that take my vdrsync.pl mpeg created files and directly create a DVD with ONE menu, without animation and without sound, only text to choose which movie to play DIRECTLY, without going to another text menu or so. Something like : create_dvd my_firt_movie.mpg my_second_movie.mpg my_third_movie.mpg which will create a DVD with ONE text menu with choices : my firt movie my second movie my third movie And possibly with as much chapters as possible on each movies (99) and with language for first audio being set to de, second audio to en (if exist). I am aware of "videotrans" at http://sourceforge.net/projects/videotrans/ which I think make too many things (I don't want animated menu or so). Thank you very much, -- Gr/'egoire FAVRE

1 0

Softdevice plugin problem
by Norm Dressler 16 Nov '05

16 Nov '05

Hi there, I'm setting up 1.3.36 with softdevice 0.2.1. It seems to compile fine but when I try to run it I get: vdr: ./PLUGINS/lib/libvdr-softdevice.so.1.3.36: undefined symbol: dc1394_create_handle Any suggestions where to look or what could be wrong? Norm

4 6

[Feature request] directing keymacros to dedicated servicename (instead of plugin)
by Stefan Lucke 15 Nov '05

15 Nov '05

Hi, current definiton of file keymacros.conf allows simple keys to be sent to plugins only. Example User1 @plugin_name Down Down Ok As with vdr's new Service() interface a more flexible solution would allow sending arbitrary actions to plugins independent of plugin's menu layout. This could be done by either reserving a special name or a special prefix character. User1 @plugins_service service_name data_for_service or User1 §service_name data_for_service -- Stefan Lucke

2 2

[patch] load i18n versions of commands/reccmds.conf
by Lucian Muresan 15 Nov '05

15 Nov '05

Hi, I'd like to propose a patch for further i18n of VDR, namely in the commands.conf and reccmds.conf files. Even if any user can customize these files to his whishes (and language), I think that it is more consistent from the VDR user interface point of view if the commands defined there also change their language on the OSD, according to the OSD language setting. All that the user has to do, is to provide, additionally to the "normal" files commands.conf and reccmds.conf, the version(s) in the language(s) he want's to use as commands.conf.xyz and reccmds.conf.xyz, where ".xyz" is the 3-characters language code defined in VDR's i18n.c: // The 3-letter names of the language... { "eng,dos", "deu,ger", "slv,slo", "ita", "dut,nla,nld", "por", "fra,fre", "nor", "fin,smi", "pol", "esl,spa", "ell,gre", "sve,swe", "rom,rum", "hun", "cat,cln", "rus", "hrv", "est", "dan", }, Important notes: - There is *no problem* if VDR does not find such translated command files at startup or at OSD language switching on the fly, it will always fall back to the default and if those are missing, as you already know, it's not a big deal either, so the behaviour doesn't change (except that there is a syslog message if such a customized file hasn't been found, just to inform users who are reading the logs, that *they could have* multiple command file versions according to configured OSD language, for english setups it#s absolutely silent as before); - this will only work with the *first* name for languages where multiple 3-letter names are defined; - As english is all over in VDR's source code specially treated, this makes no exception here, so it's indeed treated specially: english files need *no* such 3-letter extension, and are *only* loaded in their default form. Have phun, Lucian diff -Naur vdr-1.3.36_orig/config.c vdr-1.3.36/config.c --- vdr-1.3.36_orig/config.c 2005-09-09 17:08:59.000000000 +0200 +++ vdr-1.3.36/config.c 2005-11-14 13:19:45.010000000 +0100 @@ -16,6 +16,8 @@ #include "plugin.h" #include "recording.h" +const char *ConfigDirectory = NULL; + // IMPORTANT NOTE: in the 'sscanf()' calls there is a blank after the '%d' // format characters in order to allow any number of blanks after a numeric // value! @@ -535,3 +537,31 @@ } return false; } + +bool LoadCommandsI18n(cCommands & cmds, const char *FileName, bool AllowComments, bool MustExist) +{ + bool bRet = true; + bool bLoadDefault = (Setup.OSDLanguage == 0); + if (!bLoadDefault) { + // attempt to load a translated file + char LangCode[4]; + strncpy(LangCode, I18nLanguageCode(Setup.OSDLanguage), 3); + LangCode[3] = '\0'; + char *FullPath = NULL; + asprintf(&FullPath, "%s/%s.%s", ConfigDirectory && *ConfigDirectory ? ConfigDirectory : ".", FileName, LangCode); + + if (!cmds.Load((FullPath), AllowComments, true)) { // require to exist, just to be able to log + // fallback + bLoadDefault = true; + esyslog("Failed to load translated '%s.%s' for language %d (%s)", FileName, LangCode, Setup.OSDLanguage, I18nLanguages()[Setup.OSDLanguage]); + esyslog("Falling back to default '%s' (if any)", FileName); + } + free(FullPath); + } + if (bLoadDefault) { + // let's do it the normal way + bRet = cmds.Load(AddDirectory(ConfigDirectory, FileName), AllowComments, MustExist); + } + // return status only for the default commands file + return bRet; +} diff -Naur vdr-1.3.36_orig/config.h vdr-1.3.36/config.h --- vdr-1.3.36_orig/config.h 2005-11-04 16:55:05.000000000 +0100 +++ vdr-1.3.36/config.h 2005-11-14 13:16:13.170000000 +0100 @@ -34,6 +34,8 @@ #define MaxSkinName 16 #define MaxThemeName 16 +extern const char *ConfigDirectory; + class cCommand : public cListObject { private: char *title; @@ -265,4 +267,6 @@ extern cSetup Setup; +bool LoadCommandsI18n(cCommands & cmds, const char *FileName = NULL, bool AllowComments = false, bool MustExist = false); + #endif //__CONFIG_H diff -Naur vdr-1.3.36_orig/menu.c vdr-1.3.36/menu.c --- vdr-1.3.36_orig/menu.c 2005-11-05 18:29:22.000000000 +0100 +++ vdr-1.3.36/menu.c 2005-11-14 13:21:47.450000000 +0100 @@ -1910,6 +1910,10 @@ Setup.OSDLanguage = data.OSDLanguage; cFont::SetCode(I18nCharSets()[Setup.OSDLanguage]); + // try to load translated command files if available, otherwise fallback to defaults + LoadCommandsI18n(Commands, "commands.conf", true); + LoadCommandsI18n(RecordingCommands, "reccmds.conf", true); + cSkin *Skin = Skins.Get(skinIndex); if (Skin) { char *d = themes.NumThemes() ? strdup(themes.Descriptions()[themeIndex]) : NULL; diff -Naur vdr-1.3.36_orig/vdr.c vdr-1.3.36/vdr.c --- vdr-1.3.36_orig/vdr.c 2005-11-04 14:48:39.000000000 +0100 +++ vdr-1.3.36/vdr.c 2005-11-14 13:21:32.700000000 +0100 @@ -109,7 +109,6 @@ int SVDRPport = DEFAULTSVDRPPORT; const char *AudioCommand = NULL; - const char *ConfigDirectory = NULL; const char *EpgDataFileName = DEFAULTEPGDATAFILENAME; bool DisplayHelp = false; bool DisplayVersion = false; @@ -420,8 +419,9 @@ Diseqcs.Load(AddDirectory(ConfigDirectory, "diseqc.conf"), true, Setup.DiSEqC) && Channels.Load(AddDirectory(ConfigDirectory, "channels.conf"), false, true) && Timers.Load(AddDirectory(ConfigDirectory, "timers.conf")) && - Commands.Load(AddDirectory(ConfigDirectory, "commands.conf"), true) && - RecordingCommands.Load(AddDirectory(ConfigDirectory, "reccmds.conf"), true) && + // try to load translated command files if available, otherwise fallback to defaults + LoadCommandsI18n(Commands, "commands.conf", true) && + LoadCommandsI18n(RecordingCommands, "reccmds.conf", true) && SVDRPhosts.Load(AddDirectory(ConfigDirectory, "svdrphosts.conf"), true) && CaDefinitions.Load(AddDirectory(ConfigDirectory, "ca.conf"), true) && Keys.Load(AddDirectory(ConfigDirectory, "remote.conf")) &&

5 11

changing priority of current receiving device.
by syrius.ml＠no-log.org 15 Nov '05

15 Nov '05

Hi, Could someone give me a little help to write a new function to vdr ? I'd like to be able to change the priority of the receiving device from a plugin. >From the plugin i'm able to find the index of the receiving device with cDevice::ActualDevice()->DeviceNumber() But all my attemps to change the priority have resulted in segfaults. (after having moved some things from protected/private space to public) Any example would be very appreciated. :) TIA --

1 0

create a channel by SetId(int Nid, int Tid, int Sid, int Rid = 0)
by Patrick Fischer 15 Nov '05

15 Nov '05

Hello Today I'm wondering about the SetId from cChannels.h As I know, I can specify a channel only by NID, TID and SID I know vdr don't use the way, but is there an way to switch to a channel only by this values? I have tried this code: cChannel* channel = new cChannel(); channel->SetId(8468,12801,16406); cDevice::ActualDevice()->SwitchChannel(channel,true); If I execute this code, vdr will show me only the last still Pic of the old channel. The values are from my channels.conf created with w_scan Name <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#Name> Frequenz <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#Frequenz> Parameter <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#Parameter> Quelle <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#Quelle_-_Sou…> SRate <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#SRate_-_Symb…> VPID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#VPID> APID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#APID> TPID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#TPID> CA <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#CA_-_Conditi…> SID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#SID> NID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#NID> TID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#TID> RID <http://www.vdr-wiki.de/wiki/index.php/Syntax_der_channels.conf#RID> RTL2:578000:I0C23D12M16B8T8G4Y0:T:27500:353:354:359:0:16406:8468:12801:0 If I don't find a better way, I will parse my channels.conf for the SID,NID and TID If I found the right line, then I can create a cChannel with cChannel-Parse(the found line); I don't like the alternative way. Can anybody tell me a better way? greetings Patrick

1 2

Wrong NID's in channels.conf generated with dvbscan
by bluedrops 15 Nov '05

15 Nov '05

Hi to all, I know this isn't exactly vdr related but it seems nobody has time to look at this problem on linux-dvb mailing list. I'm not a programmer so I can do some testing at home but modifying scan.c is too much for me. Perhaps there are some DVB guru's on this list who could help me ? I use dvbscan to generate channels.conf file for vdr and the problem is I'm getting wrong NID's in this file. This is a BIG issue because when I do a transponder scan in VDR, it finds the same channels twice because of different (but right) NID's BUT sets VPID and APID to 0. So there are a lot of unusable duplicate channels. Here is what Johannes Stezenbach said : 2005/10/24, Johannes Stezenbach <js(a)linuxtv.org>: > Quite likely the SI data on Hotbird is inconsistent, e.g. the NITs > sent on different transponders contain differernt NIDs for the > same multiplex. However, I don't have time to look into this, and > I don't even have SAT receiption... > dvbscan should take the NID only from the NIT entry for the > transponder it is currently tuned to (check parse_nit()). I'm using dvbscan with more than 23 orbital positions and I not only have the problem with Hotbird so I'd really like it to be fixed. Here is an example : Hotbird 6 (13.0E) - 11662.00 V Multivision channels are NID 318 instead of 176 (see www.lyngsat.com or www.kingofsat.net) I've made dumps for pids 0 and 16 with "dvbsnoop -pd 7 PID". Each one is about 30 sec. 0.dump.bz2 (4k) : http://www6.rapidupload.com/d.php?file=dl&filepath=1500 16.dump.bz2 (202k) : http://www2.rapidupload.com/d.php?file=dl&filepath=882 --------------------------------------------------------- Transponder PID-Scan... --------------------------------------------------------- PID found: 0 (0x0000) [SECTION: Program Association Table (PAT)] PID found: 16 (0x0010) [SECTION: Network Information Table (NIT) - actual network] PID found: 17 (0x0011) [SECTION: Service Description Table (SDT) - other transport stream] PID found: 18 (0x0012) [SECTION: Event Information Table (EIT) - other transport stream, present/following] ... I've noticed in 16.dump for PID 16 : Network_ID: 176 (0x00b0) [= >>ERROR: not (yet) defined... Report!<<] Also, Network_ID reported by dvbsnoop for the example I've given seems to be right (176) in 16.dump. Thanks for helping, bluedrops

1 0

which tv-out card?
by Leo Márquez 15 Nov '05

15 Nov '05

Hi all, I want help a friend to set a budget cards based (dvb-t and dvb-s) vdr box in a standard computer (not micro-atx) I don't have the hardware info (only cpu athlon 1800) but the connector to be used is s-video to scart-tv. We have planned to set softdevice and use the tv-out of a vga card to connect it to the tv. Do you recommend any vga card or any kind work with softdevice?. I need directfb or only softdevice? I have no experience with softdevice. I actually have an epia system but with a ff nexus card. Thanks.

5 7

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

vdr