Commentaires sur : Darwin Streaming Server (DSS) 6.0.3, on Solaris 10

Par : Chris

Chris — Tue, 09 Dec 2008 21:53:04 +0000

Finally, I was working on 5.5.4 all day long and got to a part that just could figure out then stumbled upon this. I got alot further this time but here are my errors, it does ask to make a username and password after it completes but I still can\’t launch the app. Anyhelp would be great!!

Inserting path to perl into scripts..
Can\’t open perl script \"perlpath.pl\": No such file or directory

Creating unprivileged user to run the server = \"qtss\".
copying DarwinStreamingServer to /usr/local/sbin/DarwinStreamingServer
cp: cannot access DarwinStreamingServer
copying PlaylistBroadcaster to /usr/local/bin/PlaylistBroadcaster
cp: cannot access PlaylistBroadcaster
copying MP3Broadcaster to /usr/local/bin/MP3Broadcaster
cp: MP3Broadcaster: is a directory
copying qtpasswd to /usr/local/bin/qtpasswd
cp: cannot access qtpasswd
cp: cannot access StreamingServerModules/*
./Install: line 243: /usr/local/sbin/DarwinStreamingServer: No such file or directory
chmod: WARNING: can\’t access /etc/streaming/streamingserver.xml
chown: /etc/streaming/streamingserver.xml: No such file or directory

copying relayconfig.xml-Sample to /etc/streaming/relayconfig.xml-Sample

copying qtusers to /etc/streaming/qtusers

copying qtgroups to /etc/streaming/qtgroups
copying readme.txt to /var/streaming/readme.txt
cp: cannot access readme.txt
copying 3rdPartyAcknowledgements.rtf to /var/streaming/3rdPartyAcknowledgements.rtf
cp: cannot access 3rdPartyAcknowledgements.rtf
copying sample_100kbit.mov into /usr/local/movies/sample_100kbit.mov
copying sample_300kbit.mov into /usr/local/movies/sample_300kbit.mov
copying sample_100kbit.mp4 into /usr/local/movies/sample_100kbit.mp4
copying sample_300kbit.mp4 into /usr/local/movies/sample_300kbit.mp4
copying sample.mp3 into /usr/local/movies/sample.mp3
copying sample_50kbit.3gp into /usr/local/movies/sample_50kbit.3gp
copying sample_h264_100kbit.mp4 into /usr/local/movies/sample_h264_100kbit.mp4
copying sample_h264_300kbit.mp4 into /usr/local/movies/sample_h264_300kbit.mp4
copying sample_h264_1mbit.mp4 into /usr/local/movies/sample_h264_1mbit.mp4
copying StreamingLoadTool to /usr/local/bin/StreamingLoadTool
cp: StreamingLoadTool: is a directory
copying streamingloadtool.conf to /etc/streaming/streamingloadtool.conf
cp: cannot access streamingloadtool.conf
copying streamingadminserver.pl into /usr/local/sbin/streamingadminserver.pl
cp: cannot access streamingadminserver.pl
copying Admin HTML to /var/streaming/AdminHtml directory
cp: cannot access AdminHtml

Launching streamingadminserver.pl
./Install: line 355: /usr/local/sbin/streamingadminserver.pl: No such file or directory

Installation Complete

Par : Croft

Croft — Wed, 19 Nov 2008 11:46:19 +0000

Sure, this is the output from gcc -v:

-bash-3.2$ gcc -v
Reading specs from /usr/sfw/lib/gcc/i386-pc-solaris2.11/3.4.3/specs
Configured with: /builds2/sfwnv-gate/usr/src/cmd/gcc/gcc-3.4.3/configure –prefix=/usr/sfw –with-as=/usr/sfw/bin/gas –with-gnu-as –with-ld=/usr/ccs/bin/ld –without-gnu-ld –enable-languages=c,c++,f77,objc –enable-shared
Thread model: posix
gcc version 3.4.3 (csl-sol210-3_4-20050802)

Par : Prune

Prune — Wed, 19 Nov 2008 10:52:21 +0000

Stefan : of course !
Croft : which version of GCC are you exactly using ?
The {} this is special for Solaris… I don’t think there is another definition for OpenSolaris in DSS…
This may be a « new » gcc change if you’re bleeding edge… If yes, this means :
- keep the solaris patch as-is and have another opensolaris patch
- try other gcc (solaris) without the {} and change the patch if everything ok

Please give you gcc revision. Thanks.

Par : Croft

Croft — Wed, 19 Nov 2008 10:16:02 +0000

The only thing I did was a quick and dirty change in OSMutex.cpp – there is a section in line 49-52 where you have two versions of PTHREAD_ONCE_INIT – one is surrounded by curly braces and the other isn’t. I just experimented and did the bare minimum – removed the curly braces making both lines identical and did another « buildtarball 6.0.3″. Built nicely and works fine to stream from, haven’t done any actual hammering of it to see if its unbreakable though.

This was on OpenSolaris 2008.5 Developer Preview, but after running « pkg image-update » to snv_101a (the latest as of yesterday.) I had other issues earlier with ld, but I suspect that was just that I had some issues when I earlier upgraded to snv_99, should have no bearing on compiling DSS.

Par : Stefan Parvu

Stefan Parvu — Wed, 19 Nov 2008 08:51:44 +0000

Sebastien: we need to evaluate the change making sure we dont break something else.
Wonder why wouldn’t compile out of box on OpenSolaris bits.

Par : Prune

Prune — Wed, 19 Nov 2008 08:39:48 +0000

Croft, could you please copy/past the change you made with every detail so we can tune the patch ?
Thanks.

Par : Stefan Parvu

Stefan Parvu — Wed, 19 Nov 2008 07:04:07 +0000

Hi,
I will soon test as well on OpenSolaris bits. I have been able to compile DSS 6.0.3
using gcc on Solaris Express and Solaris 10, using latest patch set.

Par : Croft

Croft — Tue, 18 Nov 2008 17:50:48 +0000

PS. It works. Yay. Thanks again for the patch, made life a lot easier even if I wound up with this last issue.

Par : Croft

Croft — Tue, 18 Nov 2008 17:07:41 +0000

Ah here we go:

g++ -c -o OSMutex.o -D__solaris__ -D_REENTRANT -DFD_SETSIZE=65536 -L/usr/local/lib -R/usr/local/lib -include ../PlatformHeader.h -g -Wall -O2 -I. -I.. -I../APIStubLib -I../RTPMetaInfoLib -I../APICommonCode OSMutex.cpp
OSMutex.cpp:49: error: brace-enclosed initializer used to initialize `upad64_t’
*** Error code 1
make: Fatal error: Command failed for target `OSMutex.o’

Since it was complaining about curly braces around PTHREAD_ONCE_INIT in that file, I de-curly-braced it and lo and behold, it compiled successfully but with tons of warnings (not sure if that’s par for the course as I’ve never successfully compiled DSS until now.) Now to see if it actually works…

Par : Croft

Croft — Tue, 18 Nov 2008 16:52:45 +0000

Umm yeah I was confused anyway and I was very much using gcc, not cc, for this. I am indeed getting an error when perusing the log of what went on when it comes to building libCommonUtilitiesLib.a earlier on in the process. I’ll just have to see if I can figure out just why that is.