Yesterday I finally identified the cause why the first connection was always failing and this in turn lead to resolving some major bugs in network and performer code.
All together, this should make veejay a lot more stable so if you find some time for testing, that would be great.
There are still some edge cases where veejay does not correctly shutdown, but now it is known why this happens!
> Yesterday I finally identified the cause why the first connection was always
> failing
> and this in turn lead to resolving some major bugs in network and performer
> code.
> All together, this should make veejay a lot more stable so if you find some
> time for testing, that would be great.
> There are still some edge cases where veejay does not correctly shutdown,
> but now it is known why this happens!
> I've got some testing time coming up. I'm on the road now, doing > vizuals with new veejay setup!
> My Gentoo installation reveals some things I don't experience when > running veejay in Ubuntu. I'll report on that when I can.
> thanks! > --Xen
> On Sep 7, 2:27 am, Niels <nwelb...@gmail.com> wrote: >> Hi everyone,
>> Yesterday I finally identified the cause why the first connection was always >> failing >> and this in turn lead to resolving some major bugs in network and performer >> code.
>> All together, this should make veejay a lot more stable so if you find some >> time for testing, that would be great.
>> There are still some edge cases where veejay does not correctly shutdown, >> but now it is known why this happens!
>> branch: master
>> See you, >> Niels
> -- > You received this message because you are subscribed to the Google Groups
"veejay-discussion" group.
> To post to this group, send email to veejay-discussion@googlegroups.com. > To unsubscribe from this group, send email to
>> I've got some testing time coming up. I'm on the road now, doing >> vizuals with new veejay setup!
>> My Gentoo installation reveals some things I don't experience when >> running veejay in Ubuntu. I'll report on that when I can.
>> thanks! >> --Xen
>> On Sep 7, 2:27 am, Niels <nwelb...@gmail.com> wrote: >>> Hi everyone,
>>> Yesterday I finally identified the cause why the first connection was >>> always >>> failing >>> and this in turn lead to resolving some major bugs in network and >>> performer >>> code.
>>> All together, this should make veejay a lot more stable so if you find >>> some >>> time for testing, that would be great.
>>> There are still some edge cases where veejay does not correctly shutdown, >>> but now it is known why this happens!
>>> branch: master
>>> See you, >>> Niels
>> -- >> You received this message because you are subscribed to the Google Groups >> "veejay-discussion" group. >> To post to this group, send email to veejay-discussion@googlegroups.com. >> To unsubscribe from this group, send email to >> veejay-discussion+unsubscribe@googlegroups.com. >> For more options, visit this group at >> http://groups.google.com/group/veejay-discussion?hl=en.
> -- > You received this message because you are subscribed to the Google Groups > "veejay-discussion" group. > To post to this group, send email to veejay-discussion@googlegroups.com. > To unsubscribe from this group, send email to > veejay-discussion+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/veejay-discussion?hl=en.
I found another something: I get a segfault when switching between
shared video input and another sample then back to the shared video
input. It will also segfault if I go from playing a sample, to
playing shared video, and segfaults when I click in reloaded to play a
sample again. At least that how it seems to behave if that makes
sense.... :)
Reloaded won't connect to veejay when veejay is compiled with "--
enable-strict --enable-debug" so I'm running it without those and
here's what it says (with gdb and without). Hope this helps.
thanks,
--Xen
I: Link: 0 connected with 127.0.0.1 on port 19118
I: Link: 0 connected with 127.0.0.1 on port 41113
I: Link: 1 connected with 127.0.0.1 on port 19630
I: Link: 1 connected with 127.0.0.1 on port 41625
I: Trying to create new sample from /home/neo/NEO/var/neo-33-
frames.avi
D: FFmpeg: File has 1 stream
D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V
(420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi)
D: File is AVI
D: FOURCC is MJPG
D: FFmpeg: File has 1 stream
D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V
(420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi)
D: Full name: /home/neo/NEO/var/neo-33-frames.avi
D: Frames: 34
D: Width: 720
D: Height: 480
D: Interlacing: Not interlaced
D: Frames/sec: 29.970000
D: Sampling format: 4:2:0 full range
D: FOURCC: MJPG
D: Audio samps: 0
D: Audio chans: 0
D: Audio bits: 0
D: Audio rate: 0
I: Created new sample 1 from file /home/neo/NEO/var/neo-33-
frames.avi
D: Not caching this EDL to memory (Cachesize too small)
D: try increasing cache size with -m commandline parameter
D: Allocated 0 effects
D: Sample 1 continues with frame 0
I: Playing sample 1 (FX=0, Sl=0, Speed=1, Start=0, Loop=1)
I: Playing stream 1 (FX=0) (Ff=50)
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fffdae2a700 (LWP 18528)]
0x00007ffff7847eb8 in hash_scan_next () from /home/neo/NEO/usr/lib/
libveejay-1.5.so.0
Yes there were some problems with the recent "fixes", causing all kinds of undefined behaviour This afternoon I rewrote some parts of libvjnet and the unicast stream tag threading code, these I tested tonight and there is good progress,according to helgrind and memcheck there are much less problems So you could update from git and see if its gone, but I will test the SHM plugins tomorrow :)
On Sun, Sep 11, 2011 at 12:30 AM, Xendarboh <xendar...@gmail.com> wrote: > Hi Niels,
> I found another something: I get a segfault when switching between > shared video input and another sample then back to the shared video > input. It will also segfault if I go from playing a sample, to > playing shared video, and segfaults when I click in reloaded to play a > sample again. At least that how it seems to behave if that makes > sense.... :)
> Reloaded won't connect to veejay when veejay is compiled with "-- > enable-strict --enable-debug" so I'm running it without those and > here's what it says (with gdb and without). Hope this helps.
> thanks, > --Xen
> I: Link: 0 connected with 127.0.0.1 on port 19118 > I: Link: 0 connected with 127.0.0.1 on port 41113 > I: Link: 1 connected with 127.0.0.1 on port 19630 > I: Link: 1 connected with 127.0.0.1 on port 41625 > I: Trying to create new sample from /home/neo/NEO/var/neo-33- > frames.avi > D: FFmpeg: File has 1 stream > D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V > (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) > D: File is AVI > D: FOURCC is MJPG > D: FFmpeg: File has 1 stream > D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V > (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) > D: Full name: /home/neo/NEO/var/neo-33-frames.avi > D: Frames: 34 > D: Width: 720 > D: Height: 480 > D: Interlacing: Not interlaced > D: Frames/sec: 29.970000 > D: Sampling format: 4:2:0 full range > D: FOURCC: MJPG > D: Audio samps: 0 > D: Audio chans: 0 > D: Audio bits: 0 > D: Audio rate: 0 > I: Created new sample 1 from file /home/neo/NEO/var/neo-33- > frames.avi > D: Not caching this EDL to memory (Cachesize too small) > D: try increasing cache size with -m commandline parameter > D: Allocated 0 effects > D: Sample 1 continues with frame 0 > I: Playing sample 1 (FX=0, Sl=0, Speed=1, Start=0, Loop=1) > I: Playing stream 1 (FX=0) (Ff=50)
> Program received signal SIGSEGV, Segmentation fault. > [Switching to Thread 0x7fffdae2a700 (LWP 18528)] > 0x00007ffff7847eb8 in hash_scan_next () from /home/neo/NEO/usr/lib/ > libveejay-1.5.so.0
> -- > You received this message because you are subscribed to the Google Groups "veejay-discussion" group. > To post to this group, send email to veejay-discussion@googlegroups.com. > To unsubscribe from this group, send email to veejay-discussion+unsubscribe@googlegroups.com. > For more options, visit this group at http://groups.google.com/group/veejay-discussion?hl=en.
I found some problems between switching samples and the re-allocation of plugins; It appeared that there were some problems in freeing memory and recovering from failed initializations But I fixed a whole bunch of problems and a few of them could have been stopping the show, so please try to re-build all of veejay ...
I tested this: VIMS 250 followed by VIMS 25:1; and VIMS: 251 4490, with two running veejay instances and shared video works :)
On Sun, Sep 11, 2011 at 1:32 AM, Niels Elburg <nwelb...@gmail.com> wrote: > Hi Xen,
> Yes there were some problems with the recent "fixes", causing all > kinds of undefined behaviour > This afternoon I rewrote some parts of libvjnet and the unicast stream > tag threading code, > these I tested tonight and there is good progress,according to > helgrind and memcheck there are much less problems > So you could update from git and see if its gone, but I will test the > SHM plugins tomorrow :)
> See you,
> On Sun, Sep 11, 2011 at 12:30 AM, Xendarboh <xendar...@gmail.com> wrote: >> Hi Niels,
>> I found another something: I get a segfault when switching between >> shared video input and another sample then back to the shared video >> input. It will also segfault if I go from playing a sample, to >> playing shared video, and segfaults when I click in reloaded to play a >> sample again. At least that how it seems to behave if that makes >> sense.... :)
>> Reloaded won't connect to veejay when veejay is compiled with "-- >> enable-strict --enable-debug" so I'm running it without those and >> here's what it says (with gdb and without). Hope this helps.
>> thanks, >> --Xen
>> I: Link: 0 connected with 127.0.0.1 on port 19118 >> I: Link: 0 connected with 127.0.0.1 on port 41113 >> I: Link: 1 connected with 127.0.0.1 on port 19630 >> I: Link: 1 connected with 127.0.0.1 on port 41625 >> I: Trying to create new sample from /home/neo/NEO/var/neo-33- >> frames.avi >> D: FFmpeg: File has 1 stream >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) >> D: File is AVI >> D: FOURCC is MJPG >> D: FFmpeg: File has 1 stream >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) >> D: Full name: /home/neo/NEO/var/neo-33-frames.avi >> D: Frames: 34 >> D: Width: 720 >> D: Height: 480 >> D: Interlacing: Not interlaced >> D: Frames/sec: 29.970000 >> D: Sampling format: 4:2:0 full range >> D: FOURCC: MJPG >> D: Audio samps: 0 >> D: Audio chans: 0 >> D: Audio bits: 0 >> D: Audio rate: 0 >> I: Created new sample 1 from file /home/neo/NEO/var/neo-33- >> frames.avi >> D: Not caching this EDL to memory (Cachesize too small) >> D: try increasing cache size with -m commandline parameter >> D: Allocated 0 effects >> D: Sample 1 continues with frame 0 >> I: Playing sample 1 (FX=0, Sl=0, Speed=1, Start=0, Loop=1) >> I: Playing stream 1 (FX=0) (Ff=50)
>> Program received signal SIGSEGV, Segmentation fault. >> [Switching to Thread 0x7fffdae2a700 (LWP 18528)] >> 0x00007ffff7847eb8 in hash_scan_next () from /home/neo/NEO/usr/lib/ >> libveejay-1.5.so.0
>> -- >> You received this message because you are subscribed to the Google Groups "veejay-discussion" group. >> To post to this group, send email to veejay-discussion@googlegroups.com. >> To unsubscribe from this group, send email to veejay-discussion+unsubscribe@googlegroups.com. >> For more options, visit this group at http://groups.google.com/group/veejay-discussion?hl=en.
I'm still getting the same crashing behavior with the latest veejay.
I didn't test with two veejays yet, but testing shared video with NEO
crashes when trying to select or mix another sample. Seems that
pretty much any order of switching between videos or trying to mix
them will cause a crash.
gdb said this:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fffe06ae700 (LWP 3593)]
0x00007ffff7849599 in hash_scan_next (scan=0x7fffe06ad090) at hash.c:
500
500 hash_val_t nchains = hash->nchains;
thanks,
--Xen
PS: I'm just returning from playing vizuals at an Earthdance festival
last week. I'm LOVING my new VJ setup using veejay. It's working
great! I'm still learning it, and have some more to develop and wow!
I'm excited at the possibilities. Recording samples with veejay is
really awesome. It has been great to leave my big box of gear and s-
video in favor of veejay+NEO with a VGA output. Yay! Thanks so
much! :) :)
On Sep 11, 2:22 pm, Niels Elburg <nwelb...@gmail.com> wrote:
> I found some problems between switching samples and the re-allocation
> of plugins;
> It appeared that there were some problems in freeing memory and
> recovering from failed initializations
> But I fixed a whole bunch of problems and a few of them could have
> been stopping the show, so please try to re-build all of veejay ...
> I tested this: VIMS 250 followed by VIMS 25:1;
> and VIMS: 251 4490, with two running veejay instances and shared video works :)
> On Sun, Sep 11, 2011 at 1:32 AM, Niels Elburg <nwelb...@gmail.com> wrote:
> > Hi Xen,
> > Yes there were some problems with the recent "fixes", causing all
> > kinds of undefined behaviour
> > This afternoon I rewrote some parts of libvjnet and the unicast stream
> > tag threading code,
> > these I tested tonight and there is good progress,according to
> > helgrind and memcheck there are much less problems
> > So you could update from git and see if its gone, but I will test the
> > SHM plugins tomorrow :)
> > See you,
> > On Sun, Sep 11, 2011 at 12:30 AM, Xendarboh <xendar...@gmail.com> wrote:
> >> Hi Niels,
> >> I found another something: I get a segfault when switching between
> >> shared video input and another sample then back to the shared video
> >> input. It will also segfault if I go from playing a sample, to
> >> playing shared video, and segfaults when I click in reloaded to play a
> >> sample again. At least that how it seems to behave if that makes
> >> sense.... :)
> >> Reloaded won't connect to veejay when veejay is compiled with "--
> >> enable-strict --enable-debug" so I'm running it without those and
> >> here's what it says (with gdb and without). Hope this helps.
> >> thanks,
> >> --Xen
> >> I: Link: 0 connected with 127.0.0.1 on port 19118
> >> I: Link: 0 connected with 127.0.0.1 on port 41113
> >> I: Link: 1 connected with 127.0.0.1 on port 19630
> >> I: Link: 1 connected with 127.0.0.1 on port 41625
> >> I: Trying to create new sample from /home/neo/NEO/var/neo-33-
> >> frames.avi
> >> D: FFmpeg: File has 1 stream
> >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V
> >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi)
> >> D: File is AVI
> >> D: FOURCC is MJPG
> >> D: FFmpeg: File has 1 stream
> >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V
> >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi)
> >> D: Full name: /home/neo/NEO/var/neo-33-frames.avi
> >> D: Frames: 34
> >> D: Width: 720
> >> D: Height: 480
> >> D: Interlacing: Not interlaced
> >> D: Frames/sec: 29.970000
> >> D: Sampling format: 4:2:0 full range
> >> D: FOURCC: MJPG
> >> D: Audio samps: 0
> >> D: Audio chans: 0
> >> D: Audio bits: 0
> >> D: Audio rate: 0
> >> I: Created new sample 1 from file /home/neo/NEO/var/neo-33-
> >> frames.avi
> >> D: Not caching this EDL to memory (Cachesize too small)
> >> D: try increasing cache size with -m commandline parameter
> >> D: Allocated 0 effects
> >> D: Sample 1 continues with frame 0
> >> I: Playing sample 1 (FX=0, Sl=0, Speed=1, Start=0, Loop=1)
> >> I: Playing stream 1 (FX=0) (Ff=50)
> >> Program received signal SIGSEGV, Segmentation fault.
> >> [Switching to Thread 0x7fffdae2a700 (LWP 18528)]
> >> 0x00007ffff7847eb8 in hash_scan_next () from /home/neo/NEO/usr/lib/
> >> libveejay-1.5.so.0
> >> --
> >> You received this message because you are subscribed to the Google Groups "veejay-discussion" group.
> >> To post to this group, send email to veejay-discussion@googlegroups.com.
> >> To unsubscribe from this group, send email to veejay-discussion+unsubscribe@googlegroups.com.
> >> For more options, visit this group athttp://groups.google.com/group/veejay-discussion?hl=en.
Happy to hear there is improvement :) Any visuals to share? We would like to see!
I am not sure why veejay crashes; Perhaps you can try a run in valgrind ( valgrind veejay -n -f2 -v ) and see if there is any errors (invalid read/write) shortly before the crash. It looks like memory corruption. Also, from where is hash_scan_next initially called? If it comes from vevo_port_free I probably have an idea where to look,
On Mon, Oct 3, 2011 at 6:49 AM, Xendarboh <xendar...@gmail.com> wrote: > Hi Niels,
> I'm still getting the same crashing behavior with the latest veejay. > I didn't test with two veejays yet, but testing shared video with NEO > crashes when trying to select or mix another sample. Seems that > pretty much any order of switching between videos or trying to mix > them will cause a crash.
> gdb said this: > Program received signal SIGSEGV, Segmentation fault. > [Switching to Thread 0x7fffe06ae700 (LWP 3593)] > 0x00007ffff7849599 in hash_scan_next (scan=0x7fffe06ad090) at hash.c: > 500 > 500 hash_val_t nchains = hash->nchains;
> thanks, > --Xen
> PS: I'm just returning from playing vizuals at an Earthdance festival > last week. I'm LOVING my new VJ setup using veejay. It's working > great! I'm still learning it, and have some more to develop and wow! > I'm excited at the possibilities. Recording samples with veejay is > really awesome. It has been great to leave my big box of gear and s- > video in favor of veejay+NEO with a VGA output. Yay! Thanks so > much! :) :)
> On Sep 11, 2:22 pm, Niels Elburg <nwelb...@gmail.com> wrote: > > Hi Xen
> > I found some problems between switching samples and the re-allocation > > of plugins; > > It appeared that there were some problems in freeing memory and > > recovering from failed initializations > > But I fixed a whole bunch of problems and a few of them could have > > been stopping the show, so please try to re-build all of veejay ...
> > I tested this: VIMS 250 followed by VIMS 25:1; > > and VIMS: 251 4490, with two running veejay instances and shared video > works :)
> > On Sun, Sep 11, 2011 at 1:32 AM, Niels Elburg <nwelb...@gmail.com> > wrote: > > > Hi Xen,
> > > Yes there were some problems with the recent "fixes", causing all > > > kinds of undefined behaviour > > > This afternoon I rewrote some parts of libvjnet and the unicast stream > > > tag threading code, > > > these I tested tonight and there is good progress,according to > > > helgrind and memcheck there are much less problems > > > So you could update from git and see if its gone, but I will test the > > > SHM plugins tomorrow :)
> > > See you,
> > > On Sun, Sep 11, 2011 at 12:30 AM, Xendarboh <xendar...@gmail.com> > wrote: > > >> Hi Niels,
> > >> I found another something: I get a segfault when switching between > > >> shared video input and another sample then back to the shared video > > >> input. It will also segfault if I go from playing a sample, to > > >> playing shared video, and segfaults when I click in reloaded to play a > > >> sample again. At least that how it seems to behave if that makes > > >> sense.... :)
> > >> Reloaded won't connect to veejay when veejay is compiled with "-- > > >> enable-strict --enable-debug" so I'm running it without those and > > >> here's what it says (with gdb and without). Hope this helps.
> > >> thanks, > > >> --Xen
> > >> I: Link: 0 connected with 127.0.0.1 on port 19118 > > >> I: Link: 0 connected with 127.0.0.1 on port 41113 > > >> I: Link: 1 connected with 127.0.0.1 on port 19630 > > >> I: Link: 1 connected with 127.0.0.1 on port 41625 > > >> I: Trying to create new sample from /home/neo/NEO/var/neo-33- > > >> frames.avi > > >> D: FFmpeg: File has 1 stream > > >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V > > >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) > > >> D: File is AVI > > >> D: FOURCC is MJPG > > >> D: FFmpeg: File has 1 stream > > >> D: FFmpeg reports Video [] 720x480. Pixel format: 4:2:0 planar, Y-U-V > > >> (420P JPEG) Has B frames: No (/home/neo/NEO/var/neo-33-frames.avi) > > >> D: Full name: /home/neo/NEO/var/neo-33-frames.avi > > >> D: Frames: 34 > > >> D: Width: 720 > > >> D: Height: 480 > > >> D: Interlacing: Not interlaced > > >> D: Frames/sec: 29.970000 > > >> D: Sampling format: 4:2:0 full range > > >> D: FOURCC: MJPG > > >> D: Audio samps: 0 > > >> D: Audio chans: 0 > > >> D: Audio bits: 0 > > >> D: Audio rate: 0 > > >> I: Created new sample 1 from file /home/neo/NEO/var/neo-33- > > >> frames.avi > > >> D: Not caching this EDL to memory (Cachesize too small) > > >> D: try increasing cache size with -m commandline parameter > > >> D: Allocated 0 effects > > >> D: Sample 1 continues with frame 0 > > >> I: Playing sample 1 (FX=0, Sl=0, Speed=1, Start=0, Loop=1) > > >> I: Playing stream 1 (FX=0) (Ff=50)
> > >> Program received signal SIGSEGV, Segmentation fault. > > >> [Switching to Thread 0x7fffdae2a700 (LWP 18528)] > > >> 0x00007ffff7847eb8 in hash_scan_next () from /home/neo/NEO/usr/lib/ > > >> libveejay-1.5.so.0
> > >> -- > > >> You received this message because you are subscribed to the Google > Groups "veejay-discussion" group. > > >> To post to this group, send email to > veejay-discussion@googlegroups.com. > > >> To unsubscribe from this group, send email to > veejay-discussion+unsubscribe@googlegroups.com. > > >> For more options, visit this group athttp:// > groups.google.com/group/veejay-discussion?hl=en.
> -- > You received this message because you are subscribed to the Google Groups > "veejay-discussion" group. > To post to this group, send email to veejay-discussion@googlegroups.com. > To unsubscribe from this group, send email to > veejay-discussion+unsubscribe@googlegroups.com. > For more options, visit this group at > http://groups.google.com/group/veejay-discussion?hl=en.
Using veejay for Phutureprimitive's DJ set. Out of ~30 shows, ran solid with no crashes. Awesome! I've been using the shared memory plugin with veejay and using the version of veejay that loaded the lvd_shmin plugin on startup. I haven't yet upgraded to the latest veejay because of the bug mentioned in this thread. I've got some time this Summer before on the road in the Fall and want to help fix it and more. :)
I see in that 3rd youtube vid that I overshot the screen a bit on the top... oops. The projector is actually calibrated to project on the object behind the screen which we call the GLYPH or Shpongletron 2.0 -- a video of it in action here, pretty sweet vizual offering from Lucid Technology and Design: http://www.youtube.com/watch?v=7aRhHluVdIE I fit the veejay projection onto the screen in front of the structure. I know that veejay has "Projector/Camera Calibration" feature mentioned here: http://www.veejayhq.net/2008/11/projectorcamera-calibration-augmented... I played with that feature but couldn't get the screen size and location like I wanted, not sure that I understand that feature very well. So I set VEEJAY_SCREEN_GEOMETRY and VEEJAY_SCREEN_SIZE in a startup script to calibrate veejay for each show - set variables, run veejay, restart and change variables again until it's perfect. Wandering, could the veejay Calibration Feature serve the purpose of positioning the screen? mainly setting the location of the top-left corner, then setting the size without distorting the angle of the projected image?
I've upgraded my OS to Ubuntu 12.04 and updated most of the veejay dependencies that I compile in my dev environment and was able to compile all of the latest veejay, no problem. It's running and feels improved, nice! I'm stuck without the lvd_shm* stuff.
> I am not sure why veejay crashes; Perhaps you can try a run in valgrind > ( valgrind veejay -n -f2 -v ) and see if there is any errors (invalid > read/write) > shortly before the crash. It looks like memory corruption. > Also, from where is hash_scan_next initially called? If it comes from > vevo_port_free > I probably have an idea where to look,
I'm pasting more gdb output run using the latest veejay, see below. I do see vevo_port_recursive_free in there.
I'm using veejay to gift visuals at a festival this weekend, then plan to dive deeper into this when I return and can see what valgrind has to say as well. Super excited about this project, sooo awesome to use Linux / OpenSource for visuals. Yay!!!
Program received signal SIGSEGV, Segmentation fault. 0x00007ffff78446ba in veejay_print_backtrace () at vj-common.c:198 198 veejay_addr2line_bt( i + 1, space[i],strings[i] ); (gdb) Continuing.
On Mon, Oct 3, 2011 at 6:49 AM, Xendarboh <xendar...@gmail.com> wrote: > Hi Niels,
> I'm still getting the same crashing behavior with the latest veejay. > I didn't test with two veejays yet, but testing shared video with NEO > crashes when trying to select or mix another sample. Seems that > pretty much any order of switching between videos or trying to mix > them will cause a crash.
> gdb said this: > Program received signal SIGSEGV, Segmentation fault. > [Switching to Thread 0x7fffe06ae700 (LWP 3593)] > 0x00007ffff7849599 in hash_scan_next (scan=0x7fffe06ad090) at hash.c: > 500 > 500 hash_val_t nchains = hash->nchains;
> thanks, > --Xen
> PS: I'm just returning from playing vizuals at an Earthdance festival > last week. I'm LOVING my new VJ setup using veejay. It's working > great! I'm still learning it, and have some more to develop and wow! > I'm excited at the possibilities. Recording samples with veejay is > really awesome. It has been great to leave my big box of gear and s- > video in favor of veejay+NEO with a VGA output. Yay! Thanks so > much! :) :)
> On Sep 11, 2:22 pm, Niels Elburg <nwelb...@gmail.com> wrote: > > Hi Xen
> > I found some problems between switching samples and the re-allocation > > of plugins; > > It appeared that there were some problems in freeing memory and > > recovering from failed initializations > > But I fixed a whole bunch of problems and a few of them could have > > been stopping the show, so please try to re-build all of veejay ...
> > I tested this: VIMS 250 followed by VIMS 25:1; > > and VIMS: 251 4490, with two running veejay instances and shared video > works :)
> > On Sun, Sep 11, 2011 at 1:32 AM, Niels Elburg <nwelb...@gmail.com> > wrote: > > > Hi Xen,
> > > Yes there were some problems with the recent "fixes", causing all > > > kinds of undefined behaviour > > > This afternoon I rewrote some parts of libvjnet and the unicast stream > > > tag threading code, > > > these I tested tonight and there is good progress,according to > > > helgrind and memcheck there are much less problems > > > So you could update from git and see if its gone, but I will test the > > > SHM plugins tomorrow :)
> > > See you,
> > > On Sun, Sep 11, 2011 at 12:30 AM, Xendarboh <xendar...@gmail.com> > wrote:
Thanks for the youtube videos, you have some colorful visuals there! Very
cool to see it all in action
To change the projection setup, some offset of the top left corner and the
size, have you tried to setup the quad?
You can just put the 4 marker points on each (image)corner and place the
quad some offset from the top left corner and change its width/height. This
will map the image into the defined quad.
Setting the marker points is simply left-mouse clicks, the quad is Right
Shift + Left mouse clicks.
Configuring the quad is similar to the marker points but it always stays in
a rectangular shape. Right mouse is for toggling between projection modes
Usually, when I use the projection calibration feature in veejay, I setup a
camera and point that
to the projection.Then, you need another person to emphasize the
projection's screen corners
with a white shiny LED so you can find the positions (on monitor) where to
set the marker points
If its not helping, just make a drawing of the desired setup, it is usually
easier for me,
And there's some quirks in there too.
I comitted some patches to veejay (master branch, git) that should help
debugging
the segmentation fault you are getting on vevo_port_recursive_free
It is important to configure veejay (build from scatch, using autogen.sh)
with these switches:
--enable-strict --enable-debug --enable-portleak
this will produce more detailed stack traces and an overview of unfreed
ports.
Also, since it crashes after VIMS 328 (change mode & go), knowing the
contents of the FX chain will
help (is it LVD plugin, your own, something else ?).
Getting this reproduced will get rid of the problem much quicker, recusive
port free will fail on memory corruption
and we need to find out where that happens :) valgrind will provide some
insights into this, if veejay is configured with mentioned flags.
In the latest git commit livido-plugins are compiled during the build of
veejay-server and put in a directory
relative to veejay's executable ../lib/livido-plugins.
On startup this directory is read, so probably you will need to edit your
$HOME/.veejay/plugins.cfg to
skip loading the lvd_shmin/lvd_shmout plugins.
> Using veejay for Phutureprimitive's DJ set. Out of ~30 shows, ran solid
> with no crashes. Awesome! I've been using the shared memory plugin with
> veejay and using the version of veejay that loaded the lvd_shmin plugin on
> startup. I haven't yet upgraded to the latest veejay because of the bug
> mentioned in this thread. I've got some time this Summer before on the
> road in the Fall and want to help fix it and more. :)
> I see in that 3rd youtube vid that I overshot the screen a bit on the
> top... oops. The projector is actually calibrated to project on the object
> behind the screen which we call the GLYPH or Shpongletron 2.0 -- a video
> of it in action here, pretty sweet vizual offering from Lucid Technology
> and Design:
> http://www.youtube.com/watch?v=7aRhHluVdIE > I fit the veejay projection onto the screen in front of the structure. I
> know that veejay has "Projector/Camera Calibration" feature mentioned here:
> http://www.veejayhq.net/2008/11/projectorcamera-calibration-augmented... > I played with that feature but couldn't get the screen size and location
> like I wanted, not sure that I understand that feature very well. So I set
> VEEJAY_SCREEN_GEOMETRY
> and VEEJAY_SCREEN_SIZE in a startup script to calibrate veejay for each
> show - set variables, run veejay, restart and change variables again until
> it's perfect. Wandering, could the veejay Calibration Feature serve the
> purpose of positioning the screen? mainly setting the location of the
> top-left corner, then setting the size without distorting the angle of the
> projected image?
> I've upgraded my OS to Ubuntu 12.04 and updated most of the veejay
> dependencies that I compile in my dev environment and was able to compile
> all of the latest veejay, no problem. It's running and feels improved,
> nice! I'm stuck without the lvd_shm* stuff.
>> I am not sure why veejay crashes; Perhaps you can try a run in valgrind
>> ( valgrind veejay -n -f2 -v ) and see if there is any errors (invalid
>> read/write)
>> shortly before the crash. It looks like memory corruption.
>> Also, from where is hash_scan_next initially called? If it comes from
>> vevo_port_free
>> I probably have an idea where to look,
> I'm pasting more gdb output run using the latest veejay, see below. I do
> see vevo_port_recursive_free in there.
> I'm using veejay to gift visuals at a festival this weekend, then plan to
> dive deeper into this when I return and can see what valgrind has to say as
> well. Super excited about this project, sooo awesome to use Linux /
> OpenSource for visuals. Yay!!!
> Thanks for the youtube videos, you have some colorful visuals there! Very > cool to see it all in action
Yay, I like colors :)
Thanks for the projection calibration info, I'm gonna play with it more.
I comitted some patches to veejay (master branch, git) that should help
> debugging > the segmentation fault you are getting on vevo_port_recursive_free > It is important to configure veejay (build from scatch, using autogen.sh) > with these switches: > --enable-strict --enable-debug --enable-portleak > this will produce more detailed stack traces and an overview of unfreed > ports.
Sweet. I updated the latest and when running autogen.sh for veejay-server, there seems to be a missing file?: <snip> autoreconf: running: /usr/bin/autoconf --include=m4 --force autoreconf: running: /usr/bin/autoheader --include=m4 --force autoreconf: running: automake --add-missing --copy --force-missing configure.ac:1030: required file `livido-plugins/Makefile.in' not found autoreconf: automake failed with exit status: 1
Also, since it crashes after VIMS 328 (change mode & go), knowing the
> contents of the FX chain will > help (is it LVD plugin, your own, something else ?).
The LVD shm-in plugin. nothing else in the FX chain as far as I know. The plugin works, and crash when switching to another sample.
I added the Makefile.am for the livido-plugins, it should compile now.
Also, since it crashes after VIMS 328 (change mode & go), knowing the
>> contents of the FX chain will
>> help (is it LVD plugin, your own, something else ?).
> The LVD shm-in plugin. nothing else in the FX chain as far as I know. The
> plugin works, and crash when switching to another sample.
Ok, what is feeding the LVD shm-in plugin , another veejay with LVD shm-out
plugin or veejay master (built-in shm) ?
So far, I found that the LVD shm-out writes a different header then
veejay's built-in shm writer, that's wrong enough for a segfault.
A fix for this is pending, but it may not be the same problem; depending on
what is writing on the other end of the segment.
Also, if it's possible, run the crashing veejay in valgrind:
valgrind --log-file=/tmp/veejay.log --leak-resolution=high veejay -n -v
-f2 -w352 -h288
(at 2fps, low resolution , so it wont consume all your resources)
I added the Makefile.am for the livido-plugins, it should compile now.
Great! It does compile now. To find deps and compile on my system with ffmpeg (and other things) installed in a non-standard location, I added this to livido-plugins/Makefile.am:
INCLUDES = $(FFMPEG_CFLAGS)
I've compiled veejay with the suggested flags including --enable-portleak.
Ok, what is feeding the LVD shm-in plugin , another veejay with LVD shm-out
> plugin or veejay master (built-in shm) ?
I'm using code I wrote to feed veejay, it's based on the LVD shm-out plugin -- another veejay with LVD shm-out plugin has the same issue. I haven't worked with veejay built-in shm yet, that's next. :)
I can produce this with two veejays. One headless veejay using the LVD shm-out plugin. Then another veejay with LVD shm-in plugin correctly displays the shm video, but crash-o-la when changing to another sample.
I ran it with gdb and valgrind. Here's the crash info now with gdb, I notice 3 new vevo_property_set lines.: <snip> 00:01:29.454308 D: L 0 queue [ 0]/[1] [ 8] '328:0 1;' 00:01:29.454354 D: VIMS '328:0 1;' Play sample / stream 00:01:29.454370 D: VIMS: @25 Parse message '328:0 1;' 00:01:29.454420 D: Destroy 'Shared Memory Reader Veejay' 00:01:29.454480 D: vevo_property_set: key=0x7fffd8670000, type=4, n=1, addr=0x7fffe3ffe1d0 00:01:29.454519 D: vevo_property_set: key=0x7fffd8014fd0, type=66, n=1, addr=0x7fffe3ffe1a0 00:01:29.454540 D: vevo_property_set: key=0x7fffd8006950, type=66, n=1, addr=0x7fffe3ffe150
Program received signal SIGSEGV, Segmentation fault. 0x00007ffff7843702 in veejay_print_backtrace () at vj-common.c:198 198 veejay_addr2line_bt( i + 1, space[i],strings[i] ); (gdb) Continuing. </snip>
Run with valgrind --log-file=/tmp/veejay.log --leak-resolution=high veejay -n -v -f2 -w352 -h288 <snip> 00:00:45.707454 D: L 0 queue [ 0]/[1] [ 8] '328:1 2;' 00:00:45.707925 D: VIMS '328:1 2;' Play sample / stream 00:00:45.708176 D: VIMS: @46 Parse message '328:1 2;' 00:00:45.725769 D: Destroy 'Shared Memory Reader Veejay' 00:00:45.727805 D: vevo_property_set: key=0x1972ff10, type=4, n=1, addr=0x194541d0 00:00:45.731076 D: vevo_property_set: key=0x15f54e40, type=66, n=1, addr=0x194541a0 00:00:45.733980 D: vevo_property_set: key=0x15f54f00, type=66, n=1, addr=0x19454150 veejay: hash.c:501: hash_scan_next: Assertion `hash != ((void *)0)' failed. 00:00:45.783153 E: No subtitle sequences present, nothing to save 00:00:45.805169 E: No subtitle sequences present, nothing to save 00:00:45.910710 W: Saved samplelist to /home/neo/.veejay/recovery/recovery_samplelist_p1681.sl 00:00:45.911430 W: Please report this error to http://groups.google.com/group/veejay-discussion?hl=en 00:00:45.911753 W: Send at least veejay's output and include the command(s) you have used to start it. 00:00:45.912089 W: Also, please consider sending in the recovery files if any have been created. 00:00:45.912425 W: If you compiled it yourself, please include information about your system. 00:00:45.912906 W: Dumping core file to: core.1681 00:00:45.915855 W: Please wait! Running command 'gdb -p 1681 -batch -x veejay.cmd' Could not attach to process. If your uid matches the uid of the target process, check the setting of /proc/sys/kernel/yama/ptrace_scope, or try again as the root user. For more details, see /etc/sysctl.d/10-ptrace.conf ptrace: Operation not permitted. 00:00:45.978824 W: Done! 00:00:45.979303 I: Please bzip2 and upload the coredump somewhere and tell us where to find it! 00:00:45.979900 D: Change state to 0 by veejay_change_state_save:335 </snip>
This is the the full contents of /tmp/veejay.log: ==1681== Memcheck, a memory error detector ==1681== Copyright (C) 2002-2011, and GNU GPL'd, by Julian Seward et al. ==1681== Using Valgrind-3.7.0 and LibVEX; rerun with -h for copyright info ==1681== Command: /home/neo/NEO/usr/bin/veejay -n -v -f2 -w352 -h288 ==1681== Parent PID: 1674 ==1681== ==1681== Thread 4: ==1681== Conditional jump or move depends on uninitialised value(s) ==1681== at 0x768717A: ??? (in /media/DATA/NEO/src-20120603/x/NEO/usr/lib/libswscale.so.0.12.0) ==1681== by 0x7689B42: sws_scale (in /media/DATA/NEO/src-20120603/x/NEO/usr/lib/libswscale.so.0.12.0) ==1681== by 0x4FAF0DC: yuv_convert_and_scale (yuvconv.c:1233) ==1681== by 0x4F68E38: vj_fast_picture_save_to_mem (pixbuf.c:442) ==1681== by 0x4E9F2B5: vj_event_get_scaled_image (vj-event.c:8636) ==1681== by 0x4EA87EE: vj_event_vevo_inline_fire (vj-eventman.c:162) ==1681== by 0x4E82871: vj_event_fire_net_event (vj-event.c:1391) ==1681== by 0x4E832CB: vj_event_parse_msg (vj-event.c:1662) ==1681== by 0x4E83511: vj_event_update_remote (vj-event.c:1732) ==1681== by 0x4ED2325: veejay_handle_callbacks (liblavplayvj.c:1531) ==1681== by 0x4ED24B4: veejay_event_handle (liblavplayvj.c:1567) ==1681== by 0x4ED4EC3: veejay_playback_cycle (liblavplayvj.c:2618) ==1681== ==1681== Warning: invalid file descriptor -1 in syscall write() ==1681== Warning: invalid file descriptor -1 in syscall close() ==1681== ==1681== HEAP SUMMARY: ==1681== in use at exit: 13,538,243 bytes in 30,064 blocks ==1681== total heap usage: 49,055 allocs, 18,991 frees, 26,263,972 bytes allocated ==1681== ==1681== LEAK SUMMARY: ==1681== definitely lost: 15,108 bytes in 529 blocks ==1681== indirectly lost: 22,046 bytes in 781 blocks ==1681== possibly lost: 6,548 bytes in 90 blocks ==1681== still reachable: 13,494,541 bytes in 28,664 blocks ==1681== suppressed: 0 bytes in 0 blocks ==1681== Rerun with --leak-check=full to see details of leaked memory ==1681== ==1681== For counts of detected and suppressed errors, rerun with: -v ==1681== Use --track-origins=yes to see where uninitialised values come from ==1681== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 2
...
I just did a "git pull" today and found that the crash is fixed. Awesome, much thanks! :)
The LVD Shared Memory Reader (shmin) can now be loaded and switching to other samples works without crashing.
I'm now using "-Z 1" veejay parameter to load the LVD shmin plugin as a generator on startup. If I don't do that: other samples/streams can be mixed on the shmin video (of FX Chain Index #0 on the first slot) but the shmin video can't be mixed on other slots because it just mixes the solid stream from "underneath".
-Z/--load-generators [num] Load all generator plugins as streams and start with N sweet option, some "???" displayed in reloaded and otherwise working great so far. happy to be using latest veejay :)
I found another crash, maybe related to the other cuz I see hash_scan_next in the mix as well. This one is easy (for me) to reproduce. Not dealing with shm. I start veejay from the terminal then CTRL-C in the terminal to kill it and it crashes (core dump). I'm pretty sure that this was not the behavior before. Here's some info:
00:00:04.777517 E: port 0x7fff5c25a460 freed (_new_event::208) 00:00:04.777796 D: vevo_property_set: key=0x4ce6000, type=4, n=1, addr=0x7fff5c25a430 00:00:04.777834 D: vevo_property_set: key=0x4ce6000, type=66, n=0, addr=(nil) 00:00:04.777842 E: port 0x7fff5c25a400 freed (vevo_port_register::2089) 00:00:04.777855 D: vevo_property_set: key=0x4521710, type=66, n=0, addr=(nil) 00:00:04.777862 E: port 0x7fff5c25a400 freed (deal_with_livido::1425) 00:00:04.777947 D: vevo_property_set: key=0x4521710, type=4, n=1, addr=0x7fff5c25a430 00:00:04.777969 D: vevo_property_set: key=0x4521710, type=66, n=0, addr=(nil) 00:00:04.777976 E: port 0x7fff5c25a400 freed (vevo_port_register::2089) 00:00:04.777988 D: vevo_property_set: key=0x4521e30, type=66, n=0, addr=(nil) 00:00:04.777995 E: port 0x7fff5c25a400 freed (deal_with_livido::1425) 00:00:04.778049 D: vevo_property_set: key=0x4521710, type=4, n=1, addr=0x7fff5c25a430 00:00:04.778066 D: vevo_property_set: key=0x44a6950, type=66, n=1, addr=0x7fff5c25a400 00:00:04.778074 E: ports all mangled 00:00:04.778083 D: vevo_property_set: key=0x44a6ad0, type=66, n=1, addr=0x7fff5c25a400 00:00:04.778090 E: ports all mangled 00:00:04.778099 D: vevo_property_set: key=0x44a6c50, type=66, n=1, addr=0x7fff5c25a400 00:00:04.778106 E: ports all mangled veejay: hash.c:501: hash_scan_next: Assertion `hash != ((void *)0)' failed. 00:00:04.778243 E: No subtitle sequences present, nothing to save 00:00:04.778548 W: Saved samplelist to /home/neo/.veejay/recovery/recovery_samplelist_p3852.sl 00:00:04.778563 W: Please report this error to http://groups.google.com/group/veejay-discussion?hl=en 00:00:04.778569 W: Send at least veejay's output and include the command(s) you have used to start it. 00:00:04.778575 W: Also, please consider sending in the recovery files if any have been created. 00:00:04.778581 W: If you compiled it yourself, please include information about your system. 00:00:04.778587 W: Dumping core file to: core.3852 00:00:04.778642 W: Please wait! Running command 'gdb -p 3852 -batch -x veejay.cmd' Could not attach to process. If your uid matches the uid of the target process, check the setting of /proc/sys/kernel/yama/ptrace_scope, or try again as the root user. For more details, see /etc/sysctl.d/10-ptrace.conf ptrace: Operation not permitted. 00:00:04.820901 W: Done! 00:00:04.820936 I: Please bzip2 and upload the coredump somewhere and tell us where to find it! 00:00:04.820946 D: Change state to 0 by veejay_change_state_save:335 ./start.sh: line 52: 3852 Aborted (core dumped) $PREFIX/bin/veejay -v --port 3490 --audio 0 -N 1 --output-width $SRC_WIDTH --output-height $SRC_HEIGHT $PARAMS
neo@meta:~/NEO$ gdb usr/bin/veejay ./v-0/core <snip> Reading symbols from /media/DATA/NEO/src-20120603/x/NEO/usr/bin/veejay...done. [New LWP 3852] warning: Can't read pathname for load map: Input/output error. [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Core was generated by `/home/neo/NEO/usr/bin/veejay -v --port 3490 --audio 0 -N 1 --output-width 720 -'. Program terminated with signal 6, Aborted. #0 0x00007f7e0b4c6445 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64 64 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory. (gdb) backtrace #0 0x00007f7e0b4c6445 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64 #1 0x00007f7e0b4c9b30 in __GI_abort () at abort.c:120 #2 0x00007f7e0b4bf10e in __assert_fail_base (fmt=<optimized out>, assertion=0x7f7e0b9e266c "hash != ((void *)0)", file=0x7f7e0b9e2560 "hash.c", line=<optimized out>, function=<optimized out>) at assert.c:94 #3 0x00007f7e0b4bf1b2 in __GI___assert_fail (assertion=0x7f7e0b9e266c "hash != ((void *)0)", file=0x7f7e0b9e2560 "hash.c", line=501, function=0x7f7e0b9e2779 "hash_scan_next") at assert.c:103 #4 0x00007f7e0b8fb4e4 in hash_scan_next (scan=0x7fff5c25a420) at hash.c:501 #5 0x00007f7e0b9cfcb4 in vevo_list_nodes_ (p=0x4521710, atype=66) at vevo.c:1990 #6 0x00007f7e0b9d0596 in vevo_port_recursive_free (port=0x44cab00) at vevo.c:2268 #7 0x00007f7e0b95dcca in free_plugin (plugin=0x44cab00) at plugload.c:257 #8 0x00007f7e0b95de62 in free_plugins () at plugload.c:301 #9 0x00007f7e0b95e2f0 in plug_sys_free () at plugload.c:427 #10 0x00007f7e0b90210a in vj_effect_shutdown () at vj-effect.c:697 #11 0x00007f7e0b8eab79 in veejay_free (info=0x18e1000) at liblavplayvj.c:497 #12 0x00000000004041cb in main (argc=12, argv=0x7fff5c25a7e8) at veejay.c:787 (gdb)
On Thu, Jun 28, 2012 at 9:41 AM, Xendarboh <xendar...@gmail.com> wrote:
> Hi :)
> I just did a "git pull" today and found that the crash is fixed. Awesome,
> much thanks! :)
It is not fixed entirely, otherwise you would have heard the good news :)
The error log you posted says " 00:00:04.778106 E: ports all mangled ",
so there's something very wrong still. I probably get this fixed soon tho
This is related to your crash on exiting veejay.
Does it exit without crashing if you do 'echo "600:;" |sayVIMS' ?
The LVD Shared Memory Reader (shmin) can now be loaded and switching to
> other samples works without crashing.
Excellent!
> I'm now using "-Z 1" veejay parameter to load the LVD shmin plugin as a
> generator on startup. If I don't do that: other samples/streams can be
> mixed on the shmin video (of FX Chain Index #0 on the first slot) but the
> shmin video can't be mixed on other slots because it just mixes the solid
> stream from "underneath".
This is not entirely clear, what stream would you like to feed into SHM-in
, the result of the FX Chain on some index position ?
The error log you posted says " 00:00:04.778106 E: ports all mangled ",
> so there's something very wrong still. I probably get this fixed soon tho > This is related to your crash on exiting veejay. > Does it exit without crashing if you do 'echo "600:;" |sayVIMS' ?
Nope, also crashes with similar backtrace using sayVIMS.
I'm now using "-Z 1" veejay parameter to load the LVD shmin plugin as a
>> generator on startup. If I don't do that: other samples/streams can be >> mixed on the shmin video (of FX Chain Index #0 on the first slot) but the >> shmin video can't be mixed on other slots because it just mixes the solid >> stream from "underneath".
> This is not entirely clear, what stream would you like to feed into SHM-in > , the result of the FX Chain on some index position ?
Can the result of the FX Chain of one slot be mixed with other slots? Seems that the mixing effects mix the raw Sample/Stream of a source without effects applied. I don't need that functionality, just curious - and learning more about how the FX chain works.
Basically, the older version of veejay that I've been using loaded the LVD shm-in in the first slot on startup (as a generator, I think). The current version loads a solid stream. I'm able to achieve similar with "-Z 1" -- when LVD shm-in is loaded on startup as a generator, other slots can mix it as a source. so that works for me.
I'm noticing a few things, which I can work around. When using "-Z 1", two generators are loaded; LVD shm-in in the first slot, and a solid stream in the second slot. cool. When using "-Z 0", it looks like what happens without -Z with only a solid stream loaded which doesn't seem to match the description of the option which says it loads all: "Load all generator plugins as streams and start with N".
I'm getting a crash when loading a saved sample list that was saved with generators loaded. ) starting veejay with -Z 1 ) using reloaded to load a video file as sample, save sample list ) exit veejay and reloaded, start again, load saved sample list with reloaded and veejay crashes with: 00:00:33.292887 E: 'LVDShared Memory Reader Veejay' not found. 00:00:33.292926 E: 'LVDSolid Color Fill' not found. 00:00:33.292980 I: Loaded sample list from file '/media/DATA/NEO/src-20120603/x/NEO/v-0/tt.sl' 00:00:33.293148 E: Found Gremlins in your system. 00:00:33.293193 W: No fresh ale found in the fridge. 00:00:33.293217 I: Running with sub-atomic precision... 00:00:33.293231 I: (SEGV_ACCERR) invalid access to (nil) at 88f12077 00:00:33.301724 I: 0 ??:0 ?? (@147077) 00:00:33.302035 I: register [ 0] =88e7db6b 00:00:33.302074 I: register [ 1] =891ca1a0 00:00:33.302113 I: register [ 2] =790a15f0 00:00:33.302139 I: register [ 3] =88e86274 00:00:33.302161 I: register [ 4] =850bf880 00:00:33.302187 I: register [ 5] =790a29c0 00:00:33.302210 I: register [ 6] =0 00:00:33.302257 I: register [ 7] =3 00:00:33.302280 I: register [ 8] =1 00:00:33.302306 I: register [ 9] =1 00:00:33.302328 I: register [10] =790a1920 00:00:33.302355 I: register [11] =0 00:00:33.302380 I: register [12] =0 00:00:33.302406 I: register [13] =0 00:00:33.302429 I: register [14] =88f0d6fa 00:00:33.302454 I: register [15] =790a1900 00:00:33.302476 I: register [16] =88f12077 00:00:33.302503 I: register [17] =10297 00:00:33.302526 I: register [18] =33 00:00:33.302552 I: register [19] =4 00:00:33.302579 I: register [20] =e 00:00:33.302601 I: register [21] =1000 00:00:33.302626 I: register [22] =0 00:00:33.302699 I: 1 : 0x7fe388f12077 <vj_tag_get_effect_any+100> (/home/neo/NEO/usr/lib/libveejay-1.5.so.0) 00:00:33.302773 I: 2 : 0x7fe388e48a8c <(null)+140615230982796> (/home/neo/NEO/usr/lib/libveejay-1.5.so.0) 00:00:33.302847 I: 3 : 0x7fe388e50d4f <vj_perform_queue_video_frame+519> (/home/neo/NEO/usr/lib/libveejay-1.5.so.0) 00:00:33.302923 I: 4 : 0x7fe388e6e165 <(null)+140615231136101> (/home/neo/NEO/usr/lib/libveejay-1.5.so.0) 00:00:33.302998 I: 5 : 0x7fe388e6e525 <(null)+140615231137061> (/home/neo/NEO/usr/lib/libveejay-1.5.so.0)
gdb backtrace: Program terminated with signal 11, Segmentation fault. #0 0x00007fe388e7c702 in veejay_print_backtrace () at vj-common.c:198 198 veejay_addr2line_bt( i + 1, space[i],strings[i] ); (gdb) backtrace #0 0x00007fe388e7c702 in veejay_print_backtrace () at vj-common.c:198 #1 0x00007fe388e7c9ad in veejay_backtrace_handler (n=11, dist=0x7fe3790a15f0, x=0x7fe3790a14c0) at vj-common.c:291 #2 <signal handler called> #3 0x00007fe388f12077 in vj_tag_get_effect_any (t1=1, position=0) at vj-tag.c:2115 #4 0x00007fe388e48a8c in vj_perform_verify_rows (info=0x23a0000) at vj-perform.c:509 #5 0x00007fe388e50d4f in vj_perform_queue_video_frame (info=0x23a0000, skip_incr=0) at vj-perform.c:3606 #6 0x00007fe388e6e165 in veejay_playback_cycle (info=0x23a0000) at liblavplayvj.c:2694 #7 0x00007fe388e6e525 in veejay_playback_thread (data=0x23a0000) at liblavplayvj.c:2788 #8 0x00007fe3850b5e9a in start_thread (arg=0x7fe3790a2700) at pthread_create.c:308 #9 0x00007fe388b004bd in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112 #10 0x0000000000000000 in ?? ()
I notice a difference in the .sl files that I used with the older veejay version. Inside <stream> tags for the "LVDSolid Color Fill", the old version has <source_type>4</source_type> while the latest has <source_type>7</source_type>
Mentioning that info in case it is helpful :) I can work around samplelists with some sayVIMS magic.
On Thursday, June 28, 2012 2:16:10 AM UTC-7, Niels wrote:
> Hi Xen!
> On Thu, Jun 28, 2012 at 9:41 AM, Xendarboh <xendar...@gmail.com> wrote:
>> Hi :)
>> I just did a "git pull" today and found that the crash is fixed. >> Awesome, much thanks! :)
> It is not fixed entirely, otherwise you would have heard the good news :)
> The error log you posted says " 00:00:04.778106 E: ports all mangled ", > so there's something very wrong still. I probably get this fixed soon tho > This is related to your crash on exiting veejay. > Does it exit without crashing if you do 'echo "600:;" |sayVIMS' ?
> The LVD Shared Memory Reader (shmin) can now be loaded and switching to >> other samples works without crashing.
> Excellent!
>> I'm now using "-Z 1" veejay parameter to load the LVD shmin plugin as a >> generator on startup. If I don't do that: other samples/streams can be >> mixed on the shmin video (of FX Chain Index #0 on the first slot) but the >> shmin video can't be mixed on other slots because it just mixes the solid >> stream from "underneath".
> This is not entirely clear, what stream would you like to feed into SHM-in > , the result of the FX Chain on some index position ?
The FX Chain can do recursive rendering , but its functionality is limited
to single source FX (image filters) and to slots 0, 1 and 2 of the effect
chain of the mixing sample (2nd input source).
VIMS 048:2; allows you to toggle it on/off, default is off.
Also, I have comitted many bugfixes to veejay today, let me know if that
resolves your veejay problems :)
Bugfixes:
- FX activation/deactivation (all) (segf)
- sample/stream deletion (segf)
- recursive freeing of vevo ports (LVD) (segf)
- commandline option Z[num] <= total generators
- dangling EDL pointers (segf)
- dangling plugin instance pointers
- soft reference port to port pointers (segf)
> The FX Chain can do recursive rendering , but its functionality is limited
> to single source FX (image filters) and to slots 0, 1 and 2 of the effect
> chain of the mixing sample (2nd input source).
> VIMS 048:2; allows you to toggle it on/off, default is off.
Veejay is awesome! I'm testing this feature with reloaded and
reporting some things I noticed that does not look like desired
behavior. I haven't pinned down the behavior exactly to describe it
very well, seems to vary depending on which image effects are sources
(sample or stream) are used.
) Sometimes the FX chain parameters of the mixed sample are
"inherited" by the mixing sample. For example: I load Solarize image
effect on Slot #0 (which has lvd_shmin generator) then load Normal
Overlay mixing effect in Slot #2 which has a video sample. Play Slot
#2 with Slot #0 as mixing source. I see that the solarize effect is
applied in the mix, cool. If I change the value of p0 for the Normal
Overlay effect, the Solarize effect's p0 also changes. This change of
the Solarize effect is only on visible in Slot#2 -- if I play Slot #0,
the Solarize effect's p0 is like it was before being changed in
Slot#2's mix.
) The anomalies vary depending on the image effect used. sometimes it
crashes. sometimes slot #0 is distorted or disappears when mixed in
Slot #2 and then when I switch to slot#0 it is also distorted there
and can be reset by changing p0 on Slot #0. I noticed distortion with
swirl effect at least once. and haven't tested too much, but so far
more than one image effects seem to make black video in the mix.
) Changing p0 for Normal Overlay effect sometimes causes crash. I've
also seen a crash when changing mixing sources. seems like having an
lvd plugin (either running as generator or in FX chain) is more crash
prone, not sure cuz I use the lvd plugins more often than not in
general. I saw at least one crash that happened in "Bathroom Mirrors"
effect's code when changing p0 of "Normal Overlay".
Here's some veejay output and gdb snippet of a crash:
00:23:41.724189 D: VIMS '328:0 1;' Play sample / stream
00:23:41.724214 D: VIMS: @51 Parse message '328:0 1;'
00:23:41.724262 D: Destroy 'Shared Memory Reader Veejay'
00:23:41.724312 D: vevo_property_del: key=0x7f8ad4e5d1e0
00:23:41.724416 D: port 0x7f8adbffe140 freed (init_ports_from_template::734)
00:23:41.724451 D: vevo_property_del: key=0x7f8ad4e5d140
00:23:41.724497 D: port 0x7f8adbffe190 freed (livido_plug_init::975)
00:23:41.724525 D: Not caching this EDL to memory (Cachesize too small)
00:23:41.724539 D: try increasing cache size with -m commandline parameter
00:23:41.724569 D: Not caching this EDL to memory (Cachesize too small)
00:23:41.724583 D: try increasing cache size with -m commandline parameter
00:23:41.724599 D: Allocated 1 effects
00:23:41.724614 D: Subtitling sample 1: 0 - 120
00:23:41.724635 D: No composite config for sample 1
00:23:41.724651 D: Sample 1 continues with frame 0
00:23:41.724667 I: Playing sample 1 (FX=1, Sl=0, Speed=1, Start=0, Loop=1)
veejay: vj-effman.c:52: vj_effman_apply_plug_effect: Assertion
`instance != ((void *)0)' failed.
00:23:41.729637 D: Saved EDL 'sample_00001.edl' of sample 1
<snip>
(gdb) backtrace
#0 0x00007f8aefef0445 in __GI_raise (sig=<optimized out>) at
../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1 0x00007f8aefef3b30 in __GI_abort () at abort.c:120
#2 0x00007f8aefee910e in __assert_fail_base (fmt=<optimized out>,
assertion=0x7f8af040dc2c "instance != ((void *)0)",
file=0x7f8af040dc20 "vj-effman.c", line=<optimized out>,
function=<optimized out>) at assert.c:94
#3 0x00007f8aefee91b2 in __GI___assert_fail (assertion=0x7f8af040dc2c
"instance != ((void *)0)", file=0x7f8af040dc20 "vj-effman.c", line=52,
function=0x7f8af040de70 "vj_effman_apply_plug_effect") at assert.c:103
#4 0x00007f8af032cfe5 in vj_effman_apply_plug_effect
(frames=0x7f8adbffe740, frameinfo=0x184f000, todo_info=0x1850000,
arg=0x7f8adbffe6a0, n_arg=0, entry=140, fx_id=500, instance=0x0)
at vj-effman.c:52
#5 0x00007f8af032fd65 in vj_effect_apply (frames=0x7f8adbffe740,
frameinfo=0x184f000, kf=0x1850000, selector=500,
arguments=0x7f8adbffe6a0, ptr=0x0) at vj-effman.c:718
#6 0x00007f8af02f6ef2 in vj_perform_apply_first (info=0x1846000,
todo_info=0x1850000, frames=0x7f8adbffe740, frameinfo=0x184f000,
e=500, c=0, n_frame=0, ptr=0x0) at vj-perform.c:1275
#7 0x00007f8af02f9a99 in vj_perform_preprocess_secundary
(info=0x1846000, id=2, mode=0, current_ssm=0, chain_entry=0,
frames=0x7f8adbffe8c0, frameinfo=0x184f000) at vj-perform.c:2251
#8 0x00007f8af02fa036 in vj_perform_render_chain_entry
(info=0x1846000, chain_entry=0) at vj-perform.c:2367
#9 0x00007f8af02fa327 in vj_perform_sample_complete_buffers
(info=0x1846000, hint444=0x7f8adbffe994) at vj-perform.c:2456
#10 0x00007f8af02fce38 in vj_perform_queue_video_frame
(info=0x1846000, skip_incr=0) at vj-perform.c:3579
#11 0x00007f8af031a2fd in veejay_playback_cycle (info=0x1846000) at
liblavplayvj.c:2693
#12 0x00007f8af031a6bd in veejay_playback_thread (data=0x1846000) at
liblavplayvj.c:2787
#13 0x00007f8aec561e9a in start_thread (arg=0x7f8adbfff700) at
pthread_create.c:308
#14 0x00007f8aeffac4bd in clone () at
../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#15 0x0000000000000000 in ?? ()
> Also, I have comitted many bugfixes to veejay today, let me know if that
> resolves your veejay problems :)
Great!! I synced to the latest and working well. I will play with it
more and report any anomalies. I expect to be working with it every
day for a while :)
The only thing I'm now seeing is loading sample lists is still
crashing, now in a different way than before. From the output, I see
that it deletes samples and loads the video files then crash (sniped
some of the output pasted here):
00:00:14.899347 I: Deleted all samples (0 remaining)
00:00:14.922766 D: VIMS
'125:/media/DATA/NEO/src-20120603/x/NEO/v-0/t101.sl;' Load samples
from file
00:00:14.922791 D: VIMS: @30 Parse message
'125:/media/DATA/NEO/src-20120603/x/NEO/v-0/t101.sl;'
00:00:14.923882 I: Merging
/media/DATA/NEO/src-20120603/x/NEO/v-0/t101.sl into current
samplelist, auto number starts at -1
00:00:14.923948 I: Existing streams will be deleted (samplelist
overrides active streams)
GDB:
Program terminated with signal 11, Segmentation fault.
#0 0x00007fc2a87e689a in veejay_print_backtrace () at vj-common.c:198
198 veejay_addr2line_bt( i + 1, space[i],strings[i] );
(gdb) backtrace
#0 0x00007fc2a87e689a in veejay_print_backtrace () at vj-common.c:198
#1 0x00007fc2a87e6b45 in veejay_backtrace_handler (n=11,
dist=0x7fc298890670, x=0x7fc298890540) at vj-common.c:291
#2 <signal handler called>
#3 __strncasecmp_l_sse42 () at
...
> ) The anomalies vary depending on the image effect used. sometimes it
> crashes. sometimes slot #0 is distorted or disappears when mixed in
Could you send me the recovery files veejay saves in ~/.veejay/recovery,
it would help me reproduce this problem much quicker!
Also, in your error description, you speak of slots which in my world are
FX chain entries, the rest is just sample (or stream) so I have doubt
how to reproduce what you experienced :)
However, there is a weakness in switching between streams backed with a LVD
generator,
I am currently debugging it
Cheers,
Niels
--
You received this message because you are subscribed to the Google Groups
"veejay-discussion" group.
To post to this group, send email to veejay-discussion@googlegroups.com.
To unsubscribe from this group, send email to
veejay-discussion+unsubscribe@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/veejay-discussion?hl=en.
> Could you send me the recovery files veejay saves in ~/.veejay/recovery,
> it would help me reproduce this problem much quicker!
> Also, in your error description, you speak of slots which in my world are
> FX chain entries, the rest is just sample (or stream) so I have doubt
> how to reproduce what you experienced :)
I get the term "slot" from Reloaded. In the Samplebank it says stuff
like "Slots 0 to 12" and basically a slot holds either a sample or
stream.
I'm attaching two recovery files that I got by doing this:
) start veejay with lvdshmin generator, activate Bathroom Mirror image FX on it
) load video file as sample, add Normal Overlay mixing effect
) set lvdshim as mixing source
) go crazy on Normal Overlay p0 (slide the slider up and down a few times)
) crash (output pasted below)
Thanks!
--Xen
00:02:25.068356 I: Set parameter 0 to 42 on Entry 0 of Sample 1
00:02:25.068373 D: L 0 queue [ 1]/[3] [ 13] '362:0 0 0 20;'
00:02:25.068388 D: VIMS '362:0 0 0 20;' Set a parameter value
00:02:25.068402 D: VIMS: @76 Parse message '362:0 0 0 20;'
00:02:25.068423 I: Set parameter 0 to 20 on Entry 0 of Sample 1
00:02:25.068438 D: L 0 queue [ 2]/[3] [ 12] '362:0 0 0 1;'
00:02:25.068453 D: VIMS '362:0 0 0 1;' Set a parameter value
00:02:25.068519 D: VIMS: @76 Parse message '362:0 0 0 1;'
00:02:25.068541 I: Set parameter 0 to 1 on Entry 0 of Sample 1
00:02:25.071226 D: vevo_property_set: key=fps, type=2, n=1,
addr=0x7f65f70788c0
00:02:25.071337 D: vevo_property_set: key=rowstrides, type=1, n=4,
addr=0x7f65e6b9f610
00:02:25.071365 D: vevo_property_set: key=timecode, type=2, n=1,
addr=0x7f65f70788c8
00:02:25.071381 D: vevo_property_set: key=pixel_data, type=65, n=4,
addr=0x7f65e6b9f620
00:02:25.075855 D: Saved EDL 'sample_00001.edl' of sample 1
00:02:25.075899 D: Saved sample 1 EDL 'sample_00001.edl'
00:02:25.075960 E: No subtitle sequences present, nothing to save
00:02:25.076028 E: No subtitle sequences present, nothing to save
00:02:25.076068 E: No subtitle sequences present, nothing to save
00:02:25.076459 I: Saved EDL to file
/home/neo/.veejay/recovery/recovery_editlist_p21222.edl
00:02:25.076483 W: Saved samplelist to
/home/neo/.veejay/recovery/recovery_samplelist_p21222.sl
00:02:25.076496 W: Saved Editlist to
/home/neo/.veejay/recovery/recovery_editlist_p21222.edl
00:02:25.076511 W: Please report this error to
http://groups.google.com/group/veejay-discussion?hl=en 00:02:25.076524 W: Send at least veejay's output and include the
command(s) you have used to start it.
00:02:25.076537 W: Also, please consider sending in the recovery
files if any have been created.
00:02:25.076550 W: If you compiled it yourself, please include
information about your system.
00:02:25.076562 W: Dumping core file to: core.21222
gdb:
Program terminated with signal 8, Arithmetic exception.
#0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
width=720, height=480, val=0) at effects/bathroom.c:130
130 i = ((y*width) + (y % y_val) - (y_val>>1)) + x;
(gdb) backtrace
#0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
width=720, height=480, val=0) at effects/bathroom.c:130
#1 0x00007f65f6b38882 in bathroom_apply (frame=0x7f65e6b9f7c0,
width=720, height=480, mode=1, val=0) at effects/bathroom.c:164
#2 0x00007f65f6afa523 in vj_effman_apply_image_effect
(frames=0x7f65e6b9f740, frameinfo=0xdc5000, todo_info=0xdc6000,
arg=0x7f65e6b9f6a0, entry=88, e=140) at vj-effman.c:383
#3 0x00007f65f6afbdbc in vj_effect_apply (frames=0x7f65e6b9f740,
frameinfo=0xdc5000, kf=0xdc6000, selector=140,
arguments=0x7f65e6b9f6a0, ptr=0x0) at vj-effman.c:725
#4 0x00007f65f6ac2ef2 in vj_perform_apply_first (info=0xdbc000,
todo_info=0xdc6000, frames=0x7f65e6b9f740, frameinfo=0xdc5000, e=140,
c=0, n_frame=76, ptr=0x0) at vj-perform.c:1275
#5 0x00007f65f6ac5d0e in vj_perform_preprocess_secundary
(info=0xdbc000, id=1, mode=1, current_ssm=0, chain_entry=0,
frames=0x7f65e6b9f8c0, frameinfo=0xdc5000) at vj-perform.c:2298
#6 0x00007f65f6ac6036 in vj_perform_render_chain_entry
(info=0xdbc000, chain_entry=0) at vj-perform.c:2367
#7 0x00007f65f6ac6327 in vj_perform_sample_complete_buffers
(info=0xdbc000, hint444=0x7f65e6b9f994) at vj-perform.c:2456
#8 0x00007f65f6ac8e38 in vj_perform_queue_video_frame (info=0xdbc000,
skip_incr=0) at vj-perform.c:3579
#9 0x00007f65f6ae62fd in veejay_playback_cycle (info=0xdbc000) at
liblavplayvj.c:2693
#10 0x00007f65f6ae66bd in veejay_playback_thread (data=0xdbc000) at
liblavplayvj.c:2787
#11 0x00007f65f2d2de9a in start_thread (arg=0x7f65e6ba0700) at
pthread_create.c:308
#12 0x00007f65f67784bd in clone () at
../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#13 0x0000000000000000 in ?? ()
(gdb)
Thanks, that helped finding the problem; playback mode was ignored so the
parameter values for the effect chain on the mixing sample actually came
from someplace else! In other words, normal overlay parameters were feeded
into the bathroom effect
On Tue, Jul 3, 2012 at 6:24 AM, Xendarboh <xendar...@gmail.com> wrote:
> Hi Niels,
> > Could you send me the recovery files veejay saves in ~/.veejay/recovery,
> > it would help me reproduce this problem much quicker!
> > Also, in your error description, you speak of slots which in my world are
> > FX chain entries, the rest is just sample (or stream) so I have doubt
> > how to reproduce what you experienced :)
> I get the term "slot" from Reloaded. In the Samplebank it says stuff
> like "Slots 0 to 12" and basically a slot holds either a sample or
> stream.
> I'm attaching two recovery files that I got by doing this:
> ) start veejay with lvdshmin generator, activate Bathroom Mirror image FX
> on it
> ) load video file as sample, add Normal Overlay mixing effect
> ) set lvdshim as mixing source
> ) go crazy on Normal Overlay p0 (slide the slider up and down a few times)
> ) crash (output pasted below)
> Thanks!
> --Xen
> 00:02:25.068356 I: Set parameter 0 to 42 on Entry 0 of Sample 1
> 00:02:25.068373 D: L 0 queue [ 1]/[3] [ 13] '362:0 0 0 20;'
> 00:02:25.068388 D: VIMS '362:0 0 0 20;' Set a parameter value
> 00:02:25.068402 D: VIMS: @76 Parse message '362:0 0 0 20;'
> 00:02:25.068423 I: Set parameter 0 to 20 on Entry 0 of Sample 1
> 00:02:25.068438 D: L 0 queue [ 2]/[3] [ 12] '362:0 0 0 1;'
> 00:02:25.068453 D: VIMS '362:0 0 0 1;' Set a parameter value
> 00:02:25.068519 D: VIMS: @76 Parse message '362:0 0 0 1;'
> 00:02:25.068541 I: Set parameter 0 to 1 on Entry 0 of Sample 1
> 00:02:25.071226 D: vevo_property_set: key=fps, type=2, n=1,
> addr=0x7f65f70788c0
> 00:02:25.071337 D: vevo_property_set: key=rowstrides, type=1, n=4,
> addr=0x7f65e6b9f610
> 00:02:25.071365 D: vevo_property_set: key=timecode, type=2, n=1,
> addr=0x7f65f70788c8
> 00:02:25.071381 D: vevo_property_set: key=pixel_data, type=65, n=4,
> addr=0x7f65e6b9f620
> 00:02:25.075855 D: Saved EDL 'sample_00001.edl' of sample 1
> 00:02:25.075899 D: Saved sample 1 EDL 'sample_00001.edl'
> 00:02:25.075960 E: No subtitle sequences present, nothing to save
> 00:02:25.076028 E: No subtitle sequences present, nothing to save
> 00:02:25.076068 E: No subtitle sequences present, nothing to save
> 00:02:25.076459 I: Saved EDL to file
> /home/neo/.veejay/recovery/recovery_editlist_p21222.edl
> 00:02:25.076483 W: Saved samplelist to
> /home/neo/.veejay/recovery/recovery_samplelist_p21222.sl
> 00:02:25.076496 W: Saved Editlist to
> /home/neo/.veejay/recovery/recovery_editlist_p21222.edl
> 00:02:25.076511 W: Please report this error to
> http://groups.google.com/group/veejay-discussion?hl=en > 00:02:25.076524 W: Send at least veejay's output and include the
> command(s) you have used to start it.
> 00:02:25.076537 W: Also, please consider sending in the recovery
> files if any have been created.
> 00:02:25.076550 W: If you compiled it yourself, please include
> information about your system.
> 00:02:25.076562 W: Dumping core file to: core.21222
> gdb:
> Program terminated with signal 8, Arithmetic exception.
> #0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
> width=720, height=480, val=0) at effects/bathroom.c:130
> 130 i = ((y*width) + (y % y_val) - (y_val>>1)) + x;
> (gdb) backtrace
> #0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
> width=720, height=480, val=0) at effects/bathroom.c:130
> #1 0x00007f65f6b38882 in bathroom_apply (frame=0x7f65e6b9f7c0,
> width=720, height=480, mode=1, val=0) at effects/bathroom.c:164
> #2 0x00007f65f6afa523 in vj_effman_apply_image_effect
> (frames=0x7f65e6b9f740, frameinfo=0xdc5000, todo_info=0xdc6000,
> arg=0x7f65e6b9f6a0, entry=88, e=140) at vj-effman.c:383
> #3 0x00007f65f6afbdbc in vj_effect_apply (frames=0x7f65e6b9f740,
> frameinfo=0xdc5000, kf=0xdc6000, selector=140,
> arguments=0x7f65e6b9f6a0, ptr=0x0) at vj-effman.c:725
> #4 0x00007f65f6ac2ef2 in vj_perform_apply_first (info=0xdbc000,
> todo_info=0xdc6000, frames=0x7f65e6b9f740, frameinfo=0xdc5000, e=140,
> c=0, n_frame=76, ptr=0x0) at vj-perform.c:1275
> #5 0x00007f65f6ac5d0e in vj_perform_preprocess_secundary
> (info=0xdbc000, id=1, mode=1, current_ssm=0, chain_entry=0,
> frames=0x7f65e6b9f8c0, frameinfo=0xdc5000) at vj-perform.c:2298
> #6 0x00007f65f6ac6036 in vj_perform_render_chain_entry
> (info=0xdbc000, chain_entry=0) at vj-perform.c:2367
> #7 0x00007f65f6ac6327 in vj_perform_sample_complete_buffers
> (info=0xdbc000, hint444=0x7f65e6b9f994) at vj-perform.c:2456
> #8 0x00007f65f6ac8e38 in vj_perform_queue_video_frame (info=0xdbc000,
> skip_incr=0) at vj-perform.c:3579
> #9 0x00007f65f6ae62fd in veejay_playback_cycle (info=0xdbc000) at
> liblavplayvj.c:2693
> #10 0x00007f65f6ae66bd in veejay_playback_thread (data=0xdbc000) at
> liblavplayvj.c:2787
> #11 0x00007f65f2d2de9a in start_thread (arg=0x7f65e6ba0700) at
> pthread_create.c:308
> #12 0x00007f65f67784bd in clone () at
> ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
> #13 0x0000000000000000 in ?? ()
> (gdb)
> --
> You received this message because you are subscribed to the Google Groups
> "veejay-discussion" group.
> To post to this group, send email to veejay-discussion@googlegroups.com.
> To unsubscribe from this group, send email to
> veejay-discussion+unsubscribe@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/veejay-discussion?hl=en.
On Tue, Jul 3, 2012 at 2:57 PM, Niels Elburg <nwelb...@gmail.com> wrote:
> Hi Xen,
> Thanks, that helped finding the problem; playback mode was ignored so the
> parameter values for the effect chain on the mixing sample actually came
> from someplace else! In other words, normal overlay parameters were feeded
> into the bathroom effect
> I committed a fix for this!
> See you,
> Niels
> On Tue, Jul 3, 2012 at 6:24 AM, Xendarboh <xendar...@gmail.com> wrote:
>> Hi Niels,
>> > Could you send me the recovery files veejay saves in ~/.veejay/recovery,
>> > it would help me reproduce this problem much quicker!
>> > Also, in your error description, you speak of slots which in my world
>> > are
>> > FX chain entries, the rest is just sample (or stream) so I have doubt
>> > how to reproduce what you experienced :)
>> I get the term "slot" from Reloaded. In the Samplebank it says stuff
>> like "Slots 0 to 12" and basically a slot holds either a sample or
>> stream.
>> I'm attaching two recovery files that I got by doing this:
>> ) start veejay with lvdshmin generator, activate Bathroom Mirror image FX
>> on it
>> ) load video file as sample, add Normal Overlay mixing effect
>> ) set lvdshim as mixing source
>> ) go crazy on Normal Overlay p0 (slide the slider up and down a few times)
>> ) crash (output pasted below)
>> Thanks!
>> --Xen
>> 00:02:25.068356 I: Set parameter 0 to 42 on Entry 0 of Sample 1
>> 00:02:25.068373 D: L 0 queue [ 1]/[3] [ 13] '362:0 0 0 20;'
>> 00:02:25.068388 D: VIMS '362:0 0 0 20;' Set a parameter value
>> 00:02:25.068402 D: VIMS: @76 Parse message '362:0 0 0 20;'
>> 00:02:25.068423 I: Set parameter 0 to 20 on Entry 0 of Sample 1
>> 00:02:25.068438 D: L 0 queue [ 2]/[3] [ 12] '362:0 0 0 1;'
>> 00:02:25.068453 D: VIMS '362:0 0 0 1;' Set a parameter value
>> 00:02:25.068519 D: VIMS: @76 Parse message '362:0 0 0 1;'
>> 00:02:25.068541 I: Set parameter 0 to 1 on Entry 0 of Sample 1
>> 00:02:25.071226 D: vevo_property_set: key=fps, type=2, n=1,
>> addr=0x7f65f70788c0
>> 00:02:25.071337 D: vevo_property_set: key=rowstrides, type=1, n=4,
>> addr=0x7f65e6b9f610
>> 00:02:25.071365 D: vevo_property_set: key=timecode, type=2, n=1,
>> addr=0x7f65f70788c8
>> 00:02:25.071381 D: vevo_property_set: key=pixel_data, type=65, n=4,
>> addr=0x7f65e6b9f620
>> 00:02:25.075855 D: Saved EDL 'sample_00001.edl' of sample 1
>> 00:02:25.075899 D: Saved sample 1 EDL 'sample_00001.edl'
>> 00:02:25.075960 E: No subtitle sequences present, nothing to save
>> 00:02:25.076028 E: No subtitle sequences present, nothing to save
>> 00:02:25.076068 E: No subtitle sequences present, nothing to save
>> 00:02:25.076459 I: Saved EDL to file
>> /home/neo/.veejay/recovery/recovery_editlist_p21222.edl
>> 00:02:25.076483 W: Saved samplelist to
>> /home/neo/.veejay/recovery/recovery_samplelist_p21222.sl
>> 00:02:25.076496 W: Saved Editlist to
>> /home/neo/.veejay/recovery/recovery_editlist_p21222.edl
>> 00:02:25.076511 W: Please report this error to
>> http://groups.google.com/group/veejay-discussion?hl=en >> 00:02:25.076524 W: Send at least veejay's output and include the
>> command(s) you have used to start it.
>> 00:02:25.076537 W: Also, please consider sending in the recovery
>> files if any have been created.
>> 00:02:25.076550 W: If you compiled it yourself, please include
>> information about your system.
>> 00:02:25.076562 W: Dumping core file to: core.21222
>> gdb:
>> Program terminated with signal 8, Arithmetic exception.
>> #0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
>> width=720, height=480, val=0) at effects/bathroom.c:130
>> 130 i = ((y*width) + (y % y_val) - (y_val>>1)) + x;
>> (gdb) backtrace
>> #0 0x00007f65f6b386f1 in bathroom_hori_apply (frame=0x7f65e6b9f7c0,
>> width=720, height=480, val=0) at effects/bathroom.c:130
>> #1 0x00007f65f6b38882 in bathroom_apply (frame=0x7f65e6b9f7c0,
>> width=720, height=480, mode=1, val=0) at effects/bathroom.c:164
>> #2 0x00007f65f6afa523 in vj_effman_apply_image_effect
>> (frames=0x7f65e6b9f740, frameinfo=0xdc5000, todo_info=0xdc6000,
>> arg=0x7f65e6b9f6a0, entry=88, e=140) at vj-effman.c:383
>> #3 0x00007f65f6afbdbc in vj_effect_apply (frames=0x7f65e6b9f740,
>> frameinfo=0xdc5000, kf=0xdc6000, selector=140,
>> arguments=0x7f65e6b9f6a0, ptr=0x0) at vj-effman.c:725
>> #4 0x00007f65f6ac2ef2 in vj_perform_apply_first (info=0xdbc000,
>> todo_info=0xdc6000, frames=0x7f65e6b9f740, frameinfo=0xdc5000, e=140,
>> c=0, n_frame=76, ptr=0x0) at vj-perform.c:1275
>> #5 0x00007f65f6ac5d0e in vj_perform_preprocess_secundary
>> (info=0xdbc000, id=1, mode=1, current_ssm=0, chain_entry=0,
>> frames=0x7f65e6b9f8c0, frameinfo=0xdc5000) at vj-perform.c:2298
>> #6 0x00007f65f6ac6036 in vj_perform_render_chain_entry
>> (info=0xdbc000, chain_entry=0) at vj-perform.c:2367
>> #7 0x00007f65f6ac6327 in vj_perform_sample_complete_buffers
>> (info=0xdbc000, hint444=0x7f65e6b9f994) at vj-perform.c:2456
>> #8 0x00007f65f6ac8e38 in vj_perform_queue_video_frame (info=0xdbc000,
>> skip_incr=0) at vj-perform.c:3579
>> #9 0x00007f65f6ae62fd in veejay_playback_cycle (info=0xdbc000) at
>> liblavplayvj.c:2693
>> #10 0x00007f65f6ae66bd in veejay_playback_thread (data=0xdbc000) at
>> liblavplayvj.c:2787
>> #11 0x00007f65f2d2de9a in start_thread (arg=0x7f65e6ba0700) at
>> pthread_create.c:308
>> #12 0x00007f65f67784bd in clone () at
>> ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
>> #13 0x0000000000000000 in ?? ()
>> (gdb)
>> --
>> You received this message because you are subscribed to the Google Groups
>> "veejay-discussion" group.
>> To post to this group, send email to veejay-discussion@googlegroups.com.
>> To unsubscribe from this group, send email to
>> veejay-discussion+unsubscribe@googlegroups.com.
>> For more options, visit this group at
>> http://groups.google.com/group/veejay-discussion?hl=en.
> --
> You received this message because you are subscribed to the Google Groups
> "veejay-discussion" group.
> To post to this group, send email to veejay-discussion@googlegroups.com.
> To unsubscribe from this group, send email to
> veejay-discussion+unsubscribe@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/veejay-discussion?hl=en.