Hmm, perhaps I forgot about the ld cache? I just uninstalled my SDL libs and ran ldconfig and now my binary runs as well. So definitely SDL issue. Question remains though
I tried your build which works. If I use my binary with your libs, it fails (black screen) If I use your binary with my SDL libs, it also fails. Weird. It's definitely an issue with SDL but our binaries differ as well which shouldn't IMHO: md5sums: 9aad59aadd47aeb9ecc84aa0470797eb mine 827118d2fc53f...
Still no luck getting uae4all to compile to a working condition. I'm seeing some very weird SDL behavior. (Needing interrupt before displaying, no screen on full hd). I'm using the updated Makefile. @rSI Have you done 'make clean' removed all the object files lately? I just want to find the source o...
Just compiled with sdl_sound_new and I don't get the dsp error anymore. However it still crashes: $ sudo ./uae4all adfdir loaded. currentDir=/home/pi/src/uae4all-rpi/roms/ Dispmanx: Opening display 0 Dispmanx: Physical video mode is 1280 x 720 Using internal program mode: 768 x 272 16 bpp Using phys...
K, I had some build issues with your guichan repo so I looked at what files were changed and made those changes to the original guichan. I send you an issue but I might push the original guichan source to new repo and commit changes to that. I can run uae4all now however it crashes on me. Might be s...
You need an account to access those files. Do you have a copy? I'd like to get the patch for the original source.
I also tried the latest release of guichan and the latest in git. Both to no avail.
K I just compiled uae4all-rpi from scratch using the following approach: ## clone the git repo: git clone https://github.com/sphaero/uae4all-rpi.git cd uae4all-rpi git init submodules git update submodules ## Build SDL (dispmanx) cd SDL12-kms-dispmanx ./MAC_ConfigureDISPMANX.sh make && sudo make ins...
I'm in the process of buidling from scratch. When using Raspbian supplied libs (sdl, guichan) in X I can see images in the UI which I can't see when using the prebuild version of RSi on the console(without X). This could suggest an issue with SDL_image and could also be the cause of the missing mous...
@rSI, I've created a pull request for fixing compilation on the RPi. As it is not good custom to merge your own PR can you do it?
You should have access.
Although the compile is still running I think it's OK.
@exobuzz: Indeed development is very fragmented and not using repositories anywhere which makes tracking changes almost impossible I've made some issues: No command line options: https://github.com/sphaero/uae4all-rpi/issues/2 Remove pandora specifics from ui: https://github.com/sphaero/uae4all-rpi/...
I know but when I asked him he stated that it's only the android port of uae. https://github.com/lubomyr/uae4all2/issues/1 So I started searching the forums stumblin on many different versions of uae4all. Only the original works fine on x86 but the later developments seems to have undone that. Btw h...
I'm still working my way through the makefile changing it for cross compilation. I've almost got it compiled! Would be nice if you could send your diff as a pull request. If you're not familiar with it: 1: create a github account 2: clone the repo I created 3: on your own machine install git and do:...
K, at least I know the makefile is made for older versions of gcc. Now I need to get beyond memory alignment errors: In file included from src/include/memory-uae.h:36:0, from src/audio.cpp:21: src/include/maccess.h: In function ‘uae_u32 do_get_mem_long(uae_u32*)’: src/include/maccess.h:25:46: error:...
Hey RSi, I've been going through the sources and have put them on github: https://github.com/sphaero/uae4all-rpi The makefiles are quite a mess. :) How did you build your version? On the RPI or using a cross compiler? I was hoping to build this for x86 or x86_64 architecture so it would be easier to...
Did you ever get more success with gst1.4 and the omx plugins? I'm testing 1.4.4 with omx from git latest but not having a lot success yet. Creating a deb should be trivial, I haven't done it myself yet for gstreamer but the debian package in unstable is just the latest stable release. You should ju...
Try the following: gst-launch-1.0 -v videotestsrc ! autovideosink This should give you a test video. If it doesn't try: GST_DEBUG=4 gst-launch-1.0 -v videotestsrc ! autovideosink You'll see more debug output. If you're running Xorg I think it should work with this: gst-launch-1.0 -v videotestsrc ! x...
Try first with a test source something like: gst-launch-1.0 videotestsrc ! \ jpegenc ! \ rtpjpegpay ! \ udpsink host=127.0.0.1 port=5200 And on the receiver: gst-launch-1.0 udpsrc port=5200 ! \ rtpjpegdepay ! \ jpegdec ! \ autovideosink See if that does anything. If you are new gstreamer or streamin...
Yes it was a clean install. Also it makes sense that displaying through GL is a higher overhead than displaying directly like the hello_video does. Just curious what best performance can be achieved and where the bottleneck would be. Btw I need to do (add queue) decodebin ! queue ! eglglessink to ma...
Thanks for the pointer. I missed some sources indeed.
Just tested and it is much smoother although gstreamer takes quite some overhead. 50% cpu compared to 3-7%. Anybody got lower cpu burden?