This is exactly what happens. I did include libdca for completeness, however ffmpeg has more advanced optimizations for ARMV6 and VFP (vector floating point processor) included, so it will be more performant.michalhejl wrote:The DTS plugin will by default play DTS with the LIBDCA codec, when you however delete the LIBDCA.SO file, it will start using the FFMPEG codec instead.
B Series DTS bugs & support
-
- Official SamyGO Developer
- Posts: 111
- Joined: Thu Jun 24, 2010 8:26 am
- Location: Frankfurt, Germany
Re: B Series DTS bugs & support
-
- Posts: 72
- Joined: Sat Nov 07, 2009 4:39 pm
Re: B Series DTS bugs & support
First, congratulations for this achievement, some people thought it was impossible (newagehun), with many efforts and skill smartsmurf did it! Impressive work, and impressive reverse engineering. Also thanks to doodlecz and others for brainstorming on this subject.
Now i have to say i did test the restricted firmware release with my B550 CI (T-CH5DEUC) and i cant make it work. I've copied all the needed file to /mtd_contents and i've tried with gdb trick and it's crashing. Also tried using injectso 1.2 and it says it was ok, but it's not doing anything. I can load chaninfo module using injectso though. I'll provide coredump when back home.
,Regards
Now i have to say i did test the restricted firmware release with my B550 CI (T-CH5DEUC) and i cant make it work. I've copied all the needed file to /mtd_contents and i've tried with gdb trick and it's crashing. Also tried using injectso 1.2 and it says it was ok, but it's not doing anything. I can load chaninfo module using injectso though. I'll provide coredump when back home.
,Regards
Samsung LE32B554 T-CH5DEUC 2006.0 with patched u-boot for usb boot with OpenEmbedded custom kernel with network supported and usb rootfs
-
- Official SamyGO Developer
- Posts: 111
- Joined: Thu Jun 24, 2010 8:26 am
- Location: Frankfurt, Germany
Re: B Series DTS bugs & support
(1) Do you load other modules aswell? If yes, which ones?nostromo1983 wrote:First, congratulations for this achievement, some people thought it was impossible (newagehun), with many efforts and skill smartsmurf did it! Impressive work, and impressive reverse engineering. Also thanks to doodlecz and others for brainstorming on this subject.
Now i have to say i did test the restricted firmware release with my B550 CI (T-CH5DEUC) and i cant make it work. I've copied all the needed file to /mtd_contents and i've tried with gdb trick and it's crashing. Also tried using injectso 1.2 and it says it was ok, but it's not doing anything. I can load chaninfo module using injectso though. I'll provide coredump when back home.
,Regards
(2) Can you provide the logfiles /dtv/samygodca.log and /dtv/samygodca_inj.log ?
Re: WITH DTS CODEC SERIE B NOT CHOOSE LANGUAGE
I answered your question. Do not create new topics for that.alexfr10 wrote:Hello ,
dts code works properly on my tv b7000 but we can choose the language .
LE40B653T5W,UE40D6750,UE65Q8C
Have questions? Read SamyGO Wiki, Search on forum first!
FFB (v0.8), FFB for CI+ . Get root on: C series, D series, E series, F series, H series. rooting K series, exeDSP/exeTV patches[C/D/E/F/H]
DO NOT EVER INSTALL FIRMWARE UPGRADE
Have questions? Read SamyGO Wiki, Search on forum first!
FFB (v0.8), FFB for CI+ . Get root on: C series, D series, E series, F series, H series. rooting K series, exeDSP/exeTV patches[C/D/E/F/H]
DO NOT EVER INSTALL FIRMWARE UPGRADE
-
- Posts: 72
- Joined: Sat Nov 07, 2009 4:39 pm
Re: B Series DTS bugs & support
I dont run any other modulesmartsmurf wrote:(1) Do you load other modules aswell? If yes, which ones?nostromo1983 wrote:First, congratulations for this achievement, some people thought it was impossible (newagehun), with many efforts and skill smartsmurf did it! Impressive work, and impressive reverse engineering. Also thanks to doodlecz and others for brainstorming on this subject.
Now i have to say i did test the restricted firmware release with my B550 CI (T-CH5DEUC) and i cant make it work. I've copied all the needed file to /mtd_contents and i've tried with gdb trick and it's crashing. Also tried using injectso 1.2 and it says it was ok, but it's not doing anything. I can load chaninfo module using injectso though. I'll provide coredump when back home.
,Regards
(2) Can you provide the logfiles /dtv/samygodca.log and /dtv/samygodca_inj.log ?
http://pastebin.com/HBLQKEqu
with injectso 1.2:
http://pastebin.com/J3sKZG02
# ./gdb --version
GNU gdb (GDB) 7.1
I dont understand also why sometime injection is working and sometimes not, see this:
http://pastebin.com/kS2WkMMn
Also can you explain why we cant see injected library into /proc/$(pidof exeDSP)/maps|grep .so , when using injectso, whereas i can see using gdb?
Samsung LE32B554 T-CH5DEUC 2006.0 with patched u-boot for usb boot with OpenEmbedded custom kernel with network supported and usb rootfs
-
- Official SamyGO Developer
- Posts: 111
- Joined: Thu Jun 24, 2010 8:26 am
- Location: Frankfurt, Germany
Re: B Series DTS bugs & support
nostromo1983:
You should not inject "samygodca_inj.so" and call "Game_Main", because it simply hasn't got that export. This will make gdb bail out.
Instead you should inject "samygodca.so" (which is the loader).
You should not inject "samygodca_inj.so" and call "Game_Main", because it simply hasn't got that export. This will make gdb bail out.
Instead you should inject "samygodca.so" (which is the loader).
-
- Posts: 72
- Joined: Sat Nov 07, 2009 4:39 pm
Re: B Series DTS bugs & support
My bad sorry for this mistake, i was confuse by the name of the file...just tested it using gdb trick and it's working!!!! working with injectso as well !! But i never see anything into loader.log, is it normal? At first i was feeling bad because exeDSP has crashed when loading a multimedia file but this has nothing to do with your patch but it's more a design problem from the B5xx series, a known bug. Sometime when loading a file with the mediaplayer it's crashing, even with no additional module, but this is another story.
Right now i'm watching a file with DTS soundtrack, it's really amazing how good it works, i cant believe it's real! It's really a damn good job. As previously reported the first 25-30 seconds of the movie is lagging but we can hear some sound but not clearly. I haven't tested with audio switch yet. Cpu time is oscillating between 60 and 80% CPU using default configuration which is using libdca.so. Load average: 5.36 6.31 4.84 (Status: S=sleeping R=running, W=waiting). With ffmepg directly (which is using another dts lib?) CPU usage is much better as previously said, it's around 40-50% CPU.
I observed that when lagging in the first 30 seconds of the file top is showing this:
I had 2 case of files with soundtrack that are lagging this way even after 30 s of playback (using libdca and ffmpeg is same), and seems like until the end of the file, i dont know why, i will analyze those files.
As you can see usb-storage and softirq are pretty high and there's after kswapd, this seems to says the interrupt rate due to the usb controller is too high and it's in conflict with exeDSP code which produces the laggy sound . Also you can see the free memory is pretty low ( 1696K free), maybe it's kinda swaping? or maybe kswapd is responsible of context switch swaping, i dont know well how system process are working in linux.
Also i've noticed i dont see output log with serial cable when exeDSP is patched, is it normal?
Thank you
Right now i'm watching a file with DTS soundtrack, it's really amazing how good it works, i cant believe it's real! It's really a damn good job. As previously reported the first 25-30 seconds of the movie is lagging but we can hear some sound but not clearly. I haven't tested with audio switch yet. Cpu time is oscillating between 60 and 80% CPU using default configuration which is using libdca.so. Load average: 5.36 6.31 4.84 (Status: S=sleeping R=running, W=waiting). With ffmepg directly (which is using another dts lib?) CPU usage is much better as previously said, it's around 40-50% CPU.
I observed that when lagging in the first 30 seconds of the file top is showing this:
Code: Select all
Mem: 109332K used, 1696K free, 0K shrd, 1272K buff, 60428K cached
Load average: 11.26 7.43 5.34 (Status: S=sleeping R=running, W=waiting)
PID USER STATUS RSS PPID %CPU %MEM COMMAND
45 root S N 47624 44 67.9 42.8 exeDSP
242 root DW< 0 13 16.8 0.0 usb-storage
7 root SW 0 1 5.8 0.0 softirq-block/0
486 root R N 724 339 2.1 0.6 top
19 root SW< 0 13 0.1 0.0 kswapd0
339 root S N 472 338 0.0 0.4 sh
31 root S 356 25 0.0 0.3 rc.local
25 root S 344 24 0.0 0.3 rc.local
44 root S 292 31 0.0 0.2 rc.local
24 root S 284 23 0.0 0.2 rcS
338 root S N 276 45 0.0 0.2 sh
1 root S 264 0 0.0 0.2 init
23 root S 108 1 0.0 0.0 init
20 root SW< 0 13 0.0 0.0 kswapd1
56 root SWN 0 1 0.0 0.0 aeMsgTask0
14 root SW< 0 13 0.0 0.0 kblockd/0
4 root SW 0 1 0.0 0.0 softirq-timer/0
8 root SW 0 1 0.0 0.0 softirq-tasklet
11 root SW< 0 1 0.0 0.0 events/0
16 root SW 0 1 0.0 0.0 kdtvlogd
18 root SW 0 13 0.0 0.0 pdflush
As you can see usb-storage and softirq are pretty high and there's after kswapd, this seems to says the interrupt rate due to the usb controller is too high and it's in conflict with exeDSP code which produces the laggy sound . Also you can see the free memory is pretty low ( 1696K free), maybe it's kinda swaping? or maybe kswapd is responsible of context switch swaping, i dont know well how system process are working in linux.
Also i've noticed i dont see output log with serial cable when exeDSP is patched, is it normal?
Thank you
Samsung LE32B554 T-CH5DEUC 2006.0 with patched u-boot for usb boot with OpenEmbedded custom kernel with network supported and usb rootfs