Re: [PATCH V2] RFC: ARM: DaVinci: ASoc use iram to buffer sound
Troy Kisky wrote: Yusuf Caglar AKYUZ wrote: Troy Kisky wrote: Use the sram(iram) to avoid underrun on audio. I will clean this up after someone says it works for them. Hi, First of all, thanks for looking into this. This patch makes audio usable again for me on 2.6.30 kernel and overruns are gone. What I want to ask is, there is constant noise on the output on data rates higher than 8000 khz. Anything playing with a higher rate generates random artifacts on the output. I have no idea where to look or how to test further. Do you have any pointers? Thanks, Caglar I don't have any suggestions. But if you post the codec and processor that you're using, maybe someone can verify that they see the same thing. Also, if you make a recording of the real output, and what your trying to play available, I'll take a listen to it. Could you tell if these artifacts were present before this patch? Troy Aaah! There were some problem with my player, today I tested with aplay and everything is fine. It would be nice to have this patch included in mainline. Thanks again, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH V2] RFC: ARM: DaVinci: ASoc use iram to buffer sound
Steve Chen wrote: On Tue, 2009-06-30 at 18:03 +0300, Yusuf Caglar AKYUZ wrote: Aaah! There were some problem with my player, today I tested with aplay and everything is fine. It would be nice to have this patch included in mainline. Can you provide some details on which player and the sample rate you used. A few months back, I noticed background noise with wavplay when playing at sample rates that are not supported by hardware. The issue was traced to ALSA OSS sample rate conversion code (commit id 5370d96f85962769ea3df3a81cc885f257c51589). I wonder if there are scenarios the code can still fail. Regards, Steve In the beginning, I was using gstreamer(gst-launch). My pipelines were working on my PC so I thought they were ok. Today I converted some mpeg files to wav files and then tried aplay directly. For example: $gst-launch audiotestsrc ! audio/x-raw-int,endianness=1234,rate=44100,channels=1 ! alsasink $gst-launch audiotestsrc ! audio/x-raw-int,endianness=1234,rate=48000,channels=1 ! alsasink I can say that 1st one is not working whereas 2nd one is ok. Another example is playing and mp2 file: $gst-launch filesrc location=../davincieffect.mp2 ! mad ! audio/x-raw-int,endianness=1234,signed=true,rate=44100,channels=2 ! alsasink $gst-launch filesrc location=../davincieffect48.mp2 ! mad ! audio/x-raw-int,endianness=1234,signed=true,rate=48000,channels=2 ! alsasink Again 1st one is not working but 2nd is ok. In this example davincieffect.mp2 is 44100 khz file whereas davincieffect48.mp2 is 48 khz. I tried these 2 pipelines with TI mpeg codecs as well and results were the same. So I can say that there is nothing related to 'mad' mpeg codec. From these two examples it was clear(at least for me) that there were some problem with driver, especially at 44100 khz either 1 channel or 2 channel. Later I converted davincieffect.mp2 to wav at two different sample rates and played with aplay as such: $aplay r441_c2.wav $aplay r48_c2.wav Funny thing is that both played well. 441_c2 is 44100 2-ch and r48_c2 is 48000 2-ch. I can conduct further tests if you have any suggestions since I really want to see the bottom of this problem. Regards, Caglar P.S. Tests are performed on dm6446 EVM and on a dm6446 custom board. Both have the tlv320aic33 as the codec. Custom board has a different codec clock than the EVM, 27 mhz. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: overrun occuring in arecord for 2.6.30
Mani, Arun wrote: Hi Caglar, Have you tried a loopback of aplay and arecord with this fix? I am still getting overruns with this fix. I tried both following: $arecord -f cd | aplay -f cd $arecord | aplay I receive only one *underrun*, at the time of beginning and then no more. Am I missing something? Which hardware are we talking about? I'm using dm6446, I haven't tested dm355 yet. Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: overrun occuring in arecord for 2.6.30
Mani, Arun wrote: I am using DM355 and I am getting echo most of the time. My command is arecord -r 48000 -f S16_LE -d 10 -v -c 1 | aplay -r 48000 -f S16_LE -d 10 -v -c 1 I tried this and I can confirm no problems here. Is it possible that you are hitting the problem at [1] ? :) And I guess this patch also works on dm355 by looking at [2] I will try this one on dm355 as well tomorrow. Regards, Caglar [1] http://www.mail-archive.com/davinci-linux-open-source@linux.davincidsp.com/msg12525.html [2] http://www.mail-archive.com/davinci-linux-open-source@linux.davincidsp.com/msg12570.html Thanks, Arun -Original Message- From: Yusuf Caglar AKYUZ [mailto:caglarak...@gmail.com] Sent: Tuesday, June 30, 2009 1:03 PM To: Mani, Arun Cc: davinci-linux-open-source@linux.davincidsp.com Subject: Re: overrun occuring in arecord for 2.6.30 Mani, Arun wrote: Hi Caglar, Have you tried a loopback of aplay and arecord with this fix? I am still getting overruns with this fix. I tried both following: $arecord -f cd | aplay -f cd $arecord | aplay I receive only one *underrun*, at the time of beginning and then no more. Am I missing something? Which hardware are we talking about? I'm using dm6446, I haven't tested dm355 yet. Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH V2] RFC: ARM: DaVinci: ASoc use iram to buffer sound
Steve Chen wrote: I believe GStreamer uses the OSS emulation layer. It appears that the problem is at sample rate of 44.1k. I disabled CONFIG_SND_SUPPORT_OLD_API and tried with that. Results are the same. Is there a way to disable OSS emulation layer, or should I check gstreamer handling of alsa? From these two examples it was clear(at least for me) that there were some problem with driver, especially at 44100 khz either 1 channel or 2 channel. Later I converted davincieffect.mp2 to wav at two different sample rates and played with aplay as such: $aplay r441_c2.wav Can you try aplay -D hw:0,1 r441_c2.wav If you see something like Warning: rate is not accurate (requested = 44100Hz, got = 48000Hz) please, try the plug plugin That means McBSP/AIC33 are not setup to handle sample rate of 44.1K directly. The sample rates are covered by software. The problem is likely in the area of rate conversion. If you don't see the warning, McBSP/AIC33 is setup to handle 44.1K directly. You man want to check with clock settings (make sure AIC33 is setup as the master). I tried and there was no warning: $aplay -v -D hw:0,0 r441_c2.wav Playing WAVE 'r441_c2.wav' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo Hardware PCM card 0 'DaVinci Beginning' device 0 subdevice 0 Its setup is: stream : PLAYBACK access : RW_INTERLEAVED format : S16_LE subformat: STD channels : 2 rate : 44100 exact rate : 44100 (44100/1) msbits : 16 buffer_size : 22400 period_size : 896 period_time : 20317 tstamp_mode : NONE period_step : 1 avail_min: 896 period_event : 0 start_threshold : 22400 stop_threshold : 22400 silence_threshold: 0 silence_size : 0 boundary : 1468006400 appl_ptr : 0 hw_ptr : 0 I started seeing this issue after applying this patch, maybe there is something wrong with ping-pong buffer handling as this problem is present only when sample rate is 44.1 khz. Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH V2] RFC: ARM: DaVinci: ASoc use iram to buffer sound
David Brownell wrote: On Tuesday 30 June 2009, Steve Chen wrote: Can you try aplay -D hw:0,1 r441_c2.wav If you see something like Warning: rate is not accurate (requested = 44100Hz, got = 48000Hz) please, try the plug plugin That means McBSP/AIC33 are not setup to handle sample rate of 44.1K directly. The sample rates are covered by software. The problem is likely in the area of rate conversion. And ... seeking someone who has a bit of spare time, a DM6446 EVM, and interest in solving it: that EVM can switch audio clocking at runtime, among six different sample rates (of which those are two). So it would be interesting to get that working. :) Do you mean controlling PLL1705 with i2c expander u18? I always wonder why this is necessary for evm. Isn't AIC33 able to create common audio frequencies internally from a reasonable input clock, say 27 Mhz? Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dvsdk 2.0 fails to build.
Andrea Gasparini wrote: Hi, I'm trying to build DVSK 2_00_00_22, and I'm failing in it. More precisely, it's dmai make that blame with the following line: /mnt/lavoro/gigavision/mv_pro_5.0/montavista/pro/devkit/arm/v5t_le/bin/../lib/gcc/armv5tl- montavista-linux-gnueabi/4.2.0/../../../../armv5tl-montavista-linux- gnueabi/bin/ld: cannot find /db/rtree/niclas/davinci_multimedia_application_interface/dmai/packages/ti/sdo/dmai/apps/video_loopback_copy/linux/video_loopback_copy_dm355_config/package/cfg/video_loopback_copy_dm355_x470MV.o470MV and of course I don't have /db/rtree/niclas/ path. A quick search tells me that all that kind of file: ./apps/image_decode_io1/linux/image_decode_io1_dm355_config/package.bld ./apps/image_encode_io1/linux/image_encode_io1_dm355_config/package.bld ./apps/speech_decode_io1/linux/speech_decode_io1_dm355_config/package.bld ./apps/video_decode_io2/linux/video_decode_io2_dm355_config/package.bld ./apps/video_encode_io1/linux/video_encode_io1_dm355_config/package.bld ./apps/video_encode_io_multich1/linux/video_encode_io_multich1_dm355_config/package.bld ./apps/video_loopback_copy/linux/video_loopback_copy_dm355_config/package.bld ./apps/video_loopback_resize/linux/video_loopback_resize_dm355_config/package.bld have hardcoded the wrong path, here it is a snap of one of them: var exeOpts = new Executable.Attrs({ cfgScript: '/db/rtree/niclas/davinci_multimedia_application_interface/dmai/packages/ti/sdo/dmai/apps/video_loopback_copy/linux/video_loopback_copy_dm355.cfg', profile: 'release', cfgHome: 'video_loopback_copy_dm355_config', }); Ok, the question now is: am I doing something wrong or really are all package.bld files bugged? Is there perhaps a way to regenerate all package.bld ? ( dmai version is 1_20_00_06 ) thanks. AFAIK, package.bld files are generated automatically. You need to adjust makefiles for correct toolchains. I'm using OpenEmbedded for DMAI builds. It has most of the patches necessary for building DMAI and gstreamer-ti. I'm attaching most related patches here for your reference. 1st one is from OE, 2nd one is mine local patch for dm6446 builds. Regards, Caglar __ Index: dmai_1_20_00_06/packages/config.bld === --- dmai_1_20_00_06.orig/packages/config.bld2009-02-11 19:38:51.0 -0600 +++ dmai_1_20_00_06/packages/config.bld 2009-02-11 19:39:23.0 -0600 @@ -42,7 +42,7 @@ /* location of the Codec Sourcery Arm9 tools */ var GCArmv5T = xdc.useModule('gnu.targets.arm.GCArmv5T'); -GCArmv5T.LONGNAME = 'bin/arm-none-linux-gnueabi-gcc'; +GCArmv5T.LONGNAME = 'bin/arm-angstrom-linux-gnueabi-gcc'; GCArmv5T.platform = ti.platforms.evm3530; GCArmv5T.rootDir = java.lang.System.getenv(CSTOOL_DIR); Index: dmai_1_20_00_06/packages/ti/sdo/dmai/Makefile === --- dmai_1_20_00_06.orig/packages/ti/sdo/dmai/Makefile 2009-02-11 19:39:57.0 -0600 +++ dmai_1_20_00_06/packages/ti/sdo/dmai/Makefile 2009-02-11 19:58:05.0 -0600 @@ -33,7 +33,7 @@ DMAI_INSTALL_DIR = ../../../.. TARGET = dmai -include $(DMAI_INSTALL_DIR)/Rules.make +#include $(DMAI_INSTALL_DIR)/Rules.make # Should the full command be echoed to the console during build? VERBOSE=false @@ -64,7 +64,7 @@ MVL_CPP_FLAGS = $(GNU_CPP_FLAGS) -Dxdc_target_name__=MVArm9 CS_CPP_FLAGS = $(GNU_CPP_FLAGS) -Dxdc_target_name__=codesourcery/GCArmv5T -GNU_C_FLAGS= $(C_FLAGS) -Wall -Werror +GNU_C_FLAGS= $(C_FLAGS) -Wall C64P_C_FLAGS = $(C_FLAGS) GNU_AR_FLAGS = $(AR_FLAGS) Index: dmai_1_20_00_06/packages/ti/sdo/dmai/apps/Makefile.app === --- dmai_1_20_00_06.orig/packages/ti/sdo/dmai/apps/Makefile.app 2009-02-11 19:58:44.0 -0600 +++ dmai_1_20_00_06/packages/ti/sdo/dmai/apps/Makefile.app 2009-02-11 19:59:01.0 -0600 @@ -34,7 +34,7 @@ ifndef DMAI_INSTALL_DIR DMAI_INSTALL_DIR = ../../../../../.. endif -include $(DMAI_INSTALL_DIR)/Rules.make +#include $(DMAI_INSTALL_DIR)/Rules.make # Should the full command be echoed to the console during build? VERBOSE=false __ BRIJESH_GIT_022309/davinci_multimedia_application_interface/dmai/packages/ti/sdo/dmai/apps/Makefile.app 2009-06-15 21:12:43.0 +0300 +++ BRIJESH_GIT_022309/davinci_multimedia_application_interface/dmai/packages/ti/sdo/dmai/apps/Makefile.app.mine 2009-06-15 21:49:00.0 +0300 @@ -359,7 +359,7 @@ @echo Building $(DM6446_AL_TARGET) @echo Configuring application using $ @echo - $(PRE) XDCPATH=$(XDC_PATH) $(CONFIGURO) -c $(MVTOOL_DIR) -o $(DM6446_AL_XDC_CFG) -t $(MVL_XDC_TARGET) -p
Re: dvsdk 2.0 fails to build.
Andrea Gasparini wrote: Andrea Gasparini spiffera, alle Monday 29 June 2009 circa: Anyway, the question remains valid, DO I REALLY NEED to modify almost ten config.bld to have the right path, or is there another, simpler way? ok, I was too hurried in saying so. It seems it's enough to modifiy: dvsdk_2_00_00_22/dmai_1_20_00_06/packages/config.bld :) The only strange path it search for is CODEGEN_INSTALL_DIR... but for dm355 is not needed, as far as I know... am I wrong?! Right. Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: overrun occuring in arecord for 2.6.30
Mani, Arun wrote: Hi All, Anyone tried arecord in 2.6.30. I am constantly getting a overrun error which is blocking the capture. Thanks, Arun I tried it, and it is not working for me. I see constant overruns as well. This code was working quite well on 2.6.28, I have no idea what happened since then. Anyway patches posted by Troy Kisky at [1] seems to fix overrun issues. This version of patch does not apply to 2.6.30, you can find the attached patched I modified for applying to 2.6.30. Arago tree has further modifications, if you want I can send this patch against arago tree. Regards, Caglar [1] http://www.mail-archive.com/davinci-linux-open-source@linux.davincidsp.com/msg12568.html diff --git a/sound/soc/davinci/davinci-i2s.c b/sound/soc/davinci/davinci-i2s.c index b1ea52f..70e8706 100644 --- a/sound/soc/davinci/davinci-i2s.c +++ b/sound/soc/davinci/davinci-i2s.c @@ -63,6 +63,7 @@ #define DAVINCI_MCBSP_RCR_RWDLEN1(v) ((v) 5) #define DAVINCI_MCBSP_RCR_RFRLEN1(v) ((v) 8) #define DAVINCI_MCBSP_RCR_RDATDLY(v) ((v) 16) +#define DAVINCI_MCBSP_RCR_RFIG (1 18) #define DAVINCI_MCBSP_RCR_RWDLEN2(v) ((v) 21) #define DAVINCI_MCBSP_XCR_XWDLEN1(v) ((v) 5) @@ -85,14 +86,6 @@ #define DAVINCI_MCBSP_PCR_FSRM (1 10) #define DAVINCI_MCBSP_PCR_FSXM (1 11) -#define MOD_REG_BIT(val, mask, set) do { \ - if (set) { \ - val |= mask; \ - } else { \ - val = ~mask; \ - } \ -} while (0) - enum { DAVINCI_MCBSP_WORD_8 = 0, DAVINCI_MCBSP_WORD_12, @@ -112,8 +105,13 @@ static struct davinci_pcm_dma_params davinci_i2s_pcm_in = { struct davinci_mcbsp_dev { void __iomem*base; +#define MOD_DSP_A 0 +#define MOD_DSP_B 1 + int mode; + u32 pcr; struct clk *clk; struct davinci_pcm_dma_params *dma_params[2]; + struct snd_soc_dai *codec_dai; }; static inline void davinci_mcbsp_write_reg(struct davinci_mcbsp_dev *dev, @@ -127,96 +125,100 @@ static inline u32 davinci_mcbsp_read_reg(struct davinci_mcbsp_dev *dev, int reg) return __raw_readl(dev-base + reg); } -static void davinci_mcbsp_start(struct snd_pcm_substream *substream) +static void toggle_clock(struct davinci_mcbsp_dev *dev, int playback) +{ + u32 m = playback ? DAVINCI_MCBSP_PCR_CLKXP : DAVINCI_MCBSP_PCR_CLKRP; + /* The clock needs to toggle to complete reset. +* So, fake it by toggling the clk polarity. +*/ + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_PCR_REG, dev-pcr ^ m); + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_PCR_REG, dev-pcr); +} + +static void davinci_mcbsp_start(struct davinci_mcbsp_dev *dev, + struct snd_pcm_substream *substream) { struct snd_soc_pcm_runtime *rtd = substream-private_data; - struct davinci_mcbsp_dev *dev = rtd-dai-cpu_dai-private_data; struct snd_soc_device *socdev = rtd-socdev; struct snd_soc_platform *platform = socdev-card-platform; - u32 w; - int ret; - - /* Start the sample generator and enable transmitter/receiver */ - w = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); - MOD_REG_BIT(w, DAVINCI_MCBSP_SPCR_GRST, 1); - davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, w); + int playback = (substream-stream == SNDRV_PCM_STREAM_PLAYBACK); + u32 spcr; + u32 mask = playback ? DAVINCI_MCBSP_SPCR_XRST : DAVINCI_MCBSP_SPCR_RRST; + spcr = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); + if (spcr mask) { + /* start off disabled */ + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, + spcr ~mask); + toggle_clock(dev, playback); + } + if (dev-pcr (DAVINCI_MCBSP_PCR_FSXM | DAVINCI_MCBSP_PCR_FSRM | + DAVINCI_MCBSP_PCR_CLKXM | DAVINCI_MCBSP_PCR_CLKRM)) { + /* Start the sample generator */ + spcr |= DAVINCI_MCBSP_SPCR_GRST; + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, spcr); + } - if (substream-stream == SNDRV_PCM_STREAM_PLAYBACK) { + if (playback) { /* Stop the DMA to avoid data loss */ /* while the transmitter is out of reset to handle XSYNCERR */ if (platform-pcm_ops-trigger) { - ret = platform-pcm_ops-trigger(substream, + int ret = platform-pcm_ops-trigger(substream, SNDRV_PCM_TRIGGER_STOP); if (ret 0) printk(KERN_DEBUG Playback DMA stop failed\n); } /* Enable the transmitter */ - w =
Re: overrun occuring in arecord for 2.6.30
Mani, Arun wrote: Caglar, Thanks very much. Can I get the patch against Arago tree too. That will be of great help. Thanks, Arun. Patch is attached inline. It is against Arago next branch. Regards, Caglar diff --git a/sound/soc/davinci/davinci-i2s.c b/sound/soc/davinci/davinci-i2s.c index 500d2f5..aa7753c 100644 --- a/sound/soc/davinci/davinci-i2s.c +++ b/sound/soc/davinci/davinci-i2s.c @@ -63,6 +63,7 @@ #define DAVINCI_MCBSP_RCR_RWDLEN1(v) ((v) 5) #define DAVINCI_MCBSP_RCR_RFRLEN1(v) ((v) 8) #define DAVINCI_MCBSP_RCR_RDATDLY(v) ((v) 16) +#define DAVINCI_MCBSP_RCR_RFIG (1 18) #define DAVINCI_MCBSP_RCR_RWDLEN2(v) ((v) 21) #define DAVINCI_MCBSP_XCR_XWDLEN1(v) ((v) 5) @@ -85,14 +86,6 @@ #define DAVINCI_MCBSP_PCR_FSRM (1 10) #define DAVINCI_MCBSP_PCR_FSXM (1 11) -#define MOD_REG_BIT(val, mask, set) do { \ - if (set) { \ - val |= mask; \ - } else { \ - val = ~mask; \ - } \ -} while (0) - enum { DAVINCI_MCBSP_WORD_8 = 0, DAVINCI_MCBSP_WORD_12, @@ -112,8 +105,13 @@ static struct davinci_pcm_dma_params davinci_i2s_pcm_in = { struct davinci_mcbsp_dev { void __iomem*base; +#define MOD_DSP_A 0 +#define MOD_DSP_B 1 + int mode; + u32 pcr; struct clk *clk; struct davinci_pcm_dma_params *dma_params[2]; + struct snd_soc_dai *codec_dai; }; static inline void davinci_mcbsp_write_reg(struct davinci_mcbsp_dev *dev, @@ -127,94 +125,99 @@ static inline u32 davinci_mcbsp_read_reg(struct davinci_mcbsp_dev *dev, int reg) return __raw_readl(dev-base + reg); } -static void davinci_mcbsp_start(struct snd_pcm_substream *substream) +static void toggle_clock(struct davinci_mcbsp_dev *dev, int playback) { - struct snd_soc_pcm_runtime *rtd = substream-private_data; - struct davinci_mcbsp_dev *dev = rtd-dai-cpu_dai-private_data; - struct snd_soc_device *socdev = rtd-socdev; - struct snd_soc_platform *platform = socdev-card-platform; - u32 w; - int ret; + u32 m = playback ? DAVINCI_MCBSP_PCR_CLKXP : DAVINCI_MCBSP_PCR_CLKRP; + /* The clock needs to toggle to complete reset. +* So, fake it by toggling the clk polarity. +*/ + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_PCR_REG, dev-pcr ^ m); + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_PCR_REG, dev-pcr); +} - /* Start the sample generator and enable transmitter/receiver */ - w = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); - MOD_REG_BIT(w, DAVINCI_MCBSP_SPCR_GRST, 1); - davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, w); +static void davinci_mcbsp_start(struct davinci_mcbsp_dev *dev, + struct snd_pcm_substream *substream) +{ + struct snd_soc_pcm_runtime *rtd = substream-private_data; + struct snd_soc_device *socdev = rtd-socdev; + struct snd_soc_platform *platform = socdev-card-platform; + int playback = (substream-stream == SNDRV_PCM_STREAM_PLAYBACK); + u32 spcr; + u32 mask = playback ? DAVINCI_MCBSP_SPCR_XRST : DAVINCI_MCBSP_SPCR_RRST; + spcr = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); + if (spcr mask) { + /* start off disabled */ + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, + spcr ~mask); + toggle_clock(dev, playback); + } + if (dev-pcr (DAVINCI_MCBSP_PCR_FSXM | DAVINCI_MCBSP_PCR_FSRM | + DAVINCI_MCBSP_PCR_CLKXM | DAVINCI_MCBSP_PCR_CLKRM)) { + /* Start the sample generator */ + spcr |= DAVINCI_MCBSP_SPCR_GRST; + davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, spcr); + } - if (substream-stream == SNDRV_PCM_STREAM_PLAYBACK) { + if (playback) { /* Stop the DMA to avoid data loss */ /* while the transmitter is out of reset to handle XSYNCERR */ if (platform-pcm_ops-trigger) { - ret = platform-pcm_ops-trigger(substream, + int ret = platform-pcm_ops-trigger(substream, SNDRV_PCM_TRIGGER_STOP); if (ret 0) printk(KERN_DEBUG Playback DMA stop failed\n); } /* Enable the transmitter */ - w = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); - MOD_REG_BIT(w, DAVINCI_MCBSP_SPCR_XRST, 1); - davinci_mcbsp_write_reg(dev, DAVINCI_MCBSP_SPCR_REG, w); + spcr = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG); + spcr |= DAVINCI_MCBSP_SPCR_XRST; +
Re: [PATCH V2] RFC: ARM: DaVinci: ASoc use iram to buffer sound
Troy Kisky wrote: Use the sram(iram) to avoid underrun on audio. I will clean this up after someone says it works for them. Hi, First of all, thanks for looking into this. This patch makes audio usable again for me on 2.6.30 kernel and overruns are gone. What I want to ask is, there is constant noise on the output on data rates higher than 8000 khz. Anything playing with a higher rate generates random artifacts on the output. I have no idea where to look or how to test further. Do you have any pointers? Thanks, Caglar Signed-off-by: Troy Kisky troy.ki...@boundarydevices.com --- I fixed 2 bugs in this version. 1. I ensure that buffer size is a multiple of period size. 2. Disable asp rx on shutdown. If it is not disabled on shutdown, you cannot initialize it to different settings. include/sound/soc-dai.h |2 + sound/soc/davinci/davinci-i2s.c | 337 - sound/soc/davinci/davinci-pcm.c | 521 ++- sound/soc/davinci/davinci-pcm.h |3 +- 4 files changed, 624 insertions(+), 239 deletions(-) ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 0/11 - v3] ARM: DaVinci: Video: DM355/DM6446 VPFE Capture driver
Denys Dmytriyenko wrote: On Fri, Jun 26, 2009 at 12:46:31AM +0300, Yusuf Caglar AKYUZ wrote: Karicheri, Muralidharan wrote: We have been using the arago git tree and I don't see any of the issues that you are seeing. Could you use arago tree and Let me know if you see the same issue? This tree is being tested internally by TI. So it is known to work. http://arago-project.org/git/people/sneha/linux-davinci-staging.git P.S. AFAIR, arago tree is based on 2.6.30-rc2. Eventually we have to test these patches with davinci tree. That's the confusion I was afraid of - the latest patches against 2.6.30 final were tested in the next branch, while master stayed slightly old. One of the reasons was to achieve feature-parity of the next branch before making it the new master, which happened today. But, as I explained in the previous email, that still has v2 vpfe patches, which I'll start rebasing tomorrow... Thanks for the heads up. This tree boots fine, without any issues. A minor problem is asoc code is not building without CONFIG_SND_DM6467_SOC_EVM=y. BTW, does this tree includes musb patches which was submitted to usb mailing list. but yet not included in mainline and davinci trees? Thanks, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 0/5] davinci: This patch series ports PWM driver from 2.6.10 tree.
Koen Kooi wrote: On 25-06-09 00:12, caglarak...@gmail.com wrote: From: Yusuf Caglar Akyuzcag...@bilkon-kontrol.com.tr This patch series imports 2.6.10 PWM driver into recent kernel. No functional changes have been made to the driver itself. However, I added a simple sysfs interface for functionality check. I verified this driver on 2.6.28 using dm6446 EVM. Patch currently applies on top of current DaVinci HEAD. I ran this series through checkpatch.pl after applying all of them but individual patches fails. I guess this series may be merged into 2 if they are decided to be added to current HEAD. Is it using the linux PWM kernel infrastructure? No it is not. BTW, I'm not aware of any PWM kernel infrastructure. While porting this driver I looked into to kernel tree but I couldn't find anything related to. Do you mean patches at [1]? I remember they were posted last year but I don't think they have been added to kernel tree. Regards, Caglar [1] http://lwn.net/Articles/302338/ ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 0/5] davinci: This patch series ports PWM driver from 2.6.10 tree.
David Brownell wrote: On Thursday 25 June 2009, Yusuf Caglar AKYUZ wrote: BTW, I'm not aware of any PWM kernel infrastructure. While porting this driver I looked into to kernel tree but I couldn't find anything related to. linux/pwm.h ... pretty minimal, and ISTR PXA-centric, but it can be evolved. It looks like pxa, s3c and mxc are using this infrastructure along with kernel backlight interface. Initially I was looking for a way to control backlight in a generic way. I may convert this driver as well. Thanks, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 0/11 - v3] ARM: DaVinci: Video: DM355/DM6446 VPFE Capture driver
Kevin Hilman wrote: m-kariche...@ti.com writes: From: Muralidharan Karicheri a0868...@gt516km11.gt.design.ti.com Big Thanks to all reviewers who have contributed to this driver by reviewing and offering valuable comments. VPFE Capture driver for DaVinci Media SOCs :- DM355 and DM6446 This is the version v3 of the patch series. This is the reworked version of the driver based on comments received against the last version (v2) of the patch and is expected to be final version candidate for merge to upstream kernel FYI... I've updated the staging/vpfe branch of davinci git with this series and the tvp514x v3 patch. Also, I'll be pushing the arch/arm/* patches of this series to DaVinci git master and queueing them for upstream merge. Has anyone using this staging tree? I'm having hard time to making this work. My problems are: 1) This branch does not built without attached patch. 2) tvp514x throws attached kernel oops. I guess this due missing platform data master definition in board support file. 3) Even if I correct these two and compile drivers built-in then I see an oops message related to asoc code on boot. 4) If I make these drivers kernel modules, then I can install all of them but dm6446_ccdc. Don't know why, just no messages on dmesg. I tried v1 and v2 of these patches previously, they were have the same problem of (3) but not the others. I was able to make v1 and v2 working but this one is a dead end due to (4). Thanks, Caglar diff --git a/arch/arm/mach-davinci/board-dm644x-evm.c b/arch/arm/mach-davinci/board-dm644x-evm.c index 13b73a7..c1b8747 100644 --- a/arch/arm/mach-davinci/board-dm644x-evm.c +++ b/arch/arm/mach-davinci/board-dm644x-evm.c @@ -29,6 +29,7 @@ #include linux/phy.h #include linux/clk.h #include linux/videodev2.h +#include media/v4l2-int-device.h #include media/tvp514x.h #include asm/setup.h #include asm/mach-types.h diff --git a/drivers/media/video/davinci/vpfe_capture.c b/drivers/media/video/davinci/vpfe_capture.c index 003450b..4ffb164 100644 --- a/drivers/media/video/davinci/vpfe_capture.c +++ b/drivers/media/video/davinci/vpfe_capture.c @@ -70,7 +70,7 @@ #include linux/platform_device.h #include linux/interrupt.h #include linux/version.h -#include media/v4l2-common.h +#include media/v4l2-int-device.h #include linux/io.h #include media/davinci/vpfe_capture.h #include media/tvp514x.h Unable to handle kernel NULL pointer dereference at virtual address pgd = c5dd8000 [] *pgd=85dce031, *pte=, *ppte= Internal error: Oops: 17 [#1] PREEMPT Modules linked in: tvp514x(+) videobuf_dma_contig videobuf_core ipv6 CPU: 0Not tainted (2.6.30-davinci1 #105) PC is at strcpy+0x4/0x18 LR is at tvp514x_probe+0xf0/0x18c [tvp514x] pc : [c01efd5c]lr : [bf04fce8]psr: 4013 sp : c6309e50 ip : c048579c fp : 0008b480 r10: r9 : c6308000 r8 : c049e6c4 r7 : bf051724 r6 : bf0500e8 r5 : c5c15e00 r4 : c5de1c00 r3 : 000c r2 : c5de1c3c r1 : r0 : c5de1c3c Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user Control: 0005317f Table: 85dd8000 DAC: 0015 Process insmod (pid: 993, stack limit = 0xc6308268) Stack: (0xc6309e50 to 0xc630a000) 9e40: c5c15e20 bf04fbf8 c5c15e00 9e60: c5c15e20 c0291fcc c5c15e20 c5c15e54 bf051724 c022e7c0 c5c56b70 c5c15e20 9e80: c5c15e54 bf051724 bf051724 c022e8d8 c6309ea0 c022e878 c022dca4 9ea0: c5c2efb8 c5c56b70 c5c2ef88 bf0516f8 bf051724 c632f900 bf054000 c022e258 9ec0: bf0507e0 0001 c6920088 bf0516f8 bf051724 000ac018 bf054000 c04a74cc 9ee0: c022ecb4 0021 bf0516f8 bf051aa8 000ac018 bf054000 c04a74cc 9f00: c0292f94 4a57 bf051aa8 c00292fc c691ff45 c57e3de0 9f20: c6921558 001f 0021 c69205b0 bf051ab4 c6920588 00b5 9f40: 9f60: 4a57 bf051aa8 000ac018 4a57 bf051aa8 000ac018 c0029f64 9f80: 000a96ac c0063bc0 c6308000 000ac018 4a57 01b0 be9c8ecd 9fa0: 0080 c0029de0 4a57 01b0 000ac018 4a57 000ac008 01b0 9fc0: 4a57 01b0 be9c8ecd 0080 00099f58 0046 000a96ac 0008b480 9fe0: be9c8c40 be9c8c30 00017110 40182ff0 6010 000ac018 e28dd004 e8bd4030 [c01efd5c] (strcpy+0x4/0x18) from [bf04fce8] (tvp514x_probe+0xf0/0x18c [tvp514x]) [bf04fce8] (tvp514x_probe+0xf0/0x18c [tvp514x]) from [c0291fcc] (i2c_device_probe+0x68/0x80) [c0291fcc] (i2c_device_probe+0x68/0x80) from [c022e7c0] (driver_probe_device+0xac/0x164) [c022e7c0] (driver_probe_device+0xac/0x164) from [c022e8d8] (__driver_attach+0x60/0x84) [c022e8d8] (__driver_attach+0x60/0x84) from [c022dca4] (bus_for_each_dev+0x4c/0x80) [c022dca4] (bus_for_each_dev+0x4c/0x80) from [c022e258]
Re: [PATCH 0/11 - v3] ARM: DaVinci: Video: DM355/DM6446 VPFE Capture driver
Karicheri, Muralidharan wrote: Yusuf, I am not sure what is the state of vpfe branch maintained by Kevin. We have been using the arago git tree and I don't see any of the issues that you are seeing. Could you use arago tree and Let me know if you see the same issue? This tree is being tested internally by TI. So it is known to work. http://arago-project.org/git/people/sneha/linux-davinci-staging.git I tried this one before and the results were the same. However, that time I imported vpfe stuff from arago tree into davinci tree. But I don't remember what was the result of arago tree. I guess I'll give this another try. Thanks, Caglar P.S. AFAIR, arago tree is based on 2.6.30-rc2. Eventually we have to test these patches with davinci tree. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 0/9] This patch series ports VPBE drivers from 2.6.10 tree.
Karicheri, Muralidharan wrote: Caglar, Is this based on LSP 1.20/1.30? Yes. Thanks, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Status of PWM, Resizer and VPBE Drivers
Karicheri, Muralidharan wrote: Yusuf, When you do you plan to send a patch to this list for review ? I cleaned my patches. I was waiting for latest patches on the list to be applies since some of my patches touches mach-davinci/* as well. I'm going to post in a few minutes. Thanks, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: about cmem module for kernel 2.6.30!
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jeff wrote: All, I am using the newest kernel (2.6.30) on DM355. I need to use cmem module from LSP 1.20 package. When I recompiled cmem module using kernel 2.6.30, It seemed it can not get work. (kernel was dead). I don't why it is. How to use cmem module for new kernel 2.6.30? Thanks very much. Do you mean it is locking kernel while loading? If so, you may try to reduce pool sizes while loading cmem module. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoyFu8ACgkQ/nL+S5dojeg1fwCgj5p9wtZ98kWt6qXUrSNGcdJg hAgAnAtN1wyey1LtpfMxOJTdqJyQNnnx =m1FN -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: device driver for 800x480 LCD
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 elbert shiang wrote: Hi, Dear All: Hi, I want to make dm355 connect to the TFT-LCD which has RGB666 signal. I have no idea where to start to build the device driver from tool chain. Can anyone guide me which documentation I should read and what brief procedure I should proceed to integarte the 800x480 TFT-LCD into Linex kernel? Which kernel are you using? MV or GIT? the LCD I refer to is AM800480E3TMQW-00H from AMPIRE CO. Does this LCD has a public datasheet? Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkouxT4ACgkQ/nL+S5dojeh+lQCgtTWikBtL1r1P+BS3XOfaPdVY HFcAn3WzKx4dSfM6w2yYRc6U7gZWx7a1 =wPHu -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: vpfe capture driver plans
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Karicheri, Muralidharan wrote: Hi All, As you all know that I had submitted initial version of the vpfe capture driver to V4l2 mailing list and got comments against the same. I am currently doing rework of the driver based on the comments. This is my plan for the driver going forward Is there any GIT tree where these changes are staged? As the integration process takes some time, it would be nice to use/test your changes in the mean time. Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoqCbsACgkQ/nL+S5dojeifjwCeNIHRvAGKnIeK0yPg6L8R/GtG ufgAnitRA0HeyyUAjDmeMe4WL8wajQMz =/TaE -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dvsdk 2.0 oabi
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Markus Rechberger wrote: Hi, Hi, I was trying to compile the dvsdk 1.3 and dvsdk 2.0 for eabi 0 but since there are a few binary libraries included it does not work out. Is there any way to get EABI 0 versions of those precompiled dvsdk libraries? They are EABI already: $objdump -x gt.o470MV | grep flags architecture: armv5t, flags 0x0011: private flags = 400: [Version4 EABI] /usr/arm/lib/gcc/arm-unknown-linux-gnu/3.4.6/../../../../arm-unknown-linux-gnu/bin/ld: failed to merge target specific data of file /root/dvsdk_1_30_01_41/framework_components_2_00_01/packages/ti/sdo/utils/trace/lib/gt.a470MV(gt.o470MV) CBuf.o470MV: ELF 32-bit LSB relocatable, ARM, version 1 (SYSV), not stripped gt.o470MV: ELF 32-bit LSB relocatable, ARM, version 1 (SYSV), not stripped gt_config.o470MV: ELF 32-bit LSB relocatable, ARM, version 1 (SYSV), not stripped package_ti.sdo.utils.trace.o470MV: ELF 32-bit LSB relocatable, ARM, version 1 (SYSV), not stripped I guess your toolchain is too old, hence not EABI. Have you tried Codesourcery toolchain? Regards, Caglar thanks, Markus ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkojfDkACgkQ/nL+S5dojegT3ACffnatd1VIg7qv4u2CTamtKv6S KIoAoKh7BzU69alZUMlZwxu8ZM4eTi75 =AwL6 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dvsdk 2.0 oabi
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Markus Rechberger wrote: [...] I got the dvsdk already work with a newer compiler although, my problem is that the target system uses the Ingenient Codecs and was built with OABI/EABI 0 and Linux 2.6.18. In order to be able to use the DVSDK and the gstreamer_ti framework I would need the cmem and dsplinkk driver for the# same system as well as the userspace libraries with OABI/EABI0. The main reason why I cannot change the base system is the audio driver... So you want to run EABI binaries in an OABI environment? To my understanding this is not possible... The original delivered box shows: bound device '0-001b' to driver 'OMAP+TLV320AIC23 codec' bound device 'audio-i2c' to driver 'audio-i2c' bound device '0-001a' to driver 'WM8955L I2C Driver' bound device 'davinci-audio.0' to driver 'davinci-audio' but with the DVSDK the i2c map shows: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- 0a -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- 1a -- -- -- -- -- 20: -- -- -- 23 -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- 39 UU -- -- 3d -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: 50 51 52 53 54 55 56 57 -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 6f 70: -- -- -- -- -- -- -- -- so the audio codec doesn't seem to be visible on the i2c bus The WM8955L specs are freely available, writing a driver for it wouldn't be a problem, although I'm concerned about how to enable the AIC23 codec on it... I think both have the same address, 0x1a. Looking at aic23 datasheet, its address can be 0x1a or 0x1b depending on the CS pin. Any change that CS pin is software controlled on your board and low currently? Best Regards, Caglar In order to get around this problem the OABI/EABI0 DVSDK binaries would be required. Best Regards, Markus -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoj8GgACgkQ/nL+S5dojehNCQCfe7RiDQJSUcTR3wekMCcRRw/J bekAn3UFovSNc0MLlRGFcD50yfWoycC4 =SPHy -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dvsdk 2.0 oabi
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Markus Rechberger wrote: [...] I know it's not possible.. that's why I'd need DVSDK OABI binaries .. is there any way to get those? As for the kernel modules I saw the uClib objects for DVSDK 1.3 are in the OABI format, but the other libraries are in EABI4 format, and the system which I got is not compiled for EABI4. AFAIR all the DVSDKs I saw were compiled with EABI toolchain. In fact MV toolchain was eabi. But I don't know for sure. The original delivered box shows: bound device '0-001b' to driver 'OMAP+TLV320AIC23 codec' bound device 'audio-i2c' to driver 'audio-i2c' bound device '0-001a' to driver 'WM8955L I2C Driver' bound device 'davinci-audio.0' to driver 'davinci-audio' but with the DVSDK the i2c map shows: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- 0a -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- 1a -- -- -- -- -- 20: -- -- -- 23 -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- 39 UU -- -- 3d -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: 50 51 52 53 54 55 56 57 -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 6f 70: -- -- -- -- -- -- -- -- so the audio codec doesn't seem to be visible on the i2c bus The WM8955L specs are freely available, writing a driver for it wouldn't be a problem, although I'm concerned about how to enable the AIC23 codec on it... I think both have the same address, 0x1a. Looking at aic23 datasheet, its address can be 0x1a or 0x1b depending on the CS pin. Any change that CS pin is software controlled on your board and low currently? yes I saw that too but I'm confused about the WM8955L driver sitting on 0x1a already bound device '0-001a' to driver 'WM8955L I2C Driver' bound device 'davinci-audio.0' to driver 'davinci-audio' Do you know such components can be enabled/disabled on the dm6446? I tried to play around with the GPIO configuration and i2c gpio expander on 0x3a of the dm6446 initcode but without success. If you codec's CS line is connected to a GPIO on your board then you can control that GPIO from user space. I guess this is valid for i2c expanders as well. For instance to control GPIO 5: $cd /sys/class/gpio $echo 5 export $cd gpio5 $echo out direction $echo 1/0 value I also tested the latest kernel with that board, there's a difference in the i2c map between 2.6.18mvl and the latest git kernel: 2.6.18mvl: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- -- -- 0a -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- 1a -- -- -- -- -- 20: -- -- -- 23 -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- 39 UU -- -- 3d -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: 50 51 52 53 54 55 56 57 -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 6f 70: -- -- -- -- -- -- -- -- latest git kernel: 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: -- -- -- -- -- 08 -- -- -- -- -- -- -- 10: -- -- -- -- -- -- -- -- -- -- 1a -- -- -- -- -- 20: -- -- -- UU -- -- -- -- -- -- -- -- -- -- -- -- 30: -- -- -- -- -- -- -- -- -- UU -- -- -- 3d -- -- 40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 50: UU 51 52 53 54 55 56 57 -- -- -- -- -- -- -- -- 60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 6f 70: -- -- -- -- -- -- -- -- They seem consistent. I guess you have another chip(max7310???) which is addressed via GPIO. I don't think anything there is anything wrong with I2C setup, other than chip addresses. Best Regards, Caglar thanks for your response! Markus -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkokBi8ACgkQ/nL+S5dojeiDGQCgohmzGDH+pjmoRCRK1OND47zC IPQAn1FopKhF3ymnqoMkmdb1Z0XRO0Fx =Aioq -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: SRAM allocator(s!)
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Vladimir Pantelic wrote: David Brownell wrote: If, on a DM355, the ASoC driver can't get SRAM, there's no point in continuing, since the dropout problems basically make audio unusable without it. Even on otherwise idle systems. (The same may be true on some dm6446 systems, although dropouts don't show up so readily there.) Is there any info on why DM3556 and DM6446 have issues with audio dropouts? Is this a known errata or by design? I'm asking my self a similar question since this thread has been started. On DM6446 I don't see any underruns or dropouts for a long time, even when there is heavy network or video capture activity. But there are some other issues probably not related to this. I'm searching for ways to create this problem on my system. Any methodology is welcome. Is it possible that this is related to aic codec? Troy Kisky mentioned he was using aic23 whereas newer designs seem to use aic33, though I don't know if there are any significant differences. Regards, Caglar ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoZrL0ACgkQ/nL+S5dojeib5ACfRTh6zYCP3h2gPsEJ6SeEKj33 5zEAn0V5w5YW4j5/WDD9p3zUPAWz0bC6 =zEaK -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Booting dm355 from SD Card
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 David Brownell wrote: On Sunday 17 May 2009, Yusuf Caglar AKYUZ wrote: Because I couldn't find any UBL implementation which reads u-boot from SD Card, or at least mentioning of it. I was pretty sure that everything was ok, but I should re-check since you say it works. http://sourceforge.net/project/showfiles.php?group_id=233754 The 1.50 DaVinci Serial Boot and Flashing code has code for such a UBL -- not built by default though. ISTR that the SD card boot and flash utility code you referenced is built from older (1.10) versions of that code. Looking at the code, I see no headers named sdmmc.h or sdmmcboot.h, both are referenced from sdmmcboot.c I wonder if this can be compiled. That's an other story though. I started from scratch, still no results. I'm sure RBL boots from SD, but I get no response from the board after that point. I'm checking only serial port but I assume there must be something on my uart. Anyway, I go digging. Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoRojoACgkQ/nL+S5dojejCHwCgmdeuY6XvAwIDqSHo71xZu57c 8PYAmQGlYqIzqUbs0hURWC4rWVPRBpNw =NKrI -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Booting dm355 from SD Card
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Venkatachala Upadhya wrote: Hello, Hi, See the below post, in the mailing list. If that helps! http://linux.omap.com/pipermail/davinci-linux-open-source/2009-March/011957.html That was my starting point, but no success here yet. Thanks anyway. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoRopcACgkQ/nL+S5dojejDJgCbBKIVjcOII45tXFZ7t0QXkRg2 DCYAnAp+mc6H9Gw1QyB2FyCKZ3he0C0C =bCiY -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Booting dm355 from SD Card
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi everyone, I'm trying to boot dm355 from an SD card. Only resource I have found is [1]. Tools mentioned at [1] seems to format and program the SD card correctly and evm is at least not toggling GIO61 but I see nothing on serial console and board just sits there. I even dump memory address 0x7ffc and block number seems valid. But I suspect ubl needs nand in this setup anyway. I have some questions in mind as: does anyone have pointers to boot dm355 from an SD card without any nand involvement? I think only u-boot support available is still 1.2.0, is it right? I see [2] and [3] mentions adding support but I guess there is still some time until that happens. I see no SD support in u-boot for DaVinci. I read [4] but it is for u-boot 1.2.0 and only dm6446 I guess. This means ubl has to copy linux kernel with u-boot as well for full sd boot. Is this the case? Thanks, Caglar [1] http://wiki.davincidsp.com/index.php?title=DM355_SD_card_boot_and_flash_utility [2] http://linux.omap.com/pipermail/davinci-linux-open-source/2009-April/012848.html [3] http://linux.omap.com/pipermail/davinci-linux-open-source/2008-June/007105.html [4] http://linux.omap.com/pipermail/davinci-linux-open-source/2007-August/003929.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoQgRMACgkQ/nL+S5dojejMbQCeNv8qjbTyrDD0z/Fu5BTvv4c0 6KcAnjAZpiaXnN3GAMjDJn/j/pKahb5J =K9Dg -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Booting dm355 from SD Card
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 David Brownell wrote: On Sunday 17 May 2009, Yusuf Caglar AKYUZ wrote: does anyone have pointers to boot dm355 from an SD card without any nand involvement? http://wiki.davincidsp.com/index.php?title=DM355_SD_card_boot_and_flash_utility I don't know why you think that needs NAND in any way. You must be doing something wrong... unless that broke since I grabbed a copy, it's working fine. Because I couldn't find any UBL implementation which reads u-boot from SD Card, or at least mentioning of it. I was pretty sure that everything was ok, but I should re-check since you say it works. I think only u-boot support available is still 1.2.0, is it right? No. Basic DM355 EVM support is in current U-Boot trees: http://git.denx.de/cgi-bin/gitweb.cgi?p=u-boot.git;a=shortlog;h=refs/heads/next ... that is, the u-boot next branch. That's not what I'd call a stable branch, but it's based on 2009.06-rc2 and all the patches there should carry forward to the release after that. (Meaning, 2009.06 won't support DM355, but the release after it will.) I wasn't aware of that tree. I'll try that one. Thanks for the pointers, Caglar Still missing: NAND support isn't enabled in the dm355evm configuration. The 1-bit ECC flavor will work fine, but one wants to use the 4-bit ECC. I see no SD support in u-boot for DaVinci. As they say, -ENOPATCH. :) Feel free to provide it against current U-Boot. - Dave -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoQ9iMACgkQ/nL+S5dojehj8wCfcIQ3PR/i83jdhBphKPld5vbB +CMAoKjUZY7tnmjjKRQ28o53/bYBf8/6 =EqzF -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: DM355 - No vpss register write happens in the latest kernel
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Karicheri, Muralidharan wrote: Kevin, I have ported the video driver to the latest davinci git kernel and I have built a single uImage for DM355 and DM6446. The driver works fine on DM6446, but doesn't work on DM355. When I dumped the registers, none of them were being written. I tried writing manually using writel application and still I can't see the values in the registers. The architecture code has changed recently quite a lot. I have a version of the kernel that downloaded in March and this issue was not seen there. This is blocking my driver work and any help will be appreciated. We have seen the same behavior recently. There is no davincifb registiration in board-dm355-evm.c. Apart from that, vpss clocks were getting disabled but we haven't dig further into this. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkoLHMUACgkQ/nL+S5dojejLfQCaAzG/v2Us1aZWIAxIvlzWtM7t 2MgAn1mbKnpjQPcRjjnPGe5S74LK2lRa =0/DD -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Kernel hangs while booting
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Muneendra wrote: Hi, No, Kernel itself is not comming up.. It is just uncompressing the kernel and trying to boot as shown below Starting kernel ... Uncompressing Linux... . done, booting the kernel. I think this is due to framebuffer driver. Can you try adding framebuffer arguments to your boot command? It must be smt like video=davincifb:... Sorry I don't remember exactly and my driver is not the same. You can check exact arguments in davincifb.c Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknxeR0ACgkQ/nL+S5dojejbSgCgoxlTvRpKxP5nrzJpnoIOKkp4 LcwAoImeSZSPY1atFV5jx/RErqwArYbh =eNu4 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Freeing init memory hang
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Young, Joseph wrote: Hello, Hello, So I am attempting to boot the current Davinci Linux kernel for the Davinci DM644x EVM board, and am running into some problems. I have compiled the uImage, and am using the ramdisk.gz provided by TI, but when I attempt to load the kernel, it hangs indefinitely at Freeing init memory: 124K. I have tried Davinci linux kernel versions 2.29 and 2.30-rc2, and both experience the hang. I have also built a minimal kernel by following SPRAAH2A, and that also experienced the hang. The same boot parameters work correctly with the montavista Linux image provided by TI. I thought the problem was with mounting busybox, so I wrote a Hello World program, and attempted to run that. My kernel still hangs, while the montavista linux kernel will display my program, so I think the problem is with my kernel configs setup. I also thought it was a tool chain problem, and have rebuilt the kernels with the GNU C arm compiler, as well as the RidgeRun DM6446x SDK kit, and both experienced the hang. The interesting thing is that when I unplug the cord from my TFTP server, the hung kernel will output PHY: 1:01 - Link is Down and when I plug it back in PHY: 1:01 - Link is Up - 100/Full. This means that the kernel is still somewhat responsive, just not doing what I want it to do, display the shell. I'm out of ideas at this point. Has anyone run into this issue, or can anyone point me in the right direction? My guess: Kernel is running, but there is something wrong with filesystem /dev nodes. AFAIK, this point is where control passes to init and console is redirected to /dev/console(or /dev/console must be present). This is also the same for your own application. Maybe your file system is not EABI? If I were you, I would try an NFS file system. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknxgtIACgkQ/nL+S5dojeinbQCdHzh1awssu9V73VbIxJJjUwkt QL8AnAsnCCv+JkEZ7KBtjNj07K83YNSt =4y+J -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: U-Boot fw_setenv problem.
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 srikanth bomma wrote: Hello all, Hi, [...] CRC write error on /dev/mtd1: Invalid argument Error: can't write fw_env to flash There is a bug in older u-boot fw_env.c file. I changed tools/env/fw_env.c from latest u-boot tree and it worked. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknsly0ACgkQ/nL+S5dojeilPgCgsKChebg2bgLU4bq2BYAbdxZo bQQAoLsgAPs5xCmBXf3/M1Li5P9k8d8a =OIGE -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: divinci linux 2.6.28+ with DVSDK 1.30
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ondrej Pindroch wrote: Hi Will dvsdk 1.30 work with davinci linux 2.6.28+ instead of MVL 2.6.10 delivered with dvsdk? Or there are some changes needed? Yes, but not out of the box. You need to compile dsplink and cmem for your kernel. Then you can use dvsdk dsp combos with 2.6.28+. Moreover, if you want to use applications from dvsdk, you need to make some minor modifications to examples. For instance, they use oss instead of alsa and there may be some other minor issues. Best regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknm4j0ACgkQ/nL+S5dojegmhACcCHlHdTZVOPlwmZ7A9UjfW/qv /NMAnjZFae2FrsX7owlwtzVLffa8505g =dWgi -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: GUI development for OMAP
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Mahalakshmi Gonuguntala wrote: Hi, I have downloaded QT/embedded qt-embedded-linux-eval-src-4.5.0.tar.gz and cross compiled the qt for my arm toolchain.(./configure -embedded arm , make , make install ) and qt installation is sucessful. How to run a qt example on my board ( i have OMAP3530 EVM, mistral board) . You should install your libraries(if you haven't build Qt with static option) to your target board some how. If you have an NFS development environment all you need is to copy Qt libraries into YOUR_NFS_ROOT/usr/lib. Then you can run any of your test appliacations like(on your board): $someqtapp -qws You can find further detailed information in Qt documentation. What libraries i should copy to board. I have some libraries generated at /usr/lib/qt*.*. / I'm surprised that your PC is still functioning. I would like to develop a player kind of application which will list all the files on the board and when i select a file, i should be able to play the file using my decoder application. This would be off-topic in this list. You need more libraries to achieve this. I strongly recommend that you use OpenEmbedded since it builds all the libraries as well as a rootfs for you. OMAP3 is very well supported in Openembedded, with all the necessary libraries including Qt. I would like to use the touch screen . You need to install tslib and your Qt must be configured to tslib at *compile* time. After these two requirements are met, all you need to do is to run your applications like: $QWS_MOUSE_PROTO:Tslib:/dev/touchscreen0 someqtapp -qws Where can i specify the touch screen driver in QT? Try this before configuring your Qt(before building) to get more info: $./configure --help | grep -i tslib Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknlzhUACgkQ/nL+S5dojei26ACdEk560r61QZKPWYG9ch922gP4 gmYAn1Hbi5rqWEKFB+lLKy+rdHRlnSyc =CWI7 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: moved on to a kernel cmem problem
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Mat Laibowitz wrote: I have been over the memory map a dozen times. I have set my mem parameter in the bootargs to as much as 10 megs lower than the start of my cmem region. I have given DSPLINK more than enough space in the codec servers memory map and have adjusted the dsplink's memory map to match. No matter what I do, when the dsplink code gets to either mutex_lock_interruptible or mutex_unlock it crashes with the kernel paging error. I have commented out these functions as an experiment and it stops crashing. It does not work correctly because everything tries to access it at once without the locks, but it keeps going. As another experiment I tried messing up the memory map slightly, or filling all the shared memory with the linux section. And all of these fail with a different error. It really seems to me that the dsplink is not properly linking against the kernel and it is not able to find the specific code for the mutex library. Maybe I am not setting something correctly in the make files. I have also tried using the wiki's instruction for creating a makefile with kbuild. Same problem. Thanks for your reply. Any additional help would be great. -mat As I said I had *very* similar problem and I thought there was a problem with mutex handling in my Dsplink/kernel combination. My tentavive fix is at [1]. However, later my problems did not ended, and I downgraded my Dsplink version to 1.40. 1.40 was more stable than 1.60 for me. But later on I hit a kernel bug which is present in 2.6.28 but not in 2.6.27/29, please see [2]. After applying that fix, no more Dsplink problems( at least for now). However, I never tried 1.60 again. Regards, Caglar [1] http://linux.omap.com/pipermail/davinci-linux-open-source/2009-March/011647.html [2] http://www.spinics.net/lists/arm-kernel/msg62873.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkni/NMACgkQ/nL+S5dojej/9wCguUPFo3O2nntGg00gd1X0Nu6C S74AoJM2aM4i84vBe63dWp5JcGUIhBe3 =wVlS -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: moved on to a kernel cmem problem
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Mat Laibowitz wrote: In order to debug my mpeg4dec app side program I was looking dmai to see how it accessed the viddec2 ce interface. Along the way I decided to try and compile it and also to try and upgrade some components. I modified the dmai source to support my custom board and can compile it. Now I am having a run-time issue with the codec engine. I am just trying to run the decode demo. I get the error: Unable to handle kernel paging request at virtual address c808b000 There is a lot of talk about this error, and I have read through the postings, but still have not figured it out. If I change the CMEM insert command, I can get it to say: CMEMK Error: CMEM phys_start (0x8760) overlaps kernel (0x8000 - 0x8780) [...] Looking at the kernel oops, this is dsplink problem rather than cmem. Which kernel version are you using? and is this dsplink version 1.40? I faced a similar issues with 2.6.28 recently. If we are using the same software versions, I can summarize my findings. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAknhk3UACgkQ/nL+S5dojeiE2wCfc7pxnMMdXS8xciSfGAvxfMH/ dTkAnRgmpnQLfvNY11i6tct70PfBeKdM =ey7g -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: GUI development for OMAP
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Mahalakshmi Gonuguntala wrote: Hi All, Hi, We would like to develop a GUI for MISTRAL board (OMAP3530) . The GUI has to have a player kind of look. It should list all the files in the file system and we should be able to select a file from the file and a small play button,. when we invoke the play button, our decoder has to run. Have you tried Qt? You can integrate gstreamer with Qt which also easies media application development. With a gstreamer + Qt combination, your requirements are just a matter of writing a few C++ codes. Best regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkndjnoACgkQ/nL+S5dojeirYQCfXhdR6PMko//cTNnTOvnSQVqf GHUAoIMzAdRo01LGpsZtw3rPHv2J/koq =sUGD -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: DSPLINK 1.60 Oops'es on 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Das, Samir wrote: Caglar, Are you seeing this OOPs when you do Ctrl-C? Since you say that mutex is deleted, which means that some thread has done the cleanup (Mutex got deleted). I was trying to come-up with more info on this but I had no success. So I downgraded to dsplink 1.40, then we realized a memory mapping issue with 2.6.28 kernels. ( many thanks to Kevin Hilman for pointing this) So issue seems not dsplink related, though I haven't tried 1.60 after kernel bug. One last note in this issue worth mentioning is that I realized some of oops from dsplink were false negatives. I mean if one part of kernel(or application) fails, ıt is possible to see dsplink routines causing kernel oops messages. Thanks for your help. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm8iZwACgkQ/nL+S5dojegjXACgiD583FgV3l2It7HxOSm2US1H RcQAoK9Qjr5Bf+tVy4QEVs2fIr8x9Wtf =w8FV -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Montavista Linux
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Marcelo Guedes Silva wrote: I´m still with this problem. I follow three documents step-by-step, all of them said similar things, but the final uImage file has some problem. One of this document is the quickstart in: /opt/mv_pro_4.0.1/montavista/pro/documentation/docs/quickstart/cross-development/html I guess getting started guide is also mentionioning how to compile a new kernel as welll. Basically I copied the kernel source from /opt/mv_pro_4.0.1/montavista/pro/devkit/lsp/ti-davinci_evm-arm_v5t_le/linux-2.6.10_mvl401 to a temp directory: /tmp/mykernel. How did you copy? In order to preserve sym links you have to copy with cp -aR or better, create an archive and copy that. There, I compiled the kernel with: make menuconfig and after make uImage. Did you check if target system is DaVinci? I mean your tree maybe distclean'ed, or not configured before. If you're unsure use a defconfig and then make your changes on it. I don't remember MV defconfig name but you can find with: make help | grep -i davinci or make help | grep -i dm644 Then use that defconfig with: make xxx_defconfig make menuconfig make uImage The uImage was copied to /tftpboot directory, that I configured properly to work with tftp. The transfer happened right too. U-boot # tftp uImage 0x8040 U-boot # bootm 0x8040 And now I see the diamonds. You can see that the kernel started (the jpg picture), but for some reason, stopped without any warning. I tested with parameters to boot from /dev/ram *and* from /dev/nfs with: setenv bootargs console=ttyS0,115200n8 noinitrd rw ip=10.0.0.100 root=/dev/nfs nfsroot=10.0.0.20:*filesys path*,nolock mem=120M davincihd_capture.channel0_numbuffers=4 Whatever your paramters other than console you should see smt meaningful. Both gets the same problem. Ideas? Maybe you should start from scratch. I mean take fresh tar.gz source, copy it, extract it, build it with defconfig, boot it, if succeeds menuconfig it, re-build, boot it, if fails look into changes you made and so on... I can nothing else at the moment, I guess you have a very very non-standard behavior which hints me that you're doing smt unusual. BTW, are you compiling your kernel on Windows? if not(and you shouldn't be able to) you may use minicom given that you're not on a VM. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm6UYMACgkQ/nL+S5dojeieBQCeOvFzvu5nSeXUChq1ILAHHUAC Q8AAnjLR0DdcE2vofume4ygOSOn4pvNr =ZLwU -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Davinci, MUSB, DVEVM - driver but no device?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Lars Doelle wrote: Unfortunately, the 'alike' board differs here, in that it does not have a GPIO, which might explain the different behavior of the system. On the board at hand, the USB power is hard-wired. I couldn't find David Brownell's post on this topic, but according to that post hard-wiring is not working with DaVinci usb. I saw this while I'm playing with my boards and musb driver, also. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm6WgoACgkQ/nL+S5dojegwrACgoxUqrwp2gLW8k3Q3bqtDFHgB 4ysAoKuvBUiqQ1YLiQC9AsbYCKuXLj88 =FoNp -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 6/7] platform-related-updates for vpfe capture driver on DM6446
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 m-kariche...@ti.com wrote: Add platform related changes for vpfe capture driver on DM6446 [...] + /* * MSP430 supports RTC, card detection, input from IR remote, and * a bit more. It triggers interrupts on GPIO(7) from pressing @@ -557,9 +636,12 @@ static struct i2c_board_info __initdata i2c_info[] = { I2C_BOARD_INFO(24c256, 0x50), .platform_data = eeprom_info, }, + { + I2C_BOARD_INFO(tvp5146, TVP5146_I2C_ADDR), shouldn't this be tvp514x or TVP514X_MODULE_NAME ? Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm60w8ACgkQ/nL+S5dojeiYTACgpxR6uFCiaWJYD1jfev8DhBUn DAUAn0WZuOeoLktFIvMIveiqmljHHu7L =2vv4 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [DaVinci OpenSource CCB] Re: [PATCH 6/7] platform-related-updates for vpfe capture driver on DM6446
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Karicheri, Muralidharan wrote: Caglar, tvp514x is a generic name. I think in the board setup file, we need to use specific chip name that is used in the board which is TVP5146. Remember that tvp514x driver supports 4 types of chips and I think the name should match with actual device name with out which it doesn't work. But tvp514x registers itself to i2c layer with name tvp514x. This should have been changed on recent kernels. I'm looking at 2.6.28. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm62hIACgkQ/nL+S5dojej6QgCfaiRyzzi+YArTJdtgOyxBJRNI RRMAn1XuVLFHSFC+0BFLAE6RyaTUzyY2 =rBcW -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Question about compiled kernel image
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 James Whetstone wrote: Hi, Hi, I've compiled the linux kernel from the git repository source with the CodeSourcery toolchain and the result is that there are 2 binary files in the /arch/arm/boot directory (I think that's the correct directory to look for the kernel image, but I'm not sure). There's a zImage and Image file, both executable. Which file is the image I should use? None of them. u-boot needs uImage which you can generate with(at kernel build) make uImage mkimage tool must be in your PATH. At this point, all I've done is compile the kernel and u-boot. So now I want to deploy them to the device (dm6467). Can someone please run me through the basic workflow for uploading my compiled kernel and boot loader to the device (dm6467)? Or perhaps give me some documentation links that will tell me how to do this? 1. You should write u-boot into your nand/nor flash. If you're using nand flash you'll need first stage bootloader as well. http://wiki.davincidsp.com/index.php?title=RBL_UBL_and_host_program 2. Then you can download linux kernel to your board using tftp. You should setup a tftp server on your host machine. Then to boot your compiled kernel from u-boot prompt: setenv ipaddr some-ip setenv serverip host-machine-ip setenv bootargs console=11520n8,... saveenv tftp load-address uImage bootm load-adresss 3. You'll need a filesystem as well to get started: http://wiki.davincidsp.com/index.php?title=Linux_Toolchain#Kernel Good luck, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm6/e4ACgkQ/nL+S5dojejV7QCfRFzpe0tSylcY/yiFLZJoat7s 6d8An0d5clrFLZxJ5HxglXt/fqG0uADw =EKUq -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Davinci, MUSB, DVEVM - driver but no device?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Lars Doelle wrote: Hi All, i'm trying to get an USB-stick running on a DVEVM-alike hardware on a recent kernel, but the USB host appears to be plain dead. I've digged a little into the kernel and my result is, that 'platform_driver_probe' in the end of 'musb_init' never ends calling 'musb_probe', thus resulting in the musb_driver being deregistered and ignored by the kernel. Digging deeper, the 'platform_driver_probe' enumerates the 'platform' bus but fails to find any suiting device. Thus, from the sources, it appears as if the construction misses a 'platform_device' representing the USB host controller. Because only the kernel and no hardware appears to be involved, it looks like a silly kernel configuration problem to me. I write to ask if someone knows the solution or could otherwise help. I can produce the result with 'davinci_evm_dm644x_defconfig'. You said DVEVM-alike, does it have i2c port expander as well for VBUS control? If not you must modify musb driver as of 2.6.28, don't know current head. Moreover, host-device jumper is another issue to remember. And finally, have you modified your kernel config for host mode? In defconfig it is selected as gadget(device mode). Other than these, nothing comes to my mind. Regards, Caglar Thanks Lars ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm46JYACgkQ/nL+S5dojeiZTgCgkdfyELBr2s5GZjPasrhu1Ihj dm0AoKwFuW/qU5qSO9HW5qGaWjg97rR4 =UMXS -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Montavista Linux
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Marcelo Guedes Silva wrote: Hi everybody. I created a Montavista Linux Kernel Image. When I tried to run the kernel I received it (in the picture: diamond characters). Could someone explain what is happening? I have no idea and I didn´t receive error messages, just it. I guess you should check your command line for baud rate settings. It must be something like: ... console=ttyS0,115200n8 ... Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm5JpkACgkQ/nL+S5dojegnbwCfaeSVMeHQt6yQY1u08xiPTg4D lvUAoItOBTzEu4LZ7cBBcmjvu5bywEiY =bV4o -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Need help to play MONO file(.raw) using AIC34 audio codec
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steve Chen wrote: On Thu, 2009-03-12 at 18:55 +0530, DHAVAL LAKKAD wrote: Hi, Thank you very much for your reply. I would like to inform you that I am using OSS instead of ALSA. And in OSS, I have found this issue. I'm not familiar with the OSS driver. I suspect the reason mono plays much faster is because the mono data (1 channel) is being pumped out on both left and right channels, so it plays twice as fast. It would probably be faster and easier to get a newer kernel that has ALSA driver than hacking OSS driver to make it work. I know nothing about OSS either. However, I do remember threads about OSS not supporting mono. [1] is one of them. Regards, Caglar [1] http://linux.omap.com/pipermail/davinci-linux-open-source/2008-September/008275.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm5NhgACgkQ/nL+S5dojeizfACgm68eQZ/4vYSphuhVeSDJOly2 ndwAnjPhbpWMd0I/AgTsRJD1SXvB85SZ =JUlX -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Davinci, MUSB, DVEVM - driver but no device?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sergei Shtylyov wrote: Hello. Yusuf Caglar AKYUZ wrote: i'm trying to get an USB-stick running on a DVEVM-alike hardware on a recent kernel, but the USB host appears to be plain dead. I've digged a little into the kernel and my result is, that 'platform_driver_probe' in the end of 'musb_init' never ends calling 'musb_probe', thus resulting in the musb_driver being deregistered and ignored by the kernel. Digging deeper, the 'platform_driver_probe' enumerates the 'platform' bus but fails to find any suiting device. Thus, from the sources, it appears as if the construction misses a 'platform_device' representing the USB host controller. Because only the kernel and no hardware appears to be involved, it looks like a silly kernel configuration problem to me. I write to ask if someone knows the solution or could otherwise help. I can produce the result with 'davinci_evm_dm644x_defconfig'. You said DVEVM-alike, does it have i2c port expander as well for VBUS control? There's much more wisdom in controlling DRVVBUS via plain GPIO than this horrible GPIO expander thing. Absolutely. In fact we're using GPIO in our own boards and: 1) DRVVBUS control must be board specific. 2) Simply passing GPIO with platform data is not enough because polarity may even change. 3) I guess we don't need a workqueue for direct GPIO control. What is the best way to achieve these goals? Passing a function pointer via platform data to handle pin control at board* files, or passing GPIO number and polarity via platform data while using workqueue for all the cases? Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm5hpEACgkQ/nL+S5dojehz1ACfSh9HtsgnZbMh7+DFzpMQBqyo gigAn0l5FFeFd3UPvowoxnutFasktYGA =VbWz -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: DSPLINK 1.60 Oops'es on 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Yusuf Caglar AKYUZ wrote: Hi, I'm using dsplink 1.60 along with CE 2.21 and sometimes I face kernel oops messages. [1] Then the issuing process locks-up. The error always(up to now) shows-up while process is finishing. I debugged this issue a little deeper further down and I suspect that some mutex is tried to be locked after it is deleted. Hi again, I found my problem. I *think* dsplink incorrectly handles kernel signals, actually it incorrectly handles message queue timeouts. Also when it receives the signal, it tries to use a deleted mutex also. My current solution is attached to illustrate my situation. Regards, Caglar - --- sync_orig.c 2008-10-24 22:58:01.0 +0300 +++ sync.c 2009-03-10 12:49:04.0 +0200 @@ -30,9 +30,14 @@ /* --- OS Specific Headers */ +#include linux/version.h #include linux/autoconf.h #include linux/spinlock.h +#if LINUX_VERSION_CODE KERNEL_VERSION(2,6,27) #include asm/semaphore.h +#else +#include linux/semaphore.h +#endif #include linux/timer.h #include linux/sched.h #include linux/delay.h @@ -947,19 +952,25 @@ /* Schedule out for given timeout */ osStatus = schedule_timeout (MAX_SCHEDULE_TIMEOUT) ; /* Take the lock */ - -SYNC_SpinLockStartEx (semObj-lock) ; +//SYNC_SpinLockStartEx (semObj-lock) ; + if (signal_pending(current)) { + status = -ERESTARTSYS; + SET_FAILURE_REASON ; + } +#if 0 if (osStatus == -ERESTARTSYS) { /* Interrupted? huh? */ status = -ERESTARTSYS ; SET_FAILURE_REASON ; break ; } +#endif } while (0) ; /* Remove from wait list */ remove_wait_queue (semObj-list, wait) ; /* Set the current task status as running */ - -set_current_state (TASK_RUNNING) ; +//set_current_state (TASK_RUNNING) ; /* End the lock */ - -SYNC_SpinLockEndEx (semObj-lock, 0u) ; +//SYNC_SpinLockEndEx (semObj-lock, 0u) ; } else { if (semObj-semType == SyncSemType_Binary) { -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm2SaIACgkQ/nL+S5dojejIagCfc+Si9VdNghcYguUBODFeLnY1 AugAoLhy1ozDCJZREeUxd75uZmYMBZul =CGVj -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: CE ( or DSPLINK? ) Remote Node Creation Error 0x80008018
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Please ignore this message if you haven't already :) I found I had screwed-up ce libraries as you noted while fighting with a dsplink issue. Thanks, Caglar Yusuf Caglar AKYUZ wrote: Ring, Chris wrote: 0x80008018 is RMS_EINVPROT (EINVPROT == Error, invalid protocol), found in ti/sdo/ce/rms.h. This error occurs if the stubs (ARM-side) and skeletons (DSP-side) don't speak the same protocol. That is, the marshalling protocol of arguments in the stubs doesn't match the unmarshalling protocol of the arguments in the skeletons. In short, the version of Codec Engine built into the DSP-side doesn't match the version built into the ARM-side. This explains the weird behavior here. However, I can not find the source. I double checked everything, and still looking for what I'm doing wrong. One issue I should mention is I'm creating a CE shared library and using it in my applications. I stole this from video_copy/dualcpu_seperateconfig_dll. In that example it mentions using dlopen and friends. Instead of this, I'm linking *so with ld. I hope I'm not being too smart at this. During startup, when CE_DEBUG=2 is set, the versions of components in the system should be displayed, so you should be able to compare which versions were used on each side. I'm lost at version numbers. I ran ce sanity check application(out of the box without re-building) as: CE_DEBUG=2 ./app.out in.dat out.dat | grep dsplink @0,414,540us: [+4 T:0x4001cf40] OG - package dsplink.gpp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/cetools/packages/dsplink/gpp/) [5,0,0,1224878287000] @0,415,347us: [+4 T:0x4001cf40] OG - package ti.sdo.ce.ipc.dsplink (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/) [2,0,1,1226392278611] [DSP] @0,026,140tk: [+4 T:0x8fa4794c] OG - package ti.sdo.ce.ipc.dsplink.dsp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/dsp/) [2,0,1,1226392278316] [DSP] @0,026,532tk: [+4 T:0x8fa4794c] OG - package dsplink.dsp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/cetools/packages/dsplink/dsp/) [1,4,0,1224878295000] Then I ran my own application as: LD_LIBRARY_PATH=. CE_DEBUG=3 ./begtester -a test_mpeg4 | grep dsplink @0,744,829us: [+4 T:0x4001d310 S:0xbe952b9c] OG - package dsplink.gpp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/cetools/packages/dsplink/gpp/) [5,0,0,1224878287000] @0,752,839us: [+4 T:0x4001d310 S:0xbe952b9c] OG - package ti.sdo.ce.ipc.dsplink (/home/caglar/myfs/work/common/ti/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/) [2,0,1,1226392278611] [DSP] @0,058,051tk: [+4 T:0x8b80006c S:0x8b80ffe4] OG - package ti.sdo.ce.ipc.dsplink.dsp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/dsp/) [2,0,1,1226392278316] [DSP] @0,058,376tk: [+4 T:0x8b80006c S:0x8b80ffe4] OG - package dsplink.dsp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/cetools/packages/dsplink/dsp/) [1,4,0,1224878295000] I don't see what's wrong here. Anyway, I keep digging. You're explanations are pointing us* to the right direction as always. Thanks for your time. Caglar (*) People who are directed to dead ends like grepping the source of dsplink for 80008018 error!!! Chris ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm1cU4ACgkQ/nL+S5dojegP+wCfYlg2tdLD4Azxzlfe4uCHyCZU ShcAoLC0VXGm3sPeBkJ44ycOHqCYTWOd =1JNm -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
DSPLINK 1.60 Oops'es on 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm using dsplink 1.60 along with CE 2.21 and sometimes I face kernel oops messages. [1] Then the issuing process locks-up. The error always(up to now) shows-up while process is finishing. I debugged this issue a little deeper further down and I suspect that some mutex is tried to be locked after it is deleted. I made a one line change in ldrv_msgq.c but I haven't tried it yet because error is not deterministically shows itself. I'll update my module when I success 100% failure generation. Currently it is highly dependent on code-path and random, but always in the closing phase of an application. I wonder if there is a known issue or suggestions about the problem. Thanks for advance, Caglar [1] Oops message is: Unable to handle kernel NULL pointer dereference at virtual address 0003 pgd = c63ec000 [0003] *pgd=85349031, *pte=, *ppte= Internal error: Oops: 801 [#2] Modules linked in: dsplinkk cmemk CPU: 0Tainted: G D (2.6.28-davinci1-05792-g452c406-dirty #134) PC is at mutex_lock_interruptible_nested+0x114/0x2d4 LR is at mutex_lock_interruptible_nested+0x104/0x2d4 pc : [c031dfdc]lr : [c031dfcc]psr: 6093 sp : c5673de8 ip : c5673de8 fp : c5673e44 r10: c5c9c9e0 r9 : 80008051 r8 : r7 : r6 : 6013 r5 : c8fd4000 r4 : c5673e04 r3 : c8fd4020 r2 : 0003 r1 : c5673e04 r0 : c8fd4000 Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user Control: 0005317f Table: 863ec000 DAC: 0015 Process video_decode_di (pid: 973, stack limit = 0xc5672268) Stack: (0xc5673de8 to 0xc5674000) 3de0: 0002 bf018120 c031e2d0 c5673e54 bf018120 3e00: bf018bb0 c8fd4020 c5673e04 c8fd4000 c5673e04 c5c9c9e0 3e20: 8000 c5673edc cbcb3000 80008051 bf024d50 c5673e54 c5673e48 3e40: bf018120 c031ded8 c5673e84 c5673e58 bf01060c bf018114 bf01077c 6f0b 3e60: 6f0b bee2a9c4 c57220e0 bee2a9c4 c5672000 00047504 c5673e94 c5673e88 3e80: bf01b05c bf0105d4 c5673efc c5673e98 bf01b870 bf01b05c c5830738 3ea0: c536f248 c63d3d30 6093 c5672000 41765000 c002af24 3ec0: c5672000 8000 401b2940 0001 0010 3ee0: c536f238 bee2a9c4 6f0b 6f0b c5673f14 c5673f00 c009c900 bf01b590 3f00: c57220e0 0007 c5673f7c c5673f18 c009cd5c c009c8a8 c0060e70 c0060620 3f20: 0001 c57220e0 0007 c5673f84 0036 c002af24 3f40: c5673f7c c5673f50 c0091854 c0061814 0001 0007 bee2a9c4 3f60: 6f0b c57220e0 c002af24 00047504 c5673fa4 c5673f80 c009cde8 c009c948 3f80: c0082eac 0001 bee2a8f4 000472b4 bee2aa34 0036 c5673fa8 3fa0: c002ada0 c009cdb8 bee2a8f4 000472b4 0007 6f0b bee2a9c4 bee2a9c4 3fc0: bee2a8f4 000472b4 bee2aa34 0036 0001 000332dc 00047504 00033258 3fe0: bee2a8e0 bee2a8d8 0002fcb9 4025872c 8010 0007 Backtrace: [c031dec8] (mutex_lock_interruptible_nested+0x0/0x2d4) from [bf018120] (SYNC_SpinLockStartEx+0x1c/0x28 [dsplinkk]) [bf018104] (SYNC_SpinLockStartEx+0x0/0x28 [dsplinkk]) from [bf01060c] (LDRV_MSGQ_get+0x48/0xd0 [dsplinkk]) [bf0105c4] (LDRV_MSGQ_get+0x0/0xd0 [dsplinkk]) from [bf01b05c] (PMGR_MSGQ_get+0x10/0x14 [dsplinkk]) [bf01b04c] (PMGR_MSGQ_get+0x0/0x14 [dsplinkk]) from [bf01b870] (DRV_Ioctl+0x2f0/0x780 [dsplinkk]) [bf01b580] (DRV_Ioctl+0x0/0x780 [dsplinkk]) from [c009c900] (vfs_ioctl+0x68/0x78) r6:6f0b r5:6f0b r4:bee2a9c4 [c009c898] (vfs_ioctl+0x0/0x78) from [c009cd5c] (do_vfs_ioctl+0x424/0x470) r5:0007 r4:c57220e0 [c009c938] (do_vfs_ioctl+0x0/0x470) from [c009cde8] (sys_ioctl+0x40/0x64) [c009cda8] (sys_ioctl+0x0/0x64) from [c002ada0] (ret_fast_syscall+0x0/0x2c) r7:0036 r6:bee2aa34 r5:000472b4 r4:bee2a8f4 Code: e2853020 e5932004 e50b3040 e5834004 (e5824000) - ---[ end trace 79c33c778aec290b ]--- -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkm1dA4ACgkQ/nL+S5dojehCLgCgmtc6qnptlx8jk1wPm2eNFHOh OFQAnRTA60JdWMmJpQqULveGlPrVoXjD =tGiM -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
CE ( or DSPLINK? ) Remote Node Creation Error 0x80008018
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm trying DVSDK 2.0 combos with dm6446 but I'm getting error message: ...Engine_createNode Remote node creation FAILED (0x80008018). When I try CE_DEBUG=2, I do not see any other error/warning messages before this line. Both my (own)servers and dvsdk servers throw this error. I've built my own CMEM(2.21) and DSPLINK(1.6). To track down the issue, I run dsplink and codec engine samples/examples and they are running fine in my setup. However, I still suspect there is something wrong with my kernel modules. Any help would be invaluable given that I've run out of ideas here. Thanks in advance, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmy8UsACgkQ/nL+S5dojeha2ACdHw6uieWtqIZsAqqPDr74GPV/ 7noAnjftJ1iWuEbPTqd5s6GwUC8KHDfM =1o5E -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: CE ( or DSPLINK? ) Remote Node Creation Error 0x80008018
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ring, Chris wrote: 0x80008018 is RMS_EINVPROT (EINVPROT == Error, invalid protocol), found in ti/sdo/ce/rms.h. This error occurs if the stubs (ARM-side) and skeletons (DSP-side) don't speak the same protocol. That is, the marshalling protocol of arguments in the stubs doesn't match the unmarshalling protocol of the arguments in the skeletons. In short, the version of Codec Engine built into the DSP-side doesn't match the version built into the ARM-side. This explains the weird behavior here. However, I can not find the source. I double checked everything, and still looking for what I'm doing wrong. One issue I should mention is I'm creating a CE shared library and using it in my applications. I stole this from video_copy/dualcpu_seperateconfig_dll. In that example it mentions using dlopen and friends. Instead of this, I'm linking *so with ld. I hope I'm not being too smart at this. During startup, when CE_DEBUG=2 is set, the versions of components in the system should be displayed, so you should be able to compare which versions were used on each side. I'm lost at version numbers. I ran ce sanity check application(out of the box without re-building) as: CE_DEBUG=2 ./app.out in.dat out.dat | grep dsplink @0,414,540us: [+4 T:0x4001cf40] OG - package dsplink.gpp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/cetools/packages/dsplink/gpp/) [5,0,0,1224878287000] @0,415,347us: [+4 T:0x4001cf40] OG - package ti.sdo.ce.ipc.dsplink (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/) [2,0,1,1226392278611] [DSP] @0,026,140tk: [+4 T:0x8fa4794c] OG - package ti.sdo.ce.ipc.dsplink.dsp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/dsp/) [2,0,1,1226392278316] [DSP] @0,026,532tk: [+4 T:0x8fa4794c] OG - package dsplink.dsp (/db/atree/library/trees/ce-j10x/src/codec_engine_2_21/cetools/packages/dsplink/dsp/) [1,4,0,1224878295000] Then I ran my own application as: LD_LIBRARY_PATH=. CE_DEBUG=3 ./begtester -a test_mpeg4 | grep dsplink @0,744,829us: [+4 T:0x4001d310 S:0xbe952b9c] OG - package dsplink.gpp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/cetools/packages/dsplink/gpp/) [5,0,0,1224878287000] @0,752,839us: [+4 T:0x4001d310 S:0xbe952b9c] OG - package ti.sdo.ce.ipc.dsplink (/home/caglar/myfs/work/common/ti/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/) [2,0,1,1226392278611] [DSP] @0,058,051tk: [+4 T:0x8b80006c S:0x8b80ffe4] OG - package ti.sdo.ce.ipc.dsplink.dsp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/packages/ti/sdo/ce/ipc/dsplink/dsp/) [2,0,1,1226392278316] [DSP] @0,058,376tk: [+4 T:0x8b80006c S:0x8b80ffe4] OG - package dsplink.dsp (/home/caglar/myfs/work/common/ti/codec_engine_2_21/cetools/packages/dsplink/dsp/) [1,4,0,1224878295000] I don't see what's wrong here. Anyway, I keep digging. You're explanations are pointing us* to the right direction as always. Thanks for your time. Caglar (*) People who are directed to dead ends like grepping the source of dsplink for 80008018 error!!! Chris -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmzELQACgkQ/nL+S5dojei6uQCfQuD7xK4XUJU2EgzcqOvo76Ou cLEAoKKl9XKy7nxQ93Jrm9TK7wxl2Smd =T9i+ -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: How to stop the u-boot autoboot?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Lori Gao wrote: Dear All: When i set the bootdelay parameter to zero, the u-boot will boot OS without delay. But in this situation, i can't stop the u-boot into the command model even pressing ctrl+c. How can i keep no delay to boot OS and step into the command line model of u-boot if necessary? I tried different u-boot versions. Some gave me a commandline even with a 0 bootdelay if I press 'space' several times just at the booting, but some didn't. You must persist! My other approach is to delete linux kernel partition after the boot so that u-boot finds nothing to load and falls back to command line. Regards, Caglar Lori Gao RD center(Shanghai) Rayco (Shanghai) Medical Products Company Limited Carestream Health Inc. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmg+n8ACgkQ/nL+S5dojegoWgCfcNJt9GckIJtNwBNZzL7crqzv 9rMAn0FzASkn7JPm83hiFG8a6hxfOYYD =1UV8 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Kernel and FS download through USB to EVM/custom board's NOR flash
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Joshi wrote: Hi Swami, Thanks for the quick response. The serial download which I am doing it now for kernel and Ramdisk is taking around 20 min. Method I am following is loadb(via Kermit) followed by cp.b to NOR via HyperTerminal. So, I just thought some USB tool is already available to download (to NOR) from Linux. Please let us know, Is there any other way from U-Boot/Linux for faster downloading - From Linux: You can use a USB flash if you have NFS setup. 1) Select USB host mode(Kernel config, jumper, etc) 2) Boot from NFS 3) mount your usb flash(device node may need a change): mount -t auto /dev/sda1 /mnt 4) update kernel and ramdisk using jffs2 utils or like. You can also do these using USB gadgets, but I don't know the exact procedure. I hope I'm not getting your question wrong. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmdO8sACgkQ/nL+S5dojeiYQACgpC58I9K8sW9LaWg0M0iSPCAA gJUAnihRVp4+naTOMsMIGeiGqqWFU1rv =MvtV -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Kernel and FS download through USB to EVM/custom board's NOR flash
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Joshi wrote: Hi Caglar, Thanks for the information But in our custom board we have the USB port configured for device mode only with miniAB connector mounted. So I cannot configure as HOST I guessed this. One more limitation is I do not have Ethernet port :-( I was just thinking is there any small tool which can run on PC and download the kernel/FS to Flash? Yes there is. Have you looked at USB gadgets? You can make your PC show your board as an USB disk or ethernet adapter. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmdQwcACgkQ/nL+S5dojejcegCfSRzua19gqxpJOZjZIUbTrgk4 NmQAniONmwdajjt+d9JRQnKDbxDHbiU7 =nwYN -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Reg: Cross compiled ffplay on dvevm dm6446, problems..crashes..?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ragas sag wrote: Hi All, Hi, For the following messages' [...] Well finally( I hope ) i have some more question. When i ran ./ffplay sample.flv on the board, the video starts playing and after some time get the following kernal dumps, Division by zero in kernel. # Note this line Unable to handle kernel NULL pointer dereference at virtual address 0008 pgd = c0004000 [0008] *pgd= Internal error: Oops: 17 [#1] Modules linked in: 1 ssd CPU: 0 PC is at fb_videomode_to_var+0xc/0x68 LR is at fbcon_switch+0x150/0x580 You get NULL pointer dereference error in fb_videomode_to_var function in drivers/video/modedb.c . Also there is division by zero error. I really don't understand what is going on here. [...] And i think that --disable-armvfp ( above last line ) is leading to above generated problem. You don't want armvfp. It is for floating-point hardware which is not present in dm6446 ARM core. Because if i do --enable-armvfp ( Its selected by default, if not mentioned ) when i compile ffmpegs's libavcodec library , i am get the following messages as below, arm_v5t_le-gcc -DHAVE_AV_CONFIG_H -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -I.. -I/home/sagar/temp/ffmpeg/new_ffmpeg/ffmpeg -D_ISOC99_SOURCE -D_POSIX_C_SOURCE=200112 -I/home/sagar/temp/ffmpeg/new_ffmpeg/ffmpeg/SDL-1.2.13/sdl_arm/usr/local/include/SDL -std=c99 -fomit-frame-pointer -march=armv5te -g -Wdeclaration-after-statement -Wall -march=armv5te -mtune=arm926ej-s -Wno-switch -Wdisabled-optimization -Wpointer-arith -Wredundant-decls -Wcast-qual -Wwrite-strings -Wundef -O2 -fno-math-errno -c -o arm/dsputil_vfp.o arm/dsputil_vfp.S -- arm_v5t_le-gcc: ambiguous abbreviation -- arm/dsputil_vfp.S: Assembler messages: arm/dsputil_vfp.S:46: Error: bad instruction `vpush {d8-d15}' arm/dsputil_vfp.S:52: Error: bad instruction `vldmia r3!,{s0-s3}' arm/dsputil_vfp.S:53: Error: bad instruction `vldmia r1!,{s8-s11}' arm/dsputil_vfp.S:54: Error: bad instruction `vldmia r3!,{s4-s7}' arm/dsputil_vfp.S:55: Error: bad instruction `vldmia r1!,{s12-s15}' arm/dsputil_vfp.S:56: Error: bad instruction `vmul.f32 s8,s0,s8' arm/dsputil_vfp.S:59: Error: bad instruction `vmul.f32 s12,s4,s12' arm/dsputil_vfp.S:60: Error: bad instruction `vldmiage r3!,{s16-s19}' arm/dsputil_vfp.S:61: Error: bad instruction `vldmiage r1!,{s24-s27}' arm/dsputil_vfp.S:62: Error: bad instruction `vldmiage r3!,{s20-s23}' arm/dsputil_vfp.S:63: Error: bad instruction `vldmiage r1!,{s28-s31}' arm/dsputil_vfp.S:64: Error: bad instruction `vmulge.f32 s24,s16,s24' This is normal when you --enable-armvfp. What are the above error message's ? Are my configurations wrong ? To supress these above message i did --disable-armvfp . I suspect armvfp is with respect to floating point operation. And if i can compile ffplay with support for armvfp , may be the error crossing Division by zero in kernel.' ( Error mentioned above ). And may be the flv video file will just play fine. You can't use armvfp since there is no VFP hardware. Kindly help me with this problem. And if i am not clear with the question, ask me i will put it across in detail. And let get's this flv video on board playing ( ASAP) . And this will help others too. Why don't you try mplayer? Other than that, nothing comes to my mind. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmat1gACgkQ/nL+S5dojejAOQCgq/Q6xxqTO/lWQn2NvzXHUODI sWMAn0JVtojm1vUpJ1ZibuWIDCwm3eXR =EAhM -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dm6467 nand filesystem question
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 gather bzbz wrote: Hey, all Hi, I am trying to boot kernel and filesystem from the NAND on the 6467 EVM. Please help! 1. The nand partition is 0x-0x0008 : bootloader 0x0008-0x000a : params 0x000a-0x004a : kernel 0x004a-0x0800 : filesystem 2. The filesystem and kernel have been erased and programmed with nand erase and nand write from u-boot. nboot command finds the kernel section OK and continues the boot. bootargs is console=ttyS0,115200n8 noinitrd rw root=/dev/mtdblock3 I guess you have to tell kernel your file system type with rootfstype option. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmX4/QACgkQ/nL+S5dojeggEACdGChtdMbU9dGvEow4cm9RsMEA FlsAnAkEW4ArEt+O7is4LnKSXw3OnxJD =LTNs -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Reg: Cross compiled ffplay on dvevm dm6446, problems..crashes..?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ragas sag wrote: Hi, Looks like it not a problem with ffmpeg, as these are the message's that i am getting when we play an flv file using ffplay binary on DM6336. Following are the error messages that i am getting. Looks like its got something to with framebuffer or fv device. davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1152 864 , the fb check within limits is failing. That is it going beyound the available frame buffer size. 1) Do i need to change setting in SDL library w.r.t to fb device? 2) If anybody has faced similar problem, kindly guide us in solving the problem. I guess your fb resolution is not enough for your video size. Even 720x576 fails. What is your screen configuration? Regards, Caglar # ./ffplay sample.flv # FFplay version SVN-r16816, Copyright (c) 2003-2009 Fabrice Bellard, et al. configuration: --enable-cross-compile --enable-pthreads --arch=arm --logfile=config.err --cpu=armv5te --cross-prefix=/opt/mv_pro_40/montavista/pro/devkitb libavutil 49.14. 0 / 49.14. 0 libavcodec52.11. 0 / 52.11. 0 libavformat 52.25. 0 / 52.25. 0 libavdevice 52. 1. 0 / 52. 1. 0 built on Feb 4 2009 13:34:37, gcc: 3.4.3 (MontaVista 3.4.3-25.0.30.0501131 200davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 16 5-07-23) Found davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1408 1056 FB opening : /davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1280 1024 dev/fb/0. davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1152 864 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1024 768 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 960 720 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 800 600 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 768 576 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 720 576 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1600 1200 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1408 1056 davincifb davincifb.0: davincifb_check_var :within_vid0_limits fail. 0 0 1280 1024 Warm Regards, Sagar On Thu, Feb 12, 2009 at 1:33 PM, Vladimir Pantelic p...@nt.tu-darmstadt.dewrote: Ragas sag wrote: Hi, When i typed free command, i got the folllowing details as below, total used free shared buffers Mem: 118808 37584 81224 0 0 Swap:0 00 Total: 118808 37584 81224 I dont doubt about ffmpeg leaking memory. Because ffmpeg is used everywhere. If it had such problem, people would have come back with the problem. I tried to change the memory allocated for Linux from 56MB to 120MB. I have 256MB flash. Initially when i run it runs the flv file very well, only after some time..it gradually starts to crash.. so, it is leaking some memory over time and this might be specific to the dm6446 implementation. ffmpeg allocates all memory via av_malloc/ac_free in libavutil/mem.c, maybe you could add some traces there to see if there are large chunks that are allocated and never freed. Note: Any open source to play flv/flash files/vp6 encoded files or swf files on DM6446 exist? flv/vp6: ffmpeg swf: swfdec and gnash, but both barely work on X86. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmVWNkACgkQ/nL+S5dojegmOACfUIeyhssMJO4DZElj/JfXyaJW BMcAoLa75ydJ4U1rY6hCMnniSabHnT6A =72fP -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Reg: Cross compiled ffplay on dvevm dm6446, problems..crashes..?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ragas sag wrote: On 2/13/09, Yusuf Caglar AKYUZ caglarak...@gmail.com wrote: [...] 1) Do i need to change setting in SDL library w.r.t to fb device? 2) If anybody has faced similar problem, kindly guide us in solving the problem. I guess your fb resolution is not enough for your video size. Even 720x576 fails. What is your screen configuration? Screen resolution is 640x480 . Your video resolution must be higher than 640x480. I have ffmpeg + SDL library cross compiled and tested ffplay binary. This resolution will be set by ffmpeg or SDL ? Or is it that it's problem in my display drivers itself? I don't follow what you are trying to achieve with this setup. You cannot decode this resolution of video using arm side even if you adjust your fb resolution. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmVZC0ACgkQ/nL+S5dojejywQCfVl1LhHmrCxaecTRutMO9hSph jcEAoJCxSc8FLsRtn3xAHJByuJ+Vtm+p =mjf5 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Reg: Cross compiled ffplay on dvevm dm6446, problems..crashes..?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ragas sag wrote: On 2/13/09, Yusuf Caglar AKYUZ caglarak...@gmail.com wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ragas sag wrote: On 2/13/09, Yusuf Caglar AKYUZ caglarak...@gmail.com wrote: [...] 1) Do i need to change setting in SDL library w.r.t to fb device? 2) If anybody has faced similar problem, kindly guide us in solving the problem. I guess your fb resolution is not enough for your video size. Even 720x576 fails. What is your screen configuration? Screen resolution is 640x480 . Your video resolution must be higher than 640x480. I have ffmpeg + SDL library cross compiled and tested ffplay binary. This resolution will be set by ffmpeg or SDL ? Or is it that it's problem in my display drivers itself? I don't follow what you are trying to achieve with this setup. You cannot decode this resolution of video using arm side even if you adjust your fb resolution. But my video size is xres: 320 yres: 240 , Will it still be a problem for arm to decode this size resolution ? But why ffplay trying to adjust 1400x1152(and many more up to 720x576) Correct me if i am wrong in my approach. My objective is have ffmpeg+SDL enable, and using generated ffplay binary to play *.flv file or vp6 encoded video file on DVEVM / DM6446 I don't know about 320x240 video decoding performance arm dm6446 ARM core, but I don't think it will be possible. I can not say anything for sure. However, you have to find a way to tell ffplay which screen resoution to use. Maybe mplayer can help, it can render video onto framebuffer. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmVbkYACgkQ/nL+S5dojeiUTwCgsDEdVuDkcPjmsuOWdVFnSeLy k+sAoJogXp5MBpEgP3d8I+BZhJkjTXuQ =0cSp -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: i need your help for VPFE driver
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm adding mailing list to CC so people can correct me if I'm wrong. huan...@temobi.com wrote: I want to make tvp5146 capture interface work in 2.6.28-rc6 download from montavista git . There were some build errors related it 1. tvp5146 driver is present at drivers/media/video/tvp514x.c (Please check 2 if it is not) 2. VPFE drivers available in GIT tree is broken! Please do not use them. There are new drivers present in Kevin's staging tree for v4l2-capture. you should see this tree when you look-up remote branches in your git tree i.e. git branch -r 3. New vpfe drivers are building ok, but not working completely, for me. My modifications is attached in this mail. Let me know if you have any troubles. Regards, Caglar diff --git a/drivers/media/video/davinci/vpfe_capture.c b/drivers/media/video/davinci/vpfe_capture.c index 32084e2..de242d0 100644 - --- a/drivers/media/video/davinci/vpfe_capture.c +++ b/drivers/media/video/davinci/vpfe_capture.c @@ -313,7 +313,7 @@ static int vpfe_config_format(struct channel_obj *ch) vid_ch = (ch-video); common-crop.top = 0; - - common-crop.top = 0; + common-crop.left = 0; /* first get format information from the decoder. * if not available, get it from CCDC */ @@ -554,6 +554,8 @@ static irqreturn_t vpfe_isr(int irq, void *dev_id) common-curFrm-state = VIDEOBUF_DONE; wake_up_interruptible(common-curFrm- done); + common-curFrm-size = common-fmt.fmt.pix.width * + common-fmt.fmt.pix.height * 2; common-curFrm = common-nextFrm; } /* based on whether the two fields are stored @@ -979,8 +981,8 @@ static int vpfe_try_format(struct channel_obj *channel, if ((pixfmt-height == 0) (pixfmt-field == V4L2_FIELD_NONE)) pixfmt-height = 1; - - else - - pixfmt-height = 2; + //else + // pixfmt-height = 2; pixfmt-bytesperline = ROUND32(pixfmt-width * bpp); if (pixfmt-pixelformat == V4L2_PIX_FMT_NV12) @@ -2072,9 +2074,17 @@ static int vpfe_cropcap(struct file *file, void *priv, { struct vpfe_fh *fh = file-private_data; struct channel_obj *channel = fh-channel; - - struct v4l2_int_device *dec = - - channel-decoder[channel-current_decoder]; - - return vidioc_int_cropcap(dec, crop); + //struct v4l2_int_device *dec = + // channel-decoder[channel-current_decoder]; + struct video_obj *vid_ch = (channel-video); + struct v4l2_fract pa = VPFE_PIXELASPECT_PAL; + crop-bounds.top = crop-bounds.left = 0; + crop-bounds.width = vid_ch-std_info.activepixels; + crop-bounds.height = vid_ch-std_info.activelines; + crop-defrect = crop-bounds; + crop-pixelaspect = pa; + return 0; + //return vidioc_int_cropcap(dec, crop); } static int vpfe_g_crop(struct file *file, void *priv, @@ -2156,7 +2166,7 @@ static int vpfe_s_crop(struct file *file, void *priv, } else { dev_err(vpfe_dev, Error in S_CROP params\n); ret = -EINVAL; - - goto out; + goto lock_out; } lock_out: mutex_unlock(common-lock); @@ -2209,6 +2219,14 @@ out: return ret; } +static int vpfe_s_parm(struct file *file, void *priv, + struct v4l2_streamparm *a) +{ + /* FIXME: Not supported yet */ + printk(VPFE s_parm is not implemented yet\n); + return 0; +} + /* vpfe capture ioctl operations */ static const struct v4l2_ioctl_ops vpfe_ioctl_ops = { @@ -2236,6 +2254,7 @@ static const struct v4l2_ioctl_ops vpfe_ioctl_ops = { .vidioc_g_crop = vpfe_g_crop, .vidioc_s_crop = vpfe_s_crop, .vidioc_default = vpfe_param_handler, + .vidioc_s_parm = vpfe_s_parm, }; /* vpfe_probe : This function creates device entries by register -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmL+WAACgkQ/nL+S5dojeiTpQCfX6wAfx3277xfTMcACh8WWF84 XMoAn3Dpr/R5va0ek0MkPx7uIrToZOAm =sb1J -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: u boot and framebuffer interface
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Deepak Mundra wrote: On Wed, Feb 4, 2009 at 11:01 AM, Nitin Mahajan nitin...@yahoo.com wrote: HI! ___ From: Deepak Mundra deepa...@allaboutif.com To: davinci-linux-open-source@linux.davincidsp.com davinci-linux-open-source@linux.davincidsp.com Sent: Tuesday, February 3, 2009 19:11:30 Subject: u boot and framebuffer interface Dear All, How can i open frame buffer in u-boot and display a white screen(or bmp image).. What i want is ,After switching on board when u-boot executes i want to display something on screen ie before executing the kernel.. If some one has done this before please guide me through.. Check if this is what you are looking for? http://www.denx.de/wiki/view/DULG/UBootSplashScreen regards -Nitin Thax for the reply but i have gone through that already and i have enabled bmp command..but i was not able to enable bmp display because i requires LCD or VIDEO support to be enabled in u-boot .. which i am not able to do.. Has any one able to use bmp display command in u-boot ? I haven't seen any DaVinci framebuffer support in the u-boot. Last time I checked was 2 days ago. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmJUgQACgkQ/nL+S5dojeiejQCfTmazZ8dmdrFhv59kcJNhmx5o wsUAoLPDWbvGCCGeM+RC8ZgQ6G/HpIL8 =j+NS -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dm6446evm NAND Support Broken In 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 David Brownell wrote: On Monday 02 February 2009, Yusuf Caglar AKYUZ wrote: Is it possible to use CF? I haven't seen any references to it. I'm desperately searching for info on using CF. Let me ask while I found the specialist. I was told that it should work OK, if someone provides a bit of glue code. Basically it uses the IDE driver and runs the CF in true IDE mode ... in my schematics, sheet 28 suggests that most of the setup can more or less be done at board init time by tweaking existing code: - power up the CF slot (talk to the msp430) - take the CF out of reset (U35 GPIOs) - select CF instead of ATA (U35 GPIOs again) Wow, thanks. This is the kind of information I was looking for. I'll try to implement these. You could presumably get fancy and make it act like a CF-only PCMCIA slot, using the card detect logic to do whatever is needed. Once it works in true IDE mode, that is. I never thought this. I think it worths a try. I hope it's not much of a hack for me, though. Thanks again, Caglar - Dave -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmIDBsACgkQ/nL+S5dojegkpgCgh9EFCGcA0ZVroC4epdRGzsKO NgYAoJymZEMJLsELSHhlCeP3HpPKHfMK =YzPQ -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
dm6446evm NAND Support Broken In 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I was able to use NAND flash of dm6446 evm with 2.6.27, but I can't use it in 2.6.28 anymore. Checking davinci_probe in davinci_nand.c reveals that driver is looking for two nand resources whereas evm has only one. I guess this change was introduced for 355 evm support. What is the correct way to fix this? I guess necessary change is not trivial for me, so any help or suggestion will help a lot. Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmHbfoACgkQ/nL+S5dojejNnwCgvL76og1F5hIf9aQI3+0/bJ6+ dJAAnishdt7YsRrTBKtvhKpnn1a5hu8S =+U2u -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: dm6446evm NAND Support Broken In 2.6.28
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 David Brownell wrote: On Monday 02 February 2009, Yusuf Caglar AKYUZ wrote: I was able to use NAND flash of dm6446 evm with 2.6.27, but I can't use it in 2.6.28 anymore. Checking davinci_probe in davinci_nand.c reveals that driver is looking for two nand resources whereas evm has only one. I guess this change was introduced for 355 evm support. Code in board-dm644x-evm.c has two resources. I updated my tree after your post. Now I see. Sorry for the noise. Are you confusing between NOR and NAND? Or maybe not booting right? Recall there are two ways to get NAND working: - Board reprogrammed to work boot from NAND, and you never touch the jumper after setting it to boot NAND. - Boot from NOR (no board reprogramming) into u-boot. Before booting into Linux, switch the jumper so CS0 is NAND instead of NOR. Now boot into Linux. Also, only configure one of: IDE, NAND, NOR. The first of those you configure will be the only one that works, at least in current GIT. Is it possible to use CF? I haven't seen any references to it. I'm desperately searching for info on using CF. Let me ask while I found the specialist. Thanks, Caglar - Dave -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmHoC8ACgkQ/nL+S5dojejCbQCghN/3nK8ESEgGLo3B6WROUrNf M5sAnRtJtdxLskBQgxvzuwhaEoS/z+7o =2DTK -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: 2.6.28 and VPFE Drivers
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Karicheri, Muralidharan wrote: Caglar, I am a bit confused here. I guess I already send you a patch(and Kevin also) to make it working with 2.6.28! I'm using your capture drivers at the moment with my 2.6.28 tree. I had a version sent to this mailing list and also to Kevin that was ported to work with v4l2-int-device.h interface. This was developed when initial version of tvp514x driver was sent to community for review by Vaibhav from TI. My version was for 2.6.26. When you say I'm using your capture drivers at the moment, Are you referring to this patch or your own version based on the davinci_vpfe driver existing in the davinci-git which is quite old. Yes I'm referring to that patch. I'm using the version you posted to this list in November(I guess). Current one in GIT tree is quite old, as you said. From your version Kevin created a staging tree. I guess you haven't seen my mail at [1]. Thanks, Caglar [1] http://linux.omap.com/pipermail/davinci-linux-open-source/2009-January/010681.html -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmBhwIACgkQ/nL+S5dojehhrQCgvb3W6xguEOpEfEOEWyBMfWpg IusAnjDCp8uT+Cq1a3pJzuhTQ4+A37iC =8t7r -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: DM6446 MUSB
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Subbrathnam, Swaminathan wrote: Brian, Could you provide more details to the failure ? There are sometimes reset errors with the PIO mode also. Caglar Regards swami -Original Message- From: davinci-linux-open-source-boun...@linux.davincidsp.com [mailto:davinci-linux-open-source-boun...@linux.davincidsp.com] On Behalf Of Brian Rhodes Sent: Thursday, January 22, 2009 11:46 PM To: davinci-linux-open-source@linux.davincidsp.com Subject: DM6446 MUSB Hello, Is PIO ONLY the only working USB mode in the current GIT kernel? By working, I mean without periodic reset. PIO is suitable for my application, but I just merged the latest GIT code into our kernel tree and without CONFIG_MUSB_PIO_ONLY=y USB does not work very well, so I was curious if I had possibly overlooked something. Does the CPPI DMA mode work well on DM6446 in the current GIT code? musb_hdrc: version 6.0, pio, host, debug=0 musb_hdrc: ConfigData=0x06 (UTMI-8, dyn FIFOs, SoftConn) musb_hdrc: MHDRC RTL version 1.300 musb_hdrc: setup fifo_mode 2 musb_hdrc: 7/9 max ep, 2624/4096 memory musb_hdrc: hw_ep 0shared, max 64 musb_hdrc: hw_ep 1tx, max 512 musb_hdrc: hw_ep 1rx, max 512 musb_hdrc: hw_ep 2tx, max 512 musb_hdrc: hw_ep 2rx, max 512 musb_hdrc: hw_ep 3shared, max 256 musb_hdrc: hw_ep 4shared, max 256 musb_hdrc: USB Host mode controller at c800 using PIO, IRQ 12 musb_hdrc musb_hdrc: MUSB HDRC host driver musb_hdrc musb_hdrc: new USB bus registered, assigned bus number 1 Thanks. ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmB3SgACgkQ/nL+S5dojejLwACbBPd+Y9P8vrJMU4fZ/sGdwih8 oGcAn3aeiRcDb0nzAYasAlmMipi24/5K =qi2a -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: CMEM/DSPLink and 2.6.28 kernel?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Ryan Talbot wrote: I searched the list, but couldn't find any evidence of patches to get CMEM or DSPLink working with davinci-git kernels 2.6.26. We're currently using 2.6.28-rc6-davinci1, CMEM from TI's LinuxUtils 2.21 package, and DSPLink 1.60. CMEM at least looks like it needs some patches to work with 2.6.28 kernels; I haven't tried building DSPLink yet. Are there no patches available? I just wanted to make sure I wouldn't be reproducing work, if there were. You can find all the necessary patches in the openembedded tree. I worked with dsplink 1.50, dsplink 1.60 and 1.30 with latest kernels. CMEM is the well-behaved child when compared to dsplink. It was always easy to make it work, though I haven't tried latest CMEM. Regards, Caglar Thanks, Ryan Talbot rtal...@vtti.vt.edu ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmCKXgACgkQ/nL+S5dojeiHlwCePLMuW9ifZdod0oJuxdvo58eT 0fAAnj5goTVustlHL6LT9ssHjRXrSd+P =lez1 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: 2.6.28 and VPFE Drivers
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steve Poulsen wrote: Caglar, The state of those two files is such that they are not ready to be used. V4L2 has changed enough that you will need to dig in and understand the drivers enough to make the changes. For example, the videobuf code underwent an overhaul such that videobuf_init now needs to be changed to videobuf_queue_sg_init. There doesn't need to be too many changes to these files to get them to work, but to make the changes, you will need to understand these at a level that you probably didn't expect to. You are a little late, I'm already beaten up to death by the vpfe code :) Unfortunately, all you said is correct and your last sentence is the overview of my whole weekend. After two days of meaningless porting efforts I realized it was not for the weekend player. Later, I found Kevin's staging tree and I'm using it at the moment. Thanks for the info, I wish I had them earlier. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmA23YACgkQ/nL+S5dojejLKwCgsjfyUAhbdreefGp361+CvHHk II8An1Mbom7YfCeg8eFazEOFoQs538WA =TE6z -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: 2.6.28 and VPFE Drivers
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Karicheri, Muralidharan wrote: Steve Caglar, We are already working to port the latest video drivers (capture first and then display) on DM6446 to open source kernel. I had a version of capture driver that is working with 2.6.26 kernel and I need to up port the same to 2.6.28. So please stay tuned in this mailing list for the patch that I will be submitting to the v4l2 community . I guess I already send you a patch(and Kevin also) to make it working with 2.6.28! I'm using your capture drivers at the moment with my 2.6.28 tree. I'm also using display drivers from latest 2.6.10 MV kernel which I guess you are referring to. I'm also happy with them. Thanks for all these. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkmA3MIACgkQ/nL+S5dojehVPgCdGZtCKvg1KZjMMFUldN5KW4dz O5YAnReEzLz5v0txyIByqJoQ91WOx1dW =IPHC -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: DM355.....
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Arvind wrote: Hi, I am using DM355 board I want to use DM355 serial port0 serial port1. How can i use this.. my Code: #include fcntl.h #include stdio.h #include stdlib.h #include sys/types.h #include unistd.h int main(void){ int fd; //char buf[10]; //fd = open(/dev/ttys0,O_RDONLY); fd = open(/dev/ttys0,O_WRONLY); this must be /dev/ttySO Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl+xzAACgkQ/nL+S5dojejrkgCeNUyybEYIcKgTsAFvW89d7a8J EywAoIU7oz4iJWQlrcx12fQVw63Y8lYL =q0L4 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: [PATCH 1/5] add vpfe capture driver support
Kevin Hilman wrote: m-kariche...@ti.com writes: This patch adds vpfe capture driver support. This is a new driver based on v4l2-int-device.h interface for attaching slave decoder device to v4l2 master device. This replaces the old davinci_vpfe.c and davinci_vpfe.h. Files have been renamed to reflect its usage across multiple platforms that has a VPFE to capture video from yuv/raw decoders. This uses an interface, defined in ccdc_hw_if.h, to configure ccdc in vpfe to work with the decoder attached to the interface. Hi Murali, This is great work, thank you! While we wait for this to go to the V4L2 list for discussions, I've pushed this series and some of my own changes into a staging branch on DaVinci git called 'staging/v4l2-capture' which is based on DaVinci git updated to v2.6.28-rc6. Hi Kevin, Murali, I tested this driver on EVM and it is working flawlessly. Thanks for this new vpfe driver. There are some minor issues with driver: 1. Driver returns cropcap ioctl if mutex locked if operation fails. 2. s_parm ioctl is not implemented. However, gstreamer plugins rely on this iotcl. I don't know if other user space programs require this or not. 3. I guess capture height is incorrectly set to 2 always in try_fmt ioctl. 4. cropcap ioctl always returns einval. I had to change it a little bit. 5. crop rectangle left coordinate is not set, instead top is set twice. My diff against staging/v4l2-capture branch is attached. I'm fairly sure they are not in an appropriate shape They're just trivial hacks to make driver functional at the moment. Thanks, Caglar diff --git a/drivers/media/video/davinci/vpfe_capture.c b/drivers/media/video/davinci/vpfe_capture.c index 32084e2..de242d0 100644 --- a/drivers/media/video/davinci/vpfe_capture.c +++ b/drivers/media/video/davinci/vpfe_capture.c @@ -313,7 +313,7 @@ static int vpfe_config_format(struct channel_obj *ch) vid_ch = (ch-video); common-crop.top = 0; - common-crop.top = 0; + common-crop.left = 0; /* first get format information from the decoder. * if not available, get it from CCDC */ @@ -554,6 +554,8 @@ static irqreturn_t vpfe_isr(int irq, void *dev_id) common-curFrm-state = VIDEOBUF_DONE; wake_up_interruptible(common-curFrm- done); + common-curFrm-size = common-fmt.fmt.pix.width * + common-fmt.fmt.pix.height * 2; common-curFrm = common-nextFrm; } /* based on whether the two fields are stored @@ -979,8 +981,8 @@ static int vpfe_try_format(struct channel_obj *channel, if ((pixfmt-height == 0) (pixfmt-field == V4L2_FIELD_NONE)) pixfmt-height = 1; - else - pixfmt-height = 2; + //else + // pixfmt-height = 2; pixfmt-bytesperline = ROUND32(pixfmt-width * bpp); if (pixfmt-pixelformat == V4L2_PIX_FMT_NV12) @@ -2072,9 +2074,17 @@ static int vpfe_cropcap(struct file *file, void *priv, { struct vpfe_fh *fh = file-private_data; struct channel_obj *channel = fh-channel; - struct v4l2_int_device *dec = - channel-decoder[channel-current_decoder]; - return vidioc_int_cropcap(dec, crop); + //struct v4l2_int_device *dec = + // channel-decoder[channel-current_decoder]; + struct video_obj *vid_ch = (channel-video); + struct v4l2_fract pa = VPFE_PIXELASPECT_PAL; + crop-bounds.top = crop-bounds.left = 0; + crop-bounds.width = vid_ch-std_info.activepixels; + crop-bounds.height = vid_ch-std_info.activelines; + crop-defrect = crop-bounds; + crop-pixelaspect = pa; + return 0; + //return vidioc_int_cropcap(dec, crop); } static int vpfe_g_crop(struct file *file, void *priv, @@ -2156,7 +2166,7 @@ static int vpfe_s_crop(struct file *file, void *priv, } else { dev_err(vpfe_dev, Error in S_CROP params\n); ret = -EINVAL; - goto out; + goto lock_out; } lock_out: mutex_unlock(common-lock); @@ -2209,6 +2219,14 @@ out: return ret; } +static int vpfe_s_parm(struct file *file, void *priv, + struct v4l2_streamparm *a) +{ + /* FIXME: Not supported yet */ + printk(VPFE s_parm is not implemented yet\n); + return 0; +} + /* vpfe capture ioctl operations */ static const struct v4l2_ioctl_ops vpfe_ioctl_ops = { @@ -2236,6 +2254,7 @@ static const struct v4l2_ioctl_ops vpfe_ioctl_ops = { .vidioc_g_crop = vpfe_g_crop, .vidioc_s_crop = vpfe_s_crop,
Re: emdebian and Ubuntu?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Andrea Gasparini wrote: Ben West spiffera, alle Wednesday 21 January 2009 circa: However, the decision by Ubuntu maintainers above now means I must set up parallel Debian system. Has anyone else encountered this problem with Ubuntu and emdebian? Uhm. I never tried, but can't you simply install the Debian package? ( I mean, downloaded from _debian_ repository, not ubuntu one ) Usually, with some little hacks, there's no need to have various Ubuntu/Debian systems running. Anyway, I think you'll ear about me , as I'd like to give emdebian a try, and kick out MV filesystems. Bye Hi, Recently openembedded folks published an online image builder at [1] for various targets including davinci-dvevm. In case you're interested in a new fs image... Regards, Caglar [1] http://amethyst.openembedded.net/~koen/narcissus -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4Q6sACgkQ/nL+S5dojegR2wCeIK3t7TftG0YOQi4+eEy0kZAs AIYAn0U3qvjEtZvQZSjq2dcPnD9wvDGm =YGSw -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
2.6.28 and VPFE Drivers
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm trying to make tvp5146 capture interface work in 2.6.28 with DVEVM. There were some build errors related with davinci_vpfe.c, I fixed them. Current situation is that a simple application can open video device, query capabilities and perform some IOCTLs. It can even detect my PAL analog camera. However, capturing/streaming is not working due to some bugs in the driver, and my kernel crashes. I had to make some functional changes in the driver to fix build errors, and I suspect I need to do more. But I'm not sure if I'm on the correct route or not. Am I missing something or video capture interface at the moment is broken with 2.6.28? [1] Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4VcwACgkQ/nL+S5dojejmWQCfW5/CiAdh9ppNUWl+gey5nPe4 8J8An3A5j1d4wcf7/fgMyreOq/Z7qSiP =52Mw -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Resizer Initialization
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steve Ressler wrote: Hello, I'd like to use the resizer in my application and quick boot time is also very important to me. In order to reduce boot time, I am bypassing the init scripts with init=/bin/bash in the bootargs. When I bypass the scripts, I dont get a device file in /dev for the resizer. I've tried creating it myself with mknod /dev/davinci_resizer c 254 0 but the resizer doesn't seem to work. Is there something I am missing? Are you sure that resizer takes major number 254? I guess it uses dynamic number allocation. Caglar Is it possible to use the resizer without running all the init scripts? Why not? Thanks, Steve ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4m7wACgkQ/nL+S5dojeiVdQCgn2Zd428ADkY0Lgf2uqxSB110 nz8AnjPY+5B+lRV+vjXBE/pmJtbP/7wV =aVLU -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Resizer Initialization
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Steve Ressler wrote: Hi, Are you sure that resizer takes major number 254? I guess it uses dynamic number allocation. I believe you are correct, it uses dynamic number allocation. 254 was the number I saw it using when the device file was created by the init scripts. How can I find the correct major number? I'm not running the init scripts, so there is nothing in /proc to check. Maybe I should have asked: How can I use the resizer without running the init scripts? AFAIK, you cannot check device numbers without proc or sysfs. Some options coming to mind: 1) Edit resizer driver drivers/char/davinci_resizer.c Change line: result = alloc_chrdev_region(dev, ZERO, 1, DRIVER_NAME); to: dev = MKDEV(254_or_some_number_unique_in_the_system, ZERO); result = register_chrdev_region(dev, 1, DRIVER_NAME); Don't forget to create node for the number you choose as major. 2) Create a file in your filesystem and edit, let's say /home/my_init.sh: #!/bin/sh mount /proc RES=$(cat /proc/devices | awk '/davinci_resizer/ {print $1}') rm -f /dev/resizer mknod /dev/resizer c $RES 0 /path/to/your/app/app Then start this script as init, instead of your application. 3) You can also edit your inittab. Instead of running init scripts you can just mount some filesystems and then start your application. I don't use MV fs and kernel, so I can't try these. But I guess they may give some ideas. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4mH4ACgkQ/nL+S5dojeht1QCfVRqG0HxB7zqUIs6c2l4QRhrC /3YAoJb573wRtHKDi0ArFEqk83wrOKI5 =3nas -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: omap-L1xxx
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Sergei Shtylyov wrote: Hello. Hello, Besides, how are you using DaVinci? Asking because OMAP-L1xx/ and DA8xx both seem to be targeted to the different market niches than DaVinci -- at least the former completely lack the video support that the later offered... Video applications, industrial control(mainly HMI) and data acquisition. SOC can be 6446, 355 or 335 depending on the cost/performance criteria. I agree that these are not primary areas for OMAP-L1xx but I'm sure it will be offered to many portable applications. Besides, it's unique in one way compared to DaVincis, a floating point dsp which offers developers more flexible programming. I'll not be surprised if many people prefer OMAP-L1xx to DaVinci with these compatible price levels. Best regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4rh4ACgkQ/nL+S5dojegDFQCeKD/16FQkAZWsACCIJJZOVqzj 32QAoKvQ2J79knrgT93Fh5SX0EHJO/zs =6in4 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: emdebian and Ubuntu?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Philip Balister wrote: Yusuf Caglar AKYUZ wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, Recently openembedded folks published an online image builder at [1] for various targets including davinci-dvevm. In case you're interested in a new fs image... If you use this this, please let us I know how it works for you. I can make sure any emails on this list get to the right people. Thanks, I don't think that I'm going to use this one, because I'm using OE for most of my development cycle already. OTH, I can say that this type of ready/to/use/customizable fs image tool is invaluable for anyone. I wouldn't be learning openembedded if this was published some months ago :) And some comments about general OE/Angstrom-DVEVM experience: 1) Highly customizable fs and tools for DaVinci, in a very easy way. 2) Support for dsplink and cmem modules for the DSP side. Given that building and running dsplink for latest kernels is virtually impossible for average user, this is life-saving . 3) Last but not the least, powerful staging environment for development. If you happen to compile/build many different 3rd applications, you have all the libraries, sources, etc. at your hand. Thanks for this great job. Best regards, Caglar Philip ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl4sx8ACgkQ/nL+S5dojegPAgCaAixqaDhzN7lsd94ccr/cZ64a 19sAniQp32+Ww5SjFTncShrjqOiQAQQj =aq6w -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: omap-L1xxx
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Kevin Hilman wrote: Sergei Shtylyov sshtyl...@ru.mvista.com writes: Absolutely. Here's an example of enforcing your private opinion... Yes, this is my private opinion, but I am not enforcing anything. I am attempting to share the technical arguments behind my private opinion so folks know where I'm coming from. As the person who has to maintain a single code base that will run on all these platforms, I have a *very* strong desire to have a single binary run across multiple SoCs. May be sholld also ask tbe opinion of TI and the real users? I am very eager to hear more opinions. That's what this list is all about. We have some DaVinci based boards here, using single kernel for all is quite time-and-effort saver for us. It would be really nice to be able to boot the same kernel(s) in later designs as well, given that the arch is similar. Also isn't a single kernel for multiple archs is the better for auto-builders? I also quickly looked into arm/arch/mach-* directories and I see that some archs choose one mach-* folder, some many. However, it is seen that former group is well maintained than the latter. So for the sake of future code quality I hope you don't choose to differentiate code of SOCs under discussion. Best regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl3guwACgkQ/nL+S5dojejJlgCgm6r1XubIofZP9kLO98W8HaXk E84An3qa9NRXRfpzrdLKOKqTUcDTDjoK =nLy7 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Testing TI Production Codecs(MPEG4 Decoder) for dm6446
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm trying TI productions codecs for dm6446. The codec in question is mpeg4 sp decoder.(document no spruea3a.pdf) I'm able to decode video source some time i.e. ~1 min. But then VIDDEC_process fails with error code 0x408 or 0x401. If I continue decoding after these error the whole system(ARM + DSP) deadlocks. If I stop decoding at this point, I can re-start decoding operation. Moreover, if I randomly seek video file, codec also returns with same error. I think I'm seeking into the middle of an mpeg4 frame but I think I should be able to recover from this insufficient seek somehow but don't know how. I use seeking method to re-generate my problem quickly. A couple of questions: 1) What can be the original problem? (codec process error after some time) 2) How can query the codec about how many bytes should I seek so that I'll reach to an mpeg frame boundry? I'm modifying the example program delivered with the codec, but still the same problem. 3) How can I increase stack size of algorithm in CE 1.10? I already increased system thread stack size which made my system more robust in terms of deadlocks. 4) I know this is fantastic. But is it possible to make an old codec(prepeared with CE 1.10) work with a new CE 2.10+ setup on ARM side? Also updating dsplink and cmem. My current system configuration: dm6446 + dsplink 1_30_08 + dsp/bios 5_32_04 + cmem 1_02 + GIT Kernel Any help would be appreciated, Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkl0wuEACgkQ/nL+S5dojeiz2gCePLv15aZTZAjaOMfTv3TD+SWA 4NwAoKGxsKahHaqTcoW50tzc31gXHuj4 =uMtC -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: GStreamer, fbvidesink and GIT Kernel
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Venkatachala Upadhya wrote: HI, This should be /dev/fb/3 for fbvieo sink instead of /dev/fb0 fbvideosink device=/dev/fb0 Do you mean fbvideosink doesn't work with fb0? Because I tried fb3 also. The example was to demonstrate that fbdevsink works with fb0 but fbvideosink is not. BTW, I didn't understand why TI has written a seperate plug-in for a fb sink, because already implemented fb sink in gstreamer tree(fbdevsink) is working well(until now) Maybe I'm missing the design requirement, any clues? Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAklnMugACgkQ/nL+S5dojeh5SQCeKi5Xh+xy/Dt7xk8ZGcyPCLY8 1icAn3ijtvQIN9fvtR2Sv22oMhnXXgvL =qZJA -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
GStreamer, fbvidesink and GIT Kernel
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi all, My problem is very specific to gstreamer ti plugins' code. I hope someone is still maintaining those codes and following this list. I'm trying to get started with gstreamer on git kernel 2.6.27-rc6. I'm trying to get ti-plugins working with this release, especially fbvideosink. I ported latest fb driver code from montavista kernel so I expect fbvideosink to work as it is with MV kernel. My problem is screen depth, width and height are 0 in gst_fbvideosink_setcaps function. From the debug outputs I see that gst_fbvideosink_fbimage_new function is never called and it is the only place that those variables are set. So pipeline in [1] is not working. However in [2] is working as expected and I see the video on the screen. I know nothing about gstreamer and trying to find the source of error. Any help,light,clue would be very helpful. Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAklk3AoACgkQ/nL+S5dojejeAACgj2jxLQ/YHQk/zpkPOpwhIYS6 vDQAn26D1vDgT/EtoLyMflfzb3qxjjsX =D7MX -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: GStreamer, fbvidesink and GIT Kernel
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Yusuf Caglar AKYUZ wrote: Hi all, My problem is very specific to gstreamer ti plugins' code. I hope someone is still maintaining those codes and following this list. I'm trying to get started with gstreamer on git kernel 2.6.27-rc6. I'm trying to get ti-plugins working with this release, especially fbvideosink. I ported latest fb driver code from montavista kernel so I expect fbvideosink to work as it is with MV kernel. My problem is screen depth, width and height are 0 in gst_fbvideosink_setcaps function. From the debug outputs I see that gst_fbvideosink_fbimage_new function is never called and it is the only place that those variables are set. So pipeline in [1] is not working. However in [2] is working as expected and I see the video on the screen. I know nothing about gstreamer and trying to find the source of error. Any help,light,clue would be very helpful. I forgot to write write my pipelines, adding now. Sorry for the noise. [1] gst-launch videotestsrc ! video/x-raw-rgb ! fbvideosink device=/dev/fb0 [2] gst-launch videotestsrc ! video/x-raw-rgb ! fbdevsink device=/dev/fb0 Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAklk3bcACgkQ/nL+S5dojegCVgCeN7NCBoqTqn7Qfrm+VHygsvn8 U00An0niqb+40zvxA/KZbHiVTvlU5Czg =3nU4 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: starting DSP server './decodeCombo.x64P' FAILED
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Mahajan wrote: HI! I have some queries related to the DSP server, when it is present in NAND flash file system YAFFS2. Hi, What is the source of YAFFS2 image and which kernel version are you using? Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkljWl4ACgkQ/nL+S5dojegZJACeLZkUMlkwZGlOfVsuLLUrMkGT upgAoJj8e+UHaI3EcaQG3bUxNFvTOOgW =BKJw -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: starting DSP server './decodeCombo.x64P' FAILED
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Mahajan wrote: HI! - Original Message From: Yusuf Caglar AKYUZ caglarak...@gmail.com To: Nitin Mahajan nitin...@yahoo.com Cc: davinci-linux-open-source@linux.davincidsp.com Sent: Tuesday, January 6, 2009 18:49:26 Subject: Re: starting DSP server './decodeCombo.x64P' FAILED -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Mahajan wrote: HI! I have some queries related to the DSP server, when it is present in NAND flash file system YAFFS2. Hi, What is the source of YAFFS2 image and which kernel version are you using? I am using 2.6.18 kernel and the YAFFS2 roofs image I created using help on this wiki link http://wiki.davincidsp.com/index.php?title=Create_a_JFFS2_Target_Image. I am using the dsplink version 1.50, I created the basic rootfs from ramdisk image and copied the DSP related binaries to this rootfs. Everyting works fine untill I am on NFS or ext2. Once the same rootfs is brought to YAFFS2, I get the problem. So your NFS and YAFFS2 images from the same source. Then I can not help. I hit the same problem while I was trying 1.30 with GIT kernel. The result came out to be EABI/OABI related. However, 1.30 is not compatiple with newer kernels but 1.50 is. I guess your problem is something different. regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkljZSgACgkQ/nL+S5dojei0vgCfSSYCGaipxwf/TdoushWzfIR5 Z5kAn0+QP3URZKRChvR/p6l7DbIhOXFG =IiQk -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: starting DSP server './decodeCombo.x64P' FAILED
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Mahajan wrote: I am using the dsplink version 1.50, I created the basic rootfs from ramdisk image and copied the DSP related binaries to this rootfs. Everyting works fine untill I am on NFS or ext2. Once the same rootfs is brought to YAFFS2, I get the problem. So your NFS and YAFFS2 images from the same source. Then I can not help. I hit the same problem while I was trying 1.30 with GIT kernel. The result came out to be EABI/OABI related. However, 1.30 is not compatiple with newer kernels but 1.50 is. I guess your problem is something different. Still could you please give some some more info on your problem? May be I can get some pointers from it. I found the source of problem by enabling debug in the dsplink. The problem was related to sys_lseek call somewhere in dsplink code. There was an excellent post from Troy Kisky about this at [1]. Regards, Caglar regards -Nitin regards, Caglar Get your preferred Email name! Now you can @ymail.com and @rocketmail.com. http://mail.promotions.yahoo.com/newdomains/aa/ -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkljaX4ACgkQ/nL+S5dojeioLQCfaepvuum4sbmTyQpBFwyGFbks 7nEAoKyiQhi8CxzCMVhTYjqlos2otcFw =BHvj -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: starting DSP server './decodeCombo.x64P' FAILED
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Jérôme HUET wrote: Hello, I have the same problem when I want to use dsplink 1.40 with the latest kernel from the git repository. How did you manage to avoid this error? Did you follow the remarks at post [1] and related mails? Other than those I did nothing special. I can not help you more, but I can send you my dsplink 1.30 or 1.50 sources against 2.6.27-rc6. Or better, I can send you my openembedded recipes. [1] http://www.mail-archive.com/davinci-linux-open-source@linux.davincidsp.com/msg04395.html Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkljbJsACgkQ/nL+S5dojej2CgCgmrhraMhs38igE7DGPxC/y/oX V58AnR9Cug6Hir5vM8+Yad474Q0QRDxS =C4pn -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: starting DSP server './decodeCombo.x64P' FAILED
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Nitin Mahajan wrote: Thanks, I googled out this post in the mean time and went thru it. I have one question in general, whether all your DSP related software works well with the recent git kernel maintained by Montavista? I tested dsplink, codec engine, cmem and some codecs. However I haven't verified nand flash system yet. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkljcZQACgkQ/nL+S5dojehV1ACguTxpPj5gaPHz2Qpqn/db3qBo oOsAn2Iab/N/j/O36JSytxvifQlVMqKE =BqbK -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Strange problem with Linux git version. lcd crashes ttyS0 console!
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 BlackSword wrote: Hi all, I need you help! I meet such a problem. My system works perfect with original LSP1_20 for lcd. And with the git version it works perfect for component signal. But when i tried bootargs video=davincifb:output=lcd. problem occur. stopped at below shows. Try with mem=120M Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkkHShAACgkQ/nL+S5dojeh3KQCfSuHJr2nkcMNsnhYsfzNrwkpQ zn8AnRgW920OQzD10UvjwuHrfH+23Hk+ =UPKu -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: Filesystem for NOR on 6446
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 jp wrote: Hi, Hi, I guess you're starting a new thread by replying to another thread. If this is the case, please do not because threads are mixing-up this way. I still have some problems to create a filesystem to put it on the NOR (16MBytes). I only want to have the encode/decode application and , perhaps i am wrong, i want to use EXT2 for the filesystem. What i am doing is to begin with the default filesystem (but it is too big to put it on the flash) but when i removed some components we have a kernel panic at the beginning. Not all of them is meant to be removed then... Starting kernel ... Uncompressing Linux done, booting theLinux version 2.6.10_mvl401-davinci_evm ([EMAIL PROTECTED]) (version gcc 3.4.3 (MontaVista 3.4.3-25.0.104.0600975 2006-07-06)) #11 Wed Oct 22 16:51:48 CEST 2008 CPU: ARM926EJ-Sid(wb) [41069265] revision 5 (ARMv5TEJ) CPU0: D VIVT write-back cache CPU0: I cache: 16384 bytes, associativity 4, 32 byte lines, 128 sets CPU0: D cache: 8192 bytes, associativity 4, 32 byte lines, 64 sets Machine: DaVinci EVM Memory policy: ECC disabled, Data cache writeback Built 1 zonelists Kernel command line: video=davincifb:output=pal console=ttyS0,115200n8 noinitrd rw ip=dhcp root=/dev/nfs nfsroot=172.20.101.25:/home/amesys/workdir/filesys,nolock mem=120M This is still NFS, I guess you're trying it before flashing... [...] IP-Config: Got DHCP answer from 0.0.0.0, my address is 172.20.101.33 IP-Config: Complete: device=eth0, addr=172.20.101.33, mask=255.255.255.0, gw=172.20.101.254, host=172.20.101.33, domain=artware.fr, nis-domain=(none), bootserver=0.0.0.0, rootserver=172.20.101.25, rootpath= Looking up port of RPC 13/2 on 172.20.101.25 Looking up port of RPC 15/1 on 172.20.101.25 VFS: Mounted root (nfs filesystem). Freeing init memory: 136K Kernel panic - not syncing: No init found. Try passing init= option to kernel. If your nfs rootpath is correct, then it seems like you do not have an init anymore. You can boot your kernel with init=/bin/sh to see that your file system is functional. Regards, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkkFk+cACgkQ/nL+S5dojehPAQCgiYlJYlp4FcnJrl22UYxcjfhF 04AAoLXdINbE5BiTgZXdvVTBquj4lP7w =J2ZV -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
i2c devices present in i2c_board_info are not probed
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, I'm trying to use isl-1208 i2c rtc driver present in git kernel. However, I noticed that kernel never probes this driver. I modified my board setup code to include a rtc device dv-evm board setup code like: static struct i2c_board_info __initdata i2c_info[] = { { I2C_BOARD_INFO(tsc2004, 0x4B), .type = tsc2004, .platform_data = NULL, }, { I2C_BOARD_INFO(rtc-isl1208, 0x6F), .type = rtc-isl1208, .platform_data = NULL, }, . . . However, rtc driver is never probed. I had put a printk at the start of drivers probe function. After some digging, I found that i2c_device_probe in i2c-core.c is never called. I suppose it must be called for all my board devices present in the i2c_board_info. Does anyone have any idea what I should do, or what I'm doing wrong? Thanks, Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkkBiKwACgkQ/nL+S5dojehuHgCggfLacKdK7woDTe6r2pEDza4s 4EUAn1CNYkM5BgEZtQjWOmzRtpeXiJyI =FkH5 -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source
Re: open suse 11.0
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Vijay Soni wrote: On Suse 11.0 linux host platform, when I do /usr/sbin/exportfs -av command, it says file not found. I am in bash shell. Then I tried man exportfs, and it says that command not found. What is wrong? You should run it as root. Caglar -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.9 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org iEYEARECAAYFAkj0vykACgkQ/nL+S5dojeg4VACgh+oGXYXrbBY70wrEfii4RwEd M80An3zhi6oaAwcjbsKUIfyMhuKZwBEq =FkRD -END PGP SIGNATURE- ___ Davinci-linux-open-source mailing list Davinci-linux-open-source@linux.davincidsp.com http://linux.davincidsp.com/mailman/listinfo/davinci-linux-open-source