Hi Anand， our system on M4 is highly customized based on sysbios. There is no GMACSW driver and AVB stuff in original system. our supplier ported GMAC driver from Vision SDK to IPU and implemented AVB talker based on TI's talker sample. Best Regards
↧
Forum Post: RE: RTOS/DRA746: GMAC AVB configuration
↧
Forum Post: RE: Linux/TMDSEVM572X: Arago "base-rootfs" initializes only 2 ethernet interfaces
I have separated your thread here . You are asking about multiple things in one thread. Please, close this thread and create a new thread for any new question which is not related to this thread's topic.
↧
↧
Forum Post: RE: AM5728: EtherCAT Slave timing problems
[quote user="Thomas N."] I've spent some time in the DC configuration, but I still have some problems. My Slave doesn't change to OP (pending in SAFEOP) because of synchronization errors. I investigated a little bit time in the stack details and I could find out, that CheckIfEcatError() throws an AL Status ' ALSTATUSCODE_SYNCERROR '. ecatslv.c, line 2330ff: ? 1 2 3 4 5 6 7 8 9 10 11 12 void CheckIfEcatError( void ) { ... else if (!bSmSyncSequenceValid) { AL_ControlInd(STATE_SAFEOP, ALSTATUSCODE_SYNCERROR); return ; } } } } The boolean ' bSmSyncSequenceValid ' will be set to false in DC_CheckWatchdog() , if u16SmEventMissedCounter is greater then u16SyncErrorCounterLimit (the last one is set to 4. If I increse this value, the error will be delayed - but is still happend). And the boolean 'bSmSyncSequenceValid' will be incremented and decremented in PDI_Isr() and Sync0_Isr() . But I don't know why PDI decrements with '1' and the Sync0_ISR increments with '3'. Furthermore, I received sporadically the error ' Sync0 watchdog expired' from DC_CheckWatchdog() . What have I done? I changed only the mode from SM-Synchron to DC-Synchron. EcEngineer's master will set automatically the ESC register 0x981 and change both COE objects 0x1C32 and 0x1C33 to 2. Further, I changed to the local timer for the watchdog: ? 1 2 3 4 5 #define AL_EVENT_ENABLED 1 #define DC_SUPPORTED 1 #define ECAT_TIMER_INT 0 #define ESC_SM_WD_SUPPORTED 0 #define MIN_PD_CYCLE_TIME 0xFA00 My Slave received already the Sync0 event. The PDI task has a priority of 6 and the Sync0 task has a priority of 8. Do you have an idea or any hints to solve this issue? Beckhoff doesn't provide any solutions for this problem. The default example from TI (tiecatappl) has similiar problems: If the slave is in DC mode its state cannot be set to OP. [/quote] What about this issue?
↧
Forum Post: RE: TDA2EVM5777: TIDL: How to get memory usage on TIDL running?
Hi Lee, you can press 'p' when running the usecase to get various statistics (including memory usage) printed in the log. You can also find this thread useful: Regards, Yordan
↧
Forum Post: RE: AM4379: Qt issues
Yes. The touch screen doesn't work in the Processor SKD 6. Please, switch to the previous Processor SDK 5.3.07 where it works, and do not forget to setup the Processor SDK 5.3.07 settings and paths in the Qt creator.
↧
↧
Forum Post: RE: AM5728: Kernel panic
Rio, Yes, I think you can run the mem_test diagnostic test In this kind of conditions. You might also use u-boot memory tester and/or linux memtester. For u-boot memory test: u-boot/doc/README.memory-test u-boot/common/memsize.c For linux memtester: Check also below docs: There might be also some voltage/power HW issue. Please go through below docs also: Regards, Pavel
↧
Forum Post: DRA77P: Trouble building recipe for bluez5.50 in yocto
Part Number: DRA77P Hi, We have been trying to build a recipe for bluez in our yocto and comes up with following errors, WARNING: bluez5-5.50-r0 do_fetch: Checksum mismatch for local file /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz Cleaning and trying again. WARNING: bluez5-5.50-r0 do_fetch: Renaming /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz to /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz_bad-checksum_8e35c67c81a55d3ad4c9f22280dae178 WARNING: bluez5-5.50-r0 do_fetch: Checksum failure encountered with download of cdn.kernel.org/.../bluez-5.50.tar.xz - will attempt other sources if available WARNING: bluez5-5.50-r0 do_fetch: Checksum mismatch for local file /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz Cleaning and trying again. WARNING: bluez5-5.50-r0 do_fetch: Renaming /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz to /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz_bad-checksum_8e35c67c81a55d3ad4c9f22280dae178 ERROR: bluez5-5.50-r0 do_fetch: Checksum failure fetching cdn.kernel.org/.../bluez-5.50.tar.xz ERROR: bluez5-5.50-r0 do_fetch: Fetcher failure for URL: ' cdn.kernel.org/.../bluez-5.50.tar.xz' . Checksum mismatch! File: '/home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz' has md5 checksum 8e35c67c81a55d3ad4c9f22280dae178 when 913f35d6fa4ca5772c53adb936bf1947 was expected File: '/home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/downloads/bluez-5.50.tar.xz' has sha256 checksum 5ffcaae18bbb6155f1591be8c24898dc12f062075a40b538b745bfd477481911 when ddab3d3837c1afb8ae228a94ba17709a4650bd4db24211b6771ab735c8908e28 was expected If this change is expected (e.g. you have upgraded to a new version without updating the checksums) then you can use these lines within the recipe: SRC_URI[md5sum] = "8e35c67c81a55d3ad4c9f22280dae178" SRC_URI[sha256sum] = "5ffcaae18bbb6155f1591be8c24898dc12f062075a40b538b745bfd477481911" Otherwise you should retry the download and/or check with upstream to determine if the file has become corrupted or otherwise unexpectedly modified. ERROR: bluez5-5.50-r0 do_fetch: Function failed: base_do_fetch ERROR: Logfile of failure stored in: /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/build/arago-tmp-external-linaro-toolchain/work/armv7ahf-neon-linux-gnueabi/bluez5/5.50-r0/temp/log.do_fetch.14473 ERROR: Task (/home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/sources/oe-core/meta/recipes-connectivity/bluez5/bluez5_5.50.bb:do_fetch) failed with exit code '1' We tried changing the checkum to SRC_URI[md5sum] = "8e35c67c81a55d3ad4c9f22280dae178" SRC_URI[sha256sum] = "5ffcaae18bbb6155f1591be8c24898dc12f062075a40b538b745bfd477481911" And the following errors came up, ERROR: bluez5-5.50-r0 do_patch: Command Error: 'quilt --quiltrc /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/build/arago-tmp-external-linaro-toolchain/work/armv7ahf-neon-linux-gnueabi/bluez5/5.50-r0/recipe-sysroot-native/etc/quiltrc push' exited with 0 Output: Applying patch 0001-hciattach-bcm43xx-fix-the-delay-timer-for-firmware-d.patch patching file tools/hciattach_bcm43xx.c Hunk #1 FAILED at 228. 1 out of 1 hunk FAILED -- rejects in file tools/hciattach_bcm43xx.c Patch 0001-hciattach-bcm43xx-fix-the-delay-timer-for-firmware-d.patch can be reverse-applied ERROR: bluez5-5.50-r0 do_patch: Function failed: patch_do_patch ERROR: Logfile of failure stored in: /home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/build/arago-tmp-external-linaro-toolchain/work/armv7ahf-neon-linux-gnueabi/bluez5/5.50-r0/temp/log.do_patch.13626 ERROR: Task (/home/teuser/ti-processor-sdk-linux-automotive-dra7xx-evm-5_00_00_01/yocto-layers/sources/oe-core/meta/recipes-connectivity/bluez5/bluez5_5.50.bb:do_patch) failed with exit code '1' What do we do to resolve this ?
↧
Forum Post: RE: TDA3XEVM: Use GRPX Link draw bmp has a problem when switching
I have found the issue, because I think the color 0X0000 is black, but after testing it is found to be through the first layer of the camera display Q : If I need to draw a black color in GRPX, how should the color be set?
↧
Forum Post: RE: DRA746: Need to know whether EMIF1 is a mandate or optional
Hi Franklin, current version of the TRM is: DRA75x, DRA74x SoC for Automotive Infotainment Silicon Revision 2.0, 1.1 Literature Number:SPRUI30F November 2015 – Revised October 2018 you can find it here: I will ping DDR experts for the other questions. Regards, Yordan
↧
↧
Forum Post: RE: 66AK2H14: Linux freezes when it should reboot
Hi, I have an update to this case. This patch made the reboot work: diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c index 5413ffa..24ecf08 100644 --- a/drivers/clk/clk.c +++ b/drivers/clk/clk.c @@ -1025,7 +1025,7 @@ static void clk_disable_unused_subtree(struct clk_core *core) * sequence. call .disable_unused if available, otherwise fall * back to .disable */ - if (clk_core_is_enabled(core)) { + if (clk_core_is_enabled(core) && (0 != strcmp(core->name , "ddr3-1"))) { trace_clk_disable(core); if (core->ops->disable_unused) core->ops->disable_unused(core->hw); I am now seeing if it's possible to make an equivalent change via device tree instead of a kernel source patch.
↧
Forum Post: AM3352: Change ttyS0 to ttyO0
Part Number: AM3352 customer upgrade system from SDK6.0 to Processor SDK6.0, in the old kernel 3.2, the UART default name is ttyO0, now it is ttyS0, How to map ttyS0 to ttyO0, then no need to modify application code. Or how to change the name of ttyS0 to ttyO0?
↧
Forum Post: AM5746: DWC3 information
Part Number: AM5746 Hi, USB IP of AM57 is DWC3 (Synopsys). software-dl.ti.com/.../Foundational_Components_Kernel_Drivers.html Are there any devices other than Sitara/OMAP that have Synopsysno USB IP? There is little information on the DWC3 register. Therfore I would like to refer to other device documents too. Regards, Rei
↧
Forum Post: RE: TDA2EVM5777: Linux: Does the h.264 input of nullSource support only i-frame?
Hi, Its been long since any update on this thread. I hope the issue is resolved. If not the please reply below or create a new thread for a different query. Regards, Anuj
↧
↧
Forum Post: AM5728: Camera screen is missing when running tidl sample model of object detection on TDA2
Part Number: AM5728 Hi sirs, I imported a tidl trained model ( voc0712_ssdJacintoNetV2_iter_120000.caffemodel in caffe-jacinto-models-caffe-0.16) to TDA2. The camera screen doesn't display correctly when I was running the model. Do you have any ideas to solve this problem? My VSDK version is 3.7 . Thank you, Gary
↧
Forum Post: RE: TMDSICE3359: EtherNet/IP questions
Stefan, Just wanted to let you know what we are looking into the issue as high priority, and will update you if any progress. Regards, Garrett
↧
Forum Post: RE: Compiler/PROCESSOR-SDK-AM335X: PRU communication problem
Hi Nick, thanks for the pointer. Just to be clear, mmapping to SISR does not require UIO_PRUSS or any specific driver right? I'm asking becuase firmware loading by symlinking /lib/firmware/am335x-fw-pruX doesn't seem to work without the remoteproc module, and that seemed to be allow large executive files larger than the 4K limit.
↧
Forum Post: RE: Compiler/TMS320C6654: What kind of problems occur when reading and writing large amounts of data to and from external DDR?
MN, Are you continuing to see this issue or did Todd`s response help you root cause your setup. Is it deliberate that you have kept all the tasks at same priority or are you just using the default that puts all the tasks at same priority. From your description, it appears that you want the second task to start execution periodically when woken up by the timer so it should be at a higher priority so that it starts executing as soon as it gets the ready event. similarly third task should be higher priority that first task since you want the task to execute as soon as first task generates the post event. Are you using semaphores to post the event. You can use semaphore post and pend mechanism to let each task to start execution. Instead of using Task yeild at the end, you could use Semphore pend at the end of task two and three so they go to sleep and let the first task to execute and post semaphore when the event occurs. Regards, Rahul
↧
↧
Forum Post: RE: SDK_VISION guidelines_RVC
Hi Ahmed, It looks like the Alg_drawRearview function uses draw2D. Specifically, can you look at Draw2D_drawLine and the Draw2D_LinePrm parameter passed to it for how to set color format, line thickness, and color? It looks like the example uses SYSTEM_DF_YUV420SP_UV as its color format, and in apps/src/rtos/alg_plugins/drawRearview/drawRearviewLink_priv.h there is a set of defined colors already for that. Can you try using one of these combinations? Thanks, David
↧
Forum Post: RE: Compiler/BEAGLEBOARD-X15: OpenCV compilation error
Hi Mustafa, You are absolutely right. The hack I mentioned is only for you to run a quick test. OpenCV shipped with Processor SDK is built with Yocto process. You can follow the following links for setting up PSDK build and OpenCV build: http://software-dl.ti.com/processor-sdk-linux/esd/docs/latest/linux/Overview_Building_the_SDK.html To fix the problem that you reported in this thread, you can apply the following patch to opencv: From 17b9f9f72181f0d673d8699ea931e94fd9a6e74c Mon Sep 17 00:00:00 2001 From: Yuan Zhao Date: Tue, 17 Sep 2019 15:22:50 -0500 Subject: [PATCH] morph.cl: fix compilation error - C6x intrinsic _min2(), _max2() cannot be applied to data types such as uchar3. Revert to min() and max() as the master branch. Signed-off-by: Yuan Zhao --- modules/imgproc/src/opencl/morph.cl | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/modules/imgproc/src/opencl/morph.cl b/modules/imgproc/src/opencl/morph.cl index 8e3ab2fa0..47243ca0c 100644 --- a/modules/imgproc/src/opencl/morph.cl +++ b/modules/imgproc/src/opencl/morph.cl @@ -90,11 +90,11 @@ #if defined INTEL_DEVICE && defined DEPTH_0 #define MORPH_OP(A, B) ((A) < (B) ? (A) : (B)) #else -#define MORPH_OP(A, B) _min2((A), (B)) +#define MORPH_OP(A, B) min((A), (B)) #endif #endif #ifdef OP_DILATE -#define MORPH_OP(A, B) _max2((A), (B)) +#define MORPH_OP(A, B) max((A), (B)) #endif #define PROCESS(y, x) \ -- 2.17.1 Hope this helps. - Yuan
↧
Forum Post: RE: TMS320C6678: Failed to build Image_processing openmp demo
Shin, There are some discussion on GNU make error code 2 with CreateProcess() on the forum and others. Can you try to use a absolute path for 'webpages/tool/php'? Regards, Garrett
↧