Part Number: AM5728 Tool/software: Linux Hello, We are using processor sdk 5.02 for linux in A15 & rtos for DSP1/2. Our use case is as per below We are following forward msg example available in ti-processor-sdk-rtos-am57xx-evm-05.02.00.10/ipc_3_50_03_04/examples/DRA7XX_linux_elf/ex41_forwardmsg/ While trying to run this example unable to see trace message from linux Here is the error root@am57xx-evm:~# cat /sys/kernel/debug/remoteproc/remoteproc2/trace0 -sh: /bin/cat: Accessing a corrupted shared library root@am57xx-evm:~# cat /sys/kernel/debug/remoteproc/remoteproc3/trace0 -sh: /bin/cat: Accessing a corrupted shared librar How to establish communication in our required way.
↧
Forum Post: Linux/AM5728: Unable to setup messageq in sdk 5.02
↧
Forum Post: Linux/AM3358: VDDSHV3 power supply
Part Number: AM3358 Tool/software: Linux Hi TI: I have a self-developed AM3358 board with AR8031 to realize 1000M Enet in RGMII2 When I supply 3.3V to VDDSHV3 and ”ping ip -p 0xf0 -s 65000" , the probability of time out is about 30% When I supply 2.5V to VDDSHV3 and ”ping ip -p 0xf0 -s 65000" , the probability of time out is less than 0.01% When I supply 1.8V to VDDSHV3 and ”ping ip -p 0xf0 -s 65000" , the probability of time out is about 1% According to AM3358 datasheet, VDDSHV3 can run in 3.3/1.8V power supply, can I supply VDDSHV3 with 2.5V Voltage?
↧
↧
Forum Post: Linux/TDA2PXEVM: If it should not translation the address between virtual address & physical address when we enable 1G DDR
Part Number: TDA2PXEVM Tool/software: Linux According to the chapter 6.6 of VisionSDK_UserGuide_MemoryMap.pdf, if we use the 512M DDR, it should modify the vpsInitPrms.virtBaseAddr = 0x80000000U to vpsInitPrms.virtBaseAddr = 0xA0000000U. So, if we still use 1G DDR, it does not modify the vpsInitPrms.virtBaseAddr ? That is to say, if we use 1G DDR, do we need to enable the translation between virtual address and physical address? Below is the part of patch from the chapter 6.6 of VisionSDK_UserGuide_MemoryMap.pdf, Thanks Terence ============================================================ diff --git a/links_fw/src/rtos/links_ipu/system/system_bsp_init.c b/links_fw/src/rtos/links_ipu/system/system_bsp_init.c index 97287ff..5a18372 100755 --- a/links_fw/src/rtos/links_ipu/system/system_bsp_init.c +++ b/links_fw/src/rtos/links_ipu/system/system_bsp_init.c @@ -251,8 +251,10 @@ Int32 System_bspInit(void) #ifdef A15_TARGET_OS_LINUX /* This one to one mapping is required for the 1GB Linux builds */ - vpsInitPrms.virtBaseAddr = 0x80000000U; + vpsInitPrms.virtBaseAddr = 0xA0000000U; vpsInitPrms.physBaseAddr = 0x80000000U; + /* if Virtual address != Physical address then enable translation */ + vpsInitPrms.isAddrTransReq = TRUE; #else vpsInitPrms.virtBaseAddr = 0xA0000000U; vpsInitPrms.physBaseAddr = 0x80000000U;
↧
Forum Post: AM5718: DDR3L with IS43TR16256BL-125KBLI for AM5718
Part Number: AM5718 As we find MT41K256M16HA-125 used in EVM of AM5718 obsolete ,we have replaced in with IS43TR16256BL-125KBLI in our board design. What changes we have incorporate in software (linux) to make this DDR3L compatible with AM5718 ?? please suggest recommended approach and changes needed.
↧
Forum Post: Linux/AM3352: USB overcurrent
Part Number: AM3352 Tool/software: Linux Hi Sir We used TI AM335x 05.03.00.07 SDK for development and has usb current limit request. When USB over current phenomenon happened , the log showed as below musb-hdrc musb-hdrc.1: VBUS_ERROR in a_wait_vrise (80, <SessEnd), retry #3, port1 0008010c Do we have any method to recover USB port to normal work after unplugging device except rebooting. Please advise BR Yimin
↧
↧
Forum Post: AM5746: EMIF ECC processing
Part Number: AM5746 Hi. Would you tell me how to use ECC (EMIF) ? ・How to get the number of 1-bit errors（How to use EMIF_1B_ECC_ERR_CNT） ・How to add interrupt processing at 2-bit error Regards, Rei
↧
Forum Post: Linux/DRA71XEVM: can this case work?
Part Number: DRA71XEVM Tool/software: Linux hi all, on the linux side, draw background picture to GFX pipeline by DRM (not use omapdrm or not use virtual drm), and draw other up layer pictures to pipeline VID1, VID2 and VID3 by visionSDK on M4 size. Does this case can work? best regards!
↧
Forum Post: Compiler/TMS320C6657: C6657 EMIF16 dual core boot
Part Number: TMS320C6657 Tool/software: TI C/C++ Compiler Hello, I have a single core custom design which boots from flash through EMIF16. It is based on a custom boot loader (modification of boot.c from the rts, relies on copy tables and _auto_init_elf call). We use hex6x in image mode to generate the binary. Here is the principle of our custom boot loader : extern far COPY_TABLE memcpy_ctbl; extern far COPY_TABLE boot_ctbl; extern far COPY_TABLE rts_ctbl; /*------------------------------------------------------------------------*/ /* MEMCPY TABLE HAVE TO BE COPIED FIRST */ /*------------------------------------------------------------------------*/ my_copy_in(&memcpy_ctbl); //memcpy64 /*------------------------------------------------------------------------*/ /* CALL THE LOW LEVEL INITIALIZATION */ /*------------------------------------------------------------------------*/ copy_in(&rts_ctbl); //rts6600 copy_in(&boot_ctbl); //LowLevelInit LowLevelInit(); /*------------------------------------------------------------------------*/ /* CALL THE AUTOINITIALIZATION ROUTINE. */ /*------------------------------------------------------------------------*/ _auto_init_elf(); _args_main(); This how we generate our binary : hex6x.exe -a --image --order=M --issue_remarks --map=dsp_program.map -o dsp_program.btbl configSingleCore.cmd ..\dsp_program.out b2ccs.exe dsp_program.btbl dsp_program.dat ccs2bin.exe dsp_program.dat dsp_program.bin with configSingleCore.cmd : ROMS { all_mem: o = 0x70000000, l = 0x400000,romwidth=32, memwidth=32 } Despite I've read a lot of docs and posts, I'm stuck at generating a dual core binary by merging 2 .out. It seems that bfmerge is used for another kind of boot loader which relies on boot tables instead of copy tables. I was expecting a tool which is able to merge .binit and .ovly so the custom boot loader on core 0 can load core 0 and core 1 : Am I wrong ? Which method should I use to load core 1 in my case ? Regards, David
↧
Forum Post: CCS/AM5718: Pinmux tool questions
Part Number: AM5718 Tool/software: Code Composer Studio Hi Champs, I would like to confirm following pin mux tool function. 1: Rx check box detail explain 2: Slow / slew check box detail explain 3: Is this PIN MUX tool able to use "un-used " pin setting ? Regards, Kaz
↧
↧
Forum Post: Linux/AM5728: Enabling Serial Port
Part Number: AM5728 Tool/software: Linux Hello we are using AM571x IDK board to test some communication apps in RS485 port. however t he serial port seems disabled. In one of the forum posts theres one answer to modify the am57xx-beagle-x15-revc.dts file. http://e2e.ti.com/support/processors/f/791/t/762768?Linux-AM5728-Enable-UART But no definition of serial or uart in this file. There are some other related files online, but no guide on how to modify them. Could you advice? Thanks.
↧
Forum Post: RTOS/TMS320C6670: Can I generate my own XDC modules?
Part Number: TMS320C6670 Tool/software: TI-RTOS If I create new typical SYS/BIOS project in CCS, I noticed that there are basic xdc modules such as BIOS, Task, Clock, Memory, Semaphore, etc. My question is as follows. Can I generate my own customized XDC modules? For example, if I want to generate new module which provides task and memory status simultaneously, how can I do that? My ultimate goal is that first, I generate my own customized xdc module, second, I added this module in cfg file, and third, use APIs in the module when I create SYS/BIOS based project. I am looking forward to your answer. Thank you in advance. Best regards, Heungseop
↧
Forum Post: RTOS/TMS320C6670: How can I call RTOS APIs at the host PC?
Part Number: TMS320C6670 Tool/software: TI-RTOS For example, if I want to call task_create API which is executed on the DSP core at the host PC (debugging PC), how can I do that? More specifically, I want to call task_create API through C programming at the host PC. I don't know what should I read and I have no clue for this. It would be very helpful if you let me know which document should I read. I am very looking forward to your answer. Thank you in advance. Best regards, Heungseop
↧
Forum Post: Linux/TDA2SX: How to debug spl code after relocate
Part Number: TDA2SX Tool/software: Linux Dear ti, With the help of Yordan Kamenov , I can now debug C code in the spl image before relocate. i.e. the board_init_f part. After that spl relocate itself to 0x82000000 by the function call spl_relocate_stack_gd as: Then i cannot step into board_init_r part after this operation. I have tried to load symbol with the 'code offset' & 'data offset' filled with 0x82000000 and then set hardware breakpoint at the entry of function board_init_r. But it would not stop at the breakpoint. It seem that the relocate address is wrong. I got no souce files available in the ccs. How should i solve this? Regards, Liu Gan
↧
↧
Forum Post: 66AK2G12: TIDA-01414
Part Number: 66AK2G12 Dear Sir/Ms. Customers refer to our TIDA-01414 Reference design. However, the customer's product specifications are as follows 1. Only 3 speakers, 100W, 93dB 2. Input HDMI In (eARC function) *1, SPDIF *1 (OPT + COAX) 3. Output HDMI Out *1 to TV 3. Need to support ATMOS Because the datsheet specification on the 66aK2G webpage is not complete, I would like to ask the following question, TIDA-01414 Reference design question 1. Does the 66aK2G support HDMI IN? and has an eACR function? 2. Does 66aK2G support SPDIF? 3. Dose 66aK2G support HDMI output to TV? If you do not have the eACR support function, can 66aK2G connect the PCM9211 , and support function what requested by the customer? If the 66AK2G does not have the functionality required by the customer, can you suggest external parts? Best Regards, Kami Huang
↧
Forum Post: RTOS/AM3352: socket api sendto return -1 after a few minutes
Part Number: AM3352 Tool/software: TI-RTOS Hi I have creat a project to send data in UDP,this is my test file. It can send 10 datas by udp,but it will failed after about a few minutes. the log shows error on the API sendto,it returns -1,and I have check the error code,it is 64. It will stop after sending 23,740 or so. please help me to resolve this issue. #define MAXBUF 10 #define NDK_UDPPORT 5555 /* talk to NDK echo server running on Concerto board */ #define SERVER_UDPPORT 5555 #define SERVER_IP "192.168.0.112" #define SLEEPTIME 1000 #define Sleep Task_sleep /* port the UDP echo Linux side tool "udpSendRecv" to NDK on Freon */ void udpEchoClient(void) { int i; SOCKET sockfd; int sleepTime = SLEEPTIME; struct sockaddr_in bindAddr; struct sockaddr_in servAddr; int bytes_read, bytes_sent; char *buffer; struct sockaddr_in fromAddr; /* originally was type 'sockaddr_storage' */ int fromAddrLen = sizeof(fromAddr); int ret = 0; int count = 0; int id; time_t start; unsigned int buffSize = MAXBUF; ret = fdOpenSession(TaskSelf()); UART_printf( "fdOpenSession result = %d\n",ret); Task_sleep(10000); buffer = malloc(buffSize); if (!buffer) { UART_printf("failed to malloc buffer!\n"); goto QUIT; } memset(buffer, 0, buffSize); /* initialize destination address */ memset(&servAddr, 0, sizeof(servAddr)); servAddr.sin_family = AF_INET; inet_aton(SERVER_IP, &servAddr.sin_addr); servAddr.sin_port = NDK_htons(SERVER_UDPPORT); /* create the socket */ /* TODO: assuming socket created successufully, due to cast error */ sockfd = socket(AF_INET, SOCK_DGRAM, IPPROTO_UDP); if(NULL == sockfd) { UART_printf( "socket failed\n"); } /* bind our socket to a particular port. Must bind else server reply drops! */ memset(&bindAddr, 0, sizeof(bindAddr)); bindAddr.sin_family = AF_INET; bindAddr.sin_addr.s_addr = INADDR_ANY; bindAddr.sin_port = NDK_htons(NDK_UDPPORT); if (bind(sockfd, (struct sockaddr *)&bindAddr, sizeof(bindAddr)) < 0) { UART_printf("failed to bind() the socket!\n"); goto QUIT; } /* print out sockets info */ UART_printf("Starting test with a %d uSec delay between transmits\n", sleepTime); /* loop */ i = 0; while (1) { buffer[0] = (char)(++i); buffer[buffSize - 1] = (char)~i; /* send the data */ bytes_sent = sendto(sockfd, buffer, buffSize, 0, (struct sockaddr *)&servAddr, sizeof(servAddr)); if (bytes_sent < 0 || bytes_sent != buffSize) { UART_printf("[id %d] stopping test. sendto returned %d (error %d)\n", id, bytes_sent, fdError()); goto QUIT; } Task_sleep(1000); } QUIT: UART_printf("--- at QUIT\n"); /* clean up */ if (sockfd) { fdClose(sockfd); } free(buffer); fdCloseSession(TaskSelf()); }
↧
Forum Post: AM6526: AM6526 4 port EtherCAT
Part Number: AM6526 Hi, is it possible to use the AM6526 as an EtherCAT junction device with 3 or 4 EtherCAT ports? This would mean that one port is connected to the EtherCAT master and the other ports are connected to several EtherCAT slaves. See also this picture of the Beckhoff ET1100 ASIC as reference: Best regards, Johannes
↧
Forum Post: RTOS/TDA2PXEVM: SurroundView 3D usecase - demo car image change
Part Number: TDA2PXEVM Tool/software: TI-RTOS Hello, We have project with SRV 3D use case. A request from the customer to change Embedded demo car image is received: "...If we change car image file(obj file) on Demo UI how can we change? Please, provide us some guide...." Platform used is TDA 2Px EVM + VSDK 3.05 BR Simo
↧
↧
Forum Post: Linux/AM5728: how to use the ltp tool?
Part Number: AM5728 Tool/software: Linux Hi： Software version: ti-processor-sdk-linux-rt-am57xx-evm-03.03.00.04, hardware version: AM5728-ES2.0, custom board, 1, how to use the ltp tool? 2. Is there a tutorial on using the ltp tool.
↧
Forum Post: RTOS/TDA2PXEVM: SurroundView 3D usecase - real demo car size changed
Part Number: TDA2PXEVM Tool/software: TI-RTOS Hello, We have project with SRV 3D use case. A request from the customer about real car size change is received: ".....What if size of the real demo car is changed? For example, the attached ‘TI_Demo_Car_Size.JPG’ file shows different Demo car size. Actually, we maybe should change the Demo Car for other reason. So, We want to know there should be changed some settings(ex. Calibration) or S/W part if the demo car size is changed....." BR Simo
↧
Forum Post: RTOS/AM5746: About the error of the exclusive procedure in Event_post () / Mailbox_post.
Part Number: AM5746 Tool/software: TI-RTOS Hi, TI expert. I checked the code difference of Semaphore.c (between bios_6_52_00_12 and bios_6_76_00_08). C:\ti\bios_6_52_00_12\packages\ti\ sysbios \knl\Semaphore.c C:\ti\bios_6_76_00_08\packages\ti\ sysbios \knl\Semaphore.c In bios_6_76_00_08, the task dispatch disabled section is extended. (Compare with bios_6_52_00_12) In Semaphore_post () Task_disable ()-> Hwi_disable ()-> ... ...-> Hwi_restore ()-> Task_restore () ... (Exclusive end) TI implements the above exclusive procedure. I believe that the exclusive procedure of Semaphore_post () is correct. Because these are the steps of disabling dispatching, disabling interrupts, releasing interrupts disabled, and releasing dispatches disabled. Please answer the following 4 questions based on the above. Q1. About exclusive procedure error in Event_post() Event.c: In Event_post(), the exclusive procedure is as follows. Hwi_disable ()-> Task_disable ()-> ... ...-> Hwi_restore ()-> Task_restore () ... (Exclusive end) Is the exclusive procedure in Event_post () correct compared to Semaphore_post()? I think that the Event_post () exclusive procedure should be implemented as Semaphore_post(). Please refer to attached file 1-1.About_Exclusive_section_Event.pdf. Q2. About exclusive procedure error in Mailbox_post() Mailbox.c In Mailbox_post(), the exclusive procedure is as follows. Hwi_disable ()-> ... ...-> Hwi_restore () ... (Exclusive end) Is the exclusive procedure in Event_post () correct compared to Semaphore_post ()? I think that the Event_post () exclusive procedure should be implemented as Semaphore_post (). Please refer to attached file 1-2.About_Exclusive_section_Mailbox.pdf. Q3. About the same exclusive procedure error in other APIs I think the exclusive procedures in Mailbox / Event / Semaphore should be the same in all cases below. Correct? Mailbox 0.TaskA Pri: Low, Task B Pri: High, 1.TaskB wait Mailbox(Mailbox_pend)... 2.TaskA Mailbox_Post ... Event 0.TaskA Pri: Low, Task B Pri: High, 1.TaskB wait Event (Event_pend) ... 2.TaskA Event set (Event_post) ... Semaphore 0.TaskA Pri: Low, Task B Pri: High, 1.TaskB wait Semaphore (Semaphore_pend) ... 2.TaskA Signal Semephore (Semaphore_post) ... In TI-RTOS, there is a difference in the exclusive procedure in the above function. Even if there is a difference in the type of Kernel Object (Event / Semaphore / Mailbox), I think that there should be no difference in behavior. Why is the difference in the exclusive procedure? What is your intention? （HW:AM574x / SW:bios_6_76_00_08） Best Regard. kuniBash Attachment Files (Please visit the site to view this file) (Please visit the site to view this file)
↧