Search Results

Search results 1-20 of 1,000. There are more results available, please enhance your search parameters.

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

  • No worries. Mistakes happen. Glad to hear that you were able to resolve this.

  • The IAR J-Links have been discontinued years(!) before the nRF91 was released. You also do not expect your Windows 2000 to run the latest software from 2015-2020, right? It is similar for this J-Link and MCU support… We recommend to get a new J-Link: segger.com/purchase/trade-in-program/ BR Alex

  • What exactly is the question? When starting a debug session, J-Link takes care of the potentially required flash programming automatically. If you are looking for a stand-alone flash programming solution, I doubt that VSCode is the way to go.

  • You are mixing things here…. First, you asked for debug support, now, in your latest post, you are referring to production programming. These are 2 totally different things. Production programming: The easier part. You do not have to mess around with various specifics of the device. No caches, less silicon bugs, … that can come into play. As you might have noticed, we already support a lot of ATMega, PIC16, PIC18, … devices with our Flasher PRO production programmer. Debug support: Much more wor…

  • Yes it is but you would need to create a flash loader on your own: wiki.segger.com/Open_Flashloader Sorry to say that we do not have any samples for eMMC.

  • Hello, The V10 is in the maintenance period. So bugfixes will be added but no new features. Quote from Noy: “ Are there plans to support other PIC (18, dspic) beside PIC32? V10/V11 should support ICSP but why not PIC18? Would be great if i don't need my PICkit 3 anymore.. Only a JLink for all... ” No plans for PIC16 / 18 /33 / ... debug support right now. ICSP for PIC32 is different from ICSP from PIC18 / 33 / ... (Microchip just calls it ICSP on all platforms but they are actually different pro…

  • Hi, I see. This makes sense for sure. We will add this option. BR Alex

  • Hi Phil, USB: J-Link might be connected via USB but GDB and GDBServer communicate via TCP/IP with each other. Again: It is a problem in GDB. GDB is used by KDS to load the ELF file (your target application) and to interpret the debug information. During that interpretation, something goes wrong on the GDB side that causes it to half-way crash and being left in an unusable state. So if you are looking for a solution: It is still no SEGGER problem. You still need to get in touch with NXP regarding…

  • OK, found it. It actually is not really a buffer overflow but glibc seems to detect it as one. Fact 1) If you remove libncurses5 from your system as part of autoremove, you make your GDB toolchain half-way unusable because arm-none-eabi-gdb depends on it. As this is a portable and non-installed package, the package manager does not detect that libncurses5 is still "needed". Fact 2) By forcing fact 1, you trigger the following behavior of VSCode: - Start GDBServer - Try to start GDB which should …

  • Hi, After executing the "purge" with "auto-remove" I can reproduce the issue. However, I can only reproduce it when starting a session from within VSCode. If I start JLinkGDBServer stand-alone but with exactly the same command line as VSCode uses, I *never* get a fault: Source Code (1 line) Another weird thing is that this seems to be the toolchain base address used by VSCode: Source Code (1 line) But if I try to start GDB from thar directory, I get a fault because of missing dependencies (The d…

  • Hello, I still not understand what exactly you expect from us (SEGGER). Following your description, you are using the Kinetis Design Studio (KDS) in the following combination: KDS -> GDB <-TCP/IP-> GDBServer -> J-Link KDS: Component from NXP GDB: Component from NXP or 3rd party GDBServer: Component from SEGGER J-Link: Component from SEGGER The J-Link log file is not indicating any error. GDB just closes the connection to GDBServer at some point. In your original post, you posted the following er…

  • Hi, If you switch to a vanilla Eclipse or some Eclipse port that still supports the original features, you should be able to get away with the -rtos xxx option. BR Alex

  • TI CCS is not using JLink GDBServer but loads the J-Link lib and uses the J-Link API directly. I have no idea if CCS could still be used with GDBServer. If it can be, it is nothing that has been tested. I know that CCS is Eclipse but it is far away from a vanilla Eclipse, like Renesas e2studio is as well. So not sure how much of the original Eclipse functionality is still there. BR Alex

  • Yes, this can happen when using asynchronous mechanisms. It is a bit like resetting the MCU while there is a DMA transfer in progress or before a UART was able to send out all data via Tx interrupt + buffer. Suggestion: Wait for in your software until J-Link has emptied the ring buffer (ideally with timeout) before issuing a self-reset from within your software. BR Alex

  • I also checked with another VM (Ubuntu 20.04 + VMWare) I have here but starting GDBServer (same command line as you use) and waiting for the target connection etc. works fine in this VM as well. I also checked with V6.98c V7.20a BR Alex

  • Hi Godsmack et al, I just got back on this and tried the VM you provided etc. First of all: Thanks for the VM. One of the best prepared support cases ever. Download, start, description how to reproduce etc. all just perfect! Quote from Godsmack: “Anyway - all described in head post problems were recreated and easy to reproduce - right now installed driver version 6.92 and all works like charm but in Downloads folder you can find latest (per now) 6.98c and if you will replace current driver with …

  • CPU TAP NOT FOUND

    SEGGER - Alex - - J-Link/Flasher related

    Post

    Hello, Unfortunately, for RISC-V (unlike for ARM) there are no standardized JTAG IDCODEs for the JTAG TAPs that would allow identification as a RISC-V DM TAP. Just because the IRLen is 5 for RISC-V TAPs, it does not mean that any TAP with IRLen 5 is a RISC-V TAP (there are some counter examples I could tell). Do you have 4 RISC-V cores on your target? Did you assign the TAP Ids? I am just confused because there are 4 TAPs on your device and all of them have different TAP Ids. (I would have expec…

  • Hello, I do not think that OS aware debugging is available for TI CCS. At least not for the J-Link plugin. The CCS support was initially done by TI and we took over at some point, so I would need to look things up. In case it is possible to have some kind of OS plugin etc., we will probably create one but this is nothing that will happen within the next months. So more something for 2022. BR Alex

  • Hello, It depends on the actual model. On most it is done in software, on some there is a dedicated hardware UART available. Therefore, the achievable baudrates (and the max. baudrate) depends on the actual model and HW version. If you post your S/N, I can tell you the VCOM parameters. BR Alex

  • a) Real time means that outputting a string usually does not affect the real time of your application, as it is mainly a memcpy() into a buffer, which is blazing fast. It does NOT mean that the data is immediately read out by J-Link. b) We strongly suggest to NOT use sleep modes as you can encounter all kinds of weird effects when using them. From double data, over garbage data up to crashes, everything is possible. c) 4 MHz is a quite slow TIF speed for RTT. You should be more at 15 MHz - 50 MH…