i have an enclustra SA2 card that has the cyclone V soc. i designed a hardware that mimics the enclustra PE1 card that is used to interface with the SA2 card for roject purposes.i have deployed by image on the sd card which has been partitioned and tried to boot up on enclustra platform and the processor boots up. On the SA2+my hardware, the processor does not boot up from SD card. I have debugged using the ARM DS-5 tool and checked the status of bootromswstate register and its x100. i am pretty new to using processors and would appreciate any help on how to debug this issue. What are the things i can look out for when debugging this processor?
hanging of the processor at booting could have many many reasons. I am sorry in advance for any missunderstanding, because I am used to use kits only, not cards - this is the most common source of problems.
- First of all I recommend You to check bootsel switches, if any are there.
Could You answer these points?:
- Have You really compiled the SD card image for Your target card?
- Do You use Linux-based OS or bare-matal application?
- How did You compile it?: with Yocto, with Buildroot, by Your own hands or just download it?
- Had You done any adjustments at default configuration of compilation?
Yours Jan Konečný.
thanks for getting back.
I checked the values of the bootinfo register and the bsel and csel values match the development kit. the value was 0x84 for SD/MMC 1p8V configuration on my custom card. on the dev kit, the value was x2484, though the x24 was on reserved bits. the rest of the configuration was correct
AS for the questions,
- yes, the image was compiled for my target card
- I used Linux based os
- I compiled it with buildroot
- I am not sure about this.i do not believe I did.
To add what I saw today on te custom card, at powerup, the sd card clock goes to 390KHZ and there are toggles on the cmd line. the initialization part matches on both cards with respect to the waveforms. but after that on the custom card, there are no toggles on the sd clk. I am not sure what this means.
well, it seems to be quite complicated problem.
Could I ask more if You are able to got any messages from Linux kernel, bootloader or even preloader via serial link (if any is there at Your card)?
I faced some times similar problem while booting stucked while U-Boot initializate of some hardware. It was caused by wrong timing of operations which has not been discovered from its source code.
If You could not deduce the source of problem from messages got via serial link, I recommend try to use an older version of Buildroot, Linux and bootloader. May be You has found a bug in the current release.
Which DeviceTree Blob and bootloader are You using? The DeviceTree Blob generated with Linux kernel code differs from that one generated with sopc2dts tool of Indet SoC EDS and it does not reflect configuration of HPS. U-Boot generated with Buildroot also not reflect configuration of HPS and it is of absolutelly differenct version. Moreover, U-Boot and its preloader have not to be compiled separately, because wrong combinations could cause stucking at booting, too.