Re: [OpenOCD-user] Bunch of problems

[Tarek B. <tar...@gm...>]

again on this:
Info : 715 209748 aarch64.c:1041 aarch64_post_debug_entry(): cannot read
system control register in this mode
this is caused by the fact that HYP mode is managed in
aarch64_post_debug_entry
but in this case  aarch64_post_debug_entry return ERROR_FAIL
causing aarch64_debug_entry and aarch64_poll to fail too.

I adedd this :

http://openocd.zylin.com/#/c/5686/1



but then I noticed that there is already a change regarding HYP waiting in :

http://openocd.zylin.com/#/c/5255/1


could you please check it, before I abandon my 2 small patches


Le sam. 16 mai 2020 à 14:37, Mario Palomares <pal...@gm...>
a écrit :

> Hello everyone,
>
> I am trying to debug some RTEMS executable on my RPI2 v1.2 (it has BCM2837
> with ARMv8-A53) and right know i get successfully the conection but i am
> experiencing some strange behaviours when connecting from telnet and from
> GDB.
>
> *TELNET PROBLEMS*
>
> As you may know BCM2837 SoC has 4 cores, and by default the RTEMS
> executable is loaded in core0. Whenever i halt that CPU (cpu0) everything
> goes right but when i try to halt any other it gives me this error:
>
> Debug: 695 204314 command.c:143 script_debug(): command - targets targets
> Debug: 699 208650 command.c:143 script_debug(): command - halt halt
> Debug: 701 208801 target.c:3028 handle_halt_command(): -
> Debug: 702 208802 aarch64.c:368 aarch64_halt_one(): rpi2.a53.3
> Debug: 704 209102 aarch64.c:959 aarch64_debug_entry(): rpi2.a53.3 dscr =
> 0x03044213
> Debug: 706 209288 armv8_dpm.c:665 dpmv8_read_reg(): READ: x0,
> 00000003
> Debug: 707 209341 armv8_dpm.c:665 dpmv8_read_reg(): READ: x1,
> 00000131
> Error: 708 209422 armv8.c:117 armv8_mode_name(): unrecognized psr mode:
> 0x1a
> Debug: 709 209423 armv8.c:713 armv8_set_cpsr(): set CPSR 0x600001da:
> UNRECOGNIZED mode, AArch32 state
> Debug: 710 209503 armv8_dpm.c:665 dpmv8_read_reg(): READ: pc,
> 000000a8
> Debug: 711 209585 armv8_dpm.c:665 dpmv8_read_reg(): READ: ELR_EL2,
> 393c03b1
> Debug: 712 209666 armv8_dpm.c:667 dpmv8_read_reg(): READ: ESR_EL2, 68993d79
> Debug: 714 209748 armv8_dpm.c:667 dpmv8_read_reg(): READ: SPSR_EL2,
> 600001da
> Info : 715 209748 aarch64.c:1041 aarch64_post_debug_entry(): cannot read
> system control register in this mode
>
> Another thing I think is related to this happens whenever i try to load a
> program from my host computer to memory. If load the program on the CPU0
> (the one with the RTEMS executable) y get the following error:
>
> Debug: 1190 1125169 command.c:143 script_debug(): command - load_image
> load_image /home/mario/Documents/fastblink.exe
> Debug: 1192 1125320 configuration.c:97 find_file(): found
> /home/mario/Documents/fastblink.exe
> Debug: 1193 1125322 image.c:69 autodetect_image_type(): ELF image detected.
> Debug: 1194 1125322 configuration.c:97 find_file(): found
> /home/mario/Documents/fastblink.exe
> Debug: 1195 1125322 image.c:494 image_elf_read_section(): load segment 0
> at 0x0 (sz = 0xd4)
> Debug: 1196 1125322 image.c:501 image_elf_read_section(): read elf: size =
> 0x212 at 0x8000
> Debug: 1197 1125322 target.c:2148 target_write_buffer(): writing buffer of
> 212 byte at 0x00008000
> Error: 1199 1125836 aarch64.c:1884 aarch64_write_cpu_memory(): abort
> occurred - dscr = 0x03044153
> Debug: 1200 1126035 armv8_dpm.c:1338 armv8_dpm_handle_exception():
> Exception taken to EL 1, DLR=0x0000000000200010 DSPSR=0x400f0173
> Debug: 1201 1126035 armv8_dpm.c:560 armv8_dpm_modeswitch(): restoring
> mode, cpsr = 0x400f0173
> Debug: 1202 1126035 armv8_dpm.c:598 armv8_dpm_modeswitch(): target_el = 1,
> last_el = 1
> Debug: 1203 1126035 armv8_dpm.c:611 armv8_dpm_modeswitch(): SPSR =
> 0x400f0173
> Debug: 1204 1126127 command.c:630 run_command(): Command 'load_image'
> failed with error code -4
> User : 1205 1126127 command.c:695 command_run_line():
>
> But no any other CPU it loads the program correctly
>
> *GDB PROBLEMS*
>
> Related to this i get two different errors. One is that there is a
> periodically message telling:
>
> Debug: 1281 1253785 log.c:409 keep_alive(): keep_alive() was not invoked
> in the 1000ms timelimit (1853). This may cause trouble with GDB connections.
>
> The other one i have also asked in the RTEMS mailing list but i tell it
> also here in case you know. Apparently the gdb versión the RTEMS source
> builder builds isn´t compatible with the gdbserver openOCD starts. The
> RTEMS gdb configuration is the following:
>
>  configure --host=x86_64-linux-gnu --target=arm-rtems5
>              --with-auto-load-dir=$debugdir:$datadir/auto-load
>              --with-auto-load-safe-path=$debugdir:$datadir/auto-load
>              --with-expat
>
>  --with-gdb-datadir=/home/mario/development/rtems/compiler/5/share/gdb
> (relocatable)
>
>  --with-jit-reader-dir=/home/mario/development/rtems/compiler/5/lib/gdb
> (relocatable)
>              --without-libunwind-ia64
>              --with-lzma
>              --with-babeltrace
>              --without-intel-pt
>              --disable-libmcheck
>              --with-mpfr
>              --with-python=/usr
>              --without-guile
>              --disable-source-highlight
>
>  --with-separate-debug-dir=/home/mario/development/rtems/compiler/5/lib/debug
> (relocatable)
>
> The error i get when i connect is the following
>
> warning: while parsing target description (at line 4): Target description
> specified unknown architecture "aarch64"
> warning: Could not load XML target description; ignoring
> Truncated register 16 in remote 'g' packet
>
> I have tried with gdb-multiarch but get the following error:
>
> warning: Architecture rejected target-supplied description
>
> I know those are lots of question but i would appreciate on any of them.
>
> If you need more info about my environment let me know
>
> Regards,
> Mario
>
> _______________________________________________
> OpenOCD-user mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openocd-user
>