On 9/28/20 3:42 PM, Simon Glass wrote: > Hi Heinrich, > > On Mon, 28 Sep 2020 at 07:30, Heinrich Schuchardt <[email protected]> wrote: >> >> On 28.09.20 15:22, Simon Glass wrote: >>> Hi Heinrich, >>> >>> On Mon, 28 Sep 2020 at 05:31, Heinrich Schuchardt <[email protected]> >>> wrote: >>>> >>>> On 28.09.20 06:46, Heinrich Schuchardt wrote: >>>>> Am 28. September 2020 06:24:38 MESZ schrieb Simon Glass >>>>> <[email protected]>: >>>>>> Hi Heinrich, >>>>>> >>>>>> On Sat, 19 Sep 2020 at 13:48, Heinrich Schuchardt <[email protected]> >>>>>> wrote: >>>>>>> >>>>>>> Hello Simon, >>>>>>> >>>>>>> when I try to run ./u-boot -l the sandbox stalls. Shouldn't it run >>>>>> out >>>>>>> of the box? >>>>>>> >>>>>>> $ ./u-boot -l -d arch/sandbox/dts/sandbox.dtb >>>>>> >>>>>> For the record you should be able to use -D to get the same effect as >>>>>> your -d above. >>>>>> >>>>>>> >>>>>>> U-Boot 2020.10-rc4-00018-g21a10244f9-dirty (Sep 19 2020 - 19:55:39 >>>>>> +0200) >>>>>>> >>>>>>> Model: sandbox >>>>>>> DRAM: 128 MiB >>>>>>> >>>>>>> Warning: host_lo MAC addresses don't match: >>>>>>> Address in ROM is 26:4b:ca:6c:98:f4 >>>>>>> Address in environment is 00:00:11:22:33:44 >>>>>>> >>>>>>> Warning: host_virbr0 MAC addresses don't match: >>>>>>> Address in ROM is ee:3e:c9:ce:1f:9c >>>>>>> Address in environment is 00:00:11:22:33:45 >>>>>>> >>>>>>> Warning: host_docker0 MAC addresses don't match: >>>>>>> Address in ROM is c2:85:07:7b:9a:18 >>>>>>> Address in environment is 00:00:11:22:33:46 >>>>>>> WDT: Not found! >>>>>>> MMC: >>>>>>> >>>>>>> No output after this point. >>>>>>> >>>>>>> The problem also exists with U-Boot v2020.07, v2019.10, v2018.11. >>>>>>> >>>>>>> CONFIG_SANDBOX_SDL=y >>>>>>> >>>>>>> SDL_InitSubSystem() never returns. It is looping somewhere in >>>>>> U-Boot's >>>>>>> __serial_getc(). I wonder how it gets there without returning from >>>>>> the >>>>>>> function. >>>>>>> >>>>>>> I compiled SDL2.cpp from >>>>>>> https://gist.github.com/miguelmartin75/6946310#file-sdl2-cpp-L18 >>>>>>> with >>>>>>> >>>>>>> g++ SDL2.cpp -D_REENTRANT -I/usr/include/SDL2 -lSDL2 -lGL -o test >>>>>>> >>>>>>> and it runs fine showing an X11 windows with red background. >>>>>>> >>>>>>> So there seems to be no general problem with the SDL2 library. >>>>>> >>>>>> I hit this myself on another computer and it turned out to be that SDL >>>>>> defined getc(), as does U-Boot, and things get confused. At least I >>>>>> think it is getc. >>>>>> >>>>>> I hacked around with changing the name of getc (I think it was getc) >>>>>> in U-Boot and the problem went away. >>>>> >>>>> Should we include a patched SDL2 with U-Boot for static linking? >>>>> >>>> >>>> I cannot find a symbol getc() nor a reference to it in the SDL >>>> repository. getc() is defined in stdio.h. >>>> >>>> Our getc() takes no argument, while the stdio one wants a FILE *. >>>> >>>> But changing the U-Boot definition to "int getc(void *)" does not solve >>>> the issue. >>> >>> I just tried it on the machine where it doesn't work: >>> >>> $ gdb --args /tmp/b/sandbox/u-boot -l -D >>> GNU gdb (Debian 9.2-1) 9.2 >>> Copyright (C) 2020 Free Software Foundation, Inc. >>> License GPLv3+: GNU GPL version 3 or later >>> <http://gnu.org/licenses/gpl.html> >>> This is free software: you are free to change and redistribute it. >>> There is NO WARRANTY, to the extent permitted by law. >>> Type "show copying" and "show warranty" for details. >>> This GDB was configured as "x86_64-linux-gnu". >>> Type "show configuration" for configuration details. >>> For bug reporting instructions, please see: >>> <http://www.gnu.org/software/gdb/bugs/>. >>> Find the GDB manual and other documentation resources online at: >>> <http://www.gnu.org/software/gdb/documentation/>. >>> >>> For help, type "help". >>> Type "apropos word" to search for commands related to "word"... >>> Reading symbols from /tmp/b/sandbox/u-boot... >>> (gdb) br getc >>> Breakpoint 1 at 0x5b6d6: getc. (2 locations) >>> (gdb) r >>> Starting program: /tmp/b/sandbox/u-boot -l -D >>> [Thread debugging using libthread_db enabled] >>> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". >>> >>> >>> U-Boot 2020.10-rc5-00196-g0ac83d080a0 (Sep 28 2020 - 07:11:52 -0600) >>> >>> Model: sandbox >>> DRAM: 128 MiB >>> >>> Warning: host_lo MAC addresses don't match: >>> Address in ROM is 1a:34:9b:48:aa:53 >>> Address in environment is 00:00:11:22:33:44 >>> WDT: Not found! >>> MMC: >>> >>> Breakpoint 1, getc () at >>> /home/sjg/cosarm/src/third_party/u-boot/files/common/console.c:414 >>> 414 if (!gd->have_console) >>> (gdb) up >>> #1 0x00007ffff7914d48 in ?? () from /lib/x86_64-linux-gnu/libX11.so.6 >>> (gdb) >>> >>> >>> So it seems that it is libX11 causing the problem. >>> >>> I don't think the best fix is to remove getc() from that library, >>> although I do wonder if it is a bug. Renaming the symbol in U-Boot >>> with #define (only on sandbox) might be one option. >>> >>> Regards, >>> Simon >>> >> >> #define getc _uboot_getc >> >> is what I tried but it runs you into the trouble that many other symbols >> contain the getc substring. >> >> As we always try to use the same definition in U-Boot as in glibc we >> should really replace getc() by another symbol, e.g. chget() so that we >> avoid confusion. > > Well the first question is whether this is a bug with x11. > > I'm not sure that renaming is a good idea. getc() is a pretty standard name. > > Try > > #define getc(x) _sandbox_get(x) > > which might get you closer. > > Regards, > Simon > https://gitlab.denx.de/u-boot/custodians/u-boot-efi/-/commit/7b302bedb21f94b9eb619cb38ef985f7bbb9be1f
seems to fix the crash. I get an output window. But there are still problems: * The shift key does not work on the keyboard. * The output is black and white only. What I would like to have is a 32bit color framebuffer on which I can draw with the EFI graphical output protocol. Best regards Heinrich
