Ah, that makes sense. Does configuring with `--enable-mach=arm32le` work?
Using "arm32le" instead of the inferred "tarm32le" avoids memory-fence instructions, so it should solve this problem, but I'm not certain the rest of the build will adapt correctly. At Tue, 11 Aug 2020 14:35:49 -0700 (PDT), Joey Eremondi wrote: > I'm on an old RPi B (maybe a B+). It's pretty ancient, so I might be > pushing my luck. Here's my CPU info: > > pi@raspberrypi:~ $ cat /proc/cpuinfo > processor : 0 > model name : ARMv6-compatible processor rev 7 (v6l) > BogoMIPS : 697.95 > Features : half thumb fastmult vfp edsp java tls > CPU implementer : 0x41 > CPU architecture: 7 > CPU variant : 0x0 > CPU part : 0xb76 > CPU revision : 7 > > Hardware : BCM2835 > Revision : 000e > Serial : 00000000e9b1ce2d > Model : Raspberry Pi Model B Rev 2 > > Running GDB gives the illegal instruction as > <UNDEFINED> instruction: 0xf57ff05a > which looks to be a "Data Memory Barrier" instruction? Sounds like a > plausible culprit. > > Thanks! > > > > On Tuesday, August 11, 2020 at 2:14:28 PM UTC-7, Matthew Flatt wrote: > > > > Which model Pi are you using? I'm able to build on a Pi 3, so I wonder > > if it's a difference in processors, where the Arm32 backend is using > > something that it shouldn't. > > > > Whether or not that guess is right, can you try running `gdb` in the > > "ChezScheme" directory like this? > > > > env SCHEMEHEAPDIRS=tarm32le/boot/tarm32le/ \ > > gdb tarm32le/bin/tarm32le/scheme > > > > Disassembling around the failed instruction address (with `disassemble` > > in gdb) should clarify what instruction is misused. > > > > Thanks, > > Matthew > > > > At Tue, 11 Aug 2020 10:58:03 -0700 (PDT), Joey Eremondi wrote: > > > I'm wondering, does anybody have any prebuilt 32-but ARM binaries for > > > Racket on Chez? I'm trying to run a little web-server on a raspberry pi, > > > and I'd prefer to use the Chez version, but I can't seem to get it to > > build. > > > > > > Alternately, if anybody knows why it won't build and can help, that > > would > > > help. With eiether --enable-csracket or --enable-csonly, it gets stuck > > > compiling the forked version of ChezScheme. Surprisingly, it doesn't > > seem > > > to be an out of memory error. Is the illegal instruction error because > > it's > > > ARMv6 and not AArch32? > > > > > > ... > > > gcc -Wpointer-arith -Wextra -Werror -Wno-implicit-fallthrough -O2 -g > > -O2 - > > > Wall -DELF_FIND_BOOT_SECTION -pthread -rdynamic -o > > ../bin/tarm32le/scheme > > > ../boot/tarm32le/main.o ../boot/tarm32le/libkernel.a -lm -ldl -lpthread > > -lrt > > > -luuid ../zlib/libz.a ../lz4/lib/liblz4.a -pthread > > > (cd s && make bootstrap) > > > make allx > > > rm -f *.tarm32le xpatch patch *.patch *.so *.covin *.asm script.all > > header.tmp > > > *.html > > > rm -rf nanopass > > > cp -p -f ../boot/tarm32le/petite.boot ../boot/tarm32le/sbb > > > cp -p -f ../boot/tarm32le/scheme.boot ../boot/tarm32le/scb > > > make all > > > echo '(reset-handler abort)'\ > > > '(base-exception-handler (lambda (c) (fresh-line) > > > (display-condition c) (newline) (reset)))'\ > > > '(keyboard-interrupt-handler (lambda () (display > > > "interrupted---aborting\n") (reset)))'\ > > > '(optimize-level 3)'\ > > > '(debug-level 0)'\ > > > '(commonization-level (commonization-level))'\ > > > '(compile-compressed #t)'\ > > > '(compress-format (compress-format))'\ > > > '(compress-level (compress-level))'\ > > > '(generate-inspector-information #f)'\ > > > '(subset-mode (quote system))'\ > > > '(compile-file "cmacros.ss" "cmacros.so")'\ > > > | ../bin/tarm32le/scheme -q > > > Error: illegal instruction > > > () > > > make[10]: *** [Mf-base:375: cmacros.so] Error 1 > > > make[9]: *** [Mf-base:179: allx] Error 2 > > > make[8]: *** [Mf-base:196: bootstrap] Error 2 > > > make[7]: *** [Makefile:22: build] Error 2 > > > make[6]: *** [Makefile:20: build] Error 2 > > > make[6]: Leaving directory '/home/pi/gh/racket-7.8/src/ChezScheme' > > > make[5]: *** [Makefile:158: scheme-make-finish] Error 2 > > > make[5]: Leaving directory '/home/pi/gh/racket-7.8/src/cs/c' > > > make[4]: *** [Makefile:148: scheme-make-copy] Error 2 > > > make[4]: Leaving directory '/home/pi/gh/racket-7.8/src/cs/c' > > > make[3]: *** [Makefile:137: scheme] Error 2 > > > make[3]: Leaving directory '/home/pi/gh/racket-7.8/src/cs/c' > > > make[2]: *** [Makefile:60: cs] Error 2 > > > make[2]: Leaving directory '/home/pi/gh/racket-7.8/src/cs/c' > > > make[1]: *** [Makefile:90: racketcs] Error 2 > > > make[1]: Leaving directory '/home/pi/gh/racket-7.8/src' > > > make: *** [Makefile:53: all] Error 2 > > > > > > > > > -- > > > You received this message because you are subscribed to the Google > > Groups > > > "Racket Users" group. > > > To unsubscribe from this group and stop receiving emails from it, send > > an > > > email to [email protected] <javascript:>. > > > To view this discussion on the web visit > > > > > > https://groups.google.com/d/msgid/racket-users/9b0c4921-2155-4f63-bb79-a6b9ea05 > > > > 8ef6o%40googlegroups.com. > > > > -- > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/racket-users/eca60f12-db4b-4f79-ae2a-681a0089 > 8ba4o%40googlegroups.com. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/racket-users/20200811160247.294%40sirmail.smtps.cs.utah.edu.
