RTEMS 4.11.0 build fails on Windows
Linda Huxley
lhuxley at shaw.ca
Sun Apr 9 21:33:25 UTC 2017
I am attempting to build the RTEMS tool set, following the instructions
in section "5.2.1. RTEMS Tools and Kernel" of the "RTEMS User Manual"
4.11.99. The instructions for building on Windows are flawed, but I
seem to have worked my way through the issues. I am working in the
MSYS2 MINGW64 build environment. I issue the following commands:
$ cd /c
$ mkdir -p opt/rtems
$ cd opt/rtems
$ wget
https://ftp.rtems.org/pub/rtems/releases/4.11/4.11.0/rtems-source-builder-4.11.0.tar.xz
$ tar Jxf rtems-source-builder-4.11.0.tar.xz
$ mv rtems-source-builder-4.11.0 4.11.0
$ cd 4.11.0/rtems
$ ../source-builder/sb-set-builder \
--prefix=/opt/rtems/4.11.0 4.11/rtems-sparc --jobs=none
About 45 minutes later, the build fails with the following message:
error: building srg4n2xwm1
Build FAILED
See error report:
rsb-report-sparc-rtems4.11-gcc-4.9.3-newlib-2.2.0.20150423-x86_64-w64-mingw32-1.txt
error: building srg4n2xwm1
Build Set: Time 0:40:06.527128
Build FAILED
The salient error messages at the bottom of error report would appear to be:
In file included from ../../gcc-4.9.3/gcc/cp/except.c:1013:0:
cfns.gperf: In function 'const char* libc_name_p(const char*, unsigned
int)':
cfns.gperf:101:1: error: 'const char* libc_name_p(const char*, unsigned
int)' redeclared inline with 'gnu_inline' attribute
cfns.gperf:26:14: note: 'const char* libc_name_p(const char*, unsigned
int)' previously declared here
cfns.gperf: At global scope:
cfns.gperf:26:14: warning: inline function 'const char*
libc_name_p(const char*, unsigned int)' used but never defined
make[2]: *** [Makefile:1058: cp/except.o] Error 1
make[2]: Leaving directory
'/C/opt/rtems/4.11.0/rtems/build/srg4n2xwm1/build/gcc'
make[1]: *** [Makefile:3985: all-gcc] Error 2
make[1]: Leaving directory
'/C/opt/rtems/4.11.0/rtems/build/srg4n2xwm1/build'
make: *** [Makefile:872: all] Error 2
shell cmd failed: sh -ex /C/opt/rtems/4.11.0/rtems/build/srg4n2xwm1/doit
error: building srg4n2xwm1
This would appear to be a bug in gcc itself. I tried making the obvious
changes to cfns.h, but the file is overwritten when I restart RSB.
Can anybody provide a clue as to how to resolve this error? Has anybody
successfully built RTEMS 4.11.0 recently?
More information about the users
mailing list