On 2017-Jan-21, at 2:16 PM, Mark Millard <markmi at dsl-only.net> wrote:
> For:
>
> # more main.c
>
> static volatile char
> big_area[67001] = "This is a test";
>
> int main ()
> {
> big_area[67000] = '9';
> }
>
> I get (the -fPIE is not required for the behavior):
>
> # clang -fuse-ld=lld -Wl,-t -pie -fPIE main.c
> /usr/lib/Scrt1.o
> /usr/lib/crti.o
> /usr/lib/crtbeginS.o
> /tmp/main-c6f752.o
> /usr/lib/libgcc_s.so
> /lib/libc.so.7
> /usr/lib/libgcc_s.so
> /usr/lib/crtendS.o
> /usr/lib/crtn.o
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> can't create dynamic relocation R_PPC64_REL24 against readonly segment
> clang: error: linker command failed with exit code 1 (use -v to see
> invocation)
>
> (This is difficult to study because it does not leave even a
> partial a.out and it does not report the specifics of what
> segment or what symbol or the like.)
>
> Even an empty source produces that, but also including the
> expected:
>
> undefined symbol: main in /usr/lib/Scrt1.o
>
> It appears that the R_PPC64_REL24's are some subset of. . .
>
> /usr/lib/Scrt1.o :
>
> 00000000000000cc R_PPC64_REL24 atexit
> 00000000000000d8 R_PPC64_REL24 _init_tls
> 00000000000000f8 R_PPC64_REL24 atexit
> 00000000000001b4 R_PPC64_REL24 _init
> 0000000000000240 R_PPC64_REL24 main
> 0000000000000248 R_PPC64_REL24 exit
> 00000000000002fc R_PPC64_REL24 _fini
>
> (That would be 6 by ignoring main, matching
> the message count.)
>
> /usr/lib/crtbeginS.o :
>
> 0000000000000040 R_PPC64_REL24 __cxa_finalize
> 0000000000000000 R_PPC64_REL24 .opd
> 0000000000000000 R_PPC64_REL24 .opd+0x0000000000000018
>
> /usr/lib/crtendS.o :
>
> 0000000000000000 R_PPC64_REL24 .opd
>
> main.o , if I have it produced, does not have R_PPC64_REL24
> in it.
I have submitted this as llvm bugzilla 31762 and added
it to llvm bugzilla 25780's Depends On list (the meta
submittal for using clang and [its toolchain] as the
FreeBSD powerpc and powerpc64 system compiler [and
toolchain]).
===
Mark Millard
markmi at dsl-only.net
_______________________________________________
freebsd-toolchain@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-toolchain
To unsubscribe, send any mail to "freebsd-toolchain-unsubscr...@freebsd.org"
