According to config.h we are certainly using the native mpn_store. I can't see this being anything other than a compiler bug.
Bill. On 1 April 2014 22:58, Bill Hart <goodwillh...@googlemail.com> wrote: > It looks like we don't pass it anyway when compiling test programs (which > makes sense). We only pass it when compiling library .c files. > > We do seem to link libmpir twice for good measure. I'm amazed that works. > > Bill. > > > On 1 April 2014 22:53, Jean-Pierre Flori <jpfl...@gmail.com> wrote: > >> We shouldn't but on Windows it should make no difference to GCC, and >> as far as I saw, there is no specific PIC code in the assembly files >> chosen on my setup which could be triggered by the "-DPIC" define. >> >> 2014-04-01 22:52 GMT+02:00 Bill Hart <goodwillh...@googlemail.com>: >> > On the other hand, should we be passing -fPIC to gcc? >> > >> > Bill. >> > >> > >> > On 1 April 2014 22:47, Bill Hart <goodwillh...@googlemail.com> wrote: >> >> >> >> Surely a compiler error then. >> >> >> >> I can't seem to get libspeed or try to build on Cygwin64. It looses the >> >> include and lib directories in the gcc invocation. >> >> >> >> But I think you've found the problem, for sure. >> >> >> >> Bill. >> >> >> >> >> >> On 1 April 2014 22:40, Jean-Pierre Flori <jpfl...@gmail.com> wrote: >> >>> >> >>> >> >>> >> >>> On Tuesday, April 1, 2014 10:36:05 PM UTC+2, Jean-Pierre Flori wrote: >> >>>> >> >>>> What look strange to me is the way the function is called. >> >>>> All other function calls use rip or a different address range. >> >>>> >> >>>> E.g. for the call in t-neg: >> >>>> 39 gmp_randinit_default(rands); >> >>>> >> >>>> First it goes through the import lib: >> >>>> 0x00000001004010fa <+42>: mov 0x818f(%rip),%rax # >> >>>> 0x100409290 <__imp___gmp_randinit_default> >> >>>> 0x0000000100401101 <+49>: callq *%rax >> >>>> >> >>>> And %rax looks saner: >> >>>> (gdb) info reg >> >>>> rax 0x4ff1e1550 21460030800 >> >>>> >> >>>> Just as in cygmpir-16.dll: >> >>>> 00000004ff1e1550 T __gmp_randinit_default >> >>>> >> >>>> In particularn in the mpn_store direct call, the leading 4 is >> missing... >> >>> >> >>> >> >>> And: >> >>> (gdb) r >> >>> The program being debugged has been started already. >> >>> Start it from the beginning? (y or n) y >> >>> >> >>> Starting program: /home/jp/mpir-flori.git/tests/mpn/.libs/t-neg >> >>> [New Thread 3836.0x408] >> >>> [New Thread 3836.0x26c] >> >>> >> >>> Program received signal SIGSEGV, Segmentation fault. >> >>> 0x00000000ff2746c0 in ?? () >> >>> (gdb) x/i $pc >> >>> => 0xff2746c0: Cannot access memory at address 0xff2746c0 >> >>> (gdb) x/i $pc + 0x400000000 >> >>> 0x4ff2746c0 <__gmpn_store>: lea -0x20(%rcx),%rcx >> >>> (gdb) >> >>> >> >> >> > >> >> >> >> -- >> Jean-Pierre Flori >> > > -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to mpir-devel+unsubscr...@googlegroups.com. To post to this group, send email to mpir-devel@googlegroups.com. Visit this group at http://groups.google.com/group/mpir-devel. For more options, visit https://groups.google.com/d/optout.
