Re: Error compiling gcc for fr30

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Kai Ruottu wrote:
Your "optimization level" in compiling libstdc++-v3 was just the same '-Os' I had here, so it must come from somewhere and be 'fr30' dependent. Maybe changing it to '-O2' or to '-O' could help, I will check that... Anyway the '-Os' seems to cause a crash in libstdc++-v3 with 'fr30-elf' !
Been there, seen that... Only switching to '-O0' enabled the compile to succeed! So I produced this single module with -O0 and switching back to '-Os' with the next in turn helped only a while,
then it crashed again with:

/data1/home/src/gcc-4.1.1/build/fr30-elf/libstdc++-v3/include/bits/locale_facets.h: In member function '_InIter std::money_get<_CharT, _InIter>::get(_InIter, _InIter, bool, std::ios_base&, std::_Ios_Iostate&, long double&) const [with _CharT = char, _InIter = std::istreambuf_iterator<char, std::char_traits<char> >]': /data1/home/src/gcc-4.1.1/build/fr30-elf/libstdc++-v3/include/bits/locale_facets.h:4023: error: insn does not satisfy its constraints: (insn 140 139 78 0 (set (mem/s/c:SI (plus:SI (reg/f:SI 14 fp [orig:32 D.35158 ][32])
               (const_int 4 [0x4])) [94 <result>+4 S4 A32])
       (reg:SI 2 r2 [orig:67+4 ] [67])) 6 {movsi_internal} (nil)
   (expr_list:REG_DEAD (reg:SI 2 r2 [orig:67+4 ] [67])
       (nil)))
/data1/home/src/gcc-4.1.1/build/fr30-elf/libstdc++-v3/include/bits/locale_facets.h:4023: internal compiler error: in copyprop_hardreg_forward_1, at regrename.c:1592

Compiling this too with '-O0' could help and then compiling again with '-Os' could get it forwards to the next crash... Then the same thing again and again and finally the build could succeed... But
"What is broken, that is broken until being fixed!" :-(

Looking at the function 'copyprop_hardreg_forward_1()' at 'regrename.c' and the line 1592 there didn't help much. So making a bug report could help... BTW, the fr30 port maintainer was told to be Nick Clifton, nickc@xxxxxxxxxx, so if you don't get any response, trying to contact him could be one way, but maybe these crashes could have given enough info for some other to fix the problem. I myself have thought to learn the RTL "Really Soon Now" but let's see...


[Index of Archives]     [Linux C Programming]     [Linux Kernel]     [eCos]     [Fedora Development]     [Fedora Announce]     [Autoconf]     [The DWARVES Debugging Tools]     [Yosemite Campsites]     [Yosemite News]     [Linux GCC]

  Powered by Linux