Ticket #11226: sympow-1.018.1.p8-p9.diff

File sympow-1.018.1.p8-p9.diff, 1.4 KB (added by jdemeyer, 11 years ago)

Diff for the sympow spkg, for reviewing only

  • SPKG.txt

    diff -r f281f846707a SPKG.txt
    a b  
    4646
    4747== Changelog ==
    4848
     49=== sympow-1.018.1.p9 (Jeroen Demeyer, 2 May 2011) ===
     50 * #11226: Add flag -fno-expensive-optimizations when compiling with
     51   gcc 4.6.x on a ia64 system.  See also gcc bugzilla:
     52   http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48823
     53
    4954=== sympow-1.018.1.p8 (David Kirkby, 21st August 2010) ===
    5055 * #9703 + #9166 Implement inline assembly code to set the control
    5156   word of the floating point processor to round to an
  • spkg-install

    diff -r f281f846707a spkg-install
    a b  
    2121   export LDFLAGS
    2222fi
    2323
     24# Workaround problems with gcc 4.6.0 on ia64 systems:
     25if [ "x`uname -m`" = "xia64" ] && [ "x`$SAGE_LOCAL/bin/testcc.sh $CC`" = xGCC ] ; then
     26    if $CC -dumpversion 2>/dev/null |grep >/dev/null '^4\.6\.' ; then
     27        echo "Detected GCC version `$CC -dumpversion` on Itanium. Adding"
     28        echo "work-around to fix a bug when sympow is compiled with gcc 4.6.0"
     29        CFLAGS="$CFLAGS -fno-expensive-optimizations"
     30        export CFLAGS
     31    fi
     32fi
     33
     34
    2435# Add some code to disable the extended precision of Intel
    2536# x86 based floating point processors. The code in SYMPOW
    2637# already works for Linux, so there's no need to do that.