| Message ID | CAKdteOaqrpv_C__x6osn5db=uH_u2Ci6NDJV3sH1GxFE0Xpa6Q@mail.gmail.com |
|---|---|
| State | New |
| Headers | show |
On Thu, Oct 29, 2015 at 9:25 AM, Christophe Lyon <christophe.lyon@linaro.org> wrote: > Hi, > > On arm*-*-eabi*, arm*-*-symbianelf*, arm*-*-rtems* targets, > crtfastmath.o is not part of $extra_parts, which means that it is not > built, even though the make build rule is present via the inclusion of > t-crtfm. > > The impact is that the link fails when using -ffast-math. > > The attached patch adds crtfastmath.o to $extra_parts, for > arm*-*-eabi*, arm*-*-symbianelf*, arm*-*-rtems* targets. > > I tested on arm-none-eabi, and I am not sure if this is correct for > symbian/rtems. > > OK? Or should I restrict the change to arm*-*-eabi* ? I don't see how they are inappropriate on those targets as crtfastmath.c only set flush to zero in with -ffast-math when the target architecture has floating point hardware and are only brought in if the link spec asks for it (which it does only with -ffast-math) Ok if no regressions. regards Ramana > > Thanks, > > Christophe.
On 29 October 2015 at 11:07, Ramana Radhakrishnan <ramana.gcc@googlemail.com> wrote: > On Thu, Oct 29, 2015 at 9:25 AM, Christophe Lyon > <christophe.lyon@linaro.org> wrote: >> Hi, >> >> On arm*-*-eabi*, arm*-*-symbianelf*, arm*-*-rtems* targets, >> crtfastmath.o is not part of $extra_parts, which means that it is not >> built, even though the make build rule is present via the inclusion of >> t-crtfm. >> >> The impact is that the link fails when using -ffast-math. >> >> The attached patch adds crtfastmath.o to $extra_parts, for >> arm*-*-eabi*, arm*-*-symbianelf*, arm*-*-rtems* targets. >> >> I tested on arm-none-eabi, and I am not sure if this is correct for >> symbian/rtems. >> >> OK? Or should I restrict the change to arm*-*-eabi* ? > > > I don't see how they are inappropriate on those targets as > crtfastmath.c only set flush to zero in with -ffast-math when the > target architecture has floating point hardware and are only brought > in if the link spec asks for it (which it does only with -ffast-math) > > Ok if no regressions. Now committed. For the record, I see +2500 PASS on arm-none-eabi. Christophe. > > regards > Ramana > >> >> Thanks, >> >> Christophe.
diff --git a/libgcc/config.host b/libgcc/config.host index 2ee92c1..b63cc36 100644 --- a/libgcc/config.host +++ b/libgcc/config.host @@ -414,6 +414,7 @@ arm*-*-eabi* | arm*-*-symbianelf* | arm*-*-rtems*) ;; esac tmake_file="$tmake_file t-softfp-sfdf t-softfp-excl arm/t-softfp t-softfp" + extra_parts="$extra_parts crtfastmath.o" unwind_header=config/arm/unwind-arm.h ;; avr-*-rtems*)