Message ID | 1396401134-21858-2-git-send-email-gregory.0xf0@gmail.com |
---|---|
State | Accepted |
Commit | 85891447211add633243af01e960762c9a57b435 |
Headers | show |
On Tue, Apr 01, 2014 at 06:12:13PM -0700, Gregory Fong wrote: > Using TARGET_ARCH isn't sufficient when ISA can differ. Add variable > TEST_ARCH for selecting the correct ULPs file for mips32 and mips64. > > Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com> > --- > > This is a bit clunky, anyone know how to better handle this? Applied a slightly adjusted variant, please check. Thanks!
On Fri, Apr 4, 2014 at 7:17 AM, Bernhard Reutner-Fischer <rep.dot.nop@gmail.com> wrote: > On Tue, Apr 01, 2014 at 06:12:13PM -0700, Gregory Fong wrote: >> Using TARGET_ARCH isn't sufficient when ISA can differ. Add variable >> TEST_ARCH for selecting the correct ULPs file for mips32 and mips64. >> >> Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com> >> --- >> >> This is a bit clunky, anyone know how to better handle this? > > Applied a slightly adjusted variant, please check. Looks good to me. Tangentially, has anyone else noticed that the uClibc cgit doesn't show the most recent history when you go to http://git.uclibc.org/uClibc/ and click "log" at the top or select master under "Branch"?
On Fri, Apr 04, 2014 at 03:12:15PM -0700, Gregory Fong wrote: > Tangentially, has anyone else noticed that the uClibc cgit doesn't > show the most recent history when you go to > http://git.uclibc.org/uClibc/ and click "log" at the top or select > master under "Branch"? Yes, buildroot thought it was due to caching but we turned that off ATM and the bug still is exposed, so sounds like a cgit thing..
diff --git a/test/math/Makefile.in b/test/math/Makefile.in index 0285b66..ed0a0d6 100644 --- a/test/math/Makefile.in +++ b/test/math/Makefile.in @@ -28,7 +28,17 @@ PERL := /usr/bin/perl MDEPS := $(wildcard test-*.c) $(MDEPS): libm-test.c -libm-test.c: libm-test-ulps-$(TARGET_ARCH) libm-test.inc gen-libm-test.pl - $(Q)$(PERL) ./gen-libm-test.pl -u libm-test-ulps-$(TARGET_ARCH) ./libm-test.inc -o "." 2>&1 > /dev/null +ifeq ($(TARGET_ARCH),mips) +ifneq ($(CONFIG_MIPS_ISA_MIPS64R1)$(CONFIG_MIPS_ISA_MIPS64R2),) +TEST_ARCH := mips64 +else +TEST_ARCH := mips32 +endif +else # ($(TARGET_ARCH),mips) +TEST_ARCH := $(TARGET_ARCH) +endif + +libm-test.c: libm-test-ulps-$(TEST_ARCH) libm-test.inc gen-libm-test.pl + $(Q)$(PERL) ./gen-libm-test.pl -u libm-test-ulps-$(TEST_ARCH) ./libm-test.inc -o "." 2>&1 > /dev/null EXTRA_CLEAN := libm-test.c libm-test-ulps.h
Using TARGET_ARCH isn't sufficient when ISA can differ. Add variable TEST_ARCH for selecting the correct ULPs file for mips32 and mips64. Signed-off-by: Gregory Fong <gregory.0xf0@gmail.com> --- This is a bit clunky, anyone know how to better handle this? test/math/Makefile.in | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-)