| Message ID | 5c8ce41d4341aa5c0698abbc0284359a789d2334.1456117980.git.sam.bobroff@au1.ibm.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Headers | show |
On Mon, 2016-02-22 at 16:13 +1100, Sam Bobroff wrote: > It can currently be difficult to diagnose a build that fails due to > the compiler, linker or other parts of the toolchain being unable to > build binaries of the type required by the kernel config. For example > using a little endian toolchain to build a big endian kernel may > produce: > > as: unrecognized option '-maltivec' > > This patch adds a basic compile test and error message to > arch/powerpc/Makefile so that the above error becomes: > > *** Sorry, your toolchain seems to be broken or incorrect. *** > Make sure it supports your kernel configuration (ppc64). > > Signed-off-by: Sam Bobroff <sam.bobroff@au1.ibm.com> > --- How is this more useful than getting to actually see the way in which the toolchain (or the CFLAGS) is broken? -Scott
On Mon, Feb 22, 2016 at 08:05:01PM -0600, Scott Wood wrote: > On Mon, 2016-02-22 at 16:13 +1100, Sam Bobroff wrote: > > It can currently be difficult to diagnose a build that fails due to > > the compiler, linker or other parts of the toolchain being unable to > > build binaries of the type required by the kernel config. For example > > using a little endian toolchain to build a big endian kernel may > > produce: > > > > as: unrecognized option '-maltivec' > > > > This patch adds a basic compile test and error message to > > arch/powerpc/Makefile so that the above error becomes: > > > > *** Sorry, your toolchain seems to be broken or incorrect. *** > > Make sure it supports your kernel configuration (ppc64). > > > > Signed-off-by: Sam Bobroff <sam.bobroff@au1.ibm.com> > > --- > > How is this more useful than getting to actually see the way in which the > toolchain (or the CFLAGS) is broken? My reasoning was that it would be better because it happens at the start of the build, rather than (possibly) a long way into it, and it indicates that the problem is the toolchain setup (or config) itself rather than the file it's trying to compile or link. But I agree completely with what you're saying. I'll try re-working it in a way that shows the command that fails and it's output. Cheers, Sam.
diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile index 96efd82..0041cd2 100644 --- a/arch/powerpc/Makefile +++ b/arch/powerpc/Makefile @@ -355,6 +355,13 @@ TOUT := .tmp_gas_check # - Require gcc 4.0 or above on 64-bit # - gcc-4.2.0 has issues compiling modules on 64-bit checkbin: + @if test "$(call try-run,echo 'int _start(void) { return 0; }' > \"$$TMP\"; \ + $(CC) $(KBUILD_CPPFLAGS) $(KBUILD_CFLAGS) -x c -nostdlib \"$$TMP\" \ + -o /dev/null,ok,broken)" = "broken" ; then \ + echo "*** Sorry, your toolchain seems to be broken or incorrect. ***" ; \ + echo "Make sure it supports your kernel configuration ($(UTS_MACHINE))." ; \ + false; \ + fi @if test "$(cc-name)" != "clang" \ && test "$(cc-version)" = "0304" ; then \ if ! /bin/echo mftb 5 | $(AS) -v -mppc -many -o $(TOUT) >/dev/null 2>&1 ; then \
It can currently be difficult to diagnose a build that fails due to the compiler, linker or other parts of the toolchain being unable to build binaries of the type required by the kernel config. For example using a little endian toolchain to build a big endian kernel may produce: as: unrecognized option '-maltivec' This patch adds a basic compile test and error message to arch/powerpc/Makefile so that the above error becomes: *** Sorry, your toolchain seems to be broken or incorrect. *** Make sure it supports your kernel configuration (ppc64). Signed-off-by: Sam Bobroff <sam.bobroff@au1.ibm.com> --- arch/powerpc/Makefile | 7 +++++++ 1 file changed, 7 insertions(+)