On 21/03/2018 18:02, Michael S. Tsirkin wrote: > On Wed, Mar 21, 2018 at 03:36:51PM +0000, Daniel P. Berrangé wrote: >> On Wed, Mar 21, 2018 at 05:22:07PM +0200, Michael S. Tsirkin wrote: >>> Our rule right now is to use <> for external headers, >>> "" for internal ones. The idea was to avoid conflicts >>> between e.g. a system file named <trace.h> and an >>> internal one by the same name. >>> >>> Unfortunately we use -I compiler flag so it does not >>> help: a system file doing #include <trace.h> will >>> still pick up ours first. >> >> Has that actually hit you practice ? I'm wondering if this >> is just theoretical in which case it could wait till 2.13 >> since this scenario has existed in QEMU along time, or >> a real problem right now requiring fix in 2.12. > > It's borderline. > > In particular there is at least one instance where we already have a > conflict (yes, there are probably other ways to address this, but > I consider them less robust): > > util/getauxval.c:#include <sys/auxv.h> > /usr/include/sys/auxv.h:#include <elf.h> > > So it really does build by luck. > > >>> To fix, switch to -iquote which is supported by both >>> gcc and clang and only affects #include "" directives. >> >> Fine since we don't support anything other than >> gcc and clang. I'm assuming -iquote has been supported >> by these two compilers for a long time though ? The GCC >> docs annoyingly don't ever mention what release features >> appear in :-( > > Judging by git log output: > > For clang it appears to be there since 2007 that's pretty > close to day 1. > > gcc's log is better documented, it's been there since 2004-05-03 > >>> >>> As a side effect, this catches any future uses of >>> #include <> for internal headers. >>> >>> Suggested-by: Stefan Weil <s...@weilnetz.de> >>> Signed-off-by: Michael S. Tsirkin <m...@redhat.com> >>> --- >> >> Reviewed-by: Daniel P. Berrangé <berra...@redhat.com> > > > Thanks! >
Hi Michael, Should I add this patch to the RDMA pull request? Thanks, Marcel >> >>> >>> I still think we want to switch to a more formal rule such as qemu/ >>> prefix for all includes down the road, but this will at least catch any >>> scheme violations from creeping in meanwhile. >>> >>> >>> configure | 16 ++++++++-------- >>> rules.mak | 2 +- >>> Makefile.target | 4 ++-- >>> 3 files changed, 11 insertions(+), 11 deletions(-) >>> >>> diff --git a/configure b/configure >>> index af72fc8..23a4f3b 100755 >>> --- a/configure >>> +++ b/configure >>> @@ -534,7 +534,7 @@ QEMU_CFLAGS="-fno-strict-aliasing -fno-common -fwrapv >>> $QEMU_CFLAGS" >>> QEMU_CFLAGS="-Wall -Wundef -Wwrite-strings -Wmissing-prototypes >>> $QEMU_CFLAGS" >>> QEMU_CFLAGS="-Wstrict-prototypes -Wredundant-decls $QEMU_CFLAGS" >>> QEMU_CFLAGS="-D_GNU_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE >>> $QEMU_CFLAGS" >>> -QEMU_INCLUDES="-I. -I\$(SRC_PATH) -I\$(SRC_PATH)/accel/tcg >>> -I\$(SRC_PATH)/include" >>> +QEMU_INCLUDES="-iquote . -iquote \$(SRC_PATH) -iquote >>> \$(SRC_PATH)/accel/tcg -iquote \$(SRC_PATH)/include" >>> if test "$debug_info" = "yes"; then >>> CFLAGS="-g $CFLAGS" >>> LDFLAGS="-g $LDFLAGS" >>> @@ -6560,19 +6560,19 @@ if test "$vxhs" = "yes" ; then >>> fi >>> >>> if test "$tcg_interpreter" = "yes"; then >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/tci $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/tci $QEMU_INCLUDES" >>> elif test "$ARCH" = "sparc64" ; then >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/sparc $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/sparc $QEMU_INCLUDES" >>> elif test "$ARCH" = "s390x" ; then >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/s390 $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/s390 $QEMU_INCLUDES" >>> elif test "$ARCH" = "x86_64" -o "$ARCH" = "x32" ; then >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/i386 $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/i386 $QEMU_INCLUDES" >>> elif test "$ARCH" = "ppc64" ; then >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/ppc $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/ppc $QEMU_INCLUDES" >>> else >>> - QEMU_INCLUDES="-I\$(SRC_PATH)/tcg/\$(ARCH) $QEMU_INCLUDES" >>> + QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg/\$(ARCH) $QEMU_INCLUDES" >>> fi >>> -QEMU_INCLUDES="-I\$(SRC_PATH)/tcg $QEMU_INCLUDES" >>> +QEMU_INCLUDES="-iquote \$(SRC_PATH)/tcg $QEMU_INCLUDES" >>> >>> echo "TOOLS=$tools" >> $config_host_mak >>> echo "ROMS=$roms" >> $config_host_mak >>> diff --git a/rules.mak b/rules.mak >>> index 6e94333..93a0702 100644 >>> --- a/rules.mak >>> +++ b/rules.mak >>> @@ -29,7 +29,7 @@ QEMU_DGFLAGS += -MMD -MP -MT $@ -MF $(@D)/$(*F).d >>> # dir, one absolute and the other relative to the compiler working >>> # directory. These are the same for target-independent files, but >>> # different for target-dependent ones. >>> -QEMU_LOCAL_INCLUDES = -I$(BUILD_DIR)/$(@D) -I$(@D) >>> +QEMU_LOCAL_INCLUDES = -iquote $(BUILD_DIR)/$(@D) -iquote $(@D) >>> >>> WL_U := -Wl,-u, >>> find-symbols = $(if $1, $(sort $(shell $(NM) -P -g $1 | $2))) >>> diff --git a/Makefile.target b/Makefile.target >>> index 6549481..d0ec77a 100644 >>> --- a/Makefile.target >>> +++ b/Makefile.target >>> @@ -11,9 +11,9 @@ $(call set-vpath, $(SRC_PATH):$(BUILD_DIR)) >>> ifdef CONFIG_LINUX >>> QEMU_CFLAGS += -I../linux-headers >>> endif >>> -QEMU_CFLAGS += -I.. -I$(SRC_PATH)/target/$(TARGET_BASE_ARCH) -DNEED_CPU_H >>> +QEMU_CFLAGS += -iquote .. -iquote $(SRC_PATH)/target/$(TARGET_BASE_ARCH) >>> -DNEED_CPU_H >>> >>> -QEMU_CFLAGS+=-I$(SRC_PATH)/include >>> +QEMU_CFLAGS+=-iquote $(SRC_PATH)/include >>> >>> ifdef CONFIG_USER_ONLY >>> # user emulator name >>> -- >>> MST >>> >> >> Regards, >> Daniel >> -- >> |: https://berrange.com -o- https://www.flickr.com/photos/dberrange >> :| >> |: https://libvirt.org -o- https://fstop138.berrange.com >> :| >> |: https://entangle-photo.org -o- https://www.instagram.com/dberrange >> :|