[Pkg-xen-devel] Bug#853710: xen: ftbfs with GCC-7

[Ian Jackson]

Matthias Klose of the Debian Project reports that the Xen package in
Debian stretch, which is based on a snapshot of xen.git#stable-4.8,
does not build with GCC-7.

Matthias Klose writes ("Bug#853710: xen: ftbfs with GCC-7"):
> The package fails to build in a test rebuild on at least amd64 with
> gcc-7/g++-7, but succeeds to build with gcc-6/g++-6. The
> severity of this report may be raised before the buster release.
> There is no need to fix this issue in time for the stretch release.
> 
> The full build log can be found at:
> http://people.debian.org/~doko/logs/gcc7-20170126/xen_4.8.1~pre.2017.01.23-1_unstable_gcc7.log
> The last lines of the build log are at the end of this report.
...
> gcc -m64 -DBUILD_ID -fno-strict-aliasing -std=gnu99 -Wall -Wstrict-prototypes -Wdeclaration-after-statement -Wno-unused-but-set-variable -Wno-unused-local-typedefs   -O2 -fomit-frame-pointer -nostdinc -fno-builtin -fno-common -Werror -Wredundant-decls -Wno-pointer-arith -pipe -g -D__XEN__ -include /<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/xen/config.h '-D__OBJECT_FILE__="compat.o"' -Wa,--strip-local-absolute -MMD -MF ./.compat.o.d -I/<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include -I/<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/asm-x86/mach-generic -I/<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/asm-x86/mach-default '-D__OBJECT_LABEL__=arch$x86$efi$compat.o' -msoft-float -fno-stack-protector -fno-exceptions -Wnested-externs -DHAVE_GAS_VMX -DHAVE_GAS_SSE4_2 -DHAVE_GAS_EPT -DHAVE_GAS_FSGSBASE -U__OBJECT_LABEL__ -DHAVE_GAS_QUOTED_SYM '-D__OBJECT_LABEL__=arch/x86/efi/compat.o' -mno-red-zone -mno-sse -fpic -fno-asynchronous-unwind-tables -DGCC_HAS_VISIBILITY_ATTRIBUTE -fshort-wchar -c compat.c -o compat.o
> In file included from /<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/asm/uaccess.h:12:0,
>                  from /<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/asm/guest_access.h:10,
>                  from /<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/xen/guest_access.h:10,
>                  from compat.c:1:
> runtime.c: In function 'efi_compat_get_info':
> /<<PKGBUILDDIR>>/debian/build/build-hypervisor_amd64_amd64/xen/include/asm/x86_64/uaccess.h:56:37: error: '*' in boolean context, suggest '&&' instead [-Werror=int-in-bool-context]
>       compat_access_ok(addr, (count) * (size)))

I looked at the code and the macros at the root of the message are
these:

  #define compat_array_access_ok(addr,count,size) \
      (likely((count) < (~0U / (size))) && \
       compat_access_ok(addr, (count) * (size)))

  #define __compat_access_ok(d, addr, size) \
      __compat_addr_ok(d, (unsigned long)(addr) + ((size) ? (size) - 1 : 0))

This expands to, effectively (and stripping extra parents etc.)

      addr + (count*size ? count*size - 1 : 0)

GCC-7 is complaining about count*size (ie, the result of a
multiplication) being used as a boolean (ie, for the ?: conditional).

I think the compiler is simply wrong to complain about this.  Such
idioms must be extremely common.

Ian.