Re: include/linux/build_bug.h:78:41: error: static assertion failed: "struct bpf_fib_lookup size check"

[Arnd Bergmann]

On Wed, Apr 3, 2024, at 10:10, Anton Protopopov wrote:
> On Wed, Apr 3, 2024 at 10:03 AM Naresh Kamboju
> <naresh.kamboju@xxxxxxxxxx> wrote:
>>
>> The arm footbridge_defconfig failed with gcc-13 and gcc-8 on Linux next
>> starting from next-20240328..next-20240402.
>>
>> arm:
>>   build:
>>     * gcc-8-footbridge_defconfig - Failed
>>     * gcc-13-footbridge_defconfig - Failed
>>
>> Reported-by: Linux Kernel Functional Testing <lkft@xxxxxxxxxx>
>>
>> In file included from include/linux/bitfield.h:10,
>>                  from arch/arm/include/asm/ptrace.h:13,
>>                  from arch/arm/include/asm/processor.h:14,
>>                  from include/linux/prefetch.h:15,
>>                  from arch/arm/include/asm/atomic.h:12,
>>                  from include/linux/atomic.h:7,
>>                  from net/core/filter.c:20:
>> include/linux/build_bug.h:78:41: error: static assertion failed:
>> "struct bpf_fib_lookup size check"
>>    78 | #define __static_assert(expr, msg, ...) _Static_assert(expr, msg)
>>       |                                         ^~~~~~~~~~~~~~
>
> Thanks, I will take a look today

The problem is CONFIG_AEABI=n, which changes the alignment
of sub-word struct members. I had assumed that AEABI is enabled
by default for everything already, but it looks like footbridge
and a couple of other defconfigs still have it turned off:

$ git grep -l  CONFIG_ARCH_MULTI_V7.is.not arch/arm/configs/* | xargs git grep -L AEABI
arch/arm/configs/assabet_defconfig
arch/arm/configs/collie_defconfig
arch/arm/configs/footbridge_defconfig
arch/arm/configs/h3600_defconfig
arch/arm/configs/jornada720_defconfig
arch/arm/configs/neponset_defconfig
arch/arm/configs/netwinder_defconfig
arch/arm/configs/rpc_defconfig
arch/arm/configs/spear3xx_defconfig
arch/arm/configs/spear6xx_defconfig
arch/arm/configs/spitz_defconfig

Russell still has machines with an OABI toolchain, but I'm not
aware of anyone else relying on it. It does cause other
problems as well, so I already turned it off a long time ago
for my randconfig testing.

We should probably make it the default for everything, except
whichever defconfig Russell uses:

--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -1159,7 +1159,7 @@ config ARM_PATCH_IDIV
 config AEABI
        bool "Use the ARM EABI to compile the kernel" if !CPU_V7 && \
                !CPU_V7M && !CPU_V6 && !CPU_V6K && !CC_IS_CLANG && !COMPILE_TEST
-       default CPU_V7 || CPU_V7M || CPU_V6 || CPU_V6K || CC_IS_CLANG || COMPILE_TEST
+       default y
        help
          This option allows for the kernel to be compiled using the latest
          ARM ABI (aka EABI).  This is only useful if you are using a user

Or we could go one step further and make it 'depends on
EXPERT', short of removing it entirely.

     Arnd