On 7/1/2021 12:41 PM, Nick Desaulniers wrote: > On Thu, Jul 1, 2021 at 12:07 PM Nathan Chancellor wrote: >> >> On 7/1/2021 10:40 AM, Nick Desaulniers wrote: >>> On Thu, Jul 1, 2021 at 3:20 AM Arnd Bergmann wrote: >>>> >>>> On Thu, Jul 1, 2021 at 12:38 AM kernel test robot wrote: >>>>> >>>>> tree: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.4.y >>>>> head: 38c92ba3580f0d00e57a55caf8f880aa1a0f2a50 >>>>> commit: 10a0c6746644a4fcaf1ffe2916387e17ba059517 [1401/1774] ARM: s3c: fix fiq for clang IAS >>>>> config: arm-randconfig-r034-20210630 (attached as .config) >>>>> compiler: arm-linux-gnueabi-gcc (GCC) 7.5.0 >>>>> reproduce (this is a W=1 build): >>>>> wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross >>>>> chmod +x ~/bin/make.cross >>>>> # https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git/commit/?id=10a0c6746644a4fcaf1ffe2916387e17ba059517 >>>>> git remote add linux-stable-rc https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git >>>>> git fetch --no-tags linux-stable-rc linux-4.4.y >>>>> git checkout 10a0c6746644a4fcaf1ffe2916387e17ba059517 >>>>> # save the attached .config to linux build tree >>>>> mkdir build_dir >>>>> COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-7.5.0 make.cross O=build_dir ARCH=arm SHELL=/bin/bash drivers/ >>>>> >>>>> If you fix the issue, kindly add following tag as appropriate >>>>> Reported-by: kernel test robot >>>>> >>>>> All errors (new ones prefixed by >>): >>>>> >>>>> drivers/spi/spi-s3c24xx-fiq.S: Assembler messages: >>>>>>> drivers/spi/spi-s3c24xx-fiq.S:52: Error: bad instruction `subsne pc,lr,#4' >>>>> drivers/spi/spi-s3c24xx-fiq.S:78: Error: bad instruction `subsne pc,lr,#4' >>>>> drivers/spi/spi-s3c24xx-fiq.S:102: Error: bad instruction `subsne pc,lr,#4' >>>> >>>> Ah, this must be a kernel that never got the unified assembler syntax fixes >>>> that this requires. It seems they went into linux-5.1 but nothing before that, >>>> though they may have been backported. >>>> >>>> Nick or Nathan, what is the earliest lts kernel series that you would expect >>>> to work with clang IAS? >>> >>> 5.4+ for A32. >>> >>>> Should we just drop these fixes for all 4.x.y versions? >>> >>> Sure, but I'm curious what is missing here that exists further >>> upstream? subsne is the correct syntax, isn't it? >> >> Isn't this unified syntax though? Unified syntax was not used >> universally until commit 75fea300d73a ("ARM: 8723/2: always assume the >> "unified" syntax for assembly code"). I backported that to 4.9 to fix >> this error, it seems like 4.4 needs it as well. I can test that and send >> it along. > > Ah, right, using unified syntax without the assembler directive > produces these errors in GAS. It may be simpler to drop this patch > from stable. This patch was a part of the 4.4.259 release back in March 3rd of this year so we would have to revert it. I do not have any strong opinion but the backport is trivial and all the arch/arm configs on 4.4 still pass with it against GCC 8. The choice is up to you all. Cheers, Nathan