On 8/12/22 3:08 PM, Jens Axboe wrote: > On 8/12/22 3:01 PM, Jens Axboe wrote: >> On 8/12/22 2:44 PM, Jens Axboe wrote: >>> On Aug 12, 2022, at 2:28 PM, Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: >>>> >>>> ?On Fri, Aug 12, 2022 at 5:46 AM Jens Axboe <axboe@xxxxxxxxx> wrote: >>>>> >>>>> - Small series improving type safety of the sqe fields (Stefan) >>>> >>>> This doesn't work AT ALL. >>>> >>>> A basic allmodconfig build fails with tons of errors. It starts with >>>> >>>> In function ?io_kiocb_cmd_sz_check?, >>>> inlined from ?io_prep_rw? at io_uring/rw.c:38:21: >>>> ././include/linux/compiler_types.h:354:45: error: call to >>>> ?__compiletime_assert_802? declared with attribute error: BUILD_BUG_ON >>>> failed: cmd_sz > sizeof(struct io_cmd_data) >>>> 354 | _compiletime_assert(condition, msg, >>>> __compiletime_assert_, __COUNTER__) >>>> | ^ >>>> ././include/linux/compiler_types.h:335:25: note: in definition of >>>> macro ?__compiletime_assert? >>>> 335 | prefix ## suffix(); >>>> \ >>>> | ^~~~~~ >>>> ././include/linux/compiler_types.h:354:9: note: in expansion of >>>> macro ?_compiletime_assert? >>>> 354 | _compiletime_assert(condition, msg, >>>> __compiletime_assert_, __COUNTER__) >>>> | ^~~~~~~~~~~~~~~~~~~ >>>> ./include/linux/build_bug.h:39:37: note: in expansion of macro >>>> ?compiletime_assert? >>>> 39 | #define BUILD_BUG_ON_MSG(cond, msg) compiletime_assert(!(cond), msg) >>>> | ^~~~~~~~~~~~~~~~~~ >>>> ./include/linux/build_bug.h:50:9: note: in expansion of macro >>>> ?BUILD_BUG_ON_MSG? >>>> 50 | BUILD_BUG_ON_MSG(condition, "BUILD_BUG_ON failed: " >>>> #condition) >>>> | ^~~~~~~~~~~~~~~~ >>>> ./include/linux/io_uring_types.h:496:9: note: in expansion of macro >>>> ?BUILD_BUG_ON? >>>> 496 | BUILD_BUG_ON(cmd_sz > sizeof(struct io_cmd_data)); >>>> | ^~~~~~~~~~~~ >>>> >>>> and goes downhill from there. >>>> >>>> I don't think this can have seen any testing at all. >>> >>> Wtf? I always run allmodconfig before sending and it also ran testing. >>> I?ll check shortly. Sorry about that, whatever went wrong here. >> >> My test box is still on the same sha from this morning, which is: >> >> commit 2ae08b36c06ea8df73a79f6b80ff7964e006e9e3 (origin/master, origin/HEAD) >> Merge: 21f9c8a13bb2 8bb5e7f4dcd9 >> Author: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> >> Date: Thu Aug 11 09:23:08 2022 -0700 >> >> Merge tag 'input-for-v5.20-rc0' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input >> >> with io_uring-6.0 (ff34d8d06a1f16b6a58fb41bfbaa475cc6c02497) and >> block-6.0 (aa0c680c3aa96a5f9f160d90dd95402ad578e2b0) pulled in, and it >> builds just fine for me: >> >> axboe@r7525 ~/gi/build (test)> make clean 9.827s >> HOSTCC scripts/basic/fixdep >> HOSTCC scripts/kconfig/conf.o >> HOSTCC scripts/kconfig/confdata.o >> HOSTCC scripts/kconfig/expr.o >> LEX scripts/kconfig/lexer.lex.c >> YACC scripts/kconfig/parser.tab.[ch] >> HOSTCC scripts/kconfig/lexer.lex.o >> HOSTCC scripts/kconfig/menu.o >> HOSTCC scripts/kconfig/parser.tab.o >> HOSTCC scripts/kconfig/preprocess.o >> HOSTCC scripts/kconfig/symbol.o >> HOSTCC scripts/kconfig/util.o >> HOSTLD scripts/kconfig/conf >> # >> # No change to .config >> # >> axboe@r7525 ~/gi/build (test)> time make -j256 -s >> >> ________________________________________________________ >> Executed in 172.67 secs fish external >> usr time 516.61 mins 396.00 micros 516.61 mins >> sys time 44.40 mins 0.00 micros 44.40 mins >> >> using: >> >> axboe@r7525 ~/gi/build (test)> gcc --version >> gcc (Debian 12.1.0-7) 12.1.0 >> >> Puzzled, I'll keep poking... > > I re-did an allmodconfig, and also built on arm64, and I have to say I'm > puzzled with what you are seeing. Updated to latest master as well, > nothing. Furthermore, I have the build bot send me successful build > notifications as well, not just the errors, and here's what it reported > 12h ago: > > tree/branch: https://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-block.git io_uring-6.0 > branch HEAD: ff34d8d06a1f16b6a58fb41bfbaa475cc6c02497 io_uring: add missing BUILD_BUG_ON() checks for new io_uring_sqe fields > > elapsed time: 882m > > configs tested: 53 > configs skipped: 2 > > The following configs have been built successfully. > More configs may be tested in the coming days. > > gcc tested configs: > um x86_64_defconfig > um i386_defconfig > i386 defconfig > i386 randconfig-a014 > i386 randconfig-a012 > i386 randconfig-a016 > x86_64 randconfig-a013 > x86_64 randconfig-a011 > arc randconfig-r043-20220811 > arm defconfig > x86_64 randconfig-a015 > m68k allmodconfig > arc allyesconfig > i386 allyesconfig > i386 randconfig-a001 > x86_64 rhel-8.3-func > i386 randconfig-a003 > x86_64 defconfig > alpha allyesconfig > mips allyesconfig > arm allyesconfig > x86_64 rhel-8.3-kunit > m68k allyesconfig > powerpc allnoconfig > arm64 allyesconfig > i386 randconfig-a005 > x86_64 rhel-8.3 > sh allmodconfig > x86_64 rhel-8.3-kvm > x86_64 allyesconfig > x86_64 rhel-8.3-kselftests > x86_64 rhel-8.3-syz > x86_64 randconfig-a002 > x86_64 randconfig-a004 > x86_64 randconfig-a006 > ia64 allmodconfig > powerpc allmodconfig > > clang tested configs: > i386 randconfig-a011 > i386 randconfig-a013 > hexagon randconfig-r045-20220811 > hexagon randconfig-r041-20220811 > x86_64 randconfig-a012 > s390 randconfig-r044-20220811 > i386 randconfig-a015 > riscv randconfig-r042-20220811 > x86_64 randconfig-a014 > x86_64 randconfig-a016 > i386 randconfig-a002 > i386 randconfig-a006 > i386 randconfig-a004 > x86_64 randconfig-a001 > x86_64 randconfig-a003 > x86_64 randconfig-a005 > > building that very same sha I sent you. WTF? I assure you this thing has > been both built, and not just by me, and runtime tested as per usual. > Why it fails on your end, I really have no good clue right now. Keith brought up a good point, is this some weird randomization of io_kiocb that makes it bigger? struct io_rw is already at 64-bytes as it is, if it gets re-arranged for more padding maybe that's what you're hitting? Is it just io_rw or are you seeing others? -- Jens Axboe
