linux-next: build failure while building Linus' tree
Hi all,
Building Linus' tree, today's linux-next build (powerpc ppc6xx_defconfig gcc4.9) failed like this:
ERROR: modpost: "__divdi3" [drivers/block/nbd.ko] undefined!
Caused by commit
fad7cd3310db ("nbd: add the check to prevent overflow in __nbd_ioctl()")
The added check_mul_overflow() call is being passed 64 bit values. COMPILER_HAS_GENERIC_BUILTIN_OVERFLOW is not set for this build (see include/linux/overflow.h).
Signature · 2 lines
Cheers, Stephen Rothwell
+ Rasmus
This was introduced in
commit f0907827a8a91 ("compiler.h: enable builtin overflow checkers and add fallback code")
which added division using the / operator, which is problematic when checking for overflows of 64b operands on 32b targets.
We'll probably need helpers from linux/math64.h and some combination of __builtin_choose_expr/__builtin_types_compatible_p.
That will help us fix another compiler bug for older clang releases, too.
https://github.com/ClangBuiltLinux/linux/issues/1438.
Quoted 9 lines · Nick Desaulniers
+ Rasmus
This was introduced in
commitf0907827a8a91("compiler.h: enable builtin overflow checkers and add fallback code")
which added division using the/operator, which is problematic when checking for overflows of 64b operands on 32b targets.
Quoted 5 lines
We'll probably need helpers from
linux/math64.hand some combination of __builtin_choose_expr/__builtin_types_compatible_p.That will help us fix another compiler bug for older clang releases, too.
https://github.com/ClangBuiltLinux/linux/issues/1438.
Ok, I have something hacked up that I think will work:
https://gist.github.com/nickdesaulniers/2479818f4983bbf2d688cebbab435863
This incomplete diff is a little hacked up to reproduce the issue with a known-bad revision of clang that demonstrates a similar issue to GCC 4.9. You can ignore the movement of check_mul_overflow and friends in include/linux/overflow.h.
I think I'm going to break that up into 2 or 3 patches:
1. move is_signed_type from include/linux/overflow.h to perhaps include/linux/typecheck.h.
2. add div64_x64, div_x64, and div_64 to include/linux/math64.h, use them in include/linux/overflow.h to fix GCC 4.9
3. move multiply fallbacks out of COMPILER_HAS_GENERIC_BUILTIN_OVERFLOW for clang < 14.
Signature · 2 lines
Thanks, ~Nick Desaulniers
Quoted 11 lines · Nick Desaulniers
Quoted 9 lines · Nick Desaulniers
+ Rasmus
This was introduced in
commitf0907827a8a91("compiler.h: enable builtin overflow checkers and add fallback code")
which added division using the/operator, which is problematic when checking for overflows of 64b operands on 32b targets.
Quoted 8 lines
Quoted 5 lines
We'll probably need helpers from
linux/math64.hand some combination of __builtin_choose_expr/__builtin_types_compatible_p.That will help us fix another compiler bug for older clang releases, too.
https://github.com/ClangBuiltLinux/linux/issues/1438.Ok, I have something hacked up that I think will work:
https://gist.github.com/nickdesaulniers/2479818f4983bbf2d688cebbab435863
hmm...playing around with adding some static asserts to the above, I don't think it's quite right, specifically:
#define div_64(dividend, divisor) ({ \
typeof(dividend) z;
might declare z with the expected sign, but not necessarily the correct width when the dividend is 32b but the divisor is 64b. Feels a bit like trying to encode the C type promotion rules in a macro...
Quoted 15 lines
This incomplete diff is a little hacked up to reproduce the issue with a known-bad revision of clang that demonstrates a similar issue to GCC 4.9. You can ignore the movement of
check_mul_overflowand friends ininclude/linux/overflow.h.I think I'm going to break that up into 2 or 3 patches:
1. moveis_signed_typefrominclude/linux/overflow.hto perhapsinclude/linux/typecheck.h.
2. adddiv64_x64,div_x64, anddiv_64toinclude/linux/math64.h, use them ininclude/linux/overflow.hto fix GCC 4.9
3. move multiply fallbacks out ofCOMPILER_HAS_GENERIC_BUILTIN_OVERFLOWfor clang < 14.
--
Thanks,
~Nick Desaulniers
Signature · 2 lines
Thanks, ~Nick Desaulniers
Quoted 17 lines · Nick Desaulniers
Quoted 14 lines · Nick Desaulniers
Quoted 12 lines · Nick Desaulniers
+ Rasmus
This was introduced in
commitf0907827a8a91("compiler.h: enable builtin overflow checkers and add fallback code")
which added division using the/operator, which is problematic when checking for overflows of 64b operands on 32b targets.We'll probably need helpers from
linux/math64.hand some combination of __builtin_choose_expr/__builtin_types_compatible_p.
Quoted 7 lines
Quoted 5 lines
Quoted 2 lines
That will help us fix another compiler bug for older clang releases, too.
https://github.com/ClangBuiltLinux/linux/issues/1438.Ok, I have something hacked up that I think will work:
https://gist.github.com/nickdesaulniers/2479818f4983bbf2d688cebbab435863hmm...playing around with adding some static asserts to the above, I
ah! static_assert can't be used for the type agnostic macros, BUILD_BUG_ON needs to be used in its place. Ok, let me add a few and see if that can help instill some confidence here.
Quoted 29 lines
don't think it's quite right, specifically:
#definediv_64(dividend, divisor)({ \
typeof(dividend) z;might declare z with the expected sign, but not necessarily the correct width when the dividend is 32b but the divisor is 64b. Feels a bit like trying to encode the C type promotion rules in a macro...
Quoted 5 lines
This incomplete diff is a little hacked up to reproduce the issue with a known-bad revision of clang that demonstrates a similar issue to GCC 4.9. You can ignore the movement of
check_mul_overflowand friends ininclude/linux/overflow.h.Quoted 10 lines
I think I'm going to break that up into 2 or 3 patches:
1. moveis_signed_typefrominclude/linux/overflow.hto perhapsinclude/linux/typecheck.h.
2. adddiv64_x64,div_x64, anddiv_64toinclude/linux/math64.h, use them ininclude/linux/overflow.hto fix GCC 4.9
3. move multiply fallbacks out ofCOMPILER_HAS_GENERIC_BUILTIN_OVERFLOWfor clang < 14.
--
Thanks,
~Nick Desaulniers--
Thanks,
~Nick Desaulniers
Signature · 2 lines
Thanks, ~Nick Desaulniers
Quoted 23 lines · Nick Desaulniers
Quoted 20 lines · Nick Desaulniers
Quoted 17 lines · Nick Desaulniers
Quoted 14 lines · Nick Desaulniers
+ Rasmus
This was introduced in
commitf0907827a8a91("compiler.h: enable builtin overflow checkers and add fallback code")
which added division using the/operator, which is problematic when checking for overflows of 64b operands on 32b targets.We'll probably need helpers from
linux/math64.hand some combination of __builtin_choose_expr/__builtin_types_compatible_p.That will help us fix another compiler bug for older clang releases, too.
https://github.com/ClangBuiltLinux/linux/issues/1438.
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
Ok, I have something hacked up that I think will work:
https://gist.github.com/nickdesaulniers/2479818f4983bbf2d688cebbab435863hmm...playing around with adding some static asserts to the above, I
ah!
static_assertcan't be used for the type agnostic macros,BUILD_BUG_ONneeds to be used in its place. Ok, let me add a few and see if that can help instill some confidence here.
Ah, I just saw your comment now about raising the minimum required version of gcc to 5.1.
https://github.com/ClangBuiltLinux/linux/issues/1438#issuecomment-916745801
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
This incomplete diff is a little hacked up to reproduce the issue with a known-bad revision of clang that demonstrates a similar issue to GCC 4.9. You can ignore the movement of
check_mul_overflowand friends ininclude/linux/overflow.h.
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
I think I'm going to break that up into 2 or 3 patches:
1. moveis_signed_typefrominclude/linux/overflow.hto perhapsinclude/linux/typecheck.h.
2. adddiv64_x64,div_x64, anddiv_64toinclude/linux/math64.h, use them ininclude/linux/overflow.hto fix GCC 4.9
3. move multiply fallbacks out ofCOMPILER_HAS_GENERIC_BUILTIN_OVERFLOWfor clang < 14.
In that case, I do have __mulodi4() rewritten from compiler-rt to be usable in the kernel.
Signature · 2 lines
Thanks, ~Nick Desaulniers