Various objtool warnings, then segfault (-march=piledriver)

[dymn]

These started with 4.14.9, but kernel have successfully built. Here are the first ones:

arch/x86/events/amd/uncore.o: warning: objtool: amd_uncore_cpu_starting()+0x9b: can't find jump dest instruction at .text+0x125
arch/x86/events/core.o: warning: objtool: x86_pmu_hw_config()+0x13: can't find jump dest instruction at .text+0x1495
kernel/fork.o: warning: objtool: account_kernel_stack()+0xd4: sibling call from callable instruction with modified stack frame
kernel/fork.o: warning: objtool: sysctl_max_threads()+0xee: stack state mismatch: cfa1=7+104 cfa2=7+96
arch/x86/entry/vdso/vma.o: warning: objtool: vgetcpu_cpu_init()+0x9e: return with modified stack frame
arch/x86/events/amd/ibs.o: warning: objtool: perf_ibs_init()+0x6a: stack state mismatch: cfa1=7+8 cfa2=7+0
...

While building kernel 4.14.14 and 4.14.15 I got with segfault (CONFIG_MPILEDRIVER=y, AMD FX-8320E):
...
mm/.tmp_memory.o: warning: objtool: print_bad_pte()+0x9c: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: wp_page_copy()+0x3a6: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: __follow_pte_pmd.isra.10()+0x296: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: _vm_normal_page()+0x2a: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: unmap_page_range()+0xf8: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: finish_mkwrite_fault()+0xc7: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: do_wp_page()+0xd2: return with modified stack frame
mm/.tmp_memory.o: warning: objtool: alloc_set_pte()+0x38e: return with modified stack frame
/bin/sh: line 1: 23359 Segfault ./tools/objtool/objtool check "mm/.tmp_memory.o"
make[1]: *** [scripts/Makefile.build:321: mm/memory.o] Error 139
make: *** [Makefile:1028: mm] Error 2

But with "Generic-x86-64" I have successfully built kernel 4.14.15.

[graysky2]

Hmmm... I don't have access to that chip to further debug.

[dymn]

Additional info there:
https://bugs.gentoo.org/642924

[AndyLavr]

https://patchwork.kernel.org/patch/10144853/

[dymn]

This patch did not help.

https://patchwork.kernel.org/patch/10144853/

[AndyLavr]

gcc version?

[dymn]

sys-devel/gcc-6.4.0-r1

[AndyLavr]

gcc 7.2 and 7.2.1 build fine

[ 0.000000] microcode: microcode updated early to revision 0x1b, date = 2017-11-17
[ 0.000000] Linux version 4.15.0-041500rc9-lowlatency (root@GT80-2QE) (gcc version 7.2.1 20180122 (GCC)) #201801220830 SMP PREEMPT Wed Jan 24 10:13:18 EET 2018

[dymn]

gcc-6.4.0 is in stable tree in gentoo.

[graysky2]

Does gentoo offer a 7.2 ebuild?

[dymn]

Yes, in unstable tree. But 6.4 -> 7.2 = rebuild all world. It is overhead.

[jelinekto]

On FX-6300 the same thing happens with 4.15.2 and GCC 7.3.0.

[graysky2]

I can tell you that I am unable to build with the piledriver option (building on an intel chip) but I can build all the other subarchs. I am not knowledgeable enough to know why piledriver is currently failing.

[dymn]

What do we need to do to find out?
Is there a problem with kernel or patch code?

Selecting CONFIG_MBULLDOZER instead of CONFIG_PILEDRIVER solves this bug: https://bugs.gentoo.org/642924#c21

[mfp20]

Same issue here (piledriver, FX-8310). System is Arch. Kernel vanilla 4.15.7 + graysky + UKSM + ck1.

# gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-pc-linux-gnu/7.2.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /build/gcc-multilib/src/gcc/configure --prefix=/usr --libdir=/usr/lib --libexecdir=/usr/lib --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=https://bugs.archlinux.org/ --enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c++ --enable-shared --enable-threads=posix --enable-libmpx --with-system-zlib --with-isl --enable-__cxa_atexit --disable-libunwind-exceptions --enable-clocale=gnu --disable-libstdcxx-pch --disable-libssp --enable-gnu-unique-object --enable-linker-build-id --enable-lto --enable-plugin --enable-install-libiberty --with-linker-hash-style=gnu --enable-gnu-indirect-function --enable-multilib --disable-werror --enable-checking=release --enable-default-pie --enable-default-ssp
Thread model: posix
gcc version 7.2.0 (GCC) 

Many objs build with

warning: objtool: xxxxxx: can't find jump dest instruction at yyyyyyyy
warning: objtool: xxxxxx: return with modified stack frame

Probably some workarounds are here. In particular seems like some CONFIG_x can shut off those warnings. But I didn't try yet; still building the first run.

Edit: it builds and runs. At least in my case it's just a cosmetic issue.

[graysky2]

Edit: it builds and runs. At least in my case it's just a cosmetic issue.

OK, so safe to close as not-a-bug?

[mfp20]

@graysky2
I'm experiencing issues on suspend but it could be anything (ex: one of the other patches, the new AMDGPU DC driver, and so on); I don't know anything about objtool, and I didn't have time to study and investigate further. Can't tell you anything I've already written. Sorry :(

[graysky2]

OK... closing.