clang fails on linux-next since commit 8bf705d13039

[Lukas Bulwahn]

Hi Dmitry, hi Ingo,

since commit 8bf705d13039 ("locking/atomic/x86: Switch atomic.h to use atomic-instrumented.h")
on linux-next (tested and bisected from tag next-20180316), compiling the
kernel with clang fails with:

In file included from arch/x86/entry/vdso/vdso32/vclock_gettime.c:33:
In file included from arch/x86/entry/vdso/vdso32/../vclock_gettime.c:15:
In file included from ./arch/x86/include/asm/vgtod.h:6:
In file included from ./include/linux/clocksource.h:13:
In file included from ./include/linux/timex.h:56:
In file included from ./include/uapi/linux/timex.h:56:
In file included from ./include/linux/time.h:6:
In file included from ./include/linux/seqlock.h:36:
In file included from ./include/linux/spinlock.h:51:
In file included from ./include/linux/preempt.h:81:
In file included from ./arch/x86/include/asm/preempt.h:7:
In file included from ./include/linux/thread_info.h:38:
In file included from ./arch/x86/include/asm/thread_info.h:53:
In file included from ./arch/x86/include/asm/cpufeature.h:5:
In file included from ./arch/x86/include/asm/processor.h:21:
In file included from ./arch/x86/include/asm/msr.h:67:
In file included from ./arch/x86/include/asm/atomic.h:279:
./include/asm-generic/atomic-instrumented.h:295:10: error: invalid output size for constraint '=a'
return arch_cmpxchg((u64 *)ptr, (u64)old, (u64)new);
^
./arch/x86/include/asm/cmpxchg.h:149:2: note: expanded from macro 'arch_cmpxchg'
__cmpxchg(ptr, old, new, sizeof(*(ptr)))
^
./arch/x86/include/asm/cmpxchg.h:134:2: note: expanded from macro '__cmpxchg'
__raw_cmpxchg((ptr), (old), (new), (size), LOCK_PREFIX)
^
./arch/x86/include/asm/cmpxchg.h:95:17: note: expanded from macro '__raw_cmpxchg'
: "=a" (__ret), "+m" (*__ptr) \
^

(... and some more similar and closely related errors)

Best regards,

Lukas

[Dmitry Vyukov]

Thanks for reporting, Lukas.

+more people who are more aware of the current state of clang for kernel.

Are there are known issues in '=a' constraint handling between gcc and
clang? Is there a recommended way to resolve them?

Also, Lukas what's your version of clang? Potentially there are some
fixes for kernel in the very latest versions of clang.

[Lukas Bulwahn]

Hi Dmitry,

On Mon, 19 Mar 2018, Dmitry Vyukov wrote:
>
> Also, Lukas what's your version of clang? Potentially there are some
> fixes for kernel in the very latest versions of clang.
>

I tested it on clang-5.0 and clang-6.0 in a debian buster distribution.
I just reconfirmed that it also appears on clang-7 (version
1:7~svn327768-1 [1]) in a debian sid distribution.

I also forget to mention that I simply use defconfig.

[1] https://packages.debian.org/sid/clang-7

Lukas

[Nick Desaulniers]

From the trace, looking at the source of __raw_cmpxchg(), it immediately
reminds me of this clang bug [0]

where Clang validates inline asm constraints before running dead code
elimination.

[0] https://bugs.llvm.org/show_bug.cgi?id=33587

--
Thanks,
~Nick Desaulniers

[Matthias Kaehlcke]

El Mon, Mar 19, 2018 at 09:43:25AM +0300 Dmitry Vyukov ha dit:

My impression is that the problem only occurs in code built for
32-bit (like arch/x86/entry/vdso/vdso32/*), where the use of a 64-bit
address with a '=a' constraint is indeed invalid. I think the 'root
cause' is that clang parses unreachable code before it discards it:

static __always_inline unsigned long
cmpxchg_local_size(volatile void *ptr, unsigned long old, unsigned long new,
int size)
{
...
switch (size) {
...
case 8:
BUILD_BUG_ON(sizeof(unsigned long) != 8);
return arch_cmpxchg_local((u64 *)ptr, (u64)old, (u64)new);
}
...
}

For 32-bit builds size is 4 and the code in the 'offending' branch is
unreachable, however clang still parses it.

d135b8b5060e ("arm64: uaccess: suppress spurious clang warning") fixes
a similar issue.

[Dmitry Vyukov]

Thanks!

Do I understand it correctly that this is being fixed in clang?

[Matthias Kaehlcke]

El Mon, Mar 19, 2018 at 06:39:32PM +0100 Dmitry Vyukov ha dit:

Personally I am not aware of any development on that side, however I'm
not an LLVM dev.

[Dmitry Vyukov]

This looks like something that will hit us again and again if we don't
fix this in clang.

[Nick Desaulniers]

On Mon, Mar 19, 2018 at 11:16 AM Dmitry Vyukov <dvy...@google.com> wrote:

This looks like something that will hit us again and again if we don't
fix this in clang.

I agree.  I'll bring it up with some our coworkers who hack on the integrated assembler.

--

Thanks,

~Nick Desaulniers

[sedat...@gmail.com]

When forcing to build with '-O0' instead of default '-O2' I can see this...

 ./include/asm-generic/atomic-instrumented.h:364:3: error: array size is negative
                BUILD_BUG_ON(sizeof(unsigned long) != 8);
                ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
./include/linux/build_bug.h:66:52: note: expanded from macro 'BUILD_BUG_ON'
#define BUILD_BUG_ON(condition) ((void)sizeof(char[1 - 2*!!(condition)]))
                                                   ^~~~~~~~~~~~~~~~~~~

 
- Sedat -

[Dmitry Vyukov]

With clang or gcc?

[Sedat Dilek]

clang version 7 (svn330207) and binutils/ld 2.30.

- Sedat -

[Dmitry Vyukov]

Nick, will this also be fixed as part of asm constraint checking fix?

[Sedat Dilek]

On Mon, Mar 19, 2018 at 6:29 PM, Matthias Kaehlcke <m...@chromium.org> wrote:

[ CC Jan Beulich ]

Hi Jan,

can you look at this issue [1] as you have fixed the percpu issue [2]
with [3] on the Linux-kernel side?

This problem still occurs with Linux v4.17-rc7 and reverting
x86/asm-goto support (clang-7 does not support it).

Before this gets fixed on the clang-side, do you see a possibility to
fix this on the kernel-side?

Clang fails like this as reported above and see [4] as mentioned by
Matthias before.

./arch/x86/include/asm/cpufeature.h:150:2: warning: "Compiler lacks
ASM_GOTO support. Add -D __BPF_TRACING__ to your compiler arguments"
[-W#warnings]
#warning "Compiler lacks ASM_GOTO support. Add -D __BPF_TRACING__ to
your compiler arguments"
^
In file included from arch/x86/entry/vdso/vdso32/vclock_gettime.c:31:

In file included from arch/x86/entry/vdso/vdso32/../vclock_gettime.c:15:
In file included from ./arch/x86/include/asm/vgtod.h:6:
In file included from ./include/linux/clocksource.h:13:
In file included from ./include/linux/timex.h:56:
In file included from ./include/uapi/linux/timex.h:56:
In file included from ./include/linux/time.h:6:
In file included from ./include/linux/seqlock.h:36:
In file included from ./include/linux/spinlock.h:51:
In file included from ./include/linux/preempt.h:81:
In file included from ./arch/x86/include/asm/preempt.h:7:
In file included from ./include/linux/thread_info.h:38:
In file included from ./arch/x86/include/asm/thread_info.h:53:
In file included from ./arch/x86/include/asm/cpufeature.h:5:
In file included from ./arch/x86/include/asm/processor.h:21:
In file included from ./arch/x86/include/asm/msr.h:67:

In file included from ./arch/x86/include/asm/atomic.h:283:
./include/asm-generic/atomic-instrumented.h:365:10: error: invalid

output size for constraint '=a'
return arch_cmpxchg((u64 *)ptr, (u64)old, (u64)new);
^
./arch/x86/include/asm/cmpxchg.h:149:2: note: expanded from macro 'arch_cmpxchg'
__cmpxchg(ptr, old, new, sizeof(*(ptr)))
^
./arch/x86/include/asm/cmpxchg.h:134:2: note: expanded from macro '__cmpxchg'
__raw_cmpxchg((ptr), (old), (new), (size), LOCK_PREFIX)
^
./arch/x86/include/asm/cmpxchg.h:95:17: note: expanded from macro
'__raw_cmpxchg'
: "=a" (__ret), "+m" (*__ptr) \
^

Thanks in advance.

Regards,
- Sedat -

[1] https://bugs.llvm.org/show_bug.cgi?id=33587
[2] https://bugs.llvm.org/show_bug.cgi?id=33587#c18
[3] https://git.kernel.org/linus/22636f8c9511245cb3c8412039f1dd95afb3aa59
[4] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/asm-generic/atomic-instrumented.h#n365

[Jan Beulich]

I don't see the connection between the two problems. The missing suffixes
were a latent problem with future improved assembler behavior. The issue
here is completely different. Short of the clang folks being able to point
out a suitable compiler level workaround, did anyone consider replacing the
expressions with the casts to u64 by invocations of arch_cmpxchg64() /
arch_cmpxchg64_local()? Later code in asm-generic/atomic-instrumented.h
suggests these symbols are required to be defined anyway.

The only other option I see is to break out the __X86_CASE_Q cases into
separate macros (evaluating to nothing or BUILD_BUG_ON(1) for 32-bit).
The definition of __X86_CASE_Q for 32-bit is bogus anyway - the
comment saying "sizeof will never return -1" is meaningless, because for
the comparison with the expression in switch() the constant from the case
label is converted to the type of that expression (i.e. size_t) anyway, i.e.
the value compared against is (size_t)-1, which is a value the compiler
can't prove that it won't be returned by sizeof() (despite that being
extremely unlikely).

Jan

[Sedat Dilek]

Hi Jan,

Thanks for looking at this.
Maybe one of the CCed folks with better Linux/x86 skills can answer you.

Regards,
- Sedat -

[Sedat Dilek]

On Tue, Jun 5, 2018 at 2:20 PM, Raghavan Santhanam
<raghav...@gmail.com> wrote:
> Hello,
>
> I ran into this issue when building Linux kernel v4.17.0 for my Ubuntu
> 17.10(x86_64)
> after having successfully built and brought up previous kernels(from 4.15.7
> - 4.16.13)
> without this "invalid output size for constraint" error since the new
> code(file: atomic-instrumented.h) was not part of the kernel as I saw. My
> reports
> on the successful Linux kernel bring-up on my x86_64 hardware(Lenovo laptop)
> for Ubuntu 17.10 x86_64 with NVIDIA drivers, VirtualBox drivers, VMware
> Player drivers,
> etc - all built using clang, can be found here :
> 1) https://ubuntuforums.org/showthread.php?t=2386423
> 2)
> https://www.quora.com/What-is-the-state-of-the-art-of-compiling-the-Linux-Kernel-using-LLVM-Clang/answer/Raghavan-Santhanam
> 3) https://bugs.llvm.org/show_bug.cgi?id=4068#c27
>
> So, after researching enough on this above error, I found that as
> expected (u32 instead of u64, just for double-checking) type didn't
> result in the error as I noticed "-m32" option in the clang options for
> one of the source code. So, that confirmed the unreachable code
> was indeed being unnecessarily examined by clang - this is without
> the assembler involved - regardless of whether I disabled the integrated
> assembler or not - the error appeared. Also, I verified that the GNU
> assembler was not actually invoked at all. So, that analysis of
> the dead code was within the clang static code checking logic as it
> seemed.
>
> To address this issue, after reading some of the posts including this one,
> I tried different approaches to get around including any clang options
> to ignore dead code selectively with some flag - none helped. Finally, I
> decided
> to update the build logic to selectively include/exclude this 64-bit
> specific
> code as applicable so that clang won't have to examine some dead 64-bit
> code even for 32-bit code generation. That fixed the issue. If clang
> developers
> fix this issue in future, I may give it a try later, but for now my fix
> suffices the
> need and allows me to continue build and use a clang-built Linux kernel,
> this time, it's
> v4.17.0 as I had used the earlier versions on my Ubuntu 17.10 x86_64.
>
> Cheers.
>

TMI, sorry what did you do exactly to fix this?

- sed@ -

[Raghavan Santhanam]

TMI. As I already mentioned, I had to update the build logic. 

To whoever deleting my post : This is NOT a spam. Do NOT just delete blindly if you don't understand

what I explained. Period.

[Sedat Dilek]

On Tue, Jun 5, 2018 at 4:08 PM, Raghavan Santhanam
<raghav...@gmail.com> wrote:
> TMI. As I already mentioned, I had to update the build logic.

[...]

I am interested in a workaround - really.

Again, what did you exactly do?
What do you mean by "update the build logic"?
The Linux build-environment - hacked some Makefiles?
Which files?
Single step - for the dummies, please.

- sed@ -

[Raghavan Santhanam]

There's no hacking here.

It's a build-level fix.

[Raghavan Santhanam]

There's No Hacking here.

It's a Build-level Fix.

[Sedat Dilek]

Hi,

I played a bit more with this.

I was irritated when I saw in my full make lines
-fstack-protector-strong and -fno-stack-protector simultaneously,
especially in x86/vdso.

First, I tried...

CC_STACKPROTECTOR_NONE n -> y
CC_STACKPROTECTOR_STRONG y -> n

This breaks at the same place in the same way (see attached file).

Then I tried to filter-out -fstack-protector-strong in the relevant
code parts with CC_STACKPROTECTOR_STRONG=y...

diff --git a/arch/x86/entry/vdso/Makefile b/arch/x86/entry/vdso/Makefile
index d998a487c9b1..5482e453a8ca 100644
--- a/arch/x86/entry/vdso/Makefile
+++ b/arch/x86/entry/vdso/Makefile
@@ -73,6 +73,7 @@ CFL := $(PROFILING) -mcmodel=small -fPIC -O2
-fasynchronous-unwind-tables -m64 \
$(filter -g%,$(KBUILD_CFLAGS)) $(call cc-option, -fno-stack-protector) \
-fno-omit-frame-pointer -foptimize-sibling-calls \
-DDISABLE_BRANCH_PROFILING -DBUILD_VDSO
+CFL += $(filter-out -fstack-protector-strong,$(KBUILD_CFLAGS))

$(vobjs): KBUILD_CFLAGS := $(filter-out
$(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS)) $(CFL)

@@ -141,6 +142,7 @@ KBUILD_CFLAGS_32 := $(filter-out
-mcmodel=kernel,$(KBUILD_CFLAGS_32))
KBUILD_CFLAGS_32 := $(filter-out -fno-pic,$(KBUILD_CFLAGS_32))
KBUILD_CFLAGS_32 := $(filter-out -mfentry,$(KBUILD_CFLAGS_32))
KBUILD_CFLAGS_32 := $(filter-out $(GCC_PLUGINS_CFLAGS),$(KBUILD_CFLAGS_32))
+KBUILD_CFLAGS_32 := $(filter-out -fstack-protector-strong,$(KBUILD_CFLAGS_32))
KBUILD_CFLAGS_32 += -m32 -msoft-float -mregparm=0 -fpic
KBUILD_CFLAGS_32 += $(call cc-option, -fno-stack-protector)
KBUILD_CFLAGS_32 += $(call cc-option, -foptimize-sibling-calls)

This breaks...

if [ "" = "-pg" ]; then if [ arch/x86/entry/vdso/vgetcpu.o !=
"scripts/mod/empty.o" ]; then ./scripts/recordmcount
"arch/x86/entry/vdso/vgetcpu.o"; fi; fi;
mycompiler -nostdlib -o arch/x86/entry/vdso/vdso64.so.dbg -fPIC
-shared -Wl,--hash-style=both -Wl,--build-id -Wl,-Bsymbolic -m64
-Wl,-soname=linux-vdso.so.1 -Wl,--no-undefined
-Wl,-z,max-page-size=4096 -Wl,-z,common-page-size=4096
-Wl,-T,arch/x86/entry/vdso/vdso.lds arch/x86/entry/vdso/vdso-note.o
arch/x86/entry/vdso/vclock_gettime.o arch/x86/entry/vdso/vgetcpu.o &&
sh ./arch/x86/entry/vdso/checkundef.sh 'nm'
'arch/x86/entry/vdso/vdso64.so.dbg'
/usr/bin/ld: arch/x86/entry/vdso/vclock_gettime.o: relocation
R_X86_64_32S against undefined hidden symbol `vvar_vsyscall_gtod_data'
can not be used when making a shared object
/usr/bin/ld: final link failed: Nonrepresentable section on output
clang: error: linker command failed with exit code 1 (use -v to see invocation)
objcopy -S --localize-hidden arch/x86/entry/vdso/vdso64.so.dbg
arch/x86/entry/vdso/vdso64.so
objcopy: 'arch/x86/entry/vdso/vdso64.so.dbg': No such file
arch/x86/entry/vdso/Makefile:124: recipe for target
'arch/x86/entry/vdso/vdso64.so' failed
make[6]: *** [arch/x86/entry/vdso/vdso64.so] Error 1

Cannot say stack-protector kconfig settings have an impact on this all.

Thanks.

Regards,
- Sedat -

[Sedat Dilek]

[ TO Mark Rutland ]

Hi,

I was able to build a Linux v4.18-rc6 with tip.git#locking/core [1] on
top of it here on Debian/buster AMD64.

The patch of interest is [2]...

df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation

...and some more locking/atomics[/x86] may be interesting.

I had also to apply an asm-goto fix to reduce the number of warnings
when building with clang-7 (version
7.0.0-svn337957-1~exp1+0~20180725200907.1908~1.gbpcccb1b (trunk)).

CONFIG_DRM_AMDGPU=m is BROKEN and a known issue [3].

I had to hack my fakeroot-sysv binary to workaround a fatal build-stop
by commenting the part "nested operation not yet supported" when using
bindeb-pkg make-target.

The kernel does ***not boot*** on bare metal.

More details see [4] and [5] for the clang-side.

- Sedat -

[1] https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=locking/core
[2] https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/commit/?h=locking/core&id=df79ed2c064363cdc7d2d896923c1885d4e30520
[3] https://lists.freedesktop.org/archives/amd-gfx/2018-July/024634.html
[4] https://github.com/ClangBuiltLinux/linux/issues/3
[5] https://bugs.llvm.org/show_bug.cgi?id=33587

[Mark Rutland]

On Sun, Jul 29, 2018 at 08:12:00PM +0200, Sedat Dilek wrote:
> [ TO Mark Rutland ]
>
> Hi,

Hi,

> I was able to build a Linux v4.18-rc6 with tip.git#locking/core [1] on
> top of it here on Debian/buster AMD64.
>
> The patch of interest is [2]...
>
> df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation
>
> ...and some more locking/atomics[/x86] may be interesting.
>
> I had also to apply an asm-goto fix to reduce the number of warnings
> when building with clang-7 (version
> 7.0.0-svn337957-1~exp1+0~20180725200907.1908~1.gbpcccb1b (trunk)).

Just to be clear, clang 7.0.0 has not been released yet, and this is a
trunk build of clang, right?

Do any released versions of clang (e.g. 6.0.1) build a working kernel?

> CONFIG_DRM_AMDGPU=m is BROKEN and a known issue [3].
>
> I had to hack my fakeroot-sysv binary to workaround a fatal build-stop
> by commenting the part "nested operation not yet supported" when using
> bindeb-pkg make-target.

Does upstream build at all with clang, or are you always having to apply
a number of modifications?

Which config are you using?

> The kernel does ***not boot*** on bare metal.

Ok. Does the prior commit boot?

> More details see [4] and [5] for the clang-side.

It's not clear to me how these relate to the patch in question. AFAICT,
those are build-time errors, but you say that the kernel doesn't boot
(which implies it built).

Are [4,5] relevant to this commit, or to the (unrelated) issue [3]?

My patch removes the switch, so this doesn't look like the same issue.

Thanks,
Mark.

>
> - Sedat -
>
> [1] https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/log/?h=locking/core
> [2] https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git/commit/?h=locking/core&id=df79ed2c064363cdc7d2d896923c1885d4e30520
> [3] https://lists.freedesktop.org/archives/amd-gfx/2018-July/024634.html
> [4] https://github.com/ClangBuiltLinux/linux/issues/3
> [5] https://bugs.llvm.org/show_bug.cgi?id=33587

> From 5c3485197eab808768271d72e188ad11b6fcecd4 Mon Sep 17 00:00:00 2001
> From: Sedat Dilek <sedat...@credativ.de>
> Date: Fri, 8 Jun 2018 18:23:26 +0200
> Subject: [PATCH] x86: Warn clang does not support asm-goto
>
> Signed-off-by: Sedat Dilek <sedat...@credativ.de>
> ---
> arch/x86/Makefile | 2 +-
> arch/x86/include/asm/cpufeature.h | 9 +--------
> 2 files changed, 2 insertions(+), 9 deletions(-)
>
> diff --git a/arch/x86/Makefile b/arch/x86/Makefile
> index a08e82856563..6042f6f5a1be 100644
> --- a/arch/x86/Makefile
> +++ b/arch/x86/Makefile
> @@ -181,7 +181,7 @@ ifdef CONFIG_FUNCTION_GRAPH_TRACER
> endif
>
> ifndef CC_HAVE_ASM_GOTO
> - $(error Compiler lacks asm-goto support.)
> + $(warning Compiler lacks asm-goto support.)
> endif
>
> #
> diff --git a/arch/x86/include/asm/cpufeature.h b/arch/x86/include/asm/cpufeature.h
> index aced6c9290d6..79177f0efdf1 100644
> --- a/arch/x86/include/asm/cpufeature.h
> +++ b/arch/x86/include/asm/cpufeature.h
> @@ -140,16 +140,9 @@ extern void clear_cpu_cap(struct cpuinfo_x86 *c, unsigned int bit);
>
> #define setup_force_cpu_bug(bit) setup_force_cpu_cap(bit)
>
> +/* Clang does not support asm-goto (see LLVM bug #9295). */
> #if defined(__clang__) && !defined(CC_HAVE_ASM_GOTO)
>
> -/*
> - * Workaround for the sake of BPF compilation which utilizes kernel
> - * headers, but clang does not support ASM GOTO and fails the build.
> - */
> -#ifndef __BPF_TRACING__
> -#warning "Compiler lacks ASM_GOTO support. Add -D __BPF_TRACING__ to your compiler arguments"
> -#endif
> -
> #define static_cpu_has(bit) boot_cpu_has(bit)
>
> #else
> --
> 2.18.0
>

[Sedat Dilek]

On Mon, Jul 30, 2018 at 10:21 AM, Mark Rutland <mark.r...@arm.com> wrote:
> On Sun, Jul 29, 2018 at 08:12:00PM +0200, Sedat Dilek wrote:
>> [ TO Mark Rutland ]
>>
>> Hi,
>
> Hi,
>
>> I was able to build a Linux v4.18-rc6 with tip.git#locking/core [1] on
>> top of it here on Debian/buster AMD64.
>>
>> The patch of interest is [2]...
>>
>> df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation
>>
>> ...and some more locking/atomics[/x86] may be interesting.
>>
>> I had also to apply an asm-goto fix to reduce the number of warnings
>> when building with clang-7 (version
>> 7.0.0-svn337957-1~exp1+0~20180725200907.1908~1.gbpcccb1b (trunk)).
>
> Just to be clear, clang 7.0.0 has not been released yet, and this is a
> trunk build of clang, right?
>

[ CC Kees and Colin ]

Sorry, for not being precise on my build environment and my
experimenting for followers.

My clang is a snaphot aka clang-7 (svn-trunk: svn337957) from <apt.llvm.org>.
My system is Debian/testing (which will be buster) AMD64.

> Do any released versions of clang (e.g. 6.0.1) build a working kernel?
>

I haven't tried v6.0.1 from Debian/testing.

>> CONFIG_DRM_AMDGPU=m is BROKEN and a known issue [3].
>>
>> I had to hack my fakeroot-sysv binary to workaround a fatal build-stop
>> by commenting the part "nested operation not yet supported" when using
>> bindeb-pkg make-target.
>
> Does upstream build at all with clang, or are you always having to apply
> a number of modifications?
>
> Which config are you using?
>

The config is based on the config from thr package in Debian/experimental [1].

With these modifications...

LINUX_CONFIG='./.config' ; for config in X86_X32 MODULE_SIG
MODULE_SIG_KEY SYSTEM_TRUSTED_KEYS DRM_AMDGPU ; do echo [ $config ] ;
scripts/config --file $LINUX_CONFIG --disable $config ; done

I do not use signed-certs for my modules and amdgpu is known to be
broken as module.

Both configs are attached - the one I used for my experiments and the
one from Debian/experimental.

[1] https://packages.debian.org/experimental/linux-image-4.18.0-rc4-amd64-unsigned

>> The kernel does ***not boot*** on bare metal.
>
> Ok. Does the prior commit boot?
>

I cannot say as I was not able to compile with clang since the commit
8bf705d13039 mentioned here in the subject.

Kees pointed me to issue #7 "__builtin_constant_p() does not work in
deep inline functions" which is the cause for not booting.
The issue is known as #7.

My qemu-log.txt is attached for details if you want to look at.

[1] https://github.com/ClangBuiltLinux/linux/issues/7

>> More details see [4] and [5] for the clang-side.
>
> It's not clear to me how these relate to the patch in question. AFAICT,
> those are build-time errors, but you say that the kernel doesn't boot
> (which implies it built).
>
> Are [4,5] relevant to this commit, or to the (unrelated) issue [3]?
>
> My patch removes the switch, so this doesn't look like the same issue.
>

ClangBuiltLinux issue #3 "clang validates extended assembly
constraints of dead code" is the problem on the clang-side.
Matthias and Jan commented on the thread [1] if you want to read.
You fixed the issue on the kernel-side, so that I could build a Linux
v4.18-rc6 with clang-7 (trunk).
This is a huge progress - really.

[1] https://groups.google.com/forum/#!topic/kasan-dev/oMgCP37n1vw

Is this a bit clearer, now?

Many Thanks,
Sedat

[Mark Rutland]

On Mon, Jul 30, 2018 at 11:09:54AM +0200, Sedat Dilek wrote:
> On Mon, Jul 30, 2018 at 10:21 AM, Mark Rutland <mark.r...@arm.com> wrote:
> > On Sun, Jul 29, 2018 at 08:12:00PM +0200, Sedat Dilek wrote:
> >> I was able to build a Linux v4.18-rc6 with tip.git#locking/core [1] on
> >> top of it here on Debian/buster AMD64.
> >>
> >> The patch of interest is [2]...
> >>
> >> df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation
> >>
> >> ...and some more locking/atomics[/x86] may be interesting.
> >>
> >> I had also to apply an asm-goto fix to reduce the number of warnings
> >> when building with clang-7 (version
> >> 7.0.0-svn337957-1~exp1+0~20180725200907.1908~1.gbpcccb1b (trunk)).

> >> The kernel does ***not boot*** on bare metal.
> >
> > Ok. Does the prior commit boot?
>
> I cannot say as I was not able to compile with clang since the commit
> 8bf705d13039 mentioned here in the subject.

> Kees pointed me to issue #7 "__builtin_constant_p() does not work in
> deep inline functions" which is the cause for not booting.
> The issue is known as #7.
>
> My qemu-log.txt is attached for details if you want to look at.
>
> [1] https://github.com/ClangBuiltLinux/linux/issues/7
>
> >> More details see [4] and [5] for the clang-side.
> >
> > It's not clear to me how these relate to the patch in question. AFAICT,
> > those are build-time errors, but you say that the kernel doesn't boot
> > (which implies it built).
> >
> > Are [4,5] relevant to this commit, or to the (unrelated) issue [3]?
> >
> > My patch removes the switch, so this doesn't look like the same issue.
>
> ClangBuiltLinux issue #3 "clang validates extended assembly
> constraints of dead code" is the problem on the clang-side.
> Matthias and Jan commented on the thread [1] if you want to read.
> You fixed the issue on the kernel-side, so that I could build a Linux
> v4.18-rc6 with clang-7 (trunk).
> This is a huge progress - really.
>
> [1] https://groups.google.com/forum/#!topic/kasan-dev/oMgCP37n1vw
>
> Is this a bit clearer, now?

Yes; I had misunderstood your mail as reporting a regression resulting
from my patch, rather than an improvement.

IIUC, commit df79ed2c0643 ("locking/atomics: Simplify cmpxchg()
instrumentation") happens to make the kernel compile with clang, when it
would not previously (since commit 8bf705d13039).

Given that you seem to understand the remaining issue, I take it that
there is nothing that I need to do here.

Thanks,
Mark.

> $ cat run_qemu.sh
> KPATH=$(pwd)
>
> sudo qemu-system-x86_64 -enable-kvm -M pc -kernel $KPATH/bzImage -initrd $KPATH/initrd.img -m 512 -net none -serial stdio -append "root=/dev/ram0 console=ttyS0 hung_task_panic=1 earlyprintk=ttyS0,115200"
>
> sdi@iniza:~/src/linux-kernel/archives$ ./run_qemu.sh
> Probing EDD (edd=off to disable)... ok
> [ 0.000000] Linux version 4.18.0-rc6-2-iniza-llvmlinux (sedat...@gmail.com@iniza) (clang version 7.0.0-svn338192-1~exp1+0~20180728091252.1913~1.gbpcccb1b (trunk)) #1 SMP 2018-07-29
> [ 0.000000] Command line: root=/dev/ram0 console=ttyS0 hung_task_panic=1 earlyprintk=ttyS0,115200
> [ 0.000000] x86/fpu: x87 FPU will use FXSAVE
> [ 0.000000] BIOS-provided physical RAM map:
> [ 0.000000] BIOS-e820: [mem 0x0000000000000000-0x000000000009fbff] usable
> [ 0.000000] BIOS-e820: [mem 0x000000000009fc00-0x000000000009ffff] reserved
> [ 0.000000] BIOS-e820: [mem 0x00000000000f0000-0x00000000000fffff] reserved
> [ 0.000000] BIOS-e820: [mem 0x0000000000100000-0x000000001ffdffff] usable
> [ 0.000000] BIOS-e820: [mem 0x000000001ffe0000-0x000000001fffffff] reserved
> [ 0.000000] BIOS-e820: [mem 0x00000000feffc000-0x00000000feffffff] reserved
> [ 0.000000] BIOS-e820: [mem 0x00000000fffc0000-0x00000000ffffffff] reserved
> [ 0.000000] bootconsole [earlyser0] enabled
> [ 0.000000] NX (Execute Disable) protection: active
> [ 0.000000] SMBIOS 2.8 present.
> [ 0.000000] DMI: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.1-1 04/01/2014
> [ 0.000000] Hypervisor detected: KVM
> [ 0.000000] last_pfn = 0x1ffe0 max_arch_pfn = 0x400000000
> [ 0.000000] x86/PAT: PAT not supported by CPU.
> [ 0.000000] x86/PAT: Configuration [0-7]: WB WT UC- UC WB WT UC- UC
> Memory KASLR using RDTSC...
> [ 0.000000] found SMP MP-table at [mem 0x000f5d60-0x000f5d6f] mapped at [(____ptrval____)]
> [ 0.000000] RAMDISK: [mem 0x1e555000-0x1ffdffff]
> [ 0.000000] ACPI: Early table checksum verification disabled
> [ 0.000000] ACPI: RSDP 0x00000000000F5B90 000014 (v00 BOCHS )
> [ 0.000000] ACPI: RSDT 0x000000001FFE157C 000030 (v01 BOCHS BXPCRSDT 00000001 BXPC 00000001)
> [ 0.000000] ACPI: FACP 0x000000001FFE1458 000074 (v01 BOCHS BXPCFACP 00000001 BXPC 00000001)
> [ 0.000000] ACPI: DSDT 0x000000001FFE0040 001418 (v01 BOCHS BXPCDSDT 00000001 BXPC 00000001)
> [ 0.000000] ACPI: FACS 0x000000001FFE0000 000040
> [ 0.000000] ACPI: APIC 0x000000001FFE14CC 000078 (v01 BOCHS BXPCAPIC 00000001 BXPC 00000001)
> [ 0.000000] ACPI: HPET 0x000000001FFE1544 000038 (v01 BOCHS BXPCHPET 00000001 BXPC 00000001)
> [ 0.000000] No NUMA configuration found
> [ 0.000000] Faking a node at [mem 0x0000000000000000-0x000000001ffdffff]
> [ 0.000000] NODE_DATA(0) allocated [mem 0x1e550000-0x1e554fff]
> [ 0.000000] kvm-clock: cpu 0, msr 0:1e547001, primary cpu clock
> [ 0.000000] kvm-clock: Using msrs 4b564d01 and 4b564d00
> [ 0.000000] kvm-clock: using sched offset of 614287916 cycles
> [ 0.000000] clocksource: kvm-clock: mask: 0xffffffffffffffff max_cycles: 0x1cd42e4dffb, max_idle_ns: 881590591483 ns
> [ 0.000000] Zone ranges:
> [ 0.000000] DMA [mem 0x0000000000001000-0x0000000000ffffff]
> [ 0.000000] DMA32 [mem 0x0000000001000000-0x000000001ffdffff]
> [ 0.000000] Normal empty
> [ 0.000000] Device empty
> [ 0.000000] Movable zone start for each node
> [ 0.000000] Early memory node ranges
> [ 0.000000] node 0: [mem 0x0000000000001000-0x000000000009efff]
> [ 0.000000] node 0: [mem 0x0000000000100000-0x000000001ffdffff]
> [ 0.000000] Reserved but unavailable: 130 pages
> [ 0.000000] Initmem setup node 0 [mem 0x0000000000001000-0x000000001ffdffff]
> [ 0.000000] ACPI: PM-Timer IO Port: 0x608
> [ 0.000000] ACPI: LAPIC_NMI (acpi_id[0xff] dfl dfl lint[0x1])
> [ 0.000000] IOAPIC[0]: apic_id 0, version 17, address 0xfec00000, GSI 0-23
> [ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 dfl dfl)
> [ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 5 global_irq 5 high level)
> [ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 high level)
> [ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 10 global_irq 10 high level)
> [ 0.000000] ACPI: INT_SRC_OVR (bus 0 bus_irq 11 global_irq 11 high level)
> [ 0.000000] Using ACPI (MADT) for SMP configuration information
> [ 0.000000] ACPI: HPET id: 0x8086a201 base: 0xfed00000
> [ 0.000000] smpboot: Allowing 1 CPUs, 0 hotplug CPUs
> [ 0.000000] PM: Registered nosave memory: [mem 0x00000000-0x00000fff]
> [ 0.000000] PM: Registered nosave memory: [mem 0x0009f000-0x0009ffff]
> [ 0.000000] PM: Registered nosave memory: [mem 0x000a0000-0x000effff]
> [ 0.000000] PM: Registered nosave memory: [mem 0x000f0000-0x000fffff]
> [ 0.000000] [mem 0x20000000-0xfeffbfff] available for PCI devices
> [ 0.000000] Booting paravirtualized kernel on KVM
> [ 0.000000] clocksource: refined-jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645519600211568 ns
> [ 0.000000] random: get_random_bytes called from start_kernel+0x87/0x5f0 with crng_init=0
> [ 0.000000] setup_percpu: NR_CPUS:512 nr_cpumask_bits:512 nr_cpu_ids:1 nr_node_ids:1
> [ 0.000000] percpu: Embedded 44 pages/cpu @(____ptrval____) s141528 r8192 d30504 u2097152
> [ 0.000000] KVM setup async PF for cpu 0
> [ 0.000000] kvm-stealtime: cpu 0, msr 1e216180
> [ 0.000000] Built 1 zonelists, mobility grouping on. Total pages: 128872
> [ 0.000000] Policy zone: DMA32
> [ 0.000000] Kernel command line: root=/dev/ram0 console=ttyS0 hung_task_panic=1 earlyprintk=ttyS0,115200
> [ 0.000000] Memory: 465940K/523768K available (10252K kernel code, 1053K rwdata, 2964K rodata, 1704K init, 648K bss, 57828K reserved, 0K cma-reserved)
> [ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
> [ 0.000000] Kernel/User page tables isolation: enabled
> [ 0.000000] ftrace: allocating 27902 entries in 109 pages
> [ 0.004000] Hierarchical RCU implementation.
> [ 0.004000] RCU restricting CPUs from NR_CPUS=512 to nr_cpu_ids=1.
> [ 0.004000] RCU: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
> [ 0.004000] NR_IRQS: 33024, nr_irqs: 256, preallocated irqs: 16
> [ 0.004000] Console: colour VGA+ 80x25
> [ 0.004000] console [ttyS0] enabled
> [ 0.004000] console [ttyS0] enabled
> [ 0.004000] bootconsole [earlyser0] disabled
> [ 0.004000] bootconsole [earlyser0] disabled
> [ 0.004000] ACPI: Core revision 20180531
> [ 0.004000] clocksource: hpet: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604467 ns
> [ 0.004011] APIC: Switch to symmetric I/O mode setup
> [ 0.004642] x2apic enabled
> [ 0.005068] Switched APIC routing to physical x2apic.
> [ 0.006223] ..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1
> [ 0.006688] tsc: Detected 2904.000 MHz processor
> [ 0.007038] clocksource: tsc-early: mask: 0xffffffffffffffff max_cycles: 0x29dc05e54fc, max_idle_ns: 440795291716 ns
> [ 0.007038] Calibrating delay loop (skipped) preset value.. 5808.00 BogoMIPS (lpj=11616000)
> [ 0.008002] pid_max: default: 32768 minimum: 301
> [ 0.008367] Security Framework initialized
> [ 0.008714] Yama: becoming mindful.
> [ 0.008994] AppArmor: AppArmor initialized
> [ 0.009393] Dentry cache hash table entries: 65536 (order: 7, 524288 bytes)
> [ 0.010388] Inode-cache hash table entries: 32768 (order: 6, 262144 bytes)
> [ 0.011050] Mount-cache hash table entries: 1024 (order: 1, 8192 bytes)
> [ 0.012003] Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes)
> [ 0.012666] CPU: Physical Processor ID: 0
> [ 0.012984] mce: CPU supports 10 MCE banks
> [ 0.013320] Last level iTLB entries: 4KB 0, 2MB 0, 4MB 0
> [ 0.013720] Last level dTLB entries: 4KB 0, 2MB 0, 4MB 0, 1GB 0
> [ 0.014173] Spectre V2 : Mitigation: Full generic retpoline
> [ 0.014592] Speculative Store Bypass: Vulnerable
> [ 0.021689] Freeing SMP alternatives memory: 28K
> [ 0.028000] smpboot: CPU0: Intel QEMU Virtual CPU version 2.5+ (family: 0x6, model: 0x6, stepping: 0x3)
> [ 0.028000] Performance Events: PMU not available due to virtualization, using software events only.
> [ 0.028000] Hierarchical SRCU implementation.
> [ 0.028000] NMI watchdog: Perf event create on CPU 0 failed with -2
> [ 0.028000] NMI watchdog: Perf NMI watchdog permanently disabled
> [ 0.028000] smp: Bringing up secondary CPUs ...
> [ 0.028000] smp: Brought up 1 node, 1 CPU
> [ 0.028003] smpboot: Max logical packages: 1
> [ 0.028331] smpboot: Total of 1 processors activated (5808.00 BogoMIPS)
> [ 0.028937] devtmpfs: initialized
> [ 0.029233] x86/mm: Memory block size: 128MB
> [ 0.029679] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
> [ 0.030419] futex hash table entries: 256 (order: 2, 16384 bytes)
> [ 0.030903] pinctrl core: initialized pinctrl subsystem
> [ 0.031401] NET: Registered protocol family 16
> [ 0.031784] audit: initializing netlink subsys (disabled)
> [ 0.032064] cpuidle: using governor ladder
> [ 0.032383] cpuidle: using governor menu
> [ 0.032695] ACPI: bus type PCI registered
> [ 0.033003] acpiphp: ACPI Hot Plug PCI Controller Driver version: 0.5
> [ 0.033561] PCI: Using configuration type 1 for base access
> [ 0.033990] audit: type=2000 audit(1532896057.253:1): state=initialized audit_enabled=0 res=1
> [ 0.035369] HugeTLB registered 2.00 MiB page size, pre-allocated 0 pages
> [ 0.036115] ACPI: Added _OSI(Module Device)
> [ 0.036493] ACPI: Added _OSI(Processor Device)
> [ 0.036887] ACPI: Added _OSI(3.0 _SCP Extensions)
> [ 0.037304] ACPI: Added _OSI(Processor Aggregator Device)
> [ 0.037789] ACPI: Added _OSI(Linux-Dell-Video)
> [ 0.038607] ACPI: 1 ACPI AML tables successfully acquired and loaded
> [ 0.040114] ACPI: Interpreter enabled
> [ 0.040425] ACPI: (supports S0 S3 S4 S5)
> [ 0.040779] ACPI: Using IOAPIC for interrupt routing
> [ 0.041231] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
> [ 0.042161] ACPI: Enabled 2 GPEs in block 00 to 0F
> [ 0.044126] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
> [ 0.044682] acpi PNP0A03:00: _OSC: OS supports [ASPM ClockPM Segments MSI]
> [ 0.045292] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
> [ 0.045890] acpi PNP0A03:00: fail to add MMCONFIG information, can't access extended PCI configuration space under this bridge.
> [ 0.047061] acpiphp: Slot [3] registered
> [ 0.047423] acpiphp: Slot [4] registered
> [ 0.047793] acpiphp: Slot [5] registered
> [ 0.048022] acpiphp: Slot [6] registered
> [ 0.048395] acpiphp: Slot [7] registered
> [ 0.048802] acpiphp: Slot [8] registered
> [ 0.049173] acpiphp: Slot [9] registered
> [ 0.049543] acpiphp: Slot [10] registered
> [ 0.049927] acpiphp: Slot [11] registered
> [ 0.050304] acpiphp: Slot [12] registered
> [ 0.050680] acpiphp: Slot [13] registered
> [ 0.051056] acpiphp: Slot [14] registered
> [ 0.051432] acpiphp: Slot [15] registered
> [ 0.051809] acpiphp: Slot [16] registered
> [ 0.052021] acpiphp: Slot [17] registered
> [ 0.052397] acpiphp: Slot [18] registered
> [ 0.052773] acpiphp: Slot [19] registered
> [ 0.053149] acpiphp: Slot [20] registered
> [ 0.053530] acpiphp: Slot [21] registered
> [ 0.053906] acpiphp: Slot [22] registered
> [ 0.054280] acpiphp: Slot [23] registered
> [ 0.054655] acpiphp: Slot [24] registered
> [ 0.055030] acpiphp: Slot [25] registered
> [ 0.055405] acpiphp: Slot [26] registered
> [ 0.056022] acpiphp: Slot [27] registered
> [ 0.056397] acpiphp: Slot [28] registered
> [ 0.056773] acpiphp: Slot [29] registered
> [ 0.057149] acpiphp: Slot [30] registered
> [ 0.057524] acpiphp: Slot [31] registered
> [ 0.057882] PCI host bridge to bus 0000:00
> [ 0.058195] pci_bus 0000:00: root bus resource [io 0x0000-0x0cf7 window]
> [ 0.058709] pci_bus 0000:00: root bus resource [io 0x0d00-0xffff window]
> [ 0.059222] pci_bus 0000:00: root bus resource [mem 0x000a0000-0x000bffff window]
> [ 0.059789] pci_bus 0000:00: root bus resource [mem 0x20000000-0xfebfffff window]
> [ 0.060001] pci_bus 0000:00: root bus resource [mem 0x100000000-0x17fffffff window]
> [ 0.060579] pci_bus 0000:00: root bus resource [bus 00-ff]
> [ 0.064018] pci 0000:00:01.1: legacy IDE quirk: reg 0x10: [io 0x01f0-0x01f7]
> [ 0.064624] pci 0000:00:01.1: legacy IDE quirk: reg 0x14: [io 0x03f6]
> [ 0.065215] pci 0000:00:01.1: legacy IDE quirk: reg 0x18: [io 0x0170-0x0177]
> [ 0.065813] pci 0000:00:01.1: legacy IDE quirk: reg 0x1c: [io 0x0376]
> [ 0.066755] pci 0000:00:01.3: quirk: [io 0x0600-0x063f] claimed by PIIX4 ACPI
> [ 0.067596] pci 0000:00:01.3: quirk: [io 0x0700-0x070f] claimed by PIIX4 SMB
> [ 0.074067] ACPI: PCI Interrupt Link [LNKA] (IRQs 5 *10 11)
> [ 0.074720] ACPI: PCI Interrupt Link [LNKB] (IRQs 5 *10 11)
> [ 0.075328] ACPI: PCI Interrupt Link [LNKC] (IRQs 5 10 *11)
> [ 0.076093] ACPI: PCI Interrupt Link [LNKD] (IRQs 5 10 *11)
> [ 0.077044] ACPI: PCI Interrupt Link [LNKS] (IRQs *9)
> [ 0.078314] pci 0000:00:02.0: vgaarb: setting as boot VGA device
> [ 0.079203] pci 0000:00:02.0: vgaarb: VGA device added: decodes=io+mem,owns=io+mem,locks=none
> [ 0.080005] pci 0000:00:02.0: vgaarb: bridge control possible
> [ 0.080848] vgaarb: loaded
> [ 0.081197] pps_core: LinuxPPS API ver. 1 registered
> [ 0.081729] pps_core: Software ver. 5.3.6 - Copyright 2005-2007 Rodolfo Giometti <giom...@linux.it>
> [ 0.082708] PTP clock support registered
> [ 0.083130] EDAC MC: Ver: 3.0.0
> [ 0.083592] PCI: Using ACPI for IRQ routing
> [ 0.084191] HPET: 3 timers in total, 0 timers will be used for per-cpu timer
> [ 0.084948] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0
> [ 0.085473] hpet0: 3 comparators, 64-bit 100.000000 MHz counter
> [ 0.090089] clocksource: Switched to clocksource kvm-clock
> [ 0.099366] VFS: Disk quotas dquot_6.6.0
> [ 0.099817] VFS: Dquot-cache hash table entries: 512 (order 0, 4096 bytes)
> [ 0.100406] AppArmor: AppArmor Filesystem Enabled
> [ 0.100770] pnp: PnP ACPI init
> [ 0.101409] pnp: PnP ACPI: found 6 devices
> [ 0.113805] clocksource: acpi_pm: mask: 0xffffff max_cycles: 0xffffff, max_idle_ns: 2085701024 ns
> [ 0.114557] NET: Registered protocol family 2
> [ 0.114962] tcp_listen_portaddr_hash hash table entries: 256 (order: 0, 4096 bytes)
> [ 0.115694] TCP established hash table entries: 4096 (order: 3, 32768 bytes)
> [ 0.116237] TCP bind hash table entries: 4096 (order: 4, 65536 bytes)
> [ 0.116726] TCP: Hash tables configured (established 4096 bind 4096)
> [ 0.117204] UDP hash table entries: 256 (order: 1, 8192 bytes)
> [ 0.117633] UDP-Lite hash table entries: 256 (order: 1, 8192 bytes)
> [ 0.118107] NET: Registered protocol family 1
> [ 0.118511] pci 0000:00:01.0: PIIX3: Enabling Passive Release
> [ 0.119014] pci 0000:00:00.0: Limiting direct PCI/PCI transfers
> [ 0.119652] pci 0000:00:01.0: Activating ISA DMA hang workarounds
> [ 0.120590] pci 0000:00:02.0: Video device with shadowed ROM at [mem 0x000c0000-0x000dffff]
> [ 0.121720] Unpacking initramfs...
> [ 0.539655] Freeing initrd memory: 27180K
> [ 0.540155] clocksource: tsc: mask: 0xffffffffffffffff max_cycles: 0x29dc05e54fc, max_idle_ns: 440795291716 ns
> [ 0.541190] Initialise system trusted keyrings
> [ 0.541547] workingset: timestamp_bits=40 max_order=17 bucket_order=0
> [ 0.542721] zbud: loaded
> [ 0.543099] pstore: using deflate compression
> [ 0.684799] Key type asymmetric registered
> [ 0.685185] Asymmetric key parser 'x509' registered
> [ 0.685638] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 248)
> [ 0.686287] io scheduler noop registered
> [ 0.686592] io scheduler deadline registered
> [ 0.686933] io scheduler cfq registered (default)
> [ 0.687296] io scheduler mq-deadline registered
> [ 0.687695] shpchp: Standard Hot Plug PCI Controller Driver version: 0.4
> [ 0.688303] Serial: 8250/16550 driver, 4 ports, IRQ sharing enabled
> [ 0.711943] 00:05: ttyS0 at I/O 0x3f8 (irq = 4, base_baud = 115200) is a 16550A
> [ 0.712751] Linux agpgart interface v0.103
> [ 0.713091] AMD IOMMUv2 driver by Joerg Roedel <jro...@suse.de>
> [ 0.713560] AMD IOMMUv2 functionality not available on this system
> [ 0.714144] i8042: PNP: PS/2 Controller [PNP0303:KBD,PNP0f13:MOU] at 0x60,0x64 irq 1,12
> [ 0.715315] serio: i8042 KBD port at 0x60,0x64 irq 1
> [ 0.715744] serio: i8042 AUX port at 0x60,0x64 irq 12
> [ 0.716241] mousedev: PS/2 mouse device common for all mice
> [ 0.716864] input: AT Translated Set 2 keyboard as /devices/platform/i8042/serio0/input/input0
> [ 0.717808] rtc_cmos 00:00: RTC can wake from S4
> [ 0.718941] rtc_cmos 00:00: registered as rtc0
> [ 0.719531] rtc_cmos 00:00: alarms up to one day, y3k, 114 bytes nvram, hpet irqs
> [ 0.720742] ledtrig-cpu: registered to indicate activity on CPUs
> [ 0.721732] NET: Registered protocol family 10
> [ 0.724475] Segment Routing with IPv6
> [ 0.724890] mip6: Mobile IPv6
> [ 0.725166] NET: Registered protocol family 17
> [ 0.725553] mpls_gso: MPLS GSO support
> [ 0.725942] sched_clock: Marking stable (724018763, 0)->(964060525, -240041762)
> [ 0.726626] registered taskstats version 1
> [ 0.726986] Loading compiled-in X.509 certificates
> [ 0.727475] zswap: loaded using pool lzo/zbud
> [ 0.727971] AppArmor: AppArmor sha1 policy hashing enabled
> [ 0.728756] rtc_cmos 00:00: setting system clock to 2018-07-29 20:27:37 UTC (1532896057)
> [ 0.730756] Freeing unused kernel memory: 1704K
> [ 0.736161] Write protecting the kernel read-only data: 16384k
> [ 0.737499] Freeing unused kernel memory: 2024K
> [ 0.739566] Freeing unused kernel memory: 1132K
> [ 0.745418] x86/mm: Checked W+X mappings: passed, no W+X pages found.
> [ 0.746111] x86/mm: Checking user space page tables
> [ 0.752098] x86/mm: Checked W+X mappings: passed, no W+X pages found.
> [ 0.753666] usercopy: Kernel memory exposure attempt detected from SLUB object 'task_struct' (offset 1744, size 8)!
> [ 0.754649] ------------[ cut here ]------------
> [ 0.755067] kernel BUG at mm/usercopy.c:100!
> [ 0.755684] invalid opcode: 0000 [#1] SMP PTI
> [ 0.756148] CPU: 0 PID: 1 Comm: init Not tainted 4.18.0-rc6-2-iniza-llvmlinux #1
> [ 0.756773] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.1-1 04/01/2014
> [ 0.757425] RIP: 0010:usercopy_abort+0x94/0xb0
> [ 0.757913] Code: 77 01 b7 48 0f 44 d8 48 c7 c7 5f 77 01 b7 31 c0 4c 89 ce 4c 89 d1 4d 89 e0 4d 89 d9 41 56 41 57 53 e8 10 44 e5 ff 48 83 c4 18 <0f> 0b 66 2e 0f 1f 84 00 00 00 00 00 eb fe 66 66 2e 0f 1f 84 00 00
> [ 0.759980] RSP: 0018:ffffb747c00d3cb8 EFLAGS: 00010286
> [ 0.760436] RAX: 0000000000000067 RBX: ffffffffb701774f RCX: 5ad13347f669db00
> [ 0.761009] RDX: 0000000000000000 RSI: 0000000000000002 RDI: 0000000000000246
> [ 0.761546] RBP: ffffb747c00d3cd8 R08: 000000000000000a R09: 00000000ffff0a00
> [ 0.762083] R10: 0000000000000152 R11: 0000000000000000 R12: ffffffffb701774e
> [ 0.762620] R13: 00000000fffffff2 R14: 0000000000000008 R15: 00000000000006d0
> [ 0.763199] FS: 00007f4e7c48d580(0000) GS:ffff99469e200000(0000) knlGS:0000000000000000
> [ 0.763912] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 0.764522] CR2: 00007ffc3a4e4fff CR3: 0000000016cf6000 CR4: 00000000000006f0
> [ 0.765522] Call Trace:
> [ 0.765831] __check_heap_object+0xf0/0x120
> [ 0.766431] __check_object_size+0xbe/0x1c0
> [ 0.767057] do_signal+0x4ac/0x600
> [ 0.767470] prepare_exit_to_usermode+0xd2/0x160
> [ 0.767827] syscall_return_slowpath+0x48/0x240
> [ 0.768180] ? __do_page_fault+0x2c6/0x510
> [ 0.768498] ? async_page_fault+0x8/0x30
> [ 0.768801] entry_SYSCALL_64_after_hwframe+0x44/0xa9
> [ 0.769192] RIP: 0033:0x7f4e7c3941c7
> [ 0.769470] Code: c7 1c 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 00 48 8d 05 21 78 0f 00 8b 00 85 c0 75 1b 45 31 d2 b8 3d 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 59 f3 c3 0f 1f 80 00 00 00 00 41 54 55 41 89
> [ 0.770911] RSP: 002b:00007ffc3a4e5088 EFLAGS: 00000246 ORIG_RAX: 000000000000003d
> [ 0.771544] RAX: 000000000000003a RBX: 0000556de5d30260 RCX: 00007f4e7c3941c7
> [ 0.772091] RDX: 0000000000000000 RSI: 00007ffc3a4e509c RDI: 00000000ffffffff
> [ 0.772637] RBP: 0000556de5d30260 R08: 00007f4e7c48d580 R09: fefeff7168636a6c
> [ 0.773198] R10: 0000000000000000 R11: 0000000000000246 R12: 0000556de5d316b0
> [ 0.773806] R13: 00007ffc3a4e509c R14: 00000000ffffffff R15: 0000556de5d31325
> [ 0.774345] Modules linked in:
> [ 0.774604] ---[ end trace aebaaea85d265154 ]---
> [ 0.774961] RIP: 0010:usercopy_abort+0x94/0xb0
> [ 0.775308] Code: 77 01 b7 48 0f 44 d8 48 c7 c7 5f 77 01 b7 31 c0 4c 89 ce 4c 89 d1 4d 89 e0 4d 89 d9 41 56 41 57 53 e8 10 44 e5 ff 48 83 c4 18 <0f> 0b 66 2e 0f 1f 84 00 00 00 00 00 eb fe 66 66 2e 0f 1f 84 00 00
> [ 0.776745] RSP: 0018:ffffb747c00d3cb8 EFLAGS: 00010286
> [ 0.777142] RAX: 0000000000000067 RBX: ffffffffb701774f RCX: 5ad13347f669db00
> [ 0.777711] RDX: 0000000000000000 RSI: 0000000000000002 RDI: 0000000000000246
> [ 0.778252] RBP: ffffb747c00d3cd8 R08: 000000000000000a R09: 00000000ffff0a00
> [ 0.778801] R10: 0000000000000152 R11: 0000000000000000 R12: ffffffffb701774e
> [ 0.779404] R13: 00000000fffffff2 R14: 0000000000000008 R15: 00000000000006d0
> [ 0.780183] FS: 00007f4e7c48d580(0000) GS:ffff99469e200000(0000) knlGS:0000000000000000
> [ 0.780988] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 0.781501] CR2: 00007ffc3a4e4fff CR3: 0000000016cf6000 CR4: 00000000000006f0
> [ 0.782797] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
> [ 0.782797]
> [ 0.784181] Kernel Offset: 0x35200000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
> [ 0.786094] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b
> [ 0.786094] ]---

[Sedat Dilek]

On Mon, Jul 30, 2018 at 10:21 AM, Mark Rutland <mark.r...@arm.com> wrote:

> Do any released versions of clang (e.g. 6.0.1) build a working kernel?
>

I was able to build recent Linux v4.14.y kernels with several "clang-7
trunk" which needs some additional patches.

Matthias offers a patchset [1] with Linux v4.14 vanilla as base.

Some of the patches are in linux-stable-4.14.y and obsolete now.
If you use a higher clang version like 5 or 6, the patches labeled
with "CLANG-3" and CLANG-4" are not needed.
AFAICS the patches labeled with "UPSTREAM" are of interest.

My 4.14.54-1-iniza-llvmlinux kernel boots on bare metal as an example
(dmesg attached).

- Sedat -

[1] https://chromium.googlesource.com/chromiumos/third_party/kernel/+log/sandbox/mka/llvm/v4.14_ext

[Sedat Dilek]

What are your plans to have...

4d2b25f630c7 locking/atomics: Instrument cmpxchg_double*()
f9881cc43b11 locking/atomics: Instrument xchg()

df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation

00d5551cc4ee locking/atomics/x86: Reduce arch_cmpxchg64*() instrumentation

...for example in Linux 4.18 or 4.17.y?

Thanks,
- Sedat -

[Mark Rutland]

On Mon, Jul 30, 2018 at 11:40:49AM +0200, Sedat Dilek wrote:
> What are your plans to have...
>
> 4d2b25f630c7 locking/atomics: Instrument cmpxchg_double*()
> f9881cc43b11 locking/atomics: Instrument xchg()
> df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation
> 00d5551cc4ee locking/atomics/x86: Reduce arch_cmpxchg64*() instrumentation
>
> ...for example in Linux 4.18 or 4.17.y?

I have no plans to have these backported.

Thanks,
Mark.

[Sedat Dilek]

I guess this is 4.19 material?

Not sure, if I will try a "backport" myself or wait for the fix in clang.

Thanks Mark.

Regards,
- Sedat -

[Mark Rutland]

On Mon, Jul 30, 2018 at 02:28:42PM +0200, Sedat Dilek wrote:
> On Mon, Jul 30, 2018 at 11:46 AM, Mark Rutland <mark.r...@arm.com> wrote:
> > On Mon, Jul 30, 2018 at 11:40:49AM +0200, Sedat Dilek wrote:
> >> What are your plans to have...
> >>
> >> 4d2b25f630c7 locking/atomics: Instrument cmpxchg_double*()
> >> f9881cc43b11 locking/atomics: Instrument xchg()
> >> df79ed2c0643 locking/atomics: Simplify cmpxchg() instrumentation
> >> 00d5551cc4ee locking/atomics/x86: Reduce arch_cmpxchg64*() instrumentation
> >>
> >> ...for example in Linux 4.18 or 4.17.y?
> >
> > I have no plans to have these backported.
> >
>
> I guess this is 4.19 material?

Yes. the above 4 commits are all queued for v4.19.

Thanks,
Mark.

[Nick Desaulniers]

If they help us compile with clang, we'll backport to 4.17, 4.14, 4.9,
and 4.4 stable. From
https://github.com/ClangBuiltLinux/linux/issues/3#issuecomment-408839428,
it sounds like that is the case.

--
Thanks,
~Nick Desaulniers

[Sedat Dilek]

I am doing a CONFIG_HARDENED_USERCOPY=n build right now and see
tommorow if it boots on bare metal.

- sed@ -

[Sedat Dilek]

Looks good (beyond some ACPI errors).
My kernel-config and dmesg-log are attached.

- sed@ -

[Sedat Dilek]

On Mon, Jul 30, 2018 at 6:35 PM, Nick Desaulniers

The commit in Linus upstream tree is...

b06ed71a624b locking/atomic, asm-generic: Add asm-generic/atomic-instrumented.h

sdi@iniza:~/src/linux-kernel/linux$ git describe --contains b06ed71a624b
v4.17-rc1~180^2~9

...so backporting makes sense for linux-stable-4.17.y?

- sed@ -

[Sedat Dilek]

> Looks good (beyond some ACPI errors).

Fix pending in <linux-pm.git#acpi>...

"ACPICA: AML Parser: ignore dispatcher error status during table load"

[1] https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=acpi&id=73c2a01c52b657f4a0ead6c95f64c5279efbd000

[Sedat Dilek]

The real fix is...

"ACPICA: AML Parser: ignore control method status in module-level code"
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/commit/?h=linux-next&id=460a53106ac39b74625d1104d9a113e0f3716c18

- sed@ -