All of lore.kernel.org
 help / color / mirror / Atom feed
From: Matthias Kaehlcke <mka@chromium.org>
To: Mark Rutland <mark.rutland@arm.com>
Cc: Sami Tolvanen <samitolvanen@google.com>,
	linux-kernel@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	Nick Desaulniers <ndesaulniers@google.com>,
	Kees Cook <keescook@chromium.org>,
	Greg Hackmann <ghackmann@google.com>
Subject: Re: [PATCH 00/15] Add support for clang LTO
Date: Fri, 3 Nov 2017 11:07:39 -0700	[thread overview]
Message-ID: <20171103180739.GG96615@google.com> (raw)
In-Reply-To: <20171103175156.f44o7kuxgzym3pbt@lakrids.cambridge.arm.com>

Hi Mark,

El Fri, Nov 03, 2017 at 05:51:56PM +0000 Mark Rutland ha dit:

> On Fri, Nov 03, 2017 at 10:11:44AM -0700, Sami Tolvanen wrote:
> > This series adds build system support for compiling the kernel with clang
> > Link Time Optimization (LTO), using GNU gold with the LLVMgold plug-in
> > for linking. Some background for clang's LTO support is available here:
> > 
> >   https://llvm.org/docs/LinkTimeOptimization.html
> > 
> > With -flto, clang produces LLVM bitcode instead of object files, and
> > the compilation to native code happens at link time. In addition, clang
> > cannot use an external assembler for inline assembly when LTO is enabled,
> > which causes further compatibility issues.
> > 
> > The patches in this series remove intermediate linking steps when LTO is
> > used, postpone processing done on object files until after the LTO link
> > step, add workarounds for GNU gold incompatibilities, and address inline
> > assembly incompatibilities for arm64.
> > 
> > These changes allow arm64 defconfig to be compiled with LTO, but other
> > architectures are not enabled until compatibility issues have been
> > addressed. In particular, x86 inline assembly doesn't currently compile
> > with clang's integrated assembler due to this LLVM bug:
> > 
> >   https://bugs.llvm.org/show_bug.cgi?id=24487
> > 
> > Due to recent bug fixes in the toolchain, it's recommended to use clang
> > 5.0 or later, and GNU gold from binutils 2.27 or later
> 
> Which tree are you testing with?

In general upstream should be a good starting point.

> I had a go building the arm64 for/next/core branch defconfig using clang
> 5.0.0 from llvm.org, with the Linaro 17.05 toolchain's binutils
> (2.27.0.20161019), and I hit what appears to be a compiler bug quite
> quickly.
> 
> I had to create an aarch64-linux-gnu-clang wrapper, too. I'm not sure if
> there's build system help to avoid needing that?
> 
> ----
> [mark@lakrids:~/src/linux]% uselinaro 17.05 usellvm 5.0.0 make CROSS_COMPILE=aarch64-linux-gnu- CC=clang  ARCH=arm64 
> Segmentation fault
> arch/arm64/Makefile:27: ld does not support --fix-cortex-a53-843419; kernel may be susceptible to erratum
> arch/arm64/Makefile:48: Detected assembler with broken .inst;
> disassembly will be unreliable

This should be fixed by https://patchwork.kernel.org/patch/10039329/
and https://patchwork.kernel.org/patch/10030583/

Unfortunately we didn't encounter these earlier, in Chrome OS this
was/is hidden by invoking an architecture aware wrapper instead of
bare clang :(

>   CHK     include/config/kernel.release
>   CHK     include/generated/uapi/linux/version.h
>   CHK     include/generated/utsrelease.h
>   CHK     include/generated/bounds.h
>   CHK     include/generated/timeconst.h
>   CHK     include/generated/asm-offsets.h
>   CALL    scripts/checksyscalls.sh
>   CHK     scripts/mod/devicetable-offsets.h
>   CHK     include/generated/compile.h
>   CC      arch/arm64/crypto/aes-ce-cipher.o
> fatal error: error in backend: Do not know how to split the result of this operator!
> 
> clang-5.0: error: clang frontend command failed with exit code 70 (use -v to see invocation)
> clang version 5.0.0 (tags/RELEASE_500/final)
> Target: aarch64--linux-gnu
> Thread model: posix
> InstalledDir: /mnt/data/opt/toolchain/llvm/5.0.0/bin
> clang-5.0: note: diagnostic msg: PLEASE submit a bug report to http://llvm.org/bugs/ and include the crash backtrace, preprocessed source, and associated run script.
> clang-5.0: note: diagnostic msg: 
> ********************
> 
> PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
> Preprocessed source(s) and associated run script(s) are located at:
> clang-5.0: note: diagnostic msg: /tmp/aes-ce-cipher-7850ed.c
> clang-5.0: note: diagnostic msg: /tmp/aes-ce-cipher-7850ed.sh
> clang-5.0: note: diagnostic msg: 
> 
> ********************
> scripts/Makefile.build:313: recipe for target 'arch/arm64/crypto/aes-ce-cipher.o' failed
> make[1]: *** [arch/arm64/crypto/aes-ce-cipher.o] Error 70
> Makefile:1019: recipe for target 'arch/arm64/crypto' failed
> make: *** [arch/arm64/crypto] Error 2

This is fixed in newer clang versions, otherwise you need

[PATCH 07/15] arm64: use -mno-implicit-float instead of
-mgeneral-regs-only

of this series.

WARNING: multiple messages have this Message-ID (diff)
From: mka@chromium.org (Matthias Kaehlcke)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 00/15] Add support for clang LTO
Date: Fri, 3 Nov 2017 11:07:39 -0700	[thread overview]
Message-ID: <20171103180739.GG96615@google.com> (raw)
In-Reply-To: <20171103175156.f44o7kuxgzym3pbt@lakrids.cambridge.arm.com>

Hi Mark,

El Fri, Nov 03, 2017 at 05:51:56PM +0000 Mark Rutland ha dit:

> On Fri, Nov 03, 2017 at 10:11:44AM -0700, Sami Tolvanen wrote:
> > This series adds build system support for compiling the kernel with clang
> > Link Time Optimization (LTO), using GNU gold with the LLVMgold plug-in
> > for linking. Some background for clang's LTO support is available here:
> > 
> >   https://llvm.org/docs/LinkTimeOptimization.html
> > 
> > With -flto, clang produces LLVM bitcode instead of object files, and
> > the compilation to native code happens at link time. In addition, clang
> > cannot use an external assembler for inline assembly when LTO is enabled,
> > which causes further compatibility issues.
> > 
> > The patches in this series remove intermediate linking steps when LTO is
> > used, postpone processing done on object files until after the LTO link
> > step, add workarounds for GNU gold incompatibilities, and address inline
> > assembly incompatibilities for arm64.
> > 
> > These changes allow arm64 defconfig to be compiled with LTO, but other
> > architectures are not enabled until compatibility issues have been
> > addressed. In particular, x86 inline assembly doesn't currently compile
> > with clang's integrated assembler due to this LLVM bug:
> > 
> >   https://bugs.llvm.org/show_bug.cgi?id=24487
> > 
> > Due to recent bug fixes in the toolchain, it's recommended to use clang
> > 5.0 or later, and GNU gold from binutils 2.27 or later
> 
> Which tree are you testing with?

In general upstream should be a good starting point.

> I had a go building the arm64 for/next/core branch defconfig using clang
> 5.0.0 from llvm.org, with the Linaro 17.05 toolchain's binutils
> (2.27.0.20161019), and I hit what appears to be a compiler bug quite
> quickly.
> 
> I had to create an aarch64-linux-gnu-clang wrapper, too. I'm not sure if
> there's build system help to avoid needing that?
> 
> ----
> [mark at lakrids:~/src/linux]% uselinaro 17.05 usellvm 5.0.0 make CROSS_COMPILE=aarch64-linux-gnu- CC=clang  ARCH=arm64 
> Segmentation fault
> arch/arm64/Makefile:27: ld does not support --fix-cortex-a53-843419; kernel may be susceptible to erratum
> arch/arm64/Makefile:48: Detected assembler with broken .inst;
> disassembly will be unreliable

This should be fixed by https://patchwork.kernel.org/patch/10039329/
and https://patchwork.kernel.org/patch/10030583/

Unfortunately we didn't encounter these earlier, in Chrome OS this
was/is hidden by invoking an architecture aware wrapper instead of
bare clang :(

>   CHK     include/config/kernel.release
>   CHK     include/generated/uapi/linux/version.h
>   CHK     include/generated/utsrelease.h
>   CHK     include/generated/bounds.h
>   CHK     include/generated/timeconst.h
>   CHK     include/generated/asm-offsets.h
>   CALL    scripts/checksyscalls.sh
>   CHK     scripts/mod/devicetable-offsets.h
>   CHK     include/generated/compile.h
>   CC      arch/arm64/crypto/aes-ce-cipher.o
> fatal error: error in backend: Do not know how to split the result of this operator!
> 
> clang-5.0: error: clang frontend command failed with exit code 70 (use -v to see invocation)
> clang version 5.0.0 (tags/RELEASE_500/final)
> Target: aarch64--linux-gnu
> Thread model: posix
> InstalledDir: /mnt/data/opt/toolchain/llvm/5.0.0/bin
> clang-5.0: note: diagnostic msg: PLEASE submit a bug report to http://llvm.org/bugs/ and include the crash backtrace, preprocessed source, and associated run script.
> clang-5.0: note: diagnostic msg: 
> ********************
> 
> PLEASE ATTACH THE FOLLOWING FILES TO THE BUG REPORT:
> Preprocessed source(s) and associated run script(s) are located at:
> clang-5.0: note: diagnostic msg: /tmp/aes-ce-cipher-7850ed.c
> clang-5.0: note: diagnostic msg: /tmp/aes-ce-cipher-7850ed.sh
> clang-5.0: note: diagnostic msg: 
> 
> ********************
> scripts/Makefile.build:313: recipe for target 'arch/arm64/crypto/aes-ce-cipher.o' failed
> make[1]: *** [arch/arm64/crypto/aes-ce-cipher.o] Error 70
> Makefile:1019: recipe for target 'arch/arm64/crypto' failed
> make: *** [arch/arm64/crypto] Error 2

This is fixed in newer clang versions, otherwise you need

[PATCH 07/15] arm64: use -mno-implicit-float instead of
-mgeneral-regs-only

of this series.

  parent reply	other threads:[~2017-11-03 18:07 UTC|newest]

Thread overview: 126+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-11-03 17:11 [PATCH 00/15] Add support for clang LTO Sami Tolvanen
2017-11-03 17:11 ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 01/15] kbuild: add ld-name macro and support for GNU gold Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 18:24   ` Nick Desaulniers
2017-11-03 18:24     ` Nick Desaulniers
2017-11-03 17:11 ` [PATCH 02/15] kbuild: fix LD_DEAD_CODE_DATA_ELIMINATION with " Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 03/15] kbuild: add support for clang LTO Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 04/15] kbuild: fix dynamic ftrace with " Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 05/15] scripts/mod: disable LTO for empty.c Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 06/15] efi/libstub: disable clang LTO Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 20:10   ` Ard Biesheuvel
2017-11-03 20:10     ` Ard Biesheuvel
2017-11-03 17:11 ` [PATCH 06/15] efi/libstub: disable LTO Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 07/15] arm64: use -mno-implicit-float instead of -mgeneral-regs-only Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:50   ` Nick Desaulniers
2017-11-03 17:50     ` Nick Desaulniers
2017-11-03 18:02   ` Mark Rutland
2017-11-03 18:02     ` Mark Rutland
2017-11-03 18:20     ` Nick Desaulniers
2017-11-03 18:20       ` Nick Desaulniers
2017-11-03 18:31   ` Mark Rutland
2017-11-03 18:31     ` Mark Rutland
2017-11-03 18:52     ` Mark Rutland
2017-11-03 18:52       ` Mark Rutland
2017-11-03 19:06       ` Kees Cook
2017-11-03 19:06         ` Kees Cook
2017-11-03 20:18         ` Sami Tolvanen
2017-11-03 20:18           ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 08/15] arm64: don't pass -maarch64linux to GNU gold Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 20:41   ` Yury Norov
2017-11-03 20:41     ` Yury Norov
2017-11-06 16:56     ` Sami Tolvanen
2017-11-06 16:56       ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 09/15] arm64: keep .altinstructions and .altinstr_replacement Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 18:40   ` Nick Desaulniers
2017-11-03 18:40     ` Nick Desaulniers
2017-11-03 17:11 ` [PATCH 10/15] arm64: disable ARM64_ERRATUM_843419 for clang LTO Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 20:13   ` Ard Biesheuvel
2017-11-03 20:13     ` Ard Biesheuvel
2017-11-09 18:54     ` Sami Tolvanen
2017-11-09 18:54       ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 11/15] arm64: explicitly pass --no-fix-cortex-a53-843419 to GNU gold Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 11/15] arm64: fix mrs_s/msr_s macros for clang LTO Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 12/15] arm64: add a workaround for GNU gold with ARM64_MODULE_PLTS Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:11 ` [PATCH 12/15] arm64: explicitly pass --no-fix-cortex-a53-843419 to GNU gold Sami Tolvanen
2017-11-03 17:11   ` Sami Tolvanen
2017-11-03 17:12 ` [PATCH 13/15] arm64: add a workaround for GNU gold with ARM64_MODULE_PLTS Sami Tolvanen
2017-11-03 17:12   ` Sami Tolvanen
2017-11-03 20:09   ` Ard Biesheuvel
2017-11-03 20:09     ` Ard Biesheuvel
2017-11-03 20:29     ` Sami Tolvanen
2017-11-03 20:29       ` Sami Tolvanen
2017-11-03 17:12 ` [PATCH 13/15] arm64: fix mrs_s/msr_s macros for clang LTO Sami Tolvanen
2017-11-03 17:12   ` Sami Tolvanen
2017-11-03 17:53   ` Nick Desaulniers
2017-11-03 17:53     ` Nick Desaulniers
2017-11-03 18:06     ` Mark Rutland
2017-11-03 18:06       ` Mark Rutland
2017-11-03 18:15       ` Nick Desaulniers
2017-11-03 18:15         ` Nick Desaulniers
2017-11-03 18:38   ` Mark Rutland
2017-11-03 18:38     ` Mark Rutland
2017-11-08  9:25   ` Yury Norov
2017-11-08  9:25     ` Yury Norov
2017-11-09  0:02     ` Andi Kleen
2017-11-09  0:02       ` Andi Kleen
2017-11-09  4:48       ` Maxim Kuvyrkov
2017-11-09  4:48         ` Maxim Kuvyrkov
2017-11-09 16:51         ` Sami Tolvanen
2017-11-09 16:51           ` Sami Tolvanen
2017-11-09 16:50       ` Sami Tolvanen
2017-11-09 16:50         ` Sami Tolvanen
2017-11-09 16:45     ` Sami Tolvanen
2017-11-09 16:45       ` Sami Tolvanen
2017-11-03 17:12 ` [PATCH 14/15] arm64: crypto: disable LTO for aes-ce-cipher.c Sami Tolvanen
2017-11-03 17:12   ` Sami Tolvanen
2017-11-03 20:10   ` Ard Biesheuvel
2017-11-03 20:10     ` Ard Biesheuvel
2017-11-03 17:12 ` [PATCH 15/15] arm64: select ARCH_SUPPORTS_CLANG_LTO Sami Tolvanen
2017-11-03 17:12   ` Sami Tolvanen
2017-11-03 17:51 ` [PATCH 00/15] Add support for clang LTO Mark Rutland
2017-11-03 17:51   ` Mark Rutland
2017-11-03 18:07   ` Nick Desaulniers
2017-11-03 18:07     ` Nick Desaulniers
2017-11-03 18:29     ` Mark Rutland
2017-11-03 18:29       ` Mark Rutland
2017-11-03 18:36       ` Nick Desaulniers
2017-11-03 18:36         ` Nick Desaulniers
2017-11-03 18:49         ` Mark Rutland
2017-11-03 18:49           ` Mark Rutland
2017-11-03 18:07   ` Matthias Kaehlcke [this message]
2017-11-03 18:07     ` Matthias Kaehlcke
2017-11-03 18:09   ` Mark Rutland
2017-11-03 18:09     ` Mark Rutland
2017-11-03 18:11     ` Nick Desaulniers
2017-11-03 18:11       ` Nick Desaulniers
2017-11-03 19:26       ` Mark Rutland
2017-11-03 19:26         ` Mark Rutland
2017-11-03 19:56         ` Sami Tolvanen
2017-11-03 19:56           ` Sami Tolvanen
2017-11-04  2:13           ` Mark Rutland
2017-11-04  2:13             ` Mark Rutland
2017-11-09 10:54             ` AKASHI Takahiro
2017-11-09 10:54               ` AKASHI Takahiro
2017-11-09 11:03               ` Marc Zyngier
2017-11-09 11:03                 ` Marc Zyngier
2017-11-09 16:31               ` Sami Tolvanen
2017-11-09 16:31                 ` Sami Tolvanen
2017-11-03 20:23         ` Ard Biesheuvel
2017-11-03 20:23           ` Ard Biesheuvel
2017-11-03 18:22   ` Sami Tolvanen
2017-11-03 18:22     ` Sami Tolvanen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20171103180739.GG96615@google.com \
    --to=mka@chromium.org \
    --cc=ghackmann@google.com \
    --cc=keescook@chromium.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=ndesaulniers@google.com \
    --cc=samitolvanen@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.