linux-kernel - Re: [PATCH 2/3] x86/build: use -std=gnu89 for proper extern inline semantics

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Fri, 01 Jun 2018 10:36:23 -0700
From:   hpa@...or.com
To:     Nick Desaulniers <ndesaulniers@...gle.com>,
        akpm@...ux-foundation.org, ard.biesheuvel@...aro.org,
        aryabinin@...tuozzo.com, akataria@...are.com,
        boris.ostrovsky@...cle.com, brijesh.singh@....com,
        caoj.fnst@...fujitsu.com, gregkh@...uxfoundation.org,
        jan.kiszka@...mens.com, jarkko.sakkinen@...ux.intel.com,
        jgross@...e.com, jpoimboe@...hat.com,
        kirill.shutemov@...ux.intel.com, mingo@...hat.com,
        mjg59@...gle.com, mka@...omium.org, ndesaulniers@...gle.com,
        pombredanne@...b.com, rostedt@...dmis.org, tglx@...utronix.de,
        thomas.lendacky@....com, tweek@...gle.com
CC:     linux-efi@...r.kernel.org, linux-kernel@...r.kernel.org,
        x86@...nel.org, virtualization@...ts.linux-foundation.org
Subject: Re: [PATCH 2/3] x86/build: use -std=gnu89 for proper extern inline semantics

On June 1, 2018 9:55:49 AM PDT, Nick Desaulniers <ndesaulniers@...gle.com> wrote:
>The top level Makefile explicitly sets the C standard used in the
>kernel
>to gnu89. By overriding KBUILD_CFLAGS, the C standard used for this
>subdir is now implicit based on compiler and compiler version. GCC
>changes this implicit default from gnu89 to gnu11 in v5.1.
>
>This implies that depending on compiler version, parts of the kernel
>are
>being linked together from object files that were compiled with
>different C standard compiler flags.
>
>This is problematic for symbols declared as extern inline, as the
>semantics have switched since gnu89. See also:
>http://blahg.josefsipek.net/?p=529
>
>Signed-off-by: Nick Desaulniers <ndesaulniers@...gle.com>
>Tested-by: Sedat Dilek <sedat.dilek@...il.com>
>---
> arch/x86/boot/compressed/Makefile | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
>diff --git a/arch/x86/boot/compressed/Makefile
>b/arch/x86/boot/compressed/Makefile
>index fa42f895fdde..1a04c7e9add1 100644
>--- a/arch/x86/boot/compressed/Makefile
>+++ b/arch/x86/boot/compressed/Makefile
>@@ -26,7 +26,7 @@ KCOV_INSTRUMENT		:= n
>targets := vmlinux vmlinux.bin vmlinux.bin.gz vmlinux.bin.bz2
>vmlinux.bin.lzma \
> 	vmlinux.bin.xz vmlinux.bin.lzo vmlinux.bin.lz4
> 
>-KBUILD_CFLAGS := -m$(BITS) -O2
>+KBUILD_CFLAGS := -m$(BITS) -O2 -std=gnu89
> KBUILD_CFLAGS += -fno-strict-aliasing $(call cc-option, -fPIE, -fPIC)
> KBUILD_CFLAGS += -DDISABLE_BRANCH_PROFILING
> cflags-$(CONFIG_X86_32) := -march=i386

-fgnu-inlines would be a better option.

We could also simply #define inline inline __attribute__((gnu_inline))
-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.