[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20190208054147.GB9647@archlinux-ryzen>
Date: Thu, 7 Feb 2019 22:41:47 -0700
From: Nathan Chancellor <natechancellor@...il.com>
To: Mathieu Desnoyers <mathieu.desnoyers@...icios.com>
Cc: ndesaulniers@...gle.com,
yamada masahiro <yamada.masahiro@...ionext.com>,
Sami Tolvanen <samitolvanen@...gle.com>,
Andrew Morton <akpm@...ux-foundation.org>,
"Peter Zijlstra (Intel)" <peterz@...radead.org>,
Johannes Weiner <hannes@...xchg.org>,
Kees Cook <keescook@...omium.org>,
Dominik Brodowski <linux@...inikbrodowski.net>,
Nicholas Piggin <npiggin@...il.com>,
Vasily Gorbik <gor@...ux.ibm.com>,
Adrian Reber <adrian@...as.de>,
Richard Guy Briggs <rgb@...hat.com>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/4] init/Kconfig: add config support for detecting linker
On Thu, Feb 07, 2019 at 07:57:20PM -0500, Mathieu Desnoyers wrote:
>
> ----- ndesaulniers@...gle.com wrote:
> > Similar to how we differentiate between CONFIG_CC_IS_GCC and
> > CONFIG_CC_IS_CLANG, add CONFIG_LD_IS_BFD, CONFIG_LD_IS_GOLD, and
> > CONFIG_LD_IS_LLD.
> >
> > This simiplifies patches to Makefiles that need to do different things
> > for different linkers.
>
Hi Mathieu,
> What guarantees that the linker used for e.g. make defconfig is the same linker used for make ?
>
> Is it required with this patch ?
>
The build system ensures that the compiler (and after this patch, the
linker) config values are correct before compiling (and prompting the
user for the new choices that are available to them in that case).
For example:
$ make ARCH=arm64 CC=clang CROSS_COMPILE=aarch64-linux-gnu- defconfig
...
$ rg "CONFIG_CC_IS|CONFIG_LD_IS" .config
10:CONFIG_CC_IS_CLANG=y
12:CONFIG_LD_IS_BFD=y
$ make ARCH=arm64 CROSS_COMPILE=aarch64-linux-gnu- LD=ld.lld
scripts/kconfig/conf --syncconfig Kconfig
*
* Restart config...
*
*
* General architecture-dependent options
*
Kprobes (KPROBES) [N/y/?] n
Optimize very unlikely/likely branches (JUMP_LABEL) [N/y/?] (NEW)
Stack Protector buffer overflow detection (STACKPROTECTOR) [Y/n/?] y
Strong Stack Protector (STACKPROTECTOR_STRONG) [Y/n/?] y
Use a virtually-mapped stack (VMAP_STACK) [Y/n/?] y
Perform full reference count validation at the expense of speed (REFCOUNT_FULL) [Y/?] y
*
* GCC plugins
*
GCC plugins (GCC_PLUGINS) [N/y/?] (NEW)
...
$ rg "CONFIG_CC_IS|CONFIG_LD_IS" .config
9:CONFIG_CC_IS_GCC=y
13:CONFIG_LD_IS_LLD=y
> How does it work in a cross compile environment ?
>
Since $(LD) is defined based on $(CROSS_COMPILE), it should work fine.
I tested it with both arm64 and x86_64 and CONFIG_LD_IS_BFD gets
set by default and I can see CONFIG_LD_IS_GOLD get set with both
LD=aarch64-linux-gnu-ld.gold and LD=ld.gold.
If I misunderstood the questions or my explanation wasn't correct,
please let me know!
Cheers,
Nathan
> Thanks,
>
> Mathieu
>
>
> >
> > Cc: Nathan Chancellor <natechancellor@...il.com>
> > Cc: Sami Tolvanen <samitolvanen@...gle.com>
> > Signed-off-by: Nick Desaulniers <ndesaulniers@...gle.com>
> > ---
> > init/Kconfig | 9 +++++++++
> > 1 file changed, 9 insertions(+)
> >
> > diff --git a/init/Kconfig b/init/Kconfig
> > index c9386a365eea..b6046dcf7794 100644
> > --- a/init/Kconfig
> > +++ b/init/Kconfig
> > @@ -26,6 +26,15 @@ config CLANG_VERSION
> > config CC_HAS_ASM_GOTO
> > def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC))
> >
> > +config LD_IS_BFD
> > + def_bool $(success,$(LD) --version | head -n 1 | grep -q 'GNU ld')
> > +
> > +config LD_IS_GOLD
> > + def_bool $(success,$(LD) --version | head -n 1 | grep -q 'GNU gold')
> > +
> > +config LD_IS_LLD
> > + def_bool $(success,$(LD) --version | head -n 1 | grep -q 'LLD')
> > +
> > config CONSTRUCTORS
> > bool
> > depends on !UML
> > --
> > 2.20.1.791.gb4d0f1c61a-goog
> >
>
> --
> Mathieu Desnoyers
> EfficiOS Inc.
> http://www.efficios.com
Powered by blists - more mailing lists