Re: [PATCH] bug.h: Work around GCC PR82365 in BUG()

From: Vineet Gupta <Vineet.Gupta1@synopsys.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: "linux-arch@vger.kernel.org" <linux-arch@vger.kernel.org>,
	Fenghua Yu <fenghua.yu@intel.com>,
	Jesper Nilsson <jesper.nilsson@axis.com>,
	Tony Luck <tony.luck@intel.com>,
	Kees Cook <keescook@chromium.org>, Arnd Bergmann <arnd@arndb.de>,
	"linux-kbuild@vger.kernel.org" <linux-kbuild@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Christopher Li <sparse@chrisli.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Will Deacon <will.deacon@arm.com>,
	"Steven Rostedt (VMware)" <rostedt@goodmis.org>,
	Ingo Molnar <mingo@kernel.org>, Mikael Starvik <starvik@axis.com>,
	Geert Uytterhoeven <geert@linux-m68k.org>,
	Josh Poimboeuf <jpoimboe@redhat.com>,
	Thomas Gleixner <tglx@linutronix.de>,
	arcml <linux-snps-arc@lists.infradead.org>,
	"David S. Miller" <davem@davemloft.net>
Subject: Re: [PATCH] bug.h: Work around GCC PR82365 in BUG()
Date: Wed, 21 Feb 2018 13:23:52 -0800	[thread overview]
Message-ID: <2d57e8c6-069c-c4b7-0eb2-6abb8fa2181f@synopsys.com> (raw)
In-Reply-To: <e35cd0ea-b1b1-b8a1-6d01-6a17da091407@synopsys.com>

On 02/07/2018 05:20 PM, Vineet Gupta wrote:
>
>> Didn't do ;)
>>
>> Is Arnd's patch good to merge or do we need a fixup?
>>
>>
>> From: Arnd Bergmann <arnd@arndb.de>
>> Subject: bug.h: work around GCC PR82365 in BUG()
>>
>> Looking at functions with large stack frames across all architectures led
>> me discovering that BUG() suffers from the same problem as
>> fortify_panic(), which I've added a workaround for already.  In short,
>> variables that go out of scope by calling a noreturn function or
>> __builtin_unreachable() keep using stack space in functions afterwards.
>>
>> A workaround that was identified is to insert an empty assembler statement
>> just before calling the function that doesn't return.  I'm adding a macro
>> "barrier_before_unreachable()" to document this, and insert calls to that
>> in all instances of BUG() that currently suffer from this problem.
>>
..
..

>> ---
>>
>>   arch/arc/include/asm/bug.h            |    3 ++-
>>   arch/cris/include/arch-v10/arch/bug.h |   11 +++++++++--
>>   arch/ia64/include/asm/bug.h           |    6 +++++-
>>   arch/m68k/include/asm/bug.h           |    3 +++
>>   arch/sparc/include/asm/bug.h          |    6 +++++-
>>   include/asm-generic/bug.h             |    1 +
>>   include/linux/compiler-gcc.h          |   15 ++++++++++++++-
>>   include/linux/compiler.h              |    5 +++++
>>   8 files changed, 44 insertions(+), 6 deletions(-)
>>
>> diff -puN arch/arc/include/asm/bug.h~bugh-work-around-gcc-pr82365-in-bug 
>> arch/arc/include/asm/bug.h
>> --- a/arch/arc/include/asm/bug.h~bugh-work-around-gcc-pr82365-in-bug
>> +++ a/arch/arc/include/asm/bug.h
>> @@ -23,7 +23,8 @@ void die(const char *str, struct pt_regs
>>     #define BUG()    do {                                \
>>       pr_warn("BUG: failure at %s:%d/%s()!\n", __FILE__, __LINE__, __func__); \
>> -    dump_stack();                                \
>> +    barrier_before_unreachable();                        \
>> +    __builtin_trap();                            \
>>   } while (0)
>
> For ARC, it is double win.
>
> 1. Fixes 3 -Wreturn-type warnings
>
> | ../net/core/ethtool.c:311:1: warning: control reaches end of non-void function 
> [-Wreturn-type]
> | ../kernel/sched/core.c:3246:1: warning: control reaches end of non-void 
> function [-Wreturn-type]
> | ../include/linux/sunrpc/svc_xprt.h:180:1: warning: control reaches end of 
> non-void function [-Wreturn-type]
>
> 2. bloat-o-meter reports code size improvements as gcc elides the generated code 
> for stack return.
>
>
> Acked-by: Vineet Gupta <vgupta@synopsys.com>   # for arch/arc
> Tested-by: Vineet Gupta <vgupta@synopsys.com>   # for arch/arc

Ping ?


_______________________________________________
linux-snps-arc mailing list
linux-snps-arc@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-snps-arc