[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20190619121848.GE7767@fuggles.cambridge.arm.com>
Date: Wed, 19 Jun 2019 13:18:48 +0100
From: Will Deacon <will.deacon@....com>
To: jinho lim <jordan.lim@...sung.com>
Cc: mark.rutland@....com, ebiederm@...ssion.com, marc.zyngier@....com,
anshuman.khandual@....com, andreyknvl@...gle.com,
linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
seroto7@...il.com
Subject: Re: [PATCH] arm64: remove useless condition
On Wed, Jun 19, 2019 at 08:38:57PM +0900, jinho lim wrote:
> dump_instr function checks user_mode internally.
>
> Signed-off-by: jinho lim <jordan.lim@...sung.com>
> ---
> arch/arm64/kernel/traps.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
> index ccc13b45d9b1..694e78b950ca 100644
> --- a/arch/arm64/kernel/traps.c
> +++ b/arch/arm64/kernel/traps.c
> @@ -182,8 +182,7 @@ static int __die(const char *str, int err, struct pt_regs *regs)
> print_modules();
> show_regs(regs);
>
> - if (!user_mode(regs))
> - dump_instr(KERN_EMERG, regs);
> + dump_instr(KERN_EMERG, regs);
I don't think this is right, because it means we'll now dump the current
user instruction on the die() path.
Instead, we should probably rename dump_instr to dump_kernel_instr(),
and have it return immediately if user_mode(regs). We can also kill
__dump_instr completely.
Will
Powered by blists - more mailing lists