linux-kernel - Re: [PATCH] arm64: hw_breakpoint: Save privilege of access control via ptrace

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 for Android: free password hash cracker in your pocket

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

Message-ID: <9cc6d314-2431-c1b5-3d46-63c0ac80ed4d@loongson.cn>
Date: Thu, 20 Jun 2024 10:05:35 +0800
From: Tiezhu Yang <yangtiezhu@...ngson.cn>
To: Oleg Nesterov <oleg@...hat.com>
Cc: Will Deacon <will@...nel.org>, Mark Rutland <mark.rutland@....com>,
 Catalin Marinas <catalin.marinas@....com>,
 Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>,
 Arnaldo Carvalho de Melo <acme@...nel.org>,
 Namhyung Kim <namhyung@...nel.org>, linux-arm-kernel@...ts.infradead.org,
 linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] arm64: hw_breakpoint: Save privilege of access control
 via ptrace

On 06/19/2024 11:15 PM, Oleg Nesterov wrote:
> I don't understand what this patch does, but...

Thanks for your reply.

ctrl->privilege in decode_ctrl_reg() is never be used later but
it can and should be used in arch_build_bp_info().

arch/arm64/include/asm/hw_breakpoint.h
static inline void decode_ctrl_reg(u32 reg,
				   struct arch_hw_breakpoint_ctrl *ctrl)
{
     ...
	ctrl->privilege	= reg & 0x3;  // it is never be used later but ...
	...
}

arch/arm64/kernel/hw_breakpoint.c
static int arch_build_bp_info(struct perf_event *bp,
			      const struct perf_event_attr *attr,
			      struct arch_hw_breakpoint *hw)
{
	...

	if (arch_check_bp_in_kernelspace(hw))
		hw->ctrl.privilege = AARCH64_BREAKPOINT_EL1;  // ... it can and should 
be used here
	else
		hw->ctrl.privilege = AARCH64_BREAKPOINT_EL0;  // and here.
     ...
}

>
> On 06/18, Tiezhu Yang wrote:
>>
>> --- a/include/uapi/linux/perf_event.h
>> +++ b/include/uapi/linux/perf_event.h
>> @@ -467,6 +467,7 @@ struct perf_event_attr {
>>  		__u32		wakeup_watermark; /* bytes before wakeup   */
>>  	};
>>
>> +	__u8			bp_priv;
>>  	__u32			bp_type;
>
> Is it safe to add the new member in the middle of uapi struct?
> This will break userspace...

Let me put the new member "bp_priv" at the end of uapi struct
perf_event_attr in the next version if you are OK with it.

Thanks,
Tiezhu