linux-kernel - Re: [PATCH v2 5/5] arm64: irqchip/gic-v3: Select priorities at boot time

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

Date: Fri, 21 Jun 2024 16:14:34 +0800
From: "Liao, Chang" <liaochang1@...wei.com>
To: Marc Zyngier <maz@...nel.org>
CC: Mark Rutland <mark.rutland@....com>,
	<linux-arm-kernel@...ts.infradead.org>, <alexandru.elisei@....com>,
	<catalin.marinas@....com>, <linux-kernel@...r.kernel.org>,
	<tglx@...utronix.de>, <will@...nel.org>
Subject: Re: [PATCH v2 5/5] arm64: irqchip/gic-v3: Select priorities at boot
 time



在 2024/6/21 15:48, Marc Zyngier 写道:
> On Fri, 21 Jun 2024 07:23:54 +0100,
> "Liao, Chang" <liaochang1@...wei.com> wrote:
>>
>>
>>
>> 在 2024/6/17 19:18, Mark Rutland 写道:
>>>  	cpus_have_group0 = gic_has_group0();
>>
>>> +#define __gicv3_prio_to_ns(p)	(0xff & ((p) << 1))
>>> +#define __gicv3_ns_to_prio(ns)	(0x80 | ((ns) >> 1))
>>
>> What about refactoring the gic_has_group0() using the mapping macros
>> between PMR priority and GIC priority like this:
>>
>> ---------------%<-----------------
>> --- a/drivers/irqchip/irq-gic-v3.c
>> +++ b/drivers/irqchip/irq-gic-v3.c
>> @@ -882,6 +882,7 @@ static bool gic_has_group0(void)
>>  {
>>         u32 val;
>>         u32 old_pmr;
>> +       u32 prio = BIT(8 - gic_get_pribits());
>>
>>         old_pmr = gic_read_pmr();
>>
>> @@ -896,12 +897,12 @@ static bool gic_has_group0(void)
>>          * becomes 0x80. Reading it back returns 0, indicating that
>>          * we're don't have access to Group0.
>>          */
>> -       gic_write_pmr(BIT(8 - gic_get_pribits()));
>> +       gic_write_pmr(prio);
>>         val = gic_read_pmr();
>>
>>         gic_write_pmr(old_pmr);
>>
>> -       return val != 0;
>> +       return val != (__gicv3_prio_to_ns(__gicv3_ns_to_prio(prio)));
>>  }
>> --------------->%-----------------
> 
> No, that's terrible, and makes it simply impossible to understand what
> is happening without looking at 3 layers of indirection.
> 
> Read the comment, and realise that the code implements exactly that.

You are right, it is BIT(8-gic_get_pribits()) decides gic_read_pmr() must return zero
if SCR_EL3.FIQ is set, there is no need to overcomplicate things.

> 
> 	M.
> 

-- 
BR
Liao, Chang