linux-kernel - Re: About commit "io: change inX() to have their own IO barrier overrides"

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] [day] [month] [year] [list]

Date:   Wed, 11 Mar 2020 16:12:58 +0000
From:   John Garry <john.garry@...wei.com>
To:     Arnd Bergmann <arnd@...db.de>
CC:     linux-arch <linux-arch@...r.kernel.org>,
        Catalin Marinas <catalin.marinas@....com>,
        Sinan Kaya <okaya@...nel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Jiaxun Yang <jiaxun.yang@...goat.com>,
        "xuwei (O)" <xuwei5@...ilicon.com>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Will Deacon <will@...nel.org>,
        Linux ARM <linux-arm-kernel@...ts.infradead.org>
Subject: Re: About commit "io: change inX() to have their own IO barrier
 overrides"

On 06/03/2020 16:43, John Garry wrote:
> On 06/03/2020 16:29, Arnd Bergmann wrote:
>>> The idea is good, but it would be nice if we just somehow use a common
>>> asm-generic io.h definition directly in logic_pio.c, like:
>>>
>>> asm-generic io.h:
>>>
>>> #ifndef __raw_inw // name?
>>> #define __raw_inw __raw_inw
>>> static inline u16 __raw_inw(unsigned long addr)
>>> {
>>>          u16 val;
>>>
>>>          __io_pbr();
>>>          val = __le16_to_cpu(__raw_readw(addr));
>>>          __io_par(val);
>>>          return val;
>>> }
>>> #endif
>>>
>>> #include <linux/logic_pio.h>
>>>
>>> #ifndef inw
>>> #define inw __raw_inw
>>> #endif
>> Yes, makes sense. Maybe __arch_inw() then? Not great either, but I think
>> that's better than __raw_inw() because __raw_* would sound like it
>> mirrors __raw_readl() that lacks the barriers and byteswaps.
> 
> Right, I had the same concern. And maybe the "arch" prefix is 
> misleading. Just __inw could be ok, and hopefully not conflict with the 
> arch/arm/mach-* definitions.
> 

I think that it hasn't been mentioned already, but it looks like the 
outX methods also need the same treatment, from a7851aa54c.

thanks,
John