linux-kernel - Re: [PATCH v6 1/4] MIPS: Do not flush tlb page when updating PTE entry

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 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:   Wed, 27 May 2020 09:06:22 +0800
From:   maobibo <maobibo@...ngson.cn>
To:     Andrew Morton <akpm@...ux-foundation.org>
Cc:     Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
        Jiaxun Yang <jiaxun.yang@...goat.com>,
        Huacai Chen <chenhc@...ote.com>,
        Paul Burton <paulburton@...nel.org>,
        Dmitry Korotin <dkorotin@...ecomp.com>,
        Philippe Mathieu-Daudé <f4bug@...at.org>,
        Stafford Horne <shorne@...il.com>,
        Steven Price <steven.price@....com>,
        Anshuman Khandual <anshuman.khandual@....com>,
        linux-mips@...r.kernel.org, linux-kernel@...r.kernel.org,
        Mike Rapoport <rppt@...ux.ibm.com>,
        Sergei Shtylyov <sergei.shtylyov@...entembedded.com>,
        "Maciej W. Rozycki" <macro@....com>, linux-mm@...ck.org,
        David Hildenbrand <david@...hat.com>
Subject: Re: [PATCH v6 1/4] MIPS: Do not flush tlb page when updating PTE
 entry



On 05/26/2020 05:42 AM, Andrew Morton wrote:
> On Mon, 25 May 2020 10:52:37 +0800 Bibo Mao <maobibo@...ngson.cn> wrote:
> 
>> It is not necessary to flush tlb page on all CPUs if suitable PTE
>> entry exists already during page fault handling, just updating
>> TLB is fine.
>>
>> Here redefine flush_tlb_fix_spurious_fault as empty on MIPS system.
>>
>> ...
>>
>> --- a/arch/mips/include/asm/pgtable.h
>> +++ b/arch/mips/include/asm/pgtable.h
>> @@ -478,6 +478,8 @@ static inline pgprot_t pgprot_writecombine(pgprot_t _prot)
>>  	return __pgprot(prot);
>>  }
>>  
>> +#define flush_tlb_fix_spurious_fault(vma, address) do { } while (0)
>> +
> 
> static inline C would be preferred, if that works.  For a number of reasons:
> 
> - looks nicer
> 
> - more likely to get a code comment (for some reason)
> 
> - adds typechecking.  So right now a MIPS developer could do
> 
> 	struct wibble a;
> 	struct wobble b;
> 
> 	flush_tlb_fix_spurious_fault(&a, &b);
> 
>   and there would be no compiler warning.  Then the code gets merged
>   upstream and in come the embarrassing emails!
> 
> - avoids unused-var warnings
> 
> 	foo()
> 	{
> 		struct address_space a;
> 		struct vm_area_struct v;
> 
> 		flush_tlb_fix_spurious_fault(&v, &a);
> 	}
> 
> will generate unused-variable warnings if
> flush_tlb_fix_spurious_fault() is a macro.  Making
> flush_tlb_fix_spurious_fault() inlined C prevents this.
> 
Sure, I will modify this and send another version.