[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4f97491c-b0dd-406a-9ddb-4f4424571704@arm.com>
Date: Thu, 22 May 2025 13:17:02 +0530
From: Dev Jain <dev.jain@....com>
To: David Hildenbrand <david@...hat.com>, akpm@...ux-foundation.org
Cc: ryan.roberts@....com, willy@...radead.org, linux-mm@...ck.org,
linux-kernel@...r.kernel.org, catalin.marinas@....com, will@...nel.org,
Liam.Howlett@...cle.com, lorenzo.stoakes@...cle.com, vbabka@...e.cz,
jannh@...gle.com, anshuman.khandual@....com, peterx@...hat.com,
joey.gouly@....com, ioworker0@...il.com, baohua@...nel.org,
kevin.brodsky@....com, quic_zhenhuah@...cinc.com,
christophe.leroy@...roup.eu, yangyicong@...ilicon.com,
linux-arm-kernel@...ts.infradead.org, hughd@...gle.com,
yang@...amperecomputing.com, ziy@...dia.com
Subject: Re: [PATCH v3 1/5] mm: Optimize mprotect() by batch-skipping PTEs
On 22/05/25 12:43 pm, David Hildenbrand wrote:
>
>>> ... likely with a better function name,
>>
>>
>> I want to be able to reuse the folio from vm_normal_folio(), and we also
>> need
>>
>> nr_ptes to know how much to skip, so if there is no objection in passing
>> int *nr_ptes,
>>
>> or struct folio **foliop to this new function, then I'll carry on with
>> your suggestion :)
>
> Can you quickly prototype what you have in mind and paste it here?
> Will make it easier :)
if (prot_numa)
func(vma, addr, oldpte, &nr);
struct folio *folio func(vma, addr, oldpte, int *nr)
{
if (pte_protnone(oldpte))
*nr = 1, return NULL;
folio = vm_normal_folio();
if (!folio)
*nr =1, return NULL;
if the other skipping conditions happen, goto skip_batch, return
folio and set *nr = batch
if (sysctl.....) {
skip_batch:
*nr = mprotect_batch();
}
if (folio_use_access_time)....
return folio;
}
Powered by blists - more mailing lists