linux-kernel - Re: [PATCH 4/6] mm: memory: use a folio in do_cow

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]

Message-ID: <9cb900f1-6b49-4160-adc2-cdeae1b00b17@huawei.com>
Date:   Wed, 8 Nov 2023 09:41:32 +0800
From:   Kefeng Wang <wangkefeng.wang@...wei.com>
To:     Sidhartha Kumar <sidhartha.kumar@...cle.com>,
        Andrew Morton <akpm@...ux-foundation.org>
CC:     <linux-kernel@...r.kernel.org>, <linux-mm@...ck.org>,
        Matthew Wilcox <willy@...radead.org>,
        David Hildenbrand <david@...hat.com>
Subject: Re: [PATCH 4/6] mm: memory: use a folio in do_cow_page()



On 2023/11/8 2:27, Sidhartha Kumar wrote:
> On 11/7/23 5:52 AM, Kefeng Wang wrote:
>> Use folio_prealloc() helper and convert to use a folio in
>> do_cow_page(), which save five compound_head() calls.
>>
>> Signed-off-by: Kefeng Wang <wangkefeng.wang@...wei.com>
>> ---
>>   mm/memory.c | 16 ++++++----------
>>   1 file changed, 6 insertions(+), 10 deletions(-)
>>
>> diff --git a/mm/memory.c b/mm/memory.c
>> index 062136d25da3..5ecee3eac29d 100644
>> --- a/mm/memory.c
>> +++ b/mm/memory.c
>> @@ -4648,6 +4648,7 @@ static vm_fault_t do_read_fault(struct vm_fault 
>> *vmf)
>>   static vm_fault_t do_cow_fault(struct vm_fault *vmf)
>>   {
>>       struct vm_area_struct *vma = vmf->vma;
>> +    struct folio *folio;
>>       vm_fault_t ret;
>>       ret = vmf_can_call_fault(vmf);
>> @@ -4656,16 +4657,11 @@ static vm_fault_t do_cow_fault(struct vm_fault 
>> *vmf)
>>       if (ret)
>>           return ret;
>> -    vmf->cow_page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, 
>> vmf->address);
>> -    if (!vmf->cow_page)
>> +    folio = folio_prealloc(vma->vm_mm, vma, vmf->address, false);
>> +    if (!folio)
>>           return VM_FAULT_OOM;
>> -    if (mem_cgroup_charge(page_folio(vmf->cow_page), vma->vm_mm,
>> -                GFP_KERNEL)) {
>> -        put_page(vmf->cow_page);
>> -        return VM_FAULT_OOM;
>> -    }
>> -    folio_throttle_swaprate(page_folio(vmf->cow_page), GFP_KERNEL);
> 
> Hi Kefeng,
> 
> I was wondering why this block of code is no longer needed after using a 
> folio?

folio_prealloc(previous page_copy_prealloc) already with above code.

> 
> Thanks,
> Sidhartha Kumar
> 
>> +    vmf->cow_page = &folio->page;
>>       ret = __do_fault(vmf);
>>       if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE | 
>> VM_FAULT_RETRY)))
>> @@ -4674,7 +4670,7 @@ static vm_fault_t do_cow_fault(struct vm_fault 
>> *vmf)
>>           return ret;
>>       copy_user_highpage(vmf->cow_page, vmf->page, vmf->address, vma);
>> -    __SetPageUptodate(vmf->cow_page);
>> +    __folio_mark_uptodate(folio);
>>       ret |= finish_fault(vmf);
>>       unlock_page(vmf->page);
>> @@ -4683,7 +4679,7 @@ static vm_fault_t do_cow_fault(struct vm_fault 
>> *vmf)
>>           goto uncharge_out;
>>       return ret;
>>   uncharge_out:
>> -    put_page(vmf->cow_page);
>> +    folio_put(folio);
>>       return ret;
>>   }
>