lkml.org 
[lkml]   [2006]   [Mar]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: PATCH][1/8] 2.6.15 mlock: make_pages_wired/unwired
Stone Wang wrote:
> 1. Add make_pages_unwired routine.

Unfortunately you forgot wire_page and unwire_page, so this patch will
not even compile.

> 2. Replace make_pages_present with make_pages_wired, support rollback.

What does support rollback mean?

> 3. Pass 1 more param ("wire") to get_user_pages.
>

As others have pointed out, wire may be a BSD / other unix thing, but
it does not feature in Linux memory management terminology. If you
want to introduce it, you need to do a better job of specifying it.

> Signed-off-by: Shaoping Wang <pwstone@gmail.com>
>

> +void make_pages_unwired(struct mm_struct *mm,
> + unsigned long start,unsigned long end)
> +{
> + struct vm_area_struct *vma;
> + struct page *page;
> + unsigned int foll_flags;
> +
> + foll_flags =0;
> +
> + vma=find_vma(mm,start);
> + if(!vma)
> + BUG();
> + if(is_vm_hugetlb_page(vma))
> + return;
> +
> + for(; start<end ; start+=PAGE_SIZE) {
> + page=follow_page(vma,start,foll_flags);
> + if(page)
> + unwire_page(page);
> + }
> +}
> +

What happens when start goes past vma->vm_end?

> int get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
> - unsigned long start, int len, int write, int force,
> + unsigned long start, int len, int write,int force, int wire,
> struct page **pages, struct vm_area_struct **vmas)
> {
> int i;
> @@ -973,6 +995,7 @@
> if (!vma && in_gate_area(tsk, start)) {
> unsigned long pg = start & PAGE_MASK;
> struct vm_area_struct *gate_vma = get_gate_vma(tsk);
> + struct page *page;
> pgd_t *pgd;
> pud_t *pud;
> pmd_t *pmd;
> @@ -994,6 +1017,7 @@
> pte_unmap(pte);
> return i ? : -EFAULT;
> }
> + page = vm_normal_page(gate_vma, start, *pte);

You wire gate_vma pages? But it doesn't look like you can unwire them with
make_pages_unwired.

> if (pages) {
> struct page *page = vm_normal_page(gate_vma, start, *pte);

This can go now?

> pages[i] = page;
> @@ -1003,9 +1027,12 @@
> pte_unmap(pte);
> if (vmas)
> vmas[i] = gate_vma;
> + if(wire)
> + wire_page(page);
> i++;
> start += PAGE_SIZE;
> len--;
> +
> continue;
> }
>
> @@ -1013,6 +1040,7 @@
> || !(vm_flags & vma->vm_flags))
> return i ? : -EFAULT;
>
> + /* We dont account wired HugeTLB pages */

You don't account wired HugeTLB pages? If you can wire them you should be able
to unwire them as well shouldn't you?

--
SUSE Labs, Novell Inc.

Send instant messages to your online friends http://au.messenger.yahoo.com

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2006-03-21 14:09    [W:0.205 / U:0.756 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site