sambuc/codezero
1
0
Fork 0
mirror of https://github.com/drasko/codezero.git synced 2026-01-12 02:43:15 +01:00
Code Issues Actions Packages Projects Releases Wiki Activity

Few undecided changes on munmap.c

Browse Source
This commit is contained in:
Bahadir Balban
2008-10-29 14:02:21 +02:00
parent d5b5560bfa
commit 3421a29693
1 changed files with 118 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

145
tasks/mm0/src/munmap.c
View File

@@ -204,38 +204,129 @@ struct vm_area *vma_split(struct vm_area *vma, struct tcb *task,
return new;
}
void page_unmap_all_virtuals(struct page *p)
{
}
struct vmo_object *page_to_vm_object(struct page *p)
{
struct list_head *current_page;
/* Try to reach head of page list in vm object */
while (current_page->offset < p->offset) {
current_page = list_entry(p->list.prev, struct page, list);
p = current_page;
}
}
/* TODO:
* Unmapping virtual address ranges:
* Implement l4_unmap_range() which would be more efficient than unmapping per-address.
*
* Finding virtual mappings of a page:
* Find a way to obtain all virtual addresses of a physical page.
* Linux does it by reaching virtual addresses from the page itself,
* we may rather access virtual addresses from vm_object to vma to page virtual address.
*
* OR:
*
* We could do this by keeping a list of vma mapping offsets in every object. Then
* we can find the individual page offsets from those object offsets.
*
* Finding unused pages:
* We need to traverse all vma's that map the vm object to see which pages are unused.
* We could then swap out file-backed pages, and discard private and anonymous pages.
*/
/* Takes difference of 2 ranges as in the set theory meaning. */
void set_diff_of_ranges(unsigned long *minstart, unsigned long *minend, unsigned long substart, unsigned long subend)
{
if (*minstart < subend && *minstart > substart)
*minstart = subend;
if (start1 >= start2 && start1 < end) {
}
}
int vmo_unmap_pages(struct tcb *task, struct vm_area *vma, unsigned long start, unsigned long end)
{
struct vm_obj_link *vmo_link, *n, *link;
/* Go to each vm object from the vma object chain */
list_for_each_entry_safe(vmo_link, n, &vma->vm_obj_list, list) {
/* Go to each link for every object */
list_for_each_entry(link, &vmo_link->obj->link_list, linkref) {
/* Go to each vma that owns the link */
struct vm_area *vma = link->vma;
unsigned long unused_start, unused_end;
/* Try every */
unused_start = max(start, vma->file_offset);
unused_end = min(end, vma->file_offset +
(vma->pfn_end - vma->pfn_start));
}
}
}
int vm_obj_unmap_pages(struct vm_object *vmo, unsigned long start, unsigned long end)
{
}
int vma_unmap_objects(struct tcb *task, struct vm_area *vma,
unsigned long page_start, unsigned long page_end)
{
struct vm_obj_link *vmo_link, *n;
list_for_each_entry_safe(vmo_link, n, &vma->vm_obj_list, list) {
vm_obj_unmap_pages(vmo_link);
}
}
/* This shrinks the vma from *one* end only, either start or end */
int vma_shrink(struct vm_area *vma, struct tcb *task, unsigned long pfn_start,
unsigned long pfn_end)
int vma_shrink(struct vm_area *vma, struct tcb *task,
unsigned long pfn_start, unsigned long pfn_end)
{
unsigned long diff;
BUG_ON(pfn_start >= pfn_end);
/* FIXME: Shadows are currently buggy - TBD */
if (!list_empty(&vma->shadow_list)) {
BUG();
vma_swapfile_realloc(vma, pfn_start, pfn_end);
return 0;
}
/* Release the pages before modifying the original vma */
vma_release_pages(vma, task, pfn_start, pfn_end);
/* Shrink from the beginning */
if (pfn_start > vma->pfn_start) {
diff = pfn_start - vma->pfn_start;
vma->f_offset += diff;
vma->pfn_start = pfn_start;
unsigned long obj_unmap_start, obj_unmap_end;
/* Shrink from the end */
} else if (pfn_end < vma->pfn_end) {
diff = vma->pfn_end - pfn_end;
vma->pfn_end = pfn_end;
if (vma->pfn_start < pfn_start) {
/* Calculate unmap start page */
obj_unmap_start = vma->f_offset +
(pfn_start - vma->pfn_start);
/* Calculate new vma limits */
BUG_ON(pfn_start >= vma->pfn_end);
diff = vma->pfn_end - pfn_start;
vma->pfn_end = pfn_start;
/* Calculate unmap end page */
obj_unmap_end = obj_unmap_start + diff;
/* Shrink from the beginning */
} else if (vma->pfn_end > pfn_end) {
/* Calculate unmap start page */
obj_unmap_start = vma->f_offset;
/* Calculate new vma limits */
BUG_ON(pfn_end <= vma->pfn_start);
diff = pfn_end - vma->pfn_start;
vma->f_offset += diff;
vma->pfn_start = pfn_end;
/* Calculate unmap end page */
obj_unmap_end = vma->f_offset;
} else
BUG();
return vma_unmap_shadows(vma, task, pfn_start, pfn_end);
/* Unmap those pages from the task address space */
task_unmap_region(task, vma, obj_unmap_start, obj_unmap_end);
/*
* The objects underneath may now have resident pages
* with no refs, they are released here.
*/
vma_release_pages(vma, obj_unmap_start, obj_unmap_end);
}
/*
@@ -252,8 +343,8 @@ int vma_unmap(struct vm_area *vma, struct tcb *task,
list_add_tail(&vma_new->list, &vma->list);
/* Shrink needed? */
} else if (((vma->pfn_start == pfn_start) && (vma->pfn_end > pfn_end))
|| ((vma->pfn_start < pfn_start) && (vma->pfn_end == pfn_end)))
} else if (((vma->pfn_start >= pfn_start) && (vma->pfn_end > pfn_end))
|| ((vma->pfn_start < pfn_start) && (vma->pfn_end <= pfn_end)))
vma_shrink(vma, task, pfn_start, pfn_end);
/* Destroy needed? */
@@ -297,7 +388,7 @@ int sys_munmap(struct tcb *task, void *start, unsigned long length)
if ((unsigned long)start & PAGE_MASK)
return -EINVAL;
return do_munmap(task, start, __pfn(length));
return do_munmap(task, start, __pfn(page_align_up(length)));
}