Commit bfe7b00d authored by Song Liu's avatar Song Liu Committed by Linus Torvalds

mm, thp: introduce FOLL_SPLIT_PMD

Introduce a new foll_flag: FOLL_SPLIT_PMD.  As the name says
FOLL_SPLIT_PMD splits huge pmd for given mm_struct, the underlining huge
page stays as-is.

FOLL_SPLIT_PMD is useful for cases where we need to use regular pages, but
would switch back to huge page and huge pmd on.  One of such example is
uprobe.  The following patches use FOLL_SPLIT_PMD in uprobe.

Link: http://lkml.kernel.org/r/20190815164525.1848545-4-songliubraving@fb.comSigned-off-by: default avatarSong Liu <songliubraving@fb.com>
Reviewed-by: default avatarOleg Nesterov <oleg@redhat.com>
Acked-by: default avatarKirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent fb4fb04f
...@@ -2591,6 +2591,7 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address, ...@@ -2591,6 +2591,7 @@ struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
#define FOLL_COW 0x4000 /* internal GUP flag */ #define FOLL_COW 0x4000 /* internal GUP flag */
#define FOLL_ANON 0x8000 /* don't do file mappings */ #define FOLL_ANON 0x8000 /* don't do file mappings */
#define FOLL_LONGTERM 0x10000 /* mapping lifetime is indefinite: see below */ #define FOLL_LONGTERM 0x10000 /* mapping lifetime is indefinite: see below */
#define FOLL_SPLIT_PMD 0x20000 /* split huge pmd before returning */
/* /*
* NOTE on FOLL_LONGTERM: * NOTE on FOLL_LONGTERM:
......
...@@ -384,7 +384,7 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma, ...@@ -384,7 +384,7 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma,
spin_unlock(ptl); spin_unlock(ptl);
return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap); return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
} }
if (flags & FOLL_SPLIT) { if (flags & (FOLL_SPLIT | FOLL_SPLIT_PMD)) {
int ret; int ret;
page = pmd_page(*pmd); page = pmd_page(*pmd);
if (is_huge_zero_page(page)) { if (is_huge_zero_page(page)) {
...@@ -393,7 +393,7 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma, ...@@ -393,7 +393,7 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma,
split_huge_pmd(vma, pmd, address); split_huge_pmd(vma, pmd, address);
if (pmd_trans_unstable(pmd)) if (pmd_trans_unstable(pmd))
ret = -EBUSY; ret = -EBUSY;
} else { } else if (flags & FOLL_SPLIT) {
if (unlikely(!try_get_page(page))) { if (unlikely(!try_get_page(page))) {
spin_unlock(ptl); spin_unlock(ptl);
return ERR_PTR(-ENOMEM); return ERR_PTR(-ENOMEM);
...@@ -405,6 +405,10 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma, ...@@ -405,6 +405,10 @@ static struct page *follow_pmd_mask(struct vm_area_struct *vma,
put_page(page); put_page(page);
if (pmd_none(*pmd)) if (pmd_none(*pmd))
return no_page_table(vma, flags); return no_page_table(vma, flags);
} else { /* flags & FOLL_SPLIT_PMD */
spin_unlock(ptl);
split_huge_pmd(vma, pmd, address);
ret = pte_alloc(mm, pmd) ? -ENOMEM : 0;
} }
return ret ? ERR_PTR(ret) : return ret ? ERR_PTR(ret) :
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment