On 05/25/2015 11:54 PM, Naoya Horiguchi wrote:
> On Thu, May 21, 2015 at 08:47:43AM -0700, Mike Kravetz wrote:
>> This is based on the shmem version, but it has diverged quite
>> a bit. We have no swap to worry about, nor the new file sealing.
>> Add synchronication via the fault mutex table to coordinate
>> page faults, fallocate allocation and fallocate hole punch.
>>
>> What this allows us to do is move physical memory in and out of
>> a hugetlbfs file without having it mapped. This also gives us
>> the ability to support MADV_REMOVE since it is currently
>> implemented using fallocate(). MADV_REMOVE lets madvise() remove
>> pages from the middle of a hugetlbfs file, which wasn't possible
>> before.
>>
>> hugetlbfs fallocate only operates on whole huge pages.
>>
>> Based-on code-by: Dave Hansen <dave.han...@linux.intel.com>
>> Signed-off-by: Mike Kravetz <mike.krav...@oracle.com>
>
> This patch changes the behavior of user API, so please update manpage of
> fallocate(2).
Will do.
Unfortunately, I believe hugetlbfs does not follow the man page
for ftruncate. So, I will look to get that updated as well.
>> ---
>> fs/hugetlbfs/inode.c | 169
>> +++++++++++++++++++++++++++++++++++++++++++++++-
>> include/linux/hugetlb.h | 3 +
>> mm/hugetlb.c | 2 +-
>> 3 files changed, 172 insertions(+), 2 deletions(-)
>>
>> diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
>> index dfa88a5..4b1535f 100644
>> --- a/fs/hugetlbfs/inode.c
>> +++ b/fs/hugetlbfs/inode.c
>> @@ -12,6 +12,7 @@
>> #include <linux/thread_info.h>
>> #include <asm/current.h>
>> #include <linux/sched.h> /* remove ASAP */
>> +#include <linux/falloc.h>
>> #include <linux/fs.h>
>> #include <linux/mount.h>
>> #include <linux/file.h>
>> @@ -493,6 +494,171 @@ static int hugetlb_vmtruncate(struct inode *inode,
>> loff_t offset)
>> return 0;
>> }
>>
>> +static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t
>> len)
>> +{
>> + struct hstate *h = hstate_inode(inode);
>> + unsigned long hpage_size = huge_page_size(h);
>> + loff_t hole_start, hole_end;
>> +
>> + /*
>> + * For hole punch round up the beginning offset of the hole and
>> + * round down the end.
>> + */
>> + hole_start = (offset + hpage_size - 1) & huge_page_mask(h);
>> + hole_end = (offset + len) & huge_page_mask(h);
>
> We have round_up/round_up macro, so please use them here.
> Then, it's self-descriptive, so you don't have to write comment.
>
>> +
>> + if ((u64)hole_end > (u64)hole_start) {
>
> Why is this casting to u64 necessary?
It is not necessary. I will remove it.
>> + struct address_space *mapping = inode->i_mapping;
>> +
>> + mutex_lock(&inode->i_mutex);
>> + i_mmap_lock_write(mapping);
>> + if (!RB_EMPTY_ROOT(&mapping->i_mmap))
>> + hugetlb_vmdelete_list(&mapping->i_mmap,
>> + hole_start >> PAGE_SHIFT,
>> + hole_end >> PAGE_SHIFT);
>> + i_mmap_unlock_write(mapping);
>> + remove_inode_hugepages(inode, hole_start, hole_end);
>> + mutex_unlock(&inode->i_mutex);
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static long hugetlbfs_fallocate(struct file *file, int mode, loff_t offset,
>> + loff_t len)
>> +{
>> + struct inode *inode = file_inode(file);
>> + struct address_space *mapping = inode->i_mapping;
>> + struct hstate *h = hstate_inode(inode);
>> + struct vm_area_struct pseudo_vma;
>> + unsigned long hpage_size = huge_page_size(h);
>> + unsigned long hpage_shift = huge_page_shift(h);
>> + pgoff_t start, index, end;
>> + int error;
>> + u32 hash;
>> +
>> + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
>> + return -EOPNOTSUPP;
>> +
>> + if (mode & FALLOC_FL_PUNCH_HOLE)
>> + return hugetlbfs_punch_hole(inode, offset, len);
>> +
>> + /*
>> + * Default preallocate case.
>> + * For this range, start is rounded down and end is rounded up.
>> + */
>> + start = offset >> hpage_shift;
>> + end = (offset + len + hpage_size - 1) >> hpage_shift;
>> +
>> + mutex_lock(&inode->i_mutex);
>> +
>> + /* We need to check rlimit even when FALLOC_FL_KEEP_SIZE */
>> + error = inode_newsize_ok(inode, offset + len);
>> + if (error)
>> + goto out;
>> +
>> + /*
>> + * Initialize a pseudo vma that just contains the policy used
>> + * when allocating the huge pages. The actual policy field
>> + * (vm_policy) is determined based on the index in the loop below.
>> + */
>> + memset(&pseudo_vma, 0, sizeof(struct vm_area_struct));
>> + pseudo_vma.vm_start = 0;
>> + pseudo_vma.vm_flags |= (VM_HUGETLB | VM_MAYSHARE | VM_SHARED);
>
> Maybe '|' isn't necessary.
No, it is not necessary. I will remove.
>> + pseudo_vma.vm_file = file;
>> +
>> + for (index = start; index < end; index++) {
>> + /*
>> + * This is supposed to be the vaddr where the page is being
>> + * faulted in, but we have no vaddr here.
>> + */
>> + struct page *page;
>> + unsigned long addr;
>> + int avoid_reserve = 0;
>
> avoid_reserve is referred only once and never changed, so no need to use
> the variable?
It is not necessary. I will remove it.
>> +
>> + cond_resched();
>> +
>> + /*
>> + * fallocate(2) manpage permits EINTR; we may have been
>> + * interrupted because we are using up too much memory.
>> + */
>> + if (signal_pending(current)) {
>> + error = -EINTR;
>> + break;
>> + }
>> +
>> + /* Get policy based on index */
>> + pseudo_vma.vm_policy =
>> + mpol_shared_policy_lookup(&HUGETLBFS_I(inode)->policy,
>> + index);
>> +
>> + /* addr is the offset within the file (zero based) */
>> + addr = index * hpage_size;
>> +
>> + /* mutex taken here, fault path and hole punch */
>> + hash = hugetlb_fault_mutex_shared_hash(mapping, index);
>> + hugetlb_fault_mutex_lock(hash);
>> +
>> + /* see if page already exists to avoid alloc/free */
>> + page = find_get_page(mapping, index);
>> + if (page) {
>> + put_page(page);
>> + hugetlb_fault_mutex_unlock(hash);
>
> Don't you need mpol_cond_put() here?
Yes. Thank you, I will add it.
>> + continue;
>> + }
>> +
>> + page = alloc_huge_page(&pseudo_vma, addr, avoid_reserve);
>> + mpol_cond_put(pseudo_vma.vm_policy);
>> + if (IS_ERR(page)) {
>> + hugetlb_fault_mutex_unlock(hash);
>> + error = PTR_ERR(page);
>> + goto out;
>> + }
>> + clear_huge_page(page, addr, pages_per_huge_page(h));
>> + __SetPageUptodate(page);
>
> Note that recently I added page_huge_active() to mark activeness of hugepages,
> so when you rebased to v4.1-rc1+, please insert set_page_huge_active(page)
> here.
>
Yes, I noticed your change.
>> + error = huge_add_to_page_cache(page, mapping, index);
>> + if (error) {
>> + /*
>> + * An entry already exists in the cache. This implies
>> + * a region also existed in the reserve map at the time
>> + * the page was allocated above. Therefore, no use
>> + * count was added to the subpool for the page. Before
>> + * freeing the page, clear the subpool reference so
>> + * that the count is not decremented.
>> + */
>> + set_page_private(page, 0);/* clear spool reference */
>
> This looks unclear to me. Which "count" do you refer to in the comment
> "no use count was added to the subpool" or "the count is not decremented"?
> I guess spool->used_hpages or spool->rsv_hpages, but alloc_huge_page() above
> should call hugepage_subpool_get_pages(), so it's accounted, right?
> Could you write comments more specifically?
Yes, this is confusing. As I am reexamining the code, I see that
it is incorrect. This code may not be necessary. It was there to
handle a race with page faults. The code now uses the hugetlb fault
mutex table to synchronize with page faults. I will do some more
work here and expect this confusing code to go away.
Thank you for your comments,
--
Mike Kravetz
>
> Thanks,
> Naoya Horiguchi
>
>> + put_page(page);
>> +
>> + hugetlb_fault_mutex_unlock(hash);
>> + /* Keep going if we see an -EEXIST */
>> + if (error == -EEXIST) {
>> + error = 0; /* do not return to user */
>> + continue;
>> + } else
>> + goto out;
>> + }
>> +
>> + hugetlb_fault_mutex_unlock(hash);
>> +
>> + /*
>> + * page_put due to reference from alloc_huge_page()
>> + * unlock_page because locked by add_to_page_cache()
>> + */
>> + put_page(page);
>> + unlock_page(page);
>> + }
>> +
>> + if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size)
>> + i_size_write(inode, offset + len);
>> + inode->i_ctime = CURRENT_TIME;
>> + spin_lock(&inode->i_lock);
>> + inode->i_private = NULL;
>> + spin_unlock(&inode->i_lock);
>> +out:
>> + mutex_unlock(&inode->i_mutex);
>> + return error;
>> +}
>> +
>> static int hugetlbfs_setattr(struct dentry *dentry, struct iattr *attr)
>> {
>> struct inode *inode = dentry->d_inode;
>> @@ -804,7 +970,8 @@ const struct file_operations hugetlbfs_file_operations =
>> {
>> .mmap = hugetlbfs_file_mmap,
>> .fsync = noop_fsync,
>> .get_unmapped_area = hugetlb_get_unmapped_area,
>> - .llseek = default_llseek,
>> + .llseek = default_llseek,
>> + .fallocate = hugetlbfs_fallocate,
>> };
>>
>> static const struct inode_operations hugetlbfs_dir_inode_operations = {
>> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
>> index 934f339..fa36b7a 100644
>> --- a/include/linux/hugetlb.h
>> +++ b/include/linux/hugetlb.h
>> @@ -327,6 +327,8 @@ struct huge_bootmem_page {
>> #endif
>> };
>>
>> +struct page *alloc_huge_page(struct vm_area_struct *vma,
>> + unsigned long addr, int avoid_reserve);
>> struct page *alloc_huge_page_node(struct hstate *h, int nid);
>> struct page *alloc_huge_page_noerr(struct vm_area_struct *vma,
>> unsigned long addr, int avoid_reserve);
>> @@ -481,6 +483,7 @@ static inline bool hugepages_supported(void)
>>
>> #else /* CONFIG_HUGETLB_PAGE */
>> struct hstate {};
>> +#define alloc_huge_page(v, a, r) NULL
>> #define alloc_huge_page_node(h, nid) NULL
>> #define alloc_huge_page_noerr(v, a, r) NULL
>> #define alloc_bootmem_huge_page(h) NULL
>> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
>> index 94c6154..1e95038 100644
>> --- a/mm/hugetlb.c
>> +++ b/mm/hugetlb.c
>> @@ -1444,7 +1444,7 @@ static long vma_commit_reservation(struct hstate *h,
>> /* Forward declaration */
>> static int hugetlb_acct_memory(struct hstate *h, long delta);
>>
>> -static struct page *alloc_huge_page(struct vm_area_struct *vma,
>> +struct page *alloc_huge_page(struct vm_area_struct *vma,
>> unsigned long addr, int avoid_reserve)
>> {
>> struct hugepage_subpool *spool = subpool_vma(vma);
>> --
>> 2.1.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
