Re: [RFC PATCH V3] iomap: add support to track dirty state of sub pages

[yukuai (C)]

On 2020/8/19 20:56, Gao Xiang wrote:

On Wed, Aug 19, 2020 at 08:05:42PM +0800, Yu Kuai wrote:

...

+static void
+iomap_iop_set_range_dirty(struct page *page, unsigned int off,
+               unsigned int len)
+{
+       struct iomap_page *iop = to_iomap_page(page);
+       struct inode *inode = page->mapping->host;
+       unsigned int first = DIRTY_BITS(off >> inode->i_blkbits);
+       unsigned int last = DIRTY_BITS((off + len - 1) >> inode->i_blkbits);
+       unsigned long flags;
+       unsigned int i;
+
+       spin_lock_irqsave(&iop->state_lock, flags);
+       for (i = first; i <= last; i++)
+               set_bit(i, iop->state);
+
+       if (last >= first)
+               iomap_set_page_dirty(page);

set_page_dirty() in the atomic context?

Hi,

You'are right, this shouldn't be inside spin_lock.

+
+       spin_unlock_irqrestore(&iop->state_lock, flags);
+}
+
+static void
+iomap_set_range_dirty(struct page *page, unsigned int off,
+               unsigned int len)
+{
+       if (PageError(page))
+               return;
+
+       if (page_has_private(page))
+               iomap_iop_set_range_dirty(page, off, len);

I vaguely remembered iomap doesn't always set up PagePrivate.

If so, maybe I should move iomap_set_page_dirty() to
ioamp_set_range_dirty().

Thanks,
Yu Kuai

@@ -705,7 +770,7 @@ __iomap_write_end(struct inode *inode, loff_t pos, unsigned 
len,
        if (unlikely(copied < len && !PageUptodate(page)))
                return 0;
        iomap_set_range_uptodate(page, offset_in_page(pos), len);
-       iomap_set_page_dirty(page);
+       iomap_set_range_dirty(page, offset_in_page(pos), len);
        return copied;
  }

so here could be suspectable, but I might be wrong here since
I just take a quick look.

Thanks,
Gao Xiang


.