On Tue, Sep 01, 2015 at 03:25:54PM -0400, Josef Bacik wrote:
> On 09/01/2015 03:01 PM, Omar Sandoval wrote:
> >From: Omar Sandoval <osan...@fb.com>
> >
> >These are going to be used for the free space tree bitmap items.
> >
> >Signed-off-by: Omar Sandoval <osan...@fb.com>
>
> Can we get sanity tests for these operations so we know they are properly
> unit tested?
>
No problem, I'll do that.
> >---
> > fs/btrfs/extent_io.c | 101
> > +++++++++++++++++++++++++++++++++++++++++++++++++++
> > fs/btrfs/extent_io.h | 6 +++
> > 2 files changed, 107 insertions(+)
> >
> >diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
> >index 02d05817cbdf..649e3b4eeb1b 100644
> >--- a/fs/btrfs/extent_io.c
> >+++ b/fs/btrfs/extent_io.c
> >@@ -5475,6 +5475,107 @@ void copy_extent_buffer(struct extent_buffer *dst,
> >struct extent_buffer *src,
> > }
> > }
> >
> >+/*
> >+ * The extent buffer bitmap operations are done with byte granularity
> >because
> >+ * bitmap items are not guaranteed to be aligned to a word and therefore a
> >+ * single word in a bitmap may straddle two pages in the extent buffer.
> >+ */
> >+#define BIT_BYTE(nr) ((nr) / BITS_PER_BYTE)
> >+#define BYTE_MASK ((1 << BITS_PER_BYTE) - 1)
> >+#define BITMAP_FIRST_BYTE_MASK(start) \
> >+ ((BYTE_MASK << ((start) & (BITS_PER_BYTE - 1))) & BYTE_MASK)
> >+#define BITMAP_LAST_BYTE_MASK(nbits) \
> >+ (BYTE_MASK >> (-(nbits) & (BITS_PER_BYTE - 1)))
> >+
> >+int extent_buffer_test_bit(struct extent_buffer *eb, unsigned long start,
> >+ unsigned long nr)
> >+{
> >+ size_t offset;
> >+ char *kaddr;
> >+ struct page *page;
> >+ size_t byte_offset = BIT_BYTE(nr);
> >+ size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1);
> >+ unsigned long i = (start_offset + start + byte_offset) >>
> >PAGE_CACHE_SHIFT;
> >+
> >+ offset = (start_offset + start + byte_offset) & (PAGE_CACHE_SIZE - 1);
> >+ page = eb->pages[i];
> >+ WARN_ON(!PageUptodate(page));
> >+ kaddr = page_address(page);
> >+ return 1U & (kaddr[offset] >> (nr & (BITS_PER_BYTE - 1)));
> >+}
> >+
> >+void extent_buffer_bitmap_set(struct extent_buffer *eb, unsigned long start,
> >+ unsigned long pos, unsigned long len)
> >+{
> >+ size_t offset;
> >+ char *kaddr;
> >+ struct page *page;
> >+ size_t byte_offset = BIT_BYTE(pos);
> >+ size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1);
> >+ unsigned long i = (start_offset + start + byte_offset) >>
> >PAGE_CACHE_SHIFT;
> >+ const unsigned int size = pos + len;
> >+ int bits_to_set = BITS_PER_BYTE - (pos % BITS_PER_BYTE);
> >+ unsigned int mask_to_set = BITMAP_FIRST_BYTE_MASK(pos);
> >+
> >+ offset = (start_offset + start + byte_offset) & (PAGE_CACHE_SIZE - 1);
> >+ page = eb->pages[i];
> >+ WARN_ON(!PageUptodate(page));
> >+ kaddr = page_address(page);
> >+
> >+ while (len >= bits_to_set) {
> >+ kaddr[offset] |= mask_to_set;
> >+ len -= bits_to_set;
> >+ bits_to_set = BITS_PER_BYTE;
> >+ mask_to_set = ~0U;
> >+ if (++offset >= PAGE_CACHE_SIZE && len > 0) {
> >+ offset = 0;
> >+ page = eb->pages[++i];
> >+ WARN_ON(!PageUptodate(page));
> >+ kaddr = page_address(page);
> >+ }
> >+ }
> >+ if (len) {
> >+ mask_to_set &= BITMAP_LAST_BYTE_MASK(size);
> >+ kaddr[offset] |= mask_to_set;
> >+ }
> >+}
> >+
> >+void extent_buffer_bitmap_clear(struct extent_buffer *eb, unsigned long
> >start,
> >+ unsigned long pos, unsigned long len)
> >+{
> >+ size_t offset;
> >+ char *kaddr;
> >+ struct page *page;
> >+ size_t byte_offset = BIT_BYTE(pos);
> >+ size_t start_offset = eb->start & ((u64)PAGE_CACHE_SIZE - 1);
> >+ unsigned long i = (start_offset + start + byte_offset) >>
> >PAGE_CACHE_SHIFT;
> >+ const unsigned int size = pos + len;
> >+ int bits_to_clear = BITS_PER_BYTE - (pos % BITS_PER_BYTE);
> >+ unsigned int mask_to_clear = BITMAP_FIRST_BYTE_MASK(pos);
> >+
> >+ offset = (start_offset + start + byte_offset) & (PAGE_CACHE_SIZE - 1);
> >+ page = eb->pages[i];
> >+ WARN_ON(!PageUptodate(page));
> >+ kaddr = page_address(page);
> >+
>
> Abstract this offset finding logic to a helper function and then comment the
> hell out of it, I now have a migraine trying to figure out what is going on.
> Thanks,
>
> Josef
Will do, thanks.
--
Omar
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
