gfs2: iomap buffered write support
authorAndreas Gruenbacher <agruenba@redhat.com>
Sun, 24 Jun 2018 14:04:04 +0000 (15:04 +0100)
committerAndreas Gruenbacher <agruenba@redhat.com>
Mon, 2 Jul 2018 15:27:17 +0000 (16:27 +0100)
commit64bc06bb32ee9cf458f432097113c8b495d75757
treec87afdeaf0cd1c34ad4cf57c64a9a2c4678d84da
parentd505a96a3b16f46455035dc0296bc2da6014e163
gfs2: iomap buffered write support

With the traditional page-based writes, blocks are allocated separately
for each page written to.  With iomap writes, we can allocate a lot more
blocks at once, with a fraction of the allocation overhead for each
page.

Split calculating the number of blocks that can be allocated at a given
position (gfs2_alloc_size) off from gfs2_iomap_alloc: that size
determines the number of blocks to allocate and reserve in the journal.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Reviewed-by: Bob Peterson <rpeterso@redhat.com>
fs/gfs2/aops.c
fs/gfs2/aops.h [new file with mode: 0644]
fs/gfs2/bmap.c
fs/gfs2/file.c