From: Mikulas Patocka Date: Tue, 20 Feb 2024 18:11:51 +0000 (+0100) Subject: dm-verity, dm-crypt: align "struct bvec_iter" correctly X-Git-Url: http://git.maquefel.me/?a=commitdiff_plain;h=787f1b2800464aa277236a66eb3c279535edd460;p=linux.git dm-verity, dm-crypt: align "struct bvec_iter" correctly "struct bvec_iter" is defined with the __packed attribute, so it is aligned on a single byte. On X86 (and on other architectures that support unaligned addresses in hardware), "struct bvec_iter" is accessed using the 8-byte and 4-byte memory instructions, however these instructions are less efficient if they operate on unaligned addresses. (on RISC machines that don't have unaligned access in hardware, GCC generates byte-by-byte accesses that are very inefficient - see [1]) This commit reorders the entries in "struct dm_verity_io" and "struct convert_context", so that "struct bvec_iter" is aligned on 8 bytes. [1] https://lore.kernel.org/all/ZcLuWUNRZadJr0tQ@fedora/T/ Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer --- diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 6044614bab761..f54cd6714af3b 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -53,11 +53,11 @@ struct convert_context { struct completion restart; struct bio *bio_in; - struct bio *bio_out; struct bvec_iter iter_in; + struct bio *bio_out; struct bvec_iter iter_out; - u64 cc_sector; atomic_t cc_pending; + u64 cc_sector; union { struct skcipher_request *req; struct aead_request *req_aead; diff --git a/drivers/md/dm-verity.h b/drivers/md/dm-verity.h index 4620a98c99561..db93a91169d5e 100644 --- a/drivers/md/dm-verity.h +++ b/drivers/md/dm-verity.h @@ -80,12 +80,12 @@ struct dm_verity_io { /* original value of bio->bi_end_io */ bio_end_io_t *orig_bi_end_io; + struct bvec_iter iter; + sector_t block; unsigned int n_blocks; bool in_tasklet; - struct bvec_iter iter; - struct work_struct work; char *recheck_buffer;