[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH v3 4/7] block: don't use constant 512 as sector
From: |
Max Reitz |
Subject: |
Re: [Qemu-devel] [PATCH v3 4/7] block: don't use constant 512 as sector size in crypto driver |
Date: |
Sat, 16 Sep 2017 18:24:56 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.3.0 |
On 2017-09-12 13:28, Daniel P. Berrange wrote:
> Use the qcrypto_block_get_sector_size() value in the block
> crypto driver instead of hardcoding 512 as the sector size.
>
> Signed-off-by: Daniel P. Berrange <address@hidden>
> ---
> block/crypto.c | 34 ++++++++++++++++++----------------
> 1 file changed, 18 insertions(+), 16 deletions(-)
>
> diff --git a/block/crypto.c b/block/crypto.c
> index d68cbac2ac..49d6d4c058 100644
> --- a/block/crypto.c
> +++ b/block/crypto.c
> @@ -392,8 +392,9 @@ block_crypto_co_readv(BlockDriverState *bs, int64_t
> sector_num,
> uint8_t *cipher_data = NULL;
> QEMUIOVector hd_qiov;
> int ret = 0;
> + uint64_t sector_size = qcrypto_block_get_sector_size(crypto->block);
> uint64_t payload_offset =
> - qcrypto_block_get_payload_offset(crypto->block) / 512;
> + qcrypto_block_get_payload_offset(crypto->block) / sector_size;
> assert(payload_offset < (INT64_MAX / 512));
>
> qemu_iovec_init(&hd_qiov, qiov->niov);
> @@ -401,9 +402,9 @@ block_crypto_co_readv(BlockDriverState *bs, int64_t
> sector_num,
> /* Bounce buffer because we don't wish to expose cipher text
> * in qiov which points to guest memory.
> */
> - cipher_data =
> - qemu_try_blockalign(bs->file->bs, MIN(BLOCK_CRYPTO_MAX_SECTORS * 512,
> - qiov->size));
> + cipher_data = qemu_try_blockalign(
> + bs->file->bs, MIN(BLOCK_CRYPTO_MAX_SECTORS * sector_size,
> + qiov->size));
> if (cipher_data == NULL) {
> ret = -ENOMEM;
> goto cleanup;
> @@ -417,7 +418,7 @@ block_crypto_co_readv(BlockDriverState *bs, int64_t
> sector_num,
> }
>
> qemu_iovec_reset(&hd_qiov);
> - qemu_iovec_add(&hd_qiov, cipher_data, cur_nr_sectors * 512);
> + qemu_iovec_add(&hd_qiov, cipher_data, cur_nr_sectors * sector_size);
cur_nr_sectors is based on remaining_sectors; which in turn is a
parameter to this function and comes from the block layer. Therefore
its unit is BDRV_SECTOR_SIZE and not the crypto driver's sector size.
Same in the hunk below, and in block_crypto_co_writev().
>
> ret = bdrv_co_readv(bs->file,
> payload_offset + sector_num,
Same thing here, albeit in a different variation: The unit of this
parameter of bdrv_co_readv() (start sector index) is a block layer
sector, whose size is always BDRV_SECTOR_SIZE.
Therefore you cannot divide the result from
qcrypto_block_get_payload_offset() by the crypto driver's sector size
and then use it as a sector index here.
Same in block_crypto_co_writev().
I assume that you fix this in the next patch, but for now it's just wrong.
Max
signature.asc
Description: OpenPGP digital signature
- [Qemu-devel] [PATCH v3 0/7] Misc improvements to crypto block driver, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 2/7] crypto: expose encryption sector size in APIs, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 3/7] block: fix data type casting for crypto payload offset, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 1/7] block: use 1 MB bounce buffers for crypto instead of 16KB, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 4/7] block: don't use constant 512 as sector size in crypto driver, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 5/7] block: convert crypto driver to bdrv_co_preadv|pwritev, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 7/7] block: support passthrough of BDRV_REQ_FUA in crypto driver, Daniel P. Berrange, 2017/09/12
- [Qemu-devel] [PATCH v3 6/7] block: convert qcrypto_block_encrypt|decrypt to take bytes offset, Daniel P. Berrange, 2017/09/12