An additional testcase found an issue with the last series of patches applied: the fallback solution may not save the iv value after operation. This very small fix just makes sure the iv is copied back to the walk/desc struct. Signed-off-by: Harald Freudenberger <freude@xxxxxxxxxxxxxxxxxx> --- arch/s390/crypto/aes_s390.c | 3 +++ arch/s390/crypto/des_s390.c | 3 +++ 2 files changed, 6 insertions(+) diff --git a/arch/s390/crypto/aes_s390.c b/arch/s390/crypto/aes_s390.c index cf3c008..23223cd 100644 --- a/arch/s390/crypto/aes_s390.c +++ b/arch/s390/crypto/aes_s390.c @@ -820,6 +820,9 @@ static int ctr_aes_crypt(struct blkcipher_desc *desc, long func, else memcpy(walk->iv, ctrptr, AES_BLOCK_SIZE); spin_unlock(&ctrblk_lock); + } else { + if (!nbytes) + memcpy(walk->iv, ctrptr, AES_BLOCK_SIZE); } /* * final block may be < AES_BLOCK_SIZE, copy only nbytes diff --git a/arch/s390/crypto/des_s390.c b/arch/s390/crypto/des_s390.c index 0a5aac8..7acb77f 100644 --- a/arch/s390/crypto/des_s390.c +++ b/arch/s390/crypto/des_s390.c @@ -429,6 +429,9 @@ static int ctr_desall_crypt(struct blkcipher_desc *desc, long func, else memcpy(walk->iv, ctrptr, DES_BLOCK_SIZE); spin_unlock(&ctrblk_lock); + } else { + if (!nbytes) + memcpy(walk->iv, ctrptr, DES_BLOCK_SIZE); } /* final block may be < DES_BLOCK_SIZE, copy only nbytes */ if (nbytes) { -- 1.7.9.5 -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html