[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180417072932.GB15310@kroah.com>
Date: Tue, 17 Apr 2018 09:29:32 +0200
From: Greg KH <greg@...ah.com>
To: longli@...rosoft.com
Cc: Steve French <sfrench@...ba.org>, linux-cifs@...r.kernel.org,
samba-technical@...ts.samba.org, linux-kernel@...r.kernel.org,
linux-rdma@...r.kernel.org, stable@...r.kernel.org
Subject: Re: [PATCH 3/6] cifs: smbd: Avoid allocating iov on the stack
On Mon, Apr 16, 2018 at 05:49:15PM -0700, Long Li wrote:
> From: Long Li <longli@...rosoft.com>
>
> It's not necessary to allocate another iov when going through the buffers
> in smbd_send() through RDMA send.
>
> Remove it to reduce stack size.
>
> Signed-off-by: Long Li <longli@...rosoft.com>
> ---
> fs/cifs/smbdirect.c | 36 ++++++++++++------------------------
> 1 file changed, 12 insertions(+), 24 deletions(-)
>
> diff --git a/fs/cifs/smbdirect.c b/fs/cifs/smbdirect.c
> index b5c6c0d..f575e9a 100644
> --- a/fs/cifs/smbdirect.c
> +++ b/fs/cifs/smbdirect.c
> @@ -2088,7 +2088,7 @@ int smbd_send(struct smbd_connection *info, struct smb_rqst *rqst)
> int start, i, j;
> int max_iov_size =
> info->max_send_size - sizeof(struct smbd_data_transfer);
> - struct kvec iov[SMBDIRECT_MAX_SGE];
> + struct kvec *iov;
> int rc;
>
> info->smbd_send_pending++;
> @@ -2099,32 +2099,20 @@ int smbd_send(struct smbd_connection *info, struct smb_rqst *rqst)
> }
>
> /*
> - * This usually means a configuration error
> - * We use RDMA read/write for packet size > rdma_readwrite_threshold
> - * as long as it's properly configured we should never get into this
> - * situation
> - */
> - if (rqst->rq_nvec + rqst->rq_npages > SMBDIRECT_MAX_SGE) {
> - log_write(ERR, "maximum send segment %x exceeding %x\n",
> - rqst->rq_nvec + rqst->rq_npages, SMBDIRECT_MAX_SGE);
> - rc = -EINVAL;
> - goto done;
> - }
> -
> - /*
> - * Remove the RFC1002 length defined in MS-SMB2 section 2.1
> - * It is used only for TCP transport
> + * Skip the RFC1002 length defined in MS-SMB2 section 2.1
> + * It is used only for TCP transport in the iov[0]
> * In future we may want to add a transport layer under protocol
> * layer so this will only be issued to TCP transport
> */
> - iov[0].iov_base = (char *)rqst->rq_iov[0].iov_base + 4;
> - iov[0].iov_len = rqst->rq_iov[0].iov_len - 4;
> - buflen += iov[0].iov_len;
> +
> + if (rqst->rq_iov[0].iov_len != 4) {
> + log_write(ERR, "expected the pdu length in 1st iov, but got 0x%lu\n", rqst->rq_iov[0].iov_len);
> + return -EINVAL;
> + }
> + iov = &rqst->rq_iov[1];
>
> /* total up iov array first */
> - for (i = 1; i < rqst->rq_nvec; i++) {
> - iov[i].iov_base = rqst->rq_iov[i].iov_base;
> - iov[i].iov_len = rqst->rq_iov[i].iov_len;
> + for (i = 0; i < rqst->rq_nvec-1; i++) {
> buflen += iov[i].iov_len;
> }
>
> @@ -2197,14 +2185,14 @@ int smbd_send(struct smbd_connection *info, struct smb_rqst *rqst)
> goto done;
> }
> i++;
> - if (i == rqst->rq_nvec)
> + if (i == rqst->rq_nvec-1)
> break;
> }
> start = i;
> buflen = 0;
> } else {
> i++;
> - if (i == rqst->rq_nvec) {
> + if (i == rqst->rq_nvec-1) {
> /* send out all remaining vecs */
> remaining_data_length -= buflen;
> log_write(INFO,
> --
> 2.7.4
<formletter>
This is not the correct way to submit patches for inclusion in the
stable kernel tree. Please read:
https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.
</formletter>
Powered by blists - more mailing lists