> -----Original Message-----
> From: David Howells <dhowells@xxxxxxxxxx>
> Sent: Thursday, 16 March 2023 16:26
> To: Matthew Wilcox <willy@xxxxxxxxxxxxx>; David S. Miller
> <davem@xxxxxxxxxxxxx>; Eric Dumazet <edumazet@xxxxxxxxxx>; Jakub Kicinski
> <kuba@xxxxxxxxxx>; Paolo Abeni <pabeni@xxxxxxxxxx>
> Cc: David Howells <dhowells@xxxxxxxxxx>; Al Viro <viro@xxxxxxxxxxxxxxxxxx>;
> Christoph Hellwig <hch@xxxxxxxxxxxxx>; Jens Axboe <axboe@xxxxxxxxx>; Jeff
> Layton <jlayton@xxxxxxxxxx>; Christian Brauner <brauner@xxxxxxxxxx>; Linus
> Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>; netdev@xxxxxxxxxxxxxxx; linux-
> fsdevel@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; linux-mm@xxxxxxxxx;
> Bernard Metzler <BMT@xxxxxxxxxxxxxx>; Tom Talpey <tom@xxxxxxxxxx>; linux-
> rdma@xxxxxxxxxxxxxxx
> Subject: [EXTERNAL] [RFC PATCH 18/28] siw: Use sendmsg(MSG_SPLICE_PAGES)
> rather than sendpage to transmit
>
> When transmitting data, call down into TCP using a single sendmsg with
> MSG_SPLICE_PAGES to indicate that content should be spliced rather than
> performing several sendmsg and sendpage calls to transmit header, data
> pages and trailer.
>
> To make this work, the data is assembled in a bio_vec array and attached to
> a BVEC-type iterator. The header and trailer (if present) are copied into
> memory acquired from zcopy_alloc() which just breaks a page up into small
> pieces that can be freed with put_page().
>
> Signed-off-by: David Howells <dhowells@xxxxxxxxxx>
> cc: Bernard Metzler <bmt@xxxxxxxxxxxxxx>
> cc: Tom Talpey <tom@xxxxxxxxxx>
> cc: "David S. Miller" <davem@xxxxxxxxxxxxx>
> cc: Eric Dumazet <edumazet@xxxxxxxxxx>
> cc: Jakub Kicinski <kuba@xxxxxxxxxx>
> cc: Paolo Abeni <pabeni@xxxxxxxxxx>
> cc: Jens Axboe <axboe@xxxxxxxxx>
> cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
> cc: linux-rdma@xxxxxxxxxxxxxxx
> cc: netdev@xxxxxxxxxxxxxxx
> ---
> drivers/infiniband/sw/siw/siw_qp_tx.c | 231 +++++---------------------
> 1 file changed, 46 insertions(+), 185 deletions(-)
>
> diff --git a/drivers/infiniband/sw/siw/siw_qp_tx.c
> b/drivers/infiniband/sw/siw/siw_qp_tx.c
> index 8fc179321e2b..ec4f0ac324ce 100644
> --- a/drivers/infiniband/sw/siw/siw_qp_tx.c
> +++ b/drivers/infiniband/sw/siw/siw_qp_tx.c
> @@ -8,6 +8,7 @@
> #include <linux/net.h>
> #include <linux/scatterlist.h>
> #include <linux/highmem.h>
> +#include <linux/zcopy_alloc.h>
> #include <net/tcp.h>
>
> #include <rdma/iw_cm.h>
> @@ -312,114 +313,8 @@ static int siw_tx_ctrl(struct siw_iwarp_tx *c_tx,
> struct socket *s,
> return rv;
> }
>
This patch looks really great! We would get rid of all that
convoluted page mapping and in-order unmapping mechanics and
extra 0copy transmit path, making the core siw transmit
function overall a lot less complex.
> -/*
> - * 0copy TCP transmit interface: Use MSG_SPLICE_PAGES.
> - *
> - * Using sendpage to push page by page appears to be less efficient
> - * than using sendmsg, even if data are copied.
> - *
> - * A general performance limitation might be the extra four bytes
> - * trailer checksum segment to be pushed after user data.
> - */
> -static int siw_tcp_sendpages(struct socket *s, struct page **page, int
> offset,
> - size_t size)
> -{
> - struct bio_vec bvec;
> - struct msghdr msg = {
> - .msg_flags = (MSG_SPLICE_PAGES | MSG_MORE | MSG_DONTWAIT |
> - MSG_SENDPAGE_NOTLAST),
> - };
> - struct sock *sk = s->sk;
> - int i = 0, rv = 0, sent = 0;
> -
> - while (size) {
> - size_t bytes = min_t(size_t, PAGE_SIZE - offset, size);
> -
> - if (size + offset <= PAGE_SIZE)
> - msg.msg_flags = MSG_SPLICE_PAGES | MSG_MORE |
> MSG_DONTWAIT;
> -
> - tcp_rate_check_app_limited(sk);
> - bvec_set_page(&bvec, page[i], bytes, offset);
> - iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, &bvec, 1, size);
> -
> -try_page_again:
> - lock_sock(sk);
> - rv = tcp_sendmsg_locked(sk, &msg, size);
> - release_sock(sk);
> -
> - if (rv > 0) {
> - size -= rv;
> - sent += rv;
> - if (rv != bytes) {
> - offset += rv;
> - bytes -= rv;
> - goto try_page_again;
> - }
> - offset = 0;
> - } else {
> - if (rv == -EAGAIN || rv == 0)
> - break;
> - return rv;
> - }
> - i++;
> - }
> - return sent;
> -}
> -
> -/*
> - * siw_0copy_tx()
> - *
> - * Pushes list of pages to TCP socket. If pages from multiple
> - * SGE's, all referenced pages of each SGE are pushed in one
> - * shot.
> - */
> -static int siw_0copy_tx(struct socket *s, struct page **page,
> - struct siw_sge *sge, unsigned int offset,
> - unsigned int size)
> -{
> - int i = 0, sent = 0, rv;
> - int sge_bytes = min(sge->length - offset, size);
> -
> - offset = (sge->laddr + offset) & ~PAGE_MASK;
> -
> - while (sent != size) {
> - rv = siw_tcp_sendpages(s, &page[i], offset, sge_bytes);
> - if (rv >= 0) {
> - sent += rv;
> - if (size == sent || sge_bytes > rv)
> - break;
> -
> - i += PAGE_ALIGN(sge_bytes + offset) >> PAGE_SHIFT;
> - sge++;
> - sge_bytes = min(sge->length, size - sent);
> - offset = sge->laddr & ~PAGE_MASK;
> - } else {
> - sent = rv;
> - break;
> - }
> - }
> - return sent;
> -}
> -
> #define MAX_TRAILER (MPA_CRC_SIZE + 4)
>
> -static void siw_unmap_pages(struct kvec *iov, unsigned long kmap_mask, int
> len)
> -{
> - int i;
> -
> - /*
> - * Work backwards through the array to honor the kmap_local_page()
> - * ordering requirements.
> - */
> - for (i = (len-1); i >= 0; i--) {
> - if (kmap_mask & BIT(i)) {
> - unsigned long addr = (unsigned long)iov[i].iov_base;
> -
> - kunmap_local((void *)(addr & PAGE_MASK));
> - }
> - }
> -}
> -
> /*
> * siw_tx_hdt() tries to push a complete packet to TCP where all
> * packet fragments are referenced by the elements of one iovec.
> @@ -439,15 +334,13 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> {
> struct siw_wqe *wqe = &c_tx->wqe_active;
> struct siw_sge *sge = &wqe->sqe.sge[c_tx->sge_idx];
> - struct kvec iov[MAX_ARRAY];
> - struct page *page_array[MAX_ARRAY];
> + struct bio_vec bvec[MAX_ARRAY];
> struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_EOR };
>
> int seg = 0, do_crc = c_tx->do_crc, is_kva = 0, rv;
> unsigned int data_len = c_tx->bytes_unsent, hdr_len = 0, trl_len = 0,
> sge_off = c_tx->sge_off, sge_idx = c_tx->sge_idx,
> pbl_idx = c_tx->pbl_idx;
> - unsigned long kmap_mask = 0L;
>
> if (c_tx->state == SIW_SEND_HDR) {
> if (c_tx->use_sendpage) {
> @@ -457,10 +350,12 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
Can't we now also get rid of that extra hdr handling for zero copy
transmits? If we put all data pages and trailer crc in one bvec
anyway, can't we do that for the hdr as well unconditionally?
>
> c_tx->state = SIW_SEND_DATA;
> } else {
> - iov[0].iov_base =
> - (char *)&c_tx->pkt.ctrl + c_tx->ctrl_sent;
> - iov[0].iov_len = hdr_len =
> - c_tx->ctrl_len - c_tx->ctrl_sent;
> + const void *hdr = &c_tx->pkt.ctrl + c_tx->ctrl_sent;
> +
> + hdr_len = c_tx->ctrl_len - c_tx->ctrl_sent;
> + rv = zcopy_memdup(hdr_len, hdr, &bvec[0], GFP_NOFS);
> + if (rv < 0)
> + goto done;
> seg = 1;
> }
> }
> @@ -478,28 +373,9 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> } else {
> is_kva = 1;
> }
> - if (is_kva && !c_tx->use_sendpage) {
> - /*
> - * tx from kernel virtual address: either inline data
> - * or memory region with assigned kernel buffer
> - */
> - iov[seg].iov_base =
> - (void *)(uintptr_t)(sge->laddr + sge_off);
> - iov[seg].iov_len = sge_len;
> -
> - if (do_crc)
> - crypto_shash_update(c_tx->mpa_crc_hd,
> - iov[seg].iov_base,
> - sge_len);
> - sge_off += sge_len;
> - data_len -= sge_len;
> - seg++;
> - goto sge_done;
> - }
>
> while (sge_len) {
> size_t plen = min((int)PAGE_SIZE - fp_off, sge_len);
> - void *kaddr;
>
> if (!is_kva) {
> struct page *p;
> @@ -512,33 +388,12 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> p = siw_get_upage(mem->umem,
> sge->laddr + sge_off);
> if (unlikely(!p)) {
> - siw_unmap_pages(iov, kmap_mask, seg);
> wqe->processed -= c_tx->bytes_unsent;
> rv = -EFAULT;
> goto done_crc;
> }
> - page_array[seg] = p;
> -
> - if (!c_tx->use_sendpage) {
> - void *kaddr = kmap_local_page(p);
> -
> - /* Remember for later kunmap() */
> - kmap_mask |= BIT(seg);
> - iov[seg].iov_base = kaddr + fp_off;
> - iov[seg].iov_len = plen;
> -
> - if (do_crc)
> - crypto_shash_update(
> - c_tx->mpa_crc_hd,
> - iov[seg].iov_base,
> - plen);
> - } else if (do_crc) {
> - kaddr = kmap_local_page(p);
> - crypto_shash_update(c_tx->mpa_crc_hd,
> - kaddr + fp_off,
> - plen);
> - kunmap_local(kaddr);
> - }
> +
> + bvec_set_page(&bvec[seg], p, plen, fp_off);
> } else {
> /*
> * Cast to an uintptr_t to preserve all 64 bits
> @@ -552,12 +407,15 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> * bits on a 64 bit platform and 32 bits on a
> * 32 bit platform.
> */
> - page_array[seg] = virt_to_page((void *)(va &
> PAGE_MASK));
> - if (do_crc)
> - crypto_shash_update(
> - c_tx->mpa_crc_hd,
> - (void *)va,
> - plen);
> + bvec_set_virt(&bvec[seg], (void *)va, plen);
> + }
> +
> + if (do_crc) {
> + void *kaddr = kmap_local_page(bvec[seg].bv_page);
> + crypto_shash_update(c_tx->mpa_crc_hd,
> + kaddr + bvec[seg].bv_offset,
> + bvec[seg].bv_len);
> + kunmap_local(kaddr);
> }
>
> sge_len -= plen;
> @@ -567,13 +425,12 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
>
> if (++seg > (int)MAX_ARRAY) {
> siw_dbg_qp(tx_qp(c_tx), "to many fragments\n");
> - siw_unmap_pages(iov, kmap_mask, seg-1);
> wqe->processed -= c_tx->bytes_unsent;
> rv = -EMSGSIZE;
> goto done_crc;
> }
> }
> -sge_done:
> +
> /* Update SGE variables at end of SGE */
> if (sge_off == sge->length &&
> (data_len != 0 || wqe->processed < wqe->bytes)) {
> @@ -582,15 +439,8 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> sge_off = 0;
> }
> }
> - /* trailer */
> - if (likely(c_tx->state != SIW_SEND_TRAILER)) {
> - iov[seg].iov_base = &c_tx->trailer.pad[4 - c_tx->pad];
> - iov[seg].iov_len = trl_len = MAX_TRAILER - (4 - c_tx->pad);
> - } else {
> - iov[seg].iov_base = &c_tx->trailer.pad[c_tx->ctrl_sent];
> - iov[seg].iov_len = trl_len = MAX_TRAILER - c_tx->ctrl_sent;
> - }
>
> + /* Set the CRC in the trailer */
> if (c_tx->pad) {
> *(u32 *)c_tx->trailer.pad = 0;
> if (do_crc)
> @@ -603,23 +453,31 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx,
> struct socket *s)
> else if (do_crc)
> crypto_shash_final(c_tx->mpa_crc_hd, (u8 *)&c_tx->trailer.crc);
>
> - data_len = c_tx->bytes_unsent;
> + /* Copy the trailer and add it to the output list */
> + if (likely(c_tx->state != SIW_SEND_TRAILER)) {
> + void *trl = &c_tx->trailer.pad[4 - c_tx->pad];
>
> - if (c_tx->use_sendpage) {
> - rv = siw_0copy_tx(s, page_array, &wqe->sqe.sge[c_tx->sge_idx],
> - c_tx->sge_off, data_len);
> - if (rv == data_len) {
> - rv = kernel_sendmsg(s, &msg, &iov[seg], 1, trl_len);
> - if (rv > 0)
> - rv += data_len;
> - else
> - rv = data_len;
> - }
> + trl_len = MAX_TRAILER - (4 - c_tx->pad);
> + rv = zcopy_memdup(trl_len, trl, &bvec[seg], GFP_NOFS);
> + if (rv < 0)
> + goto done_crc;
> } else {
> - rv = kernel_sendmsg(s, &msg, iov, seg + 1,
> - hdr_len + data_len + trl_len);
> - siw_unmap_pages(iov, kmap_mask, seg);
> + void *trl = &c_tx->trailer.pad[c_tx->ctrl_sent];
> +
> + trl_len = MAX_TRAILER - c_tx->ctrl_sent;
> + rv = zcopy_memdup(trl_len, trl, &bvec[seg], GFP_NOFS);
> + if (rv < 0)
> + goto done_crc;
> }
> +
> + data_len = c_tx->bytes_unsent;
> +
> + if (c_tx->use_sendpage)
> + msg.msg_flags |= MSG_SPLICE_PAGES;
> + iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, bvec, seg + 1,
> + hdr_len + data_len + trl_len);
> + rv = sock_sendmsg(s, &msg);
> +
> if (rv < (int)hdr_len) {
> /* Not even complete hdr pushed or negative rv */
> wqe->processed -= data_len;
> @@ -680,6 +538,9 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct
> socket *s)
> }
> done_crc:
> c_tx->do_crc = 0;
> + if (c_tx->state == SIW_SEND_HDR)
> + folio_put(page_folio(bvec[0].bv_page));
> + folio_put(page_folio(bvec[seg].bv_page));
> done:
> return rv;
> }
