From ea08eeccae9297f7aabd2ef7f0c2517ac4549acc Mon Sep 17 00:00:00 2001
From: hc <hc@nodka.com>
Date: Tue, 20 Feb 2024 01:18:26 +0000
Subject: [PATCH] write in 30M
---
kernel/net/rds/ib_recv.c | 82 +++++++++++++++++++++++++++-------------
1 files changed, 55 insertions(+), 27 deletions(-)
diff --git a/kernel/net/rds/ib_recv.c b/kernel/net/rds/ib_recv.c
index 2f16146..cfbf0e1 100644
--- a/kernel/net/rds/ib_recv.c
+++ b/kernel/net/rds/ib_recv.c
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2006, 2017 Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2006, 2019 Oracle and/or its affiliates. All rights reserved.
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
@@ -61,7 +61,7 @@
recv->r_wr.num_sge = RDS_IB_RECV_SGE;
sge = &recv->r_sge[0];
- sge->addr = ic->i_recv_hdrs_dma + (i * sizeof(struct rds_header));
+ sge->addr = ic->i_recv_hdrs_dma[i];
sge->length = sizeof(struct rds_header);
sge->lkey = ic->i_pd->local_dma_lkey;
@@ -168,6 +168,7 @@
list_del(&inc->ii_cache_entry);
WARN_ON(!list_empty(&inc->ii_frags));
kmem_cache_free(rds_ib_incoming_slab, inc);
+ atomic_dec(&rds_ib_allocation);
}
rds_ib_cache_xfer_to_ready(&ic->i_cache_frags);
@@ -309,8 +310,8 @@
struct rds_ib_connection *ic = conn->c_transport_data;
struct ib_sge *sge;
int ret = -ENOMEM;
- gfp_t slab_mask = GFP_NOWAIT;
- gfp_t page_mask = GFP_NOWAIT;
+ gfp_t slab_mask = gfp;
+ gfp_t page_mask = gfp;
if (gfp & __GFP_DIRECT_RECLAIM) {
slab_mask = GFP_KERNEL;
@@ -342,12 +343,12 @@
WARN_ON(ret != 1);
sge = &recv->r_sge[0];
- sge->addr = ic->i_recv_hdrs_dma + (recv - ic->i_recvs) * sizeof(struct rds_header);
+ sge->addr = ic->i_recv_hdrs_dma[recv - ic->i_recvs];
sge->length = sizeof(struct rds_header);
sge = &recv->r_sge[1];
- sge->addr = ib_sg_dma_address(ic->i_cm_id->device, &recv->r_frag->f_sg);
- sge->length = ib_sg_dma_len(ic->i_cm_id->device, &recv->r_frag->f_sg);
+ sge->addr = sg_dma_address(&recv->r_frag->f_sg);
+ sge->length = sg_dma_len(&recv->r_frag->f_sg);
ret = 0;
out:
@@ -362,6 +363,7 @@
static void release_refill(struct rds_connection *conn)
{
clear_bit(RDS_RECV_REFILL, &conn->c_flags);
+ smp_mb__after_atomic();
/* We don't use wait_on_bit()/wake_up_bit() because our waking is in a
* hot path and finding waiters is very rare. We don't want to walk
@@ -384,6 +386,7 @@
unsigned int posted = 0;
int ret = 0;
bool can_wait = !!(gfp & __GFP_DIRECT_RECLAIM);
+ bool must_wake = false;
u32 pos;
/* the goal here is to just make sure that someone, somewhere
@@ -404,14 +407,13 @@
recv = &ic->i_recvs[pos];
ret = rds_ib_recv_refill_one(conn, recv, gfp);
if (ret) {
+ must_wake = true;
break;
}
rdsdebug("recv %p ibinc %p page %p addr %lu\n", recv,
recv->r_ibinc, sg_page(&recv->r_frag->f_sg),
- (long) ib_sg_dma_address(
- ic->i_cm_id->device,
- &recv->r_frag->f_sg));
+ (long)sg_dma_address(&recv->r_frag->f_sg));
/* XXX when can this fail? */
ret = ib_post_recv(ic->i_cm_id->qp, &recv->r_wr, NULL);
@@ -424,6 +426,11 @@
}
posted++;
+
+ if ((posted > 128 && need_resched()) || posted > 8192) {
+ must_wake = true;
+ break;
+ }
}
/* We're doing flow control - update the window. */
@@ -446,10 +453,13 @@
* if we should requeue.
*/
if (rds_conn_up(conn) &&
- ((can_wait && rds_ib_ring_low(&ic->i_recv_ring)) ||
+ (must_wake ||
+ (can_wait && rds_ib_ring_low(&ic->i_recv_ring)) ||
rds_ib_ring_empty(&ic->i_recv_ring))) {
queue_delayed_work(rds_wq, &conn->c_recv_w, 1);
}
+ if (can_wait)
+ cond_resched();
}
/*
@@ -653,10 +663,16 @@
seq = rds_ib_get_ack(ic);
rdsdebug("send_ack: ic %p ack %llu\n", ic, (unsigned long long) seq);
+
+ ib_dma_sync_single_for_cpu(ic->rds_ibdev->dev, ic->i_ack_dma,
+ sizeof(*hdr), DMA_TO_DEVICE);
rds_message_populate_header(hdr, 0, 0, 0);
hdr->h_ack = cpu_to_be64(seq);
hdr->h_credit = adv_credits;
rds_message_make_checksum(hdr);
+ ib_dma_sync_single_for_device(ic->rds_ibdev->dev, ic->i_ack_dma,
+ sizeof(*hdr), DMA_TO_DEVICE);
+
ic->i_ack_queued = jiffies;
ret = ib_post_send(ic->i_cm_id->qp, &ic->i_ack_wr, NULL);
@@ -774,7 +790,7 @@
unsigned long frag_off;
unsigned long to_copy;
unsigned long copied;
- uint64_t uncongested = 0;
+ __le64 uncongested = 0;
void *addr;
/* catch completely corrupt packets */
@@ -791,7 +807,7 @@
copied = 0;
while (copied < RDS_CONG_MAP_BYTES) {
- uint64_t *src, *dst;
+ __le64 *src, *dst;
unsigned int k;
to_copy = min(RDS_FRAG_SIZE - frag_off, PAGE_SIZE - map_off);
@@ -826,9 +842,7 @@
}
/* the congestion map is in little endian order */
- uncongested = le64_to_cpu(uncongested);
-
- rds_cong_map_updated(map, uncongested);
+ rds_cong_map_updated(map, le64_to_cpu(uncongested));
}
static void rds_ib_process_recv(struct rds_connection *conn,
@@ -838,6 +852,7 @@
struct rds_ib_connection *ic = conn->c_transport_data;
struct rds_ib_incoming *ibinc = ic->i_ibinc;
struct rds_header *ihdr, *hdr;
+ dma_addr_t dma_addr = ic->i_recv_hdrs_dma[recv - ic->i_recvs];
/* XXX shut down the connection if port 0,0 are seen? */
@@ -854,8 +869,10 @@
}
data_len -= sizeof(struct rds_header);
- ihdr = &ic->i_recv_hdrs[recv - ic->i_recvs];
+ ihdr = ic->i_recv_hdrs[recv - ic->i_recvs];
+ ib_dma_sync_single_for_cpu(ic->rds_ibdev->dev, dma_addr,
+ sizeof(*ihdr), DMA_FROM_DEVICE);
/* Validate the checksum. */
if (!rds_message_verify_checksum(ihdr)) {
rds_ib_conn_error(conn, "incoming message "
@@ -863,7 +880,7 @@
"forcing a reconnect\n",
&conn->c_faddr);
rds_stats_inc(s_recv_drop_bad_checksum);
- return;
+ goto done;
}
/* Process the ACK sequence which comes with every packet */
@@ -892,7 +909,7 @@
*/
rds_ib_frag_free(ic, recv->r_frag);
recv->r_frag = NULL;
- return;
+ goto done;
}
/*
@@ -926,7 +943,7 @@
hdr->h_dport != ihdr->h_dport) {
rds_ib_conn_error(conn,
"fragment header mismatch; forcing reconnect\n");
- return;
+ goto done;
}
}
@@ -958,6 +975,9 @@
rds_inc_put(&ibinc->ii_inc);
}
+done:
+ ib_dma_sync_single_for_device(ic->rds_ibdev->dev, dma_addr,
+ sizeof(*ihdr), DMA_FROM_DEVICE);
}
void rds_ib_recv_cqe_handler(struct rds_ib_connection *ic,
@@ -986,10 +1006,11 @@
} else {
/* We expect errors as the qp is drained during shutdown */
if (rds_conn_up(conn) || rds_conn_connecting(conn))
- rds_ib_conn_error(conn, "recv completion on <%pI6c,%pI6c> had status %u (%s), disconnecting and reconnecting\n",
+ rds_ib_conn_error(conn, "recv completion on <%pI6c,%pI6c, %d> had status %u (%s), vendor err 0x%x, disconnecting and reconnecting\n",
&conn->c_laddr, &conn->c_faddr,
- wc->status,
- ib_wc_status_msg(wc->status));
+ conn->c_tos, wc->status,
+ ib_wc_status_msg(wc->status),
+ wc->vendor_err);
}
/* rds_ib_process_recv() doesn't always consume the frag, and
@@ -1012,7 +1033,7 @@
rds_ib_stats_inc(s_ib_rx_ring_empty);
if (rds_ib_ring_low(&ic->i_recv_ring)) {
- rds_ib_recv_refill(conn, 0, GFP_NOWAIT);
+ rds_ib_recv_refill(conn, 0, GFP_NOWAIT | __GFP_NOWARN);
rds_ib_stats_inc(s_ib_rx_refill_from_cq);
}
}
@@ -1041,9 +1062,14 @@
si_meminfo(&si);
rds_ib_sysctl_max_recv_allocation = si.totalram / 3 * PAGE_SIZE / RDS_FRAG_SIZE;
- rds_ib_incoming_slab = kmem_cache_create("rds_ib_incoming",
- sizeof(struct rds_ib_incoming),
- 0, SLAB_HWCACHE_ALIGN, NULL);
+ rds_ib_incoming_slab =
+ kmem_cache_create_usercopy("rds_ib_incoming",
+ sizeof(struct rds_ib_incoming),
+ 0, SLAB_HWCACHE_ALIGN,
+ offsetof(struct rds_ib_incoming,
+ ii_inc.i_usercopy),
+ sizeof(struct rds_inc_usercopy),
+ NULL);
if (!rds_ib_incoming_slab)
goto out;
@@ -1061,6 +1087,8 @@
void rds_ib_recv_exit(void)
{
+ WARN_ON(atomic_read(&rds_ib_allocation));
+
kmem_cache_destroy(rds_ib_incoming_slab);
kmem_cache_destroy(rds_ib_frag_slab);
}
--
Gitblit v1.6.2