Merge branch 'eth-bnxt-fix-several-bugs-in-the-bnxt-module'

Taehee Yoo says:

====================
eth: bnxt: fix several bugs in the bnxt module

The first fixes setting incorrect skb->truesize.
When xdp-mb prog returns XDP_PASS, skb is allocated and initialized.
Currently, The truesize is calculated as BNXT_RX_PAGE_SIZE *
sinfo->nr_frags, but sinfo->nr_frags is flushed by napi_build_skb().
So, it stores sinfo before calling napi_build_skb() and then use it
for calculate truesize.

The second fixes kernel panic in the bnxt_queue_mem_alloc().
The bnxt_queue_mem_alloc() accesses rx ring descriptor.
rx ring descriptors are allocated when the interface is up and it's
freed when the interface is down.
So, if bnxt_queue_mem_alloc() is called when the interface is down,
kernel panic occurs.
This patch makes the bnxt_queue_mem_alloc() return -ENETDOWN if rx ring
descriptors are not allocated.

The third patch fixes kernel panic in the bnxt_queue_{start | stop}().
When a queue is restarted bnxt_queue_{start | stop}() are called.
These functions set MRU to 0 to stop packet flow and then to set up the
remaining things.
MRU variable is a member of vnic_info[] the first vnic_info is for
default and the second is for ntuple.
The first vnic_info is always allocated when interface is up, but the
second is allocated only when ntuple is enabled.
(ethtool -K eth0 ntuple <on | off>).
Currently, the bnxt_queue_{start | stop}() access
vnic_info[BNXT_VNIC_NTUPLE] regardless of whether ntuple is enabled or
not.
So kernel panic occurs.
This patch make the bnxt_queue_{start | stop}() use bp->nr_vnics instead
of BNXT_VNIC_NTUPLE.

The fourth patch fixes a warning due to checksum state.
The bnxt_rx_pkt() checks whether skb->ip_summed is not CHECKSUM_NONE
before updating ip_summed. if ip_summed is not CHECKSUM_NONE, it WARNS
about it. However, the bnxt_xdp_build_skb() is called in XDP-MB-PASS
path and it updates ip_summed earlier than bnxt_rx_pkt().
So, in the XDP-MB-PASS path, the bnxt_rx_pkt() always warns about
checksum.
Updating ip_summed at the bnxt_xdp_build_skb() is unnecessary and
duplicate, so it is removed.

The fifth patch fixes a kernel panic in the
bnxt_get_queue_stats{rx | tx}().
The bnxt_get_queue_stats{rx | tx}() callback functions are called when
a queue is resetting.
These internally access rx and tx rings without null check, but rings
are allocated and initialized when the interface is up.
So, these functions are called when the interface is down, it
occurs a kernel panic.

The sixth patch fixes memory leak in queue reset logic.
When a queue is resetting, tpa_info is allocated for the new queue and
tpa_info for an old queue is not used anymore.
So it should be freed, but not.

The seventh patch makes net_devmem_unbind_dmabuf() ignore -ENETDOWN.
When devmem socket is closed, net_devmem_unbind_dmabuf() is called to
unbind/release resources.
If interface is down, the driver returns -ENETDOWN.
The -ENETDOWN return value is not an actual error, because the interface
will release resources when the interface is down.
So, net_devmem_unbind_dmabuf() needs to ignore -ENETDOWN.

The last patch adds XDP testcases to
tools/testing/selftests/drivers/net/ping.py.
====================

Link: https://patch.msgid.link/20250309134219.91670-1-ap420073@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

6 files changed, 220 insertions, 31 deletions

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index 7b8b5b39c7bb..55f553debd3b 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -2038,6 +2038,7 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
 	struct rx_cmp_ext *rxcmp1;
 	u32 tmp_raw_cons = *raw_cons;
 	u16 cons, prod, cp_cons = RING_CMP(tmp_raw_cons);
+	struct skb_shared_info *sinfo;
 	struct bnxt_sw_rx_bd *rx_buf;
 	unsigned int len;
 	u8 *data_ptr, agg_bufs, cmp_type;
@@ -2164,6 +2165,7 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
 							     false);
 			if (!frag_len)
 				goto oom_next_rx;
+
 		}
 		xdp_active = true;
 	}
@@ -2173,6 +2175,12 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
 			rc = 1;
 			goto next_rx;
 		}
+		if (xdp_buff_has_frags(&xdp)) {
+			sinfo = xdp_get_shared_info_from_buff(&xdp);
+			agg_bufs = sinfo->nr_frags;
+		} else {
+			agg_bufs = 0;
+		}
 	}
 
 	if (len <= bp->rx_copybreak) {
@@ -2210,7 +2218,8 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
 			if (!skb)
 				goto oom_next_rx;
 		} else {
-			skb = bnxt_xdp_build_skb(bp, skb, agg_bufs, rxr->page_pool, &xdp, rxcmp1);
+			skb = bnxt_xdp_build_skb(bp, skb, agg_bufs,
+						 rxr->page_pool, &xdp);
 			if (!skb) {
 				/* we should be able to free the old skb here */
 				bnxt_xdp_buff_frags_free(rxr, &xdp);
@@ -15375,6 +15384,9 @@ static void bnxt_get_queue_stats_rx(struct net_device *dev, int i,
 	struct bnxt_cp_ring_info *cpr;
 	u64 *sw;
 
+	if (!bp->bnapi)
+		return;
+
 	cpr = &bp->bnapi[i]->cp_ring;
 	sw = cpr->stats.sw_stats;
 
@@ -15398,6 +15410,9 @@ static void bnxt_get_queue_stats_tx(struct net_device *dev, int i,
 	struct bnxt_napi *bnapi;
 	u64 *sw;
 
+	if (!bp->tx_ring)
+		return;
+
 	bnapi = bp->tx_ring[bp->tx_ring_map[i]].bnapi;
 	sw = bnapi->cp_ring.stats.sw_stats;
 
@@ -15439,6 +15454,9 @@ static int bnxt_queue_mem_alloc(struct net_device *dev, void *qmem, int idx)
 	struct bnxt_ring_struct *ring;
 	int rc;
 
+	if (!bp->rx_ring)
+		return -ENETDOWN;
+
 	rxr = &bp->rx_ring[idx];
 	clone = qmem;
 	memcpy(clone, rxr, sizeof(*rxr));
@@ -15521,6 +15539,7 @@ static void bnxt_queue_mem_free(struct net_device *dev, void *qmem)
 	struct bnxt_ring_struct *ring;
 
 	bnxt_free_one_rx_ring_skbs(bp, rxr);
+	bnxt_free_one_tpa_info(bp, rxr);
 
 	xdp_rxq_info_unreg(&rxr->xdp_rxq);
 
@@ -15632,7 +15651,7 @@ static int bnxt_queue_start(struct net_device *dev, void *qmem, int idx)
 	cpr = &rxr->bnapi->cp_ring;
 	cpr->sw_stats->rx.rx_resets++;
 
-	for (i = 0; i <= BNXT_VNIC_NTUPLE; i++) {
+	for (i = 0; i <= bp->nr_vnics; i++) {
 		vnic = &bp->vnic_info[i];
 
 		rc = bnxt_hwrm_vnic_set_rss_p5(bp, vnic, true);
@@ -15660,7 +15679,7 @@ static int bnxt_queue_stop(struct net_device *dev, void *qmem, int idx)
 	struct bnxt_vnic_info *vnic;
 	int i;
 
-	for (i = 0; i <= BNXT_VNIC_NTUPLE; i++) {
+	for (i = 0; i <= bp->nr_vnics; i++) {
 		vnic = &bp->vnic_info[i];
 		vnic->mru = 0;
 		bnxt_hwrm_vnic_update(bp, vnic,
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
index e6c64e4bd66c..299822cacca4 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c
@@ -460,23 +460,16 @@ int bnxt_xdp(struct net_device *dev, struct netdev_bpf *xdp)
 
 struct sk_buff *
 bnxt_xdp_build_skb(struct bnxt *bp, struct sk_buff *skb, u8 num_frags,
-		   struct page_pool *pool, struct xdp_buff *xdp,
-		   struct rx_cmp_ext *rxcmp1)
+		   struct page_pool *pool, struct xdp_buff *xdp)
 {
 	struct skb_shared_info *sinfo = xdp_get_shared_info_from_buff(xdp);
 
 	if (!skb)
 		return NULL;
-	skb_checksum_none_assert(skb);
-	if (RX_CMP_L4_CS_OK(rxcmp1)) {
-		if (bp->dev->features & NETIF_F_RXCSUM) {
-			skb->ip_summed = CHECKSUM_UNNECESSARY;
-			skb->csum_level = RX_CMP_ENCAP(rxcmp1);
-		}
-	}
+
 	xdp_update_skb_shared_info(skb, num_frags,
 				   sinfo->xdp_frags_size,
-				   BNXT_RX_PAGE_SIZE * sinfo->nr_frags,
+				   BNXT_RX_PAGE_SIZE * num_frags,
 				   xdp_buff_is_frag_pfmemalloc(xdp));
 	return skb;
 }
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h
index 0122782400b8..220285e190fc 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h
@@ -33,6 +33,5 @@ void bnxt_xdp_buff_frags_free(struct bnxt_rx_ring_info *rxr,
 			      struct xdp_buff *xdp);
 struct sk_buff *bnxt_xdp_build_skb(struct bnxt *bp, struct sk_buff *skb,
 				   u8 num_frags, struct page_pool *pool,
-				   struct xdp_buff *xdp,
-				   struct rx_cmp_ext *rxcmp1);
+				   struct xdp_buff *xdp);
 #endif
diff --git a/net/core/devmem.c b/net/core/devmem.c
index 3bba3f018df0..0e5a2c672efd 100644
--- a/net/core/devmem.c
+++ b/net/core/devmem.c
@@ -109,6 +109,7 @@ void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding)
 	struct netdev_rx_queue *rxq;
 	unsigned long xa_idx;
 	unsigned int rxq_idx;
+	int err;
 
 	if (binding->list.next)
 		list_del(&binding->list);
@@ -120,7 +121,8 @@ void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding)
 
 		rxq_idx = get_netdev_rx_queue_index(rxq);
 
-		WARN_ON(netdev_rx_queue_restart(binding->dev, rxq_idx));
+		err = netdev_rx_queue_restart(binding->dev, rxq_idx);
+		WARN_ON(err && err != -ENETDOWN);
 	}
 
 	xa_erase(&net_devmem_dmabuf_bindings, binding->id);
diff --git a/tools/testing/selftests/drivers/net/ping.py b/tools/testing/selftests/drivers/net/ping.py
index eb83e7b48797..93f4b411b378 100755
--- a/tools/testing/selftests/drivers/net/ping.py
+++ b/tools/testing/selftests/drivers/net/ping.py
@@ -1,49 +1,219 @@
 #!/usr/bin/env python3
 # SPDX-License-Identifier: GPL-2.0
 
+import os
+import random, string, time
 from lib.py import ksft_run, ksft_exit
-from lib.py import ksft_eq
-from lib.py import NetDrvEpEnv
+from lib.py import ksft_eq, KsftSkipEx, KsftFailEx
+from lib.py import EthtoolFamily, NetDrvEpEnv
 from lib.py import bkg, cmd, wait_port_listen, rand_port
+from lib.py import ethtool, ip
 
+remote_ifname=""
+no_sleep=False
 
-def test_v4(cfg) -> None:
+def _test_v4(cfg) -> None:
     cfg.require_v4()
 
     cmd(f"ping -c 1 -W0.5 {cfg.remote_v4}")
     cmd(f"ping -c 1 -W0.5 {cfg.v4}", host=cfg.remote)
+    cmd(f"ping -s 65000 -c 1 -W0.5 {cfg.remote_v4}")
+    cmd(f"ping -s 65000 -c 1 -W0.5 {cfg.v4}", host=cfg.remote)
 
-
-def test_v6(cfg) -> None:
+def _test_v6(cfg) -> None:
     cfg.require_v6()
 
-    cmd(f"ping -c 1 -W0.5 {cfg.remote_v6}")
-    cmd(f"ping -c 1 -W0.5 {cfg.v6}", host=cfg.remote)
-
+    cmd(f"ping -c 1 -W5 {cfg.remote_v6}")
+    cmd(f"ping -c 1 -W5 {cfg.v6}", host=cfg.remote)
+    cmd(f"ping -s 65000 -c 1 -W0.5 {cfg.remote_v6}")
+    cmd(f"ping -s 65000 -c 1 -W0.5 {cfg.v6}", host=cfg.remote)
 
-def test_tcp(cfg) -> None:
+def _test_tcp(cfg) -> None:
     cfg.require_cmd("socat", remote=True)
 
     port = rand_port()
     listen_cmd = f"socat -{cfg.addr_ipver} -t 2 -u TCP-LISTEN:{port},reuseport STDOUT"
 
+    test_string = ''.join(random.choice(string.ascii_lowercase) for _ in range(65536))
     with bkg(listen_cmd, exit_wait=True) as nc:
         wait_port_listen(port)
 
-        cmd(f"echo ping | socat -t 2 -u STDIN TCP:{cfg.baddr}:{port}",
+        cmd(f"echo {test_string} | socat -t 2 -u STDIN TCP:{cfg.baddr}:{port}",
             shell=True, host=cfg.remote)
-    ksft_eq(nc.stdout.strip(), "ping")
+    ksft_eq(nc.stdout.strip(), test_string)
 
+    test_string = ''.join(random.choice(string.ascii_lowercase) for _ in range(65536))
     with bkg(listen_cmd, host=cfg.remote, exit_wait=True) as nc:
         wait_port_listen(port, host=cfg.remote)
 
-        cmd(f"echo ping | socat -t 2 -u STDIN TCP:{cfg.remote_baddr}:{port}", shell=True)
-    ksft_eq(nc.stdout.strip(), "ping")
-
+        cmd(f"echo {test_string} | socat -t 2 -u STDIN TCP:{cfg.remote_baddr}:{port}", shell=True)
+    ksft_eq(nc.stdout.strip(), test_string)
+
+def _set_offload_checksum(cfg, netnl, on) -> None:
+    try:
+        ethtool(f" -K {cfg.ifname} rx {on} tx {on} ")
+    except:
+        return
+
+def _set_xdp_generic_sb_on(cfg) -> None:
+    test_dir = os.path.dirname(os.path.realpath(__file__))
+    prog = test_dir + "/../../net/lib/xdp_dummy.bpf.o"
+    cmd(f"ip link set dev {remote_ifname} mtu 1500", shell=True, host=cfg.remote)
+    cmd(f"ip link set dev {cfg.ifname} mtu 1500 xdpgeneric obj {prog} sec xdp", shell=True)
+
+    if no_sleep != True:
+        time.sleep(10)
+
+def _set_xdp_generic_mb_on(cfg) -> None:
+    test_dir = os.path.dirname(os.path.realpath(__file__))
+    prog = test_dir + "/../../net/lib/xdp_dummy.bpf.o"
+    cmd(f"ip link set dev {remote_ifname} mtu 9000", shell=True, host=cfg.remote)
+    ip("link set dev %s mtu 9000 xdpgeneric obj %s sec xdp.frags" % (cfg.ifname, prog))
+
+    if no_sleep != True:
+        time.sleep(10)
+
+def _set_xdp_native_sb_on(cfg) -> None:
+    test_dir = os.path.dirname(os.path.realpath(__file__))
+    prog = test_dir + "/../../net/lib/xdp_dummy.bpf.o"
+    cmd(f"ip link set dev {remote_ifname} mtu 1500", shell=True, host=cfg.remote)
+    cmd(f"ip -j link set dev {cfg.ifname} mtu 1500 xdp obj {prog} sec xdp", shell=True)
+    xdp_info = ip("-d link show %s" % (cfg.ifname), json=True)[0]
+    if xdp_info['xdp']['mode'] != 1:
+        """
+        If the interface doesn't support native-mode, it falls back to generic mode.
+        The mode value 1 is native and 2 is generic.
+        So it raises an exception if mode is not 1(native mode).
+        """
+        raise KsftSkipEx('device does not support native-XDP')
+
+    if no_sleep != True:
+        time.sleep(10)
+
+def _set_xdp_native_mb_on(cfg) -> None:
+    test_dir = os.path.dirname(os.path.realpath(__file__))
+    prog = test_dir + "/../../net/lib/xdp_dummy.bpf.o"
+    cmd(f"ip link set dev {remote_ifname} mtu 9000", shell=True, host=cfg.remote)
+    try:
+        cmd(f"ip link set dev {cfg.ifname} mtu 9000 xdp obj {prog} sec xdp.frags", shell=True)
+    except Exception as e:
+        cmd(f"ip link set dev {remote_ifname} mtu 1500", shell=True, host=cfg.remote)
+        raise KsftSkipEx('device does not support native-multi-buffer XDP')
+
+    if no_sleep != True:
+        time.sleep(10)
+
+def _set_xdp_offload_on(cfg) -> None:
+    test_dir = os.path.dirname(os.path.realpath(__file__))
+    prog = test_dir + "/../../net/lib/xdp_dummy.bpf.o"
+    cmd(f"ip link set dev {cfg.ifname} mtu 1500", shell=True)
+    try:
+        cmd(f"ip link set dev {cfg.ifname} xdpoffload obj {prog} sec xdp", shell=True)
+    except Exception as e:
+        raise KsftSkipEx('device does not support offloaded XDP')
+    cmd(f"ip link set dev {remote_ifname} mtu 1500", shell=True, host=cfg.remote)
+
+    if no_sleep != True:
+        time.sleep(10)
+
+def get_interface_info(cfg) -> None:
+    global remote_ifname
+    global no_sleep
+
+    remote_info = cmd(f"ip -4 -o addr show to {cfg.remote_v4} | awk '{{print $2}}'", shell=True, host=cfg.remote).stdout
+    remote_ifname = remote_info.rstrip('\n')
+    if remote_ifname == "":
+        raise KsftFailEx('Can not get remote interface')
+    local_info = ip("-d link show %s" % (cfg.ifname), json=True)[0]
+    if 'parentbus' in local_info and local_info['parentbus'] == "netdevsim":
+        no_sleep=True
+    if 'linkinfo' in local_info and local_info['linkinfo']['info_kind'] == "veth":
+        no_sleep=True
+
+def set_interface_init(cfg) -> None:
+    cmd(f"ip link set dev {cfg.ifname} mtu 1500", shell=True)
+    cmd(f"ip link set dev {cfg.ifname} xdp off ", shell=True)
+    cmd(f"ip link set dev {cfg.ifname} xdpgeneric off ", shell=True)
+    cmd(f"ip link set dev {cfg.ifname} xdpoffload off", shell=True)
+    cmd(f"ip link set dev {remote_ifname} mtu 1500", shell=True, host=cfg.remote)
+
+def test_default(cfg, netnl) -> None:
+    _set_offload_checksum(cfg, netnl, "off")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    _set_offload_checksum(cfg, netnl, "on")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+
+def test_xdp_generic_sb(cfg, netnl) -> None:
+    _set_xdp_generic_sb_on(cfg)
+    _set_offload_checksum(cfg, netnl, "off")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    _set_offload_checksum(cfg, netnl, "on")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    ip("link set dev %s xdpgeneric off" % cfg.ifname)
+
+def test_xdp_generic_mb(cfg, netnl) -> None:
+    _set_xdp_generic_mb_on(cfg)
+    _set_offload_checksum(cfg, netnl, "off")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    _set_offload_checksum(cfg, netnl, "on")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    ip("link set dev %s xdpgeneric off" % cfg.ifname)
+
+def test_xdp_native_sb(cfg, netnl) -> None:
+    _set_xdp_native_sb_on(cfg)
+    _set_offload_checksum(cfg, netnl, "off")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    _set_offload_checksum(cfg, netnl, "on")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    ip("link set dev %s xdp off" % cfg.ifname)
+
+def test_xdp_native_mb(cfg, netnl) -> None:
+    _set_xdp_native_mb_on(cfg)
+    _set_offload_checksum(cfg, netnl, "off")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    _set_offload_checksum(cfg, netnl, "on")
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    ip("link set dev %s xdp off" % cfg.ifname)
+
+def test_xdp_offload(cfg, netnl) -> None:
+    _set_xdp_offload_on(cfg)
+    _test_v4(cfg)
+    _test_v6(cfg)
+    _test_tcp(cfg)
+    ip("link set dev %s xdpoffload off" % cfg.ifname)
 
 def main() -> None:
     with NetDrvEpEnv(__file__) as cfg:
-        ksft_run(globs=globals(), case_pfx={"test_"}, args=(cfg, ))
+        get_interface_info(cfg)
+        set_interface_init(cfg)
+        ksft_run([test_default,
+                  test_xdp_generic_sb,
+                  test_xdp_generic_mb,
+                  test_xdp_native_sb,
+                  test_xdp_native_mb,
+                  test_xdp_offload],
+                 args=(cfg, EthtoolFamily()))
+        set_interface_init(cfg)
     ksft_exit()
 
 
diff --git a/tools/testing/selftests/net/lib/xdp_dummy.bpf.c b/tools/testing/selftests/net/lib/xdp_dummy.bpf.c
index d988b2e0cee8..e73fab3edd9f 100644
--- a/tools/testing/selftests/net/lib/xdp_dummy.bpf.c
+++ b/tools/testing/selftests/net/lib/xdp_dummy.bpf.c
@@ -10,4 +10,10 @@ int xdp_dummy_prog(struct xdp_md *ctx)
 	return XDP_PASS;
 }
 
+SEC("xdp.frags")
+int xdp_dummy_prog_frags(struct xdp_md *ctx)
+{
+	return XDP_PASS;
+}
+
 char _license[] SEC("license") = "GPL";