lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Date: Tue, 11 Jul 2023 19:07:43 +0800
From: menglong8.dong@...il.com
To: michael.chan@...adcom.com,
	leon@...nel.org
Cc: davem@...emloft.net,
	edumazet@...gle.com,
	kuba@...nel.org,
	pabeni@...hat.com,
	netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org,
	Menglong Dong <imagedong@...cent.com>,
	Leon Romanovsky <leonro@...dia.com>
Subject: [PATCH net-next v2] bnxt_en: use dev_consume_skb_any() in bnxt_tx_int

From: Menglong Dong <imagedong@...cent.com>

In bnxt_tx_int(), the skb in the tx ring buffer will be freed after the
transmission completes with dev_kfree_skb_any(), which will produce
the noise on the tracepoint "skb:kfree_skb":

$ perf script record -e skb:kfree_skb -a
$ perf script
  swapper     0 [014] 12814.337522: skb:kfree_skb: skbaddr=0xffff88818f145ce0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [003] 12814.338318: skb:kfree_skb: skbaddr=0xffff888108380600 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12814.375258: skb:kfree_skb: skbaddr=0xffff88818f147ce0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12814.451960: skb:kfree_skb: skbaddr=0xffff88818f145ce0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [008] 12814.562166: skb:kfree_skb: skbaddr=0xffff888112664600 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12814.732517: skb:kfree_skb: skbaddr=0xffff88818f145ce0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12814.800608: skb:kfree_skb: skbaddr=0xffff88810025d100 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12814.861501: skb:kfree_skb: skbaddr=0xffff888108295a00 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12815.377038: skb:kfree_skb: skbaddr=0xffff88818f147ce0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
  swapper     0 [014] 12815.395530: skb:kfree_skb: skbaddr=0xffff88818f145ee0 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED

And the call stack is like this:

$ perf script record -e skb:kfree_skb -a --call-graph fp
$ perf script
swapper     0 [015] 12915.386236: skb:kfree_skb: skbaddr=0xffff88b0473cd000 protocol=2048 location=dev_kfree_skb_any_reason+0x2e reason: NOT_SPECIFIED
        ffffffff81e5b0a7 kfree_skb_reason+0x117 (vmlinux)
        ffffffff81e5b0a7 kfree_skb_reason+0x117 (vmlinux)
        ffffffff81e6dd1e dev_kfree_skb_any_reason+0x2e (vmlinux)
        ffffffffa04510d5 bnxt_tx_int+0x95 ([bnxt_en])
        ffffffffa0454ee5 __bnxt_poll_work_done+0x25 ([bnxt_en])
        ffffffffa045b9d2 bnxt_poll+0x72 ([bnxt_en])
        ffffffff81e7b47e __napi_poll+0x2e (vmlinux)
        ffffffff81e7bb34 net_rx_action+0x294 (vmlinux)
        ffffffff8218ff5e __do_softirq+0xfe (vmlinux)
        ffffffff810ad51e irq_exit_rcu+0x7e (vmlinux)
        ffffffff82175bb5 common_interrupt+0xc5 (vmlinux)
        ffffffff82200d27 asm_common_interrupt+0x27 (vmlinux)
        ffffffff8217ae54 intel_idle_irq+0x64 (vmlinux)
        ffffffff81dea7ae cpuidle_enter+0x2e (vmlinux)
        ffffffff811057e3 call_cpuidle+0x23 (vmlinux)
        ffffffff8110b16a do_idle+0x1ea (vmlinux)
        ffffffff8110b3cd cpu_startup_entry+0x1d (vmlinux)
        ffffffff8106dc08 start_secondary+0x118 (vmlinux)
        ffffffff81000263 secondary_startup_64_no_verify+0x17e (vmlinux)

Replace dev_kfree_skb_any() with dev_consume_skb_any() in bnxt_tx_int()
to reduce the noise.

Signed-off-by: Menglong Dong <imagedong@...cent.com>
Reviewed-by: Leon Romanovsky <leonro@...dia.com>
---
v2:
- be more specific in the commit message
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index e5b54e6025be..d84ded8db93d 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -685,7 +685,7 @@ static void bnxt_tx_int(struct bnxt *bp, struct bnxt_napi *bnapi, int nr_pkts)
 next_tx_int:
 		cons = NEXT_TX(cons);
 
-		dev_kfree_skb_any(skb);
+		dev_consume_skb_any(skb);
 	}
 
 	WRITE_ONCE(txr->tx_cons, cons);
-- 
2.40.1


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ