Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
MIME-Version: 1.0
Date:   Wed, 09 Feb 2022 07:27:44 +0000
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
From:   yajun.deng@linux.dev
Message-ID: <e95e51e3cdf771650ae64d5295e1178a@linux.dev>
Subject: Re: [PATCH net-next] net: dev: introduce netdev_drop_inc()
To:     "Jakub Kicinski" <kuba@kernel.org>
Cc:     davem@davemloft.net, rostedt@goodmis.org, mingo@redhat.com,
        netdev@vger.kernel.org, linux-kernel@vger.kernel.org
In-Reply-To: <20220208195306.05a1760f@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
References: <20220208195306.05a1760f@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
 <20220208162729.41b62ae7@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
 <20220208064318.1075849-1-yajun.deng@linux.dev>
 <753bb02bfa8c2cf5c08c63c31f193f90@linux.dev>
Precedence: bulk

February 9, 2022 11:53 AM, "Jakub Kicinski" <kuba@kernel.org> wrote:=0A=
=0A> On Wed, 09 Feb 2022 02:20:07 +0000 yajun.deng@linux.dev wrote:=0A> =
=0A>> February 9, 2022 8:27 AM, "Jakub Kicinski" <kuba@kernel.org> wrote:=
=0A>> =0A>> On Tue, 8 Feb 2022 14:43:18 +0800 Yajun Deng wrote:=0A>> =0A>=
> We will use 'sudo perf record -g -a -e skb:kfree_skb' command to trace=
=0A>> the dropped packets when dropped increase in the output of ifconfig=
.=0A>> But there are two cases, one is only called kfree_skb(), another i=
s=0A>> increasing the dropped and called kfree_skb(). The latter is what=
=0A>> we need. So we need to separate these two cases.=0A>> =0A>> From th=
e other side, the dropped packet came from the core network and=0A>> the =
driver, we also need to separate these two cases.=0A>> =0A>> Add netdev_d=
rop_inc() and add a tracepoint for the core network dropped=0A>> packets.=
 use 'sudo perf record -g -a -e net:netdev_drop' and 'sudo perf=0A>> scri=
pt' will recored the dropped packets by the core network.=0A>> =0A>> Sign=
ed-off-by: Yajun Deng <yajun.deng@linux.dev>=0A>> =0A>> Have you seen the=
 work that's being done around kfree_skb_reason()?=0A>> =0A>> Yes, I saw =
it. The focus of kfree_skb_reason() is trace kfree_skb() and the reason,=
=0A>> but the focus of this patch only traces this case of the dropped pa=
cket.=0A>> =0A>> I don't want to trace all kfree_skb(), but I just want t=
o trace the dropped packet.=0A>> =0A>> This command 'sudo perf record -g =
-a -e skb:kfree_skb' would trace all kfree_skb(),=0A>> kfree_skb() would =
drowned out the case of dropped packets when the samples were too large.=
=0A> =0A> IIRC perf support filters, I think with -f? We can't add a trac=
epoint=0A> for every combination of attributes.=0A=0AYes, we can use a co=
mmand like this: " sudo perf record -g -a -e skb:kfree_skb --filter 'prot=
ocol =3D=3D 0x0800' ",=0AHowever, only the filter is defined in kfree_skb=
 tracepoint are available.=0A=0AThe purpose of this patch is record {rx_d=
ropped, tx_dropped, rx_nohandler} in struct net_device, to distinguish =
=0Awith struct net_device_stats. =0A=0AWe don't have any tracepoint recor=
ds {rx_dropped, tx_dropped, rx_nohandler} in struct net_device now. =0ACa=
n we add {rx_dropped, tx_dropped, rx_nohandler} in kfree_skb tracepoint? =
 like this:=0A=0A        TP_STRUCT__entry(=0A                __field(void=
 *,         skbaddr)=0A                __field(void *,         location)=
=0A                __field(unsigned short, protocol)=0A                __=
field(enum skb_drop_reason,   reason)=0A                __field(unsigned =
long,  rx_dropped)=0A                __field(unsigned long,  tx_dropped)=
=0A                __field(unsigned long,  rx_nohandler)=0A=0A        ),=
=0A=0A        TP_fast_assign(=0A                __entry->skbaddr =3D skb;=
=0A                __entry->location =3D location;=0A                __en=
try->protocol =3D ntohs(skb->protocol);=0A                __entry->reason=
 =3D reason;=0A                __entry->rx_dropped   =3D (unsigned long)a=
tomic_long_read(&skb->dev->rx_dropped);=0A                __entry->tx_dro=
pped   =3D (unsigned long)atomic_long_read(&skb->dev->tx_dropped);=0A    =
            __entry->rx_nohandler =3D (unsigned long)atomic_long_read(&sk=
b->dev->rx_nohandler);=0A        ),=0A=0AIf so, we can record this but no=
t add a tracepoint.