* [PATCH 60/64] drivers: net: uli526x: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/dec/tulip/uli526x.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/dec/tulip/uli526x.c b/drivers/net/ethernet/dec/tulip/uli526x.c
index 7fc248e..5fbbc0c 100644
--- a/drivers/net/ethernet/dec/tulip/uli526x.c
+++ b/drivers/net/ethernet/dec/tulip/uli526x.c
@@ -491,10 +491,8 @@ static int uli526x_open(struct net_device *dev)
netif_wake_queue(dev);
/* set and active a timer process */
- init_timer(&db->timer);
+ setup_timer(&db->timer, uli526x_timer, (unsigned long)dev);
db->timer.expires = ULI526X_TIMER_WUT + HZ * 2;
- db->timer.data = (unsigned long)dev;
- db->timer.function = uli526x_timer;
add_timer(&db->timer);
return 0;
--
2.7.4
^ permalink raw reply related
* [PATCH 61/64] drivers: net: enic: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/cisco/enic/enic_clsf.h | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/cisco/enic/enic_clsf.h b/drivers/net/ethernet/cisco/enic/enic_clsf.h
index 6aa9f89..4bfbf25 100644
--- a/drivers/net/ethernet/cisco/enic/enic_clsf.h
+++ b/drivers/net/ethernet/cisco/enic/enic_clsf.h
@@ -19,9 +19,8 @@ void enic_flow_may_expire(unsigned long data);
static inline void enic_rfs_timer_start(struct enic *enic)
{
- init_timer(&enic->rfs_h.rfs_may_expire);
- enic->rfs_h.rfs_may_expire.function = enic_flow_may_expire;
- enic->rfs_h.rfs_may_expire.data = (unsigned long)enic;
+ setup_timer(&enic->rfs_h.rfs_may_expire, enic_flow_may_expire,
+ (unsigned long)enic);
mod_timer(&enic->rfs_h.rfs_may_expire, jiffies + HZ/4);
}
--
2.7.4
^ permalink raw reply related
* [PATCH 62/64] drivers: net: cxgb: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/chelsio/cxgb/sge.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb/sge.c b/drivers/net/ethernet/chelsio/cxgb/sge.c
index 0f13a7f..75e4399 100644
--- a/drivers/net/ethernet/chelsio/cxgb/sge.c
+++ b/drivers/net/ethernet/chelsio/cxgb/sge.c
@@ -2075,9 +2075,8 @@ struct sge *t1_sge_create(struct adapter *adapter, struct sge_params *p)
goto nomem_port;
}
- init_timer(&sge->tx_reclaim_timer);
- sge->tx_reclaim_timer.data = (unsigned long)sge;
- sge->tx_reclaim_timer.function = sge_tx_reclaim_cb;
+ setup_timer(&sge->tx_reclaim_timer, sge_tx_reclaim_cb,
+ (unsigned long)sge);
if (is_T2(sge->adapter)) {
init_timer(&sge->espibug_timer);
--
2.7.4
^ permalink raw reply related
* [PATCH 63/64] drivers: net: bnx2x: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
index c12b4d3..54d15713 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
@@ -12414,10 +12414,8 @@ static int bnx2x_init_bp(struct bnx2x *bp)
bp->current_interval = CHIP_REV_IS_SLOW(bp) ? 5*HZ : HZ;
- init_timer(&bp->timer);
+ setup_timer(&bp->timer, bnx2x_timer, (unsigned long)bp);
bp->timer.expires = jiffies + bp->current_interval;
- bp->timer.data = (unsigned long) bp;
- bp->timer.function = bnx2x_timer;
if (SHMEM2_HAS(bp, dcbx_lldp_params_offset) &&
SHMEM2_HAS(bp, dcbx_lldp_dcbx_stat_offset) &&
--
2.7.4
^ permalink raw reply related
* [PATCH 64/64] drivers: net: lmc: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/wan/lmc/lmc_main.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/net/wan/lmc/lmc_main.c b/drivers/net/wan/lmc/lmc_main.c
index 4698450..ae69d65 100644
--- a/drivers/net/wan/lmc/lmc_main.c
+++ b/drivers/net/wan/lmc/lmc_main.c
@@ -1084,10 +1084,8 @@ static int lmc_open(struct net_device *dev)
* Setup a timer for the watchdog on probe, and start it running.
* Since lmc_ok == 0, it will be a NOP for now.
*/
- init_timer (&sc->timer);
+ setup_timer(&sc->timer, lmc_watchdog, (unsigned long)dev);
sc->timer.expires = jiffies + HZ;
- sc->timer.data = (unsigned long) dev;
- sc->timer.function = lmc_watchdog;
add_timer (&sc->timer);
lmc_trace(dev, "lmc_open out");
--
2.7.4
^ permalink raw reply related
* [PATCH 06/64] drivers: net: bcm63xx: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:04 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/broadcom/bcm63xx_enet.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bcm63xx_enet.c b/drivers/net/ethernet/broadcom/bcm63xx_enet.c
index 4f3845a..f8bbbbf 100644
--- a/drivers/net/ethernet/broadcom/bcm63xx_enet.c
+++ b/drivers/net/ethernet/broadcom/bcm63xx_enet.c
@@ -1857,9 +1857,8 @@ static int bcm_enet_probe(struct platform_device *pdev)
spin_lock_init(&priv->rx_lock);
/* init rx timeout (used for oom) */
- init_timer(&priv->rx_timeout);
- priv->rx_timeout.function = bcm_enet_refill_rx_timer;
- priv->rx_timeout.data = (unsigned long)dev;
+ setup_timer(&priv->rx_timeout, bcm_enet_refill_rx_timer,
+ (unsigned long)dev);
/* init the mib update lock&work */
mutex_init(&priv->mib_update_lock);
--
2.7.4
^ permalink raw reply related
* [PATCH 08/64] drivers: net: am79c961: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:04 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/amd/am79c961a.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/amd/am79c961a.c b/drivers/net/ethernet/amd/am79c961a.c
index b11e910..0612dbe 100644
--- a/drivers/net/ethernet/amd/am79c961a.c
+++ b/drivers/net/ethernet/amd/am79c961a.c
@@ -728,9 +728,7 @@ static int am79c961_probe(struct platform_device *pdev)
am79c961_banner();
spin_lock_init(&priv->chip_lock);
- init_timer(&priv->timer);
- priv->timer.data = (unsigned long)dev;
- priv->timer.function = am79c961_timer;
+ setup_timer(&priv->timer, am79c961_timer, (unsigned long)dev);
if (am79c961_hw_init(dev))
goto release;
--
2.7.4
^ permalink raw reply related
* Re: [PATCH net 1/3] net: mvpp2: fix the dma_mask and coherent_dma_mask settings for PPv2.2
From: David Miller @ 2017-09-21 17:07 UTC (permalink / raw)
To: antoine.tenart
Cc: andrew, gregory.clement, thomas.petazzoni, miquel.raynal, nadavh,
linux, linux-kernel, mw, stefanc, netdev
In-Reply-To: <20170921142413.GH3568@kwain>
From: Antoine Tenart <antoine.tenart@free-electrons.com>
Date: Thu, 21 Sep 2017 16:24:13 +0200
> That's also the default when the platform does not allocate dma_mask.
That's the problem that needs to be fixed then.
^ permalink raw reply
* Re: [REGRESSION] Warning in tcp_fastretrans_alert() of net/ipv4/tcp_input.c
From: Yuchung Cheng @ 2017-09-21 17:07 UTC (permalink / raw)
To: 10035198.1vE6NFrMDO
Cc: Oleksandr Natalenko, Alexey Kuznetsov, Hideaki YOSHIFUJI, netdev,
linux-kernel@vger.kernel.org
In-Reply-To: <20170921014620.GA20906@castle>
On Wed, Sep 20, 2017 at 6:46 PM, Roman Gushchin <guro@fb.com> wrote:
>
> > Hello.
> >
> > Since, IIRC, v4.11, there is some regression in TCP stack resulting in the
> > warning shown below. Most of the time it is harmless, but rarely it just
> > causes either freeze or (I believe, this is related too) panic in
> > tcp_sacktag_walk() (because sk_buff passed to this function is NULL).
> > Unfortunately, I still do not have proper stacktrace from panic, but will try
> > to capture it if possible.
> >
> > Also, I have custom settings regarding TCP stack, shown below as well. ifb is
> > used to shape traffic with tc.
> >
> > Please note this regression was already reported as BZ [1] and as a letter to
> > ML [2], but got neither attention nor resolution. It is reproducible for (not
> > only) me on my home router since v4.11 till v4.13.1 incl.
> >
> > Please advise on how to deal with it. I'll provide any additional info if
> > necessary, also ready to test patches if any.
> >
> > Thanks.
> >
> > [1] https://bugzilla.kernel.org/show_bug.cgi?id=195835
> > [2] https://www.spinics.net/lists/netdev/msg436158.html
>
> We're experiencing the same problems on some machines in our fleet.
> Exactly the same symptoms: tcp_fastretrans_alert() warnings and
> sometimes panics in tcp_sacktag_walk().
>
> Here is an example of a backtrace with the panic log:
do you still see the panics if you disable RACK?
sysctl net.ipv4.tcp_recovery=0?
also have you experience any sack reneg? could you post the output of
' nstat |grep -i TCP' thanks
>
> 978.210080] fuse
> [973978.214099] sg
> [973978.217789] loop
> [973978.221829] efivarfs
> [973978.226544] autofs4
> [973978.231109] CPU: 12 PID: 3806320 Comm: ld:srv:W20 Tainted: G W 4.11.3-7_fbk1_1174_ga56eebf #7
> [973978.250563] Hardware name: Wiwynn Leopard-Orv2/Leopard-DDR BW, BIOS LBM06 10/28/2016
> [973978.266558] Call Trace:
> [973978.271615] <IRQ>
> [973978.275817] dump_stack+0x4d/0x70
> [973978.282626] __warn+0xd3/0xf0
> [973978.288727] warn_slowpath_null+0x1e/0x20
> [973978.296910] tcp_fastretrans_alert+0xacf/0xbd0
> [973978.305974] tcp_ack+0xbce/0x1390
> [973978.312770] tcp_rcv_established+0x1ce/0x740
> [973978.321488] tcp_v6_do_rcv+0x195/0x440
> [973978.329166] tcp_v6_rcv+0x94c/0x9f0
> [973978.336323] ip6_input_finish+0xea/0x430
> [973978.344330] ip6_input+0x32/0xa0
> [973978.350968] ? ip6_rcv_finish+0xa0/0xa0
> [973978.358799] ip6_rcv_finish+0x4b/0xa0
> [973978.366289] ipv6_rcv+0x2ec/0x4f0
> [973978.373082] ? ip6_make_skb+0x1c0/0x1c0
> [973978.380919] __netif_receive_skb_core+0x2d5/0x9a0
> [973978.390505] __netif_receive_skb+0x16/0x70
> [973978.398875] netif_receive_skb_internal+0x23/0x80
> [973978.408462] napi_gro_receive+0x113/0x1d0
> [973978.416657] mlx5e_handle_rx_cqe_mpwrq+0x5b6/0x8d0
> [973978.426398] mlx5e_poll_rx_cq+0x7c/0x7f0
> [973978.434405] mlx5e_napi_poll+0x8c/0x380
> [973978.442238] ? mlx5_cq_completion+0x54/0xb0
> [973978.450770] net_rx_action+0x22e/0x380
> [973978.458447] __do_softirq+0x106/0x2e8
> [973978.465950] irq_exit+0xb0/0xc0
> [973978.472396] do_IRQ+0x4f/0xd0
> [973978.478495] common_interrupt+0x86/0x86
> [973978.486329] RIP: 0033:0x7f8dee58d8ae
> [973978.493642] RSP: 002b:00007f8cb925f078 EFLAGS: 00000206
> [973978.504251] ORIG_RAX: ffffffffffffff5f
> [973978.512085] RAX: 00007f8cb925f1a8 RBX: 0000000048000000 RCX: 00007f8764bd6a80
> [973978.526508] RDX: 00000000000001ba RSI: 00007f7cb4ca3410 RDI: 00007f7cb4ca3410
> [973978.540927] RBP: 000000000000000d R08: 00007f8764bd6b00 R09: 00007f8764bd6b80
> [973978.555347] R10: 0000000000002400 R11: 00007f8dee58e240 R12: d3273c84146e8c29
> [973978.569766] R13: 9dac83ddf04c235c R14: 00007f7cb4ca33b0 R15: 00007f7cb4ca4f50
> [973978.584189] </IRQ>
> [973978.588650] ---[ end trace 5d1c83e12a57d039 ]---
> [973995.178183] BUG: unable to handle kernel
> [973995.186385] NULL pointer dereference
> [973995.193692] at 0000000000000028
> [973995.200323] IP: tcp_sacktag_walk+0x2b7/0x460
> [973995.209032] PGD 102d856067
> [973995.214789] PUD fded0d067
> [973995.220385] PMD 0
> [973995.224577]
> [973995.227732] ------------[ cut here ]------------
> [973995.237128] Oops: 0000 [#1] SMP
> [973995.243575] Modules linked in:
> [973995.249868] mptctl
> [973995.254251] mptbase
> [973995.258792] xt_DSCP
> [973995.263331] xt_set
> [973995.267698] ip_set_hash_ip
> [973995.273452] cls_u32
> [973995.277993] sch_sfq
> [973995.282535] cls_fw
> [973995.286904] sch_htb
> [973995.291444] mpt3sas
> [973995.295982] raid_class
> [973995.301044] ip6table_mangle
> [973995.306973] iptable_mangle
> [973995.312726] cls_bpf
> [973995.317268] tcp_diag
> [973995.321983] udp_diag
> [973995.326697] inet_diag
> [973995.331585] ip6table_filter
> [973995.337513] xt_NFLOG
> [973995.342226] nfnetlink_log
> [973995.347807] xt_comment
> [973995.352866] xt_statistic
> [973995.358276] iptable_filter
> [973995.364029] xt_mark
> [973995.368572] sb_edac
> [973995.373113] edac_core
> [973995.378001] x86_pkg_temp_thermal
> [973995.384795] intel_powerclamp
> [973995.390897] coretemp
> [973995.395608] kvm_intel
> [973995.400498] kvm
> [973995.404345] irqbypass
> [973995.409235] ses
> [973995.413078] iTCO_wdt
> [973995.417794] iTCO_vendor_support
> [973995.424415] enclosure
> [973995.429301] lpc_ich
> [973995.433843] scsi_transport_sas
> [973995.440292] mfd_core
> [973995.445007] efivars
> [973995.449548] ipmi_si
> [973995.454087] ipmi_devintf
> [973995.459496] i2c_i801
> [973995.464209] ipmi_msghandler
> [973995.470138] acpi_cpufreq
> [973995.475545] button
> [973995.479914] sch_fq_codel
> [973995.485319] nfsd
> [973995.489341] auth_rpcgss
> [973995.494573] nfs_acl
> [973995.499114] oid_registry
> [973995.504524] lockd
> [973995.508717] grace
> [973995.512912] sunrpc
> [973995.517280] megaraid_sas
> [973995.522689] fuse
> [973995.526709] sg
> [973995.530382] loop
> [973995.534405] efivarfs
> [973995.539118] autofs4
> [973995.543660] CPU: 19 PID: 3806297 Comm: ld:srv:W0 Tainted: G W 4.11.3-7_fbk1_1174_ga56eebf #7
> [973995.562936] Hardware name: Wiwynn Leopard-Orv2/Leopard-DDR BW, BIOS LBM06 10/28/2016
> [973995.578914] task: ffff880129e5c380 task.stack: ffffc900210cc000
> [973995.590914] RIP: 0010:tcp_sacktag_walk+0x2b7/0x460
> [973995.600648] RSP: 0000:ffff88203ef438e8 EFLAGS: 00010207
> [973995.611254] RAX: 0000000000000001 RBX: 0000000000000000 RCX: 000000004e4ec474
> [973995.625677] RDX: 000000004e4ec2ad RSI: ffff8810361faa00 RDI: ffff881ffecf8840
> [973995.640102] RBP: ffff88203ef43940 R08: 0000000045729921 R09: 0000000000000000
> [973995.654519] R10: 00000000000085d6 R11: ffff881ffecf8998 R12: ffff881ffecf8840
> [973995.668938] R13: ffff88203ef43a48 R14: 0000000000000000 R15: ffff881ffecf8998
> [973995.683362] FS: 00007f8cc8cf7700(0000) GS:ffff88203ef40000(0000) knlGS:0000000000000000
> [973995.699686] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [973995.711331] CR2: 0000000000000028 CR3: 0000000104c1b000 CR4: 00000000003406e0
> [973995.725755] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [973995.740175] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> [973995.754595] Call Trace:
> [973995.759652] <IRQ>
> [973995.763855] ? kprobe_perf_func+0x28/0x210
> [973995.772210] tcp_sacktag_write_queue+0x5ff/0x9e0
> [973995.781615] tcp_ack+0x677/0x1390
> [973995.788408] tcp_rcv_established+0x1ce/0x740
> [973995.797112] tcp_v6_do_rcv+0x195/0x440
> [973995.804767] tcp_v6_rcv+0x94c/0x9f0
> [973995.811911] ip6_input_finish+0xea/0x430
> [973995.819917] ip6_input+0x32/0xa0
> [973995.826538] ? ip6_rcv_finish+0xa0/0xa0
> [973995.834373] ip6_rcv_finish+0x4b/0xa0
> [973995.841859] ipv6_rcv+0x2ec/0x4f0
> [973995.848653] ? ip6_fragment+0x9f0/0x9f0
> [973995.856489] ? ip6_make_skb+0x1c0/0x1c0
> [973995.864323] __netif_receive_skb_core+0x2d5/0x9a0
> [973995.873891] __netif_receive_skb+0x16/0x70
> [973995.882244] netif_receive_skb_internal+0x23/0x80
> [973995.891812] napi_gro_receive+0x113/0x1d0
> [973995.899993] mlx5e_handle_rx_cqe_mpwrq+0x5b6/0x8d0
> [973995.909735] mlx5e_poll_rx_cq+0x7c/0x7f0
> [973995.917740] mlx5e_napi_poll+0x8c/0x380
> [973995.925576] ? mlx5_cq_completion+0x54/0xb0
> [973995.934101] net_rx_action+0x22e/0x380
> [973995.941764] __do_softirq+0x106/0x2e8
> [973995.949255] irq_exit+0xb0/0xc0
> [973995.955696] do_IRQ+0x4f/0xd0
> [973995.961798] common_interrupt+0x86/0x86
> [973995.969634] RIP: 0033:0x7f8dec97a557
> [973995.976945] RSP: 002b:00007f8cc8cf2f48 EFLAGS: 00000206
> [973995.987552] ORIG_RAX: ffffffffffffff20
> [973995.995386] RAX: 00007f7fa9e15230 RBX: 00007f8c2153a160 RCX: 00007f7fa9e15230
> [973996.009810] RDX: 0000000000000000 RSI: 00007f8cc8cf3040 RDI: 00007f8c204f90c0
> [973996.024230] RBP: 00007f8cc8cf2f80 R08: 0000000000000001 R09: 000131aa4c002c01
> [973996.038652] R10: 0000000000000000 R11: 0000000000000001 R12: 00007f8c2153a170
> [973996.053073] R13: 00007f8cc8cf3040 R14: 00007f8c204f90c0 R15: 00007f8c2153a120
> [973996.067494] </IRQ>
> [973996.071858] Code:
> [973996.076051] b9
> [973996.079723] 01
> [973996.083383] 00
> [973996.087056] 00
> [973996.090730] 00
> [973996.094388] 85
> [973996.098062] c0
> [973996.101738] 0f
> [973996.105410] 8e
> [973996.109087] da
> [973996.112759] fd
> [973996.116433] ff
> [973996.120109] ff
> [973996.123783] 85
> [973996.127457] c0
> [973996.131132] 75
> [973996.134806] 28
> [973996.138481] 0f
> [973996.142156] b7
> [973996.145831] 43
> [973996.149504] 30
> [973996.153180] 41
> [973996.156835] 01
> [973996.160511] 45
> [973996.164168] 04
> [973996.167843] 48
> [973996.171517] 8b
> [973996.175190] 1b
> [973996.178848] 4c
> [973996.182521] 39
> [973996.186198] fb
> [973996.189872] 74
> [973996.193529] 8c
> [973996.197202] 49
> [973996.200877] 3b
> [973996.204534] 9c
> [973996.208209] 24
> [973996.211883] 50
> [973996.215559] 01
> [973996.219215] 00
> [973996.222889] 00
> [973996.226562] 74
> [973996.230221] c1
> [973996.233894] <8b>
> [973996.237916] 43
> [973996.241590] 28
> [973996.245264] 3b
> [973996.248921] 45
> [973996.252596] d4
> [973996.256271] 0f
> [973996.259929] 88
> [973996.263601] 9f
> [973996.267276] fd
> [973996.270935] ff
> [973996.274592] ff
> [973996.278267] eb
> [973996.281938] b3
> [973996.285614] 48
> [973996.289289] 8d
> [973996.292964] 43
> [973996.296638] 10
> [973996.300296] 8b
> [973996.303969] 4b
> [973996.307642] 28
> [973996.311325] RIP: tcp_sacktag_walk+0x2b7/0x460 RSP: ffff88203ef438e8
> [973996.324007] ------------[ cut here ]------------
> [973996.333399] CR2: 0000000000000028
> [973996.340218] ---[ end trace 5d1c83e12a57d03a ]---
> [973996.349605] Kernel panic - not syncing: Fatal exception in interrupt
> [973996.362521] Kernel Offset: disabled
> TBOOT: wait until all APs ready for txt shutdown
> TBOOT: IA32_FEATURE_CONTROL_MSR: 0000ff07
> TBOOT: CPU is SMX-capable
> TBOOT: CPU is VMX-capable
> TBOOT: SMX is enabled
> TBOOT: TXT chipset and all needed capabilities present
> TBOOT: TPM: Pcr 17 extend, return value = 0000003D
> TBOOT: TPM: Pcr 18 extend, return value = 0000003D
> TBOOT: TPM: Pcr 19 extend, return value = 0000003D
> TBOOT: cap'ed dynamic PCRs
> TBOOT: waiting for APs (0) to exit guests...
> TBOOT: .
> TBOOT:
> TBOOT: all APs exited guests
> TBOOT: calling txt_shutdown on AP
>
>
> Thanks!
^ permalink raw reply
* [PATCH 44/64] drivers: net: pxa168: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/marvell/pxa168_eth.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c
index 9937249..91b1c15 100644
--- a/drivers/net/ethernet/marvell/pxa168_eth.c
+++ b/drivers/net/ethernet/marvell/pxa168_eth.c
@@ -1496,9 +1496,8 @@ static int pxa168_eth_probe(struct platform_device *pdev)
netif_napi_add(dev, &pep->napi, pxa168_rx_poll, pep->rx_ring_size);
memset(&pep->timeout, 0, sizeof(struct timer_list));
- init_timer(&pep->timeout);
- pep->timeout.function = rxq_refill_timer_wrapper;
- pep->timeout.data = (unsigned long)pep;
+ setup_timer(&pep->timeout, rxq_refill_timer_wrapper,
+ (unsigned long)pep);
pep->smi_bus = mdiobus_alloc();
if (!pep->smi_bus) {
--
2.7.4
^ permalink raw reply related
* [PATCH 48/64] drivers: net: amd: use setup_timer() helper.
From: Allen Pais @ 2017-09-21 17:05 UTC (permalink / raw)
To: netdev
Cc: linux-kernel, m.grzeschik, dmitry.tarnyagin, wg, mkl, mark.einon,
linux, pcnet32, f.fainelli, bcm-kernel-feedback-list, venza, ajk,
jes, romieu, khc, simon, davem, linux-can, adi-buildroot-devel,
Allen Pais
In-Reply-To: <1506013525-29291-1-git-send-email-allen.lkml@gmail.com>
Use setup_timer function instead of initializing timer with the
function and data fields.
Signed-off-by: Allen Pais <allen.lkml@gmail.com>
---
drivers/net/ethernet/amd/a2065.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c
index e22f976..998d30e 100644
--- a/drivers/net/ethernet/amd/a2065.c
+++ b/drivers/net/ethernet/amd/a2065.c
@@ -733,10 +733,9 @@ static int a2065_init_one(struct zorro_dev *z,
dev->watchdog_timeo = 5*HZ;
dev->dma = 0;
- init_timer(&priv->multicast_timer);
- priv->multicast_timer.data = (unsigned long) dev;
- priv->multicast_timer.function =
- (void (*)(unsigned long))lance_set_multicast;
+ setup_timer(&priv->multicast_timer,
+ (void(*)(unsigned long))lance_set_multicast,
+ (unsigned long)dev);
err = register_netdev(dev);
if (err) {
--
2.7.4
^ permalink raw reply related
* Re: [PATCH net] net: prevent dst uses after free
From: Martin KaFai Lau @ 2017-09-21 17:12 UTC (permalink / raw)
To: Eric Dumazet
Cc: Paweł Staszewski, Wei Wang, Cong Wang,
Linux Kernel Network Developers, Eric Dumazet
In-Reply-To: <1506010546.29839.148.camel@edumazet-glaptop3.roam.corp.google.com>
On Thu, Sep 21, 2017 at 04:15:46PM +0000, Eric Dumazet wrote:
> From: Eric Dumazet <edumazet@google.com>
>
> In linux-4.13, Wei worked hard to convert dst to a traditional
> refcounted model, removing GC.
>
> We now want to make sure a dst refcount can not transition from 0 back
> to 1.
>
> The problem here is that input path attached a not refcounted dst to an
> skb. Then later, because packet is forwarded and hits skb_dst_force()
> before exiting RCU section, we might try to take a refcount on one dst
> that is about to be freed, if another cpu saw 1 -> 0 transition in
> dst_release() and queued the dst for freeing after one RCU grace period.
>
> Lets unify skb_dst_force() and skb_dst_force_safe(), since we should
> always perform the complete check against dst refcount, and not assume
> it is not zero.
Acked-by: Martin KaFai Lau <kafai@fb.com>
>
> Bugzilla : https://bugzilla.kernel.org/show_bug.cgi?id=197005
>
> [ 989.919496] skb_dst_force+0x32/0x34
> [ 989.919498] __dev_queue_xmit+0x1ad/0x482
> [ 989.919501] ? eth_header+0x28/0xc6
> [ 989.919502] dev_queue_xmit+0xb/0xd
> [ 989.919504] neigh_connected_output+0x9b/0xb4
> [ 989.919507] ip_finish_output2+0x234/0x294
> [ 989.919509] ? ipt_do_table+0x369/0x388
> [ 989.919510] ip_finish_output+0x12c/0x13f
> [ 989.919512] ip_output+0x53/0x87
> [ 989.919513] ip_forward_finish+0x53/0x5a
> [ 989.919515] ip_forward+0x2cb/0x3e6
> [ 989.919516] ? pskb_trim_rcsum.part.9+0x4b/0x4b
> [ 989.919518] ip_rcv_finish+0x2e2/0x321
> [ 989.919519] ip_rcv+0x26f/0x2eb
> [ 989.919522] ? vlan_do_receive+0x4f/0x289
> [ 989.919523] __netif_receive_skb_core+0x467/0x50b
> [ 989.919526] ? tcp_gro_receive+0x239/0x239
> [ 989.919529] ? inet_gro_receive+0x226/0x238
> [ 989.919530] __netif_receive_skb+0x4d/0x5f
> [ 989.919532] netif_receive_skb_internal+0x5c/0xaf
> [ 989.919533] napi_gro_receive+0x45/0x81
> [ 989.919536] ixgbe_poll+0xc8a/0xf09
> [ 989.919539] ? kmem_cache_free_bulk+0x1b6/0x1f7
> [ 989.919540] net_rx_action+0xf4/0x266
> [ 989.919543] __do_softirq+0xa8/0x19d
> [ 989.919545] irq_exit+0x5d/0x6b
> [ 989.919546] do_IRQ+0x9c/0xb5
> [ 989.919548] common_interrupt+0x93/0x93
> [ 989.919548] </IRQ>
>
>
> Similarly dst_clone() can use dst_hold() helper to have additional
> debugging, as a follow up to commit 44ebe79149ff ("net: add debug
> atomic_inc_not_zero() in dst_hold()")
>
> In net-next we will convert dst atomic_t to refcount_t for peace of
> mind.
>
> Fixes: a4c2fd7f7891 ("net: remove DST_NOCACHE flag")
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> Cc: Wei Wang <weiwan@google.com>
> Reported-by: Paweł Staszewski <pstaszewski@itcare.pl>
> Bisected-by: Paweł Staszewski <pstaszewski@itcare.pl>
> ---
> include/net/dst.h | 22 ++++------------------
> include/net/route.h | 2 +-
> include/net/sock.h | 2 +-
> 3 files changed, 6 insertions(+), 20 deletions(-)
>
> diff --git a/include/net/dst.h b/include/net/dst.h
> index 93568bd0a3520bb7402f04d90cf04ac99c81cfbe..06a6765da074449e6f1fe42ee05e711e898ad372 100644
> --- a/include/net/dst.h
> +++ b/include/net/dst.h
> @@ -271,7 +271,7 @@ static inline void dst_use_noref(struct dst_entry *dst, unsigned long time)
> static inline struct dst_entry *dst_clone(struct dst_entry *dst)
> {
> if (dst)
> - atomic_inc(&dst->__refcnt);
> + dst_hold(dst);
> return dst;
> }
>
> @@ -311,21 +311,6 @@ static inline void skb_dst_copy(struct sk_buff *nskb, const struct sk_buff *oskb
> __skb_dst_copy(nskb, oskb->_skb_refdst);
> }
>
> -/**
> - * skb_dst_force - makes sure skb dst is refcounted
> - * @skb: buffer
> - *
> - * If dst is not yet refcounted, let's do it
> - */
> -static inline void skb_dst_force(struct sk_buff *skb)
> -{
> - if (skb_dst_is_noref(skb)) {
> - WARN_ON(!rcu_read_lock_held());
> - skb->_skb_refdst &= ~SKB_DST_NOREF;
> - dst_clone(skb_dst(skb));
> - }
> -}
> -
> /**
> * dst_hold_safe - Take a reference on a dst if possible
> * @dst: pointer to dst entry
> @@ -339,16 +324,17 @@ static inline bool dst_hold_safe(struct dst_entry *dst)
> }
>
> /**
> - * skb_dst_force_safe - makes sure skb dst is refcounted
> + * skb_dst_force - makes sure skb dst is refcounted
> * @skb: buffer
> *
> * If dst is not yet refcounted and not destroyed, grab a ref on it.
> */
> -static inline void skb_dst_force_safe(struct sk_buff *skb)
> +static inline void skb_dst_force(struct sk_buff *skb)
> {
> if (skb_dst_is_noref(skb)) {
> struct dst_entry *dst = skb_dst(skb);
>
> + WARN_ON(!rcu_read_lock_held());
> if (!dst_hold_safe(dst))
> dst = NULL;
>
> diff --git a/include/net/route.h b/include/net/route.h
> index 1b09a9368c68d46f0c5ee8ce3cefe566000c1ec1..57dfc6850d378e4b96f13b140eef554d66c24cdf 100644
> --- a/include/net/route.h
> +++ b/include/net/route.h
> @@ -190,7 +190,7 @@ static inline int ip_route_input(struct sk_buff *skb, __be32 dst, __be32 src,
> rcu_read_lock();
> err = ip_route_input_noref(skb, dst, src, tos, devin);
> if (!err) {
> - skb_dst_force_safe(skb);
> + skb_dst_force(skb);
> if (!skb_dst(skb))
> err = -EINVAL;
> }
> diff --git a/include/net/sock.h b/include/net/sock.h
> index 03a362568357acc7278a318423dd3873103f90ca..a6b9a8d1a6df3f72df8f1aac0f577257fa6452d0 100644
> --- a/include/net/sock.h
> +++ b/include/net/sock.h
> @@ -856,7 +856,7 @@ void sk_stream_write_space(struct sock *sk);
> static inline void __sk_add_backlog(struct sock *sk, struct sk_buff *skb)
> {
> /* dont let skb dst not refcounted, we are going to leave rcu lock */
> - skb_dst_force_safe(skb);
> + skb_dst_force(skb);
>
> if (!sk->sk_backlog.tail)
> sk->sk_backlog.head = skb;
>
>
^ permalink raw reply
* Re: [PATCH net-next 2/7] rtnetlink: add helper to put master ifindex
From: David Ahern @ 2017-09-21 17:16 UTC (permalink / raw)
To: Florian Westphal, netdev
In-Reply-To: <20170921165902.10746-3-fw@strlen.de>
On 9/21/17 10:58 AM, Florian Westphal wrote:
> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index a78fd61da0ec..c801212ee40e 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -1307,6 +1307,31 @@ static u32 rtnl_get_event(unsigned long event)
> return rtnl_event_type;
> }
>
> +static int put_master_ifindex(struct sk_buff *skb, struct net_device *dev)
> +{
> + const struct net_device *upper_dev;
> + int ret = 0;
> +
> + rcu_read_lock();
> +
> + upper_dev = netdev_master_upper_dev_get_rcu(dev);
> + if (upper_dev)
> + ret = nla_put_u32(skb, IFLA_MASTER, upper_dev->ifindex);
> +
> + rcu_read_unlock();
> + return ret;
> +}
> +
> +static int nla_put_iflink(struct sk_buff *skb, const struct net_device *dev)
> +{
> + int ifindex = dev_get_iflink(dev);
> +
> + if (dev->ifindex == ifindex)
> + return 0;
> +
> + return nla_put_u32(skb, IFLA_LINK, ifindex);
> +}
> +
> static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
> int type, u32 pid, u32 seq, u32 change,
> unsigned int flags, u32 ext_filter_mask,
Subject references only change for master index, put the patch is
converting 2 things to helpers.
^ permalink raw reply
* Re: [PATCH net-next 5/7] rtnetlink: add helper to dump vf information
From: David Ahern @ 2017-09-21 17:17 UTC (permalink / raw)
To: Florian Westphal, netdev
In-Reply-To: <20170921165902.10746-6-fw@strlen.de>
On 9/21/17 10:59 AM, Florian Westphal wrote:
> similar to earlier patches, split out more parts of this function to
> better see what is happening and where we assume rtnl is locked.
>
> Signed-off-by: Florian Westphal <fw@strlen.de>
> ---
> net/core/rtnetlink.c | 80 ++++++++++++++++++++++++++++++++--------------------
> 1 file changed, 50 insertions(+), 30 deletions(-)
>
> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index 1af3ef7f329d..7503021fe308 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -1211,6 +1211,36 @@ static noinline_for_stack int rtnl_fill_vfinfo(struct sk_buff *skb,
> return -EMSGSIZE;
> }
>
> +static noinline_for_stack int rtnl_fill_vf(struct sk_buff *skb,
> + struct net_device *dev,
> + u32 ext_filter_mask)
> +{
> + struct nlattr *vfinfo;
> + int i, num_vfs;
> +
> + if (!dev->dev.parent || ((ext_filter_mask & RTEXT_FILTER_VF) == 0))
> + return 0;
> +
> + num_vfs = dev_num_vf(dev->dev.parent);
> + if (nla_put_u32(skb, IFLA_NUM_VF, num_vfs))
> + return -EMSGSIZE;
> +
> + if (!dev->netdev_ops->ndo_get_vf_config)
> + return 0;
> +
> + vfinfo = nla_nest_start(skb, IFLA_VFINFO_LIST);
> + if (!vfinfo)
> + return -EMSGSIZE;
> +
> + for (i = 0; i < num_vfs; i++) {
> + if (rtnl_fill_vfinfo(skb, dev, i, vfinfo))
> + return -EMSGSIZE;
> + }
> +
> + nla_nest_end(skb, vfinfo);
> + return 0;
> +}
> +
> static int rtnl_fill_link_ifmap(struct sk_buff *skb, struct net_device *dev)
> {
> struct rtnl_link_ifmap map;
> @@ -1355,6 +1385,23 @@ static int noinline nla_put_ifalias(struct sk_buff *skb, struct net_device *dev)
> return 0;
> }
>
> +static int noinline rtnl_fill_link_netnsid(struct sk_buff *skb,
> + const struct net_device *dev)
> +{
> + if (dev->rtnl_link_ops && dev->rtnl_link_ops->get_link_net) {
> + struct net *link_net = dev->rtnl_link_ops->get_link_net(dev);
> +
> + if (!net_eq(dev_net(dev), link_net)) {
> + int id = peernet2id_alloc(dev_net(dev), link_net);
> +
> + if (nla_put_s32(skb, IFLA_LINK_NETNSID, id))
> + return -EMSGSIZE;
> + }
> + }
> +
> + return 0;
> +}
> +
> static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
Similarly for this patch -- subject references vf helper yet the code
changes flips netnsid as well.
^ permalink raw reply
* Re: [net-next v3] bridge: trigger RTM_NEWLINK when interface is modified by bridge ioctl
From: Roopa Prabhu @ 2017-09-21 17:20 UTC (permalink / raw)
To: David Ahern
Cc: Vincent Bernat, Stephen Hemminger, David Miller, bridge,
netdev@vger.kernel.org
In-Reply-To: <0cdf06b9-e67d-c035-e0f1-c4eb143bf631@gmail.com>
On Thu, Sep 21, 2017 at 9:43 AM, David Ahern <dsahern@gmail.com> wrote:
> On 9/21/17 4:05 AM, Vincent Bernat wrote:
>> Currently, there is a difference in netlink events received when an
>> interface is modified through bridge ioctl() or through netlink. This
>> patch generates additional events when an interface is added to or
>> removed from a bridge via ioctl().
>>
>> When adding then removing an interface from a bridge with netlink, we
>> get:
>>
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>>
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> Deleted 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>>
>> When using ioctl():
>>
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>>
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue master bridge0 state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> Deleted 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 master bridge0 state UNKNOWN
>> link/ether 9e:da:60:ee:cf:c8
>> 5: dummy1: <BROADCAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN group default
>> link/ether 9e:da:60:ee:cf:c8 brd ff:ff:ff:ff:ff:ff
>>
>> Without this patch, the last netlink notification is not sent.
>>
>> Signed-off-by: Vincent Bernat <vincent@bernat.im>
>> ---
>> net/bridge/br_ioctl.c | 3 +++
>> 1 file changed, 3 insertions(+)
>>
>> diff --git a/net/bridge/br_ioctl.c b/net/bridge/br_ioctl.c
>> index 7970f8540cbb..66cd98772051 100644
>> --- a/net/bridge/br_ioctl.c
>> +++ b/net/bridge/br_ioctl.c
>> @@ -102,6 +102,9 @@ static int add_del_if(struct net_bridge *br, int ifindex, int isadd)
>> else
>> ret = br_del_if(br, dev);
>>
>> + if (!ret)
>> + rtmsg_ifinfo(RTM_NEWLINK, dev, IFF_MASTER, GFP_KERNEL);
>> +
>> return ret;
>> }
>>
>>
>
> Agreed that this is needed for userspace to know about the master change
> when done through ioctl. The bridge code is emitting a lot of what
> appears to be redundant messages for both paths (netlink and ioctl).
>
> Reviewed-by: David Ahern <dsahern@gmail.com>
>
this patch seems fine...but ideally I would have assumed
netdev_upper_dev_unlink which
is eventually called in both paths would generate the RTN_NEWLINK
IFF_MASTER in response
to the NETDEV_CHANGEUPPER notifier. If we add it there now, it might
need some more fixes
to not generate redundant notifications for the netlink case.
^ permalink raw reply
* Re: [PATCH net-next 7/7] rtnetlink: rtnl_have_link_slave_info doesn't need rtnl
From: David Ahern @ 2017-09-21 17:24 UTC (permalink / raw)
To: Florian Westphal, netdev
In-Reply-To: <20170921165902.10746-8-fw@strlen.de>
On 9/21/17 10:59 AM, Florian Westphal wrote:
> @@ -539,6 +543,8 @@ static int rtnl_link_slave_info_fill(struct sk_buff *skb,
> struct nlattr *slave_data;
> int err;
>
> + ASSERT_RTNL();
> +
> master_dev = netdev_master_upper_dev_get((struct net_device *) dev);
> if (!master_dev)
> return 0;
> @@ -570,6 +576,8 @@ static int rtnl_link_info_fill(struct sk_buff *skb,
> struct nlattr *data;
> int err;
>
> + ASSERT_RTNL();
> +
> if (!ops)
> return 0;
> if (nla_put_string(skb, IFLA_INFO_KIND, ops->kind) < 0)
> @@ -600,6 +608,8 @@ static int rtnl_link_fill(struct sk_buff *skb, const struct net_device *dev)
> struct nlattr *linkinfo;
> int err = -EMSGSIZE;
>
> + ASSERT_RTNL();
> +
> linkinfo = nla_nest_start(skb, IFLA_LINKINFO);
> if (linkinfo == NULL)
> goto out;
>
Since rtnl_link_slave_info_fill and rtnl_link_info_fill are only called
by rtnl_link_fill and rtnl_link_fill is only called rtnl_fill_ifinfo
which as the ASSERT_RTNL why add to these lower functions as well?
^ permalink raw reply
* Re: [PATCH net-next 5/7] rtnetlink: add helper to dump vf information
From: Florian Westphal @ 2017-09-21 17:25 UTC (permalink / raw)
To: David Ahern; +Cc: Florian Westphal, netdev
In-Reply-To: <bdc39ebf-e934-16ce-0834-e0795774a656@gmail.com>
David Ahern <dsahern@gmail.com> wrote:
> On 9/21/17 10:59 AM, Florian Westphal wrote:
> > similar to earlier patches, split out more parts of this function to
> > better see what is happening and where we assume rtnl is locked.
[..]
> > +static noinline_for_stack int rtnl_fill_vf(struct sk_buff *skb,
> > + struct net_device *dev,
> > + u32 ext_filter_mask)
[..]
> > +static int noinline rtnl_fill_link_netnsid(struct sk_buff *skb,
> > + const struct net_device *dev)
[..]
> > static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
>
> Similarly for this patch -- subject references vf helper yet the code
> changes flips netnsid as well.
Right, anomly due to improper squash-merge.
I'll send a v2 tomorrow (giving more time for others to comment).
^ permalink raw reply
* Re: [net-next v3] bridge: trigger RTM_NEWLINK when interface is modified by bridge ioctl
From: David Ahern @ 2017-09-21 17:29 UTC (permalink / raw)
To: Roopa Prabhu, David Ahern
Cc: netdev@vger.kernel.org, bridge, Vincent Bernat, David Miller
In-Reply-To: <CAJieiUjqttMeC24CzDa8Y8g=opRKoa2hbG2Cb+78u632vA3uqw@mail.gmail.com>
On 9/21/17 11:20 AM, Roopa Prabhu wrote:
> this patch seems fine...but ideally I would have assumed
> netdev_upper_dev_unlink which
> is eventually called in both paths would generate the RTN_NEWLINK
> IFF_MASTER in response
> to the NETDEV_CHANGEUPPER notifier. If we add it there now, it might
> need some more fixes
> to not generate redundant notifications for the netlink case.
Agreed and it is another pandora's box
^ permalink raw reply
* Re: [PATCH net-next 7/7] rtnetlink: rtnl_have_link_slave_info doesn't need rtnl
From: Florian Westphal @ 2017-09-21 17:31 UTC (permalink / raw)
To: David Ahern; +Cc: Florian Westphal, netdev
In-Reply-To: <b645a22f-9153-b327-8995-2b140cff59a0@gmail.com>
David Ahern <dsahern@gmail.com> wrote:
> On 9/21/17 10:59 AM, Florian Westphal wrote:
> > @@ -539,6 +543,8 @@ static int rtnl_link_slave_info_fill(struct sk_buff *skb,
> > struct nlattr *slave_data;
> > int err;
> >
> > + ASSERT_RTNL();
> > +
[..]
master_dev = netdev_master_upper_dev_get((struct net_device *) dev);
> Since rtnl_link_slave_info_fill and rtnl_link_info_fill are only called
> by rtnl_link_fill and rtnl_link_fill is only called rtnl_fill_ifinfo
> which as the ASSERT_RTNL why add to these lower functions as well?
I'll remove this patch in v2 and will hold it back in my private queue;
these serve more as a reminder/TODO for myself rather than anything else.
^ permalink raw reply
* Re: [PATCH iproute2 master 1/2] json: move json printer to common library
From: Julien Fortin @ 2017-09-21 17:47 UTC (permalink / raw)
To: Daniel Borkmann; +Cc: Stephen Hemminger, ast, netdev
In-Reply-To: <635a857b7b32cd0df79805f8e32b20a2674a675d.1505956723.git.daniel@iogearbox.net>
On Thu, Sep 21, 2017 at 1:42 AM, Daniel Borkmann <daniel@iogearbox.net> wrote:
> Move the json printer which is based on json writer into the
> iproute2 library, so it can be used by library code and tools
> other than ip. Should probably have been done from the beginning
> like that given json writer is in the library already anyway.
> No functional changes.
>
> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Julien Fortin <julien@cumulusnetworks.com>
> ---
> include/json_print.h | 71 ++++++++++++++++
> ip/Makefile | 2 +-
> ip/ip_common.h | 65 ++------------
> ip/ip_print.c | 233 ---------------------------------------------------
> lib/Makefile | 2 +-
> lib/json_print.c | 231 ++++++++++++++++++++++++++++++++++++++++++++++++++
> 6 files changed, 312 insertions(+), 292 deletions(-)
> create mode 100644 include/json_print.h
> delete mode 100644 ip/ip_print.c
> create mode 100644 lib/json_print.c
>
> diff --git a/include/json_print.h b/include/json_print.h
> new file mode 100644
> index 0000000..44cf5ac
> --- /dev/null
> +++ b/include/json_print.h
> @@ -0,0 +1,71 @@
> +/*
> + * json_print.h "print regular or json output, based on json_writer".
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * as published by the Free Software Foundation; either version
> + * 2 of the License, or (at your option) any later version.
> + *
> + * Authors: Julien Fortin, <julien@cumulusnetworks.com>
> + */
> +
> +#ifndef _JSON_PRINT_H_
> +#define _JSON_PRINT_H_
> +
> +#include "json_writer.h"
> +#include "color.h"
> +
> +json_writer_t *get_json_writer(void);
> +
> +/*
> + * use:
> + * - PRINT_ANY for context based output
> + * - PRINT_FP for non json specific output
> + * - PRINT_JSON for json specific output
> + */
> +enum output_type {
> + PRINT_FP = 1,
> + PRINT_JSON = 2,
> + PRINT_ANY = 4,
> +};
> +
> +void new_json_obj(int json, FILE *fp);
> +void delete_json_obj(void);
> +
> +bool is_json_context(void);
> +
> +void set_current_fp(FILE *fp);
> +
> +void fflush_fp(void);
> +
> +void open_json_object(const char *str);
> +void close_json_object(void);
> +void open_json_array(enum output_type type, const char *delim);
> +void close_json_array(enum output_type type, const char *delim);
> +
> +#define _PRINT_FUNC(type_name, type) \
> + void print_color_##type_name(enum output_type t, \
> + enum color_attr color, \
> + const char *key, \
> + const char *fmt, \
> + type value); \
> + \
> + static inline void print_##type_name(enum output_type t, \
> + const char *key, \
> + const char *fmt, \
> + type value) \
> + { \
> + print_color_##type_name(t, -1, key, fmt, value); \
> + }
> +_PRINT_FUNC(int, int);
> +_PRINT_FUNC(bool, bool);
> +_PRINT_FUNC(null, const char*);
> +_PRINT_FUNC(string, const char*);
> +_PRINT_FUNC(uint, uint64_t);
> +_PRINT_FUNC(hu, unsigned short);
> +_PRINT_FUNC(hex, unsigned int);
> +_PRINT_FUNC(0xhex, unsigned int);
> +_PRINT_FUNC(lluint, unsigned long long int);
> +#undef _PRINT_FUNC
> +
> +#endif /* _JSON_PRINT_H_ */
> diff --git a/ip/Makefile b/ip/Makefile
> index 52c9a2e..5a1c7ad 100644
> --- a/ip/Makefile
> +++ b/ip/Makefile
> @@ -9,7 +9,7 @@ IPOBJ=ip.o ipaddress.o ipaddrlabel.o iproute.o iprule.o ipnetns.o \
> link_iptnl.o link_gre6.o iplink_bond.o iplink_bond_slave.o iplink_hsr.o \
> iplink_bridge.o iplink_bridge_slave.o ipfou.o iplink_ipvlan.o \
> iplink_geneve.o iplink_vrf.o iproute_lwtunnel.o ipmacsec.o ipila.o \
> - ipvrf.o iplink_xstats.o ipseg6.o ip_print.o
> + ipvrf.o iplink_xstats.o ipseg6.o
>
> RTMONOBJ=rtmon.o
>
> diff --git a/ip/ip_common.h b/ip/ip_common.h
> index efc789c..4b8b0a7 100644
> --- a/ip/ip_common.h
> +++ b/ip/ip_common.h
> @@ -1,3 +1,10 @@
> +#ifndef _IP_COMMON_H_
> +#define _IP_COMMON_H_
> +
> +#include <stdbool.h>
> +
> +#include "json_print.h"
> +
> struct link_filter {
> int ifindex;
> int family;
> @@ -101,8 +108,6 @@ static inline int rtm_get_table(struct rtmsg *r, struct rtattr **tb)
>
> extern struct rtnl_handle rth;
>
> -#include <stdbool.h>
> -
> struct link_util {
> struct link_util *next;
> const char *id;
> @@ -141,58 +146,4 @@ int name_is_vrf(const char *name);
>
> void print_num(FILE *fp, unsigned int width, uint64_t count);
>
> -#include "json_writer.h"
> -
> -json_writer_t *get_json_writer(void);
> -/*
> - * use:
> - * - PRINT_ANY for context based output
> - * - PRINT_FP for non json specific output
> - * - PRINT_JSON for json specific output
> - */
> -enum output_type {
> - PRINT_FP = 1,
> - PRINT_JSON = 2,
> - PRINT_ANY = 4,
> -};
> -
> -void new_json_obj(int json, FILE *fp);
> -void delete_json_obj(void);
> -
> -bool is_json_context(void);
> -
> -void set_current_fp(FILE *fp);
> -
> -void fflush_fp(void);
> -
> -void open_json_object(const char *str);
> -void close_json_object(void);
> -void open_json_array(enum output_type type, const char *delim);
> -void close_json_array(enum output_type type, const char *delim);
> -
> -#include "color.h"
> -
> -#define _PRINT_FUNC(type_name, type) \
> - void print_color_##type_name(enum output_type t, \
> - enum color_attr color, \
> - const char *key, \
> - const char *fmt, \
> - type value); \
> - \
> - static inline void print_##type_name(enum output_type t, \
> - const char *key, \
> - const char *fmt, \
> - type value) \
> - { \
> - print_color_##type_name(t, -1, key, fmt, value); \
> - }
> -_PRINT_FUNC(int, int);
> -_PRINT_FUNC(bool, bool);
> -_PRINT_FUNC(null, const char*);
> -_PRINT_FUNC(string, const char*);
> -_PRINT_FUNC(uint, uint64_t);
> -_PRINT_FUNC(hu, unsigned short);
> -_PRINT_FUNC(hex, unsigned int);
> -_PRINT_FUNC(0xhex, unsigned int);
> -_PRINT_FUNC(lluint, unsigned long long int);
> -#undef _PRINT_FUNC
> +#endif /* _IP_COMMON_H_ */
> diff --git a/ip/ip_print.c b/ip/ip_print.c
> deleted file mode 100644
> index 4cd6a0b..0000000
> --- a/ip/ip_print.c
> +++ /dev/null
> @@ -1,233 +0,0 @@
> -/*
> - * ip_print.c "ip print regular or json output".
> - *
> - * This program is free software; you can redistribute it and/or
> - * modify it under the terms of the GNU General Public License
> - * as published by the Free Software Foundation; either version
> - * 2 of the License, or (at your option) any later version.
> - *
> - * Authors: Julien Fortin, <julien@cumulusnetworks.com>
> - *
> - */
> -
> -#include <stdarg.h>
> -#include <stdio.h>
> -
> -#include "utils.h"
> -#include "ip_common.h"
> -#include "json_writer.h"
> -
> -static json_writer_t *_jw;
> -static FILE *_fp;
> -
> -#define _IS_JSON_CONTEXT(type) ((type & PRINT_JSON || type & PRINT_ANY) && _jw)
> -#define _IS_FP_CONTEXT(type) (!_jw && (type & PRINT_FP || type & PRINT_ANY))
> -
> -void new_json_obj(int json, FILE *fp)
> -{
> - if (json) {
> - _jw = jsonw_new(fp);
> - if (!_jw) {
> - perror("json object");
> - exit(1);
> - }
> - jsonw_pretty(_jw, true);
> - jsonw_start_array(_jw);
> - }
> - set_current_fp(fp);
> -}
> -
> -void delete_json_obj(void)
> -{
> - if (_jw) {
> - jsonw_end_array(_jw);
> - jsonw_destroy(&_jw);
> - }
> -}
> -
> -bool is_json_context(void)
> -{
> - return _jw != NULL;
> -}
> -
> -void set_current_fp(FILE *fp)
> -{
> - if (!fp) {
> - fprintf(stderr, "Error: invalid file pointer.\n");
> - exit(1);
> - }
> - _fp = fp;
> -}
> -
> -json_writer_t *get_json_writer(void)
> -{
> - return _jw;
> -}
> -
> -void open_json_object(const char *str)
> -{
> - if (_IS_JSON_CONTEXT(PRINT_JSON)) {
> - if (str)
> - jsonw_name(_jw, str);
> - jsonw_start_object(_jw);
> - }
> -}
> -
> -void close_json_object(void)
> -{
> - if (_IS_JSON_CONTEXT(PRINT_JSON))
> - jsonw_end_object(_jw);
> -}
> -
> -/*
> - * Start json array or string array using
> - * the provided string as json key (if not null)
> - * or as array delimiter in non-json context.
> - */
> -void open_json_array(enum output_type type, const char *str)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - if (str)
> - jsonw_name(_jw, str);
> - jsonw_start_array(_jw);
> - } else if (_IS_FP_CONTEXT(type)) {
> - fprintf(_fp, "%s", str);
> - }
> -}
> -
> -/*
> - * End json array or string array
> - */
> -void close_json_array(enum output_type type, const char *str)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - jsonw_pretty(_jw, false);
> - jsonw_end_array(_jw);
> - jsonw_pretty(_jw, true);
> - } else if (_IS_FP_CONTEXT(type)) {
> - fprintf(_fp, "%s", str);
> - }
> -}
> -
> -/*
> - * pre-processor directive to generate similar
> - * functions handling different types
> - */
> -#define _PRINT_FUNC(type_name, type) \
> - void print_color_##type_name(enum output_type t, \
> - enum color_attr color, \
> - const char *key, \
> - const char *fmt, \
> - type value) \
> - { \
> - if (_IS_JSON_CONTEXT(t)) { \
> - if (!key) \
> - jsonw_##type_name(_jw, value); \
> - else \
> - jsonw_##type_name##_field(_jw, key, value); \
> - } else if (_IS_FP_CONTEXT(t)) { \
> - color_fprintf(_fp, color, fmt, value); \
> - } \
> - }
> -_PRINT_FUNC(int, int);
> -_PRINT_FUNC(hu, unsigned short);
> -_PRINT_FUNC(uint, uint64_t);
> -_PRINT_FUNC(lluint, unsigned long long int);
> -#undef _PRINT_FUNC
> -
> -void print_color_string(enum output_type type,
> - enum color_attr color,
> - const char *key,
> - const char *fmt,
> - const char *value)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - if (key && !value)
> - jsonw_name(_jw, key);
> - else if (!key && value)
> - jsonw_string(_jw, value);
> - else
> - jsonw_string_field(_jw, key, value);
> - } else if (_IS_FP_CONTEXT(type)) {
> - color_fprintf(_fp, color, fmt, value);
> - }
> -}
> -
> -/*
> - * value's type is bool. When using this function in FP context you can't pass
> - * a value to it, you will need to use "is_json_context()" to have different
> - * branch for json and regular output. grep -r "print_bool" for example
> - */
> -void print_color_bool(enum output_type type,
> - enum color_attr color,
> - const char *key,
> - const char *fmt,
> - bool value)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - if (key)
> - jsonw_bool_field(_jw, key, value);
> - else
> - jsonw_bool(_jw, value);
> - } else if (_IS_FP_CONTEXT(type)) {
> - color_fprintf(_fp, color, fmt, value ? "true" : "false");
> - }
> -}
> -
> -/*
> - * In JSON context uses hardcode %#x format: 42 -> 0x2a
> - */
> -void print_color_0xhex(enum output_type type,
> - enum color_attr color,
> - const char *key,
> - const char *fmt,
> - unsigned int hex)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - SPRINT_BUF(b1);
> -
> - snprintf(b1, sizeof(b1), "%#x", hex);
> - print_string(PRINT_JSON, key, NULL, b1);
> - } else if (_IS_FP_CONTEXT(type)) {
> - color_fprintf(_fp, color, fmt, hex);
> - }
> -}
> -
> -void print_color_hex(enum output_type type,
> - enum color_attr color,
> - const char *key,
> - const char *fmt,
> - unsigned int hex)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - SPRINT_BUF(b1);
> -
> - snprintf(b1, sizeof(b1), "%x", hex);
> - if (key)
> - jsonw_string_field(_jw, key, b1);
> - else
> - jsonw_string(_jw, b1);
> - } else if (_IS_FP_CONTEXT(type)) {
> - color_fprintf(_fp, color, fmt, hex);
> - }
> -}
> -
> -/*
> - * In JSON context we don't use the argument "value" we simply call jsonw_null
> - * whereas FP context can use "value" to output anything
> - */
> -void print_color_null(enum output_type type,
> - enum color_attr color,
> - const char *key,
> - const char *fmt,
> - const char *value)
> -{
> - if (_IS_JSON_CONTEXT(type)) {
> - if (key)
> - jsonw_null_field(_jw, key);
> - else
> - jsonw_null(_jw);
> - } else if (_IS_FP_CONTEXT(type)) {
> - color_fprintf(_fp, color, fmt, value);
> - }
> -}
> diff --git a/lib/Makefile b/lib/Makefile
> index 5e9f72f..0fbdf4c 100644
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -3,7 +3,7 @@ include ../config.mk
> CFLAGS += -fPIC
>
> UTILOBJ = utils.o rt_names.o ll_types.o ll_proto.o ll_addr.o \
> - inet_proto.o namespace.o json_writer.o \
> + inet_proto.o namespace.o json_writer.o json_print.o \
> names.o color.o bpf.o exec.o fs.o
>
> NLOBJ=libgenl.o ll_map.o libnetlink.o
> diff --git a/lib/json_print.c b/lib/json_print.c
> new file mode 100644
> index 0000000..93b4119
> --- /dev/null
> +++ b/lib/json_print.c
> @@ -0,0 +1,231 @@
> +/*
> + * json_print.c "print regular or json output, based on json_writer".
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * as published by the Free Software Foundation; either version
> + * 2 of the License, or (at your option) any later version.
> + *
> + * Authors: Julien Fortin, <julien@cumulusnetworks.com>
> + */
> +
> +#include <stdarg.h>
> +#include <stdio.h>
> +
> +#include "utils.h"
> +#include "json_print.h"
> +
> +static json_writer_t *_jw;
> +static FILE *_fp;
> +
> +#define _IS_JSON_CONTEXT(type) ((type & PRINT_JSON || type & PRINT_ANY) && _jw)
> +#define _IS_FP_CONTEXT(type) (!_jw && (type & PRINT_FP || type & PRINT_ANY))
> +
> +void new_json_obj(int json, FILE *fp)
> +{
> + if (json) {
> + _jw = jsonw_new(fp);
> + if (!_jw) {
> + perror("json object");
> + exit(1);
> + }
> + jsonw_pretty(_jw, true);
> + jsonw_start_array(_jw);
> + }
> + set_current_fp(fp);
> +}
> +
> +void delete_json_obj(void)
> +{
> + if (_jw) {
> + jsonw_end_array(_jw);
> + jsonw_destroy(&_jw);
> + }
> +}
> +
> +bool is_json_context(void)
> +{
> + return _jw != NULL;
> +}
> +
> +void set_current_fp(FILE *fp)
> +{
> + if (!fp) {
> + fprintf(stderr, "Error: invalid file pointer.\n");
> + exit(1);
> + }
> + _fp = fp;
> +}
> +
> +json_writer_t *get_json_writer(void)
> +{
> + return _jw;
> +}
> +
> +void open_json_object(const char *str)
> +{
> + if (_IS_JSON_CONTEXT(PRINT_JSON)) {
> + if (str)
> + jsonw_name(_jw, str);
> + jsonw_start_object(_jw);
> + }
> +}
> +
> +void close_json_object(void)
> +{
> + if (_IS_JSON_CONTEXT(PRINT_JSON))
> + jsonw_end_object(_jw);
> +}
> +
> +/*
> + * Start json array or string array using
> + * the provided string as json key (if not null)
> + * or as array delimiter in non-json context.
> + */
> +void open_json_array(enum output_type type, const char *str)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + if (str)
> + jsonw_name(_jw, str);
> + jsonw_start_array(_jw);
> + } else if (_IS_FP_CONTEXT(type)) {
> + fprintf(_fp, "%s", str);
> + }
> +}
> +
> +/*
> + * End json array or string array
> + */
> +void close_json_array(enum output_type type, const char *str)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + jsonw_pretty(_jw, false);
> + jsonw_end_array(_jw);
> + jsonw_pretty(_jw, true);
> + } else if (_IS_FP_CONTEXT(type)) {
> + fprintf(_fp, "%s", str);
> + }
> +}
> +
> +/*
> + * pre-processor directive to generate similar
> + * functions handling different types
> + */
> +#define _PRINT_FUNC(type_name, type) \
> + void print_color_##type_name(enum output_type t, \
> + enum color_attr color, \
> + const char *key, \
> + const char *fmt, \
> + type value) \
> + { \
> + if (_IS_JSON_CONTEXT(t)) { \
> + if (!key) \
> + jsonw_##type_name(_jw, value); \
> + else \
> + jsonw_##type_name##_field(_jw, key, value); \
> + } else if (_IS_FP_CONTEXT(t)) { \
> + color_fprintf(_fp, color, fmt, value); \
> + } \
> + }
> +_PRINT_FUNC(int, int);
> +_PRINT_FUNC(hu, unsigned short);
> +_PRINT_FUNC(uint, uint64_t);
> +_PRINT_FUNC(lluint, unsigned long long int);
> +#undef _PRINT_FUNC
> +
> +void print_color_string(enum output_type type,
> + enum color_attr color,
> + const char *key,
> + const char *fmt,
> + const char *value)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + if (key && !value)
> + jsonw_name(_jw, key);
> + else if (!key && value)
> + jsonw_string(_jw, value);
> + else
> + jsonw_string_field(_jw, key, value);
> + } else if (_IS_FP_CONTEXT(type)) {
> + color_fprintf(_fp, color, fmt, value);
> + }
> +}
> +
> +/*
> + * value's type is bool. When using this function in FP context you can't pass
> + * a value to it, you will need to use "is_json_context()" to have different
> + * branch for json and regular output. grep -r "print_bool" for example
> + */
> +void print_color_bool(enum output_type type,
> + enum color_attr color,
> + const char *key,
> + const char *fmt,
> + bool value)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + if (key)
> + jsonw_bool_field(_jw, key, value);
> + else
> + jsonw_bool(_jw, value);
> + } else if (_IS_FP_CONTEXT(type)) {
> + color_fprintf(_fp, color, fmt, value ? "true" : "false");
> + }
> +}
> +
> +/*
> + * In JSON context uses hardcode %#x format: 42 -> 0x2a
> + */
> +void print_color_0xhex(enum output_type type,
> + enum color_attr color,
> + const char *key,
> + const char *fmt,
> + unsigned int hex)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + SPRINT_BUF(b1);
> +
> + snprintf(b1, sizeof(b1), "%#x", hex);
> + print_string(PRINT_JSON, key, NULL, b1);
> + } else if (_IS_FP_CONTEXT(type)) {
> + color_fprintf(_fp, color, fmt, hex);
> + }
> +}
> +
> +void print_color_hex(enum output_type type,
> + enum color_attr color,
> + const char *key,
> + const char *fmt,
> + unsigned int hex)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + SPRINT_BUF(b1);
> +
> + snprintf(b1, sizeof(b1), "%x", hex);
> + if (key)
> + jsonw_string_field(_jw, key, b1);
> + else
> + jsonw_string(_jw, b1);
> + } else if (_IS_FP_CONTEXT(type)) {
> + color_fprintf(_fp, color, fmt, hex);
> + }
> +}
> +
> +/*
> + * In JSON context we don't use the argument "value" we simply call jsonw_null
> + * whereas FP context can use "value" to output anything
> + */
> +void print_color_null(enum output_type type,
> + enum color_attr color,
> + const char *key,
> + const char *fmt,
> + const char *value)
> +{
> + if (_IS_JSON_CONTEXT(type)) {
> + if (key)
> + jsonw_null_field(_jw, key);
> + else
> + jsonw_null(_jw);
> + } else if (_IS_FP_CONTEXT(type)) {
> + color_fprintf(_fp, color, fmt, value);
> + }
> +}
> --
> 1.9.3
>
^ permalink raw reply
* [PATCH net-next v2 0/4] cxgb4: add support to offload tc flower
From: Rahul Lakkireddy @ 2017-09-21 18:11 UTC (permalink / raw)
To: netdev; +Cc: davem, kumaras, ganeshgr, nirranjan, indranil, Rahul Lakkireddy
This series of patches add support to offload tc flower onto Chelsio
NICs.
Patch 1 adds basic skeleton to prepare for offloading tc flower flows.
Patch 2 adds support to add/remove flows for offload. Flows can have
accompanying masks. Following match and action are currently supported
for offload:
Match: ether-protocol, IPv4/IPv6 addresses, L4 ports (TCP/UDP)
Action: drop, redirect to another port on the device.
Patch 3 adds support to offload tc-flower flows having
vlan actions: pop, push, and modify.
Patch 4 adds support to fetch stats for the offloaded tc flower flows
from hardware.
Support for offloading more match and action types are to be followed
in subsequent series.
Thanks,
Rahul
---
v2:
- Setting ftid to -1 not required after bitmap_find_free_region
in cxgb4_get_free_ftid.
- Direct return can be used as jumping to error path is not needed
if flower entry allocation failed in cxgb4_tc_flower_replace.
Same applies if flower entry not found in cxgb4_tc_flower_destroy.
- Also, removed an extra return from cxgb4_tc_flower_destroy.
- Avoid wrapping line for netdev_err message. Also, use
consistent error message string.
Kumar Sanghvi (4):
cxgb4: add tc flower offload skeleton
cxgb4: add basic tc flower offload support
cxgb4: add support to offload action vlan
cxgb4: fetch stats for offloaded tc flower flows
drivers/net/ethernet/chelsio/cxgb4/Makefile | 4 +-
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 4 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c | 100 +++++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 25 ++
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c | 452 +++++++++++++++++++++
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h | 66 +++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h | 3 +
7 files changed, 653 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
--
2.14.1
^ permalink raw reply
* [PATCH net-next v2 1/4] cxgb4: add tc flower offload skeleton
From: Rahul Lakkireddy @ 2017-09-21 18:11 UTC (permalink / raw)
To: netdev; +Cc: davem, kumaras, ganeshgr, nirranjan, indranil, Rahul Lakkireddy
In-Reply-To: <cover.1506015856.git.rahul.lakkireddy@chelsio.com>
From: Kumar Sanghvi <kumaras@chelsio.com>
Add basic skeleton to prepare for offloading tc-flower flows.
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
---
v2:
- No changes.
drivers/net/ethernet/chelsio/cxgb4/Makefile | 4 +-
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 22 +++++++++
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c | 57 ++++++++++++++++++++++
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h | 46 +++++++++++++++++
4 files changed, 128 insertions(+), 1 deletion(-)
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
create mode 100644 drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
diff --git a/drivers/net/ethernet/chelsio/cxgb4/Makefile b/drivers/net/ethernet/chelsio/cxgb4/Makefile
index 817212702f0a..fecd7aab673b 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/Makefile
+++ b/drivers/net/ethernet/chelsio/cxgb4/Makefile
@@ -4,7 +4,9 @@
obj-$(CONFIG_CHELSIO_T4) += cxgb4.o
-cxgb4-objs := cxgb4_main.o l2t.o t4_hw.o sge.o clip_tbl.o cxgb4_ethtool.o cxgb4_uld.o sched.o cxgb4_filter.o cxgb4_tc_u32.o cxgb4_ptp.o
+cxgb4-objs := cxgb4_main.o l2t.o t4_hw.o sge.o clip_tbl.o cxgb4_ethtool.o \
+ cxgb4_uld.o sched.o cxgb4_filter.o cxgb4_tc_u32.o \
+ cxgb4_ptp.o cxgb4_tc_flower.o
cxgb4-$(CONFIG_CHELSIO_T4_DCB) += cxgb4_dcb.o
cxgb4-$(CONFIG_CHELSIO_T4_FCOE) += cxgb4_fcoe.o
cxgb4-$(CONFIG_DEBUG_FS) += cxgb4_debugfs.o
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 92d9d795d874..8923affbdaf8 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -79,6 +79,7 @@
#include "l2t.h"
#include "sched.h"
#include "cxgb4_tc_u32.h"
+#include "cxgb4_tc_flower.h"
#include "cxgb4_ptp.h"
char cxgb4_driver_name[] = KBUILD_MODNAME;
@@ -2873,6 +2874,25 @@ static int cxgb_set_tx_maxrate(struct net_device *dev, int index, u32 rate)
return err;
}
+static int cxgb_setup_tc_flower(struct net_device *dev,
+ struct tc_cls_flower_offload *cls_flower)
+{
+ if (!is_classid_clsact_ingress(cls_flower->common.classid) ||
+ cls_flower->common.chain_index)
+ return -EOPNOTSUPP;
+
+ switch (cls_flower->command) {
+ case TC_CLSFLOWER_REPLACE:
+ return cxgb4_tc_flower_replace(dev, cls_flower);
+ case TC_CLSFLOWER_DESTROY:
+ return cxgb4_tc_flower_destroy(dev, cls_flower);
+ case TC_CLSFLOWER_STATS:
+ return cxgb4_tc_flower_stats(dev, cls_flower);
+ default:
+ return -EOPNOTSUPP;
+ }
+}
+
static int cxgb_setup_tc_cls_u32(struct net_device *dev,
struct tc_cls_u32_offload *cls_u32)
{
@@ -2907,6 +2927,8 @@ static int cxgb_setup_tc(struct net_device *dev, enum tc_setup_type type,
switch (type) {
case TC_SETUP_CLSU32:
return cxgb_setup_tc_cls_u32(dev, type_data);
+ case TC_SETUP_CLSFLOWER:
+ return cxgb_setup_tc_flower(dev, type_data);
default:
return -EOPNOTSUPP;
}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
new file mode 100644
index 000000000000..16dff71e4d02
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
@@ -0,0 +1,57 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. 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
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#include <net/tc_act/tc_gact.h>
+#include <net/tc_act/tc_mirred.h>
+
+#include "cxgb4.h"
+#include "cxgb4_tc_flower.h"
+
+int cxgb4_tc_flower_replace(struct net_device *dev,
+ struct tc_cls_flower_offload *cls)
+{
+ return -EOPNOTSUPP;
+}
+
+int cxgb4_tc_flower_destroy(struct net_device *dev,
+ struct tc_cls_flower_offload *cls)
+{
+ return -EOPNOTSUPP;
+}
+
+int cxgb4_tc_flower_stats(struct net_device *dev,
+ struct tc_cls_flower_offload *cls)
+{
+ return -EOPNOTSUPP;
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
new file mode 100644
index 000000000000..b321fc205b5a
--- /dev/null
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
@@ -0,0 +1,46 @@
+/*
+ * This file is part of the Chelsio T4/T5/T6 Ethernet driver for Linux.
+ *
+ * Copyright (c) 2017 Chelsio Communications, Inc. 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
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+
+#ifndef __CXGB4_TC_FLOWER_H
+#define __CXGB4_TC_FLOWER_H
+
+#include <net/pkt_cls.h>
+
+int cxgb4_tc_flower_replace(struct net_device *dev,
+ struct tc_cls_flower_offload *cls);
+int cxgb4_tc_flower_destroy(struct net_device *dev,
+ struct tc_cls_flower_offload *cls);
+int cxgb4_tc_flower_stats(struct net_device *dev,
+ struct tc_cls_flower_offload *cls);
+#endif /* __CXGB4_TC_FLOWER_H */
--
2.14.1
^ permalink raw reply related
* [PATCH net-next v2 2/4] cxgb4: add basic tc flower offload support
From: Rahul Lakkireddy @ 2017-09-21 18:11 UTC (permalink / raw)
To: netdev; +Cc: davem, kumaras, ganeshgr, nirranjan, indranil, Rahul Lakkireddy
In-Reply-To: <cover.1506015856.git.rahul.lakkireddy@chelsio.com>
From: Kumar Sanghvi <kumaras@chelsio.com>
Add support to add/remove flows for offload. Following match
and action are supported for offloading a flow:
Match: ether-protocol, IPv4/IPv6 addresses, L4 ports (TCP/UDP)
Action: drop, redirect to another port on the device.
The qualifying flows can have accompanying mask information.
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
---
v2:
- Setting ftid to -1 not required after bitmap_find_free_region
in cxgb4_get_free_ftid.
- Direct return can be used as jumping to error path is not needed
if flower entry allocation failed in cxgb4_tc_flower_replace.
Same applies if flower entry not found in cxgb4_tc_flower_destroy.
- Also, removed an extra return from cxgb4_tc_flower_destroy.
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 3 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c | 24 ++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 2 +
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c | 280 ++++++++++++++++++++-
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h | 17 ++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h | 1 +
6 files changed, 325 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
index ea72d2d2e1b4..26eac599ab2c 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -904,6 +904,9 @@ struct adapter {
/* TC u32 offload */
struct cxgb4_tc_u32_table *tc_u32;
struct chcr_stats_debug chcr_stats;
+
+ /* TC flower offload */
+ DECLARE_HASHTABLE(flower_anymatch_tbl, 9);
};
/* Support for "sched-class" command to allow a TX Scheduling Class to be
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
index 45b5853ca2f1..a1f644eb0cec 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
@@ -148,6 +148,30 @@ static int get_filter_steerq(struct net_device *dev,
return iq;
}
+int cxgb4_get_free_ftid(struct net_device *dev, int family)
+{
+ struct adapter *adap = netdev2adap(dev);
+ struct tid_info *t = &adap->tids;
+ int ftid;
+
+ spin_lock_bh(&t->ftid_lock);
+ if (family == PF_INET) {
+ ftid = find_first_zero_bit(t->ftid_bmap, t->nftids);
+ if (ftid >= t->nftids)
+ ftid = -1;
+ } else {
+ ftid = bitmap_find_free_region(t->ftid_bmap, t->nftids, 2);
+ if (ftid < 0)
+ goto out_unlock;
+
+ /* this is only a lookup, keep the found region unallocated */
+ bitmap_release_region(t->ftid_bmap, ftid, 2);
+ }
+out_unlock:
+ spin_unlock_bh(&t->ftid_lock);
+ return ftid;
+}
+
static int cxgb4_set_ftid(struct tid_info *t, int fidx, int family)
{
spin_lock_bh(&t->ftid_lock);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 8923affbdaf8..3ba4e1ff8486 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -5105,6 +5105,8 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
if (!adapter->tc_u32)
dev_warn(&pdev->dev,
"could not offload tc u32, continuing\n");
+
+ cxgb4_init_tc_flower(adapter);
}
if (is_offload(adapter)) {
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
index 16dff71e4d02..dda34d5a52fb 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
@@ -38,16 +38,287 @@
#include "cxgb4.h"
#include "cxgb4_tc_flower.h"
+static struct ch_tc_flower_entry *allocate_flower_entry(void)
+{
+ struct ch_tc_flower_entry *new = kzalloc(sizeof(*new), GFP_KERNEL);
+ return new;
+}
+
+/* Must be called with either RTNL or rcu_read_lock */
+static struct ch_tc_flower_entry *ch_flower_lookup(struct adapter *adap,
+ unsigned long flower_cookie)
+{
+ struct ch_tc_flower_entry *flower_entry;
+
+ hash_for_each_possible_rcu(adap->flower_anymatch_tbl, flower_entry,
+ link, flower_cookie)
+ if (flower_entry->tc_flower_cookie == flower_cookie)
+ return flower_entry;
+ return NULL;
+}
+
+static void cxgb4_process_flow_match(struct net_device *dev,
+ struct tc_cls_flower_offload *cls,
+ struct ch_filter_specification *fs)
+{
+ u16 addr_type = 0;
+
+ if (dissector_uses_key(cls->dissector, FLOW_DISSECTOR_KEY_CONTROL)) {
+ struct flow_dissector_key_control *key =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_CONTROL,
+ cls->key);
+
+ addr_type = key->addr_type;
+ }
+
+ if (dissector_uses_key(cls->dissector, FLOW_DISSECTOR_KEY_BASIC)) {
+ struct flow_dissector_key_basic *key =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_BASIC,
+ cls->key);
+ struct flow_dissector_key_basic *mask =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_BASIC,
+ cls->mask);
+ u16 ethtype_key = ntohs(key->n_proto);
+ u16 ethtype_mask = ntohs(mask->n_proto);
+
+ if (ethtype_key == ETH_P_ALL) {
+ ethtype_key = 0;
+ ethtype_mask = 0;
+ }
+
+ fs->val.ethtype = ethtype_key;
+ fs->mask.ethtype = ethtype_mask;
+ fs->val.proto = key->ip_proto;
+ fs->mask.proto = mask->ip_proto;
+ }
+
+ if (addr_type == FLOW_DISSECTOR_KEY_IPV4_ADDRS) {
+ struct flow_dissector_key_ipv4_addrs *key =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_IPV4_ADDRS,
+ cls->key);
+ struct flow_dissector_key_ipv4_addrs *mask =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_IPV4_ADDRS,
+ cls->mask);
+ fs->type = 0;
+ memcpy(&fs->val.lip[0], &key->dst, sizeof(key->dst));
+ memcpy(&fs->val.fip[0], &key->src, sizeof(key->src));
+ memcpy(&fs->mask.lip[0], &mask->dst, sizeof(mask->dst));
+ memcpy(&fs->mask.fip[0], &mask->src, sizeof(mask->src));
+ }
+
+ if (addr_type == FLOW_DISSECTOR_KEY_IPV6_ADDRS) {
+ struct flow_dissector_key_ipv6_addrs *key =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_IPV6_ADDRS,
+ cls->key);
+ struct flow_dissector_key_ipv6_addrs *mask =
+ skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_IPV6_ADDRS,
+ cls->mask);
+
+ fs->type = 1;
+ memcpy(&fs->val.lip[0], key->dst.s6_addr, sizeof(key->dst));
+ memcpy(&fs->val.fip[0], key->src.s6_addr, sizeof(key->src));
+ memcpy(&fs->mask.lip[0], mask->dst.s6_addr, sizeof(mask->dst));
+ memcpy(&fs->mask.fip[0], mask->src.s6_addr, sizeof(mask->src));
+ }
+
+ if (dissector_uses_key(cls->dissector, FLOW_DISSECTOR_KEY_PORTS)) {
+ struct flow_dissector_key_ports *key, *mask;
+
+ key = skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_PORTS,
+ cls->key);
+ mask = skb_flow_dissector_target(cls->dissector,
+ FLOW_DISSECTOR_KEY_PORTS,
+ cls->mask);
+ fs->val.lport = cpu_to_be16(key->dst);
+ fs->mask.lport = cpu_to_be16(mask->dst);
+ fs->val.fport = cpu_to_be16(key->src);
+ fs->mask.fport = cpu_to_be16(mask->src);
+ }
+
+ /* Match only packets coming from the ingress port where this
+ * filter will be created.
+ */
+ fs->val.iport = netdev2pinfo(dev)->port_id;
+ fs->mask.iport = ~0;
+}
+
+static int cxgb4_validate_flow_match(struct net_device *dev,
+ struct tc_cls_flower_offload *cls)
+{
+ if (cls->dissector->used_keys &
+ ~(BIT(FLOW_DISSECTOR_KEY_CONTROL) |
+ BIT(FLOW_DISSECTOR_KEY_BASIC) |
+ BIT(FLOW_DISSECTOR_KEY_IPV4_ADDRS) |
+ BIT(FLOW_DISSECTOR_KEY_IPV6_ADDRS) |
+ BIT(FLOW_DISSECTOR_KEY_PORTS))) {
+ netdev_warn(dev, "Unsupported key used: 0x%x\n",
+ cls->dissector->used_keys);
+ return -EOPNOTSUPP;
+ }
+ return 0;
+}
+
+static void cxgb4_process_flow_actions(struct net_device *in,
+ struct tc_cls_flower_offload *cls,
+ struct ch_filter_specification *fs)
+{
+ const struct tc_action *a;
+ LIST_HEAD(actions);
+
+ tcf_exts_to_list(cls->exts, &actions);
+ list_for_each_entry(a, &actions, list) {
+ if (is_tcf_gact_shot(a)) {
+ fs->action = FILTER_DROP;
+ } else if (is_tcf_mirred_egress_redirect(a)) {
+ int ifindex = tcf_mirred_ifindex(a);
+ struct net_device *out = __dev_get_by_index(dev_net(in),
+ ifindex);
+ struct port_info *pi = netdev_priv(out);
+
+ fs->action = FILTER_SWITCH;
+ fs->eport = pi->port_id;
+ }
+ }
+}
+
+static int cxgb4_validate_flow_actions(struct net_device *dev,
+ struct tc_cls_flower_offload *cls)
+{
+ const struct tc_action *a;
+ LIST_HEAD(actions);
+
+ tcf_exts_to_list(cls->exts, &actions);
+ list_for_each_entry(a, &actions, list) {
+ if (is_tcf_gact_shot(a)) {
+ /* Do nothing */
+ } else if (is_tcf_mirred_egress_redirect(a)) {
+ struct adapter *adap = netdev2adap(dev);
+ struct net_device *n_dev;
+ unsigned int i, ifindex;
+ bool found = false;
+
+ ifindex = tcf_mirred_ifindex(a);
+ for_each_port(adap, i) {
+ n_dev = adap->port[i];
+ if (ifindex == n_dev->ifindex) {
+ found = true;
+ break;
+ }
+ }
+
+ /* If interface doesn't belong to our hw, then
+ * the provided output port is not valid
+ */
+ if (!found) {
+ netdev_err(dev, "%s: Out port invalid\n",
+ __func__);
+ return -EINVAL;
+ }
+ } else {
+ netdev_err(dev, "%s: Unsupported action\n", __func__);
+ return -EOPNOTSUPP;
+ }
+ }
+ return 0;
+}
+
int cxgb4_tc_flower_replace(struct net_device *dev,
struct tc_cls_flower_offload *cls)
{
- return -EOPNOTSUPP;
+ struct adapter *adap = netdev2adap(dev);
+ struct ch_tc_flower_entry *ch_flower;
+ struct ch_filter_specification *fs;
+ struct filter_ctx ctx;
+ int fidx;
+ int ret;
+
+ if (cxgb4_validate_flow_actions(dev, cls))
+ return -EOPNOTSUPP;
+
+ if (cxgb4_validate_flow_match(dev, cls))
+ return -EOPNOTSUPP;
+
+ ch_flower = allocate_flower_entry();
+ if (!ch_flower) {
+ netdev_err(dev, "%s: ch_flower alloc failed.\n", __func__);
+ return -ENOMEM;
+ }
+
+ fs = &ch_flower->fs;
+ fs->hitcnts = 1;
+ cxgb4_process_flow_actions(dev, cls, fs);
+ cxgb4_process_flow_match(dev, cls, fs);
+
+ fidx = cxgb4_get_free_ftid(dev, fs->type ? PF_INET6 : PF_INET);
+ if (fidx < 0) {
+ netdev_err(dev, "%s: No fidx for offload.\n", __func__);
+ ret = -ENOMEM;
+ goto free_entry;
+ }
+
+ init_completion(&ctx.completion);
+ ret = __cxgb4_set_filter(dev, fidx, fs, &ctx);
+ if (ret) {
+ netdev_err(dev, "%s: filter creation err %d\n",
+ __func__, ret);
+ goto free_entry;
+ }
+
+ /* Wait for reply */
+ ret = wait_for_completion_timeout(&ctx.completion, 10 * HZ);
+ if (!ret) {
+ ret = -ETIMEDOUT;
+ goto free_entry;
+ }
+
+ ret = ctx.result;
+ /* Check if hw returned error for filter creation */
+ if (ret) {
+ netdev_err(dev, "%s: filter creation err %d\n",
+ __func__, ret);
+ goto free_entry;
+ }
+
+ INIT_HLIST_NODE(&ch_flower->link);
+ ch_flower->tc_flower_cookie = cls->cookie;
+ ch_flower->filter_id = ctx.tid;
+ hash_add_rcu(adap->flower_anymatch_tbl, &ch_flower->link, cls->cookie);
+
+ return ret;
+
+free_entry:
+ kfree(ch_flower);
+ return ret;
}
int cxgb4_tc_flower_destroy(struct net_device *dev,
struct tc_cls_flower_offload *cls)
{
- return -EOPNOTSUPP;
+ struct adapter *adap = netdev2adap(dev);
+ struct ch_tc_flower_entry *ch_flower;
+ int ret;
+
+ ch_flower = ch_flower_lookup(adap, cls->cookie);
+ if (!ch_flower)
+ return -ENOENT;
+
+ ret = cxgb4_del_filter(dev, ch_flower->filter_id);
+ if (ret)
+ goto err;
+
+ hash_del_rcu(&ch_flower->link);
+ kfree_rcu(ch_flower, rcu);
+
+err:
+ return ret;
}
int cxgb4_tc_flower_stats(struct net_device *dev,
@@ -55,3 +326,8 @@ int cxgb4_tc_flower_stats(struct net_device *dev,
{
return -EOPNOTSUPP;
}
+
+void cxgb4_init_tc_flower(struct adapter *adap)
+{
+ hash_init(adap->flower_anymatch_tbl);
+}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
index b321fc205b5a..6145a9e056eb 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
@@ -37,10 +37,27 @@
#include <net/pkt_cls.h>
+struct ch_tc_flower_stats {
+ u64 packet_count;
+ u64 byte_count;
+ u64 last_used;
+};
+
+struct ch_tc_flower_entry {
+ struct ch_filter_specification fs;
+ struct ch_tc_flower_stats stats;
+ unsigned long tc_flower_cookie;
+ struct hlist_node link;
+ struct rcu_head rcu;
+ u32 filter_id;
+};
+
int cxgb4_tc_flower_replace(struct net_device *dev,
struct tc_cls_flower_offload *cls);
int cxgb4_tc_flower_destroy(struct net_device *dev,
struct tc_cls_flower_offload *cls);
int cxgb4_tc_flower_stats(struct net_device *dev,
struct tc_cls_flower_offload *cls);
+
+void cxgb4_init_tc_flower(struct adapter *adap);
#endif /* __CXGB4_TC_FLOWER_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
index 84541fce94c5..88487095d14f 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
@@ -212,6 +212,7 @@ struct filter_ctx {
struct ch_filter_specification;
+int cxgb4_get_free_ftid(struct net_device *dev, int family);
int __cxgb4_set_filter(struct net_device *dev, int filter_id,
struct ch_filter_specification *fs,
struct filter_ctx *ctx);
--
2.14.1
^ permalink raw reply related
* [PATCH net-next v2 3/4] cxgb4: add support to offload action vlan
From: Rahul Lakkireddy @ 2017-09-21 18:11 UTC (permalink / raw)
To: netdev; +Cc: davem, kumaras, ganeshgr, nirranjan, indranil, Rahul Lakkireddy
In-Reply-To: <cover.1506015856.git.rahul.lakkireddy@chelsio.com>
From: Kumar Sanghvi <kumaras@chelsio.com>
Add support for offloading tc-flower flows having
vlan actions: pop, push and modify.
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
---
v2:
- Avoid wrapping line for netdev_err message. Also, use
consistent error message string.
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c | 42 ++++++++++++++++++++++
1 file changed, 42 insertions(+)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
index dda34d5a52fb..e42d2efc9ea2 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
@@ -34,6 +34,7 @@
#include <net/tc_act/tc_gact.h>
#include <net/tc_act/tc_mirred.h>
+#include <net/tc_act/tc_vlan.h>
#include "cxgb4.h"
#include "cxgb4_tc_flower.h"
@@ -185,6 +186,27 @@ static void cxgb4_process_flow_actions(struct net_device *in,
fs->action = FILTER_SWITCH;
fs->eport = pi->port_id;
+ } else if (is_tcf_vlan(a)) {
+ u32 vlan_action = tcf_vlan_action(a);
+ u8 prio = tcf_vlan_push_prio(a);
+ u16 vid = tcf_vlan_push_vid(a);
+ u16 vlan_tci = (prio << VLAN_PRIO_SHIFT) | vid;
+
+ switch (vlan_action) {
+ case TCA_VLAN_ACT_POP:
+ fs->newvlan |= VLAN_REMOVE;
+ break;
+ case TCA_VLAN_ACT_PUSH:
+ fs->newvlan |= VLAN_INSERT;
+ fs->vlan = vlan_tci;
+ break;
+ case TCA_VLAN_ACT_MODIFY:
+ fs->newvlan |= VLAN_REWRITE;
+ fs->vlan = vlan_tci;
+ break;
+ default:
+ break;
+ }
}
}
}
@@ -222,6 +244,26 @@ static int cxgb4_validate_flow_actions(struct net_device *dev,
__func__);
return -EINVAL;
}
+ } else if (is_tcf_vlan(a)) {
+ u16 proto = be16_to_cpu(tcf_vlan_push_proto(a));
+ u32 vlan_action = tcf_vlan_action(a);
+
+ switch (vlan_action) {
+ case TCA_VLAN_ACT_POP:
+ break;
+ case TCA_VLAN_ACT_PUSH:
+ case TCA_VLAN_ACT_MODIFY:
+ if (proto != ETH_P_8021Q) {
+ netdev_err(dev, "%s: Unsupported vlan proto\n",
+ __func__);
+ return -EOPNOTSUPP;
+ }
+ break;
+ default:
+ netdev_err(dev, "%s: Unsupported vlan action\n",
+ __func__);
+ return -EOPNOTSUPP;
+ }
} else {
netdev_err(dev, "%s: Unsupported action\n", __func__);
return -EOPNOTSUPP;
--
2.14.1
^ permalink raw reply related
* [PATCH net-next v2 4/4] cxgb4: fetch stats for offloaded tc flower flows
From: Rahul Lakkireddy @ 2017-09-21 18:11 UTC (permalink / raw)
To: netdev; +Cc: davem, kumaras, ganeshgr, nirranjan, indranil, Rahul Lakkireddy
In-Reply-To: <cover.1506015856.git.rahul.lakkireddy@chelsio.com>
From: Kumar Sanghvi <kumaras@chelsio.com>
Add support to retrieve stats from hardware for offloaded tc flower
flows. Also, poll for the stats of offloaded flows via timer callback.
Signed-off-by: Kumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: Rahul Lakkireddy <rahul.lakkireddy@chelsio.com>
Signed-off-by: Ganesh Goudar <ganeshgr@chelsio.com>
---
v2:
- No changes.
drivers/net/ethernet/chelsio/cxgb4/cxgb4.h | 1 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c | 76 +++++++++++++++++++++
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 1 +
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c | 79 +++++++++++++++++++++-
.../net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h | 3 +
drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h | 2 +
6 files changed, 161 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
index 26eac599ab2c..8a94d97df025 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h
@@ -907,6 +907,7 @@ struct adapter {
/* TC flower offload */
DECLARE_HASHTABLE(flower_anymatch_tbl, 9);
+ struct timer_list flower_stats_timer;
};
/* Support for "sched-class" command to allow a TX Scheduling Class to be
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
index a1f644eb0cec..bdedf3dce51a 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c
@@ -148,6 +148,82 @@ static int get_filter_steerq(struct net_device *dev,
return iq;
}
+static int get_filter_count(struct adapter *adapter, unsigned int fidx,
+ u64 *pkts, u64 *bytes)
+{
+ unsigned int tcb_base, tcbaddr;
+ unsigned int word_offset;
+ struct filter_entry *f;
+ __be64 be64_byte_count;
+ int ret;
+
+ tcb_base = t4_read_reg(adapter, TP_CMM_TCB_BASE_A);
+ if ((fidx != (adapter->tids.nftids + adapter->tids.nsftids - 1)) &&
+ fidx >= adapter->tids.nftids)
+ return -E2BIG;
+
+ f = &adapter->tids.ftid_tab[fidx];
+ if (!f->valid)
+ return -EINVAL;
+
+ tcbaddr = tcb_base + f->tid * TCB_SIZE;
+
+ spin_lock(&adapter->win0_lock);
+ if (is_t4(adapter->params.chip)) {
+ __be64 be64_count;
+
+ /* T4 doesn't maintain byte counts in hw */
+ *bytes = 0;
+
+ /* Get pkts */
+ word_offset = 4;
+ ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+ tcbaddr + (word_offset * sizeof(__be32)),
+ sizeof(be64_count),
+ (__be32 *)&be64_count,
+ T4_MEMORY_READ);
+ if (ret < 0)
+ goto out;
+ *pkts = be64_to_cpu(be64_count);
+ } else {
+ __be32 be32_count;
+
+ /* Get bytes */
+ word_offset = 4;
+ ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+ tcbaddr + (word_offset * sizeof(__be32)),
+ sizeof(be64_byte_count),
+ &be64_byte_count,
+ T4_MEMORY_READ);
+ if (ret < 0)
+ goto out;
+ *bytes = be64_to_cpu(be64_byte_count);
+
+ /* Get pkts */
+ word_offset = 6;
+ ret = t4_memory_rw(adapter, MEMWIN_NIC, MEM_EDC0,
+ tcbaddr + (word_offset * sizeof(__be32)),
+ sizeof(be32_count),
+ &be32_count,
+ T4_MEMORY_READ);
+ if (ret < 0)
+ goto out;
+ *pkts = (u64)be32_to_cpu(be32_count);
+ }
+
+out:
+ spin_unlock(&adapter->win0_lock);
+ return ret;
+}
+
+int cxgb4_get_filter_counters(struct net_device *dev, unsigned int fidx,
+ u64 *hitcnt, u64 *bytecnt)
+{
+ struct adapter *adapter = netdev2adap(dev);
+
+ return get_filter_count(adapter, fidx, hitcnt, bytecnt);
+}
+
int cxgb4_get_free_ftid(struct net_device *dev, int family)
{
struct adapter *adap = netdev2adap(dev);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 3ba4e1ff8486..d634098d52ab 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -4637,6 +4637,7 @@ static void free_some_resources(struct adapter *adapter)
kvfree(adapter->l2t);
t4_cleanup_sched(adapter);
kvfree(adapter->tids.tid_tab);
+ cxgb4_cleanup_tc_flower(adapter);
cxgb4_cleanup_tc_u32(adapter);
kfree(adapter->sge.egr_map);
kfree(adapter->sge.ingr_map);
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
index e42d2efc9ea2..a36bd66d2834 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c
@@ -39,9 +39,12 @@
#include "cxgb4.h"
#include "cxgb4_tc_flower.h"
+#define STATS_CHECK_PERIOD (HZ / 2)
+
static struct ch_tc_flower_entry *allocate_flower_entry(void)
{
struct ch_tc_flower_entry *new = kzalloc(sizeof(*new), GFP_KERNEL);
+ spin_lock_init(&new->lock);
return new;
}
@@ -363,13 +366,87 @@ int cxgb4_tc_flower_destroy(struct net_device *dev,
return ret;
}
+void ch_flower_stats_cb(unsigned long data)
+{
+ struct adapter *adap = (struct adapter *)data;
+ struct ch_tc_flower_entry *flower_entry;
+ struct ch_tc_flower_stats *ofld_stats;
+ unsigned int i;
+ u64 packets;
+ u64 bytes;
+ int ret;
+
+ rcu_read_lock();
+ hash_for_each_rcu(adap->flower_anymatch_tbl, i, flower_entry, link) {
+ ret = cxgb4_get_filter_counters(adap->port[0],
+ flower_entry->filter_id,
+ &packets, &bytes);
+ if (!ret) {
+ spin_lock(&flower_entry->lock);
+ ofld_stats = &flower_entry->stats;
+
+ if (ofld_stats->prev_packet_count != packets) {
+ ofld_stats->prev_packet_count = packets;
+ ofld_stats->last_used = jiffies;
+ }
+ spin_unlock(&flower_entry->lock);
+ }
+ }
+ rcu_read_unlock();
+ mod_timer(&adap->flower_stats_timer, jiffies + STATS_CHECK_PERIOD);
+}
+
int cxgb4_tc_flower_stats(struct net_device *dev,
struct tc_cls_flower_offload *cls)
{
- return -EOPNOTSUPP;
+ struct adapter *adap = netdev2adap(dev);
+ struct ch_tc_flower_stats *ofld_stats;
+ struct ch_tc_flower_entry *ch_flower;
+ u64 packets;
+ u64 bytes;
+ int ret;
+
+ ch_flower = ch_flower_lookup(adap, cls->cookie);
+ if (!ch_flower) {
+ ret = -ENOENT;
+ goto err;
+ }
+
+ ret = cxgb4_get_filter_counters(dev, ch_flower->filter_id,
+ &packets, &bytes);
+ if (ret < 0)
+ goto err;
+
+ spin_lock_bh(&ch_flower->lock);
+ ofld_stats = &ch_flower->stats;
+ if (ofld_stats->packet_count != packets) {
+ if (ofld_stats->prev_packet_count != packets)
+ ofld_stats->last_used = jiffies;
+ tcf_exts_stats_update(cls->exts, bytes - ofld_stats->byte_count,
+ packets - ofld_stats->packet_count,
+ ofld_stats->last_used);
+
+ ofld_stats->packet_count = packets;
+ ofld_stats->byte_count = bytes;
+ ofld_stats->prev_packet_count = packets;
+ }
+ spin_unlock_bh(&ch_flower->lock);
+ return 0;
+
+err:
+ return ret;
}
void cxgb4_init_tc_flower(struct adapter *adap)
{
hash_init(adap->flower_anymatch_tbl);
+ setup_timer(&adap->flower_stats_timer, ch_flower_stats_cb,
+ (unsigned long)adap);
+ mod_timer(&adap->flower_stats_timer, jiffies + STATS_CHECK_PERIOD);
+}
+
+void cxgb4_cleanup_tc_flower(struct adapter *adap)
+{
+ if (adap->flower_stats_timer.function)
+ del_timer_sync(&adap->flower_stats_timer);
}
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
index 6145a9e056eb..604feffc752e 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.h
@@ -38,6 +38,7 @@
#include <net/pkt_cls.h>
struct ch_tc_flower_stats {
+ u64 prev_packet_count;
u64 packet_count;
u64 byte_count;
u64 last_used;
@@ -49,6 +50,7 @@ struct ch_tc_flower_entry {
unsigned long tc_flower_cookie;
struct hlist_node link;
struct rcu_head rcu;
+ spinlock_t lock; /* lock for stats */
u32 filter_id;
};
@@ -60,4 +62,5 @@ int cxgb4_tc_flower_stats(struct net_device *dev,
struct tc_cls_flower_offload *cls);
void cxgb4_init_tc_flower(struct adapter *adap);
+void cxgb4_cleanup_tc_flower(struct adapter *adap);
#endif /* __CXGB4_TC_FLOWER_H */
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
index 88487095d14f..52324c77a4fe 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h
@@ -221,6 +221,8 @@ int __cxgb4_del_filter(struct net_device *dev, int filter_id,
int cxgb4_set_filter(struct net_device *dev, int filter_id,
struct ch_filter_specification *fs);
int cxgb4_del_filter(struct net_device *dev, int filter_id);
+int cxgb4_get_filter_counters(struct net_device *dev, unsigned int fidx,
+ u64 *hitcnt, u64 *bytecnt);
static inline void set_wr_txq(struct sk_buff *skb, int prio, int queue)
{
--
2.14.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox