* Re: [[PATCH iproute2/net-next v2] 1/4] tc: flower: remove references to eth_type in manpage
From: Stephen Hemminger @ 2016-12-02 23:05 UTC (permalink / raw)
To: Simon Horman; +Cc: netdev, Paul Blakey
In-Reply-To: <1480668321-20875-2-git-send-email-simon.horman@netronome.com>
On Fri, 2 Dec 2016 09:45:18 +0100
Simon Horman <simon.horman@netronome.com> wrote:
> Remove references to eth_type and ether_type (spelling error) in
> the tc flower manpage.
>
> Also correct formatting of boldface text with whitespace.
>
> Cc: Paul Blakey <paulb@mellanox.com>
> Signed-off-by: Simon Horman <simon.horman@netronome.com>
Applied this one. Later ones still need rebase.
^ permalink raw reply
* Avoid deadlock situation due to use of xmit_lock
From: Lino Sanfilippo @ 2016-12-02 23:06 UTC (permalink / raw)
To: bh74.an, ks.giri, vipul.pandya, peppe.cavallaro, alexandre.torgue
Cc: linux-kernel, netdev
Hi,
after stumbling over a potential deadlock situation in the altera driver
(see http://marc.info/?l=linux-netdev&m=148054615230447&w=2), I checked
all other ethernet drivers for the same issue and actually found it in 2
more, namely stmmac, and sxgbe. Please see the commit messages for a
description of the problem.
These 2 patches fix the concerning drivers.
Regards,
Lino
^ permalink raw reply
* [PATCH 2/2] net: ethernet: stmmac: do not use xmit_lock in tx completion handler
From: Lino Sanfilippo @ 2016-12-02 23:06 UTC (permalink / raw)
To: bh74.an, ks.giri, vipul.pandya, peppe.cavallaro, alexandre.torgue
Cc: linux-kernel, netdev, Lino Sanfilippo
In-Reply-To: <1480719966-12839-1-git-send-email-LinoSanfilippo@gmx.de>
The driver already uses its private lock for synchronization between the
xmit function and the xmit completion handler, making the additional use of
the xmit_lock unnecessary.
Furthermore the driver does not set NETIF_F_LLTX resulting in xmit to be
called with the xmit_lock held and then taking the private lock.
On the other hand the xmit completion handler uses the reverse locking
order, by first taking the private lock, and then the xmit_lock, which
leads to the potential danger of a deadlock.
Fix this issue by not taking the xmit_lock in the completion handler.
By doing this also remove an unnecessary double check for a stopped tx
queue.
Signed-off-by: Lino Sanfilippo <LinoSanfilippo@gmx.de>
---
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 11 +++--------
1 file changed, 3 insertions(+), 8 deletions(-)
Please note that this patch is only compile tested.
diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
index 48a4e84..8def423 100644
--- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
+++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
@@ -1380,14 +1380,9 @@ static void stmmac_tx_clean(struct stmmac_priv *priv)
if (unlikely(netif_queue_stopped(priv->dev) &&
stmmac_tx_avail(priv) > STMMAC_TX_THRESH)) {
- netif_tx_lock(priv->dev);
- if (netif_queue_stopped(priv->dev) &&
- stmmac_tx_avail(priv) > STMMAC_TX_THRESH) {
- netif_dbg(priv, tx_done, priv->dev,
- "%s: restart transmit\n", __func__);
- netif_wake_queue(priv->dev);
- }
- netif_tx_unlock(priv->dev);
+ netif_dbg(priv, tx_done, priv->dev,
+ "%s: restart transmit\n", __func__);
+ netif_wake_queue(priv->dev);
}
if ((priv->eee_enabled) && (!priv->tx_path_in_lpi_mode)) {
--
1.9.1
^ permalink raw reply related
* Re: [[PATCH iproute2/net-next v2] 2/4] tc: flower: document SCTP ip_proto
From: Stephen Hemminger @ 2016-12-02 23:06 UTC (permalink / raw)
To: Simon Horman; +Cc: netdev
In-Reply-To: <1480668321-20875-3-git-send-email-simon.horman@netronome.com>
On Fri, 2 Dec 2016 09:45:19 +0100
Simon Horman <simon.horman@netronome.com> wrote:
> Add SCTP ip_proto to help text and man page.
>
> Signed-off-by: Simon Horman <simon.horman@netronome.com>
This doesn't apply cleanly to current net-next git.
Probably some of the other man page changes caused reject.
^ permalink raw reply
* [PATCH 1/2] net: ethernet: sxgbe: do not use xmit_lock in tx completion handler
From: Lino Sanfilippo @ 2016-12-02 23:06 UTC (permalink / raw)
To: bh74.an, ks.giri, vipul.pandya, peppe.cavallaro, alexandre.torgue
Cc: linux-kernel, netdev, Lino Sanfilippo
In-Reply-To: <1480719966-12839-1-git-send-email-LinoSanfilippo@gmx.de>
The driver already uses its private lock for synchronization between the
xmit function and the xmit completion handler, making the additional use of
the xmit_lock unnecessary.
Furthermore the driver does not set NETIF_F_LLTX resulting in xmit to be
called with the xmit_lock held and then taking the private lock.
On the other hand the xmit completion handler uses the reverse locking
order, by first taking the private lock, and then the xmit_lock, which
leads to the potential danger of a deadlock.
Fix this issue by not taking the xmit_lock in the completion handler.
By doing this also remove an unnecessary double check for a stopped tx
queue.
Signed-off-by: Lino Sanfilippo <LinoSanfilippo@gmx.de>
---
drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c | 11 +++--------
1 file changed, 3 insertions(+), 8 deletions(-)
Please note that this patch is only compile tested.
diff --git a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c
index 5dbe406..578cbec 100644
--- a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c
+++ b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c
@@ -782,14 +782,9 @@ static void sxgbe_tx_queue_clean(struct sxgbe_tx_queue *tqueue)
/* wake up queue */
if (unlikely(netif_tx_queue_stopped(dev_txq) &&
sxgbe_tx_avail(tqueue, tx_rsize) > SXGBE_TX_THRESH(priv))) {
- netif_tx_lock(priv->dev);
- if (netif_tx_queue_stopped(dev_txq) &&
- sxgbe_tx_avail(tqueue, tx_rsize) > SXGBE_TX_THRESH(priv)) {
- if (netif_msg_tx_done(priv))
- pr_debug("%s: restart transmit\n", __func__);
- netif_tx_wake_queue(dev_txq);
- }
- netif_tx_unlock(priv->dev);
+ if (netif_msg_tx_done(priv))
+ pr_debug("%s: restart transmit\n", __func__);
+ netif_tx_wake_queue(dev_txq);
}
spin_unlock(&tqueue->tx_lock);
--
1.9.1
^ permalink raw reply related
* Re: [PATCH/RFC iproute2/net-next 1/3] tc: flower: update headers for TCA_FLOWER_KEY_ICMP*
From: Stephen Hemminger @ 2016-12-02 23:06 UTC (permalink / raw)
To: Simon Horman; +Cc: netdev, Jamal Hadi Salim, Jiri Pirko
In-Reply-To: <1480672785-14570-2-git-send-email-simon.horman@netronome.com>
On Fri, 2 Dec 2016 10:59:43 +0100
Simon Horman <simon.horman@netronome.com> wrote:
> These are proposed changes for net-next.
>
> Signed-off-by: Simon Horman <simon.horman@netronome.com>
> ---
> include/linux/pkt_cls.h | 10 ++++++++++
> 1 file changed, 10 insertions(+)
>
> diff --git a/include/linux/pkt_cls.h b/include/linux/pkt_cls.h
> index a3d8a4f17d8e..fa435ea8ad21 100644
> --- a/include/linux/pkt_cls.h
> +++ b/include/linux/pkt_cls.h
> @@ -403,6 +403,16 @@ enum {
> TCA_FLOWER_KEY_ENC_UDP_SRC_PORT_MASK, /* be16 */
> TCA_FLOWER_KEY_ENC_UDP_DST_PORT, /* be16 */
> TCA_FLOWER_KEY_ENC_UDP_DST_PORT_MASK, /* be16 */
> +
> + TCA_FLOWER_KEY_ICMPV4_CODE, /* u8 */
> + TCA_FLOWER_KEY_ICMPV4_CODE_MASK,/* u8 */
> + TCA_FLOWER_KEY_ICMPV4_TYPE, /* u8 */
> + TCA_FLOWER_KEY_ICMPV4_TYPE_MASK,/* u8 */
> + TCA_FLOWER_KEY_ICMPV6_CODE, /* u8 */
> + TCA_FLOWER_KEY_ICMPV6_CODE_MASK,/* u8 */
> + TCA_FLOWER_KEY_ICMPV6_TYPE, /* u8 */
> + TCA_FLOWER_KEY_ICMPV6_TYPE_MASK,/* u8 */
> +
> __TCA_FLOWER_MAX,
> };
>
I picked this up by updating from kernel headers.
^ permalink raw reply
* Re: iproute2 public git outdated?
From: Stephen Hemminger @ 2016-12-02 23:10 UTC (permalink / raw)
To: Phil Sutter; +Cc: netdev
In-Reply-To: <20161201121806.GA21576@orbyte.nwl.cc>
On Thu, 1 Dec 2016 13:18:06 +0100
Phil Sutter <phil@nwl.cc> wrote:
> Hi,
>
> I am using iproute2's public git repo at this URL:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/shemminger/iproute2.git
>
> To my surprise, neither master nor net-next branches have received new
> commits since end of October. Did the repo location change or was it
> just not updated for a while?
>
> Thanks, Phil
I was on the road, and moving between houses. Sorry for the extended delay.
^ permalink raw reply
* Re: [PATCH] iproute2: ss: escape all null bytes in abstract unix domain socket
From: Stephen Hemminger @ 2016-12-02 23:18 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Isaac Boukris, davem, netdev, linux-kernel
In-Reply-To: <1480705196.18162.381.camel@edumazet-glaptop3.roam.corp.google.com>
On Fri, 02 Dec 2016 10:59:56 -0800
Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Sat, 2016-11-12 at 10:17 +0300, Stephen Hemminger wrote:
> > On Sat, 29 Oct 2016 22:20:19 +0300
> > Isaac Boukris <iboukris@gmail.com> wrote:
> >
> > > Abstract unix domain socket may embed null characters,
> > > these should be translated to '@' when printed by ss the
> > > same way the null prefix is currently being translated.
> > >
> > > Signed-off-by: Isaac Boukris <iboukris@gmail.com>
> >
> > Applied
>
> Probably not a good idea to have :
>
> for (int i = 0; i < len; i++)
> if (name[i] == '\0')
> name[i] = '@';
>
> ss.c: In function 'unix_show_sock':
> ss.c:3128:4: error: 'for' loop initial declarations are only allowed in C99 mode
> ss.c:3128:4: note: use option -std=c99 or -std=gnu99 to compile your code
> make[1]: *** [ss.o] Error 1
>
>
>
Thanks, fixed by patch from Simon
^ permalink raw reply
* Re: [PATCH net-next] liquidio: 'imply' ptp instead of 'select'
From: Nicolas Pitre @ 2016-12-02 23:20 UTC (permalink / raw)
To: Arnd Bergmann
Cc: David S. Miller, Felix Manlunas, Thomas Gleixner, David Daney,
Satanand Burla, Raghu Vatsavayi, Sunil Goutham, netdev,
linux-kernel
In-Reply-To: <20161202230451.1639318-1-arnd@arndb.de>
On Sat, 3 Dec 2016, Arnd Bergmann wrote:
> ptp now depends on the optional POSIX_TIMERS setting and fails to build
> if we select it without that:
>
> warning: (LIQUIDIO_VF && TI_CPTS) selects PTP_1588_CLOCK which has unmet direct dependencies (NET && POSIX_TIMERS)
> warning: (LIQUIDIO_VF && TI_CPTS) selects PTP_1588_CLOCK which has unmet direct dependencies (NET && POSIX_TIMERS)
> ERROR: "posix_clock_unregister" [drivers/ptp/ptp.ko] undefined!
> ERROR: "posix_clock_register" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_unregister_source" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_event" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_register_source" [drivers/ptp/ptp.ko] undefined!
>
> It seems that two patches have collided here, the build failure
> is a result of the combination. Changing the new option to 'imply'
> as well fixes it.
>
> Fixes: 111fc64a237f ("liquidio CN23XX: VF registration")
> Fixes: d1cbfd771ce8 ("ptp_clock: Allow for it to be optional")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Pitre <nico@linaro.org>
> ---
> drivers/net/ethernet/cavium/Kconfig | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/cavium/Kconfig b/drivers/net/ethernet/cavium/Kconfig
> index bbc8bd16cb97..dcbce6cac63e 100644
> --- a/drivers/net/ethernet/cavium/Kconfig
> +++ b/drivers/net/ethernet/cavium/Kconfig
> @@ -77,7 +77,7 @@ config OCTEON_MGMT_ETHERNET
> config LIQUIDIO_VF
> tristate "Cavium LiquidIO VF support"
> depends on 64BIT && PCI_MSI
> - select PTP_1588_CLOCK
> + imply PTP_1588_CLOCK
> ---help---
> This driver supports Cavium LiquidIO Intelligent Server Adapter
> based on CN23XX chips.
> --
> 2.9.0
>
>
^ permalink raw reply
* Re: [PATCH net-next] liquidio: 'imply' ptp instead of 'select'
From: David Daney @ 2016-12-02 23:20 UTC (permalink / raw)
To: Arnd Bergmann, David S. Miller
Cc: Felix Manlunas, Thomas Gleixner, David Daney, Satanand Burla,
Raghu Vatsavayi, Nicolas Pitre, Sunil Goutham, netdev,
linux-kernel
In-Reply-To: <20161202230451.1639318-1-arnd@arndb.de>
On 12/02/2016 03:04 PM, Arnd Bergmann wrote:
> ptp now depends on the optional POSIX_TIMERS setting and fails to build
> if we select it without that:
>
> warning: (LIQUIDIO_VF && TI_CPTS) selects PTP_1588_CLOCK which has unmet direct dependencies (NET && POSIX_TIMERS)
> warning: (LIQUIDIO_VF && TI_CPTS) selects PTP_1588_CLOCK which has unmet direct dependencies (NET && POSIX_TIMERS)
> ERROR: "posix_clock_unregister" [drivers/ptp/ptp.ko] undefined!
> ERROR: "posix_clock_register" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_unregister_source" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_event" [drivers/ptp/ptp.ko] undefined!
> ERROR: "pps_register_source" [drivers/ptp/ptp.ko] undefined!
>
> It seems that two patches have collided here, the build failure
> is a result of the combination. Changing the new option to 'imply'
> as well fixes it.
>
> Fixes: 111fc64a237f ("liquidio CN23XX: VF registration")
> Fixes: d1cbfd771ce8 ("ptp_clock: Allow for it to be optional")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
I didn't know about this new "imply" thing. This seems like a plausible
fix, so...
Acked-by: David Daney <david.daney@cavium.com>
Thanks for fixing this up.
> ---
> drivers/net/ethernet/cavium/Kconfig | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/cavium/Kconfig b/drivers/net/ethernet/cavium/Kconfig
> index bbc8bd16cb97..dcbce6cac63e 100644
> --- a/drivers/net/ethernet/cavium/Kconfig
> +++ b/drivers/net/ethernet/cavium/Kconfig
> @@ -77,7 +77,7 @@ config OCTEON_MGMT_ETHERNET
> config LIQUIDIO_VF
> tristate "Cavium LiquidIO VF support"
> depends on 64BIT && PCI_MSI
> - select PTP_1588_CLOCK
> + imply PTP_1588_CLOCK
> ---help---
> This driver supports Cavium LiquidIO Intelligent Server Adapter
> based on CN23XX chips.
>
^ permalink raw reply
* Re: [PATCH] iproute2: ss: escape all null bytes in abstract unix domain socket
From: Eric Dumazet @ 2016-12-02 23:24 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Isaac Boukris, davem, netdev
In-Reply-To: <20161202151810.4fab9504@xeon-e3>
On Fri, 2016-12-02 at 15:18 -0800, Stephen Hemminger wrote:
> name[i] = '@';
> >
> > ss.c: In function 'unix_show_sock':
> > ss.c:3128:4: error: 'for' loop initial declarations are only allowed in C99 mode
> > ss.c:3128:4: note: use option -std=c99 or -std=gnu99 to compile your code
> > make[1]: *** [ss.o] Error 1
> >
> >
> >
>
> Thanks, fixed by patch from Simon
Right, thanks !
^ permalink raw reply
* Re: [PATCH 2/7] net: ethernet: ti: cpdma: fix desc re-queuing
From: Ivan Khoronzhuk @ 2016-12-02 23:28 UTC (permalink / raw)
To: Grygorii Strashko
Cc: David S. Miller, netdev, Mugunthan V N, Sekhar Nori, linux-kernel,
linux-omap
In-Reply-To: <9171f321-32bf-81d9-3491-43d12e8fe8b2@ti.com>
On Fri, Dec 02, 2016 at 10:45:07AM -0600, Grygorii Strashko wrote:
>
>
> On 12/02/2016 05:03 AM, Ivan Khoronzhuk wrote:
> > On Thu, Dec 01, 2016 at 05:34:27PM -0600, Grygorii Strashko wrote:
> >> The currently processing cpdma descriptor with EOQ flag set may
> >> contain two values in Next Descriptor Pointer field:
> >> - valid pointer: means CPDMA missed addition of new desc in queue;
> > It shouldn't happen in normal circumstances, right?
>
> it might happen, because desc push compete with desc pop.
> You can check stats values:
> chan->stats.misqueued
> chan->stats.requeue
> under different types of net-loads.
I've done this, of-course.
By whole logic the misqueued counter has to cover all cases.
But that's not true.
>
> TRM:
> "
> If the pNext pointer is initially NULL, and more packets need to be queued for transmit, the software
> application may alter this pointer to point to a newly appended descriptor. The EMAC will use the new
> pointer value and proceed to the next descriptor unless the pNext value has already been read. In this
> latter case, the transmitter will halt on the transmit channel in question, and the software application may
> restart it at that time. The software can detect this case by checking for an end of queue (EOQ) condition
> flag on the updated packet descriptor when it is returned by the EMAC.
> "
That's true. No issues in desc.
In the code no any place to update next_desc except submit function.
And this case is supposed to be caught here:
For submit:
cpdma_chan_submit()
spin_lock_irqsave(&chan->lock);
...
--->__cpdma_chan_submit()
...
------> desc_write(prev, hw_next, desc_dma); // here next pointer is updated, it can be not in time
...
------> mode = desc_read(prev, hw_mode); // pay attention, it's read after updating next pointer
------> if ((mode & CPDMA_DESC_EOQ) &&
------> (chan->state == CPDMA_STATE_ACTIVE)) { // here checked if it was late update
---------> chan_write(chan, hdp, desc_dma); // here transmit is restarted, if needed
For process it only caught the fact of late update, but it has to be caught in
submit() already:
__cpdma_chan_process()
spin_lock_irqsave(&chan->lock);
------> if (mode & CPDMA_DESC_EOQ) // here transmit is restarted, if needed
---------> chan_write(chan, hdp, desc_dma); // but w/o updating next pointer
Seems there is no place where hw_next is updated w/o updating hdp :-| in case
of late hw_next set. And that is strange. I know it happens, I've checked it
before of-course. Then I thought, maybe there is some problem with write order,
thus out of sync, nothing more.
>
>
> > So, why it happens only for egress channels? And Does that mean
> > there is some resynchronization between submit and process function,
> > or this is h/w issue?
>
> no hw issues. this patch just removes one unnecessary I/O access
No objections against patch. Anyway it's better then before.
Just want to know the real reason why it happens, maybe there is smth else.
>
> >
> >> - null: no more descriptors in queue.
> >> In the later case, it's not required to write to HDP register, but now
> >> CPDMA does it.
> >>
> >> Hence, add additional check for Next Descriptor Pointer != null in
> >> cpdma_chan_process() function before writing in HDP register.
> >>
> >> Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
> >> ---
> >> drivers/net/ethernet/ti/davinci_cpdma.c | 2 +-
> >> 1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/net/ethernet/ti/davinci_cpdma.c b/drivers/net/ethernet/ti/davinci_cpdma.c
> >> index 0924014..379314f 100644
> >> --- a/drivers/net/ethernet/ti/davinci_cpdma.c
> >> +++ b/drivers/net/ethernet/ti/davinci_cpdma.c
> >> @@ -1152,7 +1152,7 @@ static int __cpdma_chan_process(struct cpdma_chan *chan)
> >> chan->count--;
> >> chan->stats.good_dequeue++;
> >>
> >> - if (status & CPDMA_DESC_EOQ) {
> >> + if ((status & CPDMA_DESC_EOQ) && chan->head) {
> >> chan->stats.requeue++;
> >> chan_write(chan, hdp, desc_phys(pool, chan->head));
> >> }
> >> --
> >> 2.10.1
> >>
>
> --
> regards,
> -grygorii
^ permalink raw reply
* [PATCH net-next v3 0/2] stmmac: dwmac-meson8b: configurable RGMII TX delay
From: Martin Blumenstingl @ 2016-12-02 23:32 UTC (permalink / raw)
To: davem-fT/PcQaiUtIeIZ0/mPfg9Q, netdev-u79uwXL29TY76Z2rM5mHXA,
devicetree-u79uwXL29TY76Z2rM5mHXA,
linux-amlogic-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
robh+dt-DgEjT+Ai2ygdnm+yROfE0A, mark.rutland-5wv7dgnIgG8,
carlo-KA+7E9HrN00dnm+yROfE0A, khilman-rdvid1DuHRBWk0Htik3J/w
Cc: peppe.cavallaro-qxv4g6HH51o, alexandre.torgue-qxv4g6HH51o,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Martin Blumenstingl
In-Reply-To: <20161125130156.17879-1-martin.blumenstingl-gM/Ye1E23mwN+BqQ9rBEUg@public.gmane.org>
Currently the dwmac-meson8b stmmac glue driver uses a hardcoded 1/4
cycle (= 2ns) TX clock delay. This seems to work fine for many boards
(for example Odroid-C2 or Amlogic's reference boards) but there are
some others where TX traffic is simply broken.
There are probably multiple reasons why it's working on some boards
while it's broken on others:
- some of Amlogic's reference boards are using a Micrel PHY
- hardware circuit design
- maybe more...
iperf3 results on my Mecool BB2 board (Meson GXM, RTL8211F PHY) with
TX clock delay disabled on the MAC (as it's enabled in the PHY driver).
TX throughput was virtually zero before:
$ iperf3 -c 192.168.1.100 -R
Connecting to host 192.168.1.100, port 5201
Reverse mode, remote host 192.168.1.100 is sending
[ 4] local 192.168.1.206 port 52828 connected to 192.168.1.100 port 5201
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-1.00 sec 108 MBytes 901 Mbits/sec
[ 4] 1.00-2.00 sec 94.2 MBytes 791 Mbits/sec
[ 4] 2.00-3.00 sec 96.5 MBytes 810 Mbits/sec
[ 4] 3.00-4.00 sec 96.2 MBytes 808 Mbits/sec
[ 4] 4.00-5.00 sec 96.6 MBytes 810 Mbits/sec
[ 4] 5.00-6.00 sec 96.5 MBytes 810 Mbits/sec
[ 4] 6.00-7.00 sec 96.6 MBytes 810 Mbits/sec
[ 4] 7.00-8.00 sec 96.5 MBytes 809 Mbits/sec
[ 4] 8.00-9.00 sec 105 MBytes 884 Mbits/sec
[ 4] 9.00-10.00 sec 111 MBytes 934 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-10.00 sec 1000 MBytes 839 Mbits/sec 0 sender
[ 4] 0.00-10.00 sec 998 MBytes 837 Mbits/sec receiver
iperf Done.
$ iperf3 -c 192.168.1.100
Connecting to host 192.168.1.100, port 5201
[ 4] local 192.168.1.206 port 52832 connected to 192.168.1.100 port 5201
[ ID] Interval Transfer Bandwidth Retr Cwnd
[ 4] 0.00-1.01 sec 99.5 MBytes 829 Mbits/sec 117 139 KBytes
[ 4] 1.01-2.00 sec 105 MBytes 884 Mbits/sec 129 70.7 KBytes
[ 4] 2.00-3.01 sec 107 MBytes 889 Mbits/sec 106 187 KBytes
[ 4] 3.01-4.01 sec 105 MBytes 878 Mbits/sec 92 143 KBytes
[ 4] 4.01-5.00 sec 105 MBytes 882 Mbits/sec 140 129 KBytes
[ 4] 5.00-6.01 sec 106 MBytes 883 Mbits/sec 115 195 KBytes
[ 4] 6.01-7.00 sec 102 MBytes 863 Mbits/sec 133 70.7 KBytes
[ 4] 7.00-8.01 sec 106 MBytes 884 Mbits/sec 143 97.6 KBytes
[ 4] 8.01-9.01 sec 104 MBytes 875 Mbits/sec 124 107 KBytes
[ 4] 9.01-10.01 sec 105 MBytes 876 Mbits/sec 90 139 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval Transfer Bandwidth Retr
[ 4] 0.00-10.01 sec 1.02 GBytes 874 Mbits/sec 1189 sender
[ 4] 0.00-10.01 sec 1.02 GBytes 873 Mbits/sec receiver
iperf Done.
I get similar TX throughput on my Meson GXBB "MXQ Pro+" board when I
disable the PHY's TX-delay and configure a 4ms TX-delay on the MAC.
So changes to at least the RTL8211F PHY driver are needed to get it
working properly in all situations.
Changes since v2:
- moved all .dts patches (3-7) to a separate series
- removed the default 2ns TX delay when phy-mode RGMII is specified
- (rebased against current net-next)
Changes since v1:
- renamed the devicetree property "amlogic,tx-delay" to
"amlogic,tx-delay-ns", which makes the .dts easier to read as we can
simply specify human-readable values instead of having "preprocessor
defines and calculation in human brain". Thanks to Andrew Lunn for
the suggestion!
- improved documentation to indicate when the MAC TX-delay should be
configured and how to use the PHY's TX-delay
- changed the default TX-delay in the dwmac-meson8b driver from 2ns
to 0ms when any of the rgmii-*id modes are used (the 2ns default
value still applies for phy-mode "rgmii")
- added patches to properly reset the PHY on Meson GXBB devices and to
use a similar configuration than the one we use on Meson GXL devices
(by passing a phy-handle to stmmac and defining the PHY in the mdio0
bus - patch 3-6)
- add the "amlogic,tx-delay-ns" property to all boards which are using
the RGMII PHY (patch 7)
Martin Blumenstingl (2):
net: dt-bindings: add RGMII TX delay configuration to meson8b-dwmac
net: stmmac: dwmac-meson8b: make the RGMII TX delay configurable
.../devicetree/bindings/net/meson-dwmac.txt | 14 ++++++++++++++
drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c | 21 +++++++++++++++------
2 files changed, 29 insertions(+), 6 deletions(-)
--
2.10.2
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH net-next v3 1/2] net: dt-bindings: add RGMII TX delay configuration to meson8b-dwmac
From: Martin Blumenstingl @ 2016-12-02 23:32 UTC (permalink / raw)
To: davem, netdev, devicetree, linux-amlogic, robh+dt, mark.rutland,
carlo, khilman
Cc: peppe.cavallaro, alexandre.torgue, linux-arm-kernel,
Martin Blumenstingl
In-Reply-To: <20161202233238.17561-1-martin.blumenstingl@googlemail.com>
This allows configuring the RGMII TX clock delay. The RGMII clock is
generated by underlying hardware of the the Meson 8b / GXBB DWMAC glue.
The configuration depends on the actual hardware (no delay may be
needed due to the design of the actual circuit, the PHY might add this
delay, etc.).
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Neil Armstrong <narmstrong@baylibre.com>
---
Documentation/devicetree/bindings/net/meson-dwmac.txt | 14 ++++++++++++++
1 file changed, 14 insertions(+)
diff --git a/Documentation/devicetree/bindings/net/meson-dwmac.txt b/Documentation/devicetree/bindings/net/meson-dwmac.txt
index 89e62dd..f8bc540 100644
--- a/Documentation/devicetree/bindings/net/meson-dwmac.txt
+++ b/Documentation/devicetree/bindings/net/meson-dwmac.txt
@@ -25,6 +25,20 @@ Required properties on Meson8b and newer:
- "clkin0" - first parent clock of the internal mux
- "clkin1" - second parent clock of the internal mux
+Optional properties on Meson8b and newer:
+- amlogic,tx-delay-ns: The internal RGMII TX clock delay (provided
+ by this driver) in nanoseconds. Allowed values
+ are: 0ns, 2ns, 4ns, 6ns.
+ This must be configured when the phy-mode is
+ "rgmii" (typically a value of 2ns is used in
+ this case).
+ When phy-mode is set to "rgmii-id" or
+ "rgmii-txid" the TX clock delay is already
+ provided by the PHY. In that case this
+ property should be set to 0ns (which disables
+ the TX clock delay in the MAC to prevent the
+ clock from going off because both PHY and MAC
+ are adding a delay).
Example for Meson6:
--
2.10.2
^ permalink raw reply related
* [PATCH net-next v3 2/2] net: stmmac: dwmac-meson8b: make the RGMII TX delay configurable
From: Martin Blumenstingl @ 2016-12-02 23:32 UTC (permalink / raw)
To: davem, netdev, devicetree, linux-amlogic, robh+dt, mark.rutland,
carlo, khilman
Cc: peppe.cavallaro, alexandre.torgue, linux-arm-kernel,
Martin Blumenstingl
In-Reply-To: <20161202233238.17561-1-martin.blumenstingl@googlemail.com>
Prior to this patch we were using a hardcoded RGMII TX clock delay of
2ns (= 1/4 cycle of the 125MHz RGMII TX clock). This value works for
many boards, but unfortunately not for all (due to the way the actual
circuit is designed, sometimes because the TX delay is enabled in the
PHY, etc.). Making the TX delay on the MAC side configurable allows us
to support all possible hardware combinations.
This allows fixing a compatibility issue on some boards, where the
RTL8211F PHY is configured to generate the TX delay. We can now turn
off the TX delay in the MAC, because otherwise we would be applying the
delay twice (which results in non-working TX traffic).
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Tested-by: Neil Armstrong <narmstrong@baylibre.com>
---
drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c | 21 +++++++++++++++------
1 file changed, 15 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c
index 250e4ce..dad31b0 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c
@@ -35,10 +35,6 @@
#define PRG_ETH0_TXDLY_SHIFT 5
#define PRG_ETH0_TXDLY_MASK GENMASK(6, 5)
-#define PRG_ETH0_TXDLY_OFF (0x0 << PRG_ETH0_TXDLY_SHIFT)
-#define PRG_ETH0_TXDLY_QUARTER (0x1 << PRG_ETH0_TXDLY_SHIFT)
-#define PRG_ETH0_TXDLY_HALF (0x2 << PRG_ETH0_TXDLY_SHIFT)
-#define PRG_ETH0_TXDLY_THREE_QUARTERS (0x3 << PRG_ETH0_TXDLY_SHIFT)
/* divider for the result of m250_sel */
#define PRG_ETH0_CLK_M250_DIV_SHIFT 7
@@ -69,6 +65,8 @@ struct meson8b_dwmac {
struct clk_divider m25_div;
struct clk *m25_div_clk;
+
+ u32 tx_delay_ns;
};
static void meson8b_dwmac_mask_bits(struct meson8b_dwmac *dwmac, u32 reg,
@@ -179,6 +177,7 @@ static int meson8b_init_prg_eth(struct meson8b_dwmac *dwmac)
{
int ret;
unsigned long clk_rate;
+ u8 tx_dly_val;
switch (dwmac->phy_mode) {
case PHY_INTERFACE_MODE_RGMII:
@@ -196,9 +195,13 @@ static int meson8b_init_prg_eth(struct meson8b_dwmac *dwmac)
meson8b_dwmac_mask_bits(dwmac, PRG_ETH0,
PRG_ETH0_INVERTED_RMII_CLK, 0);
- /* TX clock delay - all known boards use a 1/4 cycle delay */
+ /* TX clock delay in ns = "8ns / 4 * tx_dly_val" (where
+ * 8ns are exactly one cycle of the 125MHz RGMII TX clock):
+ * 0ns = 0x0, 2ns = 0x1, 4ns = 0x2, 6ns = 0x3
+ */
+ tx_dly_val = dwmac->tx_delay_ns >> 1;
meson8b_dwmac_mask_bits(dwmac, PRG_ETH0, PRG_ETH0_TXDLY_MASK,
- PRG_ETH0_TXDLY_QUARTER);
+ tx_dly_val << PRG_ETH0_TXDLY_SHIFT);
break;
case PHY_INTERFACE_MODE_RMII:
@@ -277,6 +280,12 @@ static int meson8b_dwmac_probe(struct platform_device *pdev)
if (dwmac->phy_mode < 0) {
dev_err(&pdev->dev, "missing phy-mode property\n");
return -EINVAL;
+ } else if (dwmac->phy_mode != PHY_INTERFACE_MODE_RMII) {
+ /* ignore errors as this is an optional property - by default
+ * we assume a TX delay of 0ns.
+ */
+ of_property_read_u32(pdev->dev.of_node, "amlogic,tx-delay-ns",
+ &dwmac->tx_delay_ns);
}
ret = meson8b_init_clk(dwmac);
--
2.10.2
^ permalink raw reply related
* [PATCH net-next 0/4]: Allow head adjustment in XDP prog
From: Martin KaFai Lau @ 2016-12-02 23:23 UTC (permalink / raw)
To: netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
This series adds a helper to allow head adjustment in XDP prog. mlx4
driver has been modified to support this feature. An example is written
to encapsulate a packet with an IPv4/v6 header and then XDP_TX it
out.
Thanks,
--Martin
^ permalink raw reply
* Re: bpf bounded loops. Was: [flamebait] xdp
From: Alexei Starovoitov @ 2016-12-02 23:34 UTC (permalink / raw)
To: Hannes Frederic Sowa
Cc: Tom Herbert, Thomas Graf, Linux Kernel Network Developers,
Daniel Borkmann, David S. Miller
In-Reply-To: <1480707761.3858425.806609417.2E2C5E07@webmail.messagingengine.com>
On Fri, Dec 02, 2016 at 08:42:41PM +0100, Hannes Frederic Sowa wrote:
> On Fri, Dec 2, 2016, at 20:25, Hannes Frederic Sowa wrote:
> > On 02.12.2016 19:39, Alexei Starovoitov wrote:
> > > On Thu, Dec 01, 2016 at 10:27:12PM +0100, Hannes Frederic Sowa wrote:
> > >> like") and the problematic of parsing DNS packets in XDP due to string
> > >> processing and looping inside eBPF.
> > >
> > > Hannes,
> > > Not too long ago you proposed a very interesting idea to add
> > > support for bounded loops without adding any new bpf instructions and
> > > changing llvm (which was way better than my 'rep' like instructions
> > > I was experimenting with). I thought systemtap guys also wanted bounded
> > > loops and you were cooperating on the design, so I gave up on my work and
> > > was expecting an imminent patch from you. I guess it sounds like you know
> > > believe that bounded loops are impossible or I misunderstand your statement ?
> >
> > Your argument was that it would need a new verifier as the current first
> > pass checks that we indeed can lay out the basic blocks as a DAG which
> > the second pass depends on. This would be violated.
yes. today the main part of verifier depends on cfg check that confirms DAG
property of the program. This was done as a simplification for the algorithm,
so any programmer that understands C can understand the verifier code.
It certainly was the case, since most of the people who hacked
verifier had zero compiler background.
Now I'm thinking to introduce proper compiler technologies to it.
On one side it will make the bar to understand higher and on the other
side it will cleanup the logic and reuse tens of years of data flow
analysis theory and will make verifier more robust and mathematically
solid.
> > Because eBPF is available by non privileged users this would need a lot
> > of effort to rewrite and verify (or indeed keep two verifiers in the
> > kernel for priv and non-priv). The verifier itself is exposed to
> > unprivileged users.
I certainly hear your concerns that people unfamiliar with it are simply
scared that more and more verification logic being added. So I don't mind
freezing current verifier for unpriv and let proper data flow analysis
to be done in root only component.
> > Also, by design, if we keep the current limits, this would not give you
> > more instructions to operate on compared to the flattened version of the
> > program, it would merely reduce the numbers of optimizations in LLVM
> > that let the verifier reject the program.
I think we most likely will keep 4k insn limit (since there were no
requests to increase it). The bounded loops will improve performance
and reduce I-cache misses.
> The only solution to protect the verifier, which I saw, would be to
> limit it by time and space, thus making loading of eBPF programs
> depending on how fast and hot (thermal throttling) one CPU thread is.
the verifier already has time and space limits.
See no reason to rely on physical cpu sensors.
> Those are the complexity problems I am talking and concerned about.
Do you have concerns when people implement encryption algorithm
that you're unfamiliar with?
Isn't it much bigger concern, since any bugs in the algorithm
are directly exploitable and when encryption is actually used
it's protecting sensitive data, whereas here the verifier
protects kernel from crashing.
^ permalink raw reply
* [PATCH net-next 4/4] bpf: xdp: Add XDP example for head adjustment
From: Martin KaFai Lau @ 2016-12-02 23:23 UTC (permalink / raw)
To: netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
In-Reply-To: <1480721013-1047541-1-git-send-email-kafai@fb.com>
The XDP prog checks if the incoming packet matches any VIP:PORT
combination in the BPF hashmap. If it is, it will encapsulate
the packet with a IPv4/v6 header as instructed by the value of
the BPF hashmap and then XDP_TX it out.
The VIP:PORT -> IP-Encap-Info can be specified by the cmd args
of the user prog.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
samples/bpf/Makefile | 4 +
samples/bpf/bpf_helpers.h | 2 +
samples/bpf/bpf_load.c | 94 ++++++++++++++
samples/bpf/bpf_load.h | 1 +
samples/bpf/xdp1_user.c | 93 --------------
samples/bpf/xdp_tx_iptnl_common.h | 37 ++++++
samples/bpf/xdp_tx_iptnl_kern.c | 232 ++++++++++++++++++++++++++++++++++
samples/bpf/xdp_tx_iptnl_user.c | 253 ++++++++++++++++++++++++++++++++++++++
8 files changed, 623 insertions(+), 93 deletions(-)
create mode 100644 samples/bpf/xdp_tx_iptnl_common.h
create mode 100644 samples/bpf/xdp_tx_iptnl_kern.c
create mode 100644 samples/bpf/xdp_tx_iptnl_user.c
diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index bfc2cb88a1f7..e4d6be8bd94b 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -32,6 +32,7 @@ hostprogs-y += trace_event
hostprogs-y += sampleip
hostprogs-y += tc_l2_redirect
hostprogs-y += lwt_len_hist
+hostprogs-y += xdp_tx_iptnl
test_lru_dist-objs := test_lru_dist.o libbpf.o
sock_example-objs := sock_example.o libbpf.o
@@ -65,6 +66,7 @@ trace_event-objs := bpf_load.o libbpf.o trace_event_user.o
sampleip-objs := bpf_load.o libbpf.o sampleip_user.o
tc_l2_redirect-objs := bpf_load.o libbpf.o tc_l2_redirect_user.o
lwt_len_hist-objs := bpf_load.o libbpf.o lwt_len_hist_user.o
+xdp_tx_iptnl-objs := bpf_load.o libbpf.o xdp_tx_iptnl_user.o
# Tell kbuild to always build the programs
always := $(hostprogs-y)
@@ -97,6 +99,7 @@ always += test_current_task_under_cgroup_kern.o
always += trace_event_kern.o
always += sampleip_kern.o
always += lwt_len_hist_kern.o
+always += xdp_tx_iptnl_kern.o
HOSTCFLAGS += -I$(objtree)/usr/include
HOSTCFLAGS += -I$(srctree)/tools/testing/selftests/bpf/
@@ -127,6 +130,7 @@ HOSTLOADLIBES_trace_event += -lelf
HOSTLOADLIBES_sampleip += -lelf
HOSTLOADLIBES_tc_l2_redirect += -l elf
HOSTLOADLIBES_lwt_len_hist += -l elf
+HOSTLOADLIBES_xdp_tx_iptnl += -lelf
# Allows pointing LLC/CLANG to a LLVM backend with bpf support, redefine on cmdline:
# make samples/bpf/ LLC=~/git/llvm/build/bin/llc CLANG=~/git/llvm/build/bin/clang
diff --git a/samples/bpf/bpf_helpers.h b/samples/bpf/bpf_helpers.h
index a246c6122629..8e9dca50b73a 100644
--- a/samples/bpf/bpf_helpers.h
+++ b/samples/bpf/bpf_helpers.h
@@ -57,6 +57,8 @@ static int (*bpf_skb_set_tunnel_opt)(void *ctx, void *md, int size) =
(void *) BPF_FUNC_skb_set_tunnel_opt;
static unsigned long long (*bpf_get_prandom_u32)(void) =
(void *) BPF_FUNC_get_prandom_u32;
+static int (*bpf_xdp_adjust_head)(void *ctx, int offset) =
+ (void *) BPF_FUNC_xdp_adjust_head;
/* llvm builtin functions that eBPF C program may use to
* emit BPF_LD_ABS and BPF_LD_IND instructions
diff --git a/samples/bpf/bpf_load.c b/samples/bpf/bpf_load.c
index 49b45ccbe153..e30b6de94f2e 100644
--- a/samples/bpf/bpf_load.c
+++ b/samples/bpf/bpf_load.c
@@ -12,6 +12,10 @@
#include <linux/bpf.h>
#include <linux/filter.h>
#include <linux/perf_event.h>
+#include <linux/netlink.h>
+#include <linux/rtnetlink.h>
+#include <sys/types.h>
+#include <sys/socket.h>
#include <sys/syscall.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
@@ -450,3 +454,93 @@ struct ksym *ksym_search(long key)
/* out of range. return _stext */
return &syms[0];
}
+
+int set_link_xdp_fd(int ifindex, int fd)
+{
+ struct sockaddr_nl sa;
+ int sock, seq = 0, len, ret = -1;
+ char buf[4096];
+ struct nlattr *nla, *nla_xdp;
+ struct {
+ struct nlmsghdr nh;
+ struct ifinfomsg ifinfo;
+ char attrbuf[64];
+ } req;
+ struct nlmsghdr *nh;
+ struct nlmsgerr *err;
+
+ memset(&sa, 0, sizeof(sa));
+ sa.nl_family = AF_NETLINK;
+
+ sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE);
+ if (sock < 0) {
+ printf("open netlink socket: %s\n", strerror(errno));
+ return -1;
+ }
+
+ if (bind(sock, (struct sockaddr *)&sa, sizeof(sa)) < 0) {
+ printf("bind to netlink: %s\n", strerror(errno));
+ goto cleanup;
+ }
+
+ memset(&req, 0, sizeof(req));
+ req.nh.nlmsg_len = NLMSG_LENGTH(sizeof(struct ifinfomsg));
+ req.nh.nlmsg_flags = NLM_F_REQUEST | NLM_F_ACK;
+ req.nh.nlmsg_type = RTM_SETLINK;
+ req.nh.nlmsg_pid = 0;
+ req.nh.nlmsg_seq = ++seq;
+ req.ifinfo.ifi_family = AF_UNSPEC;
+ req.ifinfo.ifi_index = ifindex;
+ nla = (struct nlattr *)(((char *)&req)
+ + NLMSG_ALIGN(req.nh.nlmsg_len));
+ nla->nla_type = NLA_F_NESTED | 43/*IFLA_XDP*/;
+
+ nla_xdp = (struct nlattr *)((char *)nla + NLA_HDRLEN);
+ nla_xdp->nla_type = 1/*IFLA_XDP_FD*/;
+ nla_xdp->nla_len = NLA_HDRLEN + sizeof(int);
+ memcpy((char *)nla_xdp + NLA_HDRLEN, &fd, sizeof(fd));
+ nla->nla_len = NLA_HDRLEN + nla_xdp->nla_len;
+
+ req.nh.nlmsg_len += NLA_ALIGN(nla->nla_len);
+
+ if (send(sock, &req, req.nh.nlmsg_len, 0) < 0) {
+ printf("send to netlink: %s\n", strerror(errno));
+ goto cleanup;
+ }
+
+ len = recv(sock, buf, sizeof(buf), 0);
+ if (len < 0) {
+ printf("recv from netlink: %s\n", strerror(errno));
+ goto cleanup;
+ }
+
+ for (nh = (struct nlmsghdr *)buf; NLMSG_OK(nh, len);
+ nh = NLMSG_NEXT(nh, len)) {
+ if (nh->nlmsg_pid != getpid()) {
+ printf("Wrong pid %d, expected %d\n",
+ nh->nlmsg_pid, getpid());
+ goto cleanup;
+ }
+ if (nh->nlmsg_seq != seq) {
+ printf("Wrong seq %d, expected %d\n",
+ nh->nlmsg_seq, seq);
+ goto cleanup;
+ }
+ switch (nh->nlmsg_type) {
+ case NLMSG_ERROR:
+ err = (struct nlmsgerr *)NLMSG_DATA(nh);
+ if (!err->error)
+ continue;
+ printf("nlmsg error %s\n", strerror(-err->error));
+ goto cleanup;
+ case NLMSG_DONE:
+ break;
+ }
+ }
+
+ ret = 0;
+
+cleanup:
+ close(sock);
+ return ret;
+}
diff --git a/samples/bpf/bpf_load.h b/samples/bpf/bpf_load.h
index 4adeeef53ad6..fb46a421ab41 100644
--- a/samples/bpf/bpf_load.h
+++ b/samples/bpf/bpf_load.h
@@ -31,4 +31,5 @@ struct ksym {
int load_kallsyms(void);
struct ksym *ksym_search(long key);
+int set_link_xdp_fd(int ifindex, int fd);
#endif
diff --git a/samples/bpf/xdp1_user.c b/samples/bpf/xdp1_user.c
index 2b2150d6d6f7..5f040a0d7712 100644
--- a/samples/bpf/xdp1_user.c
+++ b/samples/bpf/xdp1_user.c
@@ -5,111 +5,18 @@
* License as published by the Free Software Foundation.
*/
#include <linux/bpf.h>
-#include <linux/netlink.h>
-#include <linux/rtnetlink.h>
#include <assert.h>
#include <errno.h>
#include <signal.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
-#include <sys/socket.h>
#include <unistd.h>
#include "bpf_load.h"
#include "bpf_util.h"
#include "libbpf.h"
-static int set_link_xdp_fd(int ifindex, int fd)
-{
- struct sockaddr_nl sa;
- int sock, seq = 0, len, ret = -1;
- char buf[4096];
- struct nlattr *nla, *nla_xdp;
- struct {
- struct nlmsghdr nh;
- struct ifinfomsg ifinfo;
- char attrbuf[64];
- } req;
- struct nlmsghdr *nh;
- struct nlmsgerr *err;
-
- memset(&sa, 0, sizeof(sa));
- sa.nl_family = AF_NETLINK;
-
- sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE);
- if (sock < 0) {
- printf("open netlink socket: %s\n", strerror(errno));
- return -1;
- }
-
- if (bind(sock, (struct sockaddr *)&sa, sizeof(sa)) < 0) {
- printf("bind to netlink: %s\n", strerror(errno));
- goto cleanup;
- }
-
- memset(&req, 0, sizeof(req));
- req.nh.nlmsg_len = NLMSG_LENGTH(sizeof(struct ifinfomsg));
- req.nh.nlmsg_flags = NLM_F_REQUEST | NLM_F_ACK;
- req.nh.nlmsg_type = RTM_SETLINK;
- req.nh.nlmsg_pid = 0;
- req.nh.nlmsg_seq = ++seq;
- req.ifinfo.ifi_family = AF_UNSPEC;
- req.ifinfo.ifi_index = ifindex;
- nla = (struct nlattr *)(((char *)&req)
- + NLMSG_ALIGN(req.nh.nlmsg_len));
- nla->nla_type = NLA_F_NESTED | 43/*IFLA_XDP*/;
-
- nla_xdp = (struct nlattr *)((char *)nla + NLA_HDRLEN);
- nla_xdp->nla_type = 1/*IFLA_XDP_FD*/;
- nla_xdp->nla_len = NLA_HDRLEN + sizeof(int);
- memcpy((char *)nla_xdp + NLA_HDRLEN, &fd, sizeof(fd));
- nla->nla_len = NLA_HDRLEN + nla_xdp->nla_len;
-
- req.nh.nlmsg_len += NLA_ALIGN(nla->nla_len);
-
- if (send(sock, &req, req.nh.nlmsg_len, 0) < 0) {
- printf("send to netlink: %s\n", strerror(errno));
- goto cleanup;
- }
-
- len = recv(sock, buf, sizeof(buf), 0);
- if (len < 0) {
- printf("recv from netlink: %s\n", strerror(errno));
- goto cleanup;
- }
-
- for (nh = (struct nlmsghdr *)buf; NLMSG_OK(nh, len);
- nh = NLMSG_NEXT(nh, len)) {
- if (nh->nlmsg_pid != getpid()) {
- printf("Wrong pid %d, expected %d\n",
- nh->nlmsg_pid, getpid());
- goto cleanup;
- }
- if (nh->nlmsg_seq != seq) {
- printf("Wrong seq %d, expected %d\n",
- nh->nlmsg_seq, seq);
- goto cleanup;
- }
- switch (nh->nlmsg_type) {
- case NLMSG_ERROR:
- err = (struct nlmsgerr *)NLMSG_DATA(nh);
- if (!err->error)
- continue;
- printf("nlmsg error %s\n", strerror(-err->error));
- goto cleanup;
- case NLMSG_DONE:
- break;
- }
- }
-
- ret = 0;
-
-cleanup:
- close(sock);
- return ret;
-}
-
static int ifindex;
static void int_exit(int sig)
diff --git a/samples/bpf/xdp_tx_iptnl_common.h b/samples/bpf/xdp_tx_iptnl_common.h
new file mode 100644
index 000000000000..dd12cc35110f
--- /dev/null
+++ b/samples/bpf/xdp_tx_iptnl_common.h
@@ -0,0 +1,37 @@
+/* Copyright (c) 2016 Facebook
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+#ifndef _SAMPLES_BPF_XDP_TX_IPTNL_COMMON_H
+#define _SAMPLES_BPF_XDP_TX_IPTNL_COMMON_H
+
+#include <linux/types.h>
+
+#define MAX_IPTNL_ENTRIES 256U
+
+struct vip {
+ union {
+ __u32 v6[4];
+ __u32 v4;
+ } daddr;
+ __u16 dport;
+ __u16 family;
+ __u8 protocol;
+};
+
+struct iptnl_info {
+ union {
+ __u32 v6[4];
+ __u32 v4;
+ } saddr;
+ union {
+ __u32 v6[4];
+ __u32 v4;
+ } daddr;
+ __u16 family;
+ __u8 dmac[6];
+};
+
+#endif
diff --git a/samples/bpf/xdp_tx_iptnl_kern.c b/samples/bpf/xdp_tx_iptnl_kern.c
new file mode 100644
index 000000000000..a3c583f17f09
--- /dev/null
+++ b/samples/bpf/xdp_tx_iptnl_kern.c
@@ -0,0 +1,232 @@
+/* Copyright (c) 2016 Facebook
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+#include <uapi/linux/bpf.h>
+#include <linux/in.h>
+#include <linux/if_ether.h>
+#include <linux/if_packet.h>
+#include <linux/if_vlan.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include "bpf_helpers.h"
+#include "xdp_tx_iptnl_common.h"
+
+struct bpf_map_def SEC("maps") rxcnt = {
+ .type = BPF_MAP_TYPE_PERCPU_ARRAY,
+ .key_size = sizeof(__u32),
+ .value_size = sizeof(__u64),
+ .max_entries = 256,
+};
+
+struct bpf_map_def SEC("maps") vip2tnl = {
+ .type = BPF_MAP_TYPE_HASH,
+ .key_size = sizeof(struct vip),
+ .value_size = sizeof(struct iptnl_info),
+ .max_entries = MAX_IPTNL_ENTRIES,
+};
+
+static __always_inline void count_tx(u32 protocol)
+{
+ u64 *rxcnt_count;
+
+ rxcnt_count = bpf_map_lookup_elem(&rxcnt, &protocol);
+ if (rxcnt_count)
+ *rxcnt_count += 1;
+}
+
+static __always_inline int get_dport(void *trans_data, void *data_end,
+ u8 protocol)
+{
+ struct tcphdr *th;
+ struct udphdr *uh;
+
+ switch (protocol) {
+ case IPPROTO_TCP:
+ th = (struct tcphdr *)trans_data;
+ if (th + 1 > data_end)
+ return -1;
+ return th->dest;
+ case IPPROTO_UDP:
+ uh = (struct udphdr *)trans_data;
+ if (uh + 1 > data_end)
+ return -1;
+ return uh->dest;
+ default:
+ return 0;
+ }
+}
+
+static __always_inline void set_ethhdr(struct ethhdr *new_eth,
+ const struct ethhdr *old_eth,
+ const struct iptnl_info *tnl,
+ __be16 h_proto)
+{
+ memcpy(new_eth->h_source, old_eth->h_dest, sizeof(new_eth->h_source));
+ memcpy(new_eth->h_dest, tnl->dmac, sizeof(new_eth->h_dest));
+ new_eth->h_proto = h_proto;
+}
+
+static __always_inline int handle_ipv4(struct xdp_md *xdp)
+{
+ void *data_end = (void *)(long)xdp->data_end;
+ void *data = (void *)(long)xdp->data;
+ struct iptnl_info *tnl;
+ struct ethhdr *new_eth;
+ struct ethhdr *old_eth;
+ struct iphdr *iph = data + sizeof(struct ethhdr);
+ u16 *next_iph_u16;
+ u16 payload_len;
+ struct vip vip = {};
+ int dport;
+ u32 csum = 0;
+ int i;
+
+ if (iph + 1 > data_end)
+ return XDP_DROP;
+
+ dport = get_dport(iph + 1, data_end, iph->protocol);
+ if (dport == -1)
+ return XDP_DROP;
+
+ vip.protocol = iph->protocol;
+ vip.family = AF_INET;
+ vip.daddr.v4 = iph->daddr;
+ vip.dport = dport;
+ payload_len = ntohs(iph->tot_len);
+
+ tnl = bpf_map_lookup_elem(&vip2tnl, &vip);
+ /* It only does v4-in-v4 */
+ if (!tnl || tnl->family != AF_INET)
+ return XDP_PASS;
+
+ /* The vip key is found. Add an IP header and send it out */
+
+ if (bpf_xdp_adjust_head(xdp, 0 - (int)sizeof(struct iphdr)))
+ return XDP_DROP;
+
+ data = (void *)(long)xdp->data;
+ data_end = (void *)(long)xdp->data_end;
+
+ new_eth = data;
+ iph = data + sizeof(*new_eth);
+ old_eth = data + sizeof(*iph);
+
+ if (new_eth + 1 > data_end ||
+ old_eth + 1 > data_end ||
+ iph + 1 > data_end )
+ return XDP_DROP;
+
+ set_ethhdr(new_eth, old_eth, tnl, htons(ETH_P_IP));
+
+ iph->version = 4;
+ iph->ihl = sizeof(*iph) >> 2;
+ iph->frag_off = 0;
+ iph->protocol = IPPROTO_IPIP;
+ iph->check = 0;
+ iph->tos = 0;
+ iph->tot_len = htons(payload_len + sizeof(*iph));
+ iph->daddr = tnl->daddr.v4;
+ iph->saddr = tnl->saddr.v4;
+ iph->ttl = 8;
+
+ next_iph_u16 = (u16 *)iph;
+#pragma clang loop unroll(full)
+ for (i = 0; i < sizeof(*iph) >> 1; i++)
+ csum += *next_iph_u16++;
+
+ iph->check = ~((csum & 0xffff) + (csum >> 16));
+
+ count_tx(vip.protocol);
+
+ return XDP_TX;
+}
+
+static __always_inline int handle_ipv6(struct xdp_md *xdp)
+{
+ void *data_end = (void *)(long)xdp->data_end;
+ void *data = (void *)(long)xdp->data;
+ struct iptnl_info *tnl;
+ struct ethhdr *new_eth;
+ struct ethhdr *old_eth;
+ struct ipv6hdr *ip6h = data + sizeof(struct ethhdr);
+ __u16 payload_len;
+ struct vip vip = {};
+ int dport;
+
+ if (ip6h + 1 > data_end)
+ return XDP_DROP;
+
+ dport = get_dport(ip6h + 1, data_end, ip6h->nexthdr);
+ if (dport == -1)
+ return XDP_DROP;
+
+ vip.protocol = ip6h->nexthdr;
+ vip.family = AF_INET6;
+ memcpy(vip.daddr.v6, ip6h->daddr.s6_addr32, sizeof(vip.daddr));
+ vip.dport = dport;
+ payload_len = ip6h->payload_len;
+
+ tnl = bpf_map_lookup_elem(&vip2tnl, &vip);
+ /* It only does v6-in-v6 */
+ if (!tnl || tnl->family != AF_INET6)
+ return XDP_PASS;
+
+ /* The vip key is found. Add an IP header and send it out */
+
+ if (bpf_xdp_adjust_head(xdp, 0 - (int)sizeof(struct ipv6hdr)))
+ return XDP_DROP;
+
+ data = (void *)(long)xdp->data;
+ data_end = (void *)(long)xdp->data_end;
+
+ new_eth = data;
+ ip6h = data + sizeof(*new_eth);
+ old_eth = data + sizeof(*ip6h);
+
+ if (new_eth + 1 > data_end ||
+ old_eth + 1 > data_end ||
+ ip6h + 1 > data_end )
+ return XDP_DROP;
+
+ set_ethhdr(new_eth, old_eth, tnl, htons(ETH_P_IPV6));
+
+ ip6h->version = 6;
+ ip6h->priority = 0;
+ memset(ip6h->flow_lbl, 0, sizeof(ip6h->flow_lbl));
+ ip6h->payload_len = htons(ntohs(payload_len) + sizeof(*ip6h));
+ ip6h->nexthdr = IPPROTO_IPV6;
+ ip6h->hop_limit = 8;
+ memcpy(ip6h->saddr.s6_addr32, tnl->saddr.v6, sizeof(tnl->saddr.v6));
+ memcpy(ip6h->daddr.s6_addr32, tnl->daddr.v6, sizeof(tnl->daddr.v6));
+
+ count_tx(vip.protocol);
+
+ return XDP_TX;
+}
+
+SEC("xdp_tx_iptnl")
+int _xdp_tx_iptnl(struct xdp_md *xdp)
+{
+ void *data_end = (void *)(long)xdp->data_end;
+ void *data = (void *)(long)xdp->data;
+ struct ethhdr *eth = data;
+ __u16 h_proto;
+
+ if (eth + 1 > data_end)
+ return XDP_DROP;
+
+ h_proto = eth->h_proto;
+
+ if (h_proto == htons(ETH_P_IP))
+ return handle_ipv4(xdp);
+ else if (h_proto == htons(ETH_P_IPV6))
+
+ return handle_ipv6(xdp);
+ else
+ return XDP_PASS;
+}
+
+char _license[] SEC("license") = "GPL";
diff --git a/samples/bpf/xdp_tx_iptnl_user.c b/samples/bpf/xdp_tx_iptnl_user.c
new file mode 100644
index 000000000000..9aeef7579af4
--- /dev/null
+++ b/samples/bpf/xdp_tx_iptnl_user.c
@@ -0,0 +1,253 @@
+/* Copyright (c) 2016 Facebook
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+#include <linux/bpf.h>
+#include <assert.h>
+#include <errno.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/resource.h>
+#include <arpa/inet.h>
+#include <netinet/ether.h>
+#include <unistd.h>
+#include <time.h>
+#include "bpf_load.h"
+#include "libbpf.h"
+#include "bpf_util.h"
+#include "xdp_tx_iptnl_common.h"
+
+#define STATS_INTERVAL_S 2U
+
+static int ifindex = -1;
+
+static void int_exit(int sig)
+{
+ if (ifindex > -1)
+ set_link_xdp_fd(ifindex, -1);
+ exit(0);
+}
+
+/* simple per-protocol drop counter
+ */
+static void poll_stats(unsigned int kill_after_s)
+{
+ const unsigned int nr_protos = 256;
+ unsigned int nr_cpus = bpf_num_possible_cpus();
+ time_t started_at = time(NULL);
+ __u64 values[nr_cpus], prev[nr_protos][nr_cpus];
+ __u32 proto;
+ int i;
+
+ memset(prev, 0, sizeof(prev));
+
+ while (!kill_after_s || time(NULL) - started_at <= kill_after_s) {
+ sleep(STATS_INTERVAL_S);
+
+ for (proto = 0; proto < nr_protos; proto++) {
+ __u64 sum = 0;
+
+ assert(bpf_lookup_elem(map_fd[0], &proto, values) == 0);
+ for (i = 0; i < nr_cpus; i++)
+ sum += (values[i] - prev[proto][i]);
+
+ if (sum)
+ printf("proto %u: sum:%10llu pkts, rate:%10llu pkts/s\n",
+ proto, sum, sum / STATS_INTERVAL_S);
+ memcpy(prev[proto], values, sizeof(values));
+ }
+ }
+}
+
+static void usage(const char *cmd)
+{
+ printf("Usage: %s [...]\n", cmd);
+ printf(" -i <ifindex> Interface Index\n");
+ printf(" -a <vip-service-address> IPv4 or IPv6\n");
+ printf(" -p <vip-service-port> A port range (e.g. 433-444) is also allowed\n");
+ printf(" -s <source-ip> Used in the IPTunnel Header\n");
+ printf(" -d <dest-ip> Used in the IPTunnel header>\n");
+ printf(" -m <dest-MAC> Used in sending the IP Tunneled pkt>\n");
+ printf(" -T <stop-after-X-seconds> Default: 0 (forever)\n");
+ printf(" -P <IP-Protocol> Default is TCP\n");
+ printf(" -h Display this help\n");
+}
+
+static int parse_ipstr(const char *ipstr, unsigned int *addr)
+{
+ if (inet_pton(AF_INET6, ipstr, addr) == 1) {
+ return AF_INET6;
+ } else if (inet_pton(AF_INET, ipstr, addr) == 1) {
+ addr[1] = addr[2] = addr[3] = 0;
+ return AF_INET;
+ }
+
+ fprintf(stderr, "%s is an invalid IP\n", ipstr);
+ return AF_UNSPEC;
+}
+
+static int parse_ports(const char *port_str, int *min_port, int *max_port)
+{
+ char *end;
+ long tmp_min_port;
+ long tmp_max_port;
+
+ tmp_min_port = strtol(optarg, &end, 10);
+ if (tmp_min_port < 1 || tmp_min_port > 65535) {
+ fprintf(stderr, "Invalid port(s):%s\n", optarg);
+ return 1;
+ }
+
+ if (*end == '-') {
+ end++;
+ tmp_max_port = strtol(end, NULL, 10);
+ if (tmp_max_port < 1 || tmp_max_port > 65535) {
+ fprintf(stderr, "Invalid port(s):%s\n", optarg);
+ return 1;
+ }
+ } else {
+ tmp_max_port = tmp_min_port;
+ }
+
+ if (tmp_min_port > tmp_max_port) {
+ fprintf(stderr, "Invalid port(s):%s\n", optarg);
+ return 1;
+ }
+
+ if (tmp_max_port - tmp_min_port + 1 > MAX_IPTNL_ENTRIES) {
+ fprintf(stderr, "Port range (%s) is larger than %u\n",
+ port_str, MAX_IPTNL_ENTRIES);
+ return 1;
+ }
+ *min_port = tmp_min_port;
+ *max_port = tmp_max_port;
+
+ return 0;
+}
+
+int main(int argc, char **argv)
+{
+ unsigned char opt_flags[256] = {};
+ unsigned int kill_after_s = 0;
+ const char *optstr = "i:a:p:s:d:m:T:P:";
+ int min_port = 0, max_port = 0;
+ struct iptnl_info tnl = {};
+ struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY};
+ struct vip vip = {};
+ char filename[256];
+ int opt;
+ int i;
+
+ if (setrlimit(RLIMIT_MEMLOCK, &r)) {
+ perror("setrlimit(RLIMIT_MEMLOCK, RLIM_INFINITY)");
+ return 1;
+ }
+
+ snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
+
+ if (load_bpf_file(filename)) {
+ printf("%s", bpf_log_buf);
+ return 1;
+ }
+
+ if (!prog_fd[0]) {
+ printf("load_bpf_file: %s\n", strerror(errno));
+ return 1;
+ }
+
+ tnl.family = AF_UNSPEC;
+ vip.protocol = IPPROTO_TCP;
+
+ for (i = 0; i < strlen(optstr); i++)
+ if ('a' <= optstr[i] && optstr[i] <= 'z')
+ opt_flags[(unsigned char)optstr[i]] = 1;
+
+ while ((opt = getopt(argc, argv, optstr)) != -1) {
+ unsigned short family;
+ unsigned int *v6;
+
+ switch (opt) {
+ case 'i':
+ ifindex = atoi(optarg);
+ break;
+ case 'a':
+ vip.family = parse_ipstr(optarg, vip.daddr.v6);
+ if (vip.family == AF_UNSPEC)
+ return 1;
+ break;
+ case 'p':
+ if (parse_ports(optarg, &min_port, &max_port))
+ return 1;
+ break;
+ case 'P':
+ vip.protocol = atoi(optarg);
+ break;
+ case 's':
+ case 'd':
+ if (opt == 's')
+ v6 = tnl.saddr.v6;
+ else
+ v6 = tnl.daddr.v6;
+
+ family = parse_ipstr(optarg, v6);
+ if (family == AF_UNSPEC)
+ return 1;
+ if (tnl.family == AF_UNSPEC) {
+ tnl.family = family;
+ } else if (tnl.family != family) {
+ fprintf(stderr,
+ "The IP version of the src and dst addresses used in the IP encapsulation does not match\n");
+ return 1;
+ }
+ break;
+ case 'm':
+ if (!ether_aton_r(optarg,
+ (struct ether_addr *)tnl.dmac)) {
+ fprintf(stderr, "Invalid mac address:%s\n",
+ optarg);
+ return 1;
+ }
+ break;
+ case 'T':
+ kill_after_s = atoi(optarg);
+ break;
+ default:
+ usage(argv[0]);
+ return 1;
+ }
+ opt_flags[opt] = 0;
+ }
+
+ for (i = 0; i < strlen(optstr); i++) {
+ if (opt_flags[(unsigned int)optstr[i]]) {
+ fprintf(stderr, "Missing argument -%c\n", optstr[i]);
+ usage(argv[0]);
+ return 1;
+ }
+ }
+
+ signal(SIGINT, int_exit);
+
+ while (min_port <= max_port) {
+ vip.dport = htons(min_port++);
+ if (bpf_update_elem(map_fd[1], &vip, &tnl, BPF_NOEXIST)) {
+ perror("bpf_update_elem(&vip2tnl)");
+ return 1;
+ }
+ }
+
+ if (set_link_xdp_fd(ifindex, prog_fd[0]) < 0) {
+ printf("link set xdp fd failed\n");
+ return 1;
+ }
+
+ poll_stats(kill_after_s);
+
+ set_link_xdp_fd(ifindex, -1);
+
+ return 0;
+}
--
2.5.1
^ permalink raw reply related
* Re: [PATCH v2 0/7] stmmac: dwmac-meson8b: configurable RGMII TX delay
From: Martin Blumenstingl @ 2016-12-02 23:52 UTC (permalink / raw)
To: David Miller
Cc: linux-amlogic-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
devicetree-u79uwXL29TY76Z2rM5mHXA, netdev-u79uwXL29TY76Z2rM5mHXA,
khilman-rdvid1DuHRBWk0Htik3J/w, mark.rutland-5wv7dgnIgG8,
robh+dt-DgEjT+Ai2ygdnm+yROfE0A,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
alexandre.torgue-qxv4g6HH51o, peppe.cavallaro-qxv4g6HH51o,
will.deacon-5wv7dgnIgG8, catalin.marinas-5wv7dgnIgG8,
carlo-KA+7E9HrN00dnm+yROfE0A, f.fainelli-Re5JQEeQqe8AvxtiuMwx3w
In-Reply-To: <20161127.203324.1634866862730391239.davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>
On Mon, Nov 28, 2016 at 2:33 AM, David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org> wrote:
> From: Martin Blumenstingl <martin.blumenstingl-gM/Ye1E23mwN+BqQ9rBEUg@public.gmane.org>
> Date: Fri, 25 Nov 2016 14:01:49 +0100
>
>> Currently the dwmac-meson8b stmmac glue driver uses a hardcoded 1/4
>> cycle TX clock delay. This seems to work fine for many boards (for
>> example Odroid-C2 or Amlogic's reference boards) but there are some
>> others where TX traffic is simply broken.
>> There are probably multiple reasons why it's working on some boards
>> while it's broken on others:
>> - some of Amlogic's reference boards are using a Micrel PHY
>> - hardware circuit design
>> - maybe more...
>
> The ARM arch file changes do not apply cleanly to net-next, you probably
> want to merge them via the ARM tree instead of mine, and respin this series
> to be without the .dts file changes.
done, v3 contains only the net-next changes while the dts changes can
be found here: [0]
Regards,
Martin
[0] http://lists.infradead.org/pipermail/linux-amlogic/2016-December/001836.html
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH net-next 2/4] mlx4: xdp: Allow raising MTU up to one page minus eth and vlan hdrs
From: Martin KaFai Lau @ 2016-12-02 23:23 UTC (permalink / raw)
To: netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
In-Reply-To: <1480721013-1047541-1-git-send-email-kafai@fb.com>
When XDP prog is attached, it is currently limiting
MTU to be FRAG_SZ0 - ETH_HLEN - (2 * VLAN_HLEN) which is 1514
in x86.
AFAICT, since mlx4 is doing one page per packet for XDP,
we can at least raise the MTU limitation up to
PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN) which this patch is
doing. It will be useful in the next patch which allows
XDP program to extend the packet by adding new header(s).
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 28 +++++++++++-----
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 46 ++++++++++++++------------
2 files changed, 44 insertions(+), 30 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 091b904262bc..5df0bbd88d67 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -51,6 +51,8 @@
#include "mlx4_en.h"
#include "en_port.h"
+#define MLX4_EN_MAX_XDP_MTU ((int)(PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN)))
+
int mlx4_en_setup_tc(struct net_device *dev, u8 up)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
@@ -2262,6 +2264,19 @@ void mlx4_en_destroy_netdev(struct net_device *dev)
free_netdev(dev);
}
+static bool mlx4_en_check_xdp_mtu(struct net_device *dev, int mtu)
+{
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+
+ if (mtu > MLX4_EN_MAX_XDP_MTU) {
+ en_err(priv, "mtu:%d > max:%d when XDP prog is attached\n",
+ mtu, MLX4_EN_MAX_XDP_MTU);
+ return false;
+ }
+
+ return true;
+}
+
static int mlx4_en_change_mtu(struct net_device *dev, int new_mtu)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
@@ -2271,11 +2286,10 @@ static int mlx4_en_change_mtu(struct net_device *dev, int new_mtu)
en_dbg(DRV, priv, "Change MTU called - current:%d new:%d\n",
dev->mtu, new_mtu);
- if (priv->tx_ring_num[TX_XDP] && MLX4_EN_EFF_MTU(new_mtu) > FRAG_SZ0) {
- en_err(priv, "MTU size:%d requires frags but XDP running\n",
- new_mtu);
- return -EOPNOTSUPP;
- }
+ if (priv->tx_ring_num[TX_XDP] &&
+ !mlx4_en_check_xdp_mtu(dev, new_mtu))
+ return -ENOTSUPP;
+
dev->mtu = new_mtu;
if (netif_running(dev)) {
@@ -2723,10 +2737,8 @@ static int mlx4_xdp_set(struct net_device *dev, struct bpf_prog *prog)
return 0;
}
- if (priv->num_frags > 1) {
- en_err(priv, "Cannot set XDP if MTU requires multiple frags\n");
+ if (!mlx4_en_check_xdp_mtu(dev, dev->mtu))
return -EOPNOTSUPP;
- }
tmp = kzalloc(sizeof(*tmp), GFP_KERNEL);
if (!tmp)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 6562f78b07f4..23e9d04d1ef4 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -1164,37 +1164,39 @@ static const int frag_sizes[] = {
void mlx4_en_calc_rx_buf(struct net_device *dev)
{
- enum dma_data_direction dma_dir = PCI_DMA_FROMDEVICE;
struct mlx4_en_priv *priv = netdev_priv(dev);
int eff_mtu = MLX4_EN_EFF_MTU(dev->mtu);
- int order = MLX4_EN_ALLOC_PREFER_ORDER;
- u32 align = SMP_CACHE_BYTES;
- int buf_size = 0;
int i = 0;
/* bpf requires buffers to be set up as 1 packet per page.
* This only works when num_frags == 1.
*/
if (priv->tx_ring_num[TX_XDP]) {
- dma_dir = PCI_DMA_BIDIRECTIONAL;
- /* This will gain efficient xdp frame recycling at the expense
- * of more costly truesize accounting
+ priv->frag_info[0].order = 0;
+ priv->frag_info[0].frag_size = eff_mtu;
+ priv->frag_info[0].frag_prefix_size = 0;
+ /* This will gain efficient xdp frame recycling at the
+ * expense of more costly truesize accounting
*/
- align = PAGE_SIZE;
- order = 0;
- }
-
- while (buf_size < eff_mtu) {
- priv->frag_info[i].order = order;
- priv->frag_info[i].frag_size =
- (eff_mtu > buf_size + frag_sizes[i]) ?
- frag_sizes[i] : eff_mtu - buf_size;
- priv->frag_info[i].frag_prefix_size = buf_size;
- priv->frag_info[i].frag_stride =
- ALIGN(priv->frag_info[i].frag_size, align);
- priv->frag_info[i].dma_dir = dma_dir;
- buf_size += priv->frag_info[i].frag_size;
- i++;
+ priv->frag_info[0].frag_stride = PAGE_SIZE;
+ priv->frag_info[0].dma_dir = PCI_DMA_BIDIRECTIONAL;
+ i = 1;
+ } else {
+ int buf_size = 0;
+
+ while (buf_size < eff_mtu) {
+ priv->frag_info[i].order = MLX4_EN_ALLOC_PREFER_ORDER;
+ priv->frag_info[i].frag_size =
+ (eff_mtu > buf_size + frag_sizes[i]) ?
+ frag_sizes[i] : eff_mtu - buf_size;
+ priv->frag_info[i].frag_prefix_size = buf_size;
+ priv->frag_info[i].frag_stride =
+ ALIGN(priv->frag_info[i].frag_size,
+ SMP_CACHE_BYTES);
+ priv->frag_info[i].dma_dir = PCI_DMA_FROMDEVICE;
+ buf_size += priv->frag_info[i].frag_size;
+ i++;
+ }
}
priv->num_frags = i;
--
2.5.1
^ permalink raw reply related
* [PATCH net-next] tools: hv: Enable network manager for bonding scripts on RHEL
From: Haiyang Zhang @ 2016-12-02 23:55 UTC (permalink / raw)
To: davem, netdev; +Cc: olaf, haiyangz, driverdev-devel, linux-kernel
From: Haiyang Zhang <haiyangz@microsoft.com>
We found network manager is necessary on RHEL to make the synthetic
NIC, VF NIC bonding operations handled automatically. So, enabling
network manager here.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: K. Y. Srinivasan <kys@microsoft.com>
---
tools/hv/bondvf.sh | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/hv/bondvf.sh b/tools/hv/bondvf.sh
index 8e96023..4aa5369 100755
--- a/tools/hv/bondvf.sh
+++ b/tools/hv/bondvf.sh
@@ -74,8 +74,8 @@ function create_eth_cfg_redhat {
echo DEVICE=$1 >>$fn
echo TYPE=Ethernet >>$fn
echo BOOTPROTO=none >>$fn
+ echo UUID=`uuidgen` >>$fn
echo ONBOOT=yes >>$fn
- echo NM_CONTROLLED=no >>$fn
echo PEERDNS=yes >>$fn
echo IPV6INIT=yes >>$fn
echo MASTER=$2 >>$fn
@@ -93,8 +93,8 @@ function create_bond_cfg_redhat {
echo DEVICE=$1 >>$fn
echo TYPE=Bond >>$fn
echo BOOTPROTO=dhcp >>$fn
+ echo UUID=`uuidgen` >>$fn
echo ONBOOT=yes >>$fn
- echo NM_CONTROLLED=no >>$fn
echo PEERDNS=yes >>$fn
echo IPV6INIT=yes >>$fn
echo BONDING_MASTER=yes >>$fn
--
1.7.4.1
^ permalink raw reply related
* [PATCH net-next 1/4] bpf: xdp: Allow head adjustment in XDP prog
From: Martin KaFai Lau @ 2016-12-02 23:23 UTC (permalink / raw)
To: netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
In-Reply-To: <1480721013-1047541-1-git-send-email-kafai@fb.com>
This patch allows XDP prog to extend/remove the packet
data at the head (like adding or removing header). It is
done by adding a new XDP helper bpf_xdp_adjust_head().
It also renames bpf_helper_changes_skb_data() to
bpf_helper_changes_pkt_data() to better reflect
that XDP prog does not work on skb.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
arch/powerpc/net/bpf_jit_comp64.c | 4 ++--
arch/s390/net/bpf_jit_comp.c | 2 +-
arch/x86/net/bpf_jit_comp.c | 2 +-
include/linux/filter.h | 2 +-
include/uapi/linux/bpf.h | 11 ++++++++++-
kernel/bpf/core.c | 2 +-
kernel/bpf/verifier.c | 2 +-
net/core/filter.c | 34 ++++++++++++++++++++++++++++++++--
8 files changed, 49 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/net/bpf_jit_comp64.c b/arch/powerpc/net/bpf_jit_comp64.c
index 0fe98a567125..73a5cf18fd84 100644
--- a/arch/powerpc/net/bpf_jit_comp64.c
+++ b/arch/powerpc/net/bpf_jit_comp64.c
@@ -766,7 +766,7 @@ static int bpf_jit_build_body(struct bpf_prog *fp, u32 *image,
func = (u8 *) __bpf_call_base + imm;
/* Save skb pointer if we need to re-cache skb data */
- if (bpf_helper_changes_skb_data(func))
+ if (bpf_helper_changes_pkt_data(func))
PPC_BPF_STL(3, 1, bpf_jit_stack_local(ctx));
bpf_jit_emit_func_call(image, ctx, (u64)func);
@@ -775,7 +775,7 @@ static int bpf_jit_build_body(struct bpf_prog *fp, u32 *image,
PPC_MR(b2p[BPF_REG_0], 3);
/* refresh skb cache */
- if (bpf_helper_changes_skb_data(func)) {
+ if (bpf_helper_changes_pkt_data(func)) {
/* reload skb pointer to r3 */
PPC_BPF_LL(3, 1, bpf_jit_stack_local(ctx));
bpf_jit_emit_skb_loads(image, ctx);
diff --git a/arch/s390/net/bpf_jit_comp.c b/arch/s390/net/bpf_jit_comp.c
index bee281f3163d..167b31b186c1 100644
--- a/arch/s390/net/bpf_jit_comp.c
+++ b/arch/s390/net/bpf_jit_comp.c
@@ -981,7 +981,7 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, int i
EMIT2(0x0d00, REG_14, REG_W1);
/* lgr %b0,%r2: load return value into %b0 */
EMIT4(0xb9040000, BPF_REG_0, REG_2);
- if (bpf_helper_changes_skb_data((void *)func)) {
+ if (bpf_helper_changes_pkt_data((void *)func)) {
jit->seen |= SEEN_SKB_CHANGE;
/* lg %b1,ST_OFF_SKBP(%r15) */
EMIT6_DISP_LH(0xe3000000, 0x0004, BPF_REG_1, REG_0,
diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index fe04a04dab8e..e76d1af60f7a 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -853,7 +853,7 @@ xadd: if (is_imm8(insn->off))
func = (u8 *) __bpf_call_base + imm32;
jmp_offset = func - (image + addrs[i]);
if (seen_ld_abs) {
- reload_skb_data = bpf_helper_changes_skb_data(func);
+ reload_skb_data = bpf_helper_changes_pkt_data(func);
if (reload_skb_data) {
EMIT1(0x57); /* push %rdi */
jmp_offset += 22; /* pop, mov, sub, mov */
diff --git a/include/linux/filter.h b/include/linux/filter.h
index 97338134398f..3c02de77ad6a 100644
--- a/include/linux/filter.h
+++ b/include/linux/filter.h
@@ -590,7 +590,7 @@ void sk_filter_uncharge(struct sock *sk, struct sk_filter *fp);
u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog);
-bool bpf_helper_changes_skb_data(void *func);
+bool bpf_helper_changes_pkt_data(void *func);
struct bpf_prog *bpf_patch_insn_single(struct bpf_prog *prog, u32 off,
const struct bpf_insn *patch, u32 len);
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 6123d9b8e828..0eb0e87dbe9f 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -424,6 +424,12 @@ union bpf_attr {
* @len: length of header to be pushed in front
* @flags: Flags (unused for now)
* Return: 0 on success or negative error
+ *
+ * int bpf_xdp_adjust_head(xdp_md, delta)
+ * Adjust the xdp_md.data by delta
+ * @xdp_md: pointer to xdp_md
+ * @delta: An positive/negative integer to be added to xdp_md.data
+ * Return: 0 on success or negative on error
*/
#define __BPF_FUNC_MAPPER(FN) \
FN(unspec), \
@@ -469,7 +475,8 @@ union bpf_attr {
FN(csum_update), \
FN(set_hash_invalid), \
FN(get_numa_node_id), \
- FN(skb_change_head),
+ FN(skb_change_head), \
+ FN(xdp_adjust_head),
/* integer value in 'imm' field of BPF_CALL instruction selects which helper
* function eBPF program intends to call
@@ -576,6 +583,8 @@ struct bpf_sock {
__u32 protocol;
};
+#define XDP_PACKET_HEADROOM 256
+
/* User return codes for XDP prog type.
* A valid XDP program must return one of these defined values. All other
* return codes are reserved for future use. Unknown return codes will result
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 82a04143368e..871e2f398cf5 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -1078,7 +1078,7 @@ struct bpf_prog * __weak bpf_int_jit_compile(struct bpf_prog *prog)
return prog;
}
-bool __weak bpf_helper_changes_skb_data(void *func)
+bool __weak bpf_helper_changes_pkt_data(void *func)
{
return false;
}
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 8135cb1077ee..870b9d9ad11c 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -1216,7 +1216,7 @@ static int check_call(struct bpf_verifier_env *env, int func_id)
return -EINVAL;
}
- changes_data = bpf_helper_changes_skb_data(fn->func);
+ changes_data = bpf_helper_changes_pkt_data(fn->func);
memset(&meta, 0, sizeof(meta));
meta.pkt_access = fn->pkt_access;
diff --git a/net/core/filter.c b/net/core/filter.c
index 56b43587d200..6902e2f73e38 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -2234,7 +2234,34 @@ static const struct bpf_func_proto bpf_skb_change_head_proto = {
.arg3_type = ARG_ANYTHING,
};
-bool bpf_helper_changes_skb_data(void *func)
+BPF_CALL_2(bpf_xdp_adjust_head, struct xdp_buff *, xdp, int, offset)
+{
+ /* Both mlx4 and mlx5 driver align each packet to PAGE_SIZE when
+ * XDP prog is set.
+ * If the above is not true for the other drivers to support
+ * bpf_xdp_adjust_head, struct xdp_buff can be extended.
+ */
+ void *head = (void *)((unsigned long)xdp->data & PAGE_MASK);
+ void *new_data = xdp->data + offset;
+
+ if (new_data < head || new_data >= xdp->data_end)
+ /* The packet length must be >=1 */
+ return -EINVAL;
+
+ xdp->data = new_data;
+
+ return 0;
+}
+
+static const struct bpf_func_proto bpf_xdp_adjust_head_proto = {
+ .func = bpf_xdp_adjust_head,
+ .gpl_only = false,
+ .ret_type = RET_INTEGER,
+ .arg1_type = ARG_PTR_TO_CTX,
+ .arg2_type = ARG_ANYTHING,
+};
+
+bool bpf_helper_changes_pkt_data(void *func)
{
if (func == bpf_skb_vlan_push ||
func == bpf_skb_vlan_pop ||
@@ -2244,7 +2271,8 @@ bool bpf_helper_changes_skb_data(void *func)
func == bpf_skb_change_tail ||
func == bpf_skb_pull_data ||
func == bpf_l3_csum_replace ||
- func == bpf_l4_csum_replace)
+ func == bpf_l4_csum_replace ||
+ func == bpf_xdp_adjust_head)
return true;
return false;
@@ -2670,6 +2698,8 @@ xdp_func_proto(enum bpf_func_id func_id)
return &bpf_xdp_event_output_proto;
case BPF_FUNC_get_smp_processor_id:
return &bpf_get_smp_processor_id_proto;
+ case BPF_FUNC_xdp_adjust_head:
+ return &bpf_xdp_adjust_head_proto;
default:
return sk_filter_func_proto(func_id);
}
--
2.5.1
^ permalink raw reply related
* Re: [PATCH net-next 2/4] mlx4: xdp: Allow raising MTU up to one page minus eth and vlan hdrs
From: Rick Jones @ 2016-12-03 0:07 UTC (permalink / raw)
To: Martin KaFai Lau, netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
In-Reply-To: <1480721013-1047541-3-git-send-email-kafai@fb.com>
On 12/02/2016 03:23 PM, Martin KaFai Lau wrote:
> When XDP prog is attached, it is currently limiting
> MTU to be FRAG_SZ0 - ETH_HLEN - (2 * VLAN_HLEN) which is 1514
> in x86.
>
> AFAICT, since mlx4 is doing one page per packet for XDP,
> we can at least raise the MTU limitation up to
> PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN) which this patch is
> doing. It will be useful in the next patch which allows
> XDP program to extend the packet by adding new header(s).
Is mlx4 the only driver doing page-per-packet?
rick jones
^ permalink raw reply
* [PATCH net-next 3/4] mlx4: xdp: Reserve headroom for receiving packet when XDP prog is active
From: Martin KaFai Lau @ 2016-12-02 23:23 UTC (permalink / raw)
To: netdev
Cc: Alexei Starovoitov, Brenden Blanco, Daniel Borkmann, David Miller,
Saeed Mahameed, Tariq Toukan, Kernel Team
In-Reply-To: <1480721013-1047541-1-git-send-email-kafai@fb.com>
Reserve XDP_PACKET_HEADROOM when XDP prog is active.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
---
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 17 +++++++++++++++--
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 23 +++++++++++++++++------
drivers/net/ethernet/mellanox/mlx4/en_tx.c | 9 +++++----
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 3 ++-
4 files changed, 39 insertions(+), 13 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 5df0bbd88d67..fb6d87dbc350 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -51,7 +51,8 @@
#include "mlx4_en.h"
#include "en_port.h"
-#define MLX4_EN_MAX_XDP_MTU ((int)(PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN)))
+#define MLX4_EN_MAX_XDP_MTU ((int)(PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN) - \
+ XDP_PACKET_HEADROOM))
int mlx4_en_setup_tc(struct net_device *dev, u8 up)
{
@@ -1551,6 +1552,7 @@ int mlx4_en_start_port(struct net_device *dev)
struct mlx4_en_tx_ring *tx_ring;
int rx_index = 0;
int err = 0;
+ int mtu;
int i, t;
int j;
u8 mc_list[16] = {0};
@@ -1684,8 +1686,12 @@ int mlx4_en_start_port(struct net_device *dev)
}
/* Configure port */
+ mtu = priv->rx_skb_size + ETH_FCS_LEN;
+ if (priv->tx_ring_num[TX_XDP])
+ mtu += XDP_PACKET_HEADROOM;
+
err = mlx4_SET_PORT_general(mdev->dev, priv->port,
- priv->rx_skb_size + ETH_FCS_LEN,
+ mtu,
priv->prof->tx_pause,
priv->prof->tx_ppp,
priv->prof->rx_pause,
@@ -2268,6 +2274,13 @@ static bool mlx4_en_check_xdp_mtu(struct net_device *dev, int mtu)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
+ if (mtu + XDP_PACKET_HEADROOM > priv->max_mtu) {
+ en_err(priv,
+ "Device max mtu:%d does not allow %d bytes reserved headroom for XDP prog\n",
+ priv->max_mtu, XDP_PACKET_HEADROOM);
+ return false;
+ }
+
if (mtu > MLX4_EN_MAX_XDP_MTU) {
en_err(priv, "mtu:%d > max:%d when XDP prog is attached\n",
mtu, MLX4_EN_MAX_XDP_MTU);
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 23e9d04d1ef4..324771ac929e 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -96,7 +96,6 @@ static int mlx4_en_alloc_frags(struct mlx4_en_priv *priv,
struct mlx4_en_rx_alloc page_alloc[MLX4_EN_MAX_RX_FRAGS];
const struct mlx4_en_frag_info *frag_info;
struct page *page;
- dma_addr_t dma;
int i;
for (i = 0; i < priv->num_frags; i++) {
@@ -115,9 +114,10 @@ static int mlx4_en_alloc_frags(struct mlx4_en_priv *priv,
for (i = 0; i < priv->num_frags; i++) {
frags[i] = ring_alloc[i];
- dma = ring_alloc[i].dma + ring_alloc[i].page_offset;
+ frags[i].page_offset += priv->frag_info[i].rx_headroom;
+ rx_desc->data[i].addr = cpu_to_be64(frags[i].dma +
+ frags[i].page_offset);
ring_alloc[i] = page_alloc[i];
- rx_desc->data[i].addr = cpu_to_be64(dma);
}
return 0;
@@ -250,7 +250,8 @@ static int mlx4_en_prepare_rx_desc(struct mlx4_en_priv *priv,
if (ring->page_cache.index > 0) {
frags[0] = ring->page_cache.buf[--ring->page_cache.index];
- rx_desc->data[0].addr = cpu_to_be64(frags[0].dma);
+ rx_desc->data[0].addr = cpu_to_be64(frags[0].dma +
+ frags[0].page_offset);
return 0;
}
@@ -889,6 +890,7 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
if (xdp_prog) {
struct xdp_buff xdp;
dma_addr_t dma;
+ void *pg_addr, *orig_data;
u32 act;
dma = be64_to_cpu(rx_desc->data[0].addr);
@@ -896,11 +898,18 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
priv->frag_info[0].frag_size,
DMA_FROM_DEVICE);
- xdp.data = page_address(frags[0].page) +
- frags[0].page_offset;
+ pg_addr = page_address(frags[0].page);
+ orig_data = pg_addr + frags[0].page_offset;
+ xdp.data = orig_data;
xdp.data_end = xdp.data + length;
act = bpf_prog_run_xdp(xdp_prog, &xdp);
+
+ if (xdp.data != orig_data) {
+ length = xdp.data_end - xdp.data;
+ frags[0].page_offset = xdp.data - pg_addr;
+ }
+
switch (act) {
case XDP_PASS:
break;
@@ -1180,6 +1189,7 @@ void mlx4_en_calc_rx_buf(struct net_device *dev)
*/
priv->frag_info[0].frag_stride = PAGE_SIZE;
priv->frag_info[0].dma_dir = PCI_DMA_BIDIRECTIONAL;
+ priv->frag_info[0].rx_headroom = XDP_PACKET_HEADROOM;
i = 1;
} else {
int buf_size = 0;
@@ -1194,6 +1204,7 @@ void mlx4_en_calc_rx_buf(struct net_device *dev)
ALIGN(priv->frag_info[i].frag_size,
SMP_CACHE_BYTES);
priv->frag_info[i].dma_dir = PCI_DMA_FROMDEVICE;
+ priv->frag_info[i].rx_headroom = 0;
buf_size += priv->frag_info[i].frag_size;
i++;
}
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index 4b597dca5c52..9e5f38cefe5f 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -354,7 +354,7 @@ u32 mlx4_en_recycle_tx_desc(struct mlx4_en_priv *priv,
struct mlx4_en_rx_alloc frame = {
.page = tx_info->page,
.dma = tx_info->map0_dma,
- .page_offset = 0,
+ .page_offset = XDP_PACKET_HEADROOM,
.page_size = PAGE_SIZE,
};
@@ -1132,7 +1132,7 @@ netdev_tx_t mlx4_en_xmit_frame(struct mlx4_en_rx_ring *rx_ring,
tx_info->page = frame->page;
frame->page = NULL;
tx_info->map0_dma = dma;
- tx_info->map0_byte_count = length;
+ tx_info->map0_byte_count = length + frame->page_offset;
tx_info->nr_txbb = nr_txbb;
tx_info->nr_bytes = max_t(unsigned int, length, ETH_ZLEN);
tx_info->data_offset = (void *)data - (void *)tx_desc;
@@ -1141,9 +1141,10 @@ netdev_tx_t mlx4_en_xmit_frame(struct mlx4_en_rx_ring *rx_ring,
tx_info->linear = 1;
tx_info->inl = 0;
- dma_sync_single_for_device(priv->ddev, dma, length, PCI_DMA_TODEVICE);
+ dma_sync_single_range_for_device(priv->ddev, dma, frame->page_offset,
+ length, PCI_DMA_TODEVICE);
- data->addr = cpu_to_be64(dma);
+ data->addr = cpu_to_be64(dma + frame->page_offset);
data->lkey = ring->mr_key;
dma_wmb();
data->byte_count = cpu_to_be32(length);
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index 20a936428f4a..ba1c6cd0cc79 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -475,7 +475,8 @@ struct mlx4_en_frag_info {
u16 frag_prefix_size;
u32 frag_stride;
enum dma_data_direction dma_dir;
- int order;
+ u16 order;
+ u16 rx_headroom;
};
#ifdef CONFIG_MLX4_EN_DCB
--
2.5.1
^ permalink raw reply related
* Re: bpf bounded loops. Was: [flamebait] xdp
From: Alexei Starovoitov @ 2016-12-03 0:20 UTC (permalink / raw)
To: John Fastabend
Cc: Hannes Frederic Sowa, Tom Herbert, Thomas Graf,
Linux Kernel Network Developers, Daniel Borkmann, David S. Miller
In-Reply-To: <5841CE97.3050302@gmail.com>
On Fri, Dec 02, 2016 at 11:42:15AM -0800, John Fastabend wrote:
> >> As far as pattern search for DNS packets...
> >> it was requested by Cloudflare guys back in March:
> >> https://github.com/iovisor/bcc/issues/471
> >> and it is useful for several tracing use cases as well.
> >> Unfortunately no one had time to implement it yet.
> >
> > The string operations you proposed on the other hand, which would count
> > as one eBPF instructions, would give a lot more flexibility and allow
> > more cycles to burn, but don't help parsing binary protocols like IPv6
> > extension headers.
these are two separate things. we need pattern search regardless
of bounded loops. bpf program shouldn't be doing any complicated
algorithms. The main reasons to have loops are:
- speed up execution (smaller I-cache footprint)
- avoid forcing compiler to unroll loops (easier for users)
- support loops where unroll is not possible (like example below)
> My rough thinking on this was the verifier had to start looking for loop
> invariants and to guarantee termination. Sounds scary in general but
> LLVM could put these in some normal form for us and the verifier could
> only accept decreasing loops, the invariants could be required to be
> integers, etc. By simplifying the loop enough the problem becomes
> tractable.
yep. I think what Hannes was proposing earlier is straighforward
to implement for a compiler guy. The following:
for (int i = 0; i < (var & 0xff); i++)
sum += map->value[i]; /* map value_size >= 0xff */
is obviously bounded and dataflow analysis can easily prove
that all memory operations are valid.
Static analysis tools do way way more than this.
> I think this would be better than new instructions and/or multiple
> verifiers.
agree that it's better than new instructions that would have
required JIT changes. Though there are pros to new insns too :)
^ permalink raw reply
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