* [PATCH net] net: Don't copy pfmemalloc flag in __copy_skb_header()
From: Stefano Brivio @ 2018-07-11 12:39 UTC (permalink / raw)
To: David S . Miller
Cc: Mel Gorman, Eric Dumazet, Florian Westphal, Patrick Talbert,
netdev
The pfmemalloc flag indicates that the skb was allocated from
the PFMEMALLOC reserves, and the flag is currently copied on skb
copy and clone.
However, an skb copied from an skb flagged with pfmemalloc
wasn't necessarily allocated from PFMEMALLOC reserves, and on
the other hand an skb allocated that way might be copied from an
skb that wasn't.
So we should not copy the flag on skb copy, and rather decide
whether to allow an skb to be associated with sockets unrelated
to page reclaim depending only on how it was allocated.
Move the pfmemalloc flag before headers_start[0] using an
existing 1-bit hole, so that __copy_skb_header() doesn't copy
it.
When cloning, we'll now take care of this flag explicitly,
contravening to the warning comment of __skb_clone().
While at it, restore the newline usage introduced by commit
b19372273164 ("net: reorganize sk_buff for faster
__copy_skb_header()") to visually separate bytes used in
bitfields after headers_start[0], that was gone after commit
a9e419dc7be6 ("netfilter: merge ctinfo into nfct pointer storage
area"), and describe the pfmemalloc flag in the kernel-doc
structure comment.
This doesn't change the size of sk_buff or cacheline boundaries,
but consolidates the 15 bits hole before tc_index into a 2 bytes
hole before csum, that could now be filled more easily.
Reported-by: Patrick Talbert <ptalbert@redhat.com>
Fixes: c93bdd0e03e8 ("netvm: allow skb allocation to use PFMEMALLOC reserves")
Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
---
include/linux/skbuff.h | 10 +++++-----
net/core/skbuff.c | 2 ++
2 files changed, 7 insertions(+), 5 deletions(-)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 164cdedf6012..610a201126ee 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -630,6 +630,7 @@ typedef unsigned char *sk_buff_data_t;
* @hash: the packet hash
* @queue_mapping: Queue mapping for multiqueue devices
* @xmit_more: More SKBs are pending for this queue
+ * @pfmemalloc: skbuff was allocated from PFMEMALLOC reserves
* @ndisc_nodetype: router type (from link layer)
* @ooo_okay: allow the mapping of a socket to a queue to be changed
* @l4_hash: indicate hash is a canonical 4-tuple hash over transport
@@ -735,7 +736,7 @@ struct sk_buff {
peeked:1,
head_frag:1,
xmit_more:1,
- __unused:1; /* one bit hole */
+ pfmemalloc:1;
/* fields enclosed in headers_start/headers_end are copied
* using a single memcpy() in __copy_skb_header()
@@ -754,31 +755,30 @@ struct sk_buff {
__u8 __pkt_type_offset[0];
__u8 pkt_type:3;
- __u8 pfmemalloc:1;
__u8 ignore_df:1;
-
__u8 nf_trace:1;
__u8 ip_summed:2;
__u8 ooo_okay:1;
+
__u8 l4_hash:1;
__u8 sw_hash:1;
__u8 wifi_acked_valid:1;
__u8 wifi_acked:1;
-
__u8 no_fcs:1;
/* Indicates the inner headers are valid in the skbuff. */
__u8 encapsulation:1;
__u8 encap_hdr_csum:1;
__u8 csum_valid:1;
+
__u8 csum_complete_sw:1;
__u8 csum_level:2;
__u8 csum_not_inet:1;
-
__u8 dst_pending_confirm:1;
#ifdef CONFIG_IPV6_NDISC_NODETYPE
__u8 ndisc_nodetype:2;
#endif
__u8 ipvs_property:1;
+
__u8 inner_protocol_type:1;
__u8 remcsum_offload:1;
#ifdef CONFIG_NET_SWITCHDEV
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index eba8dae22c25..4df3164bb5fc 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -858,6 +858,8 @@ static struct sk_buff *__skb_clone(struct sk_buff *n, struct sk_buff *skb)
n->cloned = 1;
n->nohdr = 0;
n->peeked = 0;
+ if (skb->pfmemalloc)
+ n->pfmemalloc = 1;
n->destructor = NULL;
C(tail);
C(end);
--
2.15.1
^ permalink raw reply related
* [PATCH 2/2] [RESEND] nfp: avoid using getnstimeofday64()
From: Arnd Bergmann @ 2018-07-11 12:29 UTC (permalink / raw)
To: Jakub Kicinski, David S. Miller
Cc: Arnd Bergmann, John Hurley, Simon Horman,
Pieter Jansen van Vuuren, Jiri Pirko, oss-drivers, netdev,
linux-kernel
In-Reply-To: <20180711123003.453442-1-arnd@arndb.de>
getnstimeofday64 is deprecated in favor of the ktime_get() family of
functions. The direct replacement would be ktime_get_real_ts64(),
but I'm picking the basic ktime_get() instead:
- using a ktime_t simplifies the code compared to timespec64
- using monotonic time instead of real time avoids issues caused
by a concurrent settimeofday() or during a leap second adjustment.
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
drivers/net/ethernet/netronome/nfp/flower/main.h | 2 +-
drivers/net/ethernet/netronome/nfp/flower/metadata.c | 12 +++++-------
2 files changed, 6 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.h b/drivers/net/ethernet/netronome/nfp/flower/main.h
index bbe5764d26cb..ef2114d13387 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/main.h
+++ b/drivers/net/ethernet/netronome/nfp/flower/main.h
@@ -73,7 +73,7 @@ struct nfp_app;
struct nfp_fl_mask_id {
struct circ_buf mask_id_free_list;
- struct timespec64 *last_used;
+ ktime_t *last_used;
u8 init_unallocated;
};
diff --git a/drivers/net/ethernet/netronome/nfp/flower/metadata.c b/drivers/net/ethernet/netronome/nfp/flower/metadata.c
index 93fb809f50d1..c098730544b7 100644
--- a/drivers/net/ethernet/netronome/nfp/flower/metadata.c
+++ b/drivers/net/ethernet/netronome/nfp/flower/metadata.c
@@ -158,7 +158,6 @@ static int nfp_release_mask_id(struct nfp_app *app, u8 mask_id)
{
struct nfp_flower_priv *priv = app->priv;
struct circ_buf *ring;
- struct timespec64 now;
ring = &priv->mask_ids.mask_id_free_list;
/* Checking if buffer is full. */
@@ -169,8 +168,7 @@ static int nfp_release_mask_id(struct nfp_app *app, u8 mask_id)
ring->head = (ring->head + NFP_FLOWER_MASK_ELEMENT_RS) %
(NFP_FLOWER_MASK_ENTRY_RS * NFP_FLOWER_MASK_ELEMENT_RS);
- getnstimeofday64(&now);
- priv->mask_ids.last_used[mask_id] = now;
+ priv->mask_ids.last_used[mask_id] = ktime_get();
return 0;
}
@@ -178,7 +176,7 @@ static int nfp_release_mask_id(struct nfp_app *app, u8 mask_id)
static int nfp_mask_alloc(struct nfp_app *app, u8 *mask_id)
{
struct nfp_flower_priv *priv = app->priv;
- struct timespec64 delta, now;
+ ktime_t reuse_timeout;
struct circ_buf *ring;
u8 temp_id, freed_id;
@@ -198,10 +196,10 @@ static int nfp_mask_alloc(struct nfp_app *app, u8 *mask_id)
memcpy(&temp_id, &ring->buf[ring->tail], NFP_FLOWER_MASK_ELEMENT_RS);
*mask_id = temp_id;
- getnstimeofday64(&now);
- delta = timespec64_sub(now, priv->mask_ids.last_used[*mask_id]);
+ reuse_timeout = ktime_add_ns(priv->mask_ids.last_used[*mask_id],
+ NFP_FL_MASK_REUSE_TIME_NS);
- if (timespec64_to_ns(&delta) < NFP_FL_MASK_REUSE_TIME_NS)
+ if (ktime_before(ktime_get(), reuse_timeout))
goto err_not_found;
memcpy(&ring->buf[ring->tail], &freed_id, NFP_FLOWER_MASK_ELEMENT_RS);
--
2.9.0
^ permalink raw reply related
* [PATCH 1/2] [RESEND] liquidio: use ktime_get_real_ts64() instead of getnstimeofday64()
From: Arnd Bergmann @ 2018-07-11 12:29 UTC (permalink / raw)
To: Derek Chickles, Satanand Burla, Felix Manlunas, Raghu Vatsavayi,
David S. Miller
Cc: Arnd Bergmann, Intiyaz Basha, Rick Farrington,
Veerasenareddy Burru, Vijaya Mohan Guvva, Weilin Chang, netdev,
linux-kernel
The two do the same thing, but we want to have a consistent
naming in the kernel.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
drivers/net/ethernet/cavium/liquidio/lio_main.c | 2 +-
drivers/net/ethernet/cavium/liquidio/octeon_console.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_main.c b/drivers/net/ethernet/cavium/liquidio/lio_main.c
index 19d03fc24305..30884ad0a509 100644
--- a/drivers/net/ethernet/cavium/liquidio/lio_main.c
+++ b/drivers/net/ethernet/cavium/liquidio/lio_main.c
@@ -687,7 +687,7 @@ static void lio_sync_octeon_time(struct work_struct *work)
lt = (struct lio_time *)sc->virtdptr;
/* Get time of the day */
- getnstimeofday64(&ts);
+ ktime_get_real_ts64(&ts);
lt->sec = ts.tv_sec;
lt->nsec = ts.tv_nsec;
octeon_swap_8B_data((u64 *)lt, (sizeof(struct lio_time)) / 8);
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_console.c b/drivers/net/ethernet/cavium/liquidio/octeon_console.c
index 7f97ae48efed..0cc2338d8d2a 100644
--- a/drivers/net/ethernet/cavium/liquidio/octeon_console.c
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_console.c
@@ -902,7 +902,7 @@ int octeon_download_firmware(struct octeon_device *oct, const u8 *data,
*
* Octeon always uses UTC time. so timezone information is not sent.
*/
- getnstimeofday64(&ts);
+ ktime_get_real_ts64(&ts);
ret = snprintf(boottime, MAX_BOOTTIME_SIZE,
" time_sec=%lld time_nsec=%ld",
(s64)ts.tv_sec, ts.tv_nsec);
--
2.9.0
^ permalink raw reply related
* Re: [PATCH] [RESEND, net-next, v2] tcp: use monotonic timestamps for PAWS
From: Eric Dumazet @ 2018-07-11 12:17 UTC (permalink / raw)
To: Arnd Bergmann, David S. Miller, Eric Dumazet, Alexey Kuznetsov,
Hideaki YOSHIFUJI
Cc: Harsh Jain, Herbert Xu, Atul Gupta, Gustavo A. R. Silva,
Michael Werner, linux-crypto, linux-kernel, netdev
In-Reply-To: <20180711101757.3929777-1-arnd@arndb.de>
On 07/11/2018 03:16 AM, Arnd Bergmann wrote:
> Using get_seconds() for timestamps is deprecated since it can lead
> to overflows on 32-bit systems. While the interface generally doesn't
> overflow until year 2106, the specific implementation of the TCP PAWS
> algorithm breaks in 2038 when the intermediate signed 32-bit timestamps
> overflow.
>
> A related problem is that the local timestamps in CLOCK_REALTIME form
> lead to unexpected behavior when settimeofday is called to set the system
> clock backwards or forwards by more than 24 days.
>
> While the first problem could be solved by using an overflow-safe method
> of comparing the timestamps, a nicer solution is to use a monotonic
> clocksource with ktime_get_seconds() that simply doesn't overflow (at
> least not until 136 years after boot) and that doesn't change during
> settimeofday().
>
> To make 32-bit and 64-bit architectures behave the same way here, and
> also save a few bytes in the tcp_options_received structure, I'm changing
> the type to a 32-bit integer, which is now safe on all architectures.
Nit: This is only manipulated under socket lock protection, so it is safe even on
32bit kernels.
>
> Finally, the ts_recent_stamp field also (confusingly) gets used to store
> a jiffies value in tcp_synq_overflow()/tcp_synq_no_recent_overflow().
> This is currently safe, but changing the type to 32-bit requires
> some small changes there to keep it working.
>
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> ---
> v2: use time_before32()/time_after32() everywhere as suggested
> Eric Dumazet in https://lore.kernel.org/lkml/67ebb94d-c73f-6c9f-493b-00c86f595120@gmail.com/
SGTM, thanks Arnd
Signed-off-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply
* Re: [PATCH rdma-next 4/9] IB/mlx5: Introduce flow steering matcher object
From: Jason Gunthorpe @ 2018-07-11 12:02 UTC (permalink / raw)
To: Yishai Hadas
Cc: Leon Romanovsky, Doug Ledford, Leon Romanovsky, RDMA mailing list,
Yishai Hadas, Saeed Mahameed, linux-netdev, Majd Dibbiny
In-Reply-To: <ba5fca7e-814f-47f0-486e-30b27cd56797@dev.mellanox.co.il>
On Wed, Jul 11, 2018 at 12:32:56PM +0300, Yishai Hadas wrote:
> >>+static int UVERBS_HANDLER(MLX5_IB_METHOD_FLOW_MATCHER_CREATE)(struct ib_device *ib_dev,
> >>+ struct ib_uverbs_file *file,
> >>+ struct uverbs_attr_bundle *attrs)
> >>+{
> >>+ struct mlx5_ib_dev *dev = to_mdev(ib_dev);
> >
> >I have a patch changing these - when working with uobj's the ib_dev
> >argument should not be used, the ib_dev must come from the ucontext
> >instead.
> >
>
> OK, V1 will take the ib_dev from the uobj.
>
> However, does it mean that you are going to change the signatures of all the
> handlers to *not* get ib_dev ?
Yes.
Jason
^ permalink raw reply
* Re: [PATCH net-next v5 0/4] net: vhost: improve performance when enable busyloop
From: Michael S. Tsirkin @ 2018-07-11 11:59 UTC (permalink / raw)
To: Jason Wang
Cc: Tonghao Zhang, makita.toshiaki, virtualization,
Linux Kernel Network Developers
In-Reply-To: <4901593f-1d9e-2c47-b963-5fc1f48a1bfe@redhat.com>
On Wed, Jul 11, 2018 at 01:12:59PM +0800, Jason Wang wrote:
>
>
> On 2018年07月11日 11:49, Tonghao Zhang wrote:
> > On Wed, Jul 11, 2018 at 10:56 AM Jason Wang <jasowang@redhat.com> wrote:
> > >
> > >
> > > On 2018年07月04日 12:31, xiangxia.m.yue@gmail.com wrote:
> > > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > >
> > > > This patches improve the guest receive and transmit performance.
> > > > On the handle_tx side, we poll the sock receive queue at the same time.
> > > > handle_rx do that in the same way.
> > > >
> > > > For more performance report, see patch 4.
> > > >
> > > > v4 -> v5:
> > > > fix some issues
> > > >
> > > > v3 -> v4:
> > > > fix some issues
> > > >
> > > > v2 -> v3:
> > > > This patches are splited from previous big patch:
> > > > http://patchwork.ozlabs.org/patch/934673/
> > > >
> > > > Tonghao Zhang (4):
> > > > vhost: lock the vqs one by one
> > > > net: vhost: replace magic number of lock annotation
> > > > net: vhost: factor out busy polling logic to vhost_net_busy_poll()
> > > > net: vhost: add rx busy polling in tx path
> > > >
> > > > drivers/vhost/net.c | 108 ++++++++++++++++++++++++++++----------------------
> > > > drivers/vhost/vhost.c | 24 ++++-------
> > > > 2 files changed, 67 insertions(+), 65 deletions(-)
> > > >
> > > Hi, any progress on the new version?
> > >
> > > I plan to send a new series of packed virtqueue support of vhost. If you
> > > plan to send it soon, I can wait. Otherwise, I will send my series.
> > I rebase the codes. and find there is no improvement anymore, the
> > patches of makita may solve the problem. jason you may send your
> > patches, and I will do some research on busypoll.
>
> I see. Maybe you can try some bi-directional traffic.
>
> Btw, lots of optimizations could be done for busy polling. E.g integrating
> with host NAPI busy polling or a 100% busy polling vhost_net. You're welcome
> to work or propose new ideas.
>
> Thanks
It seems clear we do need adaptive polling. The difficulty with NAPI
polling is it can't access guest memory easily. But maybe
get_user_pages on the polled memory+NAPI polling can work.
> >
> > > Thanks
^ permalink raw reply
* [PATCH iproute2-next] ipaddress: fix label matching
From: Vincent Bernat @ 2018-07-11 11:36 UTC (permalink / raw)
To: netdev, stephen, serhe.popovych; +Cc: Vincent Bernat
Since 9516823051ce, "ip addr show label lo:1" doesn't work
anymore (doesn't show any address, despite a matching label).
Reverting to return 0 instead of -1 fix the issue.
However, the condition says: "if we filter by label [...] and the
label does NOT match the interface name". This makes little sense to
compare the label with the interface name. There is also a logic
around filter family being provided or not. The match against the
label is done by ifa_label_match_rta() in print_addrinfo() and
ipaddr_filter().
Just removing the condition makes "ip addr show" works as expected
with or without specifying a label, both when the label is matching
and not matching. It also works if we specify a label and the label is
the interface name. The flush operation also works as expected.
Fixes: 9516823051ce ("ipaddress: Improve print_linkinfo()")
Signed-off-by: Vincent Bernat <vincent@bernat.im>
---
ip/ipaddress.c | 5 -----
1 file changed, 5 deletions(-)
diff --git a/ip/ipaddress.c b/ip/ipaddress.c
index 5009bfe6d2e3..20ef6724944e 100644
--- a/ip/ipaddress.c
+++ b/ip/ipaddress.c
@@ -837,11 +837,6 @@ int print_linkinfo(const struct sockaddr_nl *who,
if (!name)
return -1;
- if (filter.label &&
- (!filter.family || filter.family == AF_PACKET) &&
- fnmatch(filter.label, name, 0))
- return -1;
-
if (tb[IFLA_GROUP]) {
int group = rta_getattr_u32(tb[IFLA_GROUP]);
--
2.18.0
^ permalink raw reply related
* Re: [RFC net-next 00/15] net: A socket API for LoRa
From: Helmut Tschemernjak @ 2018-07-11 11:45 UTC (permalink / raw)
To: Andreas Färber
Cc: Stefan Schmidt, netdev, linux-arm-kernel, linux-kernel,
Jian-Hong Pan, Jiri Pirko, Marcel Holtmann, David S . Miller,
Matthias Brugger, Janus Piwek, Michael Röder, Dollar Chen,
Ken Yu, Konstantin Böhm, Jan Jongboom, Jon Ortego, contact,
Ben Whitten, Brian Ray, lora, lora
In-Reply-To: <23213cab-dc87-118c-1750-cd30d60ce9f8@suse.de>
Dear Andreas,
here are a few comments in response to your answer.
The SX1276 can do MTU sizes up to 2048 bytes, there is a FIFO and
corresponding interrupts to handle this, details are in datasheet.
Regarding the LoRaWAN concept, when companies like Semtech and IBM
designing such a thing, and other companies (Stackforce) implementing
the code for it. This does not mean automatically that it is good for
most use cases. Specifically the idea of the concentrator has major
limits. Check out our article and picture: "Protocols: LoRaWANTM vs
RadioShuttle" at http://www.radioshuttle.de/en/radioshuttle-2/protocol/
Again my opinion is that that such a specific driver is not needed in
the kernel.
I am happy to provide further feedback, when you have questions about
the SX1276Generic driver which I maintain, I am available to discuss
this. And again the open source driver works on mbed, arduino and can be
easily adopted to Linux user space and other environments.
Regards
Helmut
On 11/07/18 04:07, Andreas Färber wrote:
> Dear Helmut,
>
> Am 05.07.2018 um 12:43 schrieb Helmut Tschemernjak:
>> I put the kernel support for the SX1276 LoRa chip in question. I don’t
>> think that this kind of device should be in the Linux kernel.
> Thanks for sharing your opinion.
>
>> Here are a few facts to consider:
>> - A LoRa transaction is very slow (e.g. SF7 needs about 210 ms, for
>> SF12 6280 ms) for 12 bytes user data with acknowledge.
> Where do you see a problem? If you look at my SX1276 patch, you will
> find that it queues a work item for the transmission (asynchronously)
> and has an interrupt handler to get notified of the TX interrupt. It
> surely won't get quicker in userspace - and as you point out, we're not
> speaking about DPDK performance here, so no need for polling.
>
>> - There are many different implementations for the antenna switch,
>> switching between receiving/sending, PA-BOOST, 433, 868/915 MHz. (I know
>> SX1276 Heltec ESP32, SX1276 Murata, RFM95-(1276), SX1276 Heltec
>> STM32-L4) they are all different regarding this.
> Yes, and as demonstrated with this patch series, the kernel is well able
> to handle this variety of interfaces. I would say better than userspace!
>
> As noted, I've tested both 868 MHz and 433 MHz modules. My SX1276 driver
> uses the Device Tree (radio-frequency) to configure the driver when this
> was a property of the module. RN2483 by contrast supports two, so a
> netlink interface was suggested to configure this at runtime.
>
>> - The LoRa chip device ID is only 8-bit which is not sufficient for
>> larger networks, e.g. our RadioShuttle protocol uses compressed 32-bit
>> device IDs.
> What does that have to do with anything? The whole reason that you're
> CC'ed on this patchset is to make sure that you or someone else can
> implement your custom protocol on top of the APIs being proposed. So
> my-protocol-is-better-than-theirs is no argument against this project.
>
> The way I understand LoRa and thus designed the struct sockaddr_lora is
> that there is no ID involved for LoRa at all - it is essentially a
> broadcast. The 8-byte (not 8-bit) EUIs only come into play for LoRaWAN
> AFAIU. Am I missing anything?
>
> LoRaWAN, RadioShuttle, RadioHead and LR Base all have their own ways of
> addressing - the question here is how to model that in Linux, whether as
> one PF_LORA with different protocol options and a growing union
> lora_addr inside sockaddr_lora covering all of them, or a protocol
> family (requiring global number allocation) for each one of them.
>
>> - The chip can do MTU sizes up to 2048 bytes, most protocols use less
>> than 250 bytes.
> How would 2048 bytes work? The FIFO buffer fits a maximum of 256 bytes.
>
> IIRC the recommended MTU is also restricted by the airtime, i.e. SF,
> bandwidth, preamble and other factors...
>
> Some UART based modules allow a maximum of 64 bytes.
>
>> - Applications do on-the-fly channel and spreading factor switching
>> which makes it more difficult for the configuration.
> That's exactly the question of how to implement individual options -
> Device Tree would be fixed for the system's hardware, netlink would be
> user-configurable for the whole network device, whereas socket options
> would be per socket/application, and ioctls might be yet another
> implementation layer.
>
>> - The chip supports Lora modulation as well as FSK, GFSK, MSK, GMSK,
>> and OOK modulation, for some use cases the other modulations are of
>> interest, e.g. to communicated with other devices.
> Yes, therefore I raised it in the cover letter as open point 5).
>
>> - Power saving modes, like sleep, suspend may be required for battery
>> operation.
> The Linux kernel can deal with that much better than userspace. There's
> pm hooks that one could implement. And SX1276 returns to standby mode
> whenever an operation such as TX is completed. In sleep mode not all
> settings get preserved, so they would need to be saved in the private
> struct and restored on resume.
>
>> - Cad detection is very flexible and can be differently used.
>>
>> - LoRa preamble may be different depending on the protocol used.
>>
>> - The new Lora chip SX1262 / 68 (successor of the SX1276) has different
>> hardware and all different registers, it is driver incompatible with the
>> SX1276. It needs an entire new driver.
> Not unexpected, same as SX1301 being different. As this series shows,
> multiple drivers can easily be implemented as necessary.
>
>> - The device is not multi-process capable (only a single process can
>> communicate with it).
> How is that different from other half-duplex network interfaces? The
> driver needs to take care of appropriate locking of its operations, and
> the socket subsystem should make sure packets get queued accordingly.
>
>> - There are SX1276 LoRa modules with a build-in protocol (LoRaWAN and
>> RAW) via a serial connection only, again complete different API compared
>> the SX1276 chip. Software updates for this devices are difficult.
> Drivers for such modules are already demonstrated in this series.
>
>> - I am even convinced that the LoRaWAN protocol with the concentrator
>> concept is not good, the peer to peer communication and a standard MQTT
>> gateway (what we do) is way more flexible.
> Your opinion in all honors, LoRaWAN is an open specification, whereas
> yours is closed. So whatever benefits yours may have, that way you're
> unlikely to make LoRaWAN go away. People are rolling out LoRaWAN
> gateways, and therefore the question is not which protocol is more
> flexible but rather how all of them can be enabled equally for those
> that want to use them.
>
>> For all this reasons, I feel a user level driver task implementation is
>> way more flexible. I did a lot of work/enhancements on the SX1276 link
>> level driver from Semtech, it is available and maintained on mbed.org
>> and supports mbed-os, Arduino and is prepared for Linux.
>> https://os.mbed.com/users/Helmut64/code/SX1276GenericLib/
>>
>> Protocols e.g. our RadioShuttle LoRa peer to peer protocol or the
>> LoRaWAN protocol can run on top of the SX1276GenericLib. We may should
>> focus on such a driver library getting supported for multiple OS's (Win,
>> Linux, mbed, Ardino, etc.)
>>
>> Again I feel a Linux kernel device driver for the SX1276 chip make
>> little sense for me.
> Well, feel free to use spidev if you prefer. As I pointed out, the Linux
> spi maintainers don't seem to share your view - instead of white-listing
> chipsets for concrete kernel drivers (as you appear to suggest here)
> they rather black-list chipsets to be "allowed" to use spidev if all
> else fails.
>
> Having already done part of the implementation work, for me the question
> is not whether to do a kernel driver but how to do it properly, keeping
> all corner cases in mind, such as non-standard protocols like yours.
> It's great that you're writing an Open Source library for mbed, but its
> name indicates that it offers no abstraction like proposed here.
>
> I'll drop you from v2 then, to not bother you about this Linux kernel
> implementation anymore.
>
> Regards,
> Andreas
>
--
best regards / mit freundlichen Gruessen,
Helmut Tschemernjak
HELIOS Software GmbH
Steinriede 3
30827 Garbsen (Hannover)
Phone: +49-5131-709320
Fax: +49-5131-709325
Internet Mail: helmut@helios.de
Web: http://www.helios.de
Google+: http://helios.de/helmut
^ permalink raw reply
* Re: [PATCH] net: convert gro_count to bitmask
From: Eric Dumazet @ 2018-07-11 11:32 UTC (permalink / raw)
To: Li RongQing, netdev
In-Reply-To: <1531300553-21413-1-git-send-email-lirongqing@baidu.com>
On 07/11/2018 02:15 AM, Li RongQing wrote:
> gro_hash size is 192 bytes, and uses 3 cache lines, if there is few
> flows, gro_hash may be not fully used, so it is unnecessary to iterate
> all gro_hash in napi_gro_flush(), to occupy unnecessary cacheline.
>
> convert gro_count to a bitmask, and rename it as gro_bitmask, each bit
> represents a element of gro_hash, only flush a gro_hash element if the
> related bit is set, to speed up napi_gro_flush().
>
> and update gro_bitmask only if it will be changed, to reduce cache
> update
>
> Suggested-by: Eric Dumazet <edumazet@google.com>
> Signed-off-by: Li RongQing <lirongqing@baidu.com>
> ---
> include/linux/netdevice.h | 2 +-
> net/core/dev.c | 35 +++++++++++++++++++++++------------
> 2 files changed, 24 insertions(+), 13 deletions(-)
>
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index b683971e500d..df49b36ef378 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -322,7 +322,7 @@ struct napi_struct {
>
> unsigned long state;
> int weight;
> - unsigned int gro_count;
> + unsigned long gro_bitmask;
> int (*poll)(struct napi_struct *, int);
> #ifdef CONFIG_NETPOLL
> int poll_owner;
> diff --git a/net/core/dev.c b/net/core/dev.c
> index d13cddcac41f..a08dbdd217a6 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -5171,9 +5171,11 @@ static void __napi_gro_flush_chain(struct napi_struct *napi, u32 index,
> return;
> list_del_init(&skb->list);
> napi_gro_complete(skb);
> - napi->gro_count--;
> napi->gro_hash[index].count--;
> }
> +
> + if (!napi->gro_hash[index].count)
> + clear_bit(index, &napi->gro_bitmask);
I suggest you not add an atomic operation here.
Current cpu owns this NAPI after all.
Same remark for the whole patch.
-> __clear_bit(), __set_bit() and similar operators
Ideally you should provide TCP_RR number with busy polling enabled, to eventually catch regressions.
Thanks.
^ permalink raw reply
* Re: [PATCH net-next v2 10/10] r8169: don't read chip phy status register
From: Florian Fainelli @ 2018-07-11 11:12 UTC (permalink / raw)
To: Heiner Kallweit, David Miller, Andrew Lunn,
Realtek linux nic maintainers
Cc: netdev@vger.kernel.org
In-Reply-To: <78ac9a40-c7ec-4971-0bfc-bf0b46d77866@gmail.com>
On 07/10/2018 11:40 AM, Heiner Kallweit wrote:
> Instead of accessing the PHYstatus register we can use the information
> phylib stores in the phy_device structure.
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* Re: [PATCH net-next v2 06/10] r8169: use phy_mii_ioctl
From: Florian Fainelli @ 2018-07-11 11:11 UTC (permalink / raw)
To: Heiner Kallweit, David Miller, Andrew Lunn,
Realtek linux nic maintainers
Cc: netdev@vger.kernel.org
In-Reply-To: <c901d4e7-6e6e-378c-daa6-b5e27c59d982@gmail.com>
On 07/10/2018 11:39 AM, Heiner Kallweit wrote:
> Switch to using phy_mii_ioctl().
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* [PATCH rdma-next v1 8/8] IB/mlx5: Expose vendor flow trees
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Expose mlx5 flow trees to be used by upper layers.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/infiniband/hw/mlx5/flow.c | 9 +++++++++
drivers/infiniband/hw/mlx5/main.c | 4 +++-
drivers/infiniband/hw/mlx5/mlx5_ib.h | 3 +++
3 files changed, 15 insertions(+), 1 deletion(-)
diff --git a/drivers/infiniband/hw/mlx5/flow.c b/drivers/infiniband/hw/mlx5/flow.c
index 76e76b87e5e9..ff5a02713b87 100644
--- a/drivers/infiniband/hw/mlx5/flow.c
+++ b/drivers/infiniband/hw/mlx5/flow.c
@@ -243,3 +243,12 @@ DECLARE_UVERBS_NAMED_OBJECT(MLX5_IB_OBJECT_FLOW_MATCHER,
DECLARE_UVERBS_OBJECT_TREE(flow_objects,
&UVERBS_OBJECT(MLX5_IB_OBJECT_FLOW_MATCHER));
+int mlx5_ib_get_flow_trees(const struct uverbs_object_tree_def **root)
+{
+ int i = 0;
+
+ root[i++] = &flow_objects;
+ root[i++] = &mlx5_ib_fs;
+
+ return i;
+}
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index d750562cb2c5..71f3e9677622 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -5537,7 +5537,7 @@ ADD_UVERBS_ATTRIBUTES_SIMPLE(
UVERBS_ATTR_TYPE(u64),
UA_MANDATORY));
-#define NUM_TREES 3
+#define NUM_TREES 5
static int populate_specs_root(struct mlx5_ib_dev *dev)
{
const struct uverbs_object_tree_def *default_root[NUM_TREES + 1] = {
@@ -5557,6 +5557,8 @@ static int populate_specs_root(struct mlx5_ib_dev *dev)
!WARN_ON(num_trees >= ARRAY_SIZE(default_root)))
default_root[num_trees++] = mlx5_ib_get_devx_tree();
+ num_trees += mlx5_ib_get_flow_trees(default_root + num_trees);
+
dev->ib_dev.driver_specs_root =
uverbs_alloc_spec_tree(num_trees, default_root);
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 7690dc6a897e..623f5d8bd48e 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -1244,6 +1244,7 @@ struct mlx5_ib_flow_handler *mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
int inlen, int dest_id,
int dest_type);
bool mlx5_ib_devx_is_flow_dest(void *obj, int *dest_id, int *dest_type);
+int mlx5_ib_get_flow_trees(const struct uverbs_object_tree_def **root);
#else
static inline int
mlx5_ib_devx_create(struct mlx5_ib_dev *dev,
@@ -1261,6 +1262,8 @@ mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
static inline bool
mlx5_ib_devx_is_flow_dest(void *obj, int *dest_id,
int *dest_type) { return false; };
+static inline int
+mlx5_ib_get_flow_trees(const struct uverbs_object_tree_def **root) { return 0; };
#endif
static inline void init_query_mad(struct ib_smp *mad)
{
--
2.14.4
^ permalink raw reply related
* Re: [PATCH net-next v2 05/10] r8169: use phy_ethtool_nway_reset
From: Florian Fainelli @ 2018-07-11 11:11 UTC (permalink / raw)
To: Heiner Kallweit, David Miller, Andrew Lunn,
Realtek linux nic maintainers
Cc: netdev@vger.kernel.org
In-Reply-To: <b6e1e378-ea1b-6b30-4253-292196ffb7ae@gmail.com>
On 07/10/2018 11:39 AM, Heiner Kallweit wrote:
> Switch to using phy_ethtool_nway_reset().
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* [PATCH rdma-next v1 7/8] IB/mlx5: Add support for a flow table destination
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Add support to set a destination that is a flow table, this can come
from the DEVX destination.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/infiniband/hw/mlx5/main.c | 13 ++++++++-----
1 file changed, 8 insertions(+), 5 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index dba3d529f49f..d750562cb2c5 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3791,9 +3791,6 @@ mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
if (fs_matcher->priority > MLX5_IB_FLOW_LAST_PRIO)
return ERR_PTR(-ENOMEM);
- if (dest_type != MLX5_FLOW_DESTINATION_TYPE_TIR)
- return ERR_PTR(-ENOTSUPP);
-
dst = kzalloc(sizeof(*dst), GFP_KERNEL);
if (!dst)
return ERR_PTR(-ENOMEM);
@@ -3807,8 +3804,14 @@ mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
goto unlock;
}
- dst->type = dest_type;
- dst->tir_num = dest_id;
+ if (dest_type == MLX5_FLOW_DESTINATION_TYPE_TIR) {
+ dst->type = dest_type;
+ dst->tir_num = dest_id;
+ } else {
+ dst->type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM;
+ dst->ft_num = dest_id;
+ }
+
handler = _create_raw_flow_rule(dev, ft_prio, dst, fs_matcher, cmd_in,
inlen);
--
2.14.4
^ permalink raw reply related
* [PATCH rdma-next v1 6/8] IB/mlx5: Support adding flow steering rule by raw data
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Add support to set a public flow steering rule which its destination is
a TIR by using raw specification data.
The logic follows the verbs API but instead of using ib_spec(s) the raw
input data is used.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/infiniband/hw/mlx5/main.c | 216 ++++++++++++++++++++++++++++++++---
drivers/infiniband/hw/mlx5/mlx5_ib.h | 2 +
2 files changed, 201 insertions(+), 17 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index 53e30449477e..dba3d529f49f 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -2980,11 +2980,11 @@ static void counters_clear_description(struct ib_counters *counters)
static int mlx5_ib_destroy_flow(struct ib_flow *flow_id)
{
- struct mlx5_ib_dev *dev = to_mdev(flow_id->qp->device);
struct mlx5_ib_flow_handler *handler = container_of(flow_id,
struct mlx5_ib_flow_handler,
ibflow);
struct mlx5_ib_flow_handler *iter, *tmp;
+ struct mlx5_ib_dev *dev = handler->dev;
mutex_lock(&dev->flow_db->lock);
@@ -3002,6 +3002,8 @@ static int mlx5_ib_destroy_flow(struct ib_flow *flow_id)
counters_clear_description(handler->ibcounters);
mutex_unlock(&dev->flow_db->lock);
+ if (handler->flow_matcher)
+ atomic_dec(&handler->flow_matcher->usecnt);
kfree(handler);
return 0;
@@ -3022,6 +3024,26 @@ enum flow_table_type {
#define MLX5_FS_MAX_TYPES 6
#define MLX5_FS_MAX_ENTRIES BIT(16)
+
+static struct mlx5_ib_flow_prio *_get_prio(struct mlx5_flow_namespace *ns,
+ struct mlx5_ib_flow_prio *prio,
+ int priority,
+ int num_entries, int num_groups)
+{
+ struct mlx5_flow_table *ft;
+
+ ft = mlx5_create_auto_grouped_flow_table(ns, priority,
+ num_entries,
+ num_groups,
+ 0, 0);
+ if (IS_ERR(ft))
+ return ERR_CAST(ft);
+
+ prio->flow_table = ft;
+ prio->refcount = 0;
+ return prio;
+}
+
static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev,
struct ib_flow_attr *flow_attr,
enum flow_table_type ft_type)
@@ -3034,7 +3056,6 @@ static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev,
int num_entries;
int num_groups;
int priority;
- int err = 0;
max_table_size = BIT(MLX5_CAP_FLOWTABLE_NIC_RX(dev->mdev,
log_max_ft_size));
@@ -3084,21 +3105,10 @@ static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev,
return ERR_PTR(-ENOMEM);
ft = prio->flow_table;
- if (!ft) {
- ft = mlx5_create_auto_grouped_flow_table(ns, priority,
- num_entries,
- num_groups,
- 0, 0);
-
- if (!IS_ERR(ft)) {
- prio->refcount = 0;
- prio->flow_table = ft;
- } else {
- err = PTR_ERR(ft);
- }
- }
+ if (!ft)
+ return _get_prio(ns, prio, priority, num_entries, num_groups);
- return err ? ERR_PTR(err) : prio;
+ return prio;
}
static void set_underlay_qp(struct mlx5_ib_dev *dev,
@@ -3357,6 +3367,7 @@ static struct mlx5_ib_flow_handler *_create_flow_rule(struct mlx5_ib_dev *dev,
ft_prio->refcount++;
handler->prio = ft_prio;
+ handler->dev = dev;
ft_prio->flow_table = ft;
free:
@@ -3643,13 +3654,184 @@ static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp,
return ERR_PTR(err);
}
+static struct mlx5_ib_flow_prio *_get_flow_table(struct mlx5_ib_dev *dev,
+ int priority, bool mcast)
+{
+ int max_table_size;
+ struct mlx5_flow_namespace *ns = NULL;
+ struct mlx5_ib_flow_prio *prio;
+
+ max_table_size = BIT(MLX5_CAP_FLOWTABLE_NIC_RX(dev->mdev,
+ log_max_ft_size));
+ if (max_table_size < MLX5_FS_MAX_ENTRIES)
+ return ERR_PTR(-ENOMEM);
+
+ if (mcast)
+ priority = MLX5_IB_FLOW_MCAST_PRIO;
+ else
+ priority = ib_prio_to_core_prio(priority, false);
+
+ ns = mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS);
+ if (!ns)
+ return ERR_PTR(-ENOTSUPP);
+
+ prio = &dev->flow_db->prios[priority];
+
+ if (prio->flow_table)
+ return prio;
+
+ return _get_prio(ns, prio, priority, MLX5_FS_MAX_ENTRIES,
+ MLX5_FS_MAX_TYPES);
+}
+
+static struct mlx5_ib_flow_handler *
+_create_raw_flow_rule(struct mlx5_ib_dev *dev,
+ struct mlx5_ib_flow_prio *ft_prio,
+ struct mlx5_flow_destination *dst,
+ struct mlx5_ib_flow_matcher *fs_matcher,
+ void *cmd_in, int inlen)
+{
+ struct mlx5_ib_flow_handler *handler;
+ struct mlx5_flow_act flow_act = {.flow_tag = MLX5_FS_DEFAULT_FLOW_TAG};
+ struct mlx5_flow_spec *spec;
+ struct mlx5_flow_table *ft = ft_prio->flow_table;
+ int err = 0;
+
+ spec = kvzalloc(sizeof(*spec), GFP_KERNEL);
+ handler = kzalloc(sizeof(*handler), GFP_KERNEL);
+ if (!handler || !spec) {
+ err = -ENOMEM;
+ goto free;
+ }
+
+ INIT_LIST_HEAD(&handler->list);
+
+ memcpy(spec->match_value, cmd_in, inlen);
+ memcpy(spec->match_criteria, fs_matcher->matcher_mask.match_params,
+ fs_matcher->mask_len);
+ spec->match_criteria_enable = fs_matcher->match_criteria_enable;
+
+ flow_act.action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST;
+ handler->rule = mlx5_add_flow_rules(ft, spec,
+ &flow_act, dst, 1);
+
+ if (IS_ERR(handler->rule)) {
+ err = PTR_ERR(handler->rule);
+ goto free;
+ }
+
+ ft_prio->refcount++;
+ handler->prio = ft_prio;
+ handler->dev = dev;
+ ft_prio->flow_table = ft;
+
+free:
+ if (err)
+ kfree(handler);
+ kvfree(spec);
+ return err ? ERR_PTR(err) : handler;
+}
+
+static bool raw_fs_is_multicast(struct mlx5_ib_flow_matcher *fs_matcher,
+ void *match_v)
+{
+ void *match_c;
+ void *match_v_set_lyr_2_4, *match_c_set_lyr_2_4;
+ void *dmac, *dmac_mask;
+ void *ipv4, *ipv4_mask;
+
+ if (!(fs_matcher->match_criteria_enable &
+ (1 << MATCH_CRITERIA_ENABLE_OUTER_BIT)))
+ return false;
+
+ match_c = fs_matcher->matcher_mask.match_params;
+ match_v_set_lyr_2_4 = MLX5_ADDR_OF(fte_match_param, match_v,
+ outer_headers);
+ match_c_set_lyr_2_4 = MLX5_ADDR_OF(fte_match_param, match_c,
+ outer_headers);
+
+ dmac = MLX5_ADDR_OF(fte_match_set_lyr_2_4, match_v_set_lyr_2_4,
+ dmac_47_16);
+ dmac_mask = MLX5_ADDR_OF(fte_match_set_lyr_2_4, match_c_set_lyr_2_4,
+ dmac_47_16);
+
+ if (is_multicast_ether_addr(dmac) &&
+ is_multicast_ether_addr(dmac_mask))
+ return true;
+
+ ipv4 = MLX5_ADDR_OF(fte_match_set_lyr_2_4, match_v_set_lyr_2_4,
+ dst_ipv4_dst_ipv6.ipv4_layout.ipv4);
+
+ ipv4_mask = MLX5_ADDR_OF(fte_match_set_lyr_2_4, match_c_set_lyr_2_4,
+ dst_ipv4_dst_ipv6.ipv4_layout.ipv4);
+
+ if (ipv4_is_multicast(*(__be32 *)(ipv4)) &&
+ ipv4_is_multicast(*(__be32 *)(ipv4_mask)))
+ return true;
+
+ return false;
+}
+
struct mlx5_ib_flow_handler *
mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
struct mlx5_ib_flow_matcher *fs_matcher,
void *cmd_in, int inlen, int dest_id,
int dest_type)
{
- return ERR_PTR(-EOPNOTSUPP);
+ struct mlx5_flow_destination *dst;
+ struct mlx5_ib_flow_prio *ft_prio;
+ int priority = fs_matcher->priority;
+ struct mlx5_ib_flow_handler *handler;
+ bool mcast;
+ int err;
+
+ if (fs_matcher->flow_type != MLX5_IB_FLOW_TYPE_NORMAL)
+ return ERR_PTR(-EOPNOTSUPP);
+
+ if (fs_matcher->priority > MLX5_IB_FLOW_LAST_PRIO)
+ return ERR_PTR(-ENOMEM);
+
+ if (dest_type != MLX5_FLOW_DESTINATION_TYPE_TIR)
+ return ERR_PTR(-ENOTSUPP);
+
+ dst = kzalloc(sizeof(*dst), GFP_KERNEL);
+ if (!dst)
+ return ERR_PTR(-ENOMEM);
+
+ mcast = raw_fs_is_multicast(fs_matcher, cmd_in);
+ mutex_lock(&dev->flow_db->lock);
+
+ ft_prio = _get_flow_table(dev, priority, mcast);
+ if (IS_ERR(ft_prio)) {
+ err = PTR_ERR(ft_prio);
+ goto unlock;
+ }
+
+ dst->type = dest_type;
+ dst->tir_num = dest_id;
+ handler = _create_raw_flow_rule(dev, ft_prio, dst, fs_matcher, cmd_in,
+ inlen);
+
+ if (IS_ERR(handler)) {
+ err = PTR_ERR(handler);
+ goto destroy_ft;
+ }
+
+ mutex_unlock(&dev->flow_db->lock);
+ atomic_inc(&fs_matcher->usecnt);
+ handler->flow_matcher = fs_matcher;
+
+ kfree(dst);
+
+ return handler;
+
+destroy_ft:
+ put_flow_table(dev, ft_prio, false);
+unlock:
+ mutex_unlock(&dev->flow_db->lock);
+ kfree(dst);
+
+ return ERR_PTR(err);
}
static u32 mlx5_ib_flow_action_flags_to_accel_xfrm_flags(u32 mlx5_flags)
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 7c502052cb87..7690dc6a897e 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -178,6 +178,8 @@ struct mlx5_ib_flow_handler {
struct mlx5_ib_flow_prio *prio;
struct mlx5_flow_handle *rule;
struct ib_counters *ibcounters;
+ struct mlx5_ib_dev *dev;
+ struct mlx5_ib_flow_matcher *flow_matcher;
};
struct mlx5_ib_flow_matcher {
--
2.14.4
^ permalink raw reply related
* [PATCH rdma-next v1 5/8] IB/mlx5: Introduce vendor create and destroy flow methods
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Introduce vendor create and destroy flow methods on the uverbs flow
object.
This enables the driver to get its specific device attributes to match
the underlay specification while still using the generic ib_flow object
for cleanup and code sharing.
The IB object's attributes are set via ib_set_flow() helper function.
The specific implementation for the given specification is added in
downstream patches.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/infiniband/hw/mlx5/devx.c | 22 +++++++
drivers/infiniband/hw/mlx5/flow.c | 108 +++++++++++++++++++++++++++++++
drivers/infiniband/hw/mlx5/main.c | 9 +++
drivers/infiniband/hw/mlx5/mlx5_ib.h | 15 +++++
include/rdma/ib_verbs.h | 14 ++++
include/rdma/uverbs_named_ioctl.h | 29 +++++----
include/uapi/rdma/mlx5_user_ioctl_cmds.h | 17 +++++
7 files changed, 202 insertions(+), 12 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/devx.c b/drivers/infiniband/hw/mlx5/devx.c
index 60ac1fbe940e..32c7d8c71d85 100644
--- a/drivers/infiniband/hw/mlx5/devx.c
+++ b/drivers/infiniband/hw/mlx5/devx.c
@@ -89,6 +89,28 @@ void mlx5_ib_devx_destroy(struct mlx5_ib_dev *dev,
mlx5_cmd_exec(dev->mdev, in, sizeof(in), out, sizeof(out));
}
+bool mlx5_ib_devx_is_flow_dest(void *obj, int *dest_id, int *dest_type)
+{
+ struct devx_obj *devx_obj = obj;
+ u16 opcode = MLX5_GET(general_obj_in_cmd_hdr, devx_obj->dinbox, opcode);
+
+ switch (opcode) {
+ case MLX5_CMD_OP_DESTROY_TIR:
+ *dest_type = MLX5_FLOW_DESTINATION_TYPE_TIR;
+ *dest_id = MLX5_GET(general_obj_in_cmd_hdr, devx_obj->dinbox,
+ obj_id);
+ return true;
+
+ case MLX5_CMD_OP_DESTROY_FLOW_TABLE:
+ *dest_type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
+ *dest_id = MLX5_GET(destroy_flow_table_in, devx_obj->dinbox,
+ table_id);
+ return true;
+ default:
+ return false;
+ }
+}
+
static int devx_is_valid_obj_id(struct devx_obj *obj, const void *in)
{
u16 opcode = MLX5_GET(general_obj_in_cmd_hdr, in, opcode);
diff --git a/drivers/infiniband/hw/mlx5/flow.c b/drivers/infiniband/hw/mlx5/flow.c
index f1c16defe7ff..76e76b87e5e9 100644
--- a/drivers/infiniband/hw/mlx5/flow.c
+++ b/drivers/infiniband/hw/mlx5/flow.c
@@ -38,6 +38,81 @@ static const struct uverbs_attr_spec mlx5_ib_flow_type[] = {
},
};
+static int UVERBS_HANDLER(MLX5_IB_METHOD_CREATE_FLOW)(struct ib_device *ib_dev,
+ struct ib_uverbs_file *file,
+ struct uverbs_attr_bundle *attrs)
+{
+ struct mlx5_ib_flow_handler *flow_handler;
+ struct mlx5_ib_flow_matcher *fs_matcher;
+ void *devx_obj;
+ int dest_id, dest_type;
+ void *cmd_in;
+ int inlen;
+ bool dest_devx, dest_qp;
+ struct ib_qp *qp = NULL;
+ struct ib_uobject *uobj = uverbs_attr_get_uobject(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_HANDLE);
+ struct mlx5_ib_dev *dev = to_mdev(uobj->context->device);
+
+ if (!capable(CAP_NET_RAW))
+ return -EPERM;
+
+ dest_devx = uverbs_attr_is_valid(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_DEST_DEVX);
+ dest_qp = uverbs_attr_is_valid(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_DEST_QP);
+
+ if ((dest_devx && dest_qp) || (!dest_devx && !dest_qp))
+ return -EINVAL;
+
+ if (dest_devx) {
+ devx_obj = uverbs_attr_get_obj(attrs, MLX5_IB_ATTR_CREATE_FLOW_DEST_DEVX);
+ if (IS_ERR(devx_obj))
+ return PTR_ERR(devx_obj);
+
+ /* Verify that the given DEVX object is a flow
+ * steering destination.
+ */
+ if (!mlx5_ib_devx_is_flow_dest(devx_obj, &dest_id, &dest_type))
+ return -EINVAL;
+ } else {
+ struct mlx5_ib_qp *mqp;
+
+ qp = uverbs_attr_get_obj(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_DEST_QP);
+ if (IS_ERR(qp))
+ return PTR_ERR(qp);
+
+ if (qp->qp_type != IB_QPT_RAW_PACKET)
+ return -EINVAL;
+
+ mqp = to_mqp(qp);
+ if (mqp->flags & MLX5_IB_QP_RSS)
+ dest_id = mqp->rss_qp.tirn;
+ else
+ dest_id = mqp->raw_packet_qp.rq.tirn;
+ dest_type = MLX5_FLOW_DESTINATION_TYPE_TIR;
+ }
+
+ if (dev->rep)
+ return -ENOTSUPP;
+
+ cmd_in = uverbs_attr_get_alloced_ptr(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_MATCH_VALUE);
+ inlen = uverbs_attr_get_len(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_MATCH_VALUE);
+ fs_matcher = uverbs_attr_get_obj(attrs,
+ MLX5_IB_ATTR_CREATE_FLOW_MATCHER);
+ flow_handler = mlx5_ib_raw_fs_rule_add(dev, fs_matcher, cmd_in, inlen,
+ dest_id, dest_type);
+ if (IS_ERR(flow_handler))
+ return PTR_ERR(flow_handler);
+
+ ib_set_flow(uobj, &flow_handler->ibflow, qp, ib_dev);
+
+ return 0;
+}
+
static int flow_matcher_cleanup(struct ib_uobject *uobject,
enum rdma_remove_reason why)
{
@@ -101,6 +176,39 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_FLOW_MATCHER_CREATE)(struct ib_device *
return err;
}
+DECLARE_UVERBS_NAMED_METHOD(
+ MLX5_IB_METHOD_CREATE_FLOW,
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_CREATE_FLOW_HANDLE,
+ UVERBS_OBJECT_FLOW,
+ UVERBS_ACCESS_NEW,
+ UA_MANDATORY),
+ UVERBS_ATTR_PTR_IN(MLX5_IB_ATTR_CREATE_FLOW_MATCH_VALUE,
+ UVERBS_ATTR_SIZE(1, sizeof(struct mlx5_ib_match_params)),
+ UA_MANDATORY,
+ UA_ALLOC_AND_COPY),
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_CREATE_FLOW_MATCHER,
+ MLX5_IB_OBJECT_FLOW_MATCHER,
+ UVERBS_ACCESS_READ,
+ UA_MANDATORY),
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_CREATE_FLOW_DEST_QP, UVERBS_OBJECT_QP,
+ UVERBS_ACCESS_READ),
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_CREATE_FLOW_DEST_DEVX,
+ MLX5_IB_OBJECT_DEVX_OBJ,
+ UVERBS_ACCESS_READ));
+
+
+DECLARE_UVERBS_NAMED_METHOD_DESTROY(
+ MLX5_IB_METHOD_DESTROY_FLOW,
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_CREATE_FLOW_HANDLE,
+ UVERBS_OBJECT_FLOW,
+ UVERBS_ACCESS_DESTROY,
+ UA_MANDATORY));
+
+ADD_UVERBS_METHODS(mlx5_ib_fs,
+ UVERBS_OBJECT_FLOW,
+ &UVERBS_METHOD(MLX5_IB_METHOD_CREATE_FLOW),
+ &UVERBS_METHOD(MLX5_IB_METHOD_DESTROY_FLOW));
+
DECLARE_UVERBS_NAMED_METHOD(
MLX5_IB_METHOD_FLOW_MATCHER_CREATE,
UVERBS_ATTR_IDR(MLX5_IB_ATTR_FLOW_MATCHER_CREATE_HANDLE,
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index b7f94bc3811a..53e30449477e 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -3643,6 +3643,15 @@ static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp,
return ERR_PTR(err);
}
+struct mlx5_ib_flow_handler *
+mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
+ struct mlx5_ib_flow_matcher *fs_matcher,
+ void *cmd_in, int inlen, int dest_id,
+ int dest_type)
+{
+ return ERR_PTR(-EOPNOTSUPP);
+}
+
static u32 mlx5_ib_flow_action_flags_to_accel_xfrm_flags(u32 mlx5_flags)
{
u32 flags = 0;
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 3d9544ec996f..7c502052cb87 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -1236,6 +1236,12 @@ int mlx5_ib_devx_create(struct mlx5_ib_dev *dev,
void mlx5_ib_devx_destroy(struct mlx5_ib_dev *dev,
struct mlx5_ib_ucontext *context);
const struct uverbs_object_tree_def *mlx5_ib_get_devx_tree(void);
+struct mlx5_ib_flow_handler *mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
+ struct mlx5_ib_flow_matcher *fs_matcher,
+ void *cmd_in,
+ int inlen, int dest_id,
+ int dest_type);
+bool mlx5_ib_devx_is_flow_dest(void *obj, int *dest_id, int *dest_type);
#else
static inline int
mlx5_ib_devx_create(struct mlx5_ib_dev *dev,
@@ -1244,6 +1250,15 @@ static inline void mlx5_ib_devx_destroy(struct mlx5_ib_dev *dev,
struct mlx5_ib_ucontext *context) {}
static inline const struct uverbs_object_tree_def *
mlx5_ib_get_devx_tree(void) { return NULL; }
+static inline struct mlx5_ib_flow_handler *
+mlx5_ib_raw_fs_rule_add(struct mlx5_ib_dev *dev,
+ struct mlx5_ib_flow_matcher *fs_matcher,
+ void *cmd_in,
+ int inlen, int dest_id,
+ int dest_type) { return -EOPNOTSUPP; };
+static inline bool
+mlx5_ib_devx_is_flow_dest(void *obj, int *dest_id,
+ int *dest_type) { return false; };
#endif
static inline void init_query_mad(struct ib_smp *mad)
{
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index d02bbbafbd3c..d360e6f381c8 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -4137,6 +4137,20 @@ ib_get_vector_affinity(struct ib_device *device, int comp_vector)
}
+static inline void ib_set_flow(struct ib_uobject *uobj, struct ib_flow *ibflow,
+ struct ib_qp *qp, struct ib_device *device)
+{
+ uobj->object = ibflow;
+ ibflow->uobject = uobj;
+
+ if (qp) {
+ atomic_inc(&qp->usecnt);
+ ibflow->qp = qp;
+ }
+
+ ibflow->device = device;
+}
+
/**
* rdma_roce_rescan_device - Rescan all of the network devices in the system
* and add their gids, as needed, to the relevant RoCE devices.
diff --git a/include/rdma/uverbs_named_ioctl.h b/include/rdma/uverbs_named_ioctl.h
index 2eb1767042af..b3b21733cc55 100644
--- a/include/rdma/uverbs_named_ioctl.h
+++ b/include/rdma/uverbs_named_ioctl.h
@@ -97,22 +97,14 @@
.methods = &UVERBS_OBJECT_METHODS(_object_id) \
}
-/* Used by drivers to declare a complete parsing tree for a single method that
- * differs only in having additional driver specific attributes.
+/* Used by drivers to declare a complete parsing tree for new methods
*/
-#define ADD_UVERBS_ATTRIBUTES_SIMPLE(_name, _object_id, _method_id, ...) \
- static const struct uverbs_attr_def *const UVERBS_METHOD_ATTRS( \
- _method_id)[] = { __VA_ARGS__ }; \
- static const struct uverbs_method_def UVERBS_METHOD(_method_id) = { \
- .id = _method_id, \
- .num_attrs = ARRAY_SIZE(UVERBS_METHOD_ATTRS(_method_id)), \
- .attrs = &UVERBS_METHOD_ATTRS(_method_id), \
- }; \
+#define ADD_UVERBS_METHODS(_name, _object_id, ...) \
static const struct uverbs_method_def *const UVERBS_OBJECT_METHODS( \
- _object_id)[] = { &UVERBS_METHOD(_method_id) }; \
+ _object_id)[] = { __VA_ARGS__ }; \
static const struct uverbs_object_def _name##_struct = { \
.id = _object_id, \
- .num_methods = 1, \
+ .num_methods = ARRAY_SIZE(UVERBS_OBJECT_METHODS(_object_id)), \
.methods = &UVERBS_OBJECT_METHODS(_object_id) \
}; \
static const struct uverbs_object_def *const _name##_ptrs[] = { \
@@ -123,4 +115,17 @@
.objects = &_name##_ptrs, \
}
+/* Used by drivers to declare a complete parsing tree for a single method that
+ * differs only in having additional driver specific attributes.
+ */
+#define ADD_UVERBS_ATTRIBUTES_SIMPLE(_name, _object_id, _method_id, ...) \
+ static const struct uverbs_attr_def *const UVERBS_METHOD_ATTRS( \
+ _method_id)[] = { __VA_ARGS__ }; \
+ static const struct uverbs_method_def UVERBS_METHOD(_method_id) = { \
+ .id = _method_id, \
+ .num_attrs = ARRAY_SIZE(UVERBS_METHOD_ATTRS(_method_id)), \
+ .attrs = &UVERBS_METHOD_ATTRS(_method_id), \
+ }; \
+ ADD_UVERBS_METHODS(_name, _object_id, &UVERBS_METHOD(_method_id))
+
#endif
diff --git a/include/uapi/rdma/mlx5_user_ioctl_cmds.h b/include/uapi/rdma/mlx5_user_ioctl_cmds.h
index 233d5d140179..9c51801b9e64 100644
--- a/include/uapi/rdma/mlx5_user_ioctl_cmds.h
+++ b/include/uapi/rdma/mlx5_user_ioctl_cmds.h
@@ -149,4 +149,21 @@ enum mlx5_ib_flow_type {
MLX5_IB_FLOW_TYPE_MC_DEFAULT,
};
+enum mlx5_ib_create_flow_attrs {
+ MLX5_IB_ATTR_CREATE_FLOW_HANDLE = (1U << UVERBS_ID_NS_SHIFT),
+ MLX5_IB_ATTR_CREATE_FLOW_MATCH_VALUE,
+ MLX5_IB_ATTR_CREATE_FLOW_DEST_QP,
+ MLX5_IB_ATTR_CREATE_FLOW_DEST_DEVX,
+ MLX5_IB_ATTR_CREATE_FLOW_MATCHER,
+};
+
+enum mlx5_ib_destoy_flow_attrs {
+ MLX5_IB_ATTR_DESTROY_FLOW_HANDLE = (1U << UVERBS_ID_NS_SHIFT),
+};
+
+enum mlx5_ib_flow_methods {
+ MLX5_IB_METHOD_CREATE_FLOW = (1U << UVERBS_ID_NS_SHIFT),
+ MLX5_IB_METHOD_DESTROY_FLOW,
+};
+
#endif
--
2.14.4
^ permalink raw reply related
* [PATCH rdma-next v1 4/8] IB: Consider ib_flow creation by the KABI infrastructure
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
This patch considers the case that ib_flow is created by some device
driver with its specific parameters using the KABI infrastructure.
In that case both QP and ib_uflow_resources might not be applicable.
Downstream patches from this series use the above functionality.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/infiniband/core/uverbs_cmd.c | 4 ++++
drivers/infiniband/core/uverbs_std_types.c | 5 +++--
include/rdma/ib_verbs.h | 1 +
3 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index e4c298ac5946..312f9dbec09f 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -2771,6 +2771,9 @@ void ib_uverbs_flow_resources_free(struct ib_uflow_resources *uflow_res)
{
unsigned int i;
+ if (!uflow_res)
+ return;
+
for (i = 0; i < uflow_res->collection_num; i++)
atomic_dec(&uflow_res->collection[i]->usecnt);
@@ -3590,6 +3593,7 @@ int ib_uverbs_ex_create_flow(struct ib_uverbs_file *file,
}
atomic_inc(&qp->usecnt);
flow_id->qp = qp;
+ flow_id->device = qp->device;
flow_id->uobject = uobj;
uobj->object = flow_id;
uflow = container_of(uobj, typeof(*uflow), uobject);
diff --git a/drivers/infiniband/core/uverbs_std_types.c b/drivers/infiniband/core/uverbs_std_types.c
index 718c8430d364..c1e0492cc78a 100644
--- a/drivers/infiniband/core/uverbs_std_types.c
+++ b/drivers/infiniband/core/uverbs_std_types.c
@@ -54,9 +54,10 @@ static int uverbs_free_flow(struct ib_uobject *uobject,
struct ib_qp *qp = flow->qp;
int ret;
- ret = qp->device->destroy_flow(flow);
+ ret = flow->device->destroy_flow(flow);
if (!ret) {
- atomic_dec(&qp->usecnt);
+ if (qp)
+ atomic_dec(&qp->usecnt);
ib_uverbs_flow_resources_free(uflow->resources);
}
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 2696f1d730a1..d02bbbafbd3c 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -2079,6 +2079,7 @@ struct ib_flow_attr {
struct ib_flow {
struct ib_qp *qp;
+ struct ib_device *device;
struct ib_uobject *uobject;
};
--
2.14.4
^ permalink raw reply related
* [PATCH rdma-next v1 3/8] IB/mlx5: Introduce flow steering matcher object
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Introduce flow steering matcher object and its create and destroy
methods.
This matcher object holds some mlx5 specific driver properties that
matches the underlay device specification when an mlx5 flow steering
group is created.
It will be used in downstream patches to be part of mlx5 specific create
flow method.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
.../infiniband/core/uverbs_std_types_flow_action.c | 3 +-
drivers/infiniband/hw/mlx5/Makefile | 1 +
drivers/infiniband/hw/mlx5/flow.c | 137 +++++++++++++++++++++
drivers/infiniband/hw/mlx5/mlx5_ib.h | 11 ++
include/rdma/uverbs_ioctl.h | 2 +
include/uapi/rdma/mlx5_user_ioctl_cmds.h | 33 ++++-
6 files changed, 184 insertions(+), 3 deletions(-)
create mode 100644 drivers/infiniband/hw/mlx5/flow.c
diff --git a/drivers/infiniband/core/uverbs_std_types_flow_action.c b/drivers/infiniband/core/uverbs_std_types_flow_action.c
index c753a34cd984..adb9209c4710 100644
--- a/drivers/infiniband/core/uverbs_std_types_flow_action.c
+++ b/drivers/infiniband/core/uverbs_std_types_flow_action.c
@@ -376,8 +376,7 @@ static const struct uverbs_attr_spec uverbs_flow_action_esp_keymat[] = {
static const struct uverbs_attr_spec uverbs_flow_action_esp_replay[] = {
[IB_UVERBS_FLOW_ACTION_ESP_REPLAY_NONE] = {
.type = UVERBS_ATTR_TYPE_PTR_IN,
- /* No need to specify any data */
- UVERBS_ATTR_SIZE(0, 0),
+ UVERBS_ATTR_NO_DATA(),
},
[IB_UVERBS_FLOW_ACTION_ESP_REPLAY_BMP] = {
.type = UVERBS_ATTR_TYPE_PTR_IN,
diff --git a/drivers/infiniband/hw/mlx5/Makefile b/drivers/infiniband/hw/mlx5/Makefile
index 577e4c418bae..b8e4b15e2674 100644
--- a/drivers/infiniband/hw/mlx5/Makefile
+++ b/drivers/infiniband/hw/mlx5/Makefile
@@ -4,3 +4,4 @@ mlx5_ib-y := main.o cq.o doorbell.o qp.o mem.o srq.o mr.o ah.o mad.o gsi.o ib_vi
mlx5_ib-$(CONFIG_INFINIBAND_ON_DEMAND_PAGING) += odp.o
mlx5_ib-$(CONFIG_MLX5_ESWITCH) += ib_rep.o
mlx5_ib-$(CONFIG_INFINIBAND_USER_ACCESS) += devx.o
+mlx5_ib-$(CONFIG_INFINIBAND_USER_ACCESS) += flow.o
diff --git a/drivers/infiniband/hw/mlx5/flow.c b/drivers/infiniband/hw/mlx5/flow.c
new file mode 100644
index 000000000000..f1c16defe7ff
--- /dev/null
+++ b/drivers/infiniband/hw/mlx5/flow.c
@@ -0,0 +1,137 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/*
+ * Copyright (c) 2018, Mellanox Technologies inc. All rights reserved.
+ */
+
+#include <rdma/ib_user_verbs.h>
+#include <rdma/ib_verbs.h>
+#include <rdma/uverbs_types.h>
+#include <rdma/uverbs_ioctl.h>
+#include <rdma/mlx5_user_ioctl_cmds.h>
+#include <rdma/ib_umem.h>
+#include <linux/mlx5/driver.h>
+#include <linux/mlx5/fs.h>
+#include "mlx5_ib.h"
+
+#define UVERBS_MODULE_NAME mlx5_ib
+#include <rdma/uverbs_named_ioctl.h>
+
+static const struct uverbs_attr_spec mlx5_ib_flow_type[] = {
+ [MLX5_IB_FLOW_TYPE_NORMAL] = {
+ .type = UVERBS_ATTR_TYPE_PTR_IN,
+ .u.ptr = {
+ .len = sizeof(u16), /* data is priority */
+ .min_len = sizeof(u16),
+ }
+ },
+ [MLX5_IB_FLOW_TYPE_SNIFFER] = {
+ .type = UVERBS_ATTR_TYPE_PTR_IN,
+ UVERBS_ATTR_NO_DATA(),
+ },
+ [MLX5_IB_FLOW_TYPE_ALL_DEFAULT] = {
+ .type = UVERBS_ATTR_TYPE_PTR_IN,
+ UVERBS_ATTR_NO_DATA(),
+ },
+ [MLX5_IB_FLOW_TYPE_MC_DEFAULT] = {
+ .type = UVERBS_ATTR_TYPE_PTR_IN,
+ UVERBS_ATTR_NO_DATA(),
+ },
+};
+
+static int flow_matcher_cleanup(struct ib_uobject *uobject,
+ enum rdma_remove_reason why)
+{
+ struct mlx5_ib_flow_matcher *obj = uobject->object;
+ int ret;
+
+ ret = ib_destroy_usecnt(&obj->usecnt, why, uobject);
+ if (ret)
+ return ret;
+
+ kfree(obj);
+ return 0;
+}
+
+static int UVERBS_HANDLER(MLX5_IB_METHOD_FLOW_MATCHER_CREATE)(struct ib_device *ib_dev,
+ struct ib_uverbs_file *file,
+ struct uverbs_attr_bundle *attrs)
+{
+ struct ib_uobject *uobj = uverbs_attr_get_uobject(attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_CREATE_HANDLE);
+ struct mlx5_ib_dev *dev = to_mdev(uobj->context->device);
+ struct mlx5_ib_flow_matcher *obj;
+ int err;
+
+ obj = kzalloc(sizeof(struct mlx5_ib_flow_matcher), GFP_KERNEL);
+ if (!obj)
+ return -ENOMEM;
+
+ obj->mask_len = uverbs_attr_get_len(attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_MASK);
+ err = uverbs_copy_from(obj->matcher_mask.match_params,
+ attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_MASK);
+ if (err)
+ goto end;
+
+ obj->flow_type = uverbs_attr_get_enum_id(attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_FLOW_TYPE);
+
+ if (obj->flow_type == MLX5_IB_FLOW_TYPE_NORMAL) {
+ err = uverbs_copy_from(&obj->priority,
+ attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_FLOW_TYPE);
+ if (err)
+ goto end;
+ }
+
+ err = uverbs_copy_from(&obj->match_criteria_enable,
+ attrs,
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_CRITERIA);
+ if (err)
+ goto end;
+
+ uobj->object = obj;
+ obj->mdev = dev->mdev;
+ atomic_set(&obj->usecnt, 0);
+ return 0;
+
+end:
+ kfree(obj);
+ return err;
+}
+
+DECLARE_UVERBS_NAMED_METHOD(
+ MLX5_IB_METHOD_FLOW_MATCHER_CREATE,
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_FLOW_MATCHER_CREATE_HANDLE,
+ MLX5_IB_OBJECT_FLOW_MATCHER,
+ UVERBS_ACCESS_NEW,
+ UA_MANDATORY),
+ UVERBS_ATTR_PTR_IN(
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_MASK,
+ UVERBS_ATTR_SIZE(1, sizeof(struct mlx5_ib_match_params)),
+ UA_MANDATORY),
+ UVERBS_ATTR_ENUM_IN(
+ MLX5_IB_ATTR_FLOW_MATCHER_FLOW_TYPE,
+ mlx5_ib_flow_type,
+ UA_MANDATORY),
+ UVERBS_ATTR_PTR_IN(
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_CRITERIA,
+ UVERBS_ATTR_TYPE(u8),
+ UA_MANDATORY));
+
+DECLARE_UVERBS_NAMED_METHOD_DESTROY(
+ MLX5_IB_METHOD_FLOW_MATCHER_DESTROY,
+ UVERBS_ATTR_IDR(MLX5_IB_ATTR_FLOW_MATCHER_DESTROY_HANDLE,
+ MLX5_IB_OBJECT_FLOW_MATCHER,
+ UVERBS_ACCESS_DESTROY,
+ UA_MANDATORY));
+
+DECLARE_UVERBS_NAMED_OBJECT(MLX5_IB_OBJECT_FLOW_MATCHER,
+ UVERBS_TYPE_ALLOC_IDR(flow_matcher_cleanup),
+ &UVERBS_METHOD(MLX5_IB_METHOD_FLOW_MATCHER_CREATE),
+ &UVERBS_METHOD(MLX5_IB_METHOD_FLOW_MATCHER_DESTROY));
+
+DECLARE_UVERBS_OBJECT_TREE(flow_objects,
+ &UVERBS_OBJECT(MLX5_IB_OBJECT_FLOW_MATCHER));
+
diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h
index 67e86c8304a2..3d9544ec996f 100644
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -46,6 +46,7 @@
#include <rdma/ib_user_verbs.h>
#include <rdma/mlx5-abi.h>
#include <rdma/uverbs_ioctl.h>
+#include <rdma/mlx5_user_ioctl_cmds.h>
#define mlx5_ib_dbg(dev, format, arg...) \
pr_debug("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \
@@ -179,6 +180,16 @@ struct mlx5_ib_flow_handler {
struct ib_counters *ibcounters;
};
+struct mlx5_ib_flow_matcher {
+ struct mlx5_ib_match_params matcher_mask;
+ int mask_len;
+ enum mlx5_ib_flow_type flow_type;
+ u16 priority;
+ struct mlx5_core_dev *mdev;
+ atomic_t usecnt;
+ u8 match_criteria_enable;
+};
+
struct mlx5_ib_flow_db {
struct mlx5_ib_flow_prio prios[MLX5_IB_NUM_FLOW_FT];
struct mlx5_ib_flow_prio sniffer[MLX5_IB_NUM_SNIFFER_FTS];
diff --git a/include/rdma/uverbs_ioctl.h b/include/rdma/uverbs_ioctl.h
index 017ccf75890c..7f230d1ec2b8 100644
--- a/include/rdma/uverbs_ioctl.h
+++ b/include/rdma/uverbs_ioctl.h
@@ -204,6 +204,8 @@ struct uverbs_object_tree_def {
#define UVERBS_ATTR_SIZE(_min_len, _len) \
.u.ptr.min_len = _min_len, .u.ptr.len = _len
+#define UVERBS_ATTR_NO_DATA() UVERBS_ATTR_SIZE(0, 0)
+
/*
* Specifies a uapi structure that cannot be extended. The user must always
* supply the whole structure and nothing more. The structure must be declared
diff --git a/include/uapi/rdma/mlx5_user_ioctl_cmds.h b/include/uapi/rdma/mlx5_user_ioctl_cmds.h
index 1a05bb4b0b34..233d5d140179 100644
--- a/include/uapi/rdma/mlx5_user_ioctl_cmds.h
+++ b/include/uapi/rdma/mlx5_user_ioctl_cmds.h
@@ -33,6 +33,7 @@
#ifndef MLX5_USER_IOCTL_CMDS_H
#define MLX5_USER_IOCTL_CMDS_H
+#include <linux/types.h>
#include <rdma/ib_user_ioctl_cmds.h>
enum mlx5_ib_create_flow_action_attrs {
@@ -112,10 +113,40 @@ enum mlx5_ib_devx_umem_methods {
MLX5_IB_METHOD_DEVX_UMEM_DEREG,
};
-enum mlx5_ib_devx_objects {
+enum mlx5_ib_objects {
MLX5_IB_OBJECT_DEVX = (1U << UVERBS_ID_NS_SHIFT),
MLX5_IB_OBJECT_DEVX_OBJ,
MLX5_IB_OBJECT_DEVX_UMEM,
+ MLX5_IB_OBJECT_FLOW_MATCHER,
+};
+
+enum mlx5_ib_flow_matcher_create_attrs {
+ MLX5_IB_ATTR_FLOW_MATCHER_CREATE_HANDLE = (1U << UVERBS_ID_NS_SHIFT),
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_MASK,
+ MLX5_IB_ATTR_FLOW_MATCHER_FLOW_TYPE,
+ MLX5_IB_ATTR_FLOW_MATCHER_MATCH_CRITERIA,
+};
+
+enum mlx5_ib_flow_matcher_destroy_attrs {
+ MLX5_IB_ATTR_FLOW_MATCHER_DESTROY_HANDLE = (1U << UVERBS_ID_NS_SHIFT),
+};
+
+enum mlx5_ib_flow_matcher_methods {
+ MLX5_IB_METHOD_FLOW_MATCHER_CREATE = (1U << UVERBS_ID_NS_SHIFT),
+ MLX5_IB_METHOD_FLOW_MATCHER_DESTROY,
+};
+
+#define MLX5_IB_DW_MATCH_PARAM 0x80
+
+struct mlx5_ib_match_params {
+ __u32 match_params[MLX5_IB_DW_MATCH_PARAM];
+};
+
+enum mlx5_ib_flow_type {
+ MLX5_IB_FLOW_TYPE_NORMAL,
+ MLX5_IB_FLOW_TYPE_SNIFFER,
+ MLX5_IB_FLOW_TYPE_ALL_DEFAULT,
+ MLX5_IB_FLOW_TYPE_MC_DEFAULT,
};
#endif
--
2.14.4
^ permalink raw reply related
* Re: [PATCH net-next v2 04/10] r8169: use phy_ethtool_(g|s)et_link_ksettings
From: Florian Fainelli @ 2018-07-11 11:10 UTC (permalink / raw)
To: Heiner Kallweit, David Miller, Andrew Lunn,
Realtek linux nic maintainers
Cc: netdev@vger.kernel.org
In-Reply-To: <912f8550-b0a3-61db-6a89-5d2e41435dd1@gmail.com>
On 07/10/2018 11:39 AM, Heiner Kallweit wrote:
> Use phy_ethtool_(g|s)et_link_ksettings() for the respective ethtool_ops
> callbacks.
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* [PATCH mlx5-next v1 2/8] net/mlx5: Add support for flow table destination number
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Add support to set a destination from a flow table number.
This functionality will be used in downstream patches from this
series by the DEVX stuff.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
.../mellanox/mlx5/core/diag/fs_tracepoint.c | 3 +++
drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c | 23 +++++++++++++---------
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 4 +++-
include/linux/mlx5/fs.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 1 +
5 files changed, 22 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
index b3820a34e773..0f11fff32a9b 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fs_tracepoint.c
@@ -240,6 +240,9 @@ const char *parse_fs_dst(struct trace_seq *p,
case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE:
trace_seq_printf(p, "ft=%p\n", dst->ft);
break;
+ case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM:
+ trace_seq_printf(p, "ft_num=%u\n", dst->ft_num);
+ break;
case MLX5_FLOW_DESTINATION_TYPE_TIR:
trace_seq_printf(p, "tir=%u\n", dst->tir_num);
break;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
index 5a00deff5457..3a04551696c0 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c
@@ -362,18 +362,20 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
int list_size = 0;
list_for_each_entry(dst, &fte->node.children, node.list) {
- unsigned int id;
+ unsigned int id, type = dst->dest_attr.type;
- if (dst->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_COUNTER)
+ if (type == MLX5_FLOW_DESTINATION_TYPE_COUNTER)
continue;
- MLX5_SET(dest_format_struct, in_dests, destination_type,
- dst->dest_attr.type);
- if (dst->dest_attr.type ==
- MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE) {
+ switch (type) {
+ case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM:
+ id = dst->dest_attr.ft_num;
+ type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE;
+ break;
+ case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE:
id = dst->dest_attr.ft->id;
- } else if (dst->dest_attr.type ==
- MLX5_FLOW_DESTINATION_TYPE_VPORT) {
+ break;
+ case MLX5_FLOW_DESTINATION_TYPE_VPORT:
id = dst->dest_attr.vport.num;
MLX5_SET(dest_format_struct, in_dests,
destination_eswitch_owner_vhca_id_valid,
@@ -381,9 +383,12 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev,
MLX5_SET(dest_format_struct, in_dests,
destination_eswitch_owner_vhca_id,
dst->dest_attr.vport.vhca_id);
- } else {
+ break;
+ default:
id = dst->dest_attr.tir_num;
}
+
+ MLX5_SET(dest_format_struct, in_dests, destination_type, type);
MLX5_SET(dest_format_struct, in_dests, destination_id, id);
in_dests += MLX5_ST_SZ_BYTES(dest_format_struct);
list_size++;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index eba113cf1117..69aa298a0b1c 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -1356,7 +1356,9 @@ static bool mlx5_flow_dests_cmp(struct mlx5_flow_destination *d1,
(d1->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE &&
d1->ft == d2->ft) ||
(d1->type == MLX5_FLOW_DESTINATION_TYPE_TIR &&
- d1->tir_num == d2->tir_num))
+ d1->tir_num == d2->tir_num) ||
+ (d1->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM &&
+ d1->ft_num == d2->ft_num))
return true;
}
diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h
index 757b4a30281e..a45febcf6b51 100644
--- a/include/linux/mlx5/fs.h
+++ b/include/linux/mlx5/fs.h
@@ -89,6 +89,7 @@ struct mlx5_flow_destination {
enum mlx5_flow_destination_type type;
union {
u32 tir_num;
+ u32 ft_num;
struct mlx5_flow_table *ft;
struct mlx5_fc *counter;
struct {
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index 44a6ce01c3bb..fb89cc519703 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -1181,6 +1181,7 @@ enum mlx5_flow_destination_type {
MLX5_FLOW_DESTINATION_TYPE_PORT = 0x99,
MLX5_FLOW_DESTINATION_TYPE_COUNTER = 0x100,
+ MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM = 0x101,
};
struct mlx5_ifc_dest_format_struct_bits {
--
2.14.4
^ permalink raw reply related
* [PATCH mlx5-next v1 1/8] net/mlx5: Add forward compatible support for the FTE match data
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
In-Reply-To: <20180711111045.6282-1-leon@kernel.org>
From: Yishai Hadas <yishaih@mellanox.com>
Use the PRM size including the reserved when working with the FTE
match data.
This comes to support forward compatibility for cases that current
reserved data will be exposed by the firmware and could be used by an
application by DEVX without changing the kernel.
Also drop some driver checks around the match criteria leaving the work
for firmware to enable forward compatibility for future bits there.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 77 +----------------------
1 file changed, 1 insertion(+), 76 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
index 49a75d31185e..eba113cf1117 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c
@@ -309,89 +309,17 @@ static struct fs_prio *find_prio(struct mlx5_flow_namespace *ns,
return NULL;
}
-static bool check_last_reserved(const u32 *match_criteria)
-{
- char *match_criteria_reserved =
- MLX5_ADDR_OF(fte_match_param, match_criteria, MLX5_FTE_MATCH_PARAM_RESERVED);
-
- return !match_criteria_reserved[0] &&
- !memcmp(match_criteria_reserved, match_criteria_reserved + 1,
- MLX5_FLD_SZ_BYTES(fte_match_param,
- MLX5_FTE_MATCH_PARAM_RESERVED) - 1);
-}
-
-static bool check_valid_mask(u8 match_criteria_enable, const u32 *match_criteria)
-{
- if (match_criteria_enable & ~(
- (1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_OUTER_HEADERS) |
- (1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS) |
- (1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_INNER_HEADERS) |
- (1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS_2)))
- return false;
-
- if (!(match_criteria_enable &
- 1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_OUTER_HEADERS)) {
- char *fg_type_mask = MLX5_ADDR_OF(fte_match_param,
- match_criteria, outer_headers);
-
- if (fg_type_mask[0] ||
- memcmp(fg_type_mask, fg_type_mask + 1,
- MLX5_ST_SZ_BYTES(fte_match_set_lyr_2_4) - 1))
- return false;
- }
-
- if (!(match_criteria_enable &
- 1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS)) {
- char *fg_type_mask = MLX5_ADDR_OF(fte_match_param,
- match_criteria, misc_parameters);
-
- if (fg_type_mask[0] ||
- memcmp(fg_type_mask, fg_type_mask + 1,
- MLX5_ST_SZ_BYTES(fte_match_set_misc) - 1))
- return false;
- }
-
- if (!(match_criteria_enable &
- 1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_INNER_HEADERS)) {
- char *fg_type_mask = MLX5_ADDR_OF(fte_match_param,
- match_criteria, inner_headers);
-
- if (fg_type_mask[0] ||
- memcmp(fg_type_mask, fg_type_mask + 1,
- MLX5_ST_SZ_BYTES(fte_match_set_lyr_2_4) - 1))
- return false;
- }
-
- if (!(match_criteria_enable &
- 1 << MLX5_CREATE_FLOW_GROUP_IN_MATCH_CRITERIA_ENABLE_MISC_PARAMETERS_2)) {
- char *fg_type_mask = MLX5_ADDR_OF(fte_match_param,
- match_criteria, misc_parameters_2);
-
- if (fg_type_mask[0] ||
- memcmp(fg_type_mask, fg_type_mask + 1,
- MLX5_ST_SZ_BYTES(fte_match_set_misc2) - 1))
- return false;
- }
-
- return check_last_reserved(match_criteria);
-}
-
static bool check_valid_spec(const struct mlx5_flow_spec *spec)
{
int i;
- if (!check_valid_mask(spec->match_criteria_enable, spec->match_criteria)) {
- pr_warn("mlx5_core: Match criteria given mismatches match_criteria_enable\n");
- return false;
- }
-
for (i = 0; i < MLX5_ST_SZ_DW_MATCH_PARAM; i++)
if (spec->match_value[i] & ~spec->match_criteria[i]) {
pr_warn("mlx5_core: match_value differs from match_criteria\n");
return false;
}
- return check_last_reserved(spec->match_value);
+ return true;
}
static struct mlx5_flow_root_namespace *find_root(struct fs_node *node)
@@ -1158,9 +1086,6 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft,
struct mlx5_flow_group *fg;
int err;
- if (!check_valid_mask(match_criteria_enable, match_criteria))
- return ERR_PTR(-EINVAL);
-
if (ft->autogroup.active)
return ERR_PTR(-EPERM);
--
2.14.4
^ permalink raw reply related
* [PATCH rdma-next v1 0/8] Support mlx5 flow steering with RAW data
From: Leon Romanovsky @ 2018-07-11 11:10 UTC (permalink / raw)
To: Doug Ledford, Jason Gunthorpe
Cc: Leon Romanovsky, RDMA mailing list, Yishai Hadas, Saeed Mahameed,
linux-netdev
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=UTF-8, Size: 3211 bytes --]
From: Leon Romanovsky <leonro@mellanox.com>
Changelog v0->v1:
* Fixed ADD_UVERBS_ATTRIBUTES_SIMPLE macro to pass the real address.
Â* Replaced UA_ALLOC_AND_COPY to regular copy_from
* Added UVERBS_ATTR_NO_DATA new macro for cleaner code.
* Used ib_dev from uobj when it exists.
* ib_is_destroy_retryable was replaced by ib_destroy_usecnt
-------------------------------------------------------------------------
>From Yishai:
This series introduces vendor create and destroy flow methods on the
uverbs flow object by using the KABI infra-structure.
It's done in a way that enables the driver to get its specific device
attributes in a raw data to match its underlay specification while still
using the generic ib_flow object for cleanup and code sharing.
In addition, a specific mlx5 matcher object and its create/destroy
methods were introduced. This object matches the underlay flow steering
mask specification and is used as part of mlx5 create flow input data.
This series supports IB_QP/TIR as its flow steering destination as
applicable today via the ib_create_flow API, however, it adds also an
option to work with DEVX object which its destination can be both TIR
and flow table.
Few changes were done in the mlx5 core layer to support forward
compatible for the device specification raw data and to support flow
table when the DEVX destination is used.
As part of this series the default IB destroy handler
(i.e. uverbs_destroy_def_handler()) was exposed from IB core to be
used by the drivers and existing code was refactored to use it.
Thanks
Yishai Hadas (8):
net/mlx5: Add forward compatible support for the FTE match data
net/mlx5: Add support for flow table destination number
IB/mlx5: Introduce flow steering matcher object
IB: Consider ib_flow creation by the KABI infrastructure
IB/mlx5: Introduce vendor create and destroy flow methods
IB/mlx5: Support adding flow steering rule by raw data
IB/mlx5: Add support for a flow table destination
IB/mlx5: Expose vendor flow trees
drivers/infiniband/core/uverbs_cmd.c | 4 +
drivers/infiniband/core/uverbs_std_types.c | 5 +-
.../infiniband/core/uverbs_std_types_flow_action.c | 3 +-
drivers/infiniband/hw/mlx5/Makefile | 1 +
drivers/infiniband/hw/mlx5/devx.c | 22 ++
drivers/infiniband/hw/mlx5/flow.c | 254 +++++++++++++++++++++
drivers/infiniband/hw/mlx5/main.c | 230 +++++++++++++++++--
drivers/infiniband/hw/mlx5/mlx5_ib.h | 31 +++
.../mellanox/mlx5/core/diag/fs_tracepoint.c | 3 +
drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c | 23 +-
drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 81 +------
include/linux/mlx5/fs.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 1 +
include/rdma/ib_verbs.h | 15 ++
include/rdma/uverbs_ioctl.h | 2 +
include/rdma/uverbs_named_ioctl.h | 29 ++-
include/uapi/rdma/mlx5_user_ioctl_cmds.h | 50 +++-
17 files changed, 635 insertions(+), 120 deletions(-)
create mode 100644 drivers/infiniband/hw/mlx5/flow.c
^ permalink raw reply
* Re: [PATCH net-next v2 03/10] r8169: replace open-coded PHY soft reset with genphy_soft_reset
From: Florian Fainelli @ 2018-07-11 11:10 UTC (permalink / raw)
To: Heiner Kallweit, David Miller, Andrew Lunn,
Realtek linux nic maintainers
Cc: netdev@vger.kernel.org
In-Reply-To: <ff517e17-8721-f273-ce5e-1ec513b9338a@gmail.com>
On 07/10/2018 11:39 AM, Heiner Kallweit wrote:
> Use genphy_soft_reset() instead of open-coding a PHY soft reset. We have
> to do an explicit PHY soft reset because some chips use the genphy driver
> which uses a no-op as soft_reset callback.
>
> Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
> Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
--
Florian
^ permalink raw reply
* Re: [PATCH] net: convert gro_count to bitmask
From: Stefano Brivio @ 2018-07-11 10:51 UTC (permalink / raw)
To: Li RongQing; +Cc: netdev, Eric Dumazet
In-Reply-To: <1531300553-21413-1-git-send-email-lirongqing@baidu.com>
On Wed, 11 Jul 2018 17:15:53 +0800
Li RongQing <lirongqing@baidu.com> wrote:
> @@ -5380,6 +5382,12 @@ static enum gro_result dev_gro_receive(struct napi_struct *napi, struct sk_buff
> if (grow > 0)
> gro_pull_from_frag0(skb, grow);
> ok:
> + if (napi->gro_hash[hash].count)
> + if (!test_bit(hash, &napi->gro_bitmask))
> + set_bit(hash, &napi->gro_bitmask);
> + else if (test_bit(hash, &napi->gro_bitmask))
> + clear_bit(hash, &napi->gro_bitmask);
This might not do what you want.
--
Stefano
^ permalink raw reply
* [PATCH net 2/2] sfc: hold filter_sem consistently during reset
From: Bert Kenward @ 2018-07-11 10:45 UTC (permalink / raw)
To: Dave Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <ffa1a8d2-7906-29f9-6089-f968aa376681@solarflare.com>
We should take and release the filter_sem consistently during the
reset process, in the same manner as the mac_lock and reset_lock.
For lockdep consistency we also take the filter_sem for write around
other calls to efx->type->init().
Fixes: c2bebe37c6b6 ("sfc: give ef10 its own rwsem in the filter table instead of filter_lock")
Signed-off-by: Bert Kenward <bkenward@solarflare.com>
---
drivers/net/ethernet/sfc/efx.c | 17 ++++++++---------
1 file changed, 8 insertions(+), 9 deletions(-)
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c
index 570ec72266f3..ce3a177081a8 100644
--- a/drivers/net/ethernet/sfc/efx.c
+++ b/drivers/net/ethernet/sfc/efx.c
@@ -1871,12 +1871,6 @@ static void efx_remove_filters(struct efx_nic *efx)
up_write(&efx->filter_sem);
}
-static void efx_restore_filters(struct efx_nic *efx)
-{
- down_read(&efx->filter_sem);
- efx->type->filter_table_restore(efx);
- up_read(&efx->filter_sem);
-}
/**************************************************************************
*
@@ -2688,6 +2682,7 @@ void efx_reset_down(struct efx_nic *efx, enum reset_type method)
efx_disable_interrupts(efx);
mutex_lock(&efx->mac_lock);
+ down_write(&efx->filter_sem);
mutex_lock(&efx->rss_lock);
if (efx->port_initialized && method != RESET_TYPE_INVISIBLE &&
method != RESET_TYPE_DATAPATH)
@@ -2745,9 +2740,8 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok)
if (efx->type->rx_restore_rss_contexts)
efx->type->rx_restore_rss_contexts(efx);
mutex_unlock(&efx->rss_lock);
- down_read(&efx->filter_sem);
- efx_restore_filters(efx);
- up_read(&efx->filter_sem);
+ efx->type->filter_table_restore(efx);
+ up_write(&efx->filter_sem);
if (efx->type->sriov_reset)
efx->type->sriov_reset(efx);
@@ -2764,6 +2758,7 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok)
efx->port_initialized = false;
mutex_unlock(&efx->rss_lock);
+ up_write(&efx->filter_sem);
mutex_unlock(&efx->mac_lock);
return rc;
@@ -3473,7 +3468,9 @@ static int efx_pci_probe_main(struct efx_nic *efx)
efx_init_napi(efx);
+ down_write(&efx->filter_sem);
rc = efx->type->init(efx);
+ up_write(&efx->filter_sem);
if (rc) {
netif_err(efx, probe, efx->net_dev,
"failed to initialise NIC\n");
@@ -3765,7 +3762,9 @@ static int efx_pm_resume(struct device *dev)
rc = efx->type->reset(efx, RESET_TYPE_ALL);
if (rc)
return rc;
+ down_write(&efx->filter_sem);
rc = efx->type->init(efx);
+ up_write(&efx->filter_sem);
if (rc)
return rc;
rc = efx_pm_thaw(dev);
--
2.13.6
^ 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