OSDN Git Service

uclinux-h8/linux.git
3 years agonet: macb: add userio bits as platform configuration
Claudiu Beznea [Wed, 9 Dec 2020 13:03:32 +0000 (15:03 +0200)]
net: macb: add userio bits as platform configuration

This is necessary for SAMA7G5 as it uses different values for
PHY interface and also introduces hdfctlen bit.

Signed-off-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: sched: incorrect Kconfig dependencies on Netfilter modules
Pablo Neira Ayuso [Tue, 8 Dec 2020 20:47:07 +0000 (21:47 +0100)]
net: sched: incorrect Kconfig dependencies on Netfilter modules

- NET_ACT_CONNMARK and NET_ACT_CTINFO only require conntrack support.
- NET_ACT_IPT only requires NETFILTER_XTABLES symbols, not
  IP_NF_IPTABLES. After this patch, NET_ACT_IPT becomes consistent
  with NET_EMATCH_IPT. NET_ACT_IPT dependency on IP_NF_IPTABLES predates
  Linux-2.6.12-rc2 (initial git repository build).

Fixes: 22a5dc0e5e3e ("net: sched: Introduce connmark action")
Fixes: 24ec483cec98 ("net: sched: Introduce act_ctinfo action")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Link: https://lore.kernel.org/r/20201208204707.11268-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'xdp-redirect-implementation-for-ena-driver'
Jakub Kicinski [Wed, 9 Dec 2020 23:26:42 +0000 (15:26 -0800)]
Merge branch 'xdp-redirect-implementation-for-ena-driver'

Shay Agroskin says:

====================
XDP Redirect implementation for ENA driver

ENA is adding XDP Redirect support for its driver and some other
small tweaks.

This series adds the following:

- Make log messages in the driver have a uniform format using
  netdev_* function
- Improve code readability
- Add support for XDP Redirect
====================

Link: https://lore.kernel.org/r/20201208180208.26111-1-shayagr@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: introduce ndo_xdp_xmit() function for XDP_REDIRECT
Shay Agroskin [Tue, 8 Dec 2020 18:02:08 +0000 (20:02 +0200)]
net: ena: introduce ndo_xdp_xmit() function for XDP_REDIRECT

This patch implements the ndo_xdp_xmit() net_device function which is
called when a packet is redirected to this driver using an
XDP_REDIRECT directive.

The function receives an array of xdp frames that it needs to xmit.
The TX queues that are used to xmit these frames are the XDP
queues used by the XDP_TX flow. Therefore a lock is added to synchronize
both flows (XDP_TX and XDP_REDIRECT).

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: use xdp_return_frame() to free xdp frames
Shay Agroskin [Tue, 8 Dec 2020 18:02:07 +0000 (20:02 +0200)]
net: ena: use xdp_return_frame() to free xdp frames

XDP subsystem has a function to free XDP frames and their associated
pages. Using this function would help the driver's XDP implementation to
adjust to new changes in the XDP subsystem in the kernel (e.g.
introduction of XDP MB).

Also, remove 'xdp_rx_page' field from ena_tx_buffer struct since it is
no longer used.

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: introduce XDP redirect implementation
Shay Agroskin [Tue, 8 Dec 2020 18:02:06 +0000 (20:02 +0200)]
net: ena: introduce XDP redirect implementation

This patch adds a partial support for the XDP_REDIRECT directive which
instructs the driver to pass the packet to an interface specified by the
program. The directive is passed to the driver by calling bpf_redirect()
or bpf_redirect_map() functions from the eBPF program.

To lay the ground for integration with the existing XDP TX
implementation the patch removes the redundant page ref count increase
in ena_xdp_xmit_frame() and then decrease in ena_clean_rx_irq(). Instead
it only DMA unmaps descriptors for which XDP TX or REDIRECT directive
was received.

The XDP Redirect support is still missing .ndo_xdp_xmit function
implementation, which allows to redirect packet to an ENA interface,
which would be added in a later patch.

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: use xdp_frame in XDP TX flow
Shay Agroskin [Tue, 8 Dec 2020 18:02:05 +0000 (20:02 +0200)]
net: ena: use xdp_frame in XDP TX flow

Rename the ena_xdp_xmit_buff() function to ena_xdp_xmit_frame() and pass
it an xdp_frame struct instead of xdp_buff.
This change lays the ground for XDP redirect implementation which uses
xdp_frames when 'xmit'ing packets.

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: aggregate stats increase into a function
Shay Agroskin [Tue, 8 Dec 2020 18:02:04 +0000 (20:02 +0200)]
net: ena: aggregate stats increase into a function

Introduce ena_increase_stat() function to increase statistics by a
certain number.
The function includes the
    - lock aquire (on 32bit machines)
    - stat increase
    - lock release (on 32bit machines)

line sequence that is ubiquitous across the driver.

The function increases a single stat at a time and several stats which
are increased together weren't put into a function to avoid
calling the function several times for each stat which looks bad and
might decrease performance.

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: fix coding style nits
Shay Agroskin [Tue, 8 Dec 2020 18:02:03 +0000 (20:02 +0200)]
net: ena: fix coding style nits

This commit fixes two nits, but it does not generate any change to binary
because of the optimization of gcc.

  - use `count` instead of `channels->combined_count`
  - change return type from `int` to `bool`

Also add spaces and change macro order in OR assignment to make the code
easier to read.

Signed-off-by: Sameeh Jubran <sameehj@amazon.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.co.jp>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: store values in their appropriate variables types
Shay Agroskin [Tue, 8 Dec 2020 18:02:02 +0000 (20:02 +0200)]
net: ena: store values in their appropriate variables types

This patch changes some of the variables types to match the values they
hold. These wrong types fail some of our static checkers that search for
accidental conversions in our driver.

Signed-off-by: Ido Segev <idose@amazon.com>
Signed-off-by: Igor Chauskin <igorch@amazon.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: add device distinct log prefix to files
Shay Agroskin [Tue, 8 Dec 2020 18:02:01 +0000 (20:02 +0200)]
net: ena: add device distinct log prefix to files

ENA logs are adjusted to display the full ENA representation to
distinct each ENA device in case of multiple interfaces.
Using netdev_err/warn and dev_info functions for logging provides
uniform printing with clear distinction of the device and interface.

This patch changes all printing in ena_com files to use netev_* logging
functions except for messages of info level. Log functions of that level
would be printed with dev_info because of the early stage they are
called in when net_device struct isn't yet registered.

To allow using netdev_* functions in all ena_com functions, a pointer to
the net_device was added to ena_com_dev struct.

The patch also adds some log messages to make driver debugging easier.

Signed-off-by: Amit Bernstein <amitbern@amazon.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: ena: use constant value for net_device allocation
Shay Agroskin [Tue, 8 Dec 2020 18:02:00 +0000 (20:02 +0200)]
net: ena: use constant value for net_device allocation

The patch changes the maximum number of RX/TX queues it advertises to
the kernel (via alloc_etherdev_mq()) from a value received from the
device to a constant value which is the minimum between 128 and the
number of CPUs in the system.

By allocating the net_device struct with a constant number of queues,
the driver is able to allocate it at a much earlier stage, before
calling any ena_com functions. This would allow to make all log prints
in ena_com to use netdev_* log functions instead or current pr_* ones.

Note:
netdev_* prints in ena_com functions that are called before
net_device registration in ena_probe() might print messages that are
a bit ugly (with strings like "(unnamed net_device) (uninitialized)").
However we decided to use netdev_* prints in these functions anyway,
for the sake of getting better messages later, when ena_com functions
are called after ena_probe() form other parts of the driver.
See discussion about this decision in [1].

[1] http://www.mail-archive.com/netdev@vger.kernel.org/msg353590.html

Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoptp: clockmatrix: deprecate firmware older than 4.8.7
Min Li [Tue, 8 Dec 2020 15:41:57 +0000 (10:41 -0500)]
ptp: clockmatrix: deprecate firmware older than 4.8.7

Add deprecated flag to indicate < v4.8.7.
Fix idtcm_enable_tod() call correct settime().

Signed-off-by: Min Li <min.li.xe@renesas.com>
Link: https://lore.kernel.org/r/1607442117-13661-4-git-send-email-min.li.xe@renesas.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoptp: clockmatrix: Fix non-zero phase_adj is lost after snap
Min Li [Tue, 8 Dec 2020 15:41:56 +0000 (10:41 -0500)]
ptp: clockmatrix: Fix non-zero phase_adj is lost after snap

Fix non-zero phase_adj is lost after snap. Use ktime_sub
to do ktime_t subtraction.

Signed-off-by: Min Li <min.li.xe@renesas.com>
Link: https://lore.kernel.org/r/1607442117-13661-3-git-send-email-min.li.xe@renesas.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoptp: clockmatrix: remove 5 second delay before entering write phase mode
Min Li [Tue, 8 Dec 2020 15:41:55 +0000 (10:41 -0500)]
ptp: clockmatrix: remove 5 second delay before entering write phase mode

Remove write phase mode 5 second setup delay, not needed.

Signed-off-by: Min Li <min.li.xe@renesas.com>
Link: https://lore.kernel.org/r/1607442117-13661-2-git-send-email-min.li.xe@renesas.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoptp: clockmatrix: reset device and check BOOT_STATUS
Min Li [Tue, 8 Dec 2020 15:41:54 +0000 (10:41 -0500)]
ptp: clockmatrix: reset device and check BOOT_STATUS

SM_RESET device only when loading full configuration and check
for BOOT_STATUS. Also remove polling for write trigger done in
_idtcm_settime().

Changes since v1:
-Correct warnings from strict checkpatch

Signed-off-by: Min Li <min.li.xe@renesas.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Link: https://lore.kernel.org/r/1607442117-13661-1-git-send-email-min.li.xe@renesas.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agovrf: handle CONFIG_IPV6 not set for vrf_add_mac_header_if_unset()
Andrea Mayer [Tue, 8 Dec 2020 17:52:10 +0000 (18:52 +0100)]
vrf: handle CONFIG_IPV6 not set for vrf_add_mac_header_if_unset()

The vrf_add_mac_header_if_unset() is defined within a conditional
compilation block which depends on the CONFIG_IPV6 macro.
However, the vrf_add_mac_header_if_unset() needs to be called also by IPv4
related code and when the CONFIG_IPV6 is not set, this function is missing.
As a consequence, the build process stops reporting the error:

 ERROR: implicit declaration of function 'vrf_add_mac_header_if_unset'

The problem is solved by *only* moving functions
vrf_add_mac_header_if_unset() and vrf_prepare_mac_header() out of the
conditional block.

Reported-by: kernel test robot <lkp@intel.com>
Fixes: 0489390882202 ("vrf: add mac header for tunneled packets when sniffer is attached")
Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20201208175210.8906-1-andrea.mayer@uniroma2.it
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: atheros: simplify the return expression of atl2_phy_setup_autoneg_adv()
Zheng Yongjun [Tue, 8 Dec 2020 13:57:30 +0000 (21:57 +0800)]
net: atheros: simplify the return expression of atl2_phy_setup_autoneg_adv()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodrivers: net: qlcnic: simplify the return expression of qlcnic_sriov_vf_shutdown()
Zheng Yongjun [Tue, 8 Dec 2020 13:54:37 +0000 (21:54 +0800)]
drivers: net: qlcnic: simplify the return expression of qlcnic_sriov_vf_shutdown()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agodrivers: net: ionic: simplify the return expression of ionic_set_rxfh()
Zheng Yongjun [Tue, 8 Dec 2020 13:53:53 +0000 (21:53 +0800)]
drivers: net: ionic: simplify the return expression of ionic_set_rxfh()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: openvswitch: conntrack: simplify the return expression of ovs_ct_limit_get_defau...
Zheng Yongjun [Tue, 8 Dec 2020 12:13:53 +0000 (20:13 +0800)]
net: openvswitch: conntrack: simplify the return expression of ovs_ct_limit_get_default_limit()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Reviewed-by: Eelco Chaudron <echaudro@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: core: devlink: simplify the return expression of devlink_nl_cmd_trap_set_doit()
Zheng Yongjun [Tue, 8 Dec 2020 12:10:46 +0000 (20:10 +0800)]
net: core: devlink: simplify the return expression of devlink_nl_cmd_trap_set_doit()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: ipv6: rpl_iptunnel: simplify the return expression of rpl_do_srh()
Zheng Yongjun [Tue, 8 Dec 2020 12:09:49 +0000 (20:09 +0800)]
net: ipv6: rpl_iptunnel: simplify the return expression of rpl_do_srh()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/sched: cls_u32: simplify the return expression of u32_reoffload_knode()
Zheng Yongjun [Tue, 8 Dec 2020 12:08:22 +0000 (20:08 +0800)]
net/sched: cls_u32: simplify the return expression of u32_reoffload_knode()

Simplify the return expression.

Signed-off-by: Zheng Yongjun <zhengyongjun3@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: dsa: mt7530: support setting ageing time
DENG Qingfang [Tue, 8 Dec 2020 07:00:28 +0000 (15:00 +0800)]
net: dsa: mt7530: support setting ageing time

MT7530 has a global address age control register, so use it to set
ageing time.

The applied timer is (AGE_CNT + 1) * (AGE_UNIT + 1) seconds

Signed-off-by: DENG Qingfang <dqfext@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'GVE-Raw-Addressing'
David S. Miller [Wed, 9 Dec 2020 00:06:29 +0000 (16:06 -0800)]
Merge branch 'GVE-Raw-Addressing'

David Awogbemil says:

====================
GVE Raw Addressing
Patchset description:
This  patchset introduces "raw addressing" mode to the GVE driver.
Previously (in "queue_page_list" or "qpl" mode), the driver would
pre-allocate and dma_map buffers to be used on egress and ingress.
On egress, it would copy data from the skb provided to the
pre-allocated buffers - this was expensive.
In raw addressing mode, the driver can avoid this copy and simply
dma_map the skb's data so that the NIC can use it.
On ingress, the driver passes buffers up to the networking stack and
then frees and reallocates buffers when necessary instead of using
skb_copy_to_linear_data.
Patch 3 separates the page refcount tracking mechanism
into a function gve_rx_can_recycle_buffer which uses get_page - this will
be changed in a future patch to eliminate the use of get_page in tracking
page refcounts.

Changes from v9:
  Patch 4: Use u64, not u32 for new tx stat counters.
====================

Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agogve: Add support for raw addressing in the tx path
Catherine Sullivan [Mon, 7 Dec 2020 22:45:26 +0000 (14:45 -0800)]
gve: Add support for raw addressing in the tx path

During TX, skbs' data addresses are dma_map'ed and passed to the NIC.
This means that the device can perform DMA directly from these addresses
and the driver does not have to copy the buffer content into
pre-allocated buffers/qpls (as in qpl mode).

Reviewed-by: Yangchun Fu <yangchun@google.com>
Signed-off-by: Catherine Sullivan <csully@google.com>
Signed-off-by: David Awogbemila <awogbemila@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agogve: Rx Buffer Recycling
David Awogbemila [Mon, 7 Dec 2020 22:45:25 +0000 (14:45 -0800)]
gve: Rx Buffer Recycling

This patch lets the driver reuse buffers that have been freed by the
networking stack.

In the raw addressing case, this allows the driver avoid allocating new
buffers.
In the qpl case, the driver can avoid copies.

This patch separates the page refcount tracking mechanism
into a function gve_rx_can_recycle_buffer which uses get_page - this will
be changed in a future patch to entirely eliminate the use of get_page in
tracking page refcounts.

Signed-off-by: David Awogbemila <awogbemila@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agogve: Add support for raw addressing to the rx path
Catherine Sullivan [Mon, 7 Dec 2020 22:45:24 +0000 (14:45 -0800)]
gve: Add support for raw addressing to the rx path

Add support to use raw dma addresses in the rx path. Due to this new
support we can alloc a new buffer instead of making a copy.

RX buffers are handed to the networking stack and are
re-allocated as needed, avoiding the need to use
skb_copy_to_linear_data() as in "qpl" mode.

Reviewed-by: Yangchun Fu <yangchun@google.com>
Signed-off-by: Catherine Sullivan <csully@google.com>
Signed-off-by: David Awogbemila <awogbemila@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agogve: Add support for raw addressing device option
Catherine Sullivan [Mon, 7 Dec 2020 22:45:23 +0000 (14:45 -0800)]
gve: Add support for raw addressing device option

Add support to describe device for parsing device options. As
the first device option, add raw addressing.

"Raw Addressing" mode (as opposed to the current "qpl" mode) is an
operational mode which allows the driver avoid bounce buffer copies
which it currently performs using pre-allocated qpls (queue_page_lists)
when sending and receiving packets.
For egress packets, the provided skb data addresses will be dma_map'ed and
passed to the device, allowing the NIC can perform DMA directly - the
driver will not have to copy the buffer content into pre-allocated
buffers/qpls (as in qpl mode).
For ingress packets, copies are also eliminated as buffers are handed to
the networking stack and then recycled or re-allocated as
necessary, avoiding the use of skb_copy_to_linear_data().

This patch only introduces the option to the driver.
Subsequent patches will add the ingress and egress functionality.

Reviewed-by: Yangchun Fu <yangchun@google.com>
Signed-off-by: Catherine Sullivan <csully@google.com>
Signed-off-by: David Awogbemila <awogbemila@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet: sched: fix spelling mistake in Kconfig "trys" -> "tries"
Colin Ian King [Mon, 7 Dec 2020 16:12:31 +0000 (16:12 +0000)]
net: sched: fix spelling mistake in Kconfig "trys" -> "tries"

There is a spelling mistake in the Kconfig help text. Fix it.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'for-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetoot...
David S. Miller [Tue, 8 Dec 2020 23:58:49 +0000 (15:58 -0800)]
Merge branch 'for-upstream' of git://git./linux/kernel/git/bluetooth/bluetooth-next

Johan Hedberg says:

====================
pull request: bluetooth-next 2020-12-07

Here's the main bluetooth-next pull request for the 5.11 kernel.

 - Updated Bluetooth entries in MAINTAINERS to include Luiz von Dentz
 - Added support for Realtek 8822CE and 8852A devices
 - Added support for MediaTek MT7615E device
 - Improved workarounds for fake CSR devices
 - Fix Bluetooth qualification test case L2CAP/COS/CFD/BV-14-C
 - Fixes for LL Privacy support
 - Enforce 16 byte encryption key size for FIPS security level
 - Added new mgmt commands for extended advertising support
 - Multiple other smaller fixes & improvements

Please let me know if there are any issues pulling. Thanks.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agonet/af_iucv: use DECLARE_SOCKADDR to cast from sockaddr
Julian Wiedmann [Mon, 7 Dec 2020 12:53:07 +0000 (13:53 +0100)]
net/af_iucv: use DECLARE_SOCKADDR to cast from sockaddr

This gets us compile-time size checking.

Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'mlxsw-Add-support-for-Q-in-VNI'
David S. Miller [Tue, 8 Dec 2020 23:45:57 +0000 (15:45 -0800)]
Merge branch 'mlxsw-Add-support-for-Q-in-VNI'

Ido Schimmel says:

====================
mlxsw: Add support for Q-in-VNI

This patch set adds support for Q-in-VNI over Spectrum-{2,3} ASICs.
Q-in-VNI is like regular VxLAN encapsulation with the sole difference
that overlay packets can contain a VLAN tag. In Linux, this is achieved
by adding the VxLAN device to a 802.1ad bridge instead of a 802.1q
bridge.

From mlxsw perspective, Q-in-VNI support entails two main changes:

1. An outer VLAN tag should always be pushed to the overlay packet
during decapsulation

2. The EtherType used during decapsulation should be 802.1ad (0x88a8)
instead of the default 802.1q (0x8100)

Patch set overview:

Patches #1-#3 add required device registers and fields

Patch #4 performs small refactoring to allow code re-use

Patches #5-#7 make the EtherType used during decapsulation a property of
the tunnel port (i.e., VxLAN). This leads to the driver vetoing
configurations in which VxLAN devices are member in both 802.1ad and
802.1q/802.1d bridges. Will be handled in the future by determining the
overlay EtherType on the egress port instead

Patch #8 adds support for Q-in-VNI for Spectrum-2 and newer ASICs

Patches #9-#10 veto Q-in-VNI for Spectrum-1 ASICs due to some hardware
limitations. Can be worked around, but decided not to support it for now

Patch #11 adjusts mlxsw to stop vetoing addition of VXLAN devices to
802.1ad bridges

Patch #12 adds a generic forwarding test that can be used with both veth
pairs and physical ports with a loopback

Patch #13 adds a test to make sure mlxsw vetoes unsupported Q-in-VNI
configurations
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoselftests: mlxsw: Add Q-in-VNI veto tests
Amit Cohen [Tue, 8 Dec 2020 09:22:53 +0000 (11:22 +0200)]
selftests: mlxsw: Add Q-in-VNI veto tests

Add tests to ensure that the forbidden and unsupported cases are indeed
vetoed by mlxsw driver.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoselftests: forwarding: Add Q-in-VNI test
Petr Machata [Tue, 8 Dec 2020 09:22:52 +0000 (11:22 +0200)]
selftests: forwarding: Add Q-in-VNI test

Add test to check Q-in-VNI traffic.

Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_switchdev: Allow joining VxLAN to 802.1ad bridge
Amit Cohen [Tue, 8 Dec 2020 09:22:51 +0000 (11:22 +0200)]
mlxsw: spectrum_switchdev: Allow joining VxLAN to 802.1ad bridge

The previous patches added support for VxLAN device enslaved to 802.1ad
bridge in Spectrum-2 ASIC and vetoed it in Spectrum-1.

Do not veto VxLAN with 802.1ad bridge.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: Veto Q-in-VNI for Spectrum-1 ASIC
Amit Cohen [Tue, 8 Dec 2020 09:22:50 +0000 (11:22 +0200)]
mlxsw: Veto Q-in-VNI for Spectrum-1 ASIC

Implementation of Q-in-VNI is different between ASIC types, this set adds
support only for Spectrum-2.

Return an error when trying to create VxLAN device and enslave it to
802.1ad bridge in Spectrum-1.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_switchdev: Use ops->vxlan_join() when adding VLAN to VxLAN device
Amit Cohen [Tue, 8 Dec 2020 09:22:49 +0000 (11:22 +0200)]
mlxsw: spectrum_switchdev: Use ops->vxlan_join() when adding VLAN to VxLAN device

Currently mlxsw_sp_switchdev_vxlan_vlan_add() always calls
mlxsw_sp_bridge_8021q_vxlan_join() because VLANs were only ever added to
a VLAN-filtering bridge, which is only 802.1q bridge.

This set adds support for VxLAN with 802.1ad bridge, so VLAN-filtering
bridge is not only 802.1q.

Call ops->vxlan_join(), so mlxsw_sp_bridge_802{1q, 1ad}_vxlan_join()
will be called according to bridge type.

This is needed to ensure that VxLAN with 802.1ad bridge will be vetoed
in Spectrum-1 with the next patch.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_nve_vxlan: Add support for Q-in-VNI for Spectrum-2 ASIC
Amit Cohen [Tue, 8 Dec 2020 09:22:48 +0000 (11:22 +0200)]
mlxsw: spectrum_nve_vxlan: Add support for Q-in-VNI for Spectrum-2 ASIC

On Spectrum-2, the default setting is not to push VLAN to the decapsulated
packet. This is controlled by SPVTR.ipvid_mode.
Set SPVTR.ipvid_mode to always push VLAN.
Without this setting, Spectrum-2 overtakes the VLAN tag of decapsulated
packet for bridging.

In addition, set SPVID register to use EtherType saved in
mlxsw_sp_nve_config when VLAN is pushed for the NVE tunnel.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum: Publish mlxsw_sp_ethtype_to_sver_type()
Amit Cohen [Tue, 8 Dec 2020 09:22:47 +0000 (11:22 +0200)]
mlxsw: spectrum: Publish mlxsw_sp_ethtype_to_sver_type()

Declare mlxsw_sp_ethtype_to_sver_type() in spectrum.h to enable using it
in other files.

It will be used in the next patch to map between EtherType and the
relevant value configured by SVER register.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: Save EtherType as part of mlxsw_sp_nve_config
Amit Cohen [Tue, 8 Dec 2020 09:22:46 +0000 (11:22 +0200)]
mlxsw: Save EtherType as part of mlxsw_sp_nve_config

Add EtherType field to mlxsw_sp_nve_config struct.
Set EtherType according to mlxsw_sp_nve_params.ethertype.

Pass 'mlxsw_sp_nve_params' instead of 'mlxsw_sp_nve_params->dev' to the
function which initializes mlxsw_sp_nve_config struct to know which
EtherType to use.

This field is needed to configure which EtherType will be used when
VLAN is pushed at ingress of the tunnel port.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: Save EtherType as part of mlxsw_sp_nve_params
Amit Cohen [Tue, 8 Dec 2020 09:22:45 +0000 (11:22 +0200)]
mlxsw: Save EtherType as part of mlxsw_sp_nve_params

Add EtherType field to mlxsw_sp_nve_params struct.
Set it when VxLAN device is added to bridge device.

This field is needed to configure which EtherType will be used when
VLAN is pushed at ingress of the tunnel port.

Use ETH_P_8021Q for tunnel port enslaved to 802.1d and 802.1q bridges.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: spectrum_switchdev: Create common function for joining VxLAN to VLAN-aware...
Amit Cohen [Tue, 8 Dec 2020 09:22:44 +0000 (11:22 +0200)]
mlxsw: spectrum_switchdev: Create common function for joining VxLAN to VLAN-aware bridge

The code in mlxsw_sp_bridge_8021q_vxlan_join() can be used also for
802.1ad bridge.

Move the code to function called mlxsw_sp_bridge_vlan_aware_vxlan_join()
and call it from mlxsw_sp_bridge_8021q_vxlan_join() to enable code
reuse.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: reg: Add support for tunnel port in SPVID register
Amit Cohen [Tue, 8 Dec 2020 09:22:43 +0000 (11:22 +0200)]
mlxsw: reg: Add support for tunnel port in SPVID register

Add spvid_tport field which indicates if the port is tunnel port.
When spvid_tport is true, local_port field supposed to be tunnel port
type.

It will be used to configure which Ethertype will be used when VLAN is
pushed at ingress for tunnel port.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: reg: Add Switch Port VLAN Stacking Register
Amit Cohen [Tue, 8 Dec 2020 09:22:42 +0000 (11:22 +0200)]
mlxsw: reg: Add Switch Port VLAN Stacking Register

SPVTR register configures the VLAN mode of the port to enable VLAN
stacking.

It will be used to configure VxLAN to push VLAN to the decapsulated packet.
Without this setting, Spectrum-2 overtakes the VLAN tag of decapsulated
packet for bridging.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agomlxsw: Use one enum for all registers that contain tunnel_port field
Amit Cohen [Tue, 8 Dec 2020 09:22:41 +0000 (11:22 +0200)]
mlxsw: Use one enum for all registers that contain tunnel_port field

Currently SFN, TNUMT and TNPC registers use separate enums for
tunnel_port.

Create one enum with a neutral name and use it.
Remove the enums that are not currently required.

The next patches add two more registers that contain tunnel_port field,
the new enum can be used for them also.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge tag 'mlx5-updates-2020-12-01' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Tue, 8 Dec 2020 23:39:01 +0000 (15:39 -0800)]
Merge tag 'mlx5-updates-2020-12-01' of git://git./linux/kernel/git/saeed/linux

mlx5-updates-2020-12-01

mlx5e port TX timestamping support and MISC updates

1) Add support for port TX timestamping, for better PTP accuracy.

Currently in mlx5 HW TX timestamping is done on CQE (TX completion)
generation, which much earlier than when the packet actually goes out to
the wire, in this series Eran implements the option to do timestamping on
the port using a special SQ (Send Queue), such Send Queue will generate 2
CQEs (TX completions), the original one and a new one when the packet
leaves the port, due to the nature of this special handling, such mechanism
is an opt-in only and it is off by default to avoid any performance
degradation on normal traffic flows.

This patchset improves TX Hardware timestamping offset to be less than
40ns at a 100Gbps line rate, compared to 600ns before.

With that, making our HW compliant with G.8273.2 class C, and allow Linux
systems to be deployed in the 5G telco edge, where this standard is a must.

2) Misc updates and trivial improvements.

Signed-off-by: David S. Miller <davem@davemloft.net>
3 years agoMerge branch 'net-hns3-updates-for-next'
Jakub Kicinski [Tue, 8 Dec 2020 19:58:55 +0000 (11:58 -0800)]
Merge branch 'net-hns3-updates-for-next'

Huazhong Tan says:

====================
net: hns3: updates for -next

There are some updates for the HNS3 ethernet driver.

promiscuous configuration more flexible, #2 adds ethtool
private flags to control whether enable tx unicast promisc.

previous version:
https://patchwork.kernel.org/project/netdevbpf/cover/1606997936-22166-1-git-send-email-tanhuazhong@huawei.com/
====================

Link: https://lore.kernel.org/r/1607227575-56689-1-git-send-email-tanhuazhong@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: hns3: refine the VLAN tag handle for port based VLAN
Guojia Liao [Sun, 6 Dec 2020 04:06:15 +0000 (12:06 +0800)]
net: hns3: refine the VLAN tag handle for port based VLAN

For DEVICE_VERSION_V2, the hardware only supports max two layer
VLAN tags, including port based tag inserted by hardware, tag in
tx buffer descriptor(get from skb->tci) and tag in packet.

For transmit packet:
If port based VLAN disabled, and vf driver gets a VLAN tag from
skb, the VLAN tag must be filled to the Outer_VLAN_TAG field
(tag near to DMAC) of tx buffer descriptor, otherwise it may
be inserted after the tag in packet.

If port based VLAN enabled, and vf driver gets a VLAN tag from
skb, the VLAN tag must be filled to the VLAN_TAG field (tag
far to DMAC) of tx buffer descriptor, otherwise it may be
conflicted with port based VLAN, and raise a hardware error.

For receive packet:
The hardware will strip the VLAN tags and fill them in the rx
buffer descriptor, no matter port based VLAN enable or not.
Because port based VLAN tag is useless for stack, so vf driver
needs to discard the port based VLAN tag get from rx buffer
descriptor when port based VLAN enabled.

So vf must know about the port based VLAN state.

For DEVICE_VERSION_V3, the hardware provides some new
configuration to improve it.

For transmit packet:
When enable tag shift mode, hardware will handle the VLAN tag
in outer_VLAN_TAG field as VLAN_TAG, so it won't conflict with
port based VLAN. And hardware also make sure the tag before
the tag in packet. So vf driver doesn't need to specify the tag
position according to the port based VLAN state anymore.

For receive packet:
When enable discard mode, hardware will strip and discard the
port based VLAN tag, so vf driver doesn't need to identify it
from rx buffer descriptor.

So modify the port based VLAN configuration, simplify the process
for vf handling the VLAN tag.

Signed-off-by: Guojia Liao <liaoguojia@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: hns3: add priv flags support to switch limit promisc mode
Jian Shen [Sun, 6 Dec 2020 04:06:14 +0000 (12:06 +0800)]
net: hns3: add priv flags support to switch limit promisc mode

Currently, the tx unicast promisc is always enabled when promisc
mode on. If tx unicast promisc on, a function will receive all
unicast packet from other functions belong to the same port.
Add a ethtool private flag to control whether enable tx
unicast promisc. Then the function is able to filter the
unknown unicast packets from other function.

Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: hns3: add support for extended promiscuous command
Guojia Liao [Sun, 6 Dec 2020 04:06:13 +0000 (12:06 +0800)]
net: hns3: add support for extended promiscuous command

For DEVICE_VERSION_V2, the hardware supports enable tx and rx
promiscuous separately. But tx or rx promiscuous is active for
unicast, multicast and broadcast promiscuous simultaneously.
To support traffics between functions belong to the same port,
we always enable tx promiscuous for broadcast promiscuous, so
tx promiscuous for unicast and multicast promiscuous is also
enabled.

For DEVICE_VERSION_V3, the hardware decouples the above
relationship. Tx unicast promiscuous, rx unicast promiscuous,
tx multicast promiscuous, rx multicast promiscuous, tx broadcast
promiscuous and rx broadcast promiscuous can be enabled separately.

So add support for the new promiscuous command.

Signed-off-by: Guojia Liao <liaoguojia@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5e: Fill mlx5e_create_cq_param in a function
Maxim Mikityanskiy [Tue, 1 Dec 2020 22:42:08 +0000 (14:42 -0800)]
net/mlx5e: Fill mlx5e_create_cq_param in a function

Create a function to fill the fields of struct mlx5e_create_cq_param
based on a channel. The purpose is code reuse between normal CQs, XSK
CQs and the upcoming QoS CQs.

Signed-off-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Split between RX/TX tunnel FW support indication
Aya Levin [Tue, 1 Dec 2020 22:42:07 +0000 (14:42 -0800)]
net/mlx5e: Split between RX/TX tunnel FW support indication

Use the new FW caps to advertise for ip-in-ip tunnel support separately
for RX and TX.

Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: Fix passing zero to 'PTR_ERR'
YueHaibing [Tue, 1 Dec 2020 22:42:06 +0000 (14:42 -0800)]
net/mlx5: Fix passing zero to 'PTR_ERR'

Fix smatch warnings:

drivers/net/ethernet/mellanox/mlx5/core/esw/acl/egress_lgcy.c:105 esw_acl_egress_lgcy_setup() warn: passing zero to 'PTR_ERR'
drivers/net/ethernet/mellanox/mlx5/core/esw/acl/egress_ofld.c:177 esw_acl_egress_ofld_setup() warn: passing zero to 'PTR_ERR'
drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_lgcy.c:184 esw_acl_ingress_lgcy_setup() warn: passing zero to 'PTR_ERR'
drivers/net/ethernet/mellanox/mlx5/core/esw/acl/ingress_ofld.c:262 esw_acl_ingress_ofld_setup() warn: passing zero to 'PTR_ERR'

esw_acl_table_create() never returns NULL, so
NULL test should be removed.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5: Arm only EQs with EQEs
Shay Drory [Tue, 1 Dec 2020 22:42:05 +0000 (14:42 -0800)]
net/mlx5: Arm only EQs with EQEs

Currently, when more than one EQ is sharing an IRQ, and this IRQ is
being interrupted, all the EQs sharing the IRQ will be armed. This is
done regardless of whether an EQ has EQE.
When multiple EQs are sharing an IRQ, one or more EQs can have valid
EQEs.

Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Parav Pandit <parav@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Remove duplicated include
YueHaibing [Tue, 1 Dec 2020 22:42:04 +0000 (14:42 -0800)]
net/mlx5e: Remove duplicated include

Remove duplicated include.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: remove unnecessary memset
Zhu Yanjun [Tue, 1 Dec 2020 22:42:03 +0000 (14:42 -0800)]
net/mlx5e: remove unnecessary memset

Since kvzalloc will initialize the allocated memory, it is not
necessary to initialize it once again.

Fixes: 11b717d61526 ("net/mlx5: E-Switch, Get reg_c0 value on CQE")
Signed-off-by: Zhu Yanjun <yanjunz@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Add TX port timestamp support
Eran Ben Elisha [Tue, 1 Dec 2020 22:42:02 +0000 (14:42 -0800)]
net/mlx5e: Add TX port timestamp support

Transmitted packet timestamping accuracy can be improved when using
timestamp from the port, instead of packet CQE creation timestamp, as
it better reflects the actual time of a packet's transmit.

TX port timestamping is supported starting from ConnectX6-DX hardware.
Although at the original completion, only CQE timestamp can be attached,
we are able to get TX port timestamping via an additional completion over
a special CQ associated with the SQ (in addition to the regular CQ).

Driver to ignore the original packet completion timestamp, and report
back the timestamp of the special CQ completion. If the absolute timestamp

diff between the two completions is greater than 1 / 128 second, ignore
the TX port timestamp as it has a jitter which is too big.
No skb will be generate out of the extra completion.

Allocate additional CQ per ptpsq, to receive the TX port timestamp.

Driver to hold an skb FIFO in order to map between transmitted skb to
the two expected completions. When using ptpsq, hold double refcount on
the skb, to gaurantee it will not get released before both completions
arrive.

Expose dedicated counters of the ptp additional CQ and connect it to the
TX health reporter.

This patch improves TX Hardware timestamping offset to be less than 40ns
at a 100Gbps line rate, compared to 600ns before.

With that, making our HW compliant with G.8273.2 class C, and allow Linux
systems to be deployed in the 5G telco edge, where this standard is a
must.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Add TX PTP port object support
Eran Ben Elisha [Tue, 1 Dec 2020 22:42:01 +0000 (14:42 -0800)]
net/mlx5e: Add TX PTP port object support

Add TX PTP port object support for better TX timestamping accuracy.
Currently, driver supports CQE based TX port timestamp. Device
also offers TX port timestamp, which has less jitter and better
reflects the actual time of a packet's transmit.

Define new driver layout called ptpsq, on which driver will create
SQs that will support TX port timestamp for their transmitted packets.
Driver to identify PTP TX skbs and steer them to these dedicated SQs
as part of the select queue ndo.

Driver to hold ptpsq per TC and report them at
netif_set_real_num_tx_queues().

Add support for all needed functionality in order to xmit and poll
completions received via ptpsq.

Add ptpsq to the TX reporter recover, diagnose and dump methods.

Creation of ptpsqs is disabled by default, and can be enabled via
tx_port_ts private flag.

This patch steer all timestamp related packets to a ptpsq, but it
does not open the port timestamp support for it. The support will
be added in the following patch.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Move MLX5E_RX_ERR_CQE macro
Eran Ben Elisha [Tue, 1 Dec 2020 22:42:00 +0000 (14:42 -0800)]
net/mlx5e: Move MLX5E_RX_ERR_CQE macro

MLX5E_RX_ERR_CQE Macro is used only in data-path, move it to the
appropriate header file.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Split SW group counters update function
Eran Ben Elisha [Tue, 1 Dec 2020 22:41:59 +0000 (14:41 -0800)]
net/mlx5e: Split SW group counters update function

SW group counter update function aggregates sw stats out of many
mlx5e_*_stats resides in a given mlx5e_channel_stats struct.
Split the function into a few helper functions.

This will be used later in the series to calculate specific
mlx5e_*_stats which are not defined inside mlx5e_channel_stats.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Change skb fifo push/pop API to be used without SQ
Eran Ben Elisha [Tue, 1 Dec 2020 22:41:58 +0000 (14:41 -0800)]
net/mlx5e: Change skb fifo push/pop API to be used without SQ

The skb fifo push/pop API used pre-defined attributes within the
mlx5e_txqsq.
In order to share the skb fifo API with other non-SQ use cases,
change the API input to get newly defined mlx5e_skb_fifo struct.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Allow SQ outside of channel context
Eran Ben Elisha [Tue, 1 Dec 2020 22:41:57 +0000 (14:41 -0800)]
net/mlx5e: Allow SQ outside of channel context

In order to be able to create an SQ outside of a channel context, remove
sq->channel direct pointer. This requires adding a direct pointer to:
netdevice, priv and mlx5_core in order to support SQs that are part of
mlx5e_channel. Use channel_stats from the corresponding CQ.

Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Allow RQ outside of channel context
Aya Levin [Tue, 1 Dec 2020 22:41:56 +0000 (14:41 -0800)]
net/mlx5e: Allow RQ outside of channel context

In order to be able to create an RQ outside of a channel context, remove
rq->channel direct pointer. This requires adding a direct pointer to:
ICOSQ and priv in order to support RQs that are part of mlx5e_channel.
Use channel_stats from the corresponding CQ.

Signed-off-by: Aya Levin <ayal@nvidia.com>
Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Allow CQ outside of channel context
Aya Levin [Tue, 1 Dec 2020 22:41:55 +0000 (14:41 -0800)]
net/mlx5e: Allow CQ outside of channel context

In order to be able to create a CQ outside of a channel context, remove
cq->channel direct pointer. This requires adding a direct pointer to
channel statistics, netdevice, priv and to mlx5_core in order to support
CQs that are a part of mlx5e_channel.
In addition, parameters the were previously derived from the channel
like napi, NUMA node, channel stats and index are now assembled in
struct mlx5e_create_cq_param which is given to mlx5e_open_cq() instead
of channel pointer. Generalizing mlx5e_open_cq() allows opening CQ
outside of channel context which will be used in following patches in
the patch-set.

Signed-off-by: Aya Levin <ayal@nvidia.com>
Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet/mlx5e: Free drop RQ in a dedicated function
Tariq Toukan [Tue, 1 Dec 2020 22:41:54 +0000 (14:41 -0800)]
net/mlx5e: Free drop RQ in a dedicated function

The drop RQ has very limited objects to be freed, and differs
from regular RQs in the context that it is freed from.
Add a dedicated function for it, use it where needed, and remove
the drop_rq-specific checks in the generic function.

Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Maxim Mikityanskiy <maximmi@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agonet: dsa: print the MTU value that could not be set
Rasmus Villemoes [Sat, 5 Dec 2020 13:39:44 +0000 (14:39 +0100)]
net: dsa: print the MTU value that could not be set

These warnings become somewhat more informative when they include the
MTU value that could not be set and not just the errno.

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Link: https://lore.kernel.org/r/20201205133944.10182-1-rasmus.villemoes@prevas.dk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox...
Jakub Kicinski [Tue, 8 Dec 2020 02:36:21 +0000 (18:36 -0800)]
Merge branch 'mlx5-next' of git://git./linux/kernel/git/mellanox/linux

Saeed Mahameed says:

====================
mlx5-next auxbus support

This pull request is targeting net-next and rdma-next branches.

This series provides mlx5 support for auxiliary bus devices.

It starts with a merge commit of tag 'auxbus-5.11-rc1' from
gregkh/driver-core into mlx5-next, then the mlx5 patches that will convert
mlx5 ulp devices (netdev, rdma, vdpa) to use the proper auxbus
infrastructure instead of the internal mlx5 device and interface management
implementation, which Leon is deleting at the end of this patchset.

Link: https://lore.kernel.org/alsa-devel/20201026111849.1035786-1-leon@kernel.org/
Thanks to everyone for the joint effort !

* 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
  RDMA/mlx5: Remove IB representors dead code
  net/mlx5: Simplify eswitch mode check
  net/mlx5: Delete custom device management logic
  RDMA/mlx5: Convert mlx5_ib to use auxiliary bus
  net/mlx5e: Connect ethernet part to auxiliary bus
  vdpa/mlx5: Connect mlx5_vdpa to auxiliary bus
  net/mlx5: Register mlx5 devices to auxiliary virtual bus
  vdpa/mlx5: Make hardware definitions visible to all mlx5 devices
  net/mlx5_core: Clean driver version and name
  net/mlx5: Properly convey driver version to firmware
  driver core: auxiliary bus: minor coding style tweaks
  driver core: auxiliary bus: make remove function return void
  driver core: auxiliary bus: move slab.h from include file
  Add auxiliary bus support
====================

Link: https://lore.kernel.org/r/20201207053349.402772-1-saeed@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx4: Remove unused #define MAX_MSIX_P_PORT
Tariq Toukan [Sun, 6 Dec 2020 09:12:54 +0000 (11:12 +0200)]
net/mlx4: Remove unused #define MAX_MSIX_P_PORT

All usages of the definition MAX_MSIX_P_PORT were removed.
It's not in use anymore. Remove it.

Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@mellanox.com>
Link: https://lore.kernel.org/r/20201206091254.12476-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobonding: set xfrm feature flags more sanely
Jarod Wilson [Sat, 5 Dec 2020 17:40:03 +0000 (12:40 -0500)]
bonding: set xfrm feature flags more sanely

We can remove one of the ifdef blocks here, and instead of setting both
the xfrm hw_features and features flags, then unsetting the feature
flags if not in AB, wait to set the features flags if we're actually in AB
mode.

Signed-off-by: Jarod Wilson <jarod@redhat.com>
Link: https://lore.kernel.org/r/20201205174003.578267-1-jarod@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoibmvnic: add some debugs
Sukadev Bhattiprolu [Sat, 5 Dec 2020 02:22:35 +0000 (18:22 -0800)]
ibmvnic: add some debugs

We sometimes run into situations where a soft/hard reset of the adapter
takes a long time or fails to complete. Having additional messages that
include important adapter state info will hopefully help understand what
is happening, reduce the guess work and minimize requests to reproduce
problems with debug patches.

Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Link: https://lore.kernel.org/r/20201205022235.2414110-1-sukadev@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMAINTAINERS: Update Bluetooth entries
Marcel Holtmann [Mon, 7 Dec 2020 10:08:42 +0000 (11:08 +0100)]
MAINTAINERS: Update Bluetooth entries

Update the status to Supported to reflect the current state of affairs
and add Luiz as additional maintainer.

Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Increment management interface revision
Marcel Holtmann [Mon, 7 Dec 2020 10:02:09 +0000 (11:02 +0100)]
Bluetooth: Increment management interface revision

Increment the mgmt revision due to the recently added new commands.

Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Add workaround for remote-wakeup issues with Barrot 8041a02 fake...
Hans de Goede [Sat, 5 Dec 2020 15:02:01 +0000 (16:02 +0100)]
Bluetooth: btusb: Add workaround for remote-wakeup issues with Barrot 8041a02 fake CSR controllers

With the recent btusb change to detect and deal with more fake CSR
controllers, I decided to see if fake CSR controllers with Barrot
8041a02 chips would now work.

After much experimentation I came to the conclusion that it works, if I
have autosuspend enabled initially and then disable it after the device
has suspended at least once. Yes this is very weird, but I've tried many
things, like manually clearing the remote-wakeup feature. Doing a
runtime-resume + runtime suspend is the only way to get the receiver
to actually report received data (and/or pairing info) through its
bulk rx endpoint.

But the funkyness of the bulk-endpoint does not stop there, I mainly
found out about this problem, because with autosuspend enabled
(which usually ensures the suspend at least once condition is met),
the receiver stops reporting received data through its bulk rx endpoint
as soon as autosuspend kicks in. So I initially just disabled
autosuspend, but then the receiver does not work at all.

This was with a fake CSR receiver with a Barrot 8041a02 chip with a
bcdDevice value of 0x8891, a lmp_subver of 0x1012, a hci_rev of 0x0810
and a hci_ver of BLUETOOTH_VER_4_0.

Summarizing this specific fake CSR receiver has the following 2 issues:

1. The bulk rx endpoint will never report any data unless
the device was suspended at least once.

2. They will not wakeup when autosuspended and receiving data on their
bulk rx endpoint from e.g. a keyboard or mouse (IOW remote-wakeup support
is broken for the bulk endpoint).

Add a workaround for 1. which enables runtime-suspend, force-suspends
the hci and then wakes-it up by disabling runtime-suspend again.

Add a workaround for 2. which clears the hci's can_wake flag, this way
the hci will still be autosuspended when it is not open.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Fix detection of some fake CSR controllers with a bcdDevice val...
Hans de Goede [Sat, 5 Dec 2020 15:02:00 +0000 (16:02 +0100)]
Bluetooth: btusb: Fix detection of some fake CSR controllers with a bcdDevice val of 0x0134

Commit cde1a8a99287 ("Bluetooth: btusb: Fix and detect most of the
Chinese Bluetooth controllers") made the detection of fake controllers
more generic fixing it for much of the newer fakes / clones.

But this does not work for a fake CSR controller with a bcdDevice
value of 0x0134, which was correctly identified as fake before
this change.

Add an extra check for this special case, checking for a combination
of a bcdDevice value of 0x0134, together with a lmp_subver of 0x0c5c
and a hci_ver of BLUETOOTH_VER_2_0.

The chip inside this fake dongle is marked as with "clockwise cw6629d".

Fixes: cde1a8a99287 ("Bluetooth: btusb: Fix and detect most of the Chinese Bluetooth controllers")
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Set missing suspend task bits
Abhishek Pandit-Subedi [Fri, 4 Dec 2020 03:14:31 +0000 (11:14 +0800)]
Bluetooth: Set missing suspend task bits

When suspending, mark SUSPEND_SCAN_ENABLE and SUSPEND_SCAN_DISABLE tasks
correctly when either classic or le scanning is modified.

Signed-off-by: Abhishek Pandit-Subedi <abhishekpandit@chromium.org>
Signed-off-by: Howard Chung <howardchung@google.com>
Reviewed-by: Alain Michaud <alainm@chromium.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Change MGMT security info CMD to be more generic
Daniel Winkler [Thu, 3 Dec 2020 20:12:52 +0000 (12:12 -0800)]
Bluetooth: Change MGMT security info CMD to be more generic

For advertising, we wish to know the LE tx power capabilities of the
controller in userspace, so this patch edits the Security Info MGMT
command to be more generic, such that other various controller
capabilities can be included in the EIR data. This change also includes
the LE min and max tx power into this newly-named command.

The change was tested by manually verifying that the MGMT command
returns the tx power range as expected in userspace.

Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org>
Signed-off-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Query LE tx power on startup
Daniel Winkler [Thu, 3 Dec 2020 20:12:51 +0000 (12:12 -0800)]
Bluetooth: Query LE tx power on startup

Queries tx power via HCI_LE_Read_Transmit_Power command when the hci
device is initialized, and stores resulting min/max LE power in hdev
struct. If command isn't available (< BT5 support), min/max values
both default to HCI_TX_POWER_INVALID.

This patch is manually verified by ensuring BT5 devices correctly query
and receive controller tx power range.

Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org>
Signed-off-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Use intervals and tx power from mgmt cmds
Daniel Winkler [Thu, 3 Dec 2020 20:12:50 +0000 (12:12 -0800)]
Bluetooth: Use intervals and tx power from mgmt cmds

This patch takes the min/max intervals and tx power optionally provided
in mgmt interface, stores them in the advertisement struct, and uses
them when configuring the hci requests. While tx power is not used if
extended advertising is unavailable, software rotation will use the min
and max advertising intervals specified by the client.

This change is validated manually by ensuring the min/max intervals are
propagated to the controller on both hatch (extended advertising) and
kukui (no extended advertising) chromebooks, and that tx power is
propagated correctly on hatch. These tests are performed with multiple
advertisements simultaneously.

Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org>
Signed-off-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Break add adv into two mgmt commands
Daniel Winkler [Thu, 3 Dec 2020 20:12:49 +0000 (12:12 -0800)]
Bluetooth: Break add adv into two mgmt commands

This patch adds support for the new advertising add interface, with the
first command setting advertising parameters and the second to set
advertising data. The set parameters command allows the caller to leave
some fields "unset", with a params bitfield defining which params were
purposefully set. Unset parameters will be given defaults when calling
hci_add_adv_instance. The data passed to the param mgmt command is
allowed to be flexible, so in the future if bluetoothd passes a larger
structure with new params, the mgmt command will ignore the unknown
members at the end.

This change has been validated on both hatch (extended advertising) and
kukui (no extended advertising) chromebooks running bluetoothd that
support this new interface. I ran the following manual tests:
- Set several (3) advertisements using modified test_advertisement.py
- For each, validate correct data and parameters in btmon trace
- Verified both for software rotation and extended adv

Automatic test suite also run, testing many (25) scenarios of single and
multi-advertising for data/parameter correctness.

Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org>
Signed-off-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Add helper to set adv data
Daniel Winkler [Thu, 3 Dec 2020 20:12:48 +0000 (12:12 -0800)]
Bluetooth: Add helper to set adv data

We wish to handle advertising data separately from advertising
parameters in our new MGMT requests. This change adds a helper that
allows the advertising data and scan response to be updated for an
existing advertising instance.

Reviewed-by: Sonny Sasaka <sonnysasaka@chromium.org>
Signed-off-by: Daniel Winkler <danielwinkler@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: support download nvm with different board id for wcn6855
Tim Jiang [Thu, 3 Dec 2020 04:57:14 +0000 (12:57 +0800)]
Bluetooth: btusb: support download nvm with different board id for wcn6855

we define many nvm files for wcn6855 btsoc and host driver
should find the correct nvm file based on board ID and then
download it.

Signed-off-by: Tim Jiang <tjiang@codeaurora.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Map Typhoon peak controller to BTUSB_INTEL_NEWGEN
Kiran K [Thu, 19 Nov 2020 11:39:43 +0000 (17:09 +0530)]
Bluetooth: btusb: Map Typhoon peak controller to BTUSB_INTEL_NEWGEN

Map Typhoon peak Intel controller to BTUSB_INTEL_NEWGEN

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Helper function to download firmware to Intel adapters
Kiran K [Thu, 19 Nov 2020 11:39:42 +0000 (17:09 +0530)]
Bluetooth: btusb: Helper function to download firmware to Intel adapters

Define a helper function to download firmware for new generation Intel
controllers

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Define a function to construct firmware filename
Kiran K [Thu, 19 Nov 2020 11:39:41 +0000 (17:09 +0530)]
Bluetooth: btusb: Define a function to construct firmware filename

Define a new function to construct firmware/ddc filename for new
generation Intel controllers

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Add *setup* function for new generation Intel controllers
Kiran K [Thu, 19 Nov 2020 11:39:40 +0000 (17:09 +0530)]
Bluetooth: btusb: Add *setup* function for new generation Intel controllers

Define a new  *setup* function for new generation Intel controllers

Signed-off-by: Kiran K <kiran.k@intel.com>
Signed-off-by: Amit K Bag <amit.k.bag@intel.com>
Signed-off-by: Raghuram Hegde <raghuram.hegde@intel.com>
Reviewed-by: Sathish Narasimman <Sathish.Narasimman@intel.com>
Reviewed-by: Chethan T N <chethan.tumkur.narayan@intel.com>
Reviewed-by: Srivatsa Ravishankar <ravishankar.srivatsa@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btintel: Fix endianness issue for TLV version information
Kiran K [Thu, 19 Nov 2020 11:39:39 +0000 (17:09 +0530)]
Bluetooth: btintel: Fix endianness issue for TLV version information

do __le32_to_cpu to tlv data fields to make sure driver runs
correctly when running on big endian system.

Signed-off-by: Kiran K <kiran.k@intel.com>
Reviewed-by: Chethan T N <chethan.tumkur.narayan@intel.com>
Reviewed-by: Srivatsa Ravishankar <ravishankar.srivatsa@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Add toggle to switch off interleave scan
Howard Chung [Thu, 26 Nov 2020 04:22:25 +0000 (12:22 +0800)]
Bluetooth: Add toggle to switch off interleave scan

This patch add a configurable parameter to switch off the interleave
scan feature.

Signed-off-by: Howard Chung <howardchung@google.com>
Reviewed-by: Alain Michaud <alainm@chromium.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Refactor read default sys config for various types
Howard Chung [Thu, 26 Nov 2020 04:22:24 +0000 (12:22 +0800)]
Bluetooth: Refactor read default sys config for various types

Refactor read default system configuration function so that it's capable
of returning different types than u16

Signed-off-by: Howard Chung <howardchung@google.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Handle active scan case
Howard Chung [Thu, 26 Nov 2020 04:22:23 +0000 (12:22 +0800)]
Bluetooth: Handle active scan case

This patch adds code to handle the active scan during interleave
scan. The interleave scan will be canceled when users start active scan,
and it will be restarted after active scan stopped.

Signed-off-by: Howard Chung <howardchung@google.com>
Reviewed-by: Alain Michaud <alainm@chromium.org>
Reviewed-by: Manish Mandlik <mmandlik@chromium.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Handle system suspend resume case
Howard Chung [Thu, 26 Nov 2020 04:22:22 +0000 (12:22 +0800)]
Bluetooth: Handle system suspend resume case

This patch adds code to handle the system suspension during interleave
scan. The interleave scan will be canceled when the system is going to
sleep, and will be restarted after waking up.

Signed-off-by: Howard Chung <howardchung@google.com>
Reviewed-by: Alain Michaud <alainm@chromium.org>
Reviewed-by: Manish Mandlik <mmandlik@chromium.org>
Reviewed-by: Abhishek Pandit-Subedi <abhishekpandit@chromium.org>
Reviewed-by: Miao-chen Chou <mcchou@chromium.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Interleave with allowlist scan
Howard Chung [Thu, 26 Nov 2020 04:22:21 +0000 (12:22 +0800)]
Bluetooth: Interleave with allowlist scan

This patch implements the interleaving between allowlist scan and
no-filter scan. It'll be used to save power when at least one monitor is
registered and at least one pending connection or one device to be
scanned for.

The durations of the allowlist scan and the no-filter scan are
controlled by MGMT command: Set Default System Configuration. The
default values are set randomly for now.

Signed-off-by: Howard Chung <howardchung@google.com>
Reviewed-by: Alain Michaud <alainm@chromium.org>
Reviewed-by: Manish Mandlik <mmandlik@chromium.org>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Fix attempting to set RPA timeout when unsupported
Edward Vear [Tue, 27 Oct 2020 07:02:03 +0000 (00:02 -0700)]
Bluetooth: Fix attempting to set RPA timeout when unsupported

During controller initialization, an LE Set RPA Timeout command is sent
to the controller if supported. However, the value checked to determine
if the command is supported is incorrect. Page 1921 of the Bluetooth
Core Spec v5.2 shows that bit 2 of octet 35 of the Supported_Commands
field corresponds to the LE Set RPA Timeout command, but currently
bit 6 of octet 35 is checked. This patch checks the correct value
instead.

This issue led to the error seen in the following btmon output during
initialization of an adapter (rtl8761b) and prevented initialization
from completing.

< HCI Command: LE Set Resolvable Private Address Timeout (0x08|0x002e) plen 2
        Timeout: 900 seconds
> HCI Event: Command Complete (0x0e) plen 4
      LE Set Resolvable Private Address Timeout (0x08|0x002e) ncmd 2
        Status: Unsupported Remote Feature / Unsupported LMP Feature (0x1a)
= Close Index: 00:E0:4C:6B:E5:03

The error did not appear when running with this patch.

Signed-off-by: Edward Vear <edwardvear@gmail.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Rename get_adv_instance_scan_rsp
Luiz Augusto von Dentz [Sat, 14 Nov 2020 00:44:34 +0000 (16:44 -0800)]
Bluetooth: Rename get_adv_instance_scan_rsp

This renames get_adv_instance_scan_rsp to adv_instance_is_scannable and
make it return a bool since it was not actually properly return the size
of the scan response as one could expect.

Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Fix not sending Set Extended Scan Response
Luiz Augusto von Dentz [Sat, 14 Nov 2020 00:44:33 +0000 (16:44 -0800)]
Bluetooth: Fix not sending Set Extended Scan Response

Current code is actually failing on the following tests of mgmt-tester
because get_adv_instance_scan_rsp_len did not account for flags that
cause scan response data to be included resulting in non-scannable
instance when in fact it should be scannable.

Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: Fix for Bluetooth SIG test L2CAP/COS/CFD/BV-14-C
Jimmy Wahlberg [Mon, 16 Nov 2020 15:56:26 +0000 (16:56 +0100)]
Bluetooth: Fix for Bluetooth SIG test L2CAP/COS/CFD/BV-14-C

This test case is meant to verify that multiple
unknown options is included in the response.

BLUETOOTH CORE SPECIFICATION Version 5.2 | Vol 3, Part A
page 1057

'On an unknown option failure (Result=0x0003),
the option(s) that contain anoption type field that is not
understood by the recipient of the L2CAP_CONFIGURATION_REQ
packet shall be included in the L2CAP_CONFIGURATION_RSP
packet unless they are hints.'

Before this patch:

> ACL Data RX: Handle 11 flags 0x02 dlen 24
      L2CAP: Configure Request (0x04) ident 18 len 16
        Destination CID: 64
        Flags: 0x0000
        Option: Unknown (0x10) [mandatory]
        10 00 11 02 11 00 12 02 12 00
< ACL Data TX: Handle 11 flags 0x00 dlen 17
      L2CAP: Configure Response (0x05) ident 18 len 9
        Source CID: 64
        Flags: 0x0000
        Result: Failure - unknown options (0x0003)
        Option: Unknown (0x10) [mandatory]
        12

After this patch:

> ACL Data RX: Handle 11 flags 0x02 dlen 24
      L2CAP: Configure Request (0x04) ident 5 len 16
        Destination CID: 64
        Flags: 0x0000
        Option: Unknown (0x10) [mandatory]
        10 00 11 02 11 00 12 02 12 00
< ACL Data TX: Handle 11 flags 0x00 dlen 23
      L2CAP: Configure Response (0x05) ident 5 len 15
        Source CID: 64
        Flags: 0x0000
        Result: Failure - unknown options (0x0003)
        Option: Unknown (0x10) [mandatory]
        10 11 01 11 12 01 12

Signed-off-by: Jimmy Wahlberg <jimmywa@spotify.com>
Reviewed-by: Luiz Augusto Von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: sco: Fix crash when using BT_SNDMTU/BT_RCVMTU option
Wei Yongjun [Mon, 16 Nov 2020 13:24:21 +0000 (21:24 +0800)]
Bluetooth: sco: Fix crash when using BT_SNDMTU/BT_RCVMTU option

This commit add the invalid check for connected socket, without it will
causes the following crash due to sco_pi(sk)->conn being NULL:

KASAN: null-ptr-deref in range [0x0000000000000050-0x0000000000000057]
CPU: 3 PID: 4284 Comm: test_sco Not tainted 5.10.0-rc3+ #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-1ubuntu1 04/01/2014
RIP: 0010:sco_sock_getsockopt+0x45d/0x8e0
Code: 48 c1 ea 03 80 3c 02 00 0f 85 ca 03 00 00 49 8b 9d f8 04 00 00 48 b8 00
      00 00 00 00 fc ff df 48 8d 7b 50 48 89 fa 48 c1 ea 03 <0f> b6 04 02 84
      c0 74 08 3c 03 0f 8e b5 03 00 00 8b 43 50 48 8b 0c
RSP: 0018:ffff88801bb17d88 EFLAGS: 00010206
RAX: dffffc0000000000 RBX: 0000000000000000 RCX: ffffffff83a4ecdf
RDX: 000000000000000a RSI: ffffc90002fce000 RDI: 0000000000000050
RBP: 1ffff11003762fb4 R08: 0000000000000001 R09: ffff88810e1008c0
R10: ffffffffbd695dcf R11: fffffbfff7ad2bb9 R12: 0000000000000000
R13: ffff888018ff1000 R14: dffffc0000000000 R15: 000000000000000d
FS:  00007fb4f76c1700(0000) GS:ffff88811af80000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00005555e3b7a938 CR3: 00000001117be001 CR4: 0000000000770ee0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
 ? sco_skb_put_cmsg+0x80/0x80
 ? sco_skb_put_cmsg+0x80/0x80
 __sys_getsockopt+0x12a/0x220
 ? __ia32_sys_setsockopt+0x150/0x150
 ? syscall_enter_from_user_mode+0x18/0x50
 ? rcu_read_lock_bh_held+0xb0/0xb0
 __x64_sys_getsockopt+0xba/0x150
 ? syscall_enter_from_user_mode+0x1d/0x50
 do_syscall_64+0x33/0x40
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Fixes: 0fc1a726f897 ("Bluetooth: sco: new getsockopt options BT_SNDMTU/BT_RCVMTU")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Luiz Augusto Von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btmtksdio: Add the missed release_firmware() in mtk_setup_firmware()
Jing Xiangfeng [Tue, 17 Nov 2020 02:43:32 +0000 (10:43 +0800)]
Bluetooth: btmtksdio: Add the missed release_firmware() in mtk_setup_firmware()

mtk_setup_firmware() misses to call release_firmware() in an error
path. Jump to free_fw to fix it.

Fixes: 737cd06072a7 ("Bluetooth: btmtksdio: fix up firmware download sequence")
Signed-off-by: Jing Xiangfeng <jingxiangfeng@huawei.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>
3 years agoBluetooth: btusb: Add the missed release_firmware() in btusb_mtk_setup_firmware()
Jing Xiangfeng [Tue, 17 Nov 2020 02:59:17 +0000 (10:59 +0800)]
Bluetooth: btusb: Add the missed release_firmware() in btusb_mtk_setup_firmware()

btusb_mtk_setup_firmware() misses to call release_firmware() in an error
path. Jump to err_release_fw to fix it.

Fixes: f645125711c8 ("Bluetooth: btusb: fix up firmware download sequence")
Signed-off-by: Jing Xiangfeng <jingxiangfeng@huawei.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Johan Hedberg <johan.hedberg@intel.com>