linux-2.6-microblaze.git
3 years agonet/mlx5: Fix uininitialized pointer read on pointer attr
Colin Ian King [Tue, 6 Oct 2020 18:12:43 +0000 (19:12 +0100)]
net/mlx5: Fix uininitialized pointer read on pointer attr

Currently the error exit path err_free kfree's attr. In the case where
flow and parse_attr failed to be allocated this return path will free
the uninitialized pointer attr, which is not correct.  In the other
case where attr fails to allocate attr does not need to be freed. So
in both error exits via err_free attr should not be freed, so remove
it.

Addresses-Coverity: ("Uninitialized pointer read")
Fixes: ff7ea04ad579 ("net/mlx5e: Fix potential null pointer dereference")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
3 years agoMerge branch 'bnxt_en-Updates-for-net-next'
Jakub Kicinski [Mon, 12 Oct 2020 21:42:55 +0000 (14:42 -0700)]
Merge branch 'bnxt_en-Updates-for-net-next'

Michael Chan says:

====================
bnxt_en: Updates for net-next.

This series contains these main changes:

1. Change of default message level to enable more logging.
2. Some cleanups related to processing async events from firmware.
3. Allow online ethtool selftest on multi-function PFs.
4. Return stored firmware version information to devlink.

v2:
Patch 3: Change bnxt_reset_task() to silent mode.
Patch 8 & 9: Ensure we copy NULL terminated fw strings to devlink.
Patch 8 & 9: Return directly after the last bnxt_dl_info_put() call.
Patch 9: If FW call to get stored dev info fails, return success to
         devlink without the stored versions.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Add stored FW version info to devlink info_get cb.
Vasundhara Volam [Mon, 12 Oct 2020 09:10:54 +0000 (05:10 -0400)]
bnxt_en: Add stored FW version info to devlink info_get cb.

This patch adds FW versions stored in the flash to devlink info_get
callback.  Return the correct fw.psid running version using the
newly added bp->nvm_cfg_ver.

v2:
Ensure stored pkg_name string is NULL terminated when copied to
devlink.

Return directly from the last call to bnxt_dl_info_put().

If the FW call to get stored version fails for any reason, return
success immediately to devlink without the stored versions.

Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-10-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Refactor bnxt_dl_info_get().
Vasundhara Volam [Mon, 12 Oct 2020 09:10:53 +0000 (05:10 -0400)]
bnxt_en: Refactor bnxt_dl_info_get().

Add a new function bnxt_dl_info_put() to simplify the code, as there
are more stored firmware version fields to be added in the next patch.

Also, rename fw_ver variable name to ncsi_ver for better naming while
copying to devlink info_get cb.

v2:
Ensure active_pkg_name string is NULL terminated when copied to
devlink.

Return directly from the last call to bnxt_dl_info_put().

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-9-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Add bnxt_hwrm_nvm_get_dev_info() to query NVM info.
Vasundhara Volam [Mon, 12 Oct 2020 09:10:52 +0000 (05:10 -0400)]
bnxt_en: Add bnxt_hwrm_nvm_get_dev_info() to query NVM info.

Add a new bnxt_hwrm_nvm_get_dev_info() to query firmware version
information via NVM_GET_DEV_INFO firmware command.  Use it to
get the running version of the NVM configuration information.

This new function will also be used in subsequent patches to get the
stored firmware versions.

Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-8-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Log unknown link speed appropriately.
Michael Chan [Mon, 12 Oct 2020 09:10:51 +0000 (05:10 -0400)]
bnxt_en: Log unknown link speed appropriately.

If the VF virtual link is set to always enabled, the speed may be
unknown when the physical link is down.  The driver currently logs
the link speed as 4294967295 Mbps which is SPEED_UNKNOWN.  Modify
the link up log message as "speed unknown" which makes more sense.

Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-7-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Log event_data1 and event_data2 when handling RESET_NOTIFY event.
Michael Chan [Mon, 12 Oct 2020 09:10:50 +0000 (05:10 -0400)]
bnxt_en: Log event_data1 and event_data2 when handling RESET_NOTIFY event.

Log these values that contain useful firmware state information.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-6-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Simplify bnxt_async_event_process().
Michael Chan [Mon, 12 Oct 2020 09:10:49 +0000 (05:10 -0400)]
bnxt_en: Simplify bnxt_async_event_process().

event_data1 and event_data2 are used when processing most events.
Store these in local variables at the beginning of the function to
simplify many of the case statements.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-5-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Set driver default message level.
Michael Chan [Mon, 12 Oct 2020 09:10:48 +0000 (05:10 -0400)]
bnxt_en: Set driver default message level.

Currently, bp->msg_enable has default value of 0.  It is more useful
to have the commonly used NETIF_MSG_DRV and NETIF_MSG_HW enabled by
default.

v2: Change the fall back bnxt_reset_task() inside bnxt_rx_ring_reset()
to silent mode.  With older fw, we would take the fall back path and
it would be very noisy.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Reviewed-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-4-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Enable online self tests for multi-host/NPAR mode.
Vasundhara Volam [Mon, 12 Oct 2020 09:10:47 +0000 (05:10 -0400)]
bnxt_en: Enable online self tests for multi-host/NPAR mode.

Online self tests are not disruptive and can be run in NPAR mode
and in multi-host NIC as well.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-3-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobnxt_en: Return -EROFS to user space, if NVM writes are not permitted.
Vasundhara Volam [Mon, 12 Oct 2020 09:10:46 +0000 (05:10 -0400)]
bnxt_en: Return -EROFS to user space, if NVM writes are not permitted.

If NVRAM resources are locked, NVM writes are not permitted. In such
scenarios, firmware returns HWRM_ERR_CODE_RESOURCE_LOCKED error to
firmware commands.

Reviewed-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Link: https://lore.kernel.org/r/1602493854-29283-2-git-send-email-michael.chan@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'linux-can-next-for-5.10-20201012' of git://git.kernel.org/pub/scm/linux...
Jakub Kicinski [Mon, 12 Oct 2020 16:53:08 +0000 (09:53 -0700)]
Merge tag 'linux-can-next-for-5.10-20201012' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
linux-can-next-for-5.10-20201012

Both patches are by Oliver Hartkopp, the first one addresses Jakub's review
comments of the ISOTP protocol, the other one removes version strings from
various CAN protocols.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agocx82310_eth: use netdev_err instead of dev_err
Ondrej Zary [Sat, 10 Oct 2020 14:00:47 +0000 (16:00 +0200)]
cx82310_eth: use netdev_err instead of dev_err

Use netdev_err for better device identification in syslog.

Signed-off-by: Ondrej Zary <linux@zary.sk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agocx82310_eth: re-enable ethernet mode after router reboot
Ondrej Zary [Sat, 10 Oct 2020 14:00:46 +0000 (16:00 +0200)]
cx82310_eth: re-enable ethernet mode after router reboot

When the router is rebooted without a power cycle, the USB device
remains connected but its configuration is reset. This results in
a non-working ethernet connection with messages like this in syslog:
usb 2-2: RX packet too long: 65535 B

Re-enable ethernet mode when receiving a packet with invalid size of
0xffff.

Signed-off-by: Ondrej Zary <linux@zary.sk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agocan: remove obsolete version strings
Oliver Hartkopp [Mon, 12 Oct 2020 07:43:54 +0000 (09:43 +0200)]
can: remove obsolete version strings

As pointed out by Jakub Kicinski here:
http://lore.kernel.org/r/20201009175751.5c54097f@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com
this patch removes the obsolete version information of the different
CAN protocols and the AF_CAN core module.

Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201012074354.25839-2-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agocan: isotp: implement cleanups / improvements from review
Oliver Hartkopp [Mon, 12 Oct 2020 07:43:53 +0000 (09:43 +0200)]
can: isotp: implement cleanups / improvements from review

As pointed out by Jakub Kicinski here:
http://lore.kernel.org/r/20201009175751.5c54097f@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com
this patch addresses the remarked issues:

- remove empty line in comment
- remove default=y for CAN_ISOTP in Kconfig
- make use of pr_notice_once()
- use GFP_ATOMIC instead of gfp_any() in soft hrtimer context

The version strings in the CAN subsystem are removed by a separate patch.

Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201012074354.25839-1-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
3 years agoMerge branch 'Offload-tc-vlan-mangle-to-mscc_ocelot-switch'
Jakub Kicinski [Sun, 11 Oct 2020 18:19:25 +0000 (11:19 -0700)]
Merge branch 'Offload-tc-vlan-mangle-to-mscc_ocelot-switch'

Vladimir Oltean says:

====================
Offload tc-vlan mangle to mscc_ocelot switch

This series offloads one more action to the VCAP IS1 ingress TCAM, which
is to change the classified VLAN for packets, according to the VCAP IS1
keys (VLAN, source MAC, source IP, EtherType, etc).
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoselftests: net: mscc: ocelot: add test for VLAN modify action
Vladimir Oltean [Thu, 8 Oct 2020 11:57:00 +0000 (14:57 +0300)]
selftests: net: mscc: ocelot: add test for VLAN modify action

Create a test that changes a VLAN ID from 200 to 300.

We also need to modify the preferences of the filters installed for the
other rules so that they are unique, because we now install the "tc-vlan
modify" filter in VCAP IS1 only temporarily, and we need to perform the
deletion by filter preference number.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: tag_ocelot: use VLAN information from tagging header when available
Vladimir Oltean [Thu, 8 Oct 2020 11:56:59 +0000 (14:56 +0300)]
net: dsa: tag_ocelot: use VLAN information from tagging header when available

When the Extraction Frame Header contains a valid classified VLAN, use
that instead of the VLAN header present in the packet.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: mscc: ocelot: offload VLAN mangle action to VCAP IS1
Vladimir Oltean [Thu, 8 Oct 2020 11:56:58 +0000 (14:56 +0300)]
net: mscc: ocelot: offload VLAN mangle action to VCAP IS1

The VCAP_IS1_ACT_VID_REPLACE_ENA action, from the VCAP IS1 ingress TCAM,
changes the classified VLAN.

We are only exposing this ability for switch ports that are under VLAN
aware bridges. This is because in standalone ports mode and under a
bridge with vlan_filtering=0, the ocelot driver configures the switch to
operate as VLAN-unaware, so the classified VLAN is not derived from the
802.1Q header from the packet, but instead is always equal to the
port-based VLAN ID of the ingress port. We _can_ still change the
classified VLAN for packets when operating in this mode, but the end
result will most likely be a drop, since both the ingress and the egress
port need to be members of the modified VLAN. And even if we install the
new classified VLAN into the VLAN table of the switch, the result would
still not be as expected: we wouldn't see, on the output port, the
modified VLAN tag, but the original one, even though the classified VLAN
was indeed modified. This is because of how the hardware works: on
egress, what is pushed to the frame is a "port tag", which gives us the
following options:

- Tag all frames with port tag (derived from the classified VLAN)
- Tag all frames with port tag, except if the classified VLAN is 0 or
  equal to the native VLAN of the egress port
- No port tag

Needless to say, in VLAN-unaware mode we are disabling the port tag.
Otherwise, the existing VLAN tag would be ignored, and a second VLAN
tag (the port tag), holding the classified VLAN, would be pushed
(instead of replacing the existing 802.1Q tag). This is definitely not
what the user wanted when installing a "vlan modify" action.

So it is simply not worth bothering with VLAN modify rules under other
configurations except when the ports are fully VLAN-aware.

Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'enetc-Migrate-to-PHYLINK-and-PCS_LYNX'
Jakub Kicinski [Sun, 11 Oct 2020 18:04:56 +0000 (11:04 -0700)]
Merge branch 'enetc-Migrate-to-PHYLINK-and-PCS_LYNX'

Claudiu Manoil says:

====================
enetc: Migrate to PHYLINK and PCS_LYNX

Transitioning the enetc driver from phylib to phylink.
Offloading the serdes configuration to the PCS_LYNX
module is a mandatory part of this transition. Aiming
for a cleaner, more maintainable design, and better
code reuse.
The first 2 patches are clean up prerequisites.

Tested on a p1028rdb board.

v2: validate() explicitly rejects now all interface modes not
supported by the driver instead of relying on the device tree
to provide only supported interfaces, and dropped redundant
activation of pcs_poll (addressing Ioana's findings)
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoenetc: Migrate to PHYLINK and PCS_LYNX
Claudiu Manoil [Wed, 7 Oct 2020 09:48:23 +0000 (12:48 +0300)]
enetc: Migrate to PHYLINK and PCS_LYNX

This is a methodical transition of the driver from phylib
to phylink, following the guidelines from sfp-phylink.rst.
The MAC register configurations based on interface mode
were moved from the probing path to the mac_config() hook.
MAC enable and disable commands (enabling Rx and Tx paths
at MAC level) were also extracted and assigned to their
corresponding phylink hooks.
As part of the migration to phylink, the serdes configuration
from the driver was offloaded to the PCS_LYNX module,
introduced in commit 0da4c3d393e4 ("net: phy: add Lynx PCS module"),
the PCS_LYNX module being a mandatory component required to
make the enetc driver work with phylink.

Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.cionei@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoarm64: dts: fsl-ls1028a-rdb: Specify in-band mode for ENETC port 0
Claudiu Manoil [Wed, 7 Oct 2020 09:48:22 +0000 (12:48 +0300)]
arm64: dts: fsl-ls1028a-rdb: Specify in-band mode for ENETC port 0

As part of the transition of the enetc ethernet driver from phylib
to phylink, the in-band operation mode of the SGMII interface
from enetc port 0 needs to be specified explicitly for phylink.

Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoenetc: Clean up serdes configuration
Claudiu Manoil [Wed, 7 Oct 2020 09:48:21 +0000 (12:48 +0300)]
enetc: Clean up serdes configuration

Decouple internal mdio bus creation from serdes
configuration, as a prerequisite to offloading
serdes configuration to a different module.
Group together mdio bus creation routines, cleanup.

Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoenetc: Clean up MAC and link configuration
Claudiu Manoil [Wed, 7 Oct 2020 09:48:20 +0000 (12:48 +0300)]
enetc: Clean up MAC and link configuration

Decouple level MAC configuration based on phy interface type
from general port configuration.
Group together MAC and link configuration code.
Decouple external mdio bus creation from interface type
parsing.  No longer return an (unhandled) error code when
phy_node not found, use phy_node to indicate whether the
port has a phy or not.  No longer fall-through when serdes
configuration fails for the link modes that require
internal link configuration.

Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodrivers/net/wan/hdlc_fr: Move the skb_headroom check out of fr_hard_header
Xie He [Wed, 7 Oct 2020 18:32:03 +0000 (11:32 -0700)]
drivers/net/wan/hdlc_fr: Move the skb_headroom check out of fr_hard_header

Move the skb_headroom check out of fr_hard_header and into pvc_xmit.
This has two benefits:

1. Originally we only do this check for skbs sent by users on Ethernet-
emulating PVC devices. After the change we do this check for skbs sent on
normal PVC devices, too.
(Also add a comment to make it clear that this is only a protection
against upper layers that don't take dev->needed_headroom into account.
Such upper layers should be rare and I believe they should be fixed.)

2. After the change we can simplify the parameter list of fr_hard_header.
We no longer need to use a pointer to pointers (skb_p) because we no
longer need to replace the skb inside fr_hard_header.

Cc: Krzysztof Halasa <khc@pm.waw.pl>
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: rtl8366rb: Roof MTU for switch
Linus Walleij [Thu, 8 Oct 2020 21:03:40 +0000 (23:03 +0200)]
net: dsa: rtl8366rb: Roof MTU for switch

The MTU setting for this DSA switch is global so we need
to keep track of the MTU set for each port, then as soon
as any MTU changes, roof the MTU to the biggest common
denominator and poke that into the switch MTU setting.

To achieve this we need a per-chip-variant state container
for the RTL8366RB to use for the RTL8366RB-specific
stuff. Other SMI switches does seem to have per-port
MTU setting capabilities.

Fixes: 5f4a8ef384db ("net: dsa: rtl8366rb: Support setting MTU")
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: phy: Move of_mdio from drivers/of to drivers/net/mdio
Calvin Johnson [Thu, 8 Oct 2020 14:47:06 +0000 (20:17 +0530)]
net: phy: Move of_mdio from drivers/of to drivers/net/mdio

Better place for of_mdio.c is drivers/net/mdio.
Move of_mdio.c from drivers/of to drivers/net/mdio

Signed-off-by: Calvin Johnson <calvin.johnson@oss.nxp.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodpaa_eth: enable NETIF_MSG_HW by default
Maxim Kochetkov [Thu, 8 Oct 2020 12:03:12 +0000 (15:03 +0300)]
dpaa_eth: enable NETIF_MSG_HW by default

When packets are received on the error queue, this function under
net_ratelimit():

netif_err(priv, hw, net_dev, "Err FD status = 0x%08x\n");

does not get printed. Instead we only see:

[ 3658.845592] net_ratelimit: 244 callbacks suppressed
[ 3663.969535] net_ratelimit: 230 callbacks suppressed
[ 3669.085478] net_ratelimit: 228 callbacks suppressed

Enabling NETIF_MSG_HW fixes this issue, and we can see some information
about the frame descriptors of packets.

Signed-off-by: Maxim Kochetkov <fido_max@inbox.ru>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Madalin Bucur <madalin.bucur@oss.nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agor8169: factor out handling rtl8169_stats
Heiner Kallweit [Fri, 9 Oct 2020 14:20:34 +0000 (16:20 +0200)]
r8169: factor out handling rtl8169_stats

Factor out handling the private packet/byte counters to new
functions rtl_get_priv_stats() and rtl_inc_priv_stats().

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: usbnet: remove driver version
Heiner Kallweit [Fri, 9 Oct 2020 12:10:57 +0000 (14:10 +0200)]
net: usbnet: remove driver version

Obviously this driver version doesn't make sense. Go with the default
and let ethtool display the kernel version.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: thunderx: Use struct_size() helper in kmalloc()
Gustavo A. R. Silva [Thu, 8 Oct 2020 14:28:06 +0000 (09:28 -0500)]
net: thunderx: Use struct_size() helper in kmalloc()

Make use of the new struct_size() helper instead of the offsetof() idiom.

Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'wireless-drivers-next-2020-10-09' of git://git.kernel.org/pub/scm/linux...
Jakub Kicinski [Sat, 10 Oct 2020 16:36:54 +0000 (09:36 -0700)]
Merge tag 'wireless-drivers-next-2020-10-09' of git://git./linux/kernel/git/kvalo/wireless-drivers-next

Kalle Valo says:

====================
wireless-drivers-next patches for v5.10

Fourth and last set of patches for v5.10. Most of these are iwlwifi
patches, but few small fixes to other drivers as well.

Major changes:

iwlwifi

* PNVM support (platform-specific phy config data)

* bump the FW API support to 59
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'mac80211-next-for-net-next-2020-10-08' of git://git.kernel.org/pub/scm...
Jakub Kicinski [Sat, 10 Oct 2020 16:12:52 +0000 (09:12 -0700)]
Merge tag 'mac80211-next-for-net-next-2020-10-08' of git://git./linux/kernel/git/jberg/mac80211-next

Johannes Berg says:

====================
A handful of changes:
 * fixes for the recent S1G work
 * a docbook build time improvement
 * API to pass beacon rate to lower-level driver
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'netlink-export-policy-on-validation-failures'
Jakub Kicinski [Sat, 10 Oct 2020 03:22:49 +0000 (20:22 -0700)]
Merge branch 'netlink-export-policy-on-validation-failures'

Johannes Berg says:

====================
netlink: export policy on validation failures

Export the policy used for attribute validation when it fails,
so e.g. for an out-of-range attribute userspace immediately gets
the valid ranges back.

v2 incorporates the suggestion from Jakub to have a function to
estimate the size (netlink_policy_dump_attr_size_estimate()) and
check that it does the right thing on the *normal* policy dumps,
not (just) when calling it from the error scenario.

v3 only addresses a few minor style issues.

v4 fixes up a forgotten 'git add' ... sorry.

v5 is a resend, I messed up v4's cover letter subject (saying v3)
and apparently the second patch didn't go out at all.

Tested using nl80211/iw in a few scenarios, seems to work fine
and return the policy back, e.g.

kernel reports: integer out of range
policy: 04 00 0b 00 0c 00 04 00 01 00 00 00 00 00 00 00
        ^ padding
                    ^ minimum allowed value
policy: 04 00 0b 00 0c 00 05 00 ff ff ff ff 00 00 00 00
        ^ padding
                    ^ maximum allowed value
policy: 08 00 01 00 04 00 00 00
        ^ type 4 == U32

for an out-of-range case.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonetlink: export policy in extended ACK
Johannes Berg [Thu, 8 Oct 2020 10:45:17 +0000 (12:45 +0200)]
netlink: export policy in extended ACK

Add a new attribute NLMSGERR_ATTR_POLICY to the extended ACK
to advertise the policy, e.g. if an attribute was out of range,
you'll know the range that's permissible.

Add new NL_SET_ERR_MSG_ATTR_POL() and NL_SET_ERR_MSG_ATTR_POL()
macros to set this, since realistically it's only useful to do
this when the bad attribute (offset) is also returned.

Use it in lib/nlattr.c which practically does all the policy
validation.

v2:
 - add and use netlink_policy_dump_attr_size_estimate()
v3:
 - remove redundant break
v4:
 - really remove redundant break ... sorry

Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonetlink: policy: refactor per-attr policy writing
Johannes Berg [Thu, 8 Oct 2020 10:45:16 +0000 (12:45 +0200)]
netlink: policy: refactor per-attr policy writing

Refactor the per-attribute policy writing into a new
helper function, to be used later for dumping out the
policy of a rejected attribute.

v2:
 - fix some indentation
v3:
 - change variable order in netlink_policy_dump_write()

Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'net-smc-updates-2020-10-07'
Jakub Kicinski [Sat, 10 Oct 2020 01:15:59 +0000 (18:15 -0700)]
Merge branch 'net-smc-updates-2020-10-07'

Karsten Graul says:

====================
net/smc: updates 2020-10-07

Patch 1 and 2 address warnings from static code checkers, and patch 3
handles a case when all proposed ISM V2 devices fail to init and no V1
devices are tried afterwards.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/smc: restore smcd_version when all ISM V2 devices failed to init
Karsten Graul [Wed, 7 Oct 2020 20:57:43 +0000 (22:57 +0200)]
net/smc: restore smcd_version when all ISM V2 devices failed to init

Field ini->smcd_version is set to SMC_V2 before calling
smc_listen_ism_init(). This clears the V1 bit that may be set. When all
matching ISM V2 devices fail to initialize then the smcd_version field
needs to get restored to allow any possible V1 devices to initialize.
And be consistent, always go to the not_found label when no device was
found.

Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/smc: cleanup buffer usage in smc_listen_work()
Karsten Graul [Wed, 7 Oct 2020 20:57:42 +0000 (22:57 +0200)]
net/smc: cleanup buffer usage in smc_listen_work()

coccinelle informs about
net/smc/af_smc.c:1770:10-11: WARNING: opportunity for kzfree/kvfree_sensitive

Its not that kzfree() would help here, the memset() is done to prepare
the buffer for another socket receive.
Fix that warning message by reordering the calls, while at it eliminate
the unneeded variable cclc2 and use sizeof(*buf) as above in the same
function. No functional changes.

Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/smc: consolidate unlocking in same function
Karsten Graul [Wed, 7 Oct 2020 20:57:41 +0000 (22:57 +0200)]
net/smc: consolidate unlocking in same function

Static code checkers warn of inconsistent returns because the lgr mutex
is locked in one function and unlocked in a function called by the
locking function:
net/smc/af_smc.c:823 smc_connect_rdma() warn: inconsistent returns 'smc_client_lgr_pending'.
net/smc/af_smc.c:897 smc_connect_ism() warn: inconsistent returns 'smc_server_lgr_pending'.

Make the code consistent by doing the unlock in the same function that
fetches the lock. No functional changes.

Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'linux-can-next-for-5.10-20201007' of git://git.kernel.org/pub/scm/linux...
Jakub Kicinski [Sat, 10 Oct 2020 00:58:46 +0000 (17:58 -0700)]
Merge tag 'linux-can-next-for-5.10-20201007' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
linux-can-next-for-5.10-20201007

The first 3 patches are by me and fix several warnings found
when compiling the  kernel with W=1.

Lukas Bulwahn's patch adjusts the MAINTAINERS file, to accommodate
the renaming of the mcp251xfd driver.

Vincent Mailhol contributes 3 patches for the CAN networking layer.
First error queue support is added the the CAN RAW protocol.
The second patch converts the get_can_dlc() and get_canfd_dlc()
in-Kernel-only macros from using __u8 to u8.
The third patch adds a helper function to calculate the length of
one bit in in multiple of time quanta.

Oliver Hartkopp's patch add support for the ISO 15765-2:2016
transport protocol to the CAN stack.

Three patches by Lad Prabhakar add documentation for various
new rcar controllers to the device tree bindings of the rcar_can
and rcan_canfd driver.

Michael Walle's patch adds various processors to the flexcan
driver binding documentation.

The next two patches are by me and target the flexcan driver aswell.
The remove the ack_grp and ack_bit from the fsl,stop-mode DT property
and the driver, as they are not used anymore. As these are the last
two arguments this change will not break existing device trees.

The last three patches are by Srinivas Neeli and target
the xilinx_can driver.
The first one increases the lower limit for the bit rate
prescaler to 2, the other two fix sparse and coverity findings.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch '100GbE-Intel-Wired-LAN-Driver-Updates-2020-10-07'
Jakub Kicinski [Fri, 9 Oct 2020 20:15:32 +0000 (13:15 -0700)]
Merge branch '100GbE-Intel-Wired-LAN-Driver-Updates-2020-10-07'

Tony Nguyen says:

====================
100GbE Intel Wired LAN Driver Updates 2020-10-07

This series contains updates to ice driver only.

Andy Shevchenko changes usage to %*phD format to print small buffer as hex
string.

Bruce removes repeated words reported by checkpatch.

Ani changes ice_info_get_dsn() to return void as it always returns
success.

Jake adds devlink reporting of fw.app.bundle_id. Moves devlink_port
structure to ice_vsi to resolve issues with cleanup. Adds additional
debug info for firmware updates.

Bixuan Cui resolves -Wpointer-to-int-cast warnings.

Dan adds additional packet type masks and checks to prevent overwriting
existing Flow Director rules.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: fix adding IP4 IP6 Flow Director rules
Dan Nowlin [Wed, 7 Oct 2020 17:54:47 +0000 (10:54 -0700)]
ice: fix adding IP4 IP6 Flow Director rules

A subsequent addition of an IP4 or IP6 rule after other rules would
overwrite any existing TCAM entries of related L4 protocols(ex: tcp4 or
udp6). This was due to the mask including too many TCAM entries. Add new
packet type masks with bits properly excluded so rules are not overwritten.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Henry Tieman <henry.w.tieman@intel.com>
Tested-by: Brijesh Behera <brijeshx.behera@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: Fix pointer cast warnings
Bixuan Cui [Wed, 7 Oct 2020 17:54:46 +0000 (10:54 -0700)]
ice: Fix pointer cast warnings

pointers should be casted to unsigned long to avoid
-Wpointer-to-int-cast warnings:

drivers/net/ethernet/intel/ice/ice_flow.h:197:33: warning:
    cast from pointer to integer of different size
drivers/net/ethernet/intel/ice/ice_flow.h:198:32: warning:
    cast to pointer from integer of different size

Signed-off-by: Bixuan Cui <cuibixuan@huawei.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: add additional debug logging for firmware update
Jacob Keller [Wed, 7 Oct 2020 17:54:45 +0000 (10:54 -0700)]
ice: add additional debug logging for firmware update

While debugging a recent failure to update the flash of an ice device,
I found it helpful to add additional logging which helped determine the
root cause of the problem being a timeout issue.

Add some extra dev_dbg() logging messages which can be enabled using the
dynamic debug facility, including one for ice_aq_wait_for_event that
will use jiffies to capture a rough estimate of how long we waited for
the completion of a firmware command.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Brijesh Behera <brijeshx.behera@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: refactor devlink_port to be per-VSI
Jacob Keller [Wed, 7 Oct 2020 17:54:44 +0000 (10:54 -0700)]
ice: refactor devlink_port to be per-VSI

Currently, the devlink_port structure is stored within the ice_pf. This
made sense because we create a single devlink_port for each PF. This
setup does not mesh with the abstractions in the driver very well, and
led to a flow where we accidentally call devlink_port_unregister twice
during error cleanup.

In particular, if devlink_port_register or devlink_port_unregister are
called twice, this leads to a kernel panic. This appears to occur during
some possible flows while cleaning up from a failure during driver
probe.

If register_netdev fails, then we will call devlink_port_unregister in
ice_cfg_netdev as it cleans up. Later, we again call
devlink_port_unregister since we assume that we must cleanup the port
that is associated with the PF structure.

This occurs because we cleanup the devlink_port for the main PF even
though it was not allocated. We allocated the port within a per-VSI
function for managing the main netdev, but did not release the port when
cleaning up that VSI, the allocation and destruction are not aligned.

Instead of attempting to manage the devlink_port as part of the PF
structure, manage it as part of the PF VSI. Doing this has advantages,
as we can match the de-allocation of the devlink_port with the
unregister_netdev associated with the main PF VSI.

Moving the port to the VSI is preferable as it paves the way for
handling devlink ports allocated for other purposes such as SR-IOV VFs.

Since we're changing up how we allocate the devlink_port, also change
the indexing. Originally, we indexed the port using the PF id number.
This came from an old goal of sharing a devlink for each physical
function. Managing devlink instances across multiple function drivers is
not workable. Instead, lets set the port number to the logical port
number returned by firmware and set the index using the VSI index
(sometimes referred to as VSI handle).

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: add the DDP Track ID to devlink info
Jacob Keller [Wed, 7 Oct 2020 17:54:43 +0000 (10:54 -0700)]
ice: add the DDP Track ID to devlink info

Add "fw.app.bundle_id" to display the DDP Track ID of the active DDP
package. This id is similar to "fw.bundle_id" and is a unique identifier
for the DDP package that is loaded in the device. Each new DDP has
a unique Track ID generated for it, and the ID can be used to identify
and track the DDP package.

Add documentation for the new devlink info version.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: Change ice_info_get_dsn to be void
Anirudh Venkataramanan [Wed, 7 Oct 2020 17:54:42 +0000 (10:54 -0700)]
ice: Change ice_info_get_dsn to be void

ice_info_get_dsn always returns 0, so just make it void.

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: remove repeated words
Bruce Allan [Wed, 7 Oct 2020 17:54:41 +0000 (10:54 -0700)]
ice: remove repeated words

A new test in checkpatch detects repeated words; cleanup all pre-existing
occurrences of those now.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Co-developed-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoice: devlink: use %*phD to print small buffer
Andy Shevchenko [Wed, 7 Oct 2020 17:54:40 +0000 (10:54 -0700)]
ice: devlink: use %*phD to print small buffer

Use %*phD format to print small buffer as hex string.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: dsa: microchip: add ksz9563 to ksz9477 I2C driver
Christian Eggers [Wed, 7 Oct 2020 09:30:49 +0000 (11:30 +0200)]
net: dsa: microchip: add ksz9563 to ksz9477 I2C driver

Add support for the KSZ9563 3-Port Gigabit Ethernet Switch to the
ksz9477 driver. The KSZ9563 supports both SPI (already in) and I2C. The
ksz9563 is already in the device tree binding documentation.

Signed-off-by: Christian Eggers <ceggers@arri.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/rds: suppress page allocation failure error in recv buffer refill
Manjunath Patil [Tue, 6 Oct 2020 21:31:37 +0000 (14:31 -0700)]
net/rds: suppress page allocation failure error in recv buffer refill

RDS/IB tries to refill the recv buffer in softirq context using
GFP_NOWAIT flag. However alloc failure is handled by queueing a work to
refill the recv buffer with GFP_KERNEL flag. This means failure to
allocate with GFP_NOWAIT isn't fatal. Do not print the PAF warnings if
softirq context fails to refill the recv buffer. We will see the PAF
warnings when worker also fails to allocate.

Signed-off-by: Manjunath Patil <manjunath.b.patil@oracle.com>
Reviewed-by: Aruna Ramakrishna <aruna.ramakrishna@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge branch 'devlink-add-reload-action-and-limit-options'
Jakub Kicinski [Fri, 9 Oct 2020 19:07:10 +0000 (12:07 -0700)]
Merge branch 'devlink-add-reload-action-and-limit-options'

Moshe Shemesh says:

====================
Add devlink reload action and limit options

Introduce new options on devlink reload API to enable the user to select
the reload action required and constrains limits on these actions that he
may want to ensure. Complete support for reload actions in mlx5.
The following reload actions are supported:
  driver_reinit: driver entities re-initialization, applying devlink-param
                 and devlink-resource values.
  fw_activate: firmware activate.

The uAPI is backward compatible, if the reload action option is omitted
from the reload command, the driver reinit action will be used.
Note that when required to do firmware activation some drivers may need
to reload the driver. On the other hand some drivers may need to reset
the firmware to reinitialize the driver entities. Therefore, the devlink
reload command returns the actions which were actually performed.

By default reload actions are not limited and driver implementation may
include reset or downtime as needed to perform the actions.
However, if reload limit is selected, the driver should perform only if
it can do it while keeping the limit constraints.
Reload limit added:
  no_reset: No reset allowed, no down time allowed, no link flap and no
            configuration is lost.

Each driver which supports devlink reload command should expose the
reload actions and limits supported.

Add reload stats to hold the history per reload action per limit.
For example, the number of times fw_activate has been done on this
device since the driver module was added or if the firmware activation
was done with or without reset.

Patch 1 changes devlink_reload_supported() param type to enable using
        it before allocating devlink.
Patch 2-3 add the new API reload action and reload limit options to
          devlink reload.
Patch 4-5 add reload stats and remote reload stats. These stats are
          exposed through devlink dev get.
Patches 6-11 add support on mlx5 for devlink reload action fw_activate
            and handle the firmware reset events.
Patches 12-13 add devlink enable remote dev reset parameter and use it
             in mlx5.
Patches 14-15 mlx5 add devlink reload limit no_reset support for
              fw_activate reload action.
Patch 16 adds documentation file devlink-reload.rst
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add Documentation/networking/devlink/devlink-reload.rst
Moshe Shemesh [Wed, 7 Oct 2020 06:00:57 +0000 (09:00 +0300)]
devlink: Add Documentation/networking/devlink/devlink-reload.rst

Add devlink reload rst documentation file.
Update index file to include it.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Add support for devlink reload limit no reset
Moshe Shemesh [Wed, 7 Oct 2020 06:00:56 +0000 (09:00 +0300)]
net/mlx5: Add support for devlink reload limit no reset

Add support for devlink reload action fw_activate with reload limit
no_reset which does firmware live patching, updating the firmware image
without reset, no downtime and no configuration lose. The driver checks
if the firmware is capable of handling the pending firmware changes as a
live patch. If it is then it triggers firmware live patching flow.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Add support for fw live patch event
Moshe Shemesh [Wed, 7 Oct 2020 06:00:55 +0000 (09:00 +0300)]
net/mlx5: Add support for fw live patch event

Firmware live patch event notifies the driver that the firmware was just
updated using live patch. In such case the driver should not reload or
re-initiate entities, part to updating the firmware version and
re-initiate the firmware tracer which can be updated by live patch with
new strings database to help debugging an issue.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Add devlink param enable_remote_dev_reset support
Moshe Shemesh [Wed, 7 Oct 2020 06:00:54 +0000 (09:00 +0300)]
net/mlx5: Add devlink param enable_remote_dev_reset support

The enable_remote_dev_reset devlink param flags that the host admin
allows resets by other hosts. In case it is cleared mlx5 host PF driver
will send NACK on pci sync for firmware update reset request and the
command will fail.
By default enable_remote_dev_reset parameter is true, so pci sync for
firmware update reset is enabled.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add enable_remote_dev_reset generic parameter
Moshe Shemesh [Wed, 7 Oct 2020 06:00:53 +0000 (09:00 +0300)]
devlink: Add enable_remote_dev_reset generic parameter

The enable_remote_dev_reset devlink param flags that the host admin
allows device resets that can be initiated by other hosts. This
parameter is useful for setups where a device is shared by different
hosts, such as multi-host setup. Once the user set this parameter to
false, the driver should NACK any attempt to reset the device while the
driver is loaded.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Add support for devlink reload action fw activate
Moshe Shemesh [Wed, 7 Oct 2020 06:00:52 +0000 (09:00 +0300)]
net/mlx5: Add support for devlink reload action fw activate

Add support for devlink reload action fw_activate. To activate firmware
image the mlx5 driver resets the firmware and reloads it from flash. If
a new image was stored on flash it will be loaded. Once this reload
command is executed the driver initiates fw sync reset flow, where the
firmware synchronizes all PFs on coming reset and driver reload.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Handle sync reset abort event
Moshe Shemesh [Wed, 7 Oct 2020 06:00:51 +0000 (09:00 +0300)]
net/mlx5: Handle sync reset abort event

If firmware sends sync_reset_abort to driver the driver should clear the
reset requested mode as reset is not expected any more.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Handle sync reset now event
Moshe Shemesh [Wed, 7 Oct 2020 06:00:50 +0000 (09:00 +0300)]
net/mlx5: Handle sync reset now event

On sync_reset_now event the driver does reload and PCI link toggle to
activate firmware upgrade reset. When the firmware sends this event it
syncs the event on all PFs, so all PFs will do PCI link toggle at once.
To do PCI link toggle, the driver ensures that no other device ID under
the same bridge by checking that all the PF functions under the same PCI
bridge have same device ID. If no other device it uses PCI bridge link
control to turn link down and up.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Handle sync reset request event
Moshe Shemesh [Wed, 7 Oct 2020 06:00:49 +0000 (09:00 +0300)]
net/mlx5: Handle sync reset request event

Once the driver gets sync_reset_request from firmware it prepares for the
coming reset and sends acknowledge.
After getting this event the driver expects device reset, either it will
trigger PCI reset on sync_reset_now event or such PCI reset will be
triggered by another PF of the same device. So it moves to reset
requested mode and if it gets PCI reset triggered by the other PF it
detect the reset and reloads.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Set cap for pci sync for fw update event
Moshe Shemesh [Wed, 7 Oct 2020 06:00:48 +0000 (09:00 +0300)]
net/mlx5: Set cap for pci sync for fw update event

Set capability to notify the firmware that this host driver is capable
of handling pci sync for firmware update events.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/mlx5: Add functions to set/query MFRL register
Moshe Shemesh [Wed, 7 Oct 2020 06:00:47 +0000 (09:00 +0300)]
net/mlx5: Add functions to set/query MFRL register

Add functions to query and set the MFRL reset options supported by
firmware.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add remote reload stats
Moshe Shemesh [Wed, 7 Oct 2020 06:00:46 +0000 (09:00 +0300)]
devlink: Add remote reload stats

Add remote reload stats to hold the history of actions performed due
devlink reload commands initiated by remote host. For example, in case
firmware activation with reset finished successfully but was initiated
by remote host.

The function devlink_remote_reload_actions_performed() is exported to
enable drivers update on remote reload actions performed as it was not
initiated by their own devlink instance.

Expose devlink remote reload stats to the user through devlink dev get
command.

Examples:
$ devlink dev show
pci/0000:82:00.0:
  stats:
      reload:
        driver_reinit 2 fw_activate 1 fw_activate_no_reset 0
      remote_reload:
        driver_reinit 0 fw_activate 0 fw_activate_no_reset 0
pci/0000:82:00.1:
  stats:
      reload:
        driver_reinit 1 fw_activate 0 fw_activate_no_reset 0
      remote_reload:
        driver_reinit 1 fw_activate 1 fw_activate_no_reset 0

$ devlink dev show -jp
{
    "dev": {
        "pci/0000:82:00.0": {
            "stats": {
                "reload": {
                    "driver_reinit": 2,
                    "fw_activate": 1,
                    "fw_activate_no_reset": 0
                },
                "remote_reload": {
                    "driver_reinit": 0,
                    "fw_activate": 0,
                    "fw_activate_no_reset": 0
                }
            }
        },
        "pci/0000:82:00.1": {
            "stats": {
                "reload": {
                    "driver_reinit": 1,
                    "fw_activate": 0,
                    "fw_activate_no_reset": 0
                },
                "remote_reload": {
                    "driver_reinit": 1,
                    "fw_activate": 1,
                    "fw_activate_no_reset": 0
                }
            }
        }
    }
}

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add reload stats
Moshe Shemesh [Wed, 7 Oct 2020 06:00:45 +0000 (09:00 +0300)]
devlink: Add reload stats

Add reload stats to hold the history per reload action type and limit.

For example, the number of times fw_activate has been performed on this
device since the driver module was added or if the firmware activation
was performed with or without reset.

Add devlink notification on stats update.

Expose devlink reload stats to the user through devlink dev get command.

Examples:
$ devlink dev show
pci/0000:82:00.0:
  stats:
      reload:
        driver_reinit 2 fw_activate 1 fw_activate_no_reset 0
pci/0000:82:00.1:
  stats:
      reload:
        driver_reinit 1 fw_activate 0 fw_activate_no_reset 0

$ devlink dev show -jp
{
    "dev": {
        "pci/0000:82:00.0": {
            "stats": {
                "reload": {
                    "driver_reinit": 2,
                    "fw_activate": 1,
                    "fw_activate_no_reset": 0
                }
            }
        },
        "pci/0000:82:00.1": {
            "stats": {
                "reload": {
                    "driver_reinit": 1,
                    "fw_activate": 0,
                    "fw_activate_no_reset": 0
                }
            }
        }
    }
}

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add devlink reload limit option
Moshe Shemesh [Wed, 7 Oct 2020 06:00:44 +0000 (09:00 +0300)]
devlink: Add devlink reload limit option

Add reload limit to demand restrictions on reload actions.
Reload limits supported:
no_reset: No reset allowed, no down time allowed, no link flap and no
          configuration is lost.

By default reload limit is unspecified and so no constraints on reload
actions are required.

Some combinations of action and limit are invalid. For example, driver
can not reinitialize its entities without any downtime.

The no_reset reload limit will have usecase in this patchset to
implement restricted fw_activate on mlx5.

Have the uapi parameter of reload limit ready for future support of
multiselection.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Add reload action option to devlink reload command
Moshe Shemesh [Wed, 7 Oct 2020 06:00:43 +0000 (09:00 +0300)]
devlink: Add reload action option to devlink reload command

Add devlink reload action to allow the user to request a specific reload
action. The action parameter is optional, if not specified then devlink
driver re-init action is used (backward compatible).
Note that when required to do firmware activation some drivers may need
to reload the driver. On the other hand some drivers may need to reset
the firmware to reinitialize the driver entities. Therefore, the devlink
reload command returns the actions which were actually performed.
Reload actions supported are:
driver_reinit: driver entities re-initialization, applying devlink-param
               and devlink-resource values.
fw_activate: firmware activate.

command examples:
$devlink dev reload pci/0000:82:00.0 action driver_reinit
reload_actions_performed:
  driver_reinit

$devlink dev reload pci/0000:82:00.0 action fw_activate
reload_actions_performed:
  driver_reinit fw_activate

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agodevlink: Change devlink_reload_supported() param type
Moshe Shemesh [Wed, 7 Oct 2020 06:00:42 +0000 (09:00 +0300)]
devlink: Change devlink_reload_supported() param type

Change devlink_reload_supported() function to get devlink_ops pointer
param instead of devlink pointer param.
This change will be used in the next patch to check if devlink reload is
supported before devlink instance is allocated.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet/sched: get rid of qdisc->padded
Eric Dumazet [Wed, 7 Oct 2020 16:51:11 +0000 (09:51 -0700)]
net/sched: get rid of qdisc->padded

kmalloc() of sufficiently big portion of memory is cache-aligned
in regular conditions. If some debugging options are used,
there is no reason qdisc structures would need 64-byte alignment
if most other kernel structures are not aligned.

This get rid of QDISC_ALIGN and QDISC_ALIGNTO.

Addition of privdata field will help implementing
the reverse of qdisc_priv() and documents where
the private data is.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Allen Pais <allen.lkml@gmail.com>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoRevert "iwlwifi: remove wide_cmd_header field"
Luca Coelho [Fri, 9 Oct 2020 09:21:23 +0000 (12:21 +0300)]
Revert "iwlwifi: remove wide_cmd_header field"

This reverts commit 0a8159cbd11abf586693f1ec5264d0df587b71d8.

It turns out that this flag is used by iwldvm, so we can't get rid of
it.  This broke iwldvm devices with BAD_COMMAND errors.

Fixes: 0a8159cbd11a ("iwlwifi: remove wide_cmd_header field")
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201009122123.3e4ee0ad7a71.Id6d95ae601f048aeb4d2ed63a1712e469da84369@changeid
3 years agoselftests: mptcp: interpret \n as a new line
Matthieu Baerts [Tue, 6 Oct 2020 16:06:30 +0000 (18:06 +0200)]
selftests: mptcp: interpret \n as a new line

In case of errors, this message was printed:

  (...)
  balanced bwidth with unbalanced delay       5233 max 5005  [ fail ]
  client exit code 0, server 0
  \nnetns ns3-0-EwnkPH socket stat for 10003:
  (...)

Obviously, the idea was to add a new line before the socket stat and not
print "\nnetns".

The commit 8b974778f998 ("selftests: mptcp: interpret \n as a new line")
is very similar to this one. But the modification in simult_flows.sh was
missed because this commit above was done in parallel to one here below.

Fixes: 1a418cb8e888 ("mptcp: simult flow self-tests")
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agonet: mscc: ocelot: add missing VCAP ES0 and IS1 regmaps for VSC7514
Vladimir Oltean [Tue, 6 Oct 2020 15:02:48 +0000 (18:02 +0300)]
net: mscc: ocelot: add missing VCAP ES0 and IS1 regmaps for VSC7514

Without these definitions, the driver will crash in:
mscc_ocelot_probe
-> ocelot_init
   -> ocelot_vcap_init
     -> __ocelot_target_read_ix

I missed this because I did not have the VSC7514 hardware to test, only
the VSC9959 and VSC9953, and the probing part is different.

Fixes: e3aea296d86f ("net: mscc: ocelot: add definitions for VCAP ES0 keys, actions and target")
Fixes: a61e365d7c18 ("net: mscc: ocelot: add definitions for VCAP IS1 keys, actions and target")
Reported-by: Divya Koppera <Divya.Koppera@microchip.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agomptcp: fix infinite loop on recvmsg()/worker() race.
Paolo Abeni [Tue, 6 Oct 2020 06:27:34 +0000 (08:27 +0200)]
mptcp: fix infinite loop on recvmsg()/worker() race.

If recvmsg() and the workqueue race to dequeue the data
pending on some subflow, the current mapping for such
subflow covers several skbs and some of them have not
reached yet the received, either the worker or recvmsg()
can find a subflow with the data_avail flag set - since
the current mapping is valid and in sequence - but no
skbs in the receive queue - since the other entity just
processed them.

The above will lead to an unbounded loop in __mptcp_move_skbs()
and a subsequent hang of any task trying to acquiring the msk
socket lock.

This change addresses the issue stopping the __mptcp_move_skbs()
loop as soon as we detect the above race (empty receive queue
with data_avail set).

Reported-and-tested-by: syzbot+fcf8ca5817d6e92c6567@syzkaller.appspotmail.com
Fixes: ab174ad8ef76 ("mptcp: move ooo skbs into msk out of order queue.")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agocxgb4: convert tasklets to use new tasklet_setup() API
Allen Pais [Tue, 6 Oct 2020 05:53:32 +0000 (11:23 +0530)]
cxgb4: convert tasklets to use new tasklet_setup() API

In preparation for unconditionally passing the
struct tasklet_struct pointer to all tasklet
callbacks, switch to using the new tasklet_setup()
and from_tasklet() to pass the tasklet pointer explicitly.

Signed-off-by: Romain Perier <romain.perier@gmail.com>
Signed-off-by: Allen Pais <apais@linux.microsoft.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoethtool: correct policy for ETHTOOL_MSG_CHANNELS_SET
Johannes Berg [Wed, 7 Oct 2020 10:53:51 +0000 (12:53 +0200)]
ethtool: correct policy for ETHTOOL_MSG_CHANNELS_SET

This accidentally got wired up to the *get* policy instead
of the *set* policy, causing operations to be rejected. Fix
it by wiring up the correct policy instead.

Fixes: 5028588b62cb ("ethtool: wire up set policies to ops")
Reported-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Tested-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoethtool: strset: allow ETHTOOL_A_STRSET_COUNTS_ONLY attr
Johannes Berg [Wed, 7 Oct 2020 10:53:50 +0000 (12:53 +0200)]
ethtool: strset: allow ETHTOOL_A_STRSET_COUNTS_ONLY attr

The ETHTOOL_A_STRSET_COUNTS_ONLY flag attribute was previously
not allowed to be used, but now due to the policy size reduction
we would access the tb[] array out of bounds since we tried to
check for the attribute despite it not being accepted.

Fix both issues by adding it correctly to the appropriate policy.

Fixes: ff419afa4310 ("ethtool: trim policy tables")
Fixes: 71921690f974 ("ethtool: provide string sets with STRSET_GET request")
Reported-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Tested-by: Leon Romanovsky <leonro@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Jakub Kicinski [Thu, 8 Oct 2020 22:44:50 +0000 (15:44 -0700)]
Merge git://git./linux/kernel/git/netdev/net

Small conflict around locking in rxrpc_process_event() -
channel_lock moved to bundle in next, while state lock
needs _bh() from net.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
Linus Torvalds [Thu, 8 Oct 2020 21:25:46 +0000 (14:25 -0700)]
Merge tag 'for_linus' of git://git./linux/kernel/git/mst/vhost

Pull vhost fixes from Michael Tsirkin:
 "Some last minute vhost,vdpa fixes.

  The last two of them haven't been in next but they do seem kind of
  obvious, very small and safe, fix bugs reported in the field, and they
  are both in a new mlx5 vdpa driver, so it's not like we can introduce
  regressions"

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
  vdpa/mlx5: Fix dependency on MLX5_CORE
  vdpa/mlx5: should keep avail_index despite device status
  vhost-vdpa: fix page pinning leakage in error path
  vhost-vdpa: fix vhost_vdpa_map() on error condition
  vhost: Don't call log_access_ok() when using IOTLB
  vhost: Use vhost_get_used_size() in vhost_vring_set_addr()
  vhost: Don't call access_ok() when using IOTLB
  vhost vdpa: fix vhost_vdpa_open error handling

3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 8 Oct 2020 21:11:21 +0000 (14:11 -0700)]
Merge git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "One more set of fixes from the networking tree:

   - add missing input validation in nl80211_del_key(), preventing
     out-of-bounds access

   - last minute fix / improvement of a MRP netlink (uAPI) interface
     introduced in 5.9 (current) release

   - fix "unresolved symbol" build error under CONFIG_NET w/o
     CONFIG_INET due to missing tcp_timewait_sock and inet_timewait_sock
     BTF.

   - fix 32 bit sub-register bounds tracking in the bpf verifier for OR
     case

   - tcp: fix receive window update in tcp_add_backlog()

   - openvswitch: handle DNAT tuple collision in conntrack-related code

   - r8169: wait for potential PHY reset to finish after applying a FW
     file, avoiding unexpected PHY behaviour and failures later on

   - mscc: fix tail dropping watermarks for Ocelot switches

   - avoid use-after-free in macsec code after a call to the GRO layer

   - avoid use-after-free in sctp error paths

   - add a device id for Cellient MPL200 WWAN card

   - rxrpc fixes:
      - fix the xdr encoding of the contents read from an rxrpc key
      - fix a BUG() for a unsupported encoding type.
      - fix missing _bh lock annotations.
      - fix acceptance handling for an incoming call where the incoming
        call is encrypted.
      - the server token keyring isn't network namespaced - it belongs
        to the server, so there's no need. Namespacing it means that
        request_key() fails to find it.
      - fix a leak of the server keyring"

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (21 commits)
  net: usb: qmi_wwan: add Cellient MPL200 card
  macsec: avoid use-after-free in macsec_handle_frame()
  r8169: consider that PHY reset may still be in progress after applying firmware
  openvswitch: handle DNAT tuple collision
  sctp: fix sctp_auth_init_hmacs() error path
  bridge: Netlink interface fix.
  net: wireless: nl80211: fix out-of-bounds access in nl80211_del_key()
  bpf: Fix scalar32_min_max_or bounds tracking
  tcp: fix receive window update in tcp_add_backlog()
  net: usb: rtl8150: set random MAC address when set_ethernet_addr() fails
  mptcp: more DATA FIN fixes
  net: mscc: ocelot: warn when encoding an out-of-bounds watermark value
  net: mscc: ocelot: divide watermark value by 60 when writing to SYS_ATOP
  net: qrtr: ns: Fix the incorrect usage of rcu_read_lock()
  rxrpc: Fix server keyring leak
  rxrpc: The server keyring isn't network-namespaced
  rxrpc: Fix accept on a connection that need securing
  rxrpc: Fix some missing _bh annotations on locking conn->state_lock
  rxrpc: Downgrade the BUG() for unsupported token type in rxrpc_read()
  rxrpc: Fix rxkad token xdr encoding
  ...

3 years agovdpa/mlx5: Fix dependency on MLX5_CORE
Eli Cohen [Wed, 7 Oct 2020 06:40:11 +0000 (09:40 +0300)]
vdpa/mlx5: Fix dependency on MLX5_CORE

Remove propmt for selecting MLX5_VDPA by the user and modify
MLX5_VDPA_NET to select MLX5_VDPA. Also modify MLX5_VDPA_NET to depend
on mlx5_core.

This fixes an issue where configuration sets 'y' for MLX5_VDPA_NET while
MLX5_CORE is compiled as a module causing link errors.

Reported-by: kernel test robot <lkp@intel.com>
Fixes: 1a86b377aa21 ("vdpa/mlx5: Add VDPA driver for supported mlx5 device")s
Signed-off-by: Eli Cohen <elic@nvidia.com>
Link: https://lore.kernel.org/r/20201007064011.GA50074@mtl-vdi-166.wap.labs.mlnx
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
3 years agovdpa/mlx5: should keep avail_index despite device status
Si-Wei Liu [Thu, 1 Oct 2020 20:18:31 +0000 (16:18 -0400)]
vdpa/mlx5: should keep avail_index despite device status

A VM with mlx5 vDPA has below warnings while being reset:

vhost VQ 0 ring restore failed: -1: Resource temporarily unavailable (11)
vhost VQ 1 ring restore failed: -1: Resource temporarily unavailable (11)

We should allow userspace emulating the virtio device be
able to get to vq's avail_index, regardless of vDPA device
status. Save the index that was last seen when virtq was
stopped, so that userspace doesn't complain.

Signed-off-by: Si-Wei Liu <si-wei.liu@oracle.com>
Link: https://lore.kernel.org/r/1601583511-15138-1-git-send-email-si-wei.liu@oracle.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Eli Cohen <elic@nvidia.com>
3 years agonet: usb: qmi_wwan: add Cellient MPL200 card
Wilken Gottwalt [Thu, 8 Oct 2020 07:21:38 +0000 (09:21 +0200)]
net: usb: qmi_wwan: add Cellient MPL200 card

Add usb ids of the Cellient MPL200 card.

Signed-off-by: Wilken Gottwalt <wilken.gottwalt@mailbox.org>
Acked-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agomacsec: avoid use-after-free in macsec_handle_frame()
Eric Dumazet [Wed, 7 Oct 2020 08:42:46 +0000 (01:42 -0700)]
macsec: avoid use-after-free in macsec_handle_frame()

De-referencing skb after call to gro_cells_receive() is not allowed.
We need to fetch skb->len earlier.

Fixes: 5491e7c6b1a9 ("macsec: enable GRO and RPS on macsec devices")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agor8169: consider that PHY reset may still be in progress after applying firmware
Heiner Kallweit [Wed, 7 Oct 2020 11:34:51 +0000 (13:34 +0200)]
r8169: consider that PHY reset may still be in progress after applying firmware

Some firmware files trigger a PHY soft reset and don't wait for it to
be finished. PHY register writes directly after applying the firmware
may fail or provide unexpected results therefore. Fix this by waiting
for bit BMCR_RESET to be cleared after applying firmware.

There's nothing wrong with the referenced change, it's just that the
fix will apply cleanly only after this change.

Fixes: 89fbd26cca7e ("r8169: fix firmware not resetting tp->ocp_base")
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoopenvswitch: handle DNAT tuple collision
Dumitru Ceara [Wed, 7 Oct 2020 15:48:03 +0000 (17:48 +0200)]
openvswitch: handle DNAT tuple collision

With multiple DNAT rules it's possible that after destination
translation the resulting tuples collide.

For example, two openvswitch flows:
nw_dst=10.0.0.10,tp_dst=10, actions=ct(commit,table=2,nat(dst=20.0.0.1:20))
nw_dst=10.0.0.20,tp_dst=10, actions=ct(commit,table=2,nat(dst=20.0.0.1:20))

Assuming two TCP clients initiating the following connections:
10.0.0.10:5000->10.0.0.10:10
10.0.0.10:5000->10.0.0.20:10

Both tuples would translate to 10.0.0.10:5000->20.0.0.1:20 causing
nf_conntrack_confirm() to fail because of tuple collision.

Netfilter handles this case by allocating a null binding for SNAT at
egress by default.  Perform the same operation in openvswitch for DNAT
if no explicit SNAT is requested by the user and allocate a null binding
for SNAT for packets in the "original" direction.

Reported-at: https://bugzilla.redhat.com/1877128
Suggested-by: Florian Westphal <fw@strlen.de>
Fixes: 05752523e565 ("openvswitch: Interface with NAT.")
Signed-off-by: Dumitru Ceara <dceara@redhat.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agosctp: fix sctp_auth_init_hmacs() error path
Eric Dumazet [Thu, 8 Oct 2020 08:38:31 +0000 (01:38 -0700)]
sctp: fix sctp_auth_init_hmacs() error path

After freeing ep->auth_hmacs we have to clear the pointer
or risk use-after-free as reported by syzbot:

BUG: KASAN: use-after-free in sctp_auth_destroy_hmacs net/sctp/auth.c:509 [inline]
BUG: KASAN: use-after-free in sctp_auth_destroy_hmacs net/sctp/auth.c:501 [inline]
BUG: KASAN: use-after-free in sctp_auth_free+0x17e/0x1d0 net/sctp/auth.c:1070
Read of size 8 at addr ffff8880a8ff52c0 by task syz-executor941/6874

CPU: 0 PID: 6874 Comm: syz-executor941 Not tainted 5.9.0-rc8-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x198/0x1fd lib/dump_stack.c:118
 print_address_description.constprop.0.cold+0xae/0x497 mm/kasan/report.c:383
 __kasan_report mm/kasan/report.c:513 [inline]
 kasan_report.cold+0x1f/0x37 mm/kasan/report.c:530
 sctp_auth_destroy_hmacs net/sctp/auth.c:509 [inline]
 sctp_auth_destroy_hmacs net/sctp/auth.c:501 [inline]
 sctp_auth_free+0x17e/0x1d0 net/sctp/auth.c:1070
 sctp_endpoint_destroy+0x95/0x240 net/sctp/endpointola.c:203
 sctp_endpoint_put net/sctp/endpointola.c:236 [inline]
 sctp_endpoint_free+0xd6/0x110 net/sctp/endpointola.c:183
 sctp_destroy_sock+0x9c/0x3c0 net/sctp/socket.c:4981
 sctp_v6_destroy_sock+0x11/0x20 net/sctp/socket.c:9415
 sk_common_release+0x64/0x390 net/core/sock.c:3254
 sctp_close+0x4ce/0x8b0 net/sctp/socket.c:1533
 inet_release+0x12e/0x280 net/ipv4/af_inet.c:431
 inet6_release+0x4c/0x70 net/ipv6/af_inet6.c:475
 __sock_release+0xcd/0x280 net/socket.c:596
 sock_close+0x18/0x20 net/socket.c:1277
 __fput+0x285/0x920 fs/file_table.c:281
 task_work_run+0xdd/0x190 kernel/task_work.c:141
 exit_task_work include/linux/task_work.h:25 [inline]
 do_exit+0xb7d/0x29f0 kernel/exit.c:806
 do_group_exit+0x125/0x310 kernel/exit.c:903
 __do_sys_exit_group kernel/exit.c:914 [inline]
 __se_sys_exit_group kernel/exit.c:912 [inline]
 __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:912
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x43f278
Code: Bad RIP value.
RSP: 002b:00007fffe0995c38 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 000000000043f278
RDX: 0000000000000000 RSI: 000000000000003c RDI: 0000000000000000
RBP: 00000000004bf068 R08: 00000000000000e7 R09: ffffffffffffffd0
R10: 0000000020000000 R11: 0000000000000246 R12: 0000000000000001
R13: 00000000006d1180 R14: 0000000000000000 R15: 0000000000000000

Allocated by task 6874:
 kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48
 kasan_set_track mm/kasan/common.c:56 [inline]
 __kasan_kmalloc.constprop.0+0xbf/0xd0 mm/kasan/common.c:461
 kmem_cache_alloc_trace+0x174/0x300 mm/slab.c:3554
 kmalloc include/linux/slab.h:554 [inline]
 kmalloc_array include/linux/slab.h:593 [inline]
 kcalloc include/linux/slab.h:605 [inline]
 sctp_auth_init_hmacs+0xdb/0x3b0 net/sctp/auth.c:464
 sctp_auth_init+0x8a/0x4a0 net/sctp/auth.c:1049
 sctp_setsockopt_auth_supported net/sctp/socket.c:4354 [inline]
 sctp_setsockopt+0x477e/0x97f0 net/sctp/socket.c:4631
 __sys_setsockopt+0x2db/0x610 net/socket.c:2132
 __do_sys_setsockopt net/socket.c:2143 [inline]
 __se_sys_setsockopt net/socket.c:2140 [inline]
 __x64_sys_setsockopt+0xba/0x150 net/socket.c:2140
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Freed by task 6874:
 kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48
 kasan_set_track+0x1c/0x30 mm/kasan/common.c:56
 kasan_set_free_info+0x1b/0x30 mm/kasan/generic.c:355
 __kasan_slab_free+0xd8/0x120 mm/kasan/common.c:422
 __cache_free mm/slab.c:3422 [inline]
 kfree+0x10e/0x2b0 mm/slab.c:3760
 sctp_auth_destroy_hmacs net/sctp/auth.c:511 [inline]
 sctp_auth_destroy_hmacs net/sctp/auth.c:501 [inline]
 sctp_auth_init_hmacs net/sctp/auth.c:496 [inline]
 sctp_auth_init_hmacs+0x2b7/0x3b0 net/sctp/auth.c:454
 sctp_auth_init+0x8a/0x4a0 net/sctp/auth.c:1049
 sctp_setsockopt_auth_supported net/sctp/socket.c:4354 [inline]
 sctp_setsockopt+0x477e/0x97f0 net/sctp/socket.c:4631
 __sys_setsockopt+0x2db/0x610 net/socket.c:2132
 __do_sys_setsockopt net/socket.c:2143 [inline]
 __se_sys_setsockopt net/socket.c:2140 [inline]
 __x64_sys_setsockopt+0xba/0x150 net/socket.c:2140
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Fixes: 1f485649f529 ("[SCTP]: Implement SCTP-AUTH internals")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Vlad Yasevich <vyasevich@gmail.com>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'mac80211-for-net-2020-10-08' of git://git.kernel.org/pub/scm/linux/kernel...
Jakub Kicinski [Thu, 8 Oct 2020 19:18:34 +0000 (12:18 -0700)]
Merge tag 'mac80211-for-net-2020-10-08' of git://git./linux/kernel/git/jberg/mac80211

Johannes Berg says:

====================
pull-request: mac80211 2020-10-08

A single fix for missing input validation in nl80211.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Jakub Kicinski [Thu, 8 Oct 2020 19:05:37 +0000 (12:05 -0700)]
Merge git://git./pub/scm/linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2020-10-08

The main changes are:

1) Fix "unresolved symbol" build error under CONFIG_NET w/o CONFIG_INET due
   to missing tcp_timewait_sock and inet_timewait_sock BTF, from Yonghong Song.

2) Fix 32 bit sub-register bounds tracking for OR case, from Daniel Borkmann.
====================

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agobridge: Netlink interface fix.
Henrik Bjoernlund [Wed, 7 Oct 2020 12:07:00 +0000 (12:07 +0000)]
bridge: Netlink interface fix.

This commit is correcting NETLINK br_fill_ifinfo() to be able to
handle 'filter_mask' with multiple flags asserted.

Fixes: 36a8e8e265420 ("bridge: Extend br_fill_ifinfo to return MPR status")

Signed-off-by: Henrik Bjoernlund <henrik.bjoernlund@microchip.com>
Reviewed-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Suggested-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Tested-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
3 years agoMerge tag 'drm-fixes-2020-10-08' of git://anongit.freedesktop.org/drm/drm
Linus Torvalds [Thu, 8 Oct 2020 18:14:17 +0000 (11:14 -0700)]
Merge tag 'drm-fixes-2020-10-08' of git://anongit.freedesktop.org/drm/drm

Pull drm nouveau fixes from Dave Airlie:
 "Karol found two last minute nouveau fixes, they both fix crashes, the
  TTM one follows what other drivers do already, and the other is for
  bailing on load on unrecognised chipsets.

   - fix crash in TTM alloc fail path

   - return error earlier for unknown chipsets"

* tag 'drm-fixes-2020-10-08' of git://anongit.freedesktop.org/drm/drm:
  drm/nouveau/mem: guard against NULL pointer access in mem_del
  drm/nouveau/device: return error for unknown chipsets

3 years agoMerge tag 'exfat-for-5.9-rc9' of git://git.kernel.org/pub/scm/linux/kernel/git/linkin...
Linus Torvalds [Thu, 8 Oct 2020 18:10:13 +0000 (11:10 -0700)]
Merge tag 'exfat-for-5.9-rc9' of git://git./linux/kernel/git/linkinjeon/exfat

Pull exfat fixes from Namjae Jeon:

 - Fix use of uninitialized spinlock on error path

 - Fix missing err assignment in exfat_build_inode()

* tag 'exfat-for-5.9-rc9' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfat:
  exfat: fix use of uninitialized spinlock on error path
  exfat: fix pointer error checking

3 years agoMerge tag 'for-linus-5.9b-rc9-tag' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 8 Oct 2020 18:01:53 +0000 (11:01 -0700)]
Merge tag 'for-linus-5.9b-rc9-tag' of git://git./linux/kernel/git/xen/tip

Pull xen fix from Juergen Gross:
 "One fix for a regression when booting as a Xen guest on ARM64
  introduced probably during the 5.9 cycle. It is very low risk as it is
  modifying Xen specific code only.

  The exact commit introducing the bug hasn't been identified yet, but
  everything was fine in 5.8 and only in 5.9 some configurations started
  to fail"

* tag 'for-linus-5.9b-rc9-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  arm/arm64: xen: Fix to convert percpu address to gfn correctly

3 years agoafs: Fix deadlock between writeback and truncate
David Howells [Wed, 7 Oct 2020 13:22:12 +0000 (14:22 +0100)]
afs: Fix deadlock between writeback and truncate

The afs filesystem has a lock[*] that it uses to serialise I/O operations
going to the server (vnode->io_lock), as the server will only perform one
modification operation at a time on any given file or directory.  This
prevents the the filesystem from filling up all the call slots to a server
with calls that aren't going to be executed in parallel anyway, thereby
allowing operations on other files to obtain slots.

  [*] Note that is probably redundant for directories at least since
      i_rwsem is used to serialise directory modifications and
      lookup/reading vs modification.  The server does allow parallel
      non-modification ops, however.

When a file truncation op completes, we truncate the in-memory copy of the
file to match - but we do it whilst still holding the io_lock, the idea
being to prevent races with other operations.

However, if writeback starts in a worker thread simultaneously with
truncation (whilst notify_change() is called with i_rwsem locked, writeback
pays it no heed), it may manage to set PG_writeback bits on the pages that
will get truncated before afs_setattr_success() manages to call
truncate_pagecache().  Truncate will then wait for those pages - whilst
still inside io_lock:

    # cat /proc/8837/stack
    [<0>] wait_on_page_bit_common+0x184/0x1e7
    [<0>] truncate_inode_pages_range+0x37f/0x3eb
    [<0>] truncate_pagecache+0x3c/0x53
    [<0>] afs_setattr_success+0x4d/0x6e
    [<0>] afs_wait_for_operation+0xd8/0x169
    [<0>] afs_do_sync_operation+0x16/0x1f
    [<0>] afs_setattr+0x1fb/0x25d
    [<0>] notify_change+0x2cf/0x3c4
    [<0>] do_truncate+0x7f/0xb2
    [<0>] do_sys_ftruncate+0xd1/0x104
    [<0>] do_syscall_64+0x2d/0x3a
    [<0>] entry_SYSCALL_64_after_hwframe+0x44/0xa9

The writeback operation, however, stalls indefinitely because it needs to
get the io_lock to proceed:

    # cat /proc/5940/stack
    [<0>] afs_get_io_locks+0x58/0x1ae
    [<0>] afs_begin_vnode_operation+0xc7/0xd1
    [<0>] afs_store_data+0x1b2/0x2a3
    [<0>] afs_write_back_from_locked_page+0x418/0x57c
    [<0>] afs_writepages_region+0x196/0x224
    [<0>] afs_writepages+0x74/0x156
    [<0>] do_writepages+0x2d/0x56
    [<0>] __writeback_single_inode+0x84/0x207
    [<0>] writeback_sb_inodes+0x238/0x3cf
    [<0>] __writeback_inodes_wb+0x68/0x9f
    [<0>] wb_writeback+0x145/0x26c
    [<0>] wb_do_writeback+0x16a/0x194
    [<0>] wb_workfn+0x74/0x177
    [<0>] process_one_work+0x174/0x264
    [<0>] worker_thread+0x117/0x1b9
    [<0>] kthread+0xec/0xf1
    [<0>] ret_from_fork+0x1f/0x30

and thus deadlock has occurred.

Note that whilst afs_setattr() calls filemap_write_and_wait(), the fact
that the caller is holding i_rwsem doesn't preclude more pages being
dirtied through an mmap'd region.

Fix this by:

 (1) Use the vnode validate_lock to mediate access between afs_setattr()
     and afs_writepages():

     (a) Exclusively lock validate_lock in afs_setattr() around the whole
       RPC operation.

     (b) If WB_SYNC_ALL isn't set on entry to afs_writepages(), trying to
       shared-lock validate_lock and returning immediately if we couldn't
       get it.

     (c) If WB_SYNC_ALL is set, wait for the lock.

     The validate_lock is also used to validate a file and to zap its cache
     if the file was altered by a third party, so it's probably a good fit
     for this.

 (2) Move the truncation outside of the io_lock in setattr, using the same
     hook as is used for local directory editing.

     This requires the old i_size to be retained in the operation record as
     we commit the revised status to the inode members inside the io_lock
     still, but we still need to know if we reduced the file size.

Fixes: d2ddc776a458 ("afs: Overhaul volume and server record caching and fileserver rotation")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoiwlwifi: bump FW API to 59 for AX devices
Luca Coelho [Thu, 8 Oct 2020 15:12:50 +0000 (18:12 +0300)]
iwlwifi: bump FW API to 59 for AX devices

Start supporting API version 59 for AX devices.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201008181047.0359498107fe.I559b5140be6abbee5eb759ec6e5ca046f7178f8f@changeid
3 years agoiwlwifi: read and parse PNVM file
Luca Coelho [Thu, 8 Oct 2020 15:12:49 +0000 (18:12 +0300)]
iwlwifi: read and parse PNVM file

The driver looks for a PNVM file that contains FW configuration data
for each different HW combination.  The FW requests the data for a
certain SKU_ID and the driver tries to find it in the PNVM file.

Read the file, parse its contents and send it to the trans.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201008181047.826bc607e57a.I1d93dd6e6651586878db57fac3e7c3f09d742c42@changeid
3 years agoiwlwifi: pcie: implement set_pnvm op
Luca Coelho [Thu, 8 Oct 2020 15:12:48 +0000 (18:12 +0300)]
iwlwifi: pcie: implement set_pnvm op

Implement the set_pnvm op to store the PNVM settings to the context
info and the corresponding code to free the DRAM block when the
context is freed.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201008181047.85847cfb0972.I202d90e99779f722df14b2d4102d3e466343a6f6@changeid
3 years agoiwlwifi: add trans op to set PNVM
Luca Coelho [Thu, 8 Oct 2020 15:12:47 +0000 (18:12 +0300)]
iwlwifi: add trans op to set PNVM

Add an op to allow setting the PNVM data in the trans.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201008181047.77300b00d3ce.I89d7e45e71f19fd36d89f363a28b4759b59b8765@changeid
3 years agoiwlwifi: move PNVM implementation to common code
Luca Coelho [Thu, 8 Oct 2020 15:12:46 +0000 (18:12 +0300)]
iwlwifi: move PNVM implementation to common code

The PNVM code is generic and can be used by other opmodes.  Move it to
a common file and include it in the relevant opmodes.

Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/iwlwifi.20201008181047.232aa310693b.I03a18ffa4162753af38e759d88e27509007c7bca@changeid