Marc Kleine-Budde [Thu, 19 Nov 2020 10:09:17 +0000 (11:09 +0100)]
can: flexcan: flexcan_close(): change order if commands to properly shut down the controller
There haven been reports, that the flexcan_close() soradically hangs during
simultanious ifdown, sending of CAN messages and probably open CAN bus:
| (__schedule) from [<
808bbd34>] (schedule+0x90/0xb8)
| (schedule) from [<
808bf274>] (schedule_timeout+0x1f8/0x24c)
| (schedule_timeout) from [<
8016be44>] (msleep+0x18/0x1c)
| (msleep) from [<
80746a64>] (napi_disable+0x60/0x70)
| (napi_disable) from [<
8052fdd0>] (flexcan_close+0x2c/0x140)
| (flexcan_close) from [<
80744930>] (__dev_close_many+0xb8/0xd8)
| (__dev_close_many) from [<
8074db9c>] (__dev_change_flags+0xd0/0x1a0)
| (__dev_change_flags) from [<
8074dc84>] (dev_change_flags+0x18/0x48)
| (dev_change_flags) from [<
80760c24>] (do_setlink+0x44c/0x7b4)
| (do_setlink) from [<
80761560>] (rtnl_newlink+0x374/0x68c)
I was unable to reproduce the issue, but a cleanup of the flexcan close
sequence has probably fixed the problem at the reporting user.
This patch changes the sequence in flexcan_close() to:
- stop the TX queue
- disable the interrupts on the chip level and wait via free_irq()
synchronously for the interrupt handler to finish
- disable RX offload, which disables synchronously NAPI
- disable the flexcan on the chip level
- free RX offload
- disable the transceiver
- close the CAN device
- disable the clocks
Link: https://lore.kernel.org/r/20201119100917.3013281-6-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Marc Kleine-Budde [Thu, 19 Nov 2020 10:09:16 +0000 (11:09 +0100)]
can: flexcan: flexcan_open(): completely initialize controller before requesting IRQ
This patch changes the order in which the flexcan controller is brought up
during flexcan_open(). It makes sure that the chip is completely initialized
before the IRQs are requested and finally enabled.
Link: https://lore.kernel.org/r/20201119100917.3013281-5-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Marc Kleine-Budde [Thu, 19 Nov 2020 10:09:15 +0000 (11:09 +0100)]
can: flexcan: flexcan_rx_offload_setup(): factor out mailbox and rx-offload setup into separate function
In an upcoming patch the order of operations in flexcan_open() are changed.
Introduce convenience function to make that patch simpler.
Link: https://lore.kernel.org/r/20201119100917.3013281-4-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Marc Kleine-Budde [Thu, 19 Nov 2020 10:09:14 +0000 (11:09 +0100)]
can: flexcan: move enabling/disabling of interrupts from flexcan_chip_{start,stop}() to callers
The function flexcan_chip_start() first configures the CAN controller and then
enables the interrupt, flexcan_chip_stop() does the opposite.
In an upcoming patch the order of operations in flexcan_open() and
flexcan_close() are changed. This requires
flexcan_chip_start()/flexcan_chip_stop_disable_on_error() and
flexcan_chip_interrupts_{enable,disable}() to be independent of each other.
This patch moves the enabling of the interrupts from flexcan_chip_start() to
its callers flexcan_open() and flexcan_resume(). Likewise the disabling of the
interrupts is moved from __flexcan_chip_stop() to its indirect callers
flexcan_close() and flexcan_suspend().
Link: https://lore.kernel.org/r/20201119100917.3013281-3-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Marc Kleine-Budde [Thu, 19 Nov 2020 10:09:13 +0000 (11:09 +0100)]
can: flexcan: factor out enabling and disabling of interrupts into separate function
The upcoming patches are going to move the enabling and disabling of the
interrupts. Introduce convenience functions to make these patches simpler.
Link: https://lore.kernel.org/r/20201119100917.3013281-2-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Joakim Zhang [Fri, 6 Nov 2020 10:56:25 +0000 (18:56 +0800)]
can: flexcan: rename macro FLEXCAN_QUIRK_SETUP_STOP_MODE -> FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR
This patch intends to rename FLEXCAN_QUIRK_SETUP_STOP_MODE quirk
to FLEXCAN_QUIRK_SETUP_STOP_MODE_GRP for non-scu SoCs, coming patch will
add quirk for scu SoCs.
For non-scu SoCs, setup stop mode with GPR register.
For scu SoCs, setup stop mode with SCU firmware.
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20201106105627.31061-4-qiangqing.zhang@nxp.com
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Joakim Zhang [Fri, 6 Nov 2020 10:56:26 +0000 (18:56 +0800)]
dt-bindings: firmware: add IMX_SC_R_CAN(x) macro for CAN
Add IMX_SC_R_CAN(x) macro for CAN.
Suggested-by: Marc Kleine-Budde <mkl@pengutronix.de>
Acked-by: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20201106105627.31061-5-qiangqing.zhang@nxp.com
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Joakim Zhang [Fri, 6 Nov 2020 10:56:23 +0000 (18:56 +0800)]
dt-bindings: can: fsl,flexcan: fix fsl,clk-source property
Correct fsl,clk-source example since flexcan driver uses "of_property_read_u8"
to get this property.
Fixes:
9d733992772d ("dt-bindings: can: flexcan: add PE clock source property to device tree")
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Link: https://lore.kernel.org/r/20201106105627.31061-2-qiangqing.zhang@nxp.com
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Marc Kleine-Budde [Wed, 11 Nov 2020 08:08:28 +0000 (09:08 +0100)]
dt-bindings: can: fsl,flexcan: add uint32 reference to clock-frequency property
This patch adds the missing reference to the clock-frequency property. The
driver uses a of_property_read_u32() to read the property to mark it as uint32.
Fixes:
e5ab9aa7e49b ("dt-bindings: can: flexcan: convert fsl,*flexcan bindings to yaml")
Link: https://lore.kernel.org/r/20201119073357.2858925-1-mkl@pengutronix.de
Cc: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: Oleksij Rempel <ore@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Thu, 19 Nov 2020 08:49:21 +0000 (09:49 +0100)]
can: gw: support modification of Classical CAN DLCs
Add support for data length code modifications for Classical CAN.
The netlink configuration interface always allowed to pass any value
that fits into a byte, therefore only the modification process had to be
extended to handle the raw DLC represenation of Classical CAN frames.
When a DLC value from 0 .. F is provided for Classical CAN frame
modifications the 'len' value is modified as-is with the exception that
potentially existing 9 .. F DLC values in the len8_dlc element are moved
to the 'len' element for the modification operation by mod_retrieve_ccdlc().
After the modification the Classical CAN frame DLC information is brought
back into the correct format by mod_store_ccdlc() which is filling 'len'
and 'len8_dlc' accordingly.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201119084921.2621-1-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Stefan Mätje [Mon, 16 Nov 2020 18:44:30 +0000 (19:44 +0100)]
can: drivers: add len8_dlc support for esd_usb2 CAN adapter
Support the Classical CAN raw DLC functionality to send and receive DLC values
from 9 .. 15 for the Classical CAN capable CAN network driver esd_usb that
supports the esd CAN-USB/2 and CAN-USB/Micro devices:
- esd_usb2
Signed-off-by: Stefan Mätje <stefan.maetje@esd.eu>
Tested-by: Stefan Mätje <stefan.maetje@esd.eu>
Link: https://lore.kernel.org/r/20201116184430.25462-2-stefan.maetje@esd.eu
[mkl: rewrapped some long lines]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Wed, 11 Nov 2020 09:59:23 +0000 (10:59 +0100)]
can: drivers: add len8_dlc support for various CAN adapters
Support the Classical CAN raw DLC functionality to send and receive DLC values
from 9 .. 15 on various Classical CAN capable CAN network drivers:
- sja1000
- gs_usb
- pcan_usb
- pcan_usb_fd
- usb_8dev
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Tested-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201111095923.2535-1-socketcan@hartkopp.net
[mkl: usb_8dev: changed indention]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 15:49:12 +0000 (16:49 +0100)]
can: drivers: introduce helpers to access Classical CAN DLC values
This patch adds the following helper to functions to access Classical CAN DLC
values.
can_get_cc_dlc(): get the data length code for Classical CAN raw DLC access
can_frame_set_cc_len(): set len and len8_dlc value for Classical CAN raw DLC access
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110154913.1404582-2-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 10:18:50 +0000 (11:18 +0100)]
can: update documentation for DLC usage in Classical CAN
The extension of struct can_frame with the len8_dlc element and the
can_dlc naming issue required an update of the documentation.
Additionally introduce the term 'Classical CAN' which has been established
by CAN in Automation to separate the original CAN2.0 A/B from CAN FD.
Updated some data structures and flags.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110101852.1973-7-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 10:18:49 +0000 (11:18 +0100)]
can: rename CAN FD related can_len2dlc and can_dlc2len helpers
The helper functions can_len2dlc and can_dlc2len are only relevant for
CAN FD data length code (DLC) conversion.
To fit the introduced can_cc_dlc2len for Classical CAN we rename:
can_dlc2len -> can_fd_dlc2len to get the payload length from the DLC
can_len2dlc -> can_fd_len2dlc to get the DLC from the payload length
Suggested-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110101852.1973-6-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Fri, 20 Nov 2020 10:04:44 +0000 (11:04 +0100)]
can: replace can_dlc as variable/element for payload length
The naming of can_dlc as element of struct can_frame and also as variable
name is misleading as it claims to be a 'data length CODE' but in reality
it always was a plain data length.
With the indroduction of a new 'len' element in struct can_frame we can now
remove can_dlc as name and make clear which of the former uses was a plain
length (-> 'len') or a data length code (-> 'dlc') value.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201120100444.3199-1-socketcan@hartkopp.net
[mkl: gs_usb: keep struct gs_host_frame::can_dlc as is]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 10:18:47 +0000 (11:18 +0100)]
can: remove obsolete get_canfd_dlc() macro
The macro was always used together with can_dlc2len() which sanitizes the
given dlc value on its own.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110101852.1973-4-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 10:18:46 +0000 (11:18 +0100)]
can: rename get_can_dlc() macro with can_cc_dlc2len()
The get_can_dlc() macro is used to ensure the payload length information of
the Classical CAN frame to be max 8 bytes (the CAN_MAX_DLEN).
Rename the macro and use the correct constant in preparation of the len/dlc
cleanup for Classical CAN frames.
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110101852.1973-3-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oliver Hartkopp [Tue, 10 Nov 2020 10:18:45 +0000 (11:18 +0100)]
can: add optional DLC element to Classical CAN frame structure
ISO 11898-1 Chapter 8.4.2.3 defines a 4 bit data length code (DLC) table which
maps the DLC to the payload length of the CAN frame in bytes:
DLC -> payload length
0 .. 8 -> 0 .. 8
9 .. 15 -> 8
Although the DLC values 8 .. 15 in Classical CAN always result in a payload
length of 8 bytes these DLC values are transparently transmitted on the CAN
bus. As the struct can_frame only provides a 'len' element (formerly 'can_dlc')
which contains the plain payload length ( 0 .. 8 ) of the CAN frame, the raw
DLC is not visible to the application programmer, e.g. for testing use-cases.
To access the raw DLC values 9 .. 15 the len8_dlc element is introduced, which
is only valid when the payload length 'len' is 8 and the DLC is greater than 8.
The len8_dlc element is filled by the CAN interface driver and used for CAN
frame creation by the CAN driver when the CAN_CTRLMODE_CC_LEN8_DLC flag is
supported by the driver and enabled via netlink configuration interface.
Reported-by: Vincent Mailhol <mailhol.vincent@wanadoo.fr>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
Link: https://lore.kernel.org/r/20201110101852.1973-2-socketcan@hartkopp.net
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Yegor Yefremov [Wed, 4 Nov 2020 15:57:30 +0000 (16:57 +0100)]
can: j1939: add tables for the CAN identifier and its fields
Use table markup to show the structure of the CAN identifier, PGN, PDU1, and
PDU2 formats. Also add introductory sentence.
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Link: https://lore.kernel.org/r/20201104155730.25196-1-yegorslists@googlemail.com
[mkl: removed trailing whitespace]
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Jakub Kicinski [Fri, 20 Nov 2020 06:05:44 +0000 (22:05 -0800)]
Merge branch 'enetc-clean-endianness-warnings-up'
Claudiu Manoil says:
====================
enetc: Clean endianness warnings up
Cleanup patches to address the outstanding endianness issues
in the driver reported by sparse.
====================
Link: https://lore.kernel.org/r/20201119101215.19223-1-claudiu.manoil@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Claudiu Manoil [Thu, 19 Nov 2020 10:12:15 +0000 (12:12 +0200)]
enetc: Fix endianness issues for enetc_qos
Currently the control buffer descriptor (cbd) fields have endianness
restrictions while the commands passed into the control buffers
don't (with one exception). This patch fixes offending code,
by adding endianness accessors for cbd fields and removing the
unnecessary ones in case of data buffer fields. Currently there's
no need to convert all commands to little endian format, the patch
only focuses on fixing current endianness issues reported by sparse.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Claudiu Manoil [Thu, 19 Nov 2020 10:12:14 +0000 (12:12 +0200)]
enetc: Fix endianness issues for enetc_ethtool
These particular fields are specified in the H/W reference
manual as having network byte order format, so enforce big
endian annotation for them and clear the related sparse
warnings in the process.
Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Oliver Herms [Wed, 18 Nov 2020 23:06:51 +0000 (00:06 +0100)]
IPv6: RTM_GETROUTE: Add RTA_ENCAP to result
This patch adds an IPv6 routes encapsulation attribute
to the result of netlink RTM_GETROUTE requests
(i.e. ip route get 2001:db8::).
Signed-off-by: Oliver Herms <oliver.peter.herms@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20201118230651.GA8861@tws
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paolo Abeni [Wed, 18 Nov 2020 22:05:34 +0000 (23:05 +0100)]
mptcp: update rtx timeout only if required.
We must start the retransmission timer only there are
pending data in the rtx queue.
Otherwise we can hit a WARN_ON in mptcp_reset_timer(),
as syzbot demonstrated.
Reported-and-tested-by: syzbot+42aa53dafb66a07e5a24@syzkaller.appspotmail.com
Fixes:
d9ca1de8c0cd ("mptcp: move page frag allocation in mptcp_sendmsg()")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Mat Martineau <mathew.j.martineau@linux.intel.com>
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Tested-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Link: https://lore.kernel.org/r/1a72039f112cae048c44d398ffa14e0a1432db3d.1605737083.git.pabeni@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Fri, 20 Nov 2020 05:41:08 +0000 (21:41 -0800)]
Merge branch 'devlink-move-common-flash_update-calls-to-core'
Jacob Keller says:
====================
devlink: move common flash_update calls to core
This series moves a couple common pieces done by all drivers of the
->flash_update interface into devlink.c flash update handler. Specifically,
the core code will now request_firmware and
devlink_flash_update_(begin|end)_notify.
This cleanup is intended to simplify driver implementations so that they
have less work to do and are less capable of doing the "wrong" thing.
For request_firmware, this simplification is done as it is not expected that
drivers would do anything else. It also standardizes all drivers so that
they use the same interface (request_firmware, as opposed to
request_firmware_direct), and allows reporting the netlink extended ack with
the file name attribute.
For status notification, this change prevents drivers from sending a status
message without properly sending the status end notification. The current
userspace implementation of devlink relies on this end notification to
properly close the flash update channel. Without this, the flash update
process may hang indefinitely. By moving the begin and end calls into the
core code, it is no longer possible for a driver author to get this wrong.
Changes since v3
* picked up acked-by and reviewed-by comments
* fixed the ionic driver to leave the print statement in place
For the original patch that moved request_firmware, see [1]. For the v2 see
[2]. For further discussion of the issues with devlink flash status see [3].
For v3 see [4].
[1] https://lore.kernel.org/netdev/
20201113000142.3563690-1-jacob.e.keller@intel.com/
[2] https://lore.kernel.org/netdev/
20201113224559.3910864-1-jacob.e.keller@intel.com/
[3] https://lore.kernel.org/netdev/
6352e9d3-02af-721e-3a54-
ef99a666be29@intel.com/
[4] https://lore.kernel.org/netdev/
20201117200820.854115-1-jacob.e.keller@intel.com/
====================
Link: https://lore.kernel.org/r/20201118190636.1235045-1-jacob.e.keller@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jacob Keller [Wed, 18 Nov 2020 19:06:36 +0000 (11:06 -0800)]
devlink: move flash end and begin to core devlink
When performing a flash update via devlink, device drivers may inform
user space of status updates via
devlink_flash_update_(begin|end|timeout|status)_notify functions.
It is expected that drivers do not send any status notifications unless
they send a begin and end message. If a driver sends a status
notification without sending the appropriate end notification upon
finishing (regardless of success or failure), the current implementation
of the devlink userspace program can get stuck endlessly waiting for the
end notification that will never come.
The current ice driver implementation may send such a status message
without the appropriate end notification in rare cases.
Fixing the ice driver is relatively simple: we just need to send the
begin_notify at the start of the function and always send an end_notify
no matter how the function exits.
Rather than assuming driver authors will always get this right in the
future, lets just fix the API so that it is not possible to get wrong.
Make devlink_flash_update_begin_notify and
devlink_flash_update_end_notify static, and call them in devlink.c core
code. Always send the begin_notify just before calling the driver's
flash_update routine. Always send the end_notify just after the routine
returns regardless of success or failure.
Doing this makes the status notification easier to use from the driver,
as it no longer needs to worry about catching failures and cleaning up
by calling devlink_flash_update_end_notify. It is now no longer possible
to do the wrong thing in this regard. We also save a couple of lines of
code in each driver.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Acked-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jacob Keller [Wed, 18 Nov 2020 19:06:35 +0000 (11:06 -0800)]
devlink: move request_firmware out of driver
All drivers which implement the devlink flash update support, with the
exception of netdevsim, use either request_firmware or
request_firmware_direct to locate the firmware file. Rather than having
each driver do this separately as part of its .flash_update
implementation, perform the request_firmware within net/core/devlink.c
Replace the file_name parameter in the struct devlink_flash_update_params
with a pointer to the fw object.
Use request_firmware rather than request_firmware_direct. Although most
Linux distributions today do not have the fallback mechanism
implemented, only about half the drivers used the _direct request, as
compared to the generic request_firmware. In the event that
a distribution does support the fallback mechanism, the devlink flash
update ought to be able to use it to provide the firmware contents. For
distributions which do not support the fallback userspace mechanism,
there should be essentially no difference between request_firmware and
request_firmware_direct.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Acked-by: Shannon Nelson <snelson@pensando.io>
Acked-by: Vasundhara Volam <vasundhara-v.volam@broadcom.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Fri, 20 Nov 2020 03:08:46 +0000 (19:08 -0800)]
Merge https://git./linux/kernel/git/netdev/net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Thu, 19 Nov 2020 21:33:16 +0000 (13:33 -0800)]
Merge tag 'net-5.10-rc5' of git://git./linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Networking fixes for 5.10-rc5, including fixes from the WiFi
(mac80211), can and bpf (including the strncpy_from_user fix).
Current release - regressions:
- mac80211: fix memory leak of filtered powersave frames
- mac80211: free sta in sta_info_insert_finish() on errors to avoid
sleeping in atomic context
- netlabel: fix an uninitialized variable warning added in -rc4
Previous release - regressions:
- vsock: forward all packets to the host when no H2G is registered,
un-breaking AWS Nitro Enclaves
- net: Exempt multicast addresses from five-second neighbor lifetime
requirement, decreasing the chances neighbor tables fill up
- net/tls: fix corrupted data in recvmsg
- qed: fix ILT configuration of SRC block
- can: m_can: process interrupt only when not runtime suspended
Previous release - always broken:
- page_frag: Recover from memory pressure by not recycling pages
allocating from the reserves
- strncpy_from_user: Mask out bytes after NUL terminator
- ip_tunnels: Set tunnel option flag only when tunnel metadata is
present, always setting it confuses Open vSwitch
- bpf, sockmap:
- Fix partial copy_page_to_iter so progress can still be made
- Fix socket memory accounting and obeying SO_RCVBUF
- net: Have netpoll bring-up DSA management interface
- net: bridge: add missing counters to ndo_get_stats64 callback
- tcp: brr: only postpone PROBE_RTT if RTT is < current min_rtt
- enetc: Workaround MDIO register access HW bug
- net/ncsi: move netlink family registration to a subsystem init,
instead of tying it to driver probe
- net: ftgmac100: unregister NC-SI when removing driver to avoid
crash
- lan743x:
- prevent interrupt storm on open
- fix freeing skbs in the wrong context
- net/mlx5e: Fix socket refcount leak on kTLS RX resync
- net: dsa: mv88e6xxx: Avoid VLAN database corruption on 6097
- fix 21 unset return codes and other mistakes on error paths, mostly
detected by the Hulk Robot"
* tag 'net-5.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (115 commits)
fail_function: Remove a redundant mutex unlock
selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL
lib/strncpy_from_user.c: Mask out bytes after NUL terminator.
net/smc: fix direct access to ib_gid_addr->ndev in smc_ib_determine_gid()
net/smc: fix matching of existing link groups
ipv6: Remove dependency of ipv6_frag_thdr_truncated on ipv6 module
libbpf: Fix VERSIONED_SYM_COUNT number parsing
net/mlx4_core: Fix init_hca fields offset
atm: nicstar: Unmap DMA on send error
page_frag: Recover from memory pressure
net: dsa: mv88e6xxx: Wait for EEPROM done after HW reset
mlxsw: core: Use variable timeout for EMAD retries
mlxsw: Fix firmware flashing
net: Have netpoll bring-up DSA management interface
atl1e: fix error return code in atl1e_probe()
atl1c: fix error return code in atl1c_probe()
ah6: fix error return code in ah6_input()
net: usb: qmi_wwan: Set DTR quirk for MR400
can: m_can: process interrupt only when not runtime suspended
can: flexcan: flexcan_chip_start(): fix erroneous flexcan_transceiver_enable() during bus-off recovery
...
Linus Torvalds [Thu, 19 Nov 2020 21:01:53 +0000 (13:01 -0800)]
Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma
Pull rdma fixes from Jason Gunthorpe:
"The last two weeks have been quiet here, just the usual smattering of
long standing bug fixes.
A collection of error case bug fixes:
- Improper nesting of spinlock types in cm
- Missing error codes and kfree()
- Ensure dma_virt_ops users have the right kconfig symbols to work
properly
- Compilation failure of tools/testing"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
tools/testing/scatterlist: Fix test to compile and run
IB/hfi1: Fix error return code in hfi1_init_dd()
RMDA/sw: Don't allow drivers using dma_virt_ops on highmem configs
RDMA/pvrdma: Fix missing kfree() in pvrdma_register_device()
RDMA/cm: Make the local_id_table xarray non-irq
Jakub Kicinski [Thu, 19 Nov 2020 20:26:09 +0000 (12:26 -0800)]
Merge https://git./linux/kernel/git/bpf/bpf
Alexei Starovoitov says:
====================
1) libbpf should not attempt to load unused subprogs, from Andrii.
2) Make strncpy_from_user() mask out bytes after NUL terminator, from Daniel.
3) Relax return code check for subprograms in the BPF verifier, from Dmitrii.
4) Fix several sockmap issues, from John.
* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
fail_function: Remove a redundant mutex unlock
selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL
lib/strncpy_from_user.c: Mask out bytes after NUL terminator.
libbpf: Fix VERSIONED_SYM_COUNT number parsing
bpf, sockmap: Avoid failures from skb_to_sgvec when skb has frag_list
bpf, sockmap: Handle memory acct if skb_verdict prog redirects to self
bpf, sockmap: Avoid returning unneeded EAGAIN when redirecting to self
bpf, sockmap: Use truesize with sk_rmem_schedule()
bpf, sockmap: Ensure SO_RCVBUF memory is observed on ingress redirect
bpf, sockmap: Fix partial copy_page_to_iter so progress can still be made
selftests/bpf: Fix error return code in run_getsockopt_test()
bpf: Relax return code check for subprograms
tools, bpftool: Add missing close before bpftool net attach exit
MAINTAINERS/bpf: Update Andrii's entry.
selftests/bpf: Fix unused attribute usage in subprogs_unused test
bpf: Fix unsigned 'datasec_id' compared with zero in check_pseudo_btf_id
bpf: Fix passing zero to PTR_ERR() in bpf_btf_printf_prepare
libbpf: Don't attempt to load unused subprog as an entry-point BPF program
====================
Link: https://lore.kernel.org/r/20201119200721.288-1-alexei.starovoitov@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Luo Meng [Wed, 18 Nov 2020 13:49:31 +0000 (22:49 +0900)]
fail_function: Remove a redundant mutex unlock
Fix a mutex_unlock() issue where before copy_from_user() is
not called mutex_locked.
Fixes:
4b1a29a7f542 ("error-injection: Support fault injection framework")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Luo Meng <luomeng12@huawei.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Link: https://lore.kernel.org/bpf/160570737118.263807.8358435412898356284.stgit@devnote2
Alexei Starovoitov [Thu, 19 Nov 2020 18:45:55 +0000 (10:45 -0800)]
Merge branch 'Fix bpf_probe_read_user_str() overcopying'
Daniel Xu says:
====================
6ae08ae3dea2 ("bpf: Add probe_read_{user, kernel} and probe_read_{user,
kernel}_str helpers") introduced a subtle bug where
bpf_probe_read_user_str() would potentially copy a few extra bytes after
the NUL terminator.
This issue is particularly nefarious when strings are used as map keys,
as seemingly identical strings can occupy multiple entries in a map.
This patchset fixes the issue and introduces a selftest to prevent
future regressions.
v6 -> v7:
* Add comments
v5 -> v6:
* zero-pad up to sizeof(unsigned long) after NUL
v4 -> v5:
* don't read potentially uninitialized memory
v3 -> v4:
* directly pass userspace pointer to prog
* test more strings of different length
v2 -> v3:
* set pid filter before attaching prog in selftest
* use long instead of int as bpf_probe_read_user_str() retval
* style changes
v1 -> v2:
* add Fixes: tag
* add selftest
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Daniel Xu [Tue, 17 Nov 2020 20:05:46 +0000 (12:05 -0800)]
selftest/bpf: Test bpf_probe_read_user_str() strips trailing bytes after NUL
Previously, bpf_probe_read_user_str() could potentially overcopy the
trailing bytes after the NUL due to how do_strncpy_from_user() does the
copy in long-sized strides. The issue has been fixed in the previous
commit.
This commit adds a selftest that ensures we don't regress
bpf_probe_read_user_str() again.
Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/4d977508fab4ec5b7b574b85bdf8b398868b6ee9.1605642949.git.dxu@dxuuu.xyz
Daniel Xu [Tue, 17 Nov 2020 20:05:45 +0000 (12:05 -0800)]
lib/strncpy_from_user.c: Mask out bytes after NUL terminator.
do_strncpy_from_user() may copy some extra bytes after the NUL
terminator into the destination buffer. This usually does not matter for
normal string operations. However, when BPF programs key BPF maps with
strings, this matters a lot.
A BPF program may read strings from user memory by calling the
bpf_probe_read_user_str() helper which eventually calls
do_strncpy_from_user(). The program can then key a map with the
destination buffer. BPF map keys are fixed-width and string-agnostic,
meaning that map keys are treated as a set of bytes.
The issue is when do_strncpy_from_user() overcopies bytes after the NUL
terminator, it can result in seemingly identical strings occupying
multiple slots in a BPF map. This behavior is subtle and totally
unexpected by the user.
This commit masks out the bytes following the NUL while preserving
long-sized stride in the fast path.
Fixes:
6ae08ae3dea2 ("bpf: Add probe_read_{user, kernel} and probe_read_{user, kernel}_str helpers")
Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/21efc982b3e9f2f7b0379eed642294caaa0c27a7.1605642949.git.dxu@dxuuu.xyz
Linus Torvalds [Thu, 19 Nov 2020 19:32:31 +0000 (11:32 -0800)]
Merge tag 'powerpc-cve-2020-4788' of git://git./linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
"Fixes for CVE-2020-4788.
From Daniel's cover letter:
IBM Power9 processors can speculatively operate on data in the L1
cache before it has been completely validated, via a way-prediction
mechanism. It is not possible for an attacker to determine the
contents of impermissible memory using this method, since these
systems implement a combination of hardware and software security
measures to prevent scenarios where protected data could be leaked.
However these measures don't address the scenario where an attacker
induces the operating system to speculatively execute instructions
using data that the attacker controls. This can be used for example to
speculatively bypass "kernel user access prevention" techniques, as
discovered by Anthony Steinhauser of Google's Safeside Project. This
is not an attack by itself, but there is a possibility it could be
used in conjunction with side-channels or other weaknesses in the
privileged code to construct an attack.
This issue can be mitigated by flushing the L1 cache between privilege
boundaries of concern.
This patch series flushes the L1 cache on kernel entry (patch 2) and
after the kernel performs any user accesses (patch 3). It also adds a
self-test and performs some related cleanups"
* tag 'powerpc-cve-2020-4788' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s: rename pnv|pseries_setup_rfi_flush to _setup_security_mitigations
selftests/powerpc: refactor entry and rfi_flush tests
selftests/powerpc: entry flush test
powerpc: Only include kup-radix.h for 64-bit Book3S
powerpc/64s: flush L1D after user accesses
powerpc/64s: flush L1D on kernel entry
selftests/powerpc: rfi_flush: disable entry flush if present
Linus Torvalds [Thu, 19 Nov 2020 19:22:33 +0000 (11:22 -0800)]
Merge tag 'xtensa-
20201119' of git://github.com/jcmvbkbc/linux-xtensa
Pull xtensa fixes from Max Filippov:
- fix placement of cache alias remapping area
- disable preemption around cache alias management calls
- add missing __user annotation to strncpy_from_user argument
* tag 'xtensa-
20201119' of git://github.com/jcmvbkbc/linux-xtensa:
xtensa: uaccess: Add missing __user to strncpy_from_user() prototype
xtensa: disable preemption around cache alias management calls
xtensa: fix TLBTEMP area placement
Linus Torvalds [Thu, 19 Nov 2020 19:15:15 +0000 (11:15 -0800)]
Merge tag 'acpi-5.10-rc5' of git://git./linux/kernel/git/rafael/linux-pm
Pull ACPI fixes from Rafael Wysocki:
"These fix recent regression in the APEI code and initialization issue
in the ACPI fan driver.
Specifics:
- Make the APEI code avoid attempts to obtain logical addresses for
registers located in the I/O address space to fix initialization
issues (Aili Yao)
- Fix sysfs attribute initialization in the ACPI fan driver (Guenter
Roeck)"
* tag 'acpi-5.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI, APEI, Fix error return value in apei_map_generic_address()
ACPI: fan: Initialize performance state sysfs attribute
Linus Torvalds [Thu, 19 Nov 2020 19:11:50 +0000 (11:11 -0800)]
Merge tag 'pm-5.10-rc5' of git://git./linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
"These fix two issues in ARM cpufreq drivers and one cpuidle driver
issue.
Specifics:
- Add missing RCU_NONIDLE() annotations to the Tegra cpuidle driver
(Dmitry Osipenko)
- Fix boot frequency computation in the tegra186 cpufreq driver (Jon
Hunter)
- Make the SCMI cpufreq driver register a dummy clock provider to
avoid OPP addition failures (Sudeep Holla)"
* tag 'pm-5.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
cpufreq: scmi: Fix OPP addition failure with a dummy clock provider
cpufreq: tegra186: Fix get frequency callback
cpuidle: tegra: Annotate tegra_pm_set_cpu_in_lp2() with RCU_NONIDLE
Linus Torvalds [Thu, 19 Nov 2020 19:05:28 +0000 (11:05 -0800)]
Merge tag 'spi-fix-v5.10-rc4' of git://git./linux/kernel/git/broonie/spi
Pull spi fixes from Mark Brown:
"This is a relatively large set of fixes, the bulk of it being a series
from Lukas Wunner which fixes confusion with the lifetime of driver
data allocated along with the SPI controller structure that's been
created as part of the conversion to devm APIs.
The simplest fix, explained in detail in Lukas' commit message, is to
move to a devm_ function for allocation of the controller and hence
driver data in order to push the free of that after anything tries to
reference the driver data in the remove path. This results in a
relatively large diff due to the addition of a new function but isn't
particularly complex.
There's also a fix from Sven van Asbroeck which fixes yet more fallout
from the conflicts between the various different places one can
configure the polarity of GPIOs in modern systems.
Otherwise everything is fairly small and driver specific"
* tag 'spi-fix-v5.10-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
spi: npcm-fiu: Don't leak SPI master in probe error path
spi: dw: Set transfer handler before unmasking the IRQs
spi: cadence-quadspi: Fix error return code in cqspi_probe
spi: bcm2835aux: Restore err assignment in bcm2835aux_spi_probe
spi: lpspi: Fix use-after-free on unbind
spi: bcm-qspi: Fix use-after-free on unbind
spi: bcm2835aux: Fix use-after-free on unbind
spi: bcm2835: Fix use-after-free on unbind
spi: Introduce device-managed SPI controller allocation
spi: fsi: Fix transfer returning without finalizing message
spi: fix client driver breakages when using GPIO descriptors
Jakub Kicinski [Thu, 19 Nov 2020 18:59:22 +0000 (10:59 -0800)]
Merge branch 'net-smc-fixes-2020-11-18'
Karsten Graul says:
====================
net/smc: fixes 2020-11-18
Patch 1 fixes the matching of link groups because with SMC-Dv2 the vlanid
should no longer be part of this matching. Patch 2 removes a sparse message.
====================
Link: https://lore.kernel.org/r/20201118214038.24039-1-kgraul@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Karsten Graul [Wed, 18 Nov 2020 21:40:38 +0000 (22:40 +0100)]
net/smc: fix direct access to ib_gid_addr->ndev in smc_ib_determine_gid()
Sparse complaints 3 times about:
net/smc/smc_ib.c:203:52: warning: incorrect type in argument 1 (different address spaces)
net/smc/smc_ib.c:203:52: expected struct net_device const *dev
net/smc/smc_ib.c:203:52: got struct net_device [noderef] __rcu *const ndev
Fix that by using the existing and validated ndev variable instead of
accessing attr->ndev directly.
Fixes:
5102eca9039b ("net/smc: Use rdma_read_gid_l2_fields to L2 fields")
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Karsten Graul [Wed, 18 Nov 2020 21:40:37 +0000 (22:40 +0100)]
net/smc: fix matching of existing link groups
With the multi-subnet support of SMC-Dv2 the match for existing link
groups should not include the vlanid of the network device.
Set ini->smcd_version accordingly before the call to smc_conn_create()
and use this value in smc_conn_create() to skip the vlanid check.
Fixes:
5c21c4ccafe8 ("net/smc: determine accepted ISM devices")
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Thu, 19 Nov 2020 18:55:54 +0000 (10:55 -0800)]
Merge tag 'regulator-fix-v5.10-rc4' of git://git./linux/kernel/git/broonie/regulator
Pull regulator fixes from Mark Brown:
"Mostly core fixes here, one set from Michał Mirosław which cleans up
some issues introduced as part of the coupled regulators work, one
memory leak during probe and two due to regulators which have an input
supply name and regulator name which are identical, which is very
unusual.
There's also a fix for our handling of the similarly unusual case
where we can't determine if a regulator is enabled during boot"
* tag 'regulator-fix-v5.10-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:
regulator: ti-abb: Fix array out of bound read access on the first transition
regulator: workaround self-referent regulators
regulator: avoid resolve_supply() infinite recursion
regulator: fix memory leak with repeated set_machine_constraints()
regulator: pfuze100: limit pfuze-support-disable-sw to pfuze{100,200}
regulator: core: don't disable regulator if is_enabled return error.
Georg Kohmann [Thu, 19 Nov 2020 09:58:33 +0000 (10:58 +0100)]
ipv6: Remove dependency of ipv6_frag_thdr_truncated on ipv6 module
IPV6=m
NF_DEFRAG_IPV6=y
ld: net/ipv6/netfilter/nf_conntrack_reasm.o: in function
`nf_ct_frag6_gather':
net/ipv6/netfilter/nf_conntrack_reasm.c:462: undefined reference to
`ipv6_frag_thdr_truncated'
Netfilter is depending on ipv6 symbol ipv6_frag_thdr_truncated. This
dependency is forcing IPV6=y.
Remove this dependency by moving ipv6_frag_thdr_truncated out of ipv6. This
is the same solution as used with a similar issues: Referring to
commit
70b095c843266 ("ipv6: remove dependency of nf_defrag_ipv6 on ipv6
module")
Fixes:
9d9e937b1c8b ("ipv6/netfilter: Discard first fragment not including all headers")
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Georg Kohmann <geokohma@cisco.com>
Acked-by: Pablo Neira Ayuso <pablo@netfilter.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/r/20201119095833.8409-1-geokohma@cisco.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Thu, 19 Nov 2020 18:49:35 +0000 (10:49 -0800)]
Merge tag 'thermal-v5.10-rc5' of git://git./linux/kernel/git/thermal/linux
Pull thermal fix from Daniel Lezcano:
"Disable the CPU PM notifier for OMAP4430 for suspend in order to
prevent wrong temperature leading to a critical shutdown (Peter
Ujfalusi)"
* tag 'thermal-v5.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/thermal/linux:
thermal: ti-soc-thermal: Disable the CPU PM notifier for OMAP4430
Jiri Olsa [Wed, 18 Nov 2020 21:13:50 +0000 (22:13 +0100)]
libbpf: Fix VERSIONED_SYM_COUNT number parsing
We remove "other info" from "readelf -s --wide" output when
parsing GLOBAL_SYM_COUNT variable, which was added in [1].
But we don't do that for VERSIONED_SYM_COUNT and it's failing
the check_abi target on powerpc Fedora 33.
The extra "other info" wasn't problem for VERSIONED_SYM_COUNT
parsing until commit [2] added awk in the pipe, which assumes
that the last column is symbol, but it can be "other info".
Adding "other info" removal for VERSIONED_SYM_COUNT the same
way as we did for GLOBAL_SYM_COUNT parsing.
[1]
aa915931ac3e ("libbpf: Fix readelf output parsing for Fedora")
[2]
746f534a4809 ("tools/libbpf: Avoid counting local symbols in ABI check")
Fixes:
746f534a4809 ("tools/libbpf: Avoid counting local symbols in ABI check")
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20201118211350.1493421-1-jolsa@kernel.org
Hayes Wang [Wed, 18 Nov 2020 06:43:58 +0000 (14:43 +0800)]
r8153_ecm: avoid to be prior to r8152 driver
Avoid r8153_ecm is compiled as built-in, if r8152 driver is compiled
as modules. Otherwise, the r8153_ecm would be used, even though the
device is supported by r8152 driver.
Fixes:
c1aedf015ebd ("net/usb/r8153_ecm: support ECM mode for RTL8153")
Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://lore.kernel.org/r/1394712342-15778-394-Taiwan-albertk@realtek.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Rafael J. Wysocki [Thu, 19 Nov 2020 13:03:52 +0000 (14:03 +0100)]
Merge branch 'acpi-fan'
* acpi-fan:
ACPI: fan: Initialize performance state sysfs attribute
Rafael J. Wysocki [Thu, 19 Nov 2020 13:02:18 +0000 (14:02 +0100)]
Merge branch 'pm-cpuidle'
* pm-cpuidle:
cpuidle: tegra: Annotate tegra_pm_set_cpu_in_lp2() with RCU_NONIDLE
Daniel Axtens [Tue, 17 Nov 2020 05:59:16 +0000 (16:59 +1100)]
powerpc/64s: rename pnv|pseries_setup_rfi_flush to _setup_security_mitigations
pseries|pnv_setup_rfi_flush already does the count cache flush setup, and
we just added entry and uaccess flushes. So the name is not very accurate
any more. In both platforms we then also immediately setup the STF flush.
Rename them to _setup_security_mitigations and fold the STF flush in.
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Daniel Axtens [Tue, 17 Nov 2020 05:59:15 +0000 (16:59 +1100)]
selftests/powerpc: refactor entry and rfi_flush tests
For simplicity in backporting, the original entry_flush test contained
a lot of duplicated code from the rfi_flush test. De-duplicate that code.
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Daniel Axtens [Tue, 17 Nov 2020 05:59:14 +0000 (16:59 +1100)]
selftests/powerpc: entry flush test
Add a test modelled on the RFI flush test which counts the number
of L1D misses doing a simple syscall with the entry flush on and off.
For simplicity of backporting, this test duplicates a lot of code from
rfi_flush. We clean that up in the next patch.
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Michael Ellerman [Thu, 19 Nov 2020 12:43:53 +0000 (23:43 +1100)]
powerpc: Only include kup-radix.h for 64-bit Book3S
In kup.h we currently include kup-radix.h for all 64-bit builds, which
includes Book3S and Book3E. The latter doesn't make sense, Book3E
never uses the Radix MMU.
This has worked up until now, but almost by accident, and the recent
uaccess flush changes introduced a build breakage on Book3E because of
the bad structure of the code.
So disentangle things so that we only use kup-radix.h for Book3S. This
requires some more stubs in kup.h and fixing an include in
syscall_64.c.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 17 Nov 2020 05:59:13 +0000 (16:59 +1100)]
powerpc/64s: flush L1D after user accesses
IBM Power9 processors can speculatively operate on data in the L1 cache
before it has been completely validated, via a way-prediction mechanism. It
is not possible for an attacker to determine the contents of impermissible
memory using this method, since these systems implement a combination of
hardware and software security measures to prevent scenarios where
protected data could be leaked.
However these measures don't address the scenario where an attacker induces
the operating system to speculatively execute instructions using data that
the attacker controls. This can be used for example to speculatively bypass
"kernel user access prevention" techniques, as discovered by Anthony
Steinhauser of Google's Safeside Project. This is not an attack by itself,
but there is a possibility it could be used in conjunction with
side-channels or other weaknesses in the privileged code to construct an
attack.
This issue can be mitigated by flushing the L1 cache between privilege
boundaries of concern. This patch flushes the L1 cache after user accesses.
This is part of the fix for CVE-2020-4788.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Nicholas Piggin [Tue, 17 Nov 2020 05:59:12 +0000 (16:59 +1100)]
powerpc/64s: flush L1D on kernel entry
IBM Power9 processors can speculatively operate on data in the L1 cache
before it has been completely validated, via a way-prediction mechanism. It
is not possible for an attacker to determine the contents of impermissible
memory using this method, since these systems implement a combination of
hardware and software security measures to prevent scenarios where
protected data could be leaked.
However these measures don't address the scenario where an attacker induces
the operating system to speculatively execute instructions using data that
the attacker controls. This can be used for example to speculatively bypass
"kernel user access prevention" techniques, as discovered by Anthony
Steinhauser of Google's Safeside Project. This is not an attack by itself,
but there is a possibility it could be used in conjunction with
side-channels or other weaknesses in the privileged code to construct an
attack.
This issue can be mitigated by flushing the L1 cache between privilege
boundaries of concern. This patch flushes the L1 cache on kernel entry.
This is part of the fix for CVE-2020-4788.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Russell Currey [Tue, 17 Nov 2020 05:59:11 +0000 (16:59 +1100)]
selftests/powerpc: rfi_flush: disable entry flush if present
We are about to add an entry flush. The rfi (exit) flush test measures
the number of L1D flushes over a syscall with the RFI flush enabled and
disabled. But if the entry flush is also enabled, the effect of enabling
and disabling the RFI flush is masked.
If there is a debugfs entry for the entry flush, disable it during the RFI
flush and restore it later.
Reported-by: Spoorthy S <spoorts2@in.ibm.com>
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Jakub Kicinski [Thu, 19 Nov 2020 01:49:24 +0000 (17:49 -0800)]
Merge tag 'linux-can-fixes-for-5.10-
20201118' of git://git./linux/kernel/git/mkl/linux-can
Marc Kleine-Budde says:
====================
pull-request: can 2020-11-18
Jimmy Assarsson provides two patches for the kvaser_pciefd and kvaser_usb
drivers, where the can_bittiming_const are fixed.
The next patch is by me and fixes an erroneous flexcan_transceiver_enable()
during bus-off recovery in the flexcan driver.
Jarkko Nikula's patch for the m_can driver fixes the IRQ handler to only
process the interrupts if the device is not suspended.
* tag 'linux-can-fixes-for-5.10-
20201118' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: m_can: process interrupt only when not runtime suspended
can: flexcan: flexcan_chip_start(): fix erroneous flexcan_transceiver_enable() during bus-off recovery
can: kvaser_usb: kvaser_usb_hydra: Fix KCAN bittiming limits
can: kvaser_pciefd: Fix KCAN bittiming limits
====================
Link: https://lore.kernel.org/r/20201118160414.2731659-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Aya Levin [Wed, 18 Nov 2020 08:19:22 +0000 (10:19 +0200)]
net/mlx4_core: Fix init_hca fields offset
Slave function read the following capabilities from the wrong offset:
1. log_mc_entry_sz
2. fs_log_entry_sz
3. log_mc_hash_sz
Fix that by adjusting these capabilities offset to match firmware
layout.
Due to the wrong offset read, the following issues might occur:
1+2. Negative value reported at max_mcast_qp_attach.
3. Driver to init FW with multicast hash size of zero.
Fixes:
a40ded604365 ("net/mlx4_core: Add masking for a few queries on HCA caps")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Eran Ben Elisha <eranbe@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20201118081922.553-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Bongsu Jeon [Tue, 17 Nov 2020 08:11:37 +0000 (17:11 +0900)]
nfc: s3fwrn5: Change the error code
ENOTSUPP is not a SUSV4 error code, prefer EOPNOTSUPP.
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Bongsu Jeon <bongsu.jeon@samsung.com>
Link: https://lore.kernel.org/r/20201117081137epcms2p84b5dd8d84ca608f44b0bb722b48f50b1@epcms2p8
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Bongsu Jeon [Tue, 17 Nov 2020 08:09:50 +0000 (17:09 +0900)]
nfc: s3fwrn5: Fix the misspelling in a comment
stucture should be replaced by structure.
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Bongsu Jeon <bongsu.jeon@samsung.com>
Link: https://lore.kernel.org/r/20201117080950epcms2p4d332688ce351321c3222734bca680b57@epcms2p4
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Bongsu Jeon [Tue, 17 Nov 2020 08:08:24 +0000 (17:08 +0900)]
nfc: s3fwrn5: Remove the max_payload
max_payload is unused.
Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Bongsu Jeon <bongsu.jeon@samsung.com>
Link: https://lore.kernel.org/r/20201117080824epcms2p36f70e06e2d8bd51d1af278b26ca65725@epcms2p3
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 19 Nov 2020 01:34:21 +0000 (17:34 -0800)]
Merge branch 's390-qeth-updates-2020-11-17'
Julian Wiedmann says:
====================
s390/qeth: updates 2020-11-17
This brings some cleanups, and a bunch of improvements for our
.get_link_ksettings() code.
====================
Link: https://lore.kernel.org/r/20201117161520.1089-1-jwi@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:20 +0000 (17:15 +0100)]
s390/qeth: improve selection of ethtool link modes
The link mode is a combination of port speed and port mode. But we
currently only consider the speed, and then typically select the
corresponding TP-based link mode. For 1G and 10G Fibre links this means
we display the wrong link modes.
Move the SPEED_* switch statements inside the PORT_* cases, and only
consider valid combinations where we can select the corresponding
link mode. Add the relevant link modes (1000baseX, 10000baseSR and
1000baseLR) that were introduced back with
commit
5711a9822144 ("net: ethtool: add support for 1000BaseX and missing 10G link modes").
To differentiate between 10000baseSR and 10000baseLR, use the detailed
media_type information that QUERY OAT provides.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:19 +0000 (17:15 +0100)]
s390/qeth: use QUERY OAT for initial link info
Improve the initial link info with data obtained from QUERY OAT.
Doing so _only_ at initialization time avoids
1. dealing with multi-part replies, and
2. sifting through all the data that may get returned at runtime.
This allows us to determine the correct port type for the 1000BT variant
of recent OSA adapter generations (where the .card_type field in
QUERY CARD INFO is no longer sufficient).
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:18 +0000 (17:15 +0100)]
s390/qeth: clean up default cases for ethtool link mode
Remove the default case for PORT_* and SPEED_* in our ethtool code.
The only time these could be hit is if qeth_init_link_info() was unable
to determine the port type from an OSA adapter's link_type.
We already throw a message in this case, so reduce the noise and don't
report bad data (ie. it's much more likely that any future link_type
will represent a PORT_FIBRE link ...).
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:17 +0000 (17:15 +0100)]
s390/qeth: set static link info during initialization
Hard-code the minimal link info at initialization time, after we
obtained the link_type. qeth_get_link_ksettings() can still override
this with more accurate data from QUERY CARD INFO later on.
Don't set arbitrary defaults for unknown OSA link types, they
certainly won't match any future type.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:16 +0000 (17:15 +0100)]
s390/qeth: improve QUERY CARD INFO processing
Move all the HW reply data parsing into qeth_query_card_info_cb(), and
use common ethtool enums for transporting the information back to the
caller.
Also only look at the .port_speed field when we couldn't determine the
speed from the .card_type field, and introduce some 'default' cases for
SPEED_UNKNOWN, PORT_OTHER and DUPLEX_UNKNOWN.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:15 +0000 (17:15 +0100)]
s390/qeth: tolerate error when querying card info
By the time that our .get_link_ksettings() code issues a QUERY CARD INFO
cmd to get link-related information, we already set up a good amount of
static link data.
Return this data when the cmd fails, same as when the cmd is not
supported.
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Kaixu Xia [Tue, 17 Nov 2020 16:15:14 +0000 (17:15 +0100)]
s390/qeth: remove useless if/else
Fix the following coccinelle report:
./drivers/s390/net/qeth_l3_main.c:107:2-4: WARNING: possible condition with no effect (if == else)
Both branches are the same since
commit
ab29c480b194 ("s390/qeth: replace deprecated simple_stroul()"),
so remove them.
Reported-by: Tosk Robot <tencent_os_robot@tencent.com>
Signed-off-by: Kaixu Xia <kaixuxia@tencent.com>
[jwi: point to the commit that introduced this]
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Julian Wiedmann [Tue, 17 Nov 2020 16:15:13 +0000 (17:15 +0100)]
s390/qeth: reduce rtnl locking for switchdev events
call_switchdev_notifiers() doesn't require holding the RTNL lock since
commit
ff5cf100110c ("net: switchdev: Change notifier chain to be atomic").
We still need it for the "lost event" slow path, to avoid racing against
a concurrent .ndo_bridge_setlink().
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Alexandra Winter <wintera@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Heiner Kallweit [Tue, 17 Nov 2020 20:34:09 +0000 (21:34 +0100)]
r8169: remove not needed check in rtl8169_start_xmit
In rtl_tx() the released descriptors are zero'ed by
rtl8169_unmap_tx_skb(). And in the beginning of rtl8169_start_xmit()
we check that enough descriptors are free, therefore there's no way
the DescOwn bit can be set here.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/6965d665-6c50-90c5-70e6-0bb335d4ea47@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Heiner Kallweit [Tue, 17 Nov 2020 20:25:42 +0000 (21:25 +0100)]
net: bridge: replace struct br_vlan_stats with pcpu_sw_netstats
Struct br_vlan_stats duplicates pcpu_sw_netstats (apart from
br_vlan_stats not defining an alignment requirement), therefore
switch to using the latter one.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/04d25c3d-c5f6-3611-6d37-c2f40243dae2@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 19 Nov 2020 01:17:32 +0000 (17:17 -0800)]
Merge tag 'mlx5-fixes-2020-11-17' of git://git./linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5 fixes 2020-11-17
This series introduces some fixes to mlx5 driver.
* tag 'mlx5-fixes-2020-11-17' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux:
net/mlx5: fix error return code in mlx5e_tc_nic_init()
net/mlx5: E-Switch, Fail mlx5_esw_modify_vport_rate if qos disabled
net/mlx5: Disable QoS when min_rates on all VFs are zero
net/mlx5: Clear bw_share upon VF disable
net/mlx5: Add handling of port type in rule deletion
net/mlx5e: Fix check if netdev is bond slave
net/mlx5e: Fix IPsec packet drop by mlx5e_tc_update_skb
net/mlx5e: Set IPsec WAs only in IP's non checksum partial case.
net/mlx5e: Fix refcount leak on kTLS RX resync
====================
Link: https://lore.kernel.org/r/20201117195702.386113-1-saeedm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 19 Nov 2020 00:44:02 +0000 (16:44 -0800)]
Merge branch 'atm-replace-in_interrupt-usage'
Sebastian Andrzej Siewior says:
====================
atm: Replace in_interrupt usage
this mini series contains the removal of in_interrupt() in drivers/atm
====================
Link: https://lore.kernel.org/r/20201116162117.387191-1-bigeasy@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Sebastian Andrzej Siewior [Mon, 16 Nov 2020 16:21:16 +0000 (17:21 +0100)]
atm: lanai: Remove in_interrupt() usage
lanai_shutdown_tx_vci() uses in_interrupt() to issue a warning message
if the function was used in context in which it is not safe to sleep.
The usage of in_interrupt() in driver code is deprecated as it can not always
detect all states where it is not allowed to sleep.
msleep() has debug code which will trigger a warning if used in bad
context.
Remove in_interrupt().
Cc: Chas Williams <3chas3@gmail.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Sebastian Andrzej Siewior [Mon, 16 Nov 2020 16:21:15 +0000 (17:21 +0100)]
atm: nicstar: Replace in_interrupt() usage
push_scqe() uses in_interrupt() to figure out if it is allowed to sleep.
The usage of in_interrupt() in drivers is phased out and Linus clearly
requested that code which changes behaviour depending on context should
either be separated or the context be conveyed in an argument passed by the
caller, which usually knows the context.
Aside of that in_interrupt() is not correct as it does not catch preempt
disabled regions which neither can sleep.
ns_send() (the only caller of push_scqe()) has the following callers:
- vcc_sendmsg() used as proto_ops::sendmsg is expected to be invoked in
preemtible context.
-> vcc->dev->ops->send() (ns_send())
- atm_vcc::send via atmdev_ops::send either directly (pointer copied by
atm_init_aal34() or atm_init_aal5()) or via atm_send_aal0().
This is invoked by drivers (like br2684, clip, pppoatm, ...) which are
called from net_device_ops::ndo_start_xmit with BH disabled.
Add atmdev_ops::send_bh which is used by callers from BH context
(atm_send_aal*()) and if this callback missing then ::send is used
instead.
Implement this callback in nicstar and use it to replace in_interrupt().
Cc: Chas Williams <3chas3@gmail.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Sebastian Andrzej Siewior [Mon, 16 Nov 2020 16:21:14 +0000 (17:21 +0100)]
atm: nicstar: Unmap DMA on send error
The `skb' is mapped for DMA in ns_send() but does not unmap DMA in case
push_scqe() fails to submit the `skb'. The memory of the `skb' is
released so only the DMA mapping is leaking.
Unmap the DMA mapping in case push_scqe() failed.
Fixes:
864a3ff635fa7 ("atm: [nicstar] remove virt_to_bus() and support 64-bit platforms")
Cc: Chas Williams <3chas3@gmail.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Wed, 18 Nov 2020 23:53:51 +0000 (15:53 -0800)]
Merge branch 'net-ipa-ipa-register-cleanup'
Alex Elder says:
====================
net: ipa: IPA register cleanup
This series consists of cleanup patches, almost entirely related to
the definitions for IPA registers. Some comments are updated or
added to provide better information about defined IPA registers.
Other cleanups ensure symbol names and their assigned values are
defined consistently. Some essentially duplicate definitions get
consolidated for simplicity. In a few cases some minor bugs
(missing definitions) are fixed. With these changes, all IPA
register offsets and associated field masks should be correct for
IPA versions 3.5.1, 4.0, 4.1, and 4.2.
====================
Link: https://lore.kernel.org/r/20201116233805.13775-1-elder@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:05 +0000 (17:38 -0600)]
net: ipa: a few last IPA register cleanups
Some last cleanups for the existing IPA register definitions:
- Remove the definition of IPA_REG_ENABLED_PIPES_OFFSET, because
it is not used.
- Use "IPA_" instead of "BAM_" as the prefix on fields associated
with the FLAVOR_0 register. We use GSI (not BAM), but the
fields apply to both GSI and BAM.
- Get rid of the definition of IPA_CS_RSVD; it is never used.
- Add two missing field mask definitions for the INIT_DEAGGR
endpoint register.
- Eliminate a few of the defined sequencer types, because they
are unused. We can add them back when needed.
- Add a field mask to indicate which bit causes an interrupt on
the microcontroller.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:04 +0000 (17:38 -0600)]
net: ipa: move definition of enum ipa_irq_id
Move the definition of the ipa_irq_id enumerated type out of
"ipa_interrupt.h" and into "ipa_reg.h", and flesh out its set of
defined values. Each interrupt id indicates a particular type of
IPA interrupt that can be signaled. Their numeric values define bit
positions in the IPA_IRQ_* registers, so should their definitions
should accompany the definition of those register offsets.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:03 +0000 (17:38 -0600)]
net: ipa: rearrange a few IPA register definitions
Move a few things around in "ipa_reg.h":
- Move the definition of ipa_reg_state_aggr_active_offset() down
a bit in the file so definitions are ordered by offset (for
the lowest supported IPA version) like all other definitions.
- Move the definition TIMER_FREQUENCY to be immediately above
the definition of ipa_aggr_granularity_val() where it's used.
- Move each register field value enumerated type definition to
immediately follow the definitions of the register and field
it is associated with.
No code functionality is modified by this patch.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:02 +0000 (17:38 -0600)]
net: ipa: fix up IPA register comments
Revise or add comments in "ipa_reg.h" for to provide more
information, and to improve clarity and consistency.
- Always provide a comment to define when a register or field is
supported (or not) for certain versions of IPA hardware.
- Try to be specific about *which* or *how many* definitions
a comment refers to.
- Move comments stating that ipa->available defines the valid
bits in various registers *above* the register offset
definition, to avoid some checkpatch.pl warnings.
No code is changed by this patch.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:01 +0000 (17:38 -0600)]
net: ipa: define enumerated types consistently
Consistently define numeric values for enumerated type members using
hexidecimal (rather than decimal) format values. Align the values
assigned in the same column in each file.
Only assign values where they really matter, for example don't
assign IPA_ENDPOINT_AP_MODEM_TX the value 0.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:38:00 +0000 (17:38 -0600)]
net: ipa: fix BCR register field definitions
The backward compatibility register field masks are defined using
single-bit masks defined with BIT(x) rather than GENMASK(x, x).
Change this one set of definitions to follow the GENMASK() pattern
used everywhere else. Add a few missing field definitions for this
register as well.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:37:59 +0000 (17:37 -0600)]
net: ipa: use _FMASK consistently
Several IPA register field masks are defined without the "_FMASK"
suffix naming convention. Rename these, so all field masks are
consistently named.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:37:58 +0000 (17:37 -0600)]
net: ipa: fix two inconsistent IPA register names
Rename two suspend IRQ registers so they follow the IPA_REG_IRQ_xxx
naming convention used elsewhere.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:37:57 +0000 (17:37 -0600)]
net: ipa: support more versions for HOLB timer
IPA version 3.5.1 represents the timer used in avoiding head-of-line
blocking with a simple tick count. IPA v4.2 changes that, instead
splitting the timer field into two parts (base and scale) to
represent the ticks in the timer period.
IPA v4.0 and IPA v4.1 use the same method as IPA v3.5.1. Change the
test in ipa_reg_init_hol_block_timer_val() so the result is correct
for those versions as well.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:37:56 +0000 (17:37 -0600)]
net: ipa: make filter/routing hash enable register variable
For IPA v3.5.1, the IPA filter/routing hash enable register actually
does exist, but it is at offset 0x8c into the IPA register space.
For newer versions of IPA it is at offset 0x148.
Define a new inline function ipa_reg_filt_rout_hash_en_offset() to
return the appropriate value for a given version of IPA hardware.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Alex Elder [Mon, 16 Nov 2020 23:37:55 +0000 (17:37 -0600)]
net: ipa: share field mask values for IPA hash registers
The IPA filter/routing hash enable register and filter/routing hash
flush register each have four single-bit fields representing the
four hashed tables to be enabled or flushed. The field positions
are identical, so just use a single set of field masks to represent
the fields for both registers.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Dongli Zhang [Sun, 15 Nov 2020 20:10:29 +0000 (12:10 -0800)]
page_frag: Recover from memory pressure
The ethernet driver may allocate skb (and skb->data) via napi_alloc_skb().
This ends up to page_frag_alloc() to allocate skb->data from
page_frag_cache->va.
During the memory pressure, page_frag_cache->va may be allocated as
pfmemalloc page. As a result, the skb->pfmemalloc is always true as
skb->data is from page_frag_cache->va. The skb will be dropped if the
sock (receiver) does not have SOCK_MEMALLOC. This is expected behaviour
under memory pressure.
However, once kernel is not under memory pressure any longer (suppose large
amount of memory pages are just reclaimed), the page_frag_alloc() may still
re-use the prior pfmemalloc page_frag_cache->va to allocate skb->data. As a
result, the skb->pfmemalloc is always true unless page_frag_cache->va is
re-allocated, even if the kernel is not under memory pressure any longer.
Here is how kernel runs into issue.
1. The kernel is under memory pressure and allocation of
PAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,
the pfmemalloc page is allocated for page_frag_cache->va.
2: All skb->data from page_frag_cache->va (pfmemalloc) will have
skb->pfmemalloc=true. The skb will always be dropped by sock without
SOCK_MEMALLOC. This is an expected behaviour.
3. Suppose a large amount of pages are reclaimed and kernel is not under
memory pressure any longer. We expect skb->pfmemalloc drop will not happen.
4. Unfortunately, page_frag_alloc() does not proactively re-allocate
page_frag_alloc->va and will always re-use the prior pfmemalloc page. The
skb->pfmemalloc is always true even kernel is not under memory pressure any
longer.
Fix this by freeing and re-allocating the page instead of recycling it.
References: https://lore.kernel.org/lkml/
20201103193239.1807-1-dongli.zhang@oracle.com/
References: https://lore.kernel.org/linux-mm/
20201105042140.5253-1-willy@infradead.org/
Suggested-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Aruna Ramakrishna <aruna.ramakrishna@oracle.com>
Cc: Bert Barbe <bert.barbe@oracle.com>
Cc: Rama Nichanamatlu <rama.nichanamatlu@oracle.com>
Cc: Venkat Venkatsubra <venkat.x.venkatsubra@oracle.com>
Cc: Manjunath Patil <manjunath.b.patil@oracle.com>
Cc: Joe Jin <joe.jin@oracle.com>
Cc: SRINIVAS <srinivas.eeda@oracle.com>
Fixes:
79930f5892e1 ("net: do not deplete pfmemalloc reserve")
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20201115201029.11903-1-dongli.zhang@oracle.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Wed, 18 Nov 2020 20:12:18 +0000 (12:12 -0800)]
Merge tag 'gfs2-v5.10-rc4-fixes' of git://git./linux/kernel/git/gfs2/linux-gfs2
Pull gfs2 fix from Andreas Gruenbacher:
"Fix gfs2 freeze/thaw"
* tag 'gfs2-v5.10-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
gfs2: Fix regression in freeze_go_sync
Linus Torvalds [Wed, 18 Nov 2020 20:06:34 +0000 (12:06 -0800)]
Merge tag 'nfsd-5.10-2' of git://linux-nfs.org/~bfields/linux
Pull nfsd fix from Bruce Fields:
"Just one quick fix for a tracing oops"
* tag 'nfsd-5.10-2' of git://linux-nfs.org/~bfields/linux:
SUNRPC: Fix oops in the rpc_xdr_buf event class
Xie He [Wed, 18 Nov 2020 12:42:26 +0000 (04:42 -0800)]
Documentation: Remove the deleted "framerelay" document from the index
commit
f73659192b0b ("net: wan: Delete the DLCI / SDLA drivers")
deleted "Documentation/networking/framerelay.rst". However, it is still
referenced in "Documentation/networking/index.rst". We need to remove the
reference, too.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Xie He <xie.he.0141@gmail.com>
Link: https://lore.kernel.org/r/20201118124226.15588-1-xie.he.0141@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Linus Torvalds [Wed, 18 Nov 2020 19:57:55 +0000 (11:57 -0800)]
Merge tag 'linux-kselftest-kunit-fixes-5.10-rc5' of git://git./linux/kernel/git/shuah/linux-kselftest
Pull Kunit fixes from Shuah Khan:
"Several fixes to Kunit documentation and tools, and to not pollute
the source directory.
Also remove the incorrect kunit .gitattributes file"
* tag 'linux-kselftest-kunit-fixes-5.10-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest:
kunit: fix display of failed expectations for strings
kunit: tool: fix extra trailing \n in raw + parsed test output
kunit: tool: print out stderr from make (like build warnings)
KUnit: Docs: usage: wording fixes
KUnit: Docs: style: fix some Kconfig example issues
KUnit: Docs: fix a wording typo
kunit: Do not pollute source directory with generated files (test.log)
kunit: Do not pollute source directory with generated files (.kunitconfig)
kunit: tool: fix pre-existing python type annotation errors
kunit: Fix kunit.py parse subcommand (use null build_dir)
kunit: tool: unmark test_data as binary blobs
Jakub Kicinski [Wed, 18 Nov 2020 19:51:21 +0000 (11:51 -0800)]
Merge branch 'mlxsw-preparations-for-nexthop-objects-support-part-2-2'
Ido Schimmel says:
====================
mlxsw: Preparations for nexthop objects support - part 2/2
This patch set contains the second round of preparations towards nexthop
objects support in mlxsw. Follow up patches can be found here [1].
The patches are mostly small and trivial and contain non-functional
changes aimed at making it easier to integrate nexthop objects with
mlxsw.
Patch #1 is a fix for an issue introduced in previous submission. Found
by Coverity.
[1] https://github.com/idosch/linux/tree/submit/nexthop_objects
====================
Link: https://lore.kernel.org/r/20201117174704.291990-1-idosch@idosch.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ido Schimmel [Tue, 17 Nov 2020 17:47:04 +0000 (19:47 +0200)]
mlxsw: spectrum_router: Allow returning errors from mlxsw_sp_nexthop_group_refresh()
The function is responsible for allocating the adjacency entries used by
the nexthop group and populating them with the adjacency information
such as egress RIF and MAC address.
Allow the function to return an error when it encounters a problem and
have the relevant call sites check it.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ido Schimmel [Tue, 17 Nov 2020 17:47:03 +0000 (19:47 +0200)]
mlxsw: spectrum_router: Add an indication if a nexthop group can be destroyed
Currently, a nexthop group is destroyed when the last FIB entry is
detached from it.
When nexthop objects are supported, this can no longer be the case, as
the group is a separate object whose lifetime is managed by user space.
Add an indication if a nexthop group can be destroyed and always set it
to true for the existing IPv4 and IPv6 nexthop groups.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Ido Schimmel [Tue, 17 Nov 2020 17:47:02 +0000 (19:47 +0200)]
mlxsw: spectrum_router: Only clear offload indication from valid IPv6 FIB info
When the IPv6 FIB info has a nexthop object, the nexthop offload
indication is set on the nexthop object and not on the FIB info itself.
Therefore, do not try to clear the offload indication from the FIB info
when it has a nexthop object.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>