| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
If a netdev is removed from the lag the lag should be destroyed.
With downstream patches this might trigger a reconfiguration of
representors on a different eswitch and such we don't have the proper
locking to so from this path. Move the destruction to be done by the
workqueue.
As the destruction won't affect the netdev side it okay to do so.
The RDMA side will be reconfigured and it already coded to handle such
reconfiguration.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Currently when doing hardware lag we check the eswitch mode
but as this isn't done under a lock the check isn't valid.
As the code needs to sync between two different devices an extra
care is needed.
- When going to change eswitch mode, if hardware lag is active destroy it.
- While changing eswitch modes block any hardware bond creation.
- Delay handling bonding events until there are no mode changes in
progress.
- When attaching a new mdev to lag, block until there is no mode change
in progress. In order for the mode change to finish the interface lock
will have to be taken. Release the lock and sleep for 100ms to
allow forward progress. As this is a very rare condition (can happen if
the user unbinds and binds a PCI function while also changing eswitch
mode of the other PCI function) it has no real world impact.
As taking multiple eswitch mode locks is now required lockdep will
complain about a possible deadlock. Register a key per eswitch to make
lockdep happy.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
When two mlx5 devices are paired in switchdev mode, always offload the
send-to-vport rule to the peer E-Switch. This allows to abstract
the logic when this is really necessary (single FDB) and combine
the logic of both cases into one.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
This callback will allow to notify representors about relevant events
when in OFFLOADS mode. In downstream patches, this will be used to notify
about PAIR/UNPAIR devcom events.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
FTEs are added with mapped metadata which is saved per eswitch.
When uplink reps are bonded and we are in a single FDB mode,
we could fail to find metadata which was stored on one eswitch mapping
but not the other or with a different id.
To resolve this issue use shared mapping between eswitch ports.
We do not have any conflict using a single mapping, for a type,
between the ports.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
The shared mapping is identified by an id and type.
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Set the flow source param to local vport for the uplink rep
send-to-vport rule.
This will comply with the recent changes in SW steering that
use the flow source as an indication for the rule type - rx or tx.
Since the uplink send-to-vport rule is forwarding traffic to the wire
it has to indicate that it is an sx rule and can't use the any port
value in the flow source.
Signed-off-by: Ariel Levkovich <lariel@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Shared FDB allows to create a single RDMA device that holds representors
from both eswitches. As shared FDB is only active when both uplink
representors are enslaved there is a single RDMA port that represents
both uplinks.
The number of ports is the number of vports on both eswitches minus one
as we only need 1 port for both uplinks.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
In shared FDB there is only one eswitch which is active and it receives
traffic from all representors and all vports in the HCA.
While the Ethernet representor will always reside on its native PF
the IB representor will not. Extend send to vport rule creation to
support such flows. Need to account for source vport that sends the
traffic (on which the representors resides) and the target eswitch
the traffic which reach.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
In shared FDB a single RDMA device can have representors that are
connected to two different eswitches. Use the right eswitch when
preparing the response to userspace.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
As shared FDB requires changes in two subsystems first expose the needed
core functions so the RDMA side can be changed.
mlx5_lag_is_master(): return true if a given mlx5 device is the lag master.
mlx5_lag_is_shared_fdb(): Returns true if the lag mode is shared FDB.
mlx5_lag_get_peer_mdev(): Return the peer mdev in lag.
The mentioned functions will be used by downstream patches in order
to add support for shared FDB for the RDMA side.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
Export a function so users can retrieve the mellanox device that manages
the eswitch from the eswitch device.
Signed-off-by: Mark Bloch <mbloch@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
|
|
This wrapper is never used.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-6-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This one is never used.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-5-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This wrapper is a simple call to udelay(). Remove it.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-4-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This wrapper just calls mdelay(). Remove it.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-3-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
The effect of this macro is to call msleep(). Remove the wrapper.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-2-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This wrapper is just a call to yield().
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805192644.15978-1-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Add ACPI module ID to probe the driver from the ACPI based bootloader
firmware.
Signed-off-by: Akhil R <akhilrajeev@nvidia.com>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
Pull networking fixes from Jakub Kicinski:
"Including fixes from ipsec.
Current release - regressions:
- sched: taprio: fix init procedure to avoid inf loop when dumping
- sctp: move the active_key update after sh_keys is added
Current release - new code bugs:
- sparx5: fix build with old GCC & bitmask on 32-bit targets
Previous releases - regressions:
- xfrm: redo the PREEMPT_RT RCU vs hash_resize_mutex deadlock fix
- xfrm: fixes for the compat netlink attribute translator
- phy: micrel: Fix detection of ksz87xx switch
Previous releases - always broken:
- gro: set inner transport header offset in tcp/udp GRO hook to avoid
crashes when such packets reach GSO
- vsock: handle VIRTIO_VSOCK_OP_CREDIT_REQUEST, as required by spec
- dsa: sja1105: fix static FDB entries on SJA1105P/Q/R/S and SJA1110
- bridge: validate the NUD_PERMANENT bit when adding an extern_learn
FDB entry
- usb: lan78xx: don't modify phy_device state concurrently
- usb: pegasus: check for errors of IO routines"
* tag 'net-5.14-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (48 commits)
net: vxge: fix use-after-free in vxge_device_unregister
net: fec: fix use-after-free in fec_drv_remove
net: pegasus: fix uninit-value in get_interrupt_interval
net: ethernet: ti: am65-cpsw: fix crash in am65_cpsw_port_offload_fwd_mark_update()
bnx2x: fix an error code in bnx2x_nic_load()
net: wwan: iosm: fix recursive lock acquire in unregister
net: wwan: iosm: correct data protocol mask bit
net: wwan: iosm: endianness type correction
net: wwan: iosm: fix lkp buildbot warning
net: usb: lan78xx: don't modify phy_device state concurrently
docs: networking: netdevsim rules
net: usb: pegasus: Remove the changelog and DRIVER_VERSION.
net: usb: pegasus: Check the return value of get_geristers() and friends;
net/prestera: Fix devlink groups leakage in error flow
net: sched: fix lockdep_set_class() typo error for sch->seqlock
net: dsa: qca: ar9331: reorder MDIO write sequence
VSOCK: handle VIRTIO_VSOCK_OP_CREDIT_REQUEST
mptcp: drop unused rcu member in mptcp_pm_addr_entry
net: ipv6: fix returned variable type in ip6_skb_dst_mtu
nfp: update ethtool reporting of pauseframe control
...
|
|
gpiolib does not modify struct device_node, so few local pointers can
point to a const data.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
Several gpiolib functions receive pointer to struct device_node which is
later passed to OF functions. These OF functions accept already pointer
to const, so gpiolib can follow similar approach to indicate they are
not modifying the struct device_node.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
buswidth and dtr fields in spi_mem_op are only valid when the
corresponding spi_mem_op phase has a non-zero length. For example,
SPI NAND core doesn't set buswidth when using SPI_MEM_OP_NO_ADDR
phase.
Fix the dtr checks in set_protocol() and suppports_mem_op() to
ignore empty spi_mem_op phases, as checking for dtr field in
empty phase will result in false negatives.
Signed-off-by: Apurva Nandan <a-nandan@ti.com>
Link: https://lore.kernel.org/r/20210716232504.182-3-a-nandan@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add an else-if condition in the probe function to check whether ngpios is
multiple of 8.
Per AST datasheet, numbers of available serial GPIO pins in Serial GPIO
Configuration Register must be n bytes. For instance, if n = 1, it means
AST SoC supports 8 GPIO pins.
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Reviewed-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
Replace all of_property_read_u32() with device_property_read_u32().
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Acked-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
The current design initializes irq->chip from a global irqchip struct,
which causes multiple sgpio devices use the same irq_chip.
The patch moves irq_chip to aspeed_sgpio struct for initializing
irq_chip from their private gpio struct.
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Reviewed-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
AST SoC supports *retain pin state* function when wdt reset.
The patch adds set_config function for handling sgpio reset tolerance
register.
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Reviewed-by: Andrew Jeffery <andrew@aj.id.au>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
The maximum number of gpio pins of SoC is hardcoded as 80 and the gpio pin
count mask for GPIO Configuration register is hardcode as GENMASK(9,6).
However, AST2600 has 2 sgpio master interfaces, one of them supports up
to 128 gpio pins and pin count mask of GPIO Configuration Register is 5
bits.
The patch adds ast2600 compatibles, removes MAX_NR_HW_SGPIO and
corresponding design to make the gpio input/output pin base are determined
by ngpios.
The patch also removed hardcoded pin mask and adds ast2400, ast2500,
ast2600 platform data that include gpio pin count mask for GPIO
Configuration Register.
The original pin order is as follows:
(suppose MAX_NR_HW_SGPIO is 80 and ngpios is 10 as well)
Input:
0 1 2 3 ... 9
Output:
80 81 82 ... 89
The new pin order is as follows:
Input:
0 2 4 6 ... 18
Output:
1 3 5 7 ... 19
SGPIO pin id and input/output pin mapping is as follows:
SGPIO0(0,1), SGPIO1(2,3), ..., SGPIO79(158,159)
For example:
Access SGPIO5(10,11)
Get SGPIO pin 5 (suppose sgpio chip id is 2)
gpioget 2 10
Set SGPIO pin 5 (suppose sgpio chip id is 2)
gpioset 2 11=1
gpioset 2 11=0
Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
|
|
Clang warns:
drivers/staging/r8188eu/core/rtw_mlme.c:1629:28: warning: address of
array 'pmlmepriv->assoc_ssid.Ssid' will always evaluate to 'true'
[-Wpointer-bool-conversion]
if (pmlmepriv->assoc_ssid.Ssid && pmlmepriv->assoc_ssid.SsidLength) {
~~~~~~~~~~~~~~~~~~~~~~^~~~ ~~
1 warning generated.
Ssid is an array not at the beginning of a struct so its address cannot
be NULL so remove the check.
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20210805185807.3296077-4-nathan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Clang warns:
drivers/staging/r8188eu/hal/rtl8188e_rf6052.c:344:13: warning:
explicitly assigning value of variable of type 'u32' (aka 'unsigned
int') to itself [-Wself-assign]
writeVal = writeVal;
~~~~~~~~ ^ ~~~~~~~~
1 warning generated.
Remove this branch as it is useless.
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20210805185807.3296077-3-nathan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Clang warns several times across the driver along the lines of:
drivers/staging/r8188eu/core/rtw_pwrctrl.c:222:21: warning: equality
comparison with extraneous parentheses [-Wparentheses-equality]
if ((pwrpriv->rpwm == pslv)) {
~~~~~~~~~~~~~~^~~~~~~
drivers/staging/r8188eu/core/rtw_pwrctrl.c:222:21: note: remove
extraneous parentheses around the comparison to silence this warning
if ((pwrpriv->rpwm == pslv)) {
~ ^ ~
drivers/staging/r8188eu/core/rtw_pwrctrl.c:222:21: note: use '=' to turn
this equality comparison into an assignment
if ((pwrpriv->rpwm == pslv)) {
^~
=
1 warning generated.
The compilers have agreed that single parentheses are used for equality
and double parentheses are used for assignment within control flow
statements such as if and while so remove them in these places to fix
the warning.
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20210805185807.3296077-2-nathan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Add support for the LOGIC Technologies, Inc LTTD800480070-L6WH-RT
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Co-developed-by: Søren Andersen <san@skov.dk>
Signed-off-by: Søren Andersen <san@skov.dk>
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210805111944.13533-3-o.rempel@pengutronix.de
|
|
The Multi Innotechnology is a 10.1" 1280x800 panel.
The datasheet did not specify specific values for sync, back, front porch.
The values are a best guess based on values for similar panels.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Co-developed-by: Ulrich Ölmann <u.oelmann@pengutronix.de>
Signed-off-by: Ulrich Ölmann <u.oelmann@pengutronix.de>
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210805111944.13533-2-o.rempel@pengutronix.de
|
|
This routine is just a wrapper around kfree(). The wrapper also sets
buffer length to 0, and the pointer to NULL. The length variable is in
the struct to be freed, thus setting it to zero is meaningless. Setting
the pointer to NULL is also not needed.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805183717.23007-6-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
After removing some empty routines in os_dep/osdep_service.c, new routines
become empty. These include rtw_mfree_stainfo(),
_rtw_free_sta_xmit_priv_lock(), and rtw_mfree_recv_priv_lock.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805183717.23007-5-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This routine is empty, thus all calls to it can be removed.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805183717.23007-4-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
This call is so simple that there is no need, or use, for a wrapper.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805183717.23007-3-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
These wrappers involve an in_interrupt() call, which is not frowned on
in the kernel. These changes decrease the size of the driver by a trivial
amount.
Signed-off-by: Larry Finger <Larry.Finger@lwfinger.net>
Link: https://lore.kernel.org/r/20210805183717.23007-2-Larry.Finger@lwfinger.net
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
GCC + Sparse emit warnings of passing incorrect type in arguments of
some functions because of different base types. Fix them by changing
the types of the parameters of the above-mentioned functions.
In the meantime, remove the unnecessary casts of those arguments
which are then passed to memcpy() within those same functions.
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Link: https://lore.kernel.org/r/20210805131108.19775-1-fmdefrancesco@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Remove unnecessary label to fix build error introduced with
commit b398ff88aa36 ("staging: r8188eu: remove return from void functions")
drivers/staging/r8188eu/hal/rtl8188e_dm.c:182:1: error: label at end of compound statement
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20210805140941.9130-1-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
|
|
Pull pcmcia fix from Dominik Brodowski:
"Zheyu Ma found and fixed a null pointer dereference bug in the device
driver for the i82092 card reader"
* 'pcmcia-next' of git://git.kernel.org/pub/scm/linux/kernel/git/brodo/linux:
pcmcia: i82092: fix a null pointer dereference bug
|
|
We noticed that the user interface of Android devices becomes very slow
under memory pressure. This is because Android uses the zram driver on top
of the loop driver for swapping, because under memory pressure the swap
code alternates reads and writes quickly, because mq-deadline is the
default scheduler for loop devices and because mq-deadline delays writes by
five seconds for such a workload with default settings. Fix this by making
the kernel select I/O scheduler 'none' from inside add_disk() for loop
devices. This default can be overridden at any time from user space,
e.g. via a udev rule. This approach has an advantage compared to changing
the I/O scheduler from userspace from 'mq-deadline' into 'none', namely
that synchronize_rcu() does not get called.
This patch changes the default I/O scheduler for loop devices from
'mq-deadline' into 'none'.
Additionally, this patch reduces the Android boot time on my test setup
with 0.5 seconds compared to configuring the loop I/O scheduler from user
space.
Cc: Christoph Hellwig <hch@lst.de>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: Martijn Coenen <maco@android.com>
Cc: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Link: https://lore.kernel.org/r/20210805174200.3250718-3-bvanassche@acm.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The I2C-bus to the XPower AXP288 is shared between the Linux kernel and
the SoCs P-Unit. The P-Unit has a semaphore which the kernel must "lock"
before it may use the bus. If not explicitly taken by the I2C-driver,
then this semaphore is automatically taken by the I2C-bus-driver for
each I2C-transfer and this is a quite expensive operation.
Explicitly take the semaphore in probe() around the register-accesses
done during probe, so that this only needs to be done once, rather then
once per register-access.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
The I2C-bus to the XPower AXP288 is shared between the Linux kernel and
the SoCs P-Unit. The P-Unit has a semaphore which the kernel must "lock"
before it may use the bus. If not explicitly taken by the I2C-driver,
then this semaphore is automatically taken by the I2C-bus-driver for
each I2C-transfer.
Move the AXP20X_CC_CTRL check done in probe() together with the other
register-accesses done in probe, so that we can take the semaphore once
for the entire set of register-accesses.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
The I2C-bus to the XPower AXP288 is shared between the Linux kernel and
the SoCs P-Unit. The P-Unit has a semaphore which the kernel must "lock"
before it may use the bus and while the kernel holds the semaphore the CPU
and GPU power-states must not be changed otherwise the system will freeze.
This is a complex process, which is quite expensive. This is all done by
iosf_mbi_block_punit_i2c_access(). To ensure that no unguarded I2C-bus
accesses happen, iosf_mbi_block_punit_i2c_access() gets called by the
I2C-bus-driver for every I2C transfer. Because this is so expensive it
is allowed to call iosf_mbi_block_punit_i2c_access() in a nested
fashion, so that higher-level code which does multiple I2C-transfers can
call it once for a group of transfers, turning the calls done by the
I2C-bus-driver into no-ops.
Userspace power-supply API users typically will read all provided
properties in one go, refreshing the last read values when
power_supply_changed() is called by the driver and/or periodically
(e.g. every 2 minutes).
The reading of all properties in one go causes the P-Unit semaphore
to quickly be taken and released multiple times in a row. Certain
PMIC registers like AXP20X_FG_RES are even used in multiple properties
so they get read multiple times, leading to a P-Unit take + release
each time the register is read.
As already mentioned the taking of the P-Unit semaphore is a quite
expensive operation and it has also been reported that the
"hammering" of the P-Unit semaphore done by the axp288_fuel_gauge
driver can even cause stability issues with the system as a whole.
Switch over to a scheme where the axp288_fuel_gauge driver keeps
a local copy of all the registers which it uses for properties
and make it only refresh its copy of the registers if the values
are older then 1 minute; or when a fuel-gauge interrupt has
triggered since the last read.
This not only reduces the amount of reads, it also makes the code
do all the reads in one go, rather then reading specific registers
based on which property is being queried. This allows calling
iosf_mbi_block_punit_i2c_access() once before doing all the reads,
so that we now only take the P-Unit semaphore once per update.
Tested-by: Andrejus Basovas <cpp@gcc.lt>
Signed-off-by: Andrejus Basovas <cpp@gcc.lt>
Co-developed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
Accessing registers on the AXP288 is quite expensive, so we should avoid
doing unnecessary accesses.
The FG_LOW_CAP_REG never changes underneath us, so we only need to read
it once. Devices with an AXP288 do not have user-replace (let alone
hot-swappable) batteries and the only bit we care about in the
PWR_OP_MODE register is the CHRG_STAT_BAT_PRESENT bit, so we can get
away with only reading the PWR_OP_MODE register once too.
Note that the FG_LOW_CAP_REG is not marked volatile in the regmap, so we
were effectively already reading it once. This change makes this explicit,
this is done as preparation of a further patch which moves all remaining
register accesses in fuel_gauge_get_property() out of that function.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
Directly store the struct device pointer in axp288_fg_info, rather then
storing a pointer to the struct platform_device there and then using
"&info->pdev->dev" everywhere.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
The I2C-bus to the XPower AXP288 is shared between the Linux kernel and
the SoCs P-Unit. The P-Unit has a semaphore which the kernel must "lock"
before it may use the bus. This semaphore is automatically taken by the
I2C-bus-driver.
The retry on -EBUSY logic in fuel_gauge_reg_readb() likely was added to
deal with the I2C-bus-drive returning -EBUSY when it failed to take the
semaphore, but this really should never happen. The semaphore code even
has a WARN_ON(ret) to log a kernel backtrace if this does somehow happen,
when this happens something is seriously wrong and the system typically
freezes soon afterwards.
TL;DR: the regmap_read() should never fail with -EBUSY so the retries
are unnecessary.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
When fuel_gauge_reg_readb()/_writeb() fails, report which register we
were trying to read / write when the error happened.
Also reword the message a bit:
- Drop the axp288 prefix, dev_err() already prints this
- Switch from telegram / abbreviated style to a normal sentence, aligning
the message with those from fuel_gauge_read_*bit_word()
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
|
Drop the IRQ mapping messages, because they are really not
interesting at all.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
|
Mark Bloch
Roi Dayan
Ariel Levkovich
Larry Finger
Akhil R
Linus Torvalds
Krzysztof Kozlowski
Apurva Nandan
Steven Lee
Nathan Chancellor
Søren Andersen
Sam Ravnborg
Fabio M. De Francesco
Michael Straube
Bart Van Assche
Hans de Goede
Andrejus Basovas