We want to differentiate between the DFP dotclock and TMDS clock
limits. Let's convert the current thing to just give us the
dotclock limit.
v2: Use Returns: for kdoc (Lyude)
Fix up nouveau code too
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200904115354.25336-9-ville.syrjala@linux.intel.com
Reviewed-by: Lyude Paul <lyude@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
q6afe already exposes lpass clocks, however this was not presented
as proper clock controller driver. This patch basically adds clock
controller support for q6afe clocks.
This is useful for other drivers like lpass digital codec or lpass
lowpower island drivers to request or vote for these clocks.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20200910135708.14842-3-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
q6afe exposes various lpass clocks controls via q6dsp q6afe commands.
This patch adds bindings required for this clock controller.
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20200910135708.14842-2-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Add support for the hardware monitoring controller of the sl28cpld board
management controller. This driver is part of a multi-function device.
Signed-off-by: Michael Walle <michael@walle.cc>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add support for the GPIO controller of the sl28 board management
controller. This driver is part of a multi-function device.
A controller has 8 lines. There are three different flavors:
full-featured GPIO with interrupt support, input-only and output-only.
Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add support for the PWM controller of the sl28cpld board management
controller. This is part of a multi-function device driver.
The controller has one PWM channel and can just generate four distinct
frequencies.
Signed-off-by: Michael Walle <michael@walle.cc>
Acked-by: Thierry Reding <thierry.reding@gmail.com>
Reviewed-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add support for the watchdog of the sl28cpld board management
controller. This is part of a multi-function device driver.
Signed-off-by: Michael Walle <michael@walle.cc>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add support for the interrupt controller inside the sl28 CPLD management
controller.
The interrupt controller can handle at most 8 interrupts and is really
simplistic and consists only of an interrupt mask and an interrupt
pending register.
Signed-off-by: Michael Walle <michael@walle.cc>
Acked-by: Marc Zyngier <maz@kernel.org>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add the core support for the board management controller found on the
SMARC-sAL28 board.
Also add a virtual symbol which pulls in the simple-mfd-i2c driver and
provide a common symbol on which the subdevice drivers can depend on.
At the moment, this controller is used on the Kontron SMARC-sAL28 board.
Signed-off-by: Michael Walle <michael@walle.cc>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Add a device tree bindings for the board management controller found on
the Kontron SMARC-sAL28 board.
Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
There are I2C devices which contain several different functions but
doesn't require any special access functions. For these kind of drivers
an I2C regmap should be enough.
Create an I2C driver which creates an I2C regmap and enumerates its
children. If a device wants to use this as its MFD core driver, it has
to add an individual compatible string. It may provide its own regmap
configuration.
Subdevices can use dev_get_regmap() on the parent to get their regmap
instance.
Signed-off-by: Michael Walle <michael@walle.cc>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Previous commits reduced the verbosity of errors during boot
iterations, but there are still a couple remaining which generate
false positives. Errors should only be logged when after last attempt
to download firmware failed.
Duplicating logs and assigning them different levels based on the
iteration number isn't really elegant, use macro as suggested by
Guennadi.
Suggested-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-9-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
On multiple locations checks are performed of untrusted values after adding
a constant to them. This is wrong, because the addition might overflow and
the result can then pass the check, although the original value is invalid.
Fix multiple such issues by checking the actual value and not a sum of it
and a constant.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-8-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
No need to type-cast assignments between void and other pointers in C.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-7-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
In some cases no "ret" variable is even needed, those functions always
return 0 anyway, in other cases "ret" initialisation is redundant.
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-6-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
We need to avoid reporting an error for -EACCESS when pm_runtime is
not enabled.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-5-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
We need to avoid reporting an error for -EACCESS when pm_runtime is
not enabled.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-4-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This field will be used to compare ldc file with loaded fw version,
to assert validity of trace logs. Value used in sof-logger.
Signed-off-by: Karol Trzcinski <karolx.trzcinski@linux.intel.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-3-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
This patch adds debug support for imx platforms. This is important in
order to gather information about the state of the DSP in case of an
oops and the reason for the oops.
This is done by checking if a message with a panic code has been placed
in the debug box, in the imx8_dsp_handle_request function from sof/imx.
If positive, the function imx8_dump, added in common, will be called.
The first step is to gather information about the registers, filename,
line number and stack by calling the imx8_get_registers, added in common.
Then the information will be printed to the console by calling the
get_status function.
Signed-off-by: Iulian Olaru <iulianolaru249@yahoo.com>
Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Reviewed-by: Daniel Baluta <daniel.baluta@gmail.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917105633.2579047-2-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Extern is the default attribute for functions anyway.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/390972/
Deal with more cases in drm_dp_downstream_max_bpc():
- DPCD 1.0 -> assume 8bpc for non-DP
- DPCD 1.1+ DP (or DP++ with DP sink) -> allow anything
- DPCD 1.1+ TMDS -> check the caps, assume 8bpc if the value is crap
- anything else -> assume 8bpc
v2: Use Returns: for kdoc (Lyude)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200904115354.25336-8-ville.syrjala@linux.intel.com
Reviewed-by: Lyude Paul <lyude@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Stash the downstream facing port max bpc away during
intel_dp_set_edid(). We'll soon need the EDID in there so
we can't figure this out so easily during .compute_config() anymore.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200904115354.25336-6-ville.syrjala@linux.intel.com
Reviewed-by: Lyude Paul <lyude@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Non-HDMI sinks shouldn't be sent infoframes. Check for that when
using LSPCON.
FIXME: How do we turn off infoframes once enabled? Do we even
have to?
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200904115354.25336-3-ville.syrjala@linux.intel.com
Reviewed-by: Lyude Paul <lyude@redhat.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Fix the BCLK inversion for DSP modes
This is how it is defined by ASoC:
* BCLK:
* - "normal" polarity means signal is available at rising edge of BCLK
* - "inverted" polarity means signal is available at falling edge of BCLK
The adcx140 defines the BCLK edge based on coding type.
The PCM (DSP_A/B) should drive on rising and sample on falling edge, so
from ASoC pov, it is IB_NF. But from the codec pov if it is configured in
DSP mode, then the BCLK should not be inverted, defaults to the coding
standard.
For i2s, it is NB_NF from ASoC pov (drive on falling, sample on rising).
>From the codec's pov BCLK should not invert either, as this is the default
for the coding.
So, inversion must take the format into account:
IB_NF + DSP_A/B == the codec bclk inversion should be disabled
NB_NF + DSP_A/B == the codec bclk inversion should be enabled
NB_NF + I2S == the codec bclk inversion should be disabled
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200915190606.1744-2-dmurphy@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It was observed that if the device was active and register writes were
performed there were some unwanted behaviors particularly when writing
the word length and some filter options. So when writing to the device
the device should be placed in sleep mode and then exit sleep mode once
the register update is complete.
Signed-off-by: Dan Murphy <dmurphy@ti.com>
Link: https://lore.kernel.org/r/20200915190606.1744-1-dmurphy@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
According to its datasheet, the digital gain should be -100 dB when
CHx_DVOL is 1 and 27 dB when CHx_DVOL is 255. But with the current
dig_vol_tlv, "Digital CHx Out Volume" shows 27.5 dB if CHx_DVOL is 255
and -95.5 dB if CHx_DVOL is 1. This commit fixes this bug.
Fixes: 689c7655b5 ("ASoC: tlv320adcx140: Add the tlv320adcx140 codec driver family")
Signed-off-by: Camel Guo <camelg@axis.com>
Link: https://lore.kernel.org/r/20200908090417.16695-1-camel.guo@axis.com
Signed-off-by: Mark Brown <broonie@kernel.org>
When CONFIG_SND_CTL_VALIDATION is set, accesses to extended bytes
control generate spurious error messages when the size exceeds 512
bytes, such as
[ 11.224223] sof_sdw sof_sdw: control 2:0:0:EQIIR5.0 eqiir_coef_5:0:
invalid count 1024
In addition the error check returns -EINVAL which has the nasty side
effect of preventing applications accessing controls from working,
e.g.
root@plb:~# alsamixer
cannot load mixer controls: Invalid argument
It's agreed that the control interface has been abused since 2014, but
forcing a check should not prevent existing solutions from working.
This patch skips the checks conditionally if CONFIG_SND_CTL_VALIDATION
is set and the byte array provided by topology is > 512. This
preserves the checks for all other cases.
Fixes: 1a3232d2f6 ('ASoC: topology: Add support for TLV bytes controls')
BugLink: https://github.com/thesofproject/linux/issues/2430
Reported-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Jaska Uimonen <jaska.uimonen@intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20200917103912.2565907-1-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
"alsactl -f state.conf store/restore" sequence fails because setting
"mixing clock source" and "output source" requires active TDM clock
being started for configuration propagation. Make these two controls
write only so that their values are not stored at "alsactl store".
Signed-off-by: Viorel Suman <viorel.suman@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Link: https://lore.kernel.org/r/1600104274-13110-1-git-send-email-viorel.suman@oss.nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The rcar-gen3-cpg driver is also used on Renesas RZ/G2 SoC's, update the
description for the CLK_RCAR_GEN3_CPG config symbol to reflect this.
Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Chris Paterson <Chris.Paterson2@renesas.com>
Link: https://lore.kernel.org/r/20200911101703.20521-1-prabhakar.mahadev-lad.rj@bp.renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Initial support for R-Car V3U (r8a779a0), including core, module
clocks, resets, and register access, because register specification
differs from R-Car Gen2/3.
Inspired by patches in the BSP by LUU HOAI.
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Link: https://lore.kernel.org/r/1599810232-29035-4-git-send-email-yoshihiro.shimoda.uh@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Clock and Power Domain definitions for the Renesas R-Car V3U (R8A779A0)
SoC, shared by driver and DT source files.
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQQ9qaHoIs/1I4cXmEiKwlD9ZEnxcAUCX1pckgAKCRCKwlD9ZEnx
cCbUAQCBb4x5zMgZsZnAPGyK6M1CyV3KC3UvH5OBYhDQRC//uAEAqUT6R21KbZbe
fYWzy8rcmSgjHPyF737DmYzYS9nRMgE=
=/t2Y
-----END PGP SIGNATURE-----
Merge tag 'renesas-r8a779a0-dt-binding-defs-tag' into clk-renesas-for-v5.10
Renesas R-Car V3U DT Binding Definitions
Clock and Power Domain definitions for the Renesas R-Car V3U (R8A779A0)
SoC, shared by driver and DT source files.
To support other register layouts in the future, add register pointers
of {control,status,reset,reset_clear}_regs into struct cpg_mssr_priv.
After that, we can remove unused macros like MSTPSR(). No behavioral
changes.
Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com>
Link: https://lore.kernel.org/r/1599810232-29035-3-git-send-email-yoshihiro.shimoda.uh@renesas.com
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
GPU drivers need this in their shrinkers, to be able to throw out
mmap'ed buffers. Note that we also need dma_resv_lock in shrinkers,
but that loop is resolved by trylocking in shrinkers.
So full hierarchy is now (ignore some of the other branches we already
have primed):
mmap_read_lock -> dma_resv -> shrinkers -> i_mmap_lock_write
I hope that's not inconsistent with anything mm or fs does, adding
relevant people.
Reviewed-by: Thomas Hellström <thomas.hellstrom@intel.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: linux-media@vger.kernel.org
Cc: linaro-mm-sig@lists.linaro.org
Cc: Dave Chinner <david@fromorbit.com>
Cc: Qian Cai <cai@lca.pw>
Cc: linux-xfs@vger.kernel.org
Cc: linux-fsdevel@vger.kernel.org
Cc: Thomas Hellström (Intel) <thomas_os@shipmail.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Jason Gunthorpe <jgg@mellanox.com>
Cc: linux-mm@kvack.org
Cc: linux-rdma@vger.kernel.org
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200728135839.1035515-1-daniel.vetter@ffwll.ch
The spinlock ap_poll_timer_lock is initialized statically. It is
unnecessary to initialize by spin_lock_init().
Signed-off-by: Qinglang Miao <miaoqinglang@huawei.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Use DEFINE_SEQ_ATTRIBUTE macro to simplify the code.
Signed-off-by: Liu Shixin <liushixin2@huawei.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
This patch reworks the zcrypt device driver so that the set_fs()
invocation is not needed any more. Instead there is a new flag bool
userspace passed through all the functions which tells if the pointer
arguments are userspace or kernelspace. Together with the two new
inline functions z_copy_from_user() and z_copy_to_user() which either
invoke copy_from_user (userspace is true) or memcpy (userspace is
false) the zcrypt dd and the AP bus now has no requirement for
the set_fs() functionality any more.
Signed-off-by: Harald Freudenberger <freude@linux.ibm.com>
Reviewed-by: Ingo Franzki <ifranzki@linux.ibm.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
The roce path triggers a work queue that continues to touch the id_priv
but doesn't hold any reference on it. Futher, unlike in the IB case, the
work queue is not fenced during rdma_destroy_id().
This can trigger a use after free if a destroy is triggered in the
incredibly narrow window after the queue_work and the work starting and
obtaining the handler_mutex.
The only purpose of this work queue is to run the ULP event callback from
the standard context, so switch the design to use the existing
cma_work_handler() scheme. This simplifies quite a lot of the flow:
- Use the cma_work_handler() callback to launch the work for roce. This
requires generating the event synchronously inside the
rdma_join_multicast(), which in turn means the dummy struct
ib_sa_multicast can become a simple stack variable.
- cm_work_handler() used the id_priv kref, so we can entirely eliminate
the kref inside struct cma_multicast. Since the cma_multicast never
leaks into an unprotected work queue the kfree can be done at the same
time as for IB.
- Eliminating the general multicast.ib requires using cma_set_mgid() in a
few places to recompute the mgid.
Fixes: 3c86aa70bf ("RDMA/cm: Add RDMA CM support for IBoE devices")
Link: https://lore.kernel.org/r/20200902081122.745412-9-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Two places were open coding this sequence, and also pull in
cma_leave_roce_mc_group() which was called only once.
Link: https://lore.kernel.org/r/20200902081122.745412-8-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
There is no kernel user of RDMA CM multicast so this code managing the
multicast subscription of the kernel-only internal QP is dead. Remove it.
This makes the bug fixes in the next patches much simpler.
Link: https://lore.kernel.org/r/20200902081122.745412-7-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
These are the same thing, except that cma_ndev_work doesn't have a state
transition. Signal no state transition by setting old_state and new_state
== 0.
In all cases the handler function should not be called once
rdma_destroy_id() has progressed passed setting the state.
Link: https://lore.kernel.org/r/20200902081122.745412-6-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
The only place that still uses it is rdma_join_multicast() which is only
doing a sanity check that the caller hasn't done something wrong and
doesn't need the spinlock.
At least in the case of rdma_join_multicast() the information it needs
will remain until the ID is destroyed once it enters these
states. Similarly there is no reason to check for these specific states in
the handler callback, instead use the usual check for a destroyed id under
the handler_mutex.
Link: https://lore.kernel.org/r/20200902081122.745412-5-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
There is a strange unlocked read of the ID state when checking for
reuseaddr. This is because an ID cannot be reusable once it becomes a
listening ID. Instead of using the state to exclude reuse, just clear it
as part of rdma_listen()'s flow to convert reusable into not reusable.
Once a ID goes to listen there is no way back out, and the only use of
reusable is on the bind_list check.
Finally, update the checks under handler_mutex to use READ_ONCE and audit
that once RDMA_CM_LISTEN is observed in a req callback it is stable under
the handler_mutex.
Link: https://lore.kernel.org/r/20200902081122.745412-4-leon@kernel.org
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
