Merge tag 'memory-controller-drv-5.10' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux-mem-ctrl into arm/drivers

Memory controller drivers for v5.10

1. Fixes in several drivers for GCC warnings, including kerneldoc fixes
   and issues discovered while compile testing.
2. Enable compile testing of most of the drivers.
3. Use dev_err_probe() to simplify the code.
4. omap-gpmc: fix off by one errors, code cleanups and improvements.
5. tegra: remove the GPU from DRM IOMMU group so it would use its own;
   few minor fixes.
6. brcmstb_dpfe: fix memory leak and array index out of bounds.

* tag 'memory-controller-drv-5.10' of https://git.kernel.org/pub/scm/linux/kernel/git/krzk/linux-mem-ctrl: (26 commits)
  memory: fsl-corenet-cf: Fix handling of platform_get_irq() error
  memory: omap-gpmc: Fix -Wunused-function warnings
  memory: tegra: Remove GPU from DRM IOMMU group
  memory: tegra186-emc: Simplify with dev_err_probe()
  memory: brcmstb_dpfe: Simplify with dev_err_probe()
  memory: samsung: exynos5422-dmc: add missing and fix kerneldoc
  memory: samsung: exynos5422-dmc: remove unused exynos5_dmc members
  memory: samsung: exynos5422-dmc: rename timing register fields variables
  memory: emif: Remove bogus debugfs error handling
  memory: omap-gpmc: Fix build error without CONFIG_OF
  memory: omap-gpmc: Fix a couple off by ones
  memory: brcmstb_dpfe: fix array index out of bounds
  memory: brcmstb_dpfe: Fix memory leak
  memory: tegra: Correct shift value of apew
  memory: Enable compile testing for most of the drivers
  memory: brcmstb_dpfe: add separate entry for compile test
  memory: tegra: tegra210-emc: fix indentation
  memory: renesas-rpc-if: simplify with PTR_ERR_OR_ZERO
  memory: omap-gpmc: consistently use !res for NULL checks
  memory: omap-gpmc: use WARN() instead of BUG() on wrong free
  ...

Link: https://lore.kernel.org/r/20200907150611.11267-1-krzk@kernel.orgSigned-off-by: Olof Johansson <olof@lixom.net>

Merge tag 'renesas-drivers-for-v5.10-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel into arm/drivers

Renesas driver updates for v5.10

  - Improve visual Kconfig structure.

* tag 'renesas-drivers-for-v5.10-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/renesas-devel:
  soc: renesas: Align driver description titles
  soc: renesas: Use menu for Renesas SoC

Link: https://lore.kernel.org/r/20200904114819.30254-4-geert+renesas@glider.beSigned-off-by: Olof Johansson <olof@lixom.net>

Merge tag 'optee-i2c-fix-for-v5.10' of git://git.linaro.org:/people/jens.wiklander/linux-tee into arm/drivers

Make sure I2C functions used in OP-TEE are reachable with IS_REACHABLE()

* tag 'optee-i2c-fix-for-v5.10' of git://git.linaro.org:/people/jens.wiklander/linux-tee:
  drivers: optee: fix i2c build issue

Link: https://lore.kernel.org/r/20200901101806.GA3286324@jadeSigned-off-by: Olof Johansson <olof@lixom.net>

platform_get_irq() returns -ERRNO on error.  In such case comparison
to 0 would pass the check.

Fixes: 54afbec0 ("memory: Freescale CoreNet Coherency Fabric error reporting driver")
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200827073315.29351-1-krzk@kernel.org

If CONFIG_OF is not set, make W=1 warns:

  drivers/memory/omap-gpmc.c:987:12: warning: ‘gpmc_cs_remap’ defined but not used [-Wunused-function]
  drivers/memory/omap-gpmc.c:926:20: warning: ‘gpmc_cs_get_name’ defined but not used [-Wunused-function]
  drivers/memory/omap-gpmc.c:919:13: warning: ‘gpmc_cs_set_name’ defined but not used [-Wunused-function]

Move them to #ifdef CONFIG_OF block to fix this.
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Roger Quadros <rogerq@ti.com>
Link: https://lore.kernel.org/r/20200901112832.3084-1-yuehaibing%40huawei.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Commit 63a613fd ("memory: tegra: Add gr2d and gr3d to DRM IOMMU
group") added the GPU to the DRM IOMMU group, which doesn't make any
sense. This causes problems when Nouveau tries to attach to the SMMU
and causes it to fall back to using the DMA API.

Remove the GPU from the DRM groups to restore the old behaviour. The
GPU should always have its own IOMMU domain to make sure it can map
buffers into contiguous chunks (for big page support) without getting
in the way of mappings from the DRM group.

Cc: <stable@vger.kernel.org>
Fixes: 63a613fd ("memory: tegra: Add gr2d and gr3d to DRM IOMMU group")
Reported-by: Matias Zuniga <matias.nicolas.zc@gmail.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
Reviewed-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20200901153248.1831263-1-thierry.reding@gmail.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Common pattern of handling deferred probe can be simplified with
dev_err_probe().  Less code and the error value gets printed.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200828153747.22358-2-krzk@kernel.org

Common pattern of handling deferred probe can be simplified with
dev_err_probe().  Less code and the error value gets printed.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Markus Mayer <mmayer@broadcom.com>
Link: https://lore.kernel.org/r/20200828153747.22358-1-krzk@kernel.org

Add missing kerneldoc to struct exynos5_dmc and correct the existing
kerneldoc in other places to fix W=1 warnings like:

    drivers/memory/samsung/exynos5422-dmc.c:107: warning: Function parameter or member 'freq_hz' not described in 'dmc_opp_table'
    drivers/memory/samsung/exynos5422-dmc.c:154: warning: Function parameter or member 'dev' not described in 'exynos5_dmc'
    drivers/memory/samsung/exynos5422-dmc.c:357: warning: Excess function parameter 'param' description in 'exynos5_set_bypass_dram_timings'
    drivers/memory/samsung/exynos5422-dmc.c:630: warning: Function parameter or member 'flags' not described in 'exynos5_dmc_get_volt_freq'
    drivers/memory/samsung/exynos5422-dmc.c:962: warning: cannot understand function prototype: 'struct devfreq_dev_profile exynos5_dmc_df_profile = '
    drivers/memory/samsung/exynos5422-dmc.c:1011: warning: Function parameter or member 'reg_timing_row' not described in 'create_timings_aligned'
    drivers/memory/samsung/exynos5422-dmc.c:1011: warning: Excess function parameter 'idx' description in 'create_timings_aligned'
    drivers/memory/samsung/exynos5422-dmc.c:1345: warning: Excess function parameter 'set' description in 'exynos5_dmc_set_pause_on_switching'
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Lukasz Luba <lukasz.luba@arm.com>
Link: https://lore.kernel.org/r/20200822163218.21857-3-krzk@kernel.org

The struct exynos5_dmc members bypass_rate, mx_mspll_ccore_phy,
mout_mx_mspll_ccore_phy and opp_bypass are not actually used.

Apparently there was a plan to store the OPP for the bypass mode in
opp_bypass member, but driver fails to do it and instead always sets
target voltage during bypass mode.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200822163218.21857-2-krzk@kernel.org

The driver has file-scope arrays defining fields of timing registers
(e.g. struct timing_reg timing_row) and actual values for these
registers per each OPP in state container (struct
exynos5_dmc.timing_row).  The meanings of these are different so use
different names to avoid confusion.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Lukasz Luba <lukasz.luba@arm.com>
Link: https://lore.kernel.org/r/20200822163218.21857-1-krzk@kernel.org

When the optee driver is compiled into the kernel while the i2c core
is configured as a module, the i2c symbols are not available.

This commit addresses the situation by disabling the i2c support for
this use case while allowing it in all other scenarios:

 i2c=y, optee=y
 i2c=m, optee=m
 i2c=y, optee=m
 i2c=m, optee=y (not supported)

Fixes: c05210ab ("drivers: optee: allow op-tee to access devices on the i2c bus")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Jorge Ramirez-Ortiz <jorge@foundries.io>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

Now, Renesas SoC drivers are under menu, but current descriptions are
not aligned.
This patch aligns them.

    - Emma Mobile EV2
    - RZ/A1H (R7S72100)
      ...
    - R-Car H2 (R8A77900)
      ...
    - Renesas R-Car H3 ES1.x SoC Platform
      ...
    - R-Car H2 System Controller support
    - R-Car M2-W/N System Controller support
    - R-Car V2H System Controller support
    - R-Car E2 System Controller support
    - R-Car H3 System Controller support
    - R-Car M3-W System Controller support
    - R-Car M3-W+ System Controller support
    - R-Car M3-N System Controller support

    + SoC Platform support for Emma Mobile EV2
    + SoC Platform support for RZ/A1H+
      ...
    + SoC Platform support for R-Car H2
      ...
    + SoC Platform support for R-Car H3 ES1.x
      ...
    + System Controller support for R-Car H2
    + System Controller support for R-Car M2-W/N
    + System Controller support for R-Car V2H
    + System Controller support for R-Car E2
    + System Controller support for R-Car H3
    + System Controller support for R-Car M3-W
    + System Controller support for R-Car M3-W+
    + System Controller support for R-Car M3-N
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87zh6kyedc.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

Callers are generally not supposed to check the return values from
debugfs functions.  Debugfs functions never return NULL so this error
handling will never trigger.  (Historically debugfs functions used to
return a mix of NULL and error pointers but it was eventually deemed too
complicated for something which wasn't intended to be used in normal
situations).

Delete all the error handling.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Santosh Shilimkar <ssantosh@kernel.org>
Link: https://lore.kernel.org/r/20200826113759.GF393664@mwandaSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

If CONFIG_OF is n, gcc fails:

drivers/memory/omap-gpmc.o: In function `gpmc_omap_onenand_set_timings':
    omap-gpmc.c:(.text+0x2a88): undefined reference to `gpmc_read_settings_dt'

Add gpmc_read_settings_dt() helper function, which zero the gpmc_settings
so the caller doesn't proceed with random/invalid settings.

Fixes: a758f50f ("mtd: onenand: omap2: Configure driver from DT")
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Roger Quadros <rogerq@ti.com>
Link: https://lore.kernel.org/r/20200827125316.20780-1-yuehaibing@huawei.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Merge tag 'tee-pin-user-pages-for-5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee into arm/drivers

Converts tee subsystem to use pin_user_pages() instead of get_user_pages()

* tag 'tee-pin-user-pages-for-5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee:
  tee: convert get_user_pages() --> pin_user_pages()

Link: https://lore.kernel.org/r/20200825090715.GA2370775@jadeSigned-off-by: Olof Johansson <olof@lixom.net>

These comparisons should be >= instead of > to prevent reading one
element beyond the end of the gpmc_cs[] array.

Fixes: cdd6928c ("ARM: OMAP2+: Add device-tree support for NOR flash")
Fixes: f37e4580 ("ARM: OMAP2: Dynamic allocator for GPMC memory space")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Roger Quadros <rogerq@ti.com>
Link: https://lore.kernel.org/r/20200825104707.GB278587@mwandaSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Renesas related SoC settings are located on TOP level menu,
thus it is very verbose.
This patch groups Renesas related settings into
"Renesas SoC driver support" menu.

And it aligns config menu names.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87k0xxho7t.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: Geert Uytterhoeven <geert+renesas@glider.be>

This code was using get_user_pages*(), in a "Case 2" scenario
(DMA/RDMA), using the categorization from [1]. That means that it's
time to convert the get_user_pages*() + put_page() calls to
pin_user_pages*() + unpin_user_pages() calls.

Factor out a new, small release_registered_pages() function, in
order to consolidate the logic for discerning between
TEE_SHM_USER_MAPPED and TEE_SHM_KERNEL_MAPPED pages. This also
absorbs the kfree() call that is also required there.

There is some helpful background in [2]: basically, this is a small
part of fixing a long-standing disconnect between pinning pages, and
file systems' use of those pages.

[1] Documentation/core-api/pin_user_pages.rst

[2] "Explicit pinning of user-space pages":
    https://lwn.net/Articles/807108/

Cc: Jens Wiklander <jens.wiklander@linaro.org>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: tee-dev@lists.linaro.org
Cc: linux-media@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Cc: linaro-mm-sig@lists.linaro.org
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

We would overrun the error_text array if we hit a TIMEOUT condition,
because we were using the error code "ETIMEDOUT" (which is 110) as an
array index.

We fix the problem by correcting the array index and by providing a
function to retrieve error messages rather than accessing the array
directly. The function includes a bounds check that prevents the array
from being overrun.

Link: https://lore.kernel.org/linux-arm-kernel/38d00022-730c-948a-917c-d86382df8cb9@canonical.com/
Link: https://lore.kernel.org/r/20200822205000.15841-1-mmayer@broadcom.com
Fixes: 2f330caf ("memory: brcmstb: Add driver for DPFE")
Reported-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Markus Mayer <mmayer@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Merge tag 'tee-memref-null-for-v5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee into arm/drivers

Handle NULL pointer indication from tee client

Adds support to indicate NULL pointers instead of a valid buffer when
querying the needed size of a buffer.

* tag 'tee-memref-null-for-v5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee:
  driver: tee: Handle NULL pointer indication from client

Link: https://lore.kernel.org/r/20200821102535.GA1872111@jadeSigned-off-by: Olof Johansson <olof@lixom.net>

Merge tag 'optee-i2c-for-v5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee into arm/drivers

Enable i2c device access from OP-TEE RPC

Extends the OP-TEE RPC protocol to enable I2C device access. This allows
a driver in secure world to access devices on a normal world I2C bus.

* tag 'optee-i2c-for-v5.10' of git://git.linaro.org/people/jens.wiklander/linux-tee:
  drivers: optee: allow op-tee to access devices on the i2c bus

Link: https://lore.kernel.org/r/20200821100532.GA1868056@jadeSigned-off-by: Olof Johansson <olof@lixom.net>

Some secure elements like NXP's SE050 sit on I2C buses. For OP-TEE to
control this type of cryptographic devices it needs coordinated access
to the bus, so collisions and RUNTIME_PM dont get in the way.

This trampoline driver allow OP-TEE to access them.
Signed-off-by: Jorge Ramirez-Ortiz <jorge@foundries.io>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

TEE Client introduce a new capability "TEE_GEN_CAP_MEMREF_NULL"
to handle the support of the shared memory buffer with a NULL pointer.

This capability depends on TEE Capabilities and driver support.
Driver and TEE exchange capabilities at driver initialization.
Signed-off-by: Michael Whitfield <michael.whitfield@nxp.com>
Signed-off-by: Cedric Neveux <cedric.neveux@nxp.com>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Tested-by: Joakim Bech <joakim.bech@linaro.org> (QEMU)
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

In brcmstb_dpfe_download_firmware(), memory is allocated to variable fw by
firmware_request_nowarn(), but never released. Fix up to release fw on
all return paths.

Cc: <stable@vger.kernel.org>
Fixes: 2f330caf ("memory: brcmstb: Add driver for DPFE")
Signed-off-by: Alex Dewar <alex.dewar90@gmail.com>
Acked-by: Markus Mayer <mmayer@broadcom.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20200820172118.781324-1-alex.dewar90@gmail.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

According to Tegra X1 (Tegra210) TRM, the APEW field is between
[23:16] so the shift bit for apew should be 16 accordingly.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Link: https://lore.kernel.org/r/20200818095121.13645-1-nicoleotsuka@gmail.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Most of the memory controller drivers do not depend on architecture
specific code so can be compile tested to increase build coverage.

When compile tested, do not enable them by default.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-7-krzk@kernel.org

Add separate entry for Broadcom STB DPFE driver, enabled by default on
ARCH_BRCMSTB.  This allows compile testing.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>

Use tabs instead of spaces for indentation.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-25-krzk@kernel.org

Use PTR_ERR_OR_ZERO to make the code a little bit simpler.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-17-krzk@kernel.org

The driver already uses 'if (!res)' pattern in the probe function so be
consistent.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724182328.3348-11-krzk@kernel.org

Since driver tracks reserved memory, freeing a non-reserved GPMC should
not be fatal and crash the system.  Printing a warning is friendlier.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724182328.3348-9-krzk@kernel.org

The GPMC_SET_ONE_CD_MAX macro uses return statement and variable 'cs'
coming from called scope.  This is not a good practice.  Also
checkpatch complained:

    WARNING: Macros with flow control statements should be avoided
    ERROR: Macros starting with if should be enclosed by a do - while
        loop to avoid possible if/else logic defects

Since GPMC_SET_ONE_CD_MAX macro just calls one function, it can be open
coded.  The difference with original code is that function will exit on
error not after every register set, but after a group of sets.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-22-krzk@kernel.org

The callers of gpmc_cs_set_timings() expect to receive -ERRNO on errors
and they pass further what they have received.

However gpmc_cs_set_timings() was returning -1 (equal to -EPERM) which
does not make sense in this context.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-21-krzk@kernel.org

The file-scope variables phys_base and mem_size are assigned in
gpmc_probe() but never read.

This fixes build error when compile testing on x86_64 architecture:

    drivers/memory/omap-gpmc.c:246:24: error: conflicting types for ‘phys_base’
     static resource_size_t phys_base, mem_size;
    In file included from arch/x86/include/asm/page.h:12:0,
                     from arch/x86/include/asm/thread_info.h:12,
                     from include/linux/thread_info.h:38,
                     from arch/x86/include/asm/preempt.h:7,
                     from include/linux/preempt.h:78,
                     from include/linux/spinlock.h:51,
                     from include/linux/irq.h:14,
                     from drivers/memory/omap-gpmc.c:12:
    arch/x86/include/asm/page_64.h:12:22: note: previous declaration of ‘phys_base’ was here
     extern unsigned long phys_base;
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-3-krzk@kernel.org

The driver does not use macros from asm/mach-types.h (neither MACH_TYPE
nor machine_is_xxx()).  Removal of this include allows compile testing
on non-ARM architectures which lack this header.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Link: https://lore.kernel.org/r/20200724074038.5597-2-krzk@kernel.org

Document scope of the mutex used by driver.
Signed-off-by: Krzysztof Kozlowski <krzk@kernel.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Link: https://lore.kernel.org/r/20200724180857.22119-1-krzk@kernel.org

The 'curr_rate' is protected by local 'dmc->lock' in various places, but
not in a function exynos5_dmc_get_status(). The lock protects frequency
(and voltage) change process and the corresponding value stored in
'curr_rate'. Add the locking mechanism to protect the 'curr_rate' reading
also in the exynos5_dmc_get_status().
Suggested-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Link: https://lore.kernel.org/r/20200811101727.3976-1-lukasz.luba@arm.comSigned-off-by: Krzysztof Kozlowski <krzk@kernel.org>

Pull io_uring fixes from Jens Axboe:
 "A few differerent things in here.

  Seems like syzbot got some more io_uring bits wired up, and we got a
  handful of reports and the associated fixes are in here.

  General fixes too, and a lot of them marked for stable.

  Lastly, a bit of fallout from the async buffered reads, where we now
  more easily trigger short reads. Some applications don't really like
  that, so the io_read() code now handles short reads internally, and
  got a cleanup along the way so that it's now easier to read (and
  documented). We're now passing tests that failed before"

* tag 'io_uring-5.9-2020-08-15' of git://git.kernel.dk/linux-block:
  io_uring: short circuit -EAGAIN for blocking read attempt
  io_uring: sanitize double poll handling
  io_uring: internally retry short reads
  io_uring: retain iov_iter state over io_read/io_write calls
  task_work: only grab task signal lock when needed
  io_uring: enable lookup of links holding inflight files
  io_uring: fail poll arm on queue proc failure
  io_uring: hold 'ctx' reference around task_work queue + execute
  fs: RWF_NOWAIT should imply IOCB_NOIO
  io_uring: defer file table grabbing request cleanup for locked requests
  io_uring: add missing REQ_F_COMP_LOCKED for nested requests
  io_uring: fix recursive completion locking on oveflow flush
  io_uring: use TWA_SIGNAL for task_work uncondtionally
  io_uring: account locked memory before potential error case
  io_uring: set ctx sq/cq entry count earlier
  io_uring: Fix NULL pointer dereference in loop_rw_iter()
  io_uring: add comments on how the async buffered read retry works
  io_uring: io_async_buf_func() need not test page bit