Per datasheet: maximum block length is 2048 bytes,
data length field is in bits 0-23 of the Data Length Register.

Also for DMA mode we have to take into account rx/tx buffers' sizes.

In my tests this change doubles SD card I/O performance on big files.
Before the change Linux used default request size of 4 KB.
Signed-off-by: Sergei Antonov <saproj@gmail.com>
Link: https://lore.kernel.org/r/20230210143843.369943-1-saproj@gmail.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Convert platform_get_resource(), devm_ioremap_resource() to a single
call to devm_platform_get_and_ioremap_resource(), as this is exactly
what this function does.
Signed-off-by: Ye Xingchen <ye.xingchen@zte.com.cn>
Link: https://lore.kernel.org/r/202302101628321403257@zte.com.cnSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

This commit adds the missing MODULE_DEVICE_TABLE for acpi, or else
it won't be loaded automatically when compiled as a kernel module.
Reviewed-by: David Thompson <davthompson@nvidia.com>
Signed-off-by: Liming Sun <limings@nvidia.com>
Link: https://lore.kernel.org/r/f57ad0f8fdf663465bca74467c344dfa305a3199.1675305696.git.limings@nvidia.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

For Rockchip platform, DLL bypass bit and start bit need to be set if
DLL is not locked. And adjust pre-change delay to 0x3 for better signal
test result.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Link: https://lore.kernel.org/r/1675298118-64243-2-git-send-email-shawn.lin@rock-chips.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Support enabling / disabling the vqmmc power supply if it was provided
by the firmware.

Provide the .start_signal_voltage_switch callback to change the voltage
of the external vqmmc power supply.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Link: https://lore.kernel.org/r/20230131210229.68129-2-paul@crapouillou.netSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Merge the mmc fixes for v6.2-rc[n] into the next branch, to allow them to
get tested together with the new mmc changes that are targeted for v6.3.
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>

On JZ4760 and JZ4760B, SD cards fail to run if the maximum clock
rate is set to 50 MHz, even though the controller officially does
support it.

Until the actual bug is found and fixed, limit the maximum clock rate to
24 MHz.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20230131210229.68129-1-paul@crapouillou.netSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

If mmc_add_host() fails, it doesn't need to call mmc_remove_host(),
or it will cause null-ptr-deref, because of deleting a not added
device in mmc_remove_host().

To fix this, goto label 'fail_glue_init', if mmc_add_host() fails,
and change the label 'fail_add_host' to 'fail_gpiod_request'.

Fixes: 15a0580c ("mmc_spi host driver")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Cc:stable@vger.kernel.org
Link: https://lore.kernel.org/r/20230131013835.3564011-1-yangyingliang@huawei.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

If sdio_add_func() or sdio_init_func() fails, sdio_remove_func() can
not release the resources, because the sdio function is not presented
in these two cases, it won't call of_node_put() or put_device().

To fix these leaks, make sdio_func_present() only control whether
device_del() needs to be called or not, then always call of_node_put()
and put_device().

In error case in sdio_init_func(), the reference of 'card->dev' is
not get, to avoid redundant put in sdio_free_func_cis(), move the
get_device() to sdio_alloc_func() and put_device() to sdio_release_func(),
it can keep the get/put function be balanced.

Without this patch, while doing fault inject test, it can get the
following leak reports, after this fix, the leak is gone.

unreferenced object 0xffff888112514000 (size 2048):
  comm "kworker/3:2", pid 65, jiffies 4294741614 (age 124.774s)
  hex dump (first 32 bytes):
    00 e0 6f 12 81 88 ff ff 60 58 8d 06 81 88 ff ff  ..o.....`X......
    10 40 51 12 81 88 ff ff 10 40 51 12 81 88 ff ff  .@Q......@Q.....
  backtrace:
    [<000000009e5931da>] kmalloc_trace+0x21/0x110
    [<000000002f839ccb>] mmc_alloc_card+0x38/0xb0 [mmc_core]
    [<0000000004adcbf6>] mmc_sdio_init_card+0xde/0x170 [mmc_core]
    [<000000007538fea0>] mmc_attach_sdio+0xcb/0x1b0 [mmc_core]
    [<00000000d4fdeba7>] mmc_rescan+0x54a/0x640 [mmc_core]

unreferenced object 0xffff888112511000 (size 2048):
  comm "kworker/3:2", pid 65, jiffies 4294741623 (age 124.766s)
  hex dump (first 32 bytes):
    00 40 51 12 81 88 ff ff e0 58 8d 06 81 88 ff ff  .@Q......X......
    10 10 51 12 81 88 ff ff 10 10 51 12 81 88 ff ff  ..Q.......Q.....
  backtrace:
    [<000000009e5931da>] kmalloc_trace+0x21/0x110
    [<00000000fcbe706c>] sdio_alloc_func+0x35/0x100 [mmc_core]
    [<00000000c68f4b50>] mmc_attach_sdio.cold.18+0xb1/0x395 [mmc_core]
    [<00000000d4fdeba7>] mmc_rescan+0x54a/0x640 [mmc_core]

Fixes: 3d10a1ba ("sdio: fix reference counting in sdio_remove_func()")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20230130125808.3471254-1-yangyingliang@huawei.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

This makes the text read a little better.
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Acked-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20230129130059.1322858-1-j.neuschaefer@gmx.netSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Some SDIO WiFi modules stopped working after SDIO interrupt mode
was added if cap_sdio_irq isn't set in device tree. This patch was
confirmed to fix the issue.

Fixes: 066ecde6 ("mmc: meson-gx: add SDIO interrupt support")
Reported-by: Geraldo Nascimento <geraldogabriel@gmail.com>
Tested-by: Geraldo Nascimento <geraldogabriel@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/816cba9f-ff92-31a2-60f0-aca542d1d13e@gmail.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Use new function devm_mmc_alloc_host() to simplify the code.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/728f159b-885f-c78a-1a3d-f55c245250e1@gmail.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Add a device-managed version of mmc_alloc_host().

The argument order is reversed compared to mmc_alloc_host() because
device-managed functions typically have the device argument first.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/6d8f9fdc-7c9e-8e4f-e6ef-5470b971c74e@gmail.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

If we enable the memory stick block layer, use Kconfig to imply
the BFQ I/O scheduler.

As all memstick devices are single-queue, this is the scheduler that
users want so let's be helpful and make sure it gets
default-selected into a manual kernel configuration. It will still
need to be enabled at runtime (usually with udev scripts).

Cc: linux-block@vger.kernel.org
Cc: Paolo Valente <paolo.valente@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230131085220.1038241-1-linus.walleij@linaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

If we enable the MMC/SD block layer, use Kconfig to imply the BFQ
I/O scheduler.

As all MMC/SD devices are single-queue, this is the scheduler that
users want so let's be helpful and make sure it gets
default-selected into a manual kernel configuration. It will still
need to be enabled at runtime (usually with udev scripts).

Cc: linux-block@vger.kernel.org
Cc: Paolo Valente <paolo.valente@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230131084742.1038135-1-linus.walleij@linaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

SD interface logic has the register to switch UHS speed.
The default is up to SDR25 and to support SDR50 or faster,
add uniphier_sd_speed_switch() function to switch the speed mode.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Link: https://lore.kernel.org/r/20230125010201.28246-4-hayashi.kunihiko@socionext.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Transition of UHS mode needs to control the register in SD interface
logic. Add access to the register in the logic using the regmap from
"socionext,syscon-uhs-mode" property.

Define the start_signal_voltage_switch function only if UHS mode is
available.
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Link: https://lore.kernel.org/r/20230125010201.28246-3-hayashi.kunihiko@socionext.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

UHS mode control is included in the interface logic represented
by syscon. To refer this, add new property "socionext,syscon-uhs-mode".
Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230125010201.28246-2-hayashi.kunihiko@socionext.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

This commit fixes a race between completion of stop command and start of a
new command.
Previously the command ready interrupt was enabled before stop command
was written to the command register. This caused the command ready
interrupt to fire immediately since the CMDRDY flag is asserted constantly
while there is no command in progress.
Consequently the command state machine will immediately advance to the
next state when the tasklet function is executed again, no matter
actual completion state of the stop command.
Thus a new command can then be dispatched immediately, interrupting and
corrupting the stop command on the CMD line.
Fix that by dropping the command ready interrupt enable before calling
atmci_send_stop_cmd. atmci_send_stop_cmd does already enable the
command ready interrupt, no further writes to ATMCI_IER are necessary.
Signed-off-by: Tobias Schramm <t.schramm@manjaro.org>
Acked-by: Ludovic Desroches <ludovic.desroches@microchip.com>
Link: https://lore.kernel.org/r/20221230194315.809903-2-t.schramm@manjaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The IPQ5332 supports eMMC with an SDHCI controller. Add the appropriate
compatible to the documentation.
Signed-off-by: Kathiravan T <quic_kathirav@quicinc.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230125104520.89684-9-quic_kathirav@quicinc.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Document the compatible for SDHCI on IPQ9574.
Signed-off-by: devi priya <quic_devipriy@quicinc.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230124141541.8290-7-quic_devipriy@quicinc.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Convert the Amlogic SD / eMMC controller for S905/GXBB family SoCs
to dt-schema.

Take in account the used variant with amlogic,meson-gx-mmc.
Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20221117-b4-amlogic-bindings-convert-v3-6-e28dd31e3bed@linaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Node names should be generic and should not contain underscores.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Acked-by: Rob Herring <robh@kernel.org>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Link: https://lore.kernel.org/r/20230120085722.171965-2-krzysztof.kozlowski@linaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Cleanup by removing unneeded quotes from refs and redundant blank lines.
No functional impact except adjusting to preferred coding style.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> # for Renesas controllers
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Acked-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20230120085722.171965-1-krzysztof.kozlowski@linaro.orgSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Add a compatible for the pxav1 controller in the PXA168, along with
optional pinctrl properties to use for an errata workaround.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230116194401.20372-9-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The PXA168 errata recommends that the CMD signal should be detached from
the SD bus while performing the dummy CMD0 to restart the clock.
Implement this using pinctrl states.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-8-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The PXA168 has a documented silicon bug that causes SDIO card IRQs to be
missed. Implement the first half of the suggested workaround, which
involves resetting the data port logic and issuing a dummy CMD0 to
restart the clock.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-7-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Add ability to have an optional core clock just like the pxav3 driver.
The PXA168 needs this because its SDHC controllers have separate core
and io clocks that both need to be enabled. This also correctly matches
the documented devicetree bindings for this driver.
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-6-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The devicetree bindings for this driver specify that the two allowed
clock names are io and core. Change this driver to look for io, but
allow any name if it fails for backwards compatibility. Follow the same
pattern used in sdhci-pxav3, but add support for EPROBE_DEFER.

Get rid of an unnecessary pdev->dev while we're at it.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-5-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The PXA168 has a documented silicon bug that results in a data abort
exception when accessing the SDHCI_HOST_VERSION register on SDH2 and
SDH4 through a 16-bit read. Implement the workaround described in the
errata, which performs a 32-bit read from a lower address instead. This
is safe to use on all four SDH peripherals.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-4-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Enable CONFIG_MMC_SDHCI_IO_ACCESSORS for the pxav2 driver. The read_w
callback is needed for a silicon bug workaround in the PXA168.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-3-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Add a new compatible string for the version 1 controller used in the
PXA168, along with necessary quirks. Use a separate ops struct in
preparation for a silicon bug workaround only necessary on V1.
Signed-off-by: Doug Brown <doug@schmorgal.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230116194401.20372-2-doug@schmorgal.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

pwrseq_sd8787 is forced to be built as a module if its dependencies are.

That's unnecessary, it's perfectly fine for it to be built-in even
though the wireless drivers that need it are modules.

Relax the depends definition in Kconfig accordingly.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Cc: Matt Ranostay <matt@ranostay.consulting>
Cc: Lubomir Rintel <lkundrak@v3.sk>
Cc: Claudiu Beznea <claudiu.beznea@microchip.com>
Link: https://lore.kernel.org/r/8bb3d7c3a36985e030ba40e853c57578de8fb303.1673866725.git.lukas@wunner.deSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Currently make dtbs_check shows lots of errors because imx*.dtsi does
not use single compatibles but combinations of them.

Add fallbacks for imx6sll/ull which are useful for U-Boot.

This will significantly reduce noise on make dtbs_check.
Signed-off-by: Andreas Kemnade <andreas@kemnade.info>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230115195212.2530087-1-andreas@kemnade.infoSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

The SD Card/MMC interface on the Renesas RZ/V2M SoC uses 4 clocks, like
on RZ/G2L SoCs.

The current description specifies 3 clocks, like on R-Car Gen3, which
will lead to "make dtbs_check" failures:

    arch/arm64/boot/dts/renesas/r9a09g011-v2mevk2.dtb: mmc@85000000: clocks: [[3, 1, 18], [3, 1, 20], [3, 1, 19], [3, 1, 17]] is too long
	    From schema: Documentation/devicetree/bindings/mmc/renesas,sdhi.yaml

Fixes: c8dbdf2a1044951b ("dt-bindings: mmc: renesas,sdhi: Document RZ/V2M support")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Fabrizio Castro <fabrizio.castro.jz@renesas.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/054c9f4dbb0bb1525f780d1e85c724436465c20c.1673270716.git.geert+renesas@glider.beSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

This commit enables SDHCI Host V4 support on Bluefield-3 SoC to be
consistent with the default setting in firmware(UEFI).
Reviewed-by: David Woods <davwoods@nvidia.com>
Signed-off-by: Liming Sun <limings@nvidia.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/990885f566c32ac8e6888ad6b434fb70d1a5d7af.1673460632.git.limings@nvidia.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

Now that it is no longer used, remove SDHCI_QUIRK_MISSING_CAPS.

Note, from now on, __sdhci_read_caps() should be used to provide missing
capability flags.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20230113110011.129835-7-adrian.hunter@intel.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

SDHCI_QUIRK_MISSING_CAPS is not needed because __sdhci_read_caps() can be
called instead.

In preparation to get rid of SDHCI_QUIRK_MISSING_CAPS, replace
SDHCI_QUIRK_MISSING_CAPS with __sdhci_read_caps().

__sdhci_read_caps() is also called from sdhci_setup_host() via
sdhci_read_caps(), however only the first call to __sdhci_read_caps() does
anything because after that host->read_caps has been set to true.

Note, __sdhci_read_caps() does more than just set host->caps, such as do a
reset, so calling __sdhci_read_caps() earlier could have unforeseen
side-effects. However the code flow has been reviewed with that in mind.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Acked-by: Scott Branden <scott.branden@broadcom.com>
Link: https://lore.kernel.org/r/20230113110011.129835-6-adrian.hunter@intel.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

SDHCI_QUIRK_MISSING_CAPS is not needed because sdhci_read_caps() can be
called instead.

In preparation to get rid of SDHCI_QUIRK_MISSING_CAPS, replace
SDHCI_QUIRK_MISSING_CAPS with sdhci_read_caps().

__sdhci_read_caps() is also called from sdhci_setup_host() via
sdhci_read_caps(), however only the first call to __sdhci_read_caps() does
anything because after that host->read_caps has been set to true.

Note, __sdhci_read_caps() does more than just set host->caps, such as do a
reset, so calling __sdhci_read_caps() earlier could have unforeseen
side-effects. However the code flow has been reviewed with that in mind.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Jisheng Zhang <jszhang@kernel.org>
Link: https://lore.kernel.org/r/20230113110011.129835-5-adrian.hunter@intel.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>

SDHCI_QUIRK_MISSING_CAPS is not needed because sdhci_read_caps() can be
called instead.

In preparation to get rid of SDHCI_QUIRK_MISSING_CAPS, replace
SDHCI_QUIRK_MISSING_CAPS with sdhci_read_caps().

__sdhci_read_caps() is also called from sdhci_setup_host() via
sdhci_read_caps(), however only the first call to __sdhci_read_caps() does
anything because after that host->read_caps has been set to true.

Note, __sdhci_read_caps() does more than just set host->caps, such as do a
reset, so calling __sdhci_read_caps() earlier could have unforeseen
side-effects. However the code flow has been reviewed with that in mind.
Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Acked-by: Chunyan Zhang <zhang.lyra@gmail.com>
Link: https://lore.kernel.org/r/20230113110011.129835-4-adrian.hunter@intel.comSigned-off-by: Ulf Hansson <ulf.hansson@linaro.org>