- 02 May, 2016 40 commits
-
-
Wolfram Sang authored
BIT() makes it easier to match the bits to the datasheet. This is especially important here, since some variants have different names in their datasheets (like with Renesas R-Car). Reviewed-by:
Simon Horman <horms+renesas@verge.net.au> Signed-off-by:
Wolfram Sang <wsa+renesas@sang-engineering.com> Signed-off-by:
Ulf Hansson <ulf.hansson@linaro.org>
-
Wolfram Sang authored
Looking at the backlogs, I am not the only one who missed that the above functions do not read u32 from one register, but create a virtual u32 from reading to adjacent u16 registers (which depending on 'bus_shift' can be up to 8 byte apart). Because this driver supports old hardware for which we don't have documentation, I first wrongly assumed there was a variant which had a few u32 registers. Let's give the functions more descriptive names to make it more obvious what is happening. Reviewed-by:
-
Tony Lindgren authored
The legacy user space for n900 relies on the MMC slot names. Let's check if those are passed in pdata and use them. As this makes the DT booting compatible with legacy booting, we should be able to start dropping omap3 legacy booting support in v4.8. Cc: Ulf Hansson <ulf.hansson@linaro.org> Cc: Kishon Vijay Abraham I <kishon@ti.com> Cc: linux-mmc@vger.kernel.org Signed-off-by:
Tony Lindgren <tony@atomide.com> Tested-by:
Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com> Signed-off-by:
-
Wolfram Sang authored
There is no support for this platform in the kernel anymore. Reported-by:
Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by:
-
Wolfram Sang authored
There is no support for this platform in the kernel anymore. Make the Kconfig text more generic, so it won't get stale anymore. Reported-by:
-
Wolfram Sang authored
A last minute fix applied by Ulf made room for some simplification. Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Wolfram Sang authored
There is no reason to have a public and private header file. Merge them into a private one, so looking up symbols is less confusing. Signed-off-by:
-
Wolfram Sang authored
Having just one irq handler again, let's include the 'card_status' function in the main handler which is way more readable. Drop a useless debug output while here. It should be a dev_dbg in case we ever need it again. Signed-off-by:
-
Wolfram Sang authored
We removed installation of separate handlers previously, so we can also remove the separate handlers. Signed-off-by:
-
Wolfram Sang authored
There is no user left in the kernel, so this code can be removed. (Legacy, non-DT sh_mobile boards have been removed a while ago.) The diff looks more complicated than it is: The if-block for multiplexed isr is now the main code path, the rest is removed. Signed-off-by:
-
Wolfram Sang authored
We won't access an index based array to get our DT config, but create separate structs instead. So, remove the array which only wastes memory. Signed-off-by:
-
Ulf Hansson authored
Instead of using an mmc specific implementation to deal with indexes through a BITMAP, let's convert to use the IDA library. Signed-off-by: -
Ulf Hansson authored
If the allocation of a new partition fails, let's make sure to also release the previously picked device index. Signed-off-by: -
Ulf Hansson authored
As IDA is more lightweight than IDR, let's convert to use that instead. Signed-off-by: -
Masahiro Yamada authored
defined(CONFIG_LEDS_CLASS) || defined(CONFIG_LEDS_CLASS_MODULE) is equivalent to: IS_ENABLED(CONFIG_LEDS_CLASS) Signed-off-by:
Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by:
Adrian Hunter <adrian.hunter@intel.com> Signed-off-by:
-
Masahiro Yamada authored
defined(CONFIG_LEDS_CLASS) || (defined(CONFIG_LEDS_CLASS_MODULE) && \ defined(CONFIG_MMC_SDHCI_MODULE)) is equivalent to: defined(CONFIG_LEDS_CLASS) || (defined(CONFIG_LEDS_CLASS_MODULE) && \ defined(MODULE)) and it can also be written shortly as: IS_REACHABLE(CONFIG_LEDS_CLASS) Signed-off-by: -
Adrian Hunter authored
SDHCI_SDR104_NEEDS_TUNING was originally named SDHCI_HS200_NEEDS_TUNING and was added in commit 069c9f14 ("mmc: host: Adds support for eMMC 4.5 HS200 mode"). That commit conflated SDHCI_SDR50_NEEDS_TUNING and SDHCI_HS200_NEEDS_TUNING due to what appears to be misplaced parentheses. Commit 156e14b1 ("mmc: sdhci: fix caps2 for HS200") made HS200 configuration equivalent to SDR104 configuration, renaming SDHCI_HS200_NEEDS_TUNING to SDHCI_SDR104_NEEDS_TUNING despite tuning for HS200 now being non-optional. The mix-up with SDHCI_SDR50_NEEDS_TUNING remained and became more obvious after commit 4b6f37d3 ("mmc: sdhci: clean up sdhci_execute_tuning() decision") where the author noted the patch was "reflecting what the original code was doing, it shows that it may not be what the author actually intended." The way the code is currently written, SDHCI_SDR104_NEEDS_TUNING causes tuning to be done always for SDR50 mode if SDR104 mode is also supported by the host controller. That makes no sense because we already have capabilities bit SDHCI_USE_SDR50_TUNING and corresponding flag SDHCI_SDR50_NEEDS_TUNING for that purpose. Given the dubious origins of SDHCI_SDR104_NEEDS_TUNING, it seems reasonable to remove it. The benefit being SDR50 mode will now not un-nessessarily do tuning. Signed-off-by:
-
Masahiro Yamada authored
Swap the call order of sdhci_alloc_host() and platform_get_irq(). It makes sdhci_alloc_host() the last function that can fail in the sdhci_pltfm_init(). So, we can drop the sdhci_free_host() call from the failure path. Signed-off-by:
-
Masahiro Yamada authored
Call devm_ioremap_resource() right after platform_get_resource(). This saves the error check of platform_get_resource() because devm_ioremap_resource() checks if the given resource is NULL. Signed-off-by:
-
Masahiro Yamada authored
The chain of devm_request_mem_region() and devm_ioremap() can be replaced with devm_ioremap_resource(). Also, we can drop the error messages because devm_ioremap_resource() displays similar messages on error. Signed-off-by:
-
Masahiro Yamada authored
Use the managed variant of ioremap(). Signed-off-by:
-
Masahiro Yamada authored
Use the managed variant of request_mem_region(). Signed-off-by:
-
Masahiro Yamada authored
The function platform_get_irq() can fail; it returns a negative error code on failure. A negative IRQ number will make sdhci_add_host() fail to request IRQ anyway, but it makes sense to let it fail earlier here. Signed-off-by:
-
Masahiro Yamada authored
The requirement resource_size >= 0x100 may not necessarily be reasonable; for example, sdhci-dove appears to sidestep some registers in sdhci_dove_readw(). Moreover, current code displays an error message for too small resource size, but still moves forward. Every DT should be responsible for describing its properties correctly, so lets's remove this error message from the common framework. Signed-off-by:
-
Masahiro Yamada authored
This if-block is going to call mmc_card_set_blockaddr(), so mmc_card_blockaddr() right before it is redundant. I am fixing the block comment style while I am here. Signed-off-by:
-
Srinivas Kandagatla authored
simple-pwrseq and emmc-pwrseq drivers rely on platform_device structure from of_find_device_by_node(), this works mostly. But, as there is no driver associated with this devices, cases like default/init pinctrl setup would never be performed by pwrseq. This becomes problem when the gpios used in pwrseq require pinctrl setup. Currently most of the common pinctrl setup is done in drivers/base/pinctrl.c by pinctrl_bind_pins(). There are two ways to solve this issue on either convert pwrseq drivers to a proper platform drivers or copy the exact code from pcintrl_bind_pins(). I prefer converting pwrseq to proper drivers so that other cases like setting up clks/parents from dt would also be possible. Signed-off-by:
Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Signed-off-by:
-
Srinivas Kandagatla authored
This patch adds to_pwrseq_emmc() macro to make the code more readable. Signed-off-by:
-
Srinivas Kandagatla authored
This patch adds to_pwrseq_simple() macro to make the code more readable. Signed-off-by:
-
Adrian Hunter authored
ifdef's make the code more complicated and harder to read. Move all the LED code together to reduce the ifdef's to one place. Signed-off-by:
-
Adrian Hunter authored
Some error paths in sdhci_add_host() simply returned without cleaning up. Also the return value from mmc_add_host() was not being checked. Signed-off-by:
-
Adrian Hunter authored
The logic '!mmc.f_max || (mmc.f_max && mmc.f_max > max_clk)' is equivalent to '!mmc.f_max || mmc.f_max > max_clk'. Reported-by:
David Binderman <dcb314@hotmail.com> Signed-off-by:
-
Adrian Hunter authored
Set MMC_CAP_AGGRESSIVE_PM for Broxton host controllers. Signed-off-by:
-
Adrian Hunter authored
Set MMC_CAP_AGGRESSIVE_PM for Broxton host controllers. Signed-off-by:
-
Ludovic Desroches authored
SDHCI_QUIRK2_NEED_DELAY_AFTER_INT_CLK_RST quirk is not used anymore so remove it. Signed-off-by:
Ludovic Desroches <ludovic.desroches@atmel.com> Acked-by:
-
Ludovic Desroches authored
Disabling the internal clock while configuring the SD card clock can lead to internal clock stabilization issue and/or unexpected switch to the base clock when using presets. A quirk SDHCI_QUIRK2_NEED_DELAY_AFTER_INT_CLK_RST was introduced to fix these bugs. The cause was assumed to be a too long internal re-synchronisation but it seems in some cases the delay (even if longer) doesn't fix this bug. The safest workaround is to not disable/enable the internal clock during the SD card clock configuration. Signed-off-by:
-
Ludovic Desroches authored
In order to remove the SDHCI_QUIRK2_NEED_DELAY_AFTER_INT_CLK_RST and to reduce code duplication, put the code relative to the SD clock configuration in a function which can be used by hosts for the implementation of the ->set_clock() callback. Signed-off-by:
-
Ulf Hansson authored
There are no need to have two versions of sdhci_runtime_pm_bus_off|on(), depending on whether CONFIG_PM is set or unset. Thus it's easy to move the implementation of these functions a bit earlier to avoid the unnecessary pre-definition of them, so let's do that. Signed-off-by:
-
Masahiro Yamada authored
A platform_driver does not need to set an owner, it will be populated by the driver core. Signed-off-by:
-
Ulf Hansson authored
Commit 9250aea7 ("mmc: core: Enable runtime PM management of host devices"), made some calls to the runtime PM API from the driver redundant. Especially those which deals with runtime PM reference counting, so let's remove them. Moreover as SDHCI have its own wrapper functions for runtime PM these becomes superfluous, so let's remove them as well. Signed-off-by:
-
Gwendal Grignou authored
The eMMC HW reset may be implemented either via the host ops ->hw_reset() callback or through DT and the eMMC pwrseq. Additionally some eMMC cards don't support HW reset. To allow a reset to be done for the different combinations of mmc hosts and eMMC/MMC cards, let's implement a fallback via trying a regular power cycle. This improves the mmc block layer retry mechanism of failing I/O requests. Signed-off-by:
Gwendal Grignou <gwendal@chromium.org> [Ulf: Rewrote changelog] Signed-off-by:
-
