At present dm/device.h includes the linux-compatible features. This
requires including linux/compat.h which in turn includes a lot of headers.
One of these is malloc.h which we thus end up including in every file in
U-Boot. Apart from the inefficiency of this, it is problematic for sandbox
which needs to use the system malloc() in some files.
Move the compatibility features into a separate header file.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present devres.h is included in all files that include dm.h but few
make use of it. Also this pulls in linux/compat which adds several more
headers. Drop the automatic inclusion and require files to include devres
themselves. This provides a good indication of which files use devres.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Anatolij Gustschin <agust@denx.de>
The ext4 file system requires log2blksz to be set. So when setting the
block size on the block descriptor we should fill this field too.
This fixes a problem with EFI block devices providing ext4 partitions, cf.
https://lists.denx.de/pipermail/u-boot/2019-October/387702.html.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
We have a 'safe' version of this function but sometimes it is not needed.
Add a normal version too and update a few places that can use it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This reverts commit 0ebe112d09.
Most block devices have only one hwpart. Multiple hwparts only found used
by eMMC devices in u-boot. The mmc driver do blk_dselect_hwpart() at the
beginning of mmc_bread() which causes block cache being invalidated too
frequently and makes block cache useless.
So it's not a good idea to put blkcache_invalidate() in the common
functions. It should be called inside mmc_select_hwpart().
Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
Tested-by: Felix Brack <fb@ltec.ch>
Some storage devices have multiple hw partitions and both address from
zero, for example eMMC.
However currently block cache invalidation only applies to block
write/erase.
This can cause a problem that data of current hw partition is cached
before switching to another hw partition. And the following read
operation of the latter hw partition will get wrong data when reading
from the addresses that have been cached previously.
To solve this problem, invalidate block cache after a successful
select_hwpart operation.
Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
The function part_init() will only be built when we have both
CONFIG_PARTITIONS and CONFIG_HAVE_BLOCK_DEVICE set. Protect the call to
this function with both of these tests now.
Cc: Simon Glass <sjg@chromium.org>
Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: York Sun <york.sun@nxp.com>
Cc: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Cc: Mingkai Hu <mingkai.hu@nxp.com>
Cc: Stefan Roese <sr@denx.de>
Cc: Marek Behún <marek.behun@nic.cz>
Cc: Vanessa Maegima <vanessa.maegima@nxp.com>
Cc: Eugen Hristev <eugen.hristev@microchip.com>
Cc: Adam Ford <aford173@gmail.com>
Cc: Jagan Teki <jagan@amarulasolutions.com>
Cc: Tom Warren <twarren@nvidia.com>
Cc: Stephen Warren <swarren@nvidia.com>
Cc: Vitaly Andrianov <vitalya@ti.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This adds a new block interface type for VirtIO block devices.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
blk_next_free_devnum() can be helpful in some cases. Make it
a public API.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
part_init() is currently called in every DM BLK driver, either
in its bind() or probe() method. However we can use the BLK
uclass driver's post_probe() method to do it automatically.
Update all DM BLK drivers to adopt this change.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add a function to find the block device descriptor of the parent
device.
Signed-off-by: Tien Fong Chee <tien.fong.chee@intel.com>
[trini: Move function declaration to avoid warning]
Signed-off-by: Tom Rini <trini@konsulko.com>
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
This driver is no longer used on any supported platform in U-Boot and
there is no interest in maintaining it further from people that have
used it historically.
Cc: Simon Glass <sjg@chromium.org>
Cc: Michal Simek <michal.simek@xilinx.com>
c: Alexey Brodkin <alexey.brodkin@gmail.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Michal Simek <michal.simek@xilinx.com>
This patch provides
* a uclass for EFI drivers
* a EFI driver for block devices
For each EFI driver the uclass
* creates a handle
* adds the driver binding protocol
The uclass provides the bind, start, and stop entry points for the driver
binding protocol.
In bind() and stop() it checks if the controller implements the protocol
supported by the EFI driver. In the start() function it calls the bind()
function of the EFI driver. In the stop() function it destroys the child
controllers.
The EFI block driver binds to controllers implementing the block io
protocol.
When the bind function of the EFI block driver is called it creates a
new U-Boot block device. It installs child handles for all partitions and
installs the simple file protocol on these.
The read and write functions of the EFI block driver delegate calls to the
controller that it is bound to.
A usage example is as following:
U-Boot loads the iPXE snp.efi executable. iPXE connects an iSCSI drive and
exposes a handle with the block IO protocol. It calls ConnectController.
Now the EFI block driver installs the partitions with the simple file
protocol.
iPXE uses the simple file protocol to load Grub or the Linux Kernel.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
[agraf: add comment on calloc len]
Signed-off-by: Alexander Graf <agraf@suse.de>
This patch changes the calls to uclass_first/next_device() in blk_first/
next_device() to use uclass_find_first/next_device() instead. These functions
don't prepare the devices, which is correct in this case.
With this patch applied, the "usb storage" command now works again as
expected:
=> usb storage
Device 0: Vendor: SanDisk Rev: 1.00 Prod: Ultra
Type: Removable Hard Disk
Capacity: 58656.0 MB = 57.2 GB (120127488 x 512)
Without this patch, it used to generate this buggy output:
=> usb storage
Card did not respond to voltage select!
mmc_init: -95, time 26
No storage devices, perhaps not 'usb start'ed..?
Signed-off-by: Stefan Roese <sr@denx.de>
Suggested-by: Simon Glass <sjg@chromium.org>
Cc: Simon Glass <sjg@chromium.org>
Cc: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
All these places seem to inherit the codes from the MMC driver where
a FIXME was put in the comment. However the correct operation after
read should be cache invalidate, not flush.
The underlying drivers should be responsible for the cache operation.
Remove these codes completely.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: York Sun <york.sun@nxp.com>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: York Sun <york.sun@nxp.com>
There is an overflow problem when taking the size instead of the number
of blocks in blk_create_device(). This results in a wrong device size: the
device apparent size is its real size modulo 4GB.
Using the number of blocks instead of the device size fixes the problem and
is more coherent with the internals of the block layer.
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This converts the IDE driver to driver model so that block read and
write are fully functional.
Fixes: b7c6baef ("x86: Convert MMC to driver model")
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Add a function to find the name of an interface type (e.g. "sata", "scsi")
from the interface type enum.
This is useful for generic code (not specific to SATA or SCSI, for
example) that wants to display the type of interface it is dealing with.
Signed-off-by: Simon Glass <sjg@chromium.org>
This adds a new uclass id and block interface type for NVMe.
Signed-off-by: Zhikang Zhang <zhikang.zhang@nxp.com>
Signed-off-by: Wenbin Song <wenbin.song@nxp.com>
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Jon Nettleton <jon@solid-run.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Many devices support a child block device (e.g. MMC, USB). Add a
convenient way to get this device given the parent device.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
The intention with block devices is that the device number (devnum field
in its descriptor) matches the alias of its parent device. For example,
with:
aliases {
mmc0 = "/sdhci@700b0600";
mmc1 = "/sdhci@700b0400";
}
we expect that the block devices for mmc0 and mmc1 would have device
numbers of 0 and 1 respectively.
Unfortunately this does not currently always happen. If there is another
MMC device earlier in the driver model data structures its block device
will be created first. It will therefore get device number 0 and mmc0
will therefore miss out. In this case the MMC device will have sequence
number 0 but its block device will not.
To avoid this, allow a device to request a device number and bump any
existing device number that is using it. This all happens during the
binding phase so it is safe to change these numbers around. This allows
device numbers to match the aliases in all circumstances.
Add a test to verify the behaviour.
Signed-off-by: Simon Glass <sjg@chromium.org>
Sometimes it is useful to be able to find a block device without also
probing it. Add a function for this as well as the associated test.
Signed-off-by: Simon Glass <sjg@chromium.org>
This patch adds the flags parameter to device_remove() and changes all
calls to this function to provide the default value of DM_REMOVE_NORMAL
for "normal" device removal.
This is in preparation for the driver specific pre-OS (e.g. DMA
cancelling) remove support.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
All sata based drivers are bind and corresponding block
device is created. Based on this find_scsi_device() is able
to get back block device based on scsi_curr_dev pointer.
intr_scsi() is commented now but it can be replaced by calling
find_scsi_device() and scsi_scan().
scsi_dev_desc[] is commented out but common/scsi.c heavily depends on
it. That's why CONFIG_SYS_SCSI_MAX_DEVICE is hardcoded to 1 and symbol
is reassigned to a block description allocated by uclass.
There is only one block description by device now but it doesn't need to
be correct when more devices are present.
scsi_bind() ensures corresponding block device creation.
uclass post_probe (scsi_post_probe()) is doing low level init.
SCSI/SATA DM based drivers requires to have 64bit base address as
the first entry in platform data structure to setup mmio_base.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
All sata based drivers are bind and corresponding block
device is created. Based on this find_scsi_device() is able
to get back block device based on scsi_curr_dev pointer.
intr_scsi() is commented now but it can be replaced by calling
find_scsi_device() and scsi_scan().
scsi_dev_desc[] is commented out but common/scsi.c heavily depends on
it. That's why CONFIG_SYS_SCSI_MAX_DEVICE is hardcoded to 1 and symbol
is reassigned to a block description allocated by uclass.
There is only one block description by device now but it doesn't need to
be correct when more devices are present.
scsi_bind() ensures corresponding block device creation.
uclass post_probe (scsi_post_probe()) is doing low level init.
SCSI/SATA DM based drivers requires to have 64bit base address as
the first entry in platform data structure to setup mmio_base.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Series-changes: 2
- Use CONFIG_DM_SCSI instead of mix of DM_SCSI and DM_SATA
Ceva sata has never used sata commands that's why keep it in
SCSI part only.
- Separate scsi_scan() for DM_SCSI and do not change cmd/scsi.c
- Extend platdata
Series-changes: 3
- Fix scsi_scan return path
- Fix header location uclass-internal.h
- Add scsi_max_devs under !DM_SCSI
- Add new header device-internal because of device_probe()
- Redesign block device creation algorithm
- Use device_unbind in error path
- Create block device with id and lun numbers (lun was there in v2)
- Cleanup dev_num initialization in block device description
with fixing parameters in blk_create_devicef
- Create new Kconfig menu for SATA/SCSI drivers
- Extend description for DM_SCSI
- Fix Kconfig dependencies
- Fix kernel doc format in scsi_platdata
- Fix ahci_init_one - vendor variable
Series-changes: 4
- Fix Kconfig entry
- Remove SPL ifdef around SCSI uclass
- Clean ahci_print_info() ifdef logic
Current get_desc() implementation is not able to succesfully
finish and return pointer to block device descriptor.
Also function always return non zero value even device is found.
The patch fills block device descriptor and return 0 if device is found.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The block device uclass does not currently support selecting a particular
hardware partition but this is needed for MMC. Add it so that the blk API
can support MMC properly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a function that automatically builds the device name given the parent
and a supplied string. Most callers will want to do this, so putting this
functionality in one place makes more sense.
Signed-off-by: Simon Glass <sjg@chromium.org>
Allow a devnum parameter of -1 to indicate that the device number should be
alocated automatically. The next highest available device number for that
interface type is used.
Signed-off-by: Simon Glass <sjg@chromium.org>
Now that the MMC code accesses devices by number, we can implement this same
interface for driver model, allowing MMC to support using driver model for
block devices.
Add the required functions to the uclass.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a block device cache to speed up repeated reads of block devices by
various filesystems.
This small amount of cache can dramatically speed up filesystem
operations by skipping repeated reads of common areas of a block
device (typically directory structures).
This has shown to have some benefit on FAT filesystem operations of
loading a kernel and RAM disk, but more dramatic benefits on ext4
filesystems when the kernel and/or RAM disk are spread across
multiple extent header structures as described in commit fc0fc50.
The cache is implemented through a minimal list (block_cache) maintained
in most-recently-used order and count of the current number of entries
(cache_count). It uses a maximum block count setting to prevent copies
of large block reads and an upper bound on the number of cached areas.
The maximum number of entries in the cache defaults to 32 and the maximum
number of blocks per cache entry has a default of 2, which has shown to
produce the best results on testing of ext4 and FAT filesystems.
The 'blkcache' command (enabled through CONFIG_CMD_BLOCK_CACHE) allows
changing these values and can be used to tune for a particular filesystem
layout.
Signed-off-by: Eric Nelson <eric@nelint.com>
Add a uclass for block devices. These provide block-oriented data access,
supporting reading, writing and erasing of whole blocks.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
Simon Glass