fsg_common_init is a lengthy function. Factor a portion of it out.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

fsg_common_init is a lengthy function. Factor a portion of it out.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

When configfs is in place, the luns will not be represented in sysfs,
so there will be no struct device associated with a lun.
In order to maintain compatibility and allow configfs adoption
sysfs is made optional in this patch.

As a consequence some debug macros need to be adjusted. Two new
fields are added to struct fsg_lun: name and name_pfx.
The "name" is for storing a string which is presented to the user
instead of the dev_name. The "name_pfx", if non-NULL, is prepended
to the "name" at printing time.

The name_pfx is for a future lun.0, which will be a default group in
mass_storage.<name>. By design at USB function configfs group's creation
time its name is not known (but instead set a bit later in
drivers/usb/gadget/configfs.c:function_make) and it is this name that
serves the purpose of the said name prefix. So instead of copying
a yet-unknown string a pointer to it is stored in struct fsg_lun.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

When configfs is in place, gadgets will have to be able to free
fsg buffers. Add a helper function.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Prepare for handling with configfs.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This is needed to prepare for configfs integration.

So far the luns have been allocated during gadget's initialization, based
on the nluns module parameter's value; the exact number is known when the
gadget is initialized and that number of luns is allocated in one go; they
all will be used.

When configfs is in place, the luns will be created one-by-one by the user.
Once the user is satisfied with the number of luns, they activate the
gadget. The number of luns must be <= FSG_MAX_LUN (currently 8), but other
than that it is not known up front and the user need not use contiguous
numbering (apart from the default lun #0). On the other hand, the function
code uses lun numbers to identify them and the number needs to be used
as an index into an array.

Given the above, an array needs to be allocated, but it might happen that
7 out of its 8 elements will not be used. On my machine
sizeof(struct fsg_lun) == 462, so > 3k of memory is allocated but not used
in the worst case.

By adding another level of indirection (allocating an array of pointers
to struct fsg_lun and then allocating individual luns instead of an array
of struct fsg_luns) at most 7 pointers are wasted, which is much less.

This patch also changes some for/while loops to cope with the fact
that in the luns array some entries are potentially empty.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

In order to prepare for the new function interface the f_mass_storage.c
needs to be compiled as a module, and so a header file will be required.

This patch factors out some code to a new f_mass_storage.h.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Converting to configfs requires making the f_mass_storage.c a module.

But first we need to get rid of "#include "storage_common.c".

This patch makes storage_common.c a separately compiled file, which is
built as a utility module named u_ms.ko. After all mass storage users are
converted to the new function interface this module can be eliminated
by merging it with the mass storage function's module.

USB descriptors are exported so that they can be accessed from
f_mass_storage.

FSG_VENDOR_ID and FSG_PRODUCT_ID are moved to their only user.

Handling of CONFIG_USB_GADGET_DEBUG_FILES is moved to f_mass_storage.c.
The fsg_num_buffers static is moved to FSG_MODULE_PARAMETER users, so
instead of using a global variable the f_mass_storage introduces
fsg_num_buffers member in fsg_common (and fsg_config).

fsg_strings and fsg_stringtab are moved to f_mass_storage.c.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Gadgets endpoint driver data is a criteria to judge that
whether the endpoints are in use or not. When gadget gets
assigned an endpoint from endpoint list, they check its
driver data if the driver data is NULL.

If the driver data is not NULL then they regard it as in use.
Therefore all of gadgets should reset their endpoints driver
data to NULL as they are disabled. Otherwise it causes a leak
of endpoint resource.
Signed-off-by: NPeter Oh <poh@broadcom.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Convert all USB gadget sysfs attributes to use the _RO or _RW variants,
to make them easier to audit and ensure that the permissions are
correct.

Note, two are left using the DEVICE_ATTR() macro, as there is no
DEVICE_ATTR_WO() in Linus's tree, that will happen after 3.12-rc1 is
out, a follow-on patch will be sent then.
Reviewed-by: NFelipe Balbi <balbi@ti.com>
Acked-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

--
 drivers/usb/gadget/composite.c      |    8 +++-----
 drivers/usb/gadget/dummy_hcd.c      |    8 ++++----
 drivers/usb/gadget/f_mass_storage.c |   14 ++++++--------
 drivers/usb/gadget/net2272.c        |    4 ++--
 drivers/usb/gadget/net2280.c        |   18 +++++++++---------
 drivers/usb/gadget/storage_common.c |   25 ++++++++++++-------------
 drivers/usb/gadget/udc-core.c       |   14 +++++++-------
 7 files changed, 43 insertions(+), 48 deletions(-)

The local variables such as 'filename', 'vendor_name', and
'product_name' are pointers; thus, use NULL instead of 0 to fix
the following sparse warnings

drivers/usb/gadget/f_mass_storage.c:3046:27: warning: Using plain integer as NULL pointer
drivers/usb/gadget/f_mass_storage.c:3050:28: warning: Using plain integer as NULL pointer
drivers/usb/gadget/f_mass_storage.c:3051:29: warning: Using plain integer as NULL pointer
Signed-off-by: NJingoo Han <jg1.han@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

If cfg->product name is not set, a default name is chosen depending
on the common->luns->cdrom flag. If the flag is set the name should
be "File-CD Gadget", and if the flag is not set the name should be
"File-Stor Gadget".
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Without this memory barrier, the file-storage thread may fail to
escape from the following while loop, because it may observe new
common->thread_wakeup_needed and old bh->state which are updated by
the callback functions.

	/* Wait for the CBW to arrive */
	while (bh->state != BUF_STATE_FULL) {
		rc = sleep_thread(common);
		if (rc)
			return rc;
	}

Cc: stable@vger.kernel.org
Signed-off-by: NUCHINO Satoshi <satoshi.uchino@toshiba.co.jp>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

| In file included from drivers/usb/gadget/acm_ms.c:43:
| f_mass_storage.c:2199:18: warning: comparison of unsigned expression >= 0 is always true  [-Wtautological-compare]
|         if (common->lun >= 0 && common->lun < common->nluns)
|             ~~~~~~~~~~~ ^  ~

common->lun is defined as "unsigned int" so its value is always >= 0.
It is assigned via cbw->Lun which is defined as u8 so it is also not
abused as -1.

[ mina86@mina86.com : make lun unsigned int and use %u in DBG() macro for it ]
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

pre_eject and post_eject are not used by anyone. Removing them.
Signed-off-by: NAndrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

storage_common.c has been used by both file_storage.c and f_mass_storage.c
which had some different requirements in a few places.  To accomodate for
that, storage_common.c provided configuratian macros which were to be
defined (or not) prior to the file #inclusion.  Because now
file_storage.c is no longer with us, we can remove support for those
macros and thus simplify the code slightly.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

HS and SS descriptors are staticaly created. They are updated during the
bind process with the endpoint address, string id or interface numbers.

After that, the descriptor chain is linked to struct usb_function which
is used by composite in order to serve the GET_DESCRIPTOR requests,
number of available configs and so on.

There is no need to assign the HS descriptor only if the UDC supports
HS speed because composite won't report those to the host if HS support
has not been reached. The same reasoning is valid for SS.

This patch makes sure each function updates HS/SS descriptors
unconditionally and uses the newly introduced helper function to create a
copy the descriptors for the speed which is supported by the UDC.

While at that, also rename f->descriptors to f->fs_descriptors in order
to make it more explicit what that means.

Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The bcdDevice field is defined as
|Device release number in binary-coded decimal
in the USB 2.0 specification. We use this field to distinguish the UDCs
from each other. In theory this could be used on the host side to apply
certain quirks if the "special" UDC in combination with this gadget is
used. This hasn't been done as far as I am aware. In practice it would
be better to fix the UDC driver before shipping since a later release
might not need this quirk anymore.

There are some driver in tree (on the host side) which use the bcdDevice
field to figure out special workarounds for a given firmware revision.
This seems to make sense. Therefore this patch converts all gadgets
(except a few) to use the kernel version instead a random 2 or 3 plus
the UDC number. The few that don't report kernel's version are:
- webcam
  This one reports always a version 0x10 so allow it to do so in future.
- nokia
  This one reports always 0x211. The comment says that this gadget works
  only if the UDC supports altsettings so I added a check for this.
- serial
  This one reports 0x2400 + UDC number. Since the gadget version is 2.4
  this could make sense. Therefore bcdDevice is 0x2400 here.

I also remove various gadget_is_<name> macros which are unused. The
remaining few macros should be moved to feature / bug bitfield.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Some gadgets provide custom entry here. Some may override it with an
etntry that is also created by composite if there was no value sumbitted
at all.
This patch removes all "custom manufacturer" strings which are the same
as these which are created by composite. Then it moves the creation of
the default manufacturer string to usb_composite_overwrite_options() in
case no command line argument has been used and the entry is still an
empty string.
By doing this we get rid of the global variable "composite_manufacturer"
in composite.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The “file” sysfs entry for LUNs was writable even for non-removable
LUNs and the fsg_store_file() function did not check whether LUN is
removable or not.  This made it possible to change or even close
LUN's backing file.

The same is true for “ro” sysfs entry and LUNs simulating CD-ROM.
For those LUNs, the file should not be writable.

This commit introduces two new device_attribute structures for those
two special cases so that the file/ro sysfs entries are made
non-writable when not desired.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This commit adds Documentation/usb/mass-storage.txt file.  It contains
description of how to use the mass storage gadget from user space.  It
elaborates on madule parameters and sysfs interface more then it was
written in the comments in the source code.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This commit removes thread_name and lun_name_format fields from the
fsg_config structure.  Those fields are not used by any in-tree code
and their usefulness is rather theoretical.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This commit changes the default value of the removable module parameter
from “y” to “n”.  This comes with line with file_storag's default and
seems to be a better default.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

There are no in-tree fsg_add() users and it has been deprecated
since 2.6.35 [1dc90985: fsg_add() renamed to fsg_bind_config()] so
out-of-tree users had more then enough time to convert.  Removing.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch (as1539) fixes a minor bug in the mass-storage gadget
drivers.  When an unknown command is received, the error code sent
back is "Invalid Field in CDB" rather than "Invalid Command".  This is
because the bitmask of CDB bytes allowed to be nonzero is incorrect.

When handling an unknown command, we don't care which command bytes
are nonzero.  All the bits in the mask should be set, not just eight
of them.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
CC: <Michal Nazarewicz <mina86@mina86.com>
CC: <stable@vger.kernel.org>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This replaces the remaining defines which are available in "public"
include/ directory and are re-defined by the storage gadget.
This is patch is basicaly search & replace followed by the removal of
the defines.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

This moves the BOT data structures for CBW and CSW from drivers internal
header file to global include able file in include/.
The storage gadget is using the same name for CSW but a different for
CBW so I fix it up properly. The same goes for the ub driver and keucr
driver in staging.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Fix the following build warnings:

 CC [M]  drivers/usb/gadget/acm_ms.o
drivers/usb/gadget/acm_ms.c: In function ‘__check_ro’:
drivers/usb/gadget/acm_ms.c:119: warning: return from incompatible pointer type
drivers/usb/gadget/acm_ms.c: In function ‘__check_removable’:
drivers/usb/gadget/acm_ms.c:119: warning: return from incompatible pointer type
drivers/usb/gadget/acm_ms.c: In function ‘__check_cdrom’:
drivers/usb/gadget/acm_ms.c:119: warning: return from incompatible pointer type
drivers/usb/gadget/acm_ms.c: In function ‘__check_nofua’:
drivers/usb/gadget/acm_ms.c:119: warning: return from incompatible pointer type
drivers/usb/gadget/acm_ms.c: In function ‘__check_stall’:
drivers/usb/gadget/acm_ms.c:119: warning: return from incompatible pointer type
  CC [M]  drivers/usb/gadget/mass_storage.o
drivers/usb/gadget/mass_storage.c: In function ‘__check_ro’:
drivers/usb/gadget/mass_storage.c:94: warning: return from incompatible pointer type
drivers/usb/gadget/mass_storage.c: In function ‘__check_removable’:
drivers/usb/gadget/mass_storage.c:94: warning: return from incompatible pointer type
drivers/usb/gadget/mass_storage.c: In function ‘__check_cdrom’:
drivers/usb/gadget/mass_storage.c:94: warning: return from incompatible pointer type
drivers/usb/gadget/mass_storage.c: In function ‘__check_nofua’:
drivers/usb/gadget/mass_storage.c:94: warning: return from incompatible pointer type
drivers/usb/gadget/mass_storage.c: In function ‘__check_stall’:
drivers/usb/gadget/mass_storage.c:94: warning: return from incompatible pointer type

Declare the fsg_module_parameters fields as "bool" so that they can match the types
passed in FSG_MODULE_PARAM_ARRAY macro.

Since commit 493c90ef (module_param: check that bool parameters really are bool.),
moduleparam.h was changed in a way that the "bool" parameter type now really
requires "bool" type and no longer allows "unsigned int".
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Remove one define of FSG_NO_INTR_EP and we still have that we can use.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The m.nazarewicz@samsung.com email address is no longer valid,
so this commit replaces it with mina86@mina86.com which is
employer-agnostic and thus should be valid for foreseeable
feature.
Signed-off-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

Release superspeed mass storage descriptors memory
when the function is unbind.
Signed-off-by: NYu Xu <yuxu@marvell.com>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The advantage of kcalloc is, that will prevent integer overflows which could
result from the multiplication of number of elements and size and it is also
a bit nicer to read.

The semantic patch that makes this change is available
in https://lkml.org/lkml/2011/11/25/107Signed-off-by: NThomas Meyer <thomas@m3y3r.de>
Acked-by: NMichal Nazarewicz <[4]mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

With Peiyu's patch "gadget: mass_storage: adapt logic block size to bound block
devices" (http://www.spinics.net/lists/linux-usb/msg50791.html), now mass storage
can adjust logic block size dynamically based on real devices.
Then there is one issue caused by it, if two luns have different logic block size,
mass storage can't work.
Let's check the current software flow:
1. get_next_command(): call received_cbw();
2. received_cbw(): update common->lun = cbw->Lun, but common->curlen is not updated;
3. do_scsi_command(): in READ_X and WRITE_X commands, common->data_size_from_cmnd is
updated by common->curlun->blkbits;
4. check_command(): update common->curlun according to common->lun
As you can see, the step 3 uses wrong common->curlun, then wrong common->curlun->blkbits.
If the two luns have same blkbits, there isn't issue. Otherwise, both will fail.
This patch moves the common->curlun update to step 1, then make sure step 3 gets right
blkbits and data_size_from_cmnd.

Cc: Peiyu Li <peiyu.li@csr.com>
Signed-off-by: NYuPing Luo <yuping.luo@csr.com>
Signed-off-by: NBarry Song <Baohua.Song@csr.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The USB-IF CV compliance tester is getting stricter, and it would
be valid for it to fail a mass-storage device that accepts an
invalid USB_BULK_RESET_REQUEST request. Although it doesn't do
that yet, let's be proactive and fix that now.

Suggested by Alan Stern.
Signed-off-by: NPaul Zimmerman <paulz@synopsys.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

The latest USB-IF CV tester checks for a valid length for this
request.
Signed-off-by: NPaul Zimmerman <paulz@synopsys.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@suse.de>

this patch adds superspeed descriptors for the
storage gadgets.
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

The Synopsys USB device controller requires all OUT transfer request
lengths to be aligned to max packet size. The mass storage gadgets do
not meet this requirement for Super Speed. The gadgets already have a
function which performs this alignment for CBW packets, so use it for
data packets too.

The alternative would be to implement bounce buffers in the DWC3
driver, but that could have a significant impact on performance.

This version is based upon a more-correct patch written by Alan
Stern.
Signed-off-by: NPaul Zimmerman <paulz@synopsys.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

FSG_NUM_BUFFERS is set to 2 as default.
Usually 2 buffers are enough to establish a good buffering pipeline.
The number may be increased in order to compensate a for bursty VFS
behaviour.

Here follows a description of system that may require more than
2 buffers.
 * CPU ondemand governor active
 * latency cost for wake up and/or frequency change
 * DMA for IO

Use case description.
 * Data transfer from MMC via VFS to USB.
 * DMA shuffles data from MMC and to USB.
 * The CPU wakes up every now and then to pass data in and out from VFS,
   which cause the bursty VFS behaviour.

Test set up
 * Running dd on the host reading from the mass storage device
 * cmdline: dd if=/dev/sdb of=/dev/null bs=4k count=$((256*100))
 * Caches are dropped on the host and on the device before each run

Measurements on a Snowball board with ondemand_governor active.

FSG_NUM_BUFFERS 2
104857600 bytes (105 MB) copied, 5.62173 s, 18.7 MB/s
104857600 bytes (105 MB) copied, 5.61811 s, 18.7 MB/s
104857600 bytes (105 MB) copied, 5.57817 s, 18.8 MB/s

FSG_NUM_BUFFERS 4
104857600 bytes (105 MB) copied, 5.26839 s, 19.9 MB/s
104857600 bytes (105 MB) copied, 5.2691 s, 19.9 MB/s
104857600 bytes (105 MB) copied, 5.2711 s, 19.9 MB/s

There may not be one optimal number for all boards. This is why
the number is added to Kconfig. If selecting USB_GADGET_DEBUG_FILES
this value may be set by a module parameter as well.
Signed-off-by: NPer Forlin <per.forlin@linaro.org>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: NFelipe Balbi <balbi@ti.com>

This patch (as1481) fixes a problem affecting g_file_storage and
g_mass_storage when running at SuperSpeed.  The two drivers currently
assume that the bulk-out maxpacket size can evenly divide the SCSI
block size, which is 512 bytes.  But SuperSpeed bulk endpoints have a
maxpacket size of 1024, so the assumption is no longer true.

This patch removes that assumption from the drivers, by getting rid of
a small optimization (they try to align VFS reads and writes on page
cache boundaries).  If a command's starting logical block address is
512 bytes below the end of a page, it's not okay to issue a USB
command for just those 512 bytes when the maxpacket size is 1024 -- it
would result in either babble (for an OUT transfer) or a short packet
(for an IN transfer).

Also, for backward compatibility, the test for writes extending beyond
the end of the backing storage has to be changed.  If the host tries
to do this, we should accept the data that fits in the backing storage
and ignore the rest.  Because the storage's end may not align with a
USB packet boundary, this means we may have to accept a USB OUT
transfer that extends beyond the end of the storage and then write out
only the part of the data that fits.
Signed-off-by: NAlan Stern <stern@rowland.harvard.edu>
Acked-by: NMichal Nazarewicz <mina86@mina86.com>
Signed-off-by: NFelipe Balbi <balbi@ti.com>