Everything else starts with ICF so the last two should as well.
Fix places they are used to match.
Signed-off-by: Andy Grover <agrover@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
For PGR reservation of type Write Exclusive Access, allow all non
reservation holding I_T nexuses with active registrations to READ
from the device.
This addresses a bug where active registrations that attempted
to READ would result in an reservation conflict.
Signed-off-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
Now that fabric_max_sectors is no longer used to enforce the maximum
I/O size, go ahead and drop it's left-over usage in target-core and
associated backend drivers.
Cc: Christoph Hellwig <hch@lst.de>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Roland Dreier <roland@purestorage.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
This patch drops the arbitrary maximum I/O size limit in sbc_parse_cdb(),
which currently for fabric_max_sectors is hardcoded to 8192 (4 MB for 512
byte sector devices), and for hw_max_sectors is a backend driver dependent
value.
This limit is problematic because Linux initiators have only recently
started to honor block limits MAXIMUM TRANSFER LENGTH, and other non-Linux
based initiators (eg: MSFT Fibre Channel) can also generate I/Os larger
than 4 MB in size.
Currently when this happens, the following message will appear on the
target resulting in I/Os being returned with non recoverable status:
SCSI OP 28h with too big sectors 16384 exceeds fabric_max_sectors: 8192
Instead, drop both [fabric,hw]_max_sector checks in sbc_parse_cdb(),
and convert the existing hw_max_sectors into a purely informational
attribute used to represent the granuality that backend driver and/or
subsystem code is splitting I/Os upon.
Also, update FILEIO with an explicit FD_MAX_BYTES check in fd_execute_rw()
to deal with the one special iovec limitiation case.
v2 changes:
- Drop hw_max_sectors check in sbc_parse_cdb()
Reported-by: Lance Gropper <lance.gropper@qosserver.com>
Reported-by: Stefan Priebe <s.priebe@profihost.ag>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Roland Dreier <roland@purestorage.com>
Cc: stable@vger.kernel.org # 3.4
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
Commit 7854d3f749 ("mtd: spelling, capitalization, uniformity") added
a correctly spelled line, but failed to remove the wrongly spelled one.
Commit 064a7694b5 ("mtd: Fix typo mtd/tests") then fixed the spelling
again, but left the duplication.
Fixes: 7854d3f749 ("mtd: spelling, capitalization, uniformity")
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Merge misc fixes from Andrew Morton:
"12 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm, vmscan: prevent kswapd livelock due to pfmemalloc-throttled process being killed
memcg: fix destination cgroup leak on task charges migration
mm: memcontrol: switch soft limit default back to infinity
mm/debug_pagealloc: remove obsolete Kconfig options
vfs: renumber FMODE_NONOTIFY and add to uniqueness check
arch/blackfin/mach-bf533/boards/stamp.c: add linux/delay.h
ocfs2: fix the wrong directory passed to ocfs2_lookup_ino_from_name() when link file
MAINTAINERS: update rydberg's addresses
mm: protect set_page_dirty() from ongoing truncation
mm: prevent endless growth of anon_vma hierarchy
exit: fix race between wait_consider_task() and wait_task_zombie()
ocfs2: remove bogus check in dlm_process_recovery_data
H_RST bit in H_CSR register may be found lit before reset is started,
for example if preceding reset flow hasn't completed.
In that case asserting H_RST will be ignored, therefore we need to clean
H_RST bit to start a successful reset sequence.
Cc: <stable@vger.kernel.org> #3.10+
Signed-off-by: Alexander Usyskin <alexander.usyskin@intel.com>
Signed-off-by: Tomas Winkler <tomas.winkler@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
- plane handling refactoring from Matt Roper and Gustavo Padovan in prep for
atomic updates
- fixes and more patches for the seqno to request transformation from John
- docbook for fbc from Rodrigo
- prep work for dual-link dsi from Gaurav Signh
- crc fixes from Ville
- special ggtt views infrastructure from Tvrtko Ursulin
- shadow patch copying for the cmd parser from Brad Volkin
- execlist and full ppgtt by default on gen8, for testing for now
* tag 'drm-intel-next-2014-12-19' of git://anongit.freedesktop.org/drm-intel: (131 commits)
drm/i915: Update DRIVER_DATE to 20141219
drm/i915: Hold runtime PM during plane commit
drm/i915: Organize bind_vma funcs
drm/i915: Organize INSTDONE report for future.
drm/i915: Organize PDP regs report for future.
drm/i915: Organize PPGTT init
drm/i915: Organize Fence registers for future enablement.
drm/i915: tame the chattermouth (v2)
drm/i915: Warn about missing context state workarounds only once
drm/i915: Use true PPGTT in Gen8+ when execlists are enabled
drm/i915: Skip gunit save/restore for cherryview
drm/i915/chv: Use timeout mode for RC6 on chv
drm/i915: Add GPGPU_THREADS_DISPATCHED to the register whitelist
drm/i915: Tidy up execbuffer command parsing code
drm/i915: Mark shadow batch buffers as purgeable
drm/i915: Use batch length instead of object size in command parser
drm/i915: Use batch pools with the command parser
drm/i915: Implement a framework for batch buffer pools
drm/i915: fix use after free during eDP encoder destroying
drm/i915/skl: Skylake also supports DP MST
...
Currently vt_bind and vt_unbind access at least the con_driver object
and registered_con_driver array without holding the console lock. Fix
this by locking around the whole function in each case.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Peter Hurley <peter@hurleysoftware.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The default console driver (conswitchp) and busy drivers bound to a
console (as reported by con_is_bound()) shouldn't be unregistered.
System console drivers (without the CON_DRIVER_FLAG_MODULE flag) can be
unregistered, provided they are neither default nor busy. The current
code checks for the CON_DRIVER_FLAG_INIT flag but this doesn't make
sense: this flag is set for a driver whenever its associated console's
con_startup() function is called, which first happens when the console
driver is registered (so before the console gets bound) and gets cleared
when the console gets unbound. The purpose of this flag is to show if we
need to call con_startup() on a console before we use it.
Based on the above, do_unregister_con_driver() in its current form will
allow unregistering a console driver only if it was never bound, but
will refuse to unregister one that was bound and later unbound.
Fix this by dropping the CON_DRIVER_FLAG_INIT check, allowing
unregistering of any console driver provided that it's not the default
one and it's not busy.
v2:
- reword the third paragraph to clarify how the fix works (Peter Hurley)
v3:
- unchanged
v4:
- Allow unregistering a system console driver too, needed by i915 to
unregister vgacon. Update commit description accordingly. (Daniel)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Replace a misspelled function name by %s and then __func__.
This was done using Coccinelle, including the use of Levenshtein distance,
as proposed by Rasmus Villemoes.
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Replace the last argument of serial_paranoia_check by the actual function
name.
This was done using Coccinelle, including the use of Levenshtein distance,
as proposed by Rasmus Villemoes.
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some of FSL SoCs like T1040 has new version of UART controller which
can support 64byte FiFo.
To enable 64 byte support, following needs to be done:
-FCR[EN64] needs to be programmed to 1 to enable it.
-Also, when FCR[EN64]==1, RTL bits to be used as below
to define various Receive Trigger Levels:
-FCR[RTL] = 00 1 byte
-FCR[RTL] = 01 16 bytes
-FCR[RTL] = 10 32 bytes
-FCR[RTL] = 11 56 bytes
-tx_loadsz is set to 63-bytes instead of 64-bytes to implement
workaround of errata A-008006 which states that tx_loadsz should
be configured less than Maximum supported fifo bytes
Signed-off-by: Vijay Rai <vijay.rai@freescale.com>
Signed-off-by: Priyanka Jain <Priyanka.Jain@freescale.com>
Signed-off-by: Poonam Aggrwal <poonam.aggrwal@freescale.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If IGNPAR/INPCK are clear in termios->c_iflag, characters
received with parity or framing errors should be preserved
and passed to the upper layers of the tty stack.
Specifically, the decision of whether to set the character
value to zero should be made by n_tty.c/n_tty_receive_parity_error().
Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
Reviewed-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add RX DMA transfers support for samsung serial driver. It's enabled
when DMA controller for RX channel is specified in device-tree.
DMA transactions are started when number of bytes in RX FIFO reaches
trigger level, otherwise PIO mode is used. DMA transfer size is always
PAGE_SIZE which can cause large latency when smaller data amount is
transferred, so we always terminate DMA transaction on RX timeout
interrupt. Timeout interval is set to 64 frame times.
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add TX DMA transfers support for samsung serial driver. It's enabled
when "dmas" property is defined in serial device-tree node, otherwise
TX transfers are prerformed using PIO.
TX DMA is used for data segments larger than fifosize to reduce number
of interrupts during data transmission. For buffers shorter than fifosize
PIO mode is selected.
Data blocks for DMA transfers are aligned to cache line size to avoid
problems with coherency (some areas of TX circ buffer can be used by
CPU during DMA transaction, so we have to ensure that our data is always
consistent).
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add functions requesting and releasing RX and TX DMA channels. This
function are called only when "dmas" property in serial device-tree
node is defined.
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This macro definitions are necessary to implement DMA transfers
is samsung serial driver.
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When we have "dmas" property in serial node in device-tree, we do
memory alocation for dma structure which will be used in DMA handling
code.
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This struct contains DMA configuration for each serial port.
It will be used in next commits adding DMA support in driver.
Based on previous work of Sylwester Nawrocki and Lukasz Czerwinski.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After send out x_char in UART driver, x_char needs to be cleared
by UART driver itself, otherwise data in TXFIFO can no longer be
sent out.
Also tx counter needs to be increased to keep track of correct
number of transmitted data.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Other than enable Receiver Overrun Interrupt Enable (UCR4_OREN)
in start_tx interface, UCR4_OREN should be enabled before enable
of Receiver.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Terminating the DMA, make sure the interrupt is disabled, too.
This fixes random kernel Oops due to dma_tx_call() called for
invalid transmissions.
If we disable the TDMAEN, make sure it's enabled again if a TX
DMA is started.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently in dma_tx_callback(), no matter if there is still
remaining data pending in circle buffer or not, DMA transmit
will be terminated.
This will result in some data never get transmitted.
In order to fix this issue, call imx_dma_tx() again in
dma_tx_callback, when there is pending data and uart hasn't
been stopped.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This reverts commit 2ad28e3efe.
Instead of always wake up write_wait process in TX callback,
TX callback should call imx_dma_tx() again, and let imx_dma_tx
transfer the remaining data in circle buffer.
The issue with commit 2ad28e3 is, in case there is remaining
data in circle buffer, but no process is waiting on write_wait
queue, then as no following uart_write() will be called after
uart_write_wakeup(), thus cause data loss.
Moreover according to Documentation/serial/driver, uart_write_wakeup()
should be called in case the transmit buffer have dropped below
a threshold.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To synchronize between dma_tx_callback() and imx_dma_tx() use the same
variable, dma_is_txing. This prevents any race between these two functions
and ensures that a new DMA can start only after the first has been
finished.
Before the new DMA can be set up, update the circular buffer logic, first.
Therefore, change dma_is_txing after that update, instead of before.
While doing this, in dma_tx_callback() extend the locking to dma_unmap_sg()
and the update of dma_is_txing.
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
First, reformat the code to exit immediately. This allows us to add
more code in more readable format.
In case the TX DMA was still running, remove and disable it's resources.
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Inverting the logic of the if statement for the sg initialization
makes the if statement easier and better to read.
No functional change.
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use imx_start_tx() just to enable the TX interrupt. It's the job of the
TX interrupt ISR to fill the transmit buffer, then. If the transmit buffer
is empty, the TX interrupt should be executed as soon as the start_tx()
enables the interrupt, so there is no reason for the extra
imx_transmit_buffer() call, here. Remove it.
Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com>
Signed-off-by: Andy Lowe <andy_lowe@mentor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Several places are accessing the UCR1 register without locking.
This probably will cause a race issue when another thread
is accessing the same register.
Add locking to preventing concurrent access of the UCR1 register.
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Document the rfkill module parameters default_state and
master_switch_mode.
Signed-off-by: Andrew Clausen <andrew.p.clausen@gmail.com>
[rewrite commit message]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This patch fixes many bugs in the code dealing with the hardware handshake.
As an example, in atmel_set_termios(), we used to test whether the CRTSCTS
c_cflag was set. If so, we selected the "Hardware Handshake" mode through the
Mode Register. However, few lines below the mode was reset to "Normal" (0).
So there was no way to select the "Hardware Handshake" mode. To fix this issue,
we moved the CRTSCRTS c_cflag test AFTER the mode has been reset to "Normal".
Also setting the RTSEN and RTSDIS bits in the Control Register has different
results whether the USART is set in "Normal" or "Hardware Handshake" mode:
1) "Normal" mode
- the RTSEN bit forces the RTS line to low level, which tells the remote peer
that we are ready to received new data.
- the RTSDIS bit forces the RTS line to high level, which tells the remote peer
to stop sending new data.
2) "Hardware Handshake" mode
- the RTSEN bit forces the RTS line to high level.
- the RTSDIS bit lets the hardware control the RTS line.
WARNING:
when FIFOs are not available or not enabled, the RTS line is controlled by the
PDC. This is why using the Hardware Handshake mode requires using the PDC
channel for reception. However the Hardware Handshake mode DOES NOT work with
DMA controller since it cannot control the RTS line.
Future designs with FIFOs will introduce a new feature: the RTS line will be
controlled by the RX FIFO using thresholds. This patch was tested with this new
design.
Signed-off-by: Cyrille Pitchen <cyrille.pitchen@atmel.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
