The mtd var in fun_wait_rnb() is now unused, let's get rid of it and
fix the warning resulting from this unused var.
Fixes: 50a487e771 ("mtd: rawnand: Pass a nand_chip object to chip->dev_ready()")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200603134922.1352340-2-boris.brezillon@collabora.com
Instead of manipulating the statically allocated structure and copy
timings around, allocate one at identification time and save it in the
nand_chip structure once it has been initialized.
All NAND chips using the same interface configuration during reset and
startup, we define a helper to retrieve a single reset interface
configuration object, shared across all NAND chips.
We use a second pointer to always have a reference on the currently
applied interface configuration, which may either point to the "best
interface configuration" or to the "default reset interface
configuration".
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-29-miquel.raynal@bootlin.com
The ->choose_interface() hook is here for manufacturer drivers to
provide a better timing interface than the default one, this field is
not needed anymore.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-28-miquel.raynal@bootlin.com
The Kioxia/Toshiba TH58NVG2S3HBAI4 NAND memory is not ONFI compliant.
The timings of the NAND chip memory are quite close to ONFI mode 4 but
is breaking that spec.
By providing our own set of timings, erase block read speed is increased
from 6910 kiB/s to 13490 kiB/s and erase block write speed is increased
from 3350 kiB/s to 4410 kiB/s.
Tested on IMX6SX which has a NAND controller supporting EDO mode.
Signed-off-by: Rickard x Andersson <rickaran@axis.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-27-miquel.raynal@bootlin.com
This hook can be overloaded by NAND manufacturer drivers to propose
alternative timings when not following the main standards. In this
case, the manufacturer drivers is responsible for choosing the best
interface configuration that fits both the controller and chip
capabilities.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-23-miquel.raynal@bootlin.com
Extract the logic out of nand_choose_interface_config() to create a
public helper that can be reused by manufacturer drivers. Add the
possibility to provide a specific set of timings.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-22-miquel.raynal@bootlin.com
Warn the user if the parameters are wrong but basically it would mean
there is a serious issue in the NAND core. So no need to ever check
its output, let's make this helper return void.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-21-miquel.raynal@bootlin.com
The name/suffix data_interface is a bit misleading in that the field
or functions actually represent a configuration that can be applied by
the controller/chip. Let's rename all fields/functions/hooks that are
worth renaming.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
As a preparation for allocating the data interface structure
dynamically (and rename it), let's avoid accessing
chip->data_interface directly.
Instead, we introduce a helper, nand_get_interface_config(), and use
it to retrieve the current data interface configuration out of a
nand_chip object.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-19-miquel.raynal@bootlin.com
The ONFI parameter page of a chip might define more fine grained
tPROG_max and tBERS_max. When we do not have this information, we
default to the highest possible values (they are maxima anyway).
There is no point setting these fields at runtime, so explicitly move
these defaults to the main ONFI SDR timings structure. This way, we
will also be able to return a pointer to mode 0 directly when we will
create a default reset configuration.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-18-miquel.raynal@bootlin.com
These two values are already hardcoded in the default ONFI timing
structure, no need to redefine them here. Plus, we want to be able to
reference timing mode 0 easily and reliably, without extra
computation, so we get rid of the extra assignations.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-17-miquel.raynal@bootlin.com
Vendors are allowed to provide their own set of timings. In this case,
we provide a way to derive the "closest" timing mode so that, if the
NAND controller does not support tweaking these parameters, it will be
able to configure itself anyway.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-16-miquel.raynal@bootlin.com
Right now the core uses onfi_fill_data_interface() to initialize the
nand_data_interface object embedded in nand_chip, but we are about to
allocate this object dynamically and let manufacturer drivers provide
their own interface config. Let's patch the onfi_fill_data_interface()
so it can initialize an interface config that's not the one
currently attached to the nand_chip.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-14-miquel.raynal@bootlin.com
This name is a bit misleading, what we do in this helper is trying to
find the best SDR timings supported by the controller and the chip.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-12-miquel.raynal@bootlin.com
The data interface setup does not care about the default timing mode
but cares about the actual timing mode at the time of the call of this
helper.
Use this entry instead and let chip->default_timing_mode only be used
at initialization time.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-9-miquel.raynal@bootlin.com
Avoid relying just on the default timing mode to discriminate if the
data interface must be restored. This field should only be used
at initialization time by legacy chips statically defined. Do a
memcmp() instead.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-8-miquel.raynal@bootlin.com
Reorder fields in this structure and pack entries by theme:
* The main descriptive structures
* The data interface details
* Bad block information
* The device layout
* Extra buffers matching the device layout
* Internal values
* External objects like the ECC controller, the ECC engine and a
private data pointer.
While at it, adapt the documentation style.
I changed on purpose the description of @oob_poi which was weird.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-7-miquel.raynal@bootlin.com
Now that struct nand_manufacturer type is free, use it to store the
nand_manufacturer_desc and the manufacturer's private data.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-6-miquel.raynal@bootlin.com
It is currently called nand_manufacturer but could actually be called
nand_manufacturer_desc, like its instances, so that the former name is
left unused for now.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-5-miquel.raynal@bootlin.com
page_shift, phys_erase_shift, bbt_erase_shift, chip_shift, pagemask,
subpagesize and badblockbits are all positive values, so declare
them as unsigned.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529111322.7184-2-miquel.raynal@bootlin.com
Prepare the migration to a generic ECC engine by renaming the
nand_ecc_req structure into nand_ecc_props. This structure will be the
base of a wider 'nand_ecc' structure.
In nand_device, these properties are still named "eccreq" even if
"eccprops" might be more descriptive. This is just a transition step,
this field is being replaced very soon by a much wider structure. The
impact of renaming this field would be huge compared to its interest.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529002517.3546-14-miquel.raynal@bootlin.com
Both OneNAND and raw NAND bits can't be compiled if MTD is disabled
because of the if/endif logic in drivers/mtd/Kconfig. There is no need
for an extra "depends on MTD" in their respective Kconfig files.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>
Link: https://lore.kernel.org/linux-mtd/20200529002517.3546-12-miquel.raynal@bootlin.com
A few drivers use this property to describe GPIO pins used to sample
the NAND Ready/Busy state. Let's make it part of the generic binding
doc.
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/linux-mtd/20200519232454.374081-2-boris.brezillon@collabora.com
In some occasions task held spinlock (mac_filter_hash_lock),
while being rescheduled due to admin queue mutex_lock. The struct
i40e_spinlock asq_spinlock, which later expands to struct mutex
spinlock. Moved i40e_aq_set_vsi_multicast_promiscuous(),
i40e_aq_set_vsi_unicast_promiscuous(),
i40e_aq_set_vsi_mc_promisc_on_vlan(), and
i40e_aq_set_vsi_uc_promisc_on_vlan() outside of atomic context. Without
this patch there is a race condition, which might result in scheduling
while in atomic context. The race condition is between the thread, which
holds mac_filter_hash_lock, while trying to acquire an admin queue mutex
and a thread, which already has said admin queue mutex. The thread, which
holds spinlock, fails to acquire the mutex, which causes this thread to
sleep.
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Make possible for the i40e driver to bind to the new v710 for 5GBASE-T
NICs.
Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
As with other networking drivers, remove the unnecessary driver version
from the Intel drivers. The ethtool driver information and module version
will then report the kernel version instead.
For ixgbe, i40e and ice drivers, the driver passes the driver version to
the firmware to confirm that we are up and running. So we now pass the
value of UTS_RELEASE to the firmware. This adminq call is required per
the HAS document. The Device then sends an indication to the BMC that the
PF driver is present. This is done using Host NC Driver Status Indication
in NC-SI Get Link command or via the Host Network Controller Driver Status
Change AEN.
What the BMC may do with this information is implementation-dependent, but
this is a standard NC-SI 1.1 command we honor per the HAS.
CC: Bruce Allan <bruce.w.allan@intel.com>
CC: Jesse Brandeburg <jesse.brandeburg@intel.com>
CC: Alek Loktionov <aleksandr.loktionov@intel.com>
CC: Kevin Liedtke <kevin.d.liedtke@intel.com>
CC: Aaron Rowden <aaron.f.rowden@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Co-developed-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
A customer was chain-booting to provision his systems and one of the
steps was setting MFS. MFS isn't cleared by normal warm reboots
(clearing requires a GLOBR) and there was no indication of why Jumbo
Frame receives were failing.
Add a warning if MFS is set to anything lower than the default.
Signed-off-by: Todd Fujinaka <todd.fujinaka@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Detect and log information about pre-recovery mode when firmware
transitions to a recovery mode.
When a firmware transitions to a recovery mode it stores a number
of unexpected EMP resets in one of its registers. The number of EMP
resets ranging from 0x21 to 0x2A indicates that FW transitions
to recovery mode. Use these values to emit log entry about transition
process. Previously the pre-recovery mode may not have been detected
and there was no log entry when NIC was in pre-recovery mode.
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Piotr Kwapulinski <piotr.kwapulinski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Use jiffies to limit max waiting time for PF reset to succeed.
Previous wait loop was unreliable. It required unreasonably long time
to wait for PF reset after reboot when NIC was about to enter
recovery mode
Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Piotr Kwapulinski <piotr.kwapulinski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Remove all the unused defines as they are just dead weight.
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Move i40e_client.h to include/linux/net/intel/*
since its shared between i40iw and i40e.
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
