The driver remove() function prints a message when the operation
is completed. Make this a debug level message to avoid polluting
the kernel log wih too much information.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We use DPAA2_ETH_MAX_TX_QUEUES to dimension the array holding
information on Tx queues. At most, we can have one queue per cpu.
Until now we used the NR_CPUS macro to set the upper limit on number
of Tx queues. However, the platforms that the DPAA2 Ethernet driver
supports have at most 16 cores, whereas NR_CPUS is Kconfigurable and
can be much higher.
Avoid allocating memory we'll never use, by setting
DPAA2_ETH_MAX_TX_QUEUES to 16. Same for DPAA2_ETH_MAX_DPCONS.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We mistakenly allocate space for too many entries in the
scatter-gather table of multi buffer egress frames.
While it doesn't have a negative impact from a functional
point of view, it wastes resources so fix it.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cleanup code in free_tx_fd() that deals with S/G frames:
- remove local variables that aren't really needed
- in the frame sw annotation area, store the actual SG table
buffer size, which is needed on free, rather then recompute
it based on number of S/G entries
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Newer MC versions allow us to change link settings while the
interface is up. Only check interface status if we are using
an old version.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The DPAA2 Ethernet driver assumes the DPNI objects that it
uses to build network interfaces have a minimum supported
API version, but until now it did nothing to enforce this
requirement.
Add a check at probe time to make sure the DPNI object is
compatible with the set of MC commands we intend to use
on it.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Structure dpaa2_fas is naturally aligned, so no need to use
the __packed attribute explicitly.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove dpio_id field in struct dpaa2_eth_channel, which wasn't
used anywhere in the code.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Let the driver remove() function print an informative message
after it finishes removing the network interface, not at an
arbitrary point during cleanup.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For the link state interrupt, we used a dummy non-threaded
irq handler, which had the same implementation as the generic
irq_default_primary_handler() function.
Give up on using our own irq handler and let the kernel use
the generic one instead.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The "struct mc_command" is a very generic name for a global
kernel structure. Change its name in "struct fsl_mc_command".
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
MC portals may not be available at the initial probing attempt
due to dependencies on other modules.
Check the return value of the MC portal allocation function and
defer probing in case it's not available yet. For all other error
cases the behaviour stays the same.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Suggested-by: Nipun Gupta <nipun.gupta@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use netdev_alloc_frag() instead of kmalloc to allocate space for
the S/G table of egress multi-buffer frames.
This fixes a bug where an unaligned pointer received from the
allocator would be overwritten with the 64B aligned value,
leading to a wrong address being later passed to kfree.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a TODO file describing what needs to be added/changed before the driver
can be moved out of staging.
Signed-off-by: Razvan Stefanescu <razvan.stefanescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add driver information, link details and hardware statistics to be
reported via ethtool -S.
Signed-off-by: Razvan Stefanescu <razvan.stefanescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce the DPAA2 Ethernet Switch driver, which manages Datapath Switch
(DPSW) objects discovered on the MC bus.
Suggested-by: Alexandru Marginean <alexandru.marginean@nxp.com>
Signed-off-by: Razvan Stefanescu <razvan.stefanescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add the command build/parse APIs for operating on DPSW objects through
the DPAA2 Management Complex.
Signed-off-by: Razvan Stefanescu <razvan.stefanescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move the source files out of staging into their final locations:
- dpcon.c goes to drivers/bus/fsl-mc/, next to the core infrastructure
- dpcon-cmd.h gets merged into drivers/bus/fsl-mc/fsl-mc-private.h, next
to the other internally used APIs
- dpcon.h gets merged into include/linux/fsl/mc.h, exposing the public
API
Update references in the dpaa2-eth staging driver.
DPCON stands for Data Path Concentrator - an interface between DPIO
(Data Path IO) and its users (e.g. dpaa2-eth). You can read more about
DPIO in Documentation/networking/dpaa2/overview.rst
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Reviewed-by: Laurentiu Tudor <laurentiu.tudor@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move the source files out of staging into their final locations:
- dpbp.c goes to drivers/bus/fsl-mc/, next to the core infrastructure
- dpbp-cmd.h gets merged into drivers/bus/fsl-mc/fsl-mc-private.h, next
to the other internally used APIs
- dpbp.h gets merged into include/linux/fsl/mc.h, exposing the public
API
Update references in the dpaa2-eth staging driver.
DPBP stands for Data Path Buffer Pool - you can read more about the
object in Documentation/networking/dpaa2/overview.rst
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Reviewed-by: Laurentiu Tudor <laurentiu.tudor@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Drop dependency on ARCH_LAYERSCAPE (which in turn depends on ARM64),
thus allowing this driver to compile on all architectures supported
by the fsl-mc bus driver.
This was compile tested on:
- powerpc (corenet_basic_defconfig, ppc64_defconfig)
- x86 (i386_defconfig, x86_64_defconfig, needs CONFIG_OF)
- arm64 (defconfig)
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
One MC command structure got away with using uXX fields instead
of __leXX. Fix it.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The DPAA2 Ethernet driver incorrectly assumes virtual addresses
are always 64b long, which causes compiler errors when building
for a 32b platform.
Fix this by using explicit casts to uintptr_t where necessary.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move the source files out of staging into their final locations:
-mc.h include file in drivers/staging/fsl-mc/include go to include/linux/fsl
-source files in drivers/staging/fsl-mc/bus go to drivers/bus/fsl-mc
-overview.rst, providing an overview of DPAA2, goes to
Documentation/networking/dpaa2/overview.rst
Update or delete other remaining staging files -- Makefile, Kconfig, TODO.
Update dpaa2_eth and dpio staging drivers.
Add integration bits for the documentation build system.
Signed-off-by: Stuart Yoder <stuyoder@gmail.com>
[rebased, add dpaa2_eth and dpio #include updates]
Signed-off-by: Laurentiu Tudor <laurentiu.tudor@nxp.com>
[rebased, split irqchip to separate patch]
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Use the newly added DPIO service API to map cpu-affine DPIO services
to channels.
The DPAA2 Ethernet driver already had mappings of frame queues and
channels to cpus, but had no control over the DPIOs used. We can
now ensure full affinity of hotpath hardware resources to cores,
which improves performance and almost eliminates some resource
contentions (e.g. enqueue/dequeue busy counters should be close to
zero from now on).
Making the pull channel operation core affine brings the most
significant benefits. This ensures the same DPIO service will be
used for all dequeue commands issued for a certain frame queue,
which is in line with the way hardware is optimized.
Additionally, we also use affine DPIOs for the frame enqueue and
buffer release operations in order to avoid resource contention.
dpaa2_io_service_register() and dpaa2_io_service_rearm()
functions receive an affine DPIO as argument mostly for uniformity,
but this doesn't change the previous functionality.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The previous patch ensures Tx flow affinity for forwarded frames,
but for termination traffic the initial flow affinity is determined
based on the skb hash, which is expected to hit only a few Tx queues
when there is a small number of flows.
Instead, use XPS (transmit packet steering) to set netdevice queue
affinity to the sending core.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The driver xmit function chooses an egress FQ based on the current
core id. The network stack itself sets a mapping field in the skb
based on many things - the default one being a hash on packet fields,
which the current driver ignores.
This patch saves the ingress frame flow affinity information in the
skb. In case of forwarded frames, this info will then be used for Tx
and Tx confirmation hardware queue selection, ensuring all processing
of the given frame is done on a single core.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Aligning the Tx buffers at 64B is a performance optimization
recommendation, not a hard requirement.
Make optional the alignment of Tx FD buffers, without enforcing
a reallocation in case there is not enough headroom for it.
On Rx, we keep allocating buffers with enough headroom to allow
Tx alignment of forwarded frames.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For non-linear skbs we build scatter-gather frames and allocate
a new buffer for the S/G table in which we reserve the required
headroom, so the actual skb headroom size doesn't matter.
Rather than use a one-size-fits-all approach, decide when to
enforce headroom requirements on a frame by frame basis.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For Tx confirmed frames that have an error indication in the frame
descriptor, we look at the Frame Annotation Status field (in the
buffer headroom) for details on the root cause and then print
a debug message with that information.
While useful in initial development stages, it doesn't bring
enough added value to justify reserving 64B of headroom for all
Tx frames (FAS is only 8B long, but we must reserve chunks of 64B
from the hardware annotation area).
If we remove the need for FAS field from egress frames, we can
renounce hardware annotation completely, since FAS is the only
HWA field we currently use.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a counter for the number of egress frames that need to be
realloc'ed due to insufficient headroom space.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 4b2d9fe879 ("staging: fsl-dpaa2/eth: Extra headroom in RX
buffers") tried to avoid the performance penalty of doing skb
reallocations in the network stack for IP forwarded frames between
two DPAA2 Ethernet interfaces. This led to a (too) complicated
formula that relies on the stack's internal implementation.
Instead, it's safer and easier to just not request any guarantee
from the stack. We already double check in the driver the required
headroom size of egress frames and realloc the skb if needed, so
we don't need to add any extra code.
On forwarding between two of our own interfaces, there is no
functional change; for traffic forwarded from a different device or
generated on the core, skb realloc operations are moved from the stack
to our driver, with no visible impact on performance.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 4b2d9fe879 ("staging: fsl-dpaa2/eth: Extra headroom in RX
buffers") removes the software annotation (SWA) area from the RX
buffer layout, as it's not used by anyone, but fails to update the
macros for accessing hardware annotation (HWA) fields, which is
right after the SWA in the buffer headroom.
This may lead to some frame annotation status fields (e.g. indication
if L3/L4 checksum is valid) to be read incorrectly.
Turn the accessor macros into inline functions and add a bool param
to specify if SWA is present or not.
Fixes: 4b2d9fe879 ("staging: fsl-dpaa2/eth: Extra headroom in RX buffers")
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The needed headroom that we ask the stack to reserve for us in TX
skbs is larger than the headroom available in RX frames, which
leads to skb reallocations in forwarding scenarios involving two
DPNI interfaces.
Configure the hardware to reserve some extra space in the RX
frame headroom to avoid this situation. The value is chosen based
on the Tx frame data offset, the Rx buffer alignment value and the
netdevice required headroom.
The network stack will take care to reserve space for HH_DATA_MOD when
building the skb, so there's no need to account for it in the netdevice
needed headroom.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The WRIOP hardware block v1.0.0 (found on LS2080A board)
requires data in RX buffers to be aligned to 256B, but
newer revisions (e.g. on LS2088A, LS1088A) only require
64B alignment.
Check WRIOP version and decide at runtime which alignment
requirement to configure for ingress buffers.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When configuring the Tx buffer layout, the software annotation size is
mentioned, and MC accounts for it when configuring the frame
tx_data_offset. No need to handle it in the driver as well.
This results in 64B less memory allocated per frame.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since setup_dpni() became a bit too long, move the buffer layout
configuration to a separate function.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Clean up goto labels in a couple of functions, by
removing/renaming redundant ones.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Include firmware version in the driver information exported
through ethtool.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Early during probe the netdevice name is not initialized yet,
so use dev_err instead of netdev_err when printing error
messages.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
dpni_get_irq_status() also looks at the input value of its
status parameter, and if not null it automatically clears
from pending state the bits that are set there.
Use this feature to avoid a separate MC command for clearing
the interrupt event bits after reading the status.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case we fail to allocate a skb for a fragmented ingress
frame, the cleanup function will attempt to unmap again the
first frame fragment, which had already been unmapped during
early Rx processing.
Avoid this by freeing the first buffer immediately in case
we hit an error, leaving the cleanup function to free only
the subsequent buffers.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Take into consideration the return value of napi_complete_done(),
since there might be an indication that it's not suitable to
enable driver interrupts yet.
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case dpaa2_io_service_rearm() fails with an error other then
EBUSY, it will do so silently; add a check for this and a warning
message, as a failure here means we're unable to receive any
more traffic on the current cpu.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For scatter-gather ingress frames, we expect to receive
a list of fragments from the hardware, last of which is
marked with a "final" bit.
Add a check to make sure the Rx frame has this bit set
correctly; there's not much we can do in case of a
malformed frame, but at least issue a warning.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On Rx path, if we fail to build an skb from the incoming FD,
we still need to update the channel buffer count accordingly,
otherwise we risk depleting the pool while the software counter
still sees available buffers.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We incorrectly assumed that dpaa2_io_release() can only
return -EBUSY as an error code, when in fact it can also
fail in case some of its arguments don't have valid values.
Make sure we only retry the operation while the portal is
busy and abort for all other error cases, otherwise we risk
entering an endless loop.
Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: Bogdan Purcareata <bogdan.purcareata@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix the values of DPAA2_FD_CTRL_FSE and DPAA2_FD_CTRL_FAERR,
which are shifted off by one bit.
Fixes: 39163c0ce0 ("staging: fsl-dpaa2/eth: Errors checking update")
Cc: <stable@vger.kernel.org>
Signed-off-by: Horia Geantă <horia.geanta@nxp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixed alignment so that it matched open parenthesis
Signed-off-by: Srishti Sharma <srishtishar@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Ioana Radulescu