Age | Commit message (Collapse) | Author |
|
Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Change-Id: I21558e105b5c3092e96b7e98b570e5613e8bc771
Reviewed-on: http://git.am.freescale.net:8181/17399
Reviewed-by: Cristian Bercaru <cristian.bercaru@freescale.com>
|
|
fm_soc_resume
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: I888b95ca0c7f2e13266ecf81c41e779666a6e612
Reviewed-on: http://git.am.freescale.net:8181/19106
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
In FMan v3, special transmission queues, called recycle queues, were created
because it was believed that the FCO flag from the FD is not enough to
differentiate between frames that are recyclable and those who are not. The
frame queue flags were used to signal the BMI to release the buffers back to
the default buffer pool.
The last findings and FMan improvements permit recycling activation only by
using FD fields. With this patch, the recycling frame queues are not needed
anymore.
This patch also removes the code dependency on the FMan version and improves
TCP traffic in forwarding scenarios due to single transmission queues, reducing
the out-of-order segments.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Change-Id: Idd4fea4ef7883563df67cf21dea9e87435b2dee8
Reviewed-on: http://git.am.freescale.net:8181/16872
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Tested-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
|
|
.resume
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
auto_res_snmp_entry
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: I8b02d8538c564fb2deb849a2053c5a6b82661fc6
Reviewed-on: http://git.am.freescale.net:8181/17192
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
|
|
The current sequence fixes the order of operations executed
when entering deep sleep state.
It also implements a workaround for a hardware bug by writing
to fm_cld register - disabling DMA after a short idle period.
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: I2c73406543495dad5a7fb240a2e8fab01b81e05e
|
|
Due to external constraints (modules loading order and registry
configuration order in Auto Response mode) we need to implement
suspend/resume instead of suspend_noirq/resume_noirq callbacks.
Signed-off-by: Cristian Sovaiala <cristian.sovaiala@freescale.com>
|
|
Macless is a type of generic DPAA driver that does not know its immediate
underlying HW block on TX. It can be connected with an OH port, a TX port,
directly with another Macless driver in Macless-Macless scenario or with
another HW module.
Because not all HW modules know how to calculate CSUM, Macless triggers
software computation of the CSUM. If another HW block modifies the frame it
needs to trigger the Parser to recalculate CSUM.
This patch adds different default values for the parse results structure
that forbids the Parser from doing CSUM validation or computation. The _DTC and
_RPD flags inside the Frame Descriptor are enabled together. This is the only
way to deactivate HW CSUM in all the flows.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Change-Id: I6f2e27825f082c9319ff44241d19b06bda3faf47
Reviewed-on: http://git.am.freescale.net:8181/13671
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
|
|
Remove unused local variable and unnecessary CGR function calls. Macless does
not use CGR capabilities or specific port fqs.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Change-Id: I13ac2037266948a22ab16606bd65e9e965d9aa32
Reviewed-on: http://git.am.freescale.net:8181/13672
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
|
|
In unidirectional flows, like UDP, generated by kernel through onic driver, the
traffic can block because of numerous frames that are in HW and not released
fast enough by the driver.
This patch adds a timer that releases frames periodically, regardless of the
reception and transmission rates.
This mechanism was chosen instead of the workqueue mechanism because of its
higher priority, larger than RX and TX procedures.
Change-Id: I141fcb7c2857a726e5e3954cabefb923c8ac0aac
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/13569
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
|
|
This will enable increasing the total allowed fifo-size
to more than 120KB for FManV1 parts (up to 160KB)
for the purpose of better traffic performance in case
that jumbo frames are involved
Change-Id: I3e357e08656af7d4f6c7e1c071212ccc81825449
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/13297
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Thomas Trefny <Tom.Trefny@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Add debug option to divert traffic received on one DPAA interface
directly to another DPAA interface. By default this option is disabled.
When enabled special debugfs entries appear for each DPAA interface
that allow changing the divert state from disabled (-1) to enabled by
writing the Tx interface number in them.
Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Change-Id: I7c92d074980dadef55077e43616cb3caa2787838
Reviewed-on: http://git.am.freescale.net:8181/13043
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
The draining function will be executed when each CPU has added more than 8
buffers in the draining pool. In some conditions, the sock buffers waits for
the Ethernet driver to release the buffer in order to continue its
transmission.
To avoid memory depletion in the sock buffers, the Ethernet driver will release
faster the received skb from the stack.
This patch makes onic to call the drain procedure on RX path also even if there
is only one buffer in the bpool.
Change-Id: I19f800bbd17c23fda5739b7613fab3da355c6fbc
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/13095
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Bogdan Hamciuc <bogdan.hamciuc@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: If804bfb9694ad08330e7a7c083cec9970ad4954d
Reviewed-on: http://git.am.freescale.net:8181/13062
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Change-Id: I5ef4f9bd5f0cb1d9b43d8bcc4ac7b1efb3b0bd38
Reviewed-on: http://git.am.freescale.net:8181/12856
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
(cherry picked from commit 39967816be45a10552e25ec27e2d20de0a466534)
Reviewed-on: http://git.am.freescale.net:8181/13060
|
|
This fixes the problem that if at a time, a path between a keygen and
a CCtable is broken (e.g. by deleting an entry in a previous table)
then it is impossible to add any entry in that table or in any tables
that are "behind" it until a new path would be created
Also fixing the case w.r.t. to share CC nodes when
deleting entry in a CC node attached to first port will corrupt
context of ccnode shared with other ports.
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Change-Id: I089982acee71c489b6ed9275b641afa74315c18f
Reviewed-on: http://git.am.freescale.net:8181/12820
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
(cherry picked from commit e25a50bbcd56f0c1f4a725b70eab5f75604d14a3)
Reviewed-on: http://git.am.freescale.net:8181/13059
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: Ica2eaeb13ecd0a2ad3084ad427a3e2f2da86b21f
Reviewed-on: http://git.am.freescale.net:8181/13019
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
started_ar_enter variable is defined inside CONFIG_PM so
guard the variable all over it is referenced.
Signed-off-by: Cristian Sovaiala <cristian.sovaiala@freescale.com>
Change-Id: I3a9ea33b20c947909689f63fddd8152f5f38c9f9
Reviewed-on: http://git.am.freescale.net:8181/13010
Reviewed-by: Eyal Harari <Eyal.Harari@freescale.com>
Reviewed-by: Cristian Bercaru <cristian.bercaru@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: I90e7e97c9ff8a149893c8244cb0e8376c66eb2ad
Reviewed-on: http://git.am.freescale.net:8181/12925
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Tested-by: Richard Schmitt <richard.schmitt@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
The APIs are implemented in FMD so no need to have this check in
place.
Signed-off-by: Cristian Sovaiala <cristian.sovaiala@freescale.com>
Change-Id: Ia69b55b0f77600459ef211c624ff46cdbf1479e1
Reviewed-on: http://git.am.freescale.net:8181/12871
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Reviewed-by: Eyal Harari <Eyal.Harari@freescale.com>
|
|
The DPAA jumbo frames optimizations do not depend anymore on the FMan
dynamic resource allocation algorithm and are available on all
platforms.
Signed-off-by: Cristian Bercaru <cristian.bercaru@freescale.com>
Change-Id: Ia270d0aeba9801d67fa401bae4215ab1f25040fe
Reviewed-on: http://git.am.freescale.net:8181/12501
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
With this patch, the option FMAN_DISABLE_OH_TO_REUSE_RESOURCES is not
availeble anymore because it depends on
FMAN_RESOURCE_ALLOCATION_ALGORITHM, which no longer exists.
Signed-off-by: Cristian Bercaru <cristian.bercaru@freescale.com>
Change-Id: I67a1e45f8b8b11c9b191cae0d8c9b56bdc912876
Reviewed-on: http://git.am.freescale.net:8181/12502
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
49KB can accomodate buffers for 2x1G + 1xOH + 1xOP
Change-Id: Ib2da4bb8e9d393f207bb756cf1b5a3ab1e75cdd2
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12499
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
142KB can accomodate buffers for 5x1G + 1xOH + 1xOP
5*0x6400+0xa00+0x3200=140KB
Change-Id: I37d7a1713101796cf3ab942ad4981d0070df4a98
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12473
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Change-Id: Ided02763bb84061238a52d00bc3976a6681c5bfc
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12455
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Remove BQL support from Gianfar. The original upstream
commit that added BQL to Gianfar does not revert cleanly:
"d8a0f1b gianfar: Add support for byte queue limits."
BQL has been integrated in Gianfar by the upstream comunity
with no clear reason why would it be required / how would it
benefit to this particular driver. Only a few drivers have it
integrated. BQL adds processing overhead to the Tx path and
potential functional issues as it may stop the transmission,
which could result in tx timeout.
Change-Id: I71e0a99d9616bf86dd98c795a927c89f881b4b08
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12254
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Change-Id: I0080ab68d4658bf8eb855b6c4b43ed2526f2a878
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12366
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Change-Id: I54409ee49f85935ced3f9c2038a8dfcc801e6f6e
Signed-off-by: Mandy Lavi <mandy.lavi@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12365
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Nir Erez <nir.erez@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
|
|
Signed-off-by: Eyal Harari <Eyal.Harari@freesacle.com>
Change-Id: I2b775e3609510a0290e751adc1cf3726cf11904f
Reviewed-on: http://git.am.freescale.net:8181/12341
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Cristian-Constantin Sovaiala <Cristian.Sovaiala@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Signed-off-by: Scott Wood <scottwood@freescale.com>
Conflicts:
arch/sparc/Kconfig
drivers/tty/tty_buffer.c
|
|
In some complex scenarios, VSP cannot be enabled on Tx O/H port. To support
these scenarios, a new attribute was added to onic dts node.
The user should guarantee that other hw modules will release the buffers into
onic draining pool, otherwise the system will leak memory.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Signed-off-by: Ioana Tibuleac <ioana.tibuleac@freescale.com>
Change-Id: Ieb6d1c9cc7d165ad71165a58bd2dbd7b46a70dbc
Reviewed-on: http://git.am.freescale.net:8181/12332
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Reordered the fm_port_suspend/resume, set_wol API calls.
Besides that, in resume callback we do not suspend the port anymore
in case of error such as when the setting of MG bit fails.
The goto labels have been updated accordingly.
Change-Id: I1e8cbe29be4f30129162f0c152bd2914615b7b98
Signed-off-by: Cristian Sovaiala <cristian.sovaiala@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/12326
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Reviewed-by: Ruxandra Ioana Radulescu <ruxandra.radulescu@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Bogdan Hamciuc <bogdan.hamciuc@freescale.com>
Reviewed-by: Marian Cristian Rotariu <marian.rotariu@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
Set skb->data to point towards the origin of the mapped area
for recycled frames. The skb->data value is adjusted with the
fd offset on reception so it needs to point towards the origin
of the buffer as seen by hardware.
Signed-off-by: Madalin Bucur <madalin.bucur@freescale.com>
Change-Id: I33d96a6c56cc3c05a2331368c375ff0ce8431ae8
Reviewed-on: http://git.am.freescale.net:8181/12293
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Ruxandra Ioana Radulescu <ruxandra.radulescu@freescale.com>
Reviewed-by: Bogdan Hamciuc <bogdan.hamciuc@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
128 bytes"
This reverts commit 98345f7c55a5a52b30aa0982cec8644ad9636bdb.
Change-Id: I7cc841b30ed0e23728b40dd10057edb9c54d07fe
Reviewed-on: http://git.am.freescale.net:8181/12289
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
Tested-by: Jose Rivera <German.Rivera@freescale.com>
|
|
There are two FMAN V3 version: FMAN_V3H and FMAN_V3L. Some hardware
resources are different between the two version. T1040, T1042,
T1020, T1022 belong to FMAN_V3L, there are 4 Offline/Host Command
Ports (O/H n) in FMAN_V3L. But there are 6 Offline/Host Command
Ports (O/H n) in FMAN_V3H.
This patch is correcting some definition like INTG_MAX_NUM_OF_FM,
FM_MAX_NUM_OF_10G_MACS, FM_MAX_NUM_OF_1G_MACS and etc. according
to (T1040RM Rev D, 04/2014)
Signed-off-by: Jianhua Xie <jianhua.xie@freescale.com>
Change-Id: I3f7fa9b70fad30b49cd76da57429ab8ad08b5176
Reviewed-on: http://git.am.freescale.net:8181/12241
Reviewed-by: Jiafei Pan <Jiafei.Pan@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
The MTU known by the network device driver was initialized with a default value
(1500) and could not be validated in accordance with DPAA values.
This patch initializes the MTU with proper values and adds the posibility to be
changed and validated at run-time.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Change-Id: I44b9a9e18dbb45669e16913f0f5bdc290ec3e6b2
Reviewed-on: http://git.am.freescale.net:8181/11833
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
RX PCD fqs should have rx default handler not the error handler. This patch
solves this problem and clarifies some comments related to fqs initialization.
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Change-Id: I9624a3dfce594b2b4d0fe99567dfdd8ea3cab1e0
Reviewed-on: http://git.am.freescale.net:8181/11841
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
For phy devices that don't issue interrupts upon link
state changes, phylib polls the link state resulting in
repeated calls to adjust_link(), even if the link state
didn't change. As a result, some mac registers are
repeatedly read and written with the same values, which
is not ok.
To fix this, adjust_link() has been refactored to check
first whether the link state has changed and to take action
only if needed, updating mac registers and local state
variables. The 'new_state' local flag, set if one of the
link params changed (link, speed or duplex), has been
rendered useless and removed by this refactoring.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change-Id: I796d506bfd7193f9206c122d64844b43015c09a9
Reviewed-on: http://git.am.freescale.net:8181/11912
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
We need to keep the TX queues stopped throughout a reset,
without triggering the TX watchdog and regardless of the
link state. Given the way the watchdog works, the proper
way to do this is to use netif_device_{detach,attach}()
just as we do around suspend/resume.
netif_device_detach() can be called to stop TX queues in a
control operation other than ndo_stop, and in our case
this is insured by testing the driver's RESETTING state
flag.
napi_disable() is called before the stopping the TX queues
to make sure that the poll function is not running,
and therefore there can be no netif_wake_queue() calls.
(Inspired by sfc.)
Change-Id: Ie00ddaf39e454938b4ab5418d038de5d1a277b03
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/11994
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Rajan Gupta <rajan.gupta@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
ASF uses two TX queues to avoid contention. BSC9132 even though having
controller which can support two groups was having the DTS file written
for one group. To fix this we allow ASF to override the default tx queue
mapping with the "fsl,tx-bit-map" value given in the DT.
This patch resolves the crash seen in case of bsc9132 board
when the network interface is up and trying to xmit the packet.
Signed-off-by: Sandeep Malik <Sandeep.Malik@freescale.com>
---
CQ:ENGR311623
Change-Id: I9d61b0ac73674774f4b674836478048e9a0c8873
Reviewed-on: http://git.am.freescale.net:8181/12091
Reviewed-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
64 is not enough to handle IPv6-IPSec and it will impact both Linux
and ASF IPSec-IPv6 performance so updating the headroom to 128 bytes
CR:ENGR00309290
Signed-off-by: Alok Makhariya <B46187@freescale.com>
Change-Id: Iae44de3c58d616a846a5f04d2f3013f7c7581ad7
Reviewed-on: http://git.am.freescale.net:8181/12056
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Hemant Agrawal <hemant@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
This fixes a seg fault on 'ethtool -A' entry if the
interface is down. Obviously we need to have the
phy device initialized / "connected" (see of_phy_connect())
to be able to advertise pause frame capabilities.
Fixes: 23402bddf9e56eecb27bbd1e5467b3b79b3dbe58
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change-Id: I94f1a677dafca91e68029acce120e7417c75ff8f
Reviewed-on: http://git.am.freescale.net:8181/11514
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
When this option is enabled the driver will try to allocate
the buffer descriptors (BDs) to the L2 SRAM memory, using
the powerpc/85xx cache-sram support API for the P1/P2 QorIQ
platforms. For this to work, the size and address of the
SRAM region must be provided as kernel boot params.
Should the attempt to allocate in L2 SRAM fail, the driver
falls back to normal allocation of BDs as before (to DDR).
The gfar_l2sram_en driver module param is provided to allow
to disable the allocation to L2 SRAM at driver level
(enabled by default).
Change-Id: I087e97baf289111a0841fdf69592c56d3a9a0310
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/11624
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|
|
This patch adds ethtool handlers for some basic operations: set/get the
message level, get driver information, notify the user that the driver
does not have rings by setting rings parameters to 0 values and that some
MAC device related operations are not supported.
Change-Id: Id8f4d3e9152f48e94822c04e34c5894ffd00ec1b
Signed-off-by: Marian Rotariu <marian.rotariu@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/11659
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Madalin-Cristian Bucur <madalin.bucur@freescale.com>
Reviewed-by: Jose Rivera <German.Rivera@freescale.com>
|