Age | Commit message (Collapse) | Author |
|
Support PTPd 1588 stack by adding PTP 1588 clock
using the dTSEC
Signed-off-by: Yangbo Lu <yangbo.lu@freescale.com>
Change-Id: I4f6fb9a721cede75c6cf23560014b55b46c84fef
Reviewed-on: http://git.am.freescale.net:8181/20295
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Xiaobo Xie <X.Xie@freescale.com>
Reviewed-by: Mandy Lavi <Mandy.Lavi@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
Support for FSL_FMAN_CPC_STASH option was partially implemented.
The FMAN operation mapping didn't allow support for write transaction
stashing. Write transaction stashing support is required in certain
data path intensive tasks. Without this support "FMAN writes" to DDR
take more time complete, with this option FMAN transactions are stashed
to CPC. Stashing reduces the overall completion time for FMAN transactions.
This in turn prevents buffer over run issues in FMAN. Without this support
it has been observed, while handling high traffic rate FMAN starts reporting
Rx errors.
Signed-off-by: Varun Sethi <Varun.Sethi@freescale.com>
Change-Id: Iae7b2e9108204e64d336feb39517f4c72235feb4
Reviewed-on: http://git.am.freescale.net:8181/21040
Reviewed-by: Vakul Garg <vakul@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
LNI shaper is disabled by setting all 1's to both CR/ER
Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Change-Id: I7b680ecbcfcfa41860dddcca34044116719cda08
Reviewed-on: http://git.am.freescale.net:8181/22075
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Geoff Thorpe <Geoff.Thorpe@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
Pass device-tree usb node offset to each usb device structure.
This is required by rcpm driver to get mask setting for
ip exception register to make/remove usb as wake-up source
for sleep/deep-sleep
Signed-off-by: Ramneek Mehresh <ramneek.mehresh@freescale.com>
Change-Id: I4e6b0f931fb29a6311b36a97255d994c2ce0ad05
Reviewed-on: http://git.am.freescale.net:8181/22478
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Nikhil Badola <nikhil.badola@freescale.com>
Reviewed-by: Matthew Weigel <Matthew.Weigel@freescale.com>
|
|
Some functions are powerpc specific which cause error when the
driver is used on ARM platforms.
Added micro to make them only be used on powerpc platforms.
Signed-off-by: Tang Yuantian <Yuantian.Tang@freescale.com>
Change-Id: I7ab009fa13f47e23d67aef875185a23882dd1813
Reviewed-on: http://git.am.freescale.net:8181/23070
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Yang Li <LeoLi@freescale.com>
(cherry picked from commit 44a0893a220ef1906359aa283837fba8b18e20f7)
Signed-off-by: Matthew Weigel <Matthew.Weigel@freescale.com>
|
|
The hardware can automatically generate pause frames when the number
of free buffers drops under a certain threshold, but in order to do this,
the address of the last free buffer needs to be written to a specific
register for each RX queue.
This has to be done in 'gfar_clean_rx_ring' which is called for each
RX queue. In order not to impact performance, by adding a register write
for each incoming packet, this operation is done only when the PAUSE frame
transmission is enabled.
Whenever the link is readjusted, this capability is turned on or off.
Change-Id Ib4751d205a00c0813355cf23c4428bf6dcbda003
Signed-off-by: Matei Pavaluca <matei.pavaluca@freescale.com>
Change-Id: I22a836b86f256128ea1bd39e87902321030f7742
Reviewed-on: http://git.am.freescale.net:8181/23138
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
|
|
Local flow control options needed in order to resolve the negotiation
are incorrectly calculated.
Previously 'mii_advertise_flowctrl' was called to determine the local advertising
options, but these were determined based on FLOW_CTRL_RX/TX flags which are
never set through ethtool.
The patch simply translates from ethtool flow options to mii flow options.
Change-Id: I3ed7801eecaa6b5b8501ba1c8e860c92f959d559
Signed-off-by: Pavaluca Matei <matei.pavaluca@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/23137
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
The phy device supports 802.3x flow control, but the specific flags are not set
in the phy initialisation code. Flow control flags need to be added to the
supported capabilities of the phydev by the driver.
This is needed in order for ethtool to work ('ethtool -A' code checks for these
flags)
Change-Id: I71d8411a82777aff36bf98a09097eef1935ff113
Signed-off-by: Pavaluca Matei <matei.pavaluca@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/23136
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
CONFIG_MEMORY is required for IFC
Signed-off-by: Jaiprakash Singh <b44839@freescale.com>
Change-Id: I6605a908a4ee2bd0e8ef5b4f81456ee697138c63
Reviewed-on: http://git.am.freescale.net:8181/22565
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Matthew Weigel <Matthew.Weigel@freescale.com>
|
|
Define and use CONTROL_REGISTER_W1C_MASK to make sure that
w1c bits of usb control register do not get reset while
writing any other bit
Signed-off-by: Nikhil Badola <nikhil.badola@freescale.com>
Signed-off-by: Ramneek Mehresh <ramneek.mehresh@freescale.com>
Change-Id: Ie56814b6f6c4d05fbf862619330d2e505438e1c7
Reviewed-on: http://git.am.freescale.net:8181/22464
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Matthew Weigel <Matthew.Weigel@freescale.com>
|
|
Change string format for errata property in USB node that is
checked in device tree for applicablilty of corresponding erratum
Signed-off-by: Nikhil Badola <nikhil.badola@freescale.com>
Change-Id: Ic07fa5b5a8ce93ac3845332f23b3b71338956266
Reviewed-on: http://git.am.freescale.net:8181/22463
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Ramneek Mehresh <ramneek.mehresh@freescale.com>
Reviewed-by: Matthew Weigel <Matthew.Weigel@freescale.com>
|
|
The original u32 define limited the ceetm shaping speed.
Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Change-Id: I02fe21c541e5c78151b181b8ab75cc84d019661d
Reviewed-on: http://git.am.freescale.net:8181/22074
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Geoff Thorpe <Geoff.Thorpe@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
Ensure that idle line is asserted before leaving suspend function.
If not, fail the request as system could hang waiting for idle line
to be asserted which might never happen.
Signed-off-by: Jeffrey Ladouceur <Jeffrey.Ladouceur@freescale.com>
Change-Id: I95390751160b8825d58d9c04a6884b37e91424b4
Reviewed-on: http://git.am.freescale.net:8181/21956
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Geoff Thorpe <Geoff.Thorpe@freescale.com>
Reviewed-by: Haiying Wang <Haiying.Wang@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
|
|
mpc85xx_pmc_set_wake() is only for powerpc platform. For LS1 platform,
no such function. Use a weak function definition to void compile
problem.
Signed-off-by: Chenhui Zhao <chenhui.zhao@freescale.com>
Change-Id: I84ab02a65addabd2f8d7a9681e5f80c7b06c5ad0
Reviewed-on: http://git.am.freescale.net:8181/22821
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Yang Li <LeoLi@freescale.com>
Reviewed-by: Zhengxiong Jin <Jason.Jin@freescale.com>
|
|
This variable should represent the clock number per PLL.
So, rename it and re-calculate it.
Signed-off-by: Tang Yuantian <Yuantian.Tang@freescale.com>
Change-Id: I2817f76a076d468073a870edc47e41ef43ffa07f
Reviewed-on: http://git.am.freescale.net:8181/22795
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Zhengxiong Jin <Jason.Jin@freescale.com>
|
|
This patch only for LS1021A PCIe Power Managment, because there has
a PCIe hardware issue, we need a workaround to support it.
LS1021A Workaround for internal TKT228622 to fix the INTx hang issue.
If this hardware be fixed, we not need to do anything for PCIe PM.
Signed-off-by: Wang Dongsheng <dongsheng.wang@freescale.com>
Change-Id: I7b689e4780322f77bebd8671da71d82f6fdb18a3
Reviewed-on: http://git.am.freescale.net:8181/22585
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Minghuan Lian <Minghuan.Lian@freescale.com>
Reviewed-by: Zhengxiong Jin <Jason.Jin@freescale.com>
|
|
Merge remote-tracking branch 'dpaa/dpa-offload-devel'
|
|
There are 2 different VSP profile parameters in the FMD API that need to
be filled in depending whether the policer is enabled or disabled. The
dpa_classifier was filling in the enqueue VSP profile all the time, but
this was used only when the next action was "enqueue". When the next
action is "policer" the policer VSP profile must be filled in.
Change-Id: I09e2a6fad7216e7ce7b3c1190a4441b7a9825705
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/22064
Reviewed-by: Francois Massot-Pellet <Francois.Massot-Pellet@freescale.com>
|
|
The dpa_stats_init function failed to initialized the dpa_stats instance Id
by declaring it unused. This causes problems when using the rest of the
dpa_stats API as the instance Id is checked against negative values.
Change-Id: I43e460e84e2f22c995bcadf64d165b57ec6e789c
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/20956
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
|
|
This commit adds support in dpa_classifier in order to retrieve
miss statistics for a given table.
Change-Id: Iea61107e05d13abf70d19aeb8c970c401c1562fc
Signed-off-by: Bogdan Constantin Popescu <bogdan.c.popescu@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19618
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
There is an unreachable code paragraph in function "table_delete_entry_by_ref".
The function doesn't work for indexed tables. In the first half, an error
is issued in this case. However, in the second half, another code paragraph
is testing again if the table type is "indexed". This branch will never
be reachable and was removed.
Change-Id: Ib200f49eafca8131d4c1696af2e61753ede41a3d
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19135
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
There is a useless assignment of variable "hmd" in function
"hash_table_modify_entry". It was removed.
Change-Id: I2710a5d13916c8ab56a6e2c8218db721bf61abe5
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19134
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
In the DPA statistics wrapper, in the function
"do_ioctl_stats_modify_class_counter" the pointer prm.params.pair is
dereferenced immediately after it has been freed up. The assignment
espressions were useless hence they were discarded.
Change-Id: Idbfc3865413b15d9713803fceb91956cc618d930
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/18812
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
|
|
In the DPA statistics wrapper, in function
"do_ioctl_stats_create_class_counter" the following pointers were allocated
using a bad size (because of an additional '*' added by mistake to the
"sizeof" operator argument):
tbl->keys
cnode->keys
The size for these arrays of pointers should be
(NUM x sizeof(struct dpa_offload_lookup_key *)) instead of
(NUM x sizeof(struct dpa_offload_lookup_key)).
tbl->pairs
The correct size of this array should be
(NUM x sizeof(struct dpa_offload_lookup_key_pair *)) instead of
(NUM x sizeof(struct dpa_offload_lookup_key_pair)).
Change-Id: I190d76973107300171e6f3033d8a79a3482fdd73
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/18811
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
|
|
Two memory leaks were fixed in DPA statistics wrapper.
In function "do_ioctl_stats_get_counters" the allocated memory for
"prm.req_params.cnts_ids" leaks in some cases of error and even in
successful function return.
In function "dpa_stats_tbl_cls_compatcpy" the memory allocated for "us_keys"
may leak in case the key type is DPA_STATS_CLASSIF_SINGLE_KEY.
Change-Id: Ifffd8d67f972473b75a43190155754d1af68fe84
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/18810
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
|
|
In function "dpa_classif_mcast_create_group_error" in the dpa_classifier
driver, the allocated memory for "replic_grp_params" is leaking in case
of errors. This change adds a release of this memory in case it is
allocated when the function returns an error.
Change-Id: Ic0f537f3e980d498f11ecf6c2874674b7003b2a9
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/18809
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
In function "dpa_classif_mcast_create_group" from the dpa_classifier
driver there is a danger of dereferencing pgroup->entries pointer which
can be NULL in case of an error. To avoid this, in case of error this
pointer is checked before using it to release resources. Also, in
order to keep consistency of pgroup->entries in case of an allocation
error of pgroup->member_ids, the allocation of this structure and the
allocation of pgroup->member_ids have swicthed places.
Change-Id: I3d7ec1b01983a9c917f54d596e35afd8ab3539cc
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/18808
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
This patch adds a check for the maximum value of the IV
provided to an outbound SA.
Change-Id: I102aeb893100222171103a738f0bfa20ed17f33c
Signed-off-by: Radu Bulie <radu.bulie@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19121
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
This patch adds an extra check for the number of
policies defined per SA, limiting it to the maximum
possible values from a ccnode
Change-Id: I7ff1e9400e2357a53ee01d9f8d8ac5f15d3f8c40
Signed-off-by: Radu Bulie <radu.bulie@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19120
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
This patch adds a fix for alg_suite value that could be tainted
from the ipsec wrapper.
Signed-off-by: Radu Bulie <radu.bulie@freescale.com>
Change-Id: I288f71f11d0a84311dcf966f4c7a693ff754a34b
Reviewed-on: http://git.am.freescale.net:8181/19119
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
In function set_cnt_classif_node_retrieve_func the
switch was made based on enum dpa_stats_classif_node_type
but the case values were taken from enum dpa_cls_tbl_type.
Fix it by using case values from the correct enum.
Signed-off-by: Anca Jeanina FLOAREA <anca.floarea@freescale.com>
Change-Id: I041ad4ad56d84d9fd1f8424609c0efbe2da5c726
Reviewed-on: http://git.am.freescale.net:8181/5174
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Aurelian Zanoschi <Aurelian.Zanoschi@freescale.com>
Reviewed-by: Rivera Jose-B46482 <German.Rivera@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/19117
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
Fix DPA Stats issues related to tainted values by adding
constraints to user-provided parameters.
Signed-off-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
Change-Id: Ib59f89518d96c9102a71be4c673e2d065b0a71f8
Reviewed-on: http://git.am.freescale.net:8181/19070
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
Although an error code was set in case an error occurred, the function
always returned 0. Fix this by returning the error code.
Signed-off-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
Change-Id: I8fdc4c514fee0b77f4831bc976f5a8e69f28238e
Reviewed-on: http://git.am.freescale.net:8181/18804
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
In function dpa_ipsec_free the SA returned by function
get_sa_from_sa_id can be NULL in case of an error. Fixed
this by adding BUG_ON on returned SA.
Signed-off-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
Change-Id: Id663dec2c47cd2eacce8c2a5ca1f00639df08bba
Reviewed-on: http://git.am.freescale.net:8181/18802
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
The value returned by FM_PCD_PlcrProfileGetCounter function
is cast to uint64_t while the variable to store it in is
uint32_t. Changed the variable from uint32_t to uint64_t.
Signed-off-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
Change-Id: I6c6e7be411ca66ea62541ea860ad39b8d660be66
Reviewed-on: http://git.am.freescale.net:8181/18801
Reviewed-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
Tested-by: Marian-Cornel Chereji <marian.chereji@freescale.com>
|
|
management functions
The two functions that are managing the IPSec sequence numbers,
dpa_ipsec_sa_modify and dpa_ipsec_sa_request_seq_number, don't have
IOCTL support for working in compat mode (32bit user space applications
calling a 64bit kernel). This change adds this support.
Signed-off-by: Grigore Ion <ion.grigore@freescale.com>
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Change-Id: I8ce0514d0d40aee7155c2d17dd7cacb52736437a
Reviewed-on: http://git.am.freescale.net:8181/15676
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
Due to the recent aggregation that the dpa_classifier is performing for
chained header manip nodes, a problem has occured in the update HM
operation. This particular operation can consist of either one or two
HM nodes, depending on what HM options the user is selecting in the
parameters. This header manip operation can include an optional IP
fragmentation. It can also consist ONLY of IP fragmentation at user's
request. However, in this last case, the allocation of underlying HM
nodes was incorrect and it was fixed.
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Change-Id: Idb1ebe3cb518e1fd7a9d52fb0423e9e0e9b7900c
Reviewed-on: http://git.am.freescale.net:8181/15621
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
Possibility to modify at runtime the IP fragmentation params in the
update header manipulation operation (created via dpa_classif_set_update_hm)
was disabled from the DPA classifier driver. This change enables it.
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Change-Id: I4ed91e941a810235503dc23f24b6ebb4e563be90
Reviewed-on: http://git.am.freescale.net:8181/15223
Reviewed-by: Anca Jeanina Floarea <anca.floarea@freescale.com>
|
|
Added possibility to display some more debugging information.
Change-Id: I66a7a110953b6dbcea15b3ee983cd5814fc65ce7
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/15222
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
Added the reparsing option to the DPA classifier driver API so that
users can request a relaunch of the FMan parser immediatley after
the configured header manipulation operation.
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Change-Id: I3e6c9ea03da57e65e21a411fda9ba3cb5a3d5e87
Reviewed-on: http://git.am.freescale.net:8181/15221
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
The DPA classifier driver allows multiple HM operations to be aggregated
in the same HM node only when importing the HM node resources. When
creating the nodes dynamically, the driver does not search for candidates
for aggregation before creating the HM nodes.
This update improves the dpa_classifier driver to actively check for
HM operation aggregation possibilities to certain neighbor HM nodes even
when using the API in dynamic mode.
Signed-off-by: Marian Chereji <marian.chereji@freescale.com>
Change-Id: I96bb3eb62be81e3d1c3ac83947f32bee230ace38
Reviewed-on: http://git.am.freescale.net:8181/15220
Reviewed-by: Radu-Andrei Bulie <Radu.Bulie@freescale.com>
|
|
Multicast member with id 0 could not be removed.
This constraint was wrong in the context when
a new member was added to a multicast group and the user
no longer wanted to replicate the frame on member with id 0.
This patch fix the above issue.
Change-Id: Ia4c86687ef5693936933f7c67ee4c52550227d6d
Signed-off-by: Radu Bulie <radu.bulie@freescale.com>
Reviewed-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/14841
|
|
Add the GFAR_DBG_LOOP config option and debugfs support
to be able to configure packet forwarding loops in the
Gianfar driver. This config option is disabled by default
and should be used only for debugging/ benchmarking
purposes.
For details on how to configure a packet forwarding debug
loop refer to the help section of CONFIG_GFAR_DBG_LOOP.
Change-Id: I24d440edc4c7d63a055205e022b743e30f58409b
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/22861
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
Tested-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
The hm descriptor for an entry of type multicast was not fetched correctly.
This patch fixes the above issue.
Change-Id: I4ec4bad9ae3d65fe997b5b9047ef705c2eb052b5
Signed-off-by: Radu Bulie <radu.bulie@freescale.com>
Reviewed-by: Marian Chereji <marian.chereji@freescale.com>
Reviewed-on: http://git.am.freescale.net:8181/14840
|
|
If RMan dts node does not contain child node 'fsl,rman-global-cfg'
or the child node does not contain property 'reg', RMan global
register will not be initialized correctly. The patch is to fix
this issue.
Signed-off-by: Minghuan Lian <Minghuan.Lian@freescale.com>
Change-Id: Ifc781a28b3f6c59efe7fe44c37fd6f7c0e5efc1a
Reviewed-on: http://git.am.freescale.net:8181/21985
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Tiefei Zang <tie-fei.zang@freescale.com>
Reviewed-by: Richard Schmitt <richard.schmitt@freescale.com>
|
|
Merge remote-tracking branch 't1024-linux/master'
|
|
DPA Shared Ethernet has no plan to support NAPI so
far, so setup a dummy net device for NAPI.
Signed-off-by: Pan Jiafei <Jiafei.Pan@freescale.com>
Change-Id: I2cdd19c0aa7cfe64a0a7fb4f1397b5a0e24cc5f6
Reviewed-on: http://git.am.freescale.net:8181/21951
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Shengzhou Liu <Shengzhou.Liu@freescale.com>
|
|
Fqs back to outbound OP must have a higher priority compared
to fqs from core to outbound OP, otherwise the buffer will
be exhausted.
Signed-off-by: Pan Jiafei <Jiafei.Pan@freescale.com>
Change-Id: Ia4de53493d06768f86c617340927e2af4d722011
Reviewed-on: http://git.am.freescale.net:8181/21442
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Tested-by: Shengzhou Liu <Shengzhou.Liu@freescale.com>
Reviewed-by: Shengzhou Liu <Shengzhou.Liu@freescale.com>
|
|
This can avoid buffer pool is exhausted by others,
so that it can avoid deadlock for outbound OP.
Signed-off-by: Pan Jiafei <Jiafei.Pan@freescale.com>
Change-Id: I6fd1a604e0ae004837e02a4ac925360967da9827
Reviewed-on: http://git.am.freescale.net:8181/21441
Tested-by: Review Code-CDREVIEW <CDREVIEW@freescale.com>
Reviewed-by: Jianhua Xie <jianhua.xie@freescale.com>
Reviewed-by: Shengzhou Liu <Shengzhou.Liu@freescale.com>
|