The data fragments were stored in reversed order when the RX
data was saved into network buffers. This was caused by net_pkt
changes in commit db11fcd "net/net_pkt: Fully separate struct
net_pkt from struct net_buf".
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
This patch adds support for reading MAC address from I2C EEPROM.
Only chips with 7-bit I2C device address are supported.
Change-Id: Ibedc33e54e33bdb901840e104063e2f4752b9123
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
Convert code to use u{8,16,32,64}_t and s{8,16,32,64}_t instead of C99
integer types.
Jira: ZEP-2051
Change-Id: I4ec03eb2183d59ef86ea2c20d956e5d272656837
Signed-off-by: Kumar Gala <kumar.gala@linaro.org>
- net_pkt becomes a stand-alone structure with network packet meta
information.
- network packet data is still managed through net_buf, mostly named
'frag'.
- net_pkt memory management is done through k_mem_slab
- function got introduced or relevantly renamed to target eithe net_pkt
or net_buf fragments.
- net_buf's sent_list ends up in net_pkt now, and thus helps to save
memory when TCP is enabled.
Change-Id: Ibd5c17df4f75891dec79db723a4c9fc704eb843d
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
There have been long lasting confusion between net_buf and net_nbuf.
While the first is actually a buffer, the second one is not. It's a
network buffer descriptor. More precisely it provides meta data about a
network packet, and holds the chain of buffer fragments made of net_buf.
Thus renaming net_nbuf to net_pkt and all names around it as well
(function, Kconfig option, ..).
Though net_pkt if the new name, it still inherit its logic from net_buf.
'
This patch is the first of a serie that will separate completely net_pkt
from net_buf.
Change-Id: Iecb32d2a0d8f4647692e5328e54b5c35454194cd
Signed-off-by: Tomasz Bursztyka <tomasz.bursztyka@linux.intel.com>
Check net_recv_data() return value, if it returns an error release
the net_buf. Based on a fix in eth_mcux driver.
Change-Id: I44ca5fd8dfb7175620b7e8850a68443100039db6
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
Atmel SAM family GMAC Ethernet driver is implementing zero-copy
networking. As a result it has to reserve a defined amount of RX
data net buffers before bringing up the interface. Since net buffer
pool is initialized by the network stack and this driver was bringing
the interface up in its initialization function the driver initialization
was performed, as a workaround, after network stack initialization. It
is not a clean solution. This patch fixes this by bringing the
interface up in interface initialization function. The driver itself
can now be initialized before the network stack is.
Tested on Atmel SMART SAM E70 Xplained board
Change-Id: I65886fd6db6f27a10628e393cfabd8e5f78c08ff
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
Fix eth_tx function which was dereferencing a pointer before
checking that it is not null.
Tested on Atmel SMART SAM E70 Xplained board
Change-Id: Idae4cf9d9a80f6ee9f74a94dd1debe7511c5fab4
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
Networking stack has split one global DATA pool to RX and TX DATA pools
and also added net_buf pool support to each context. Update the driver
to support this new design. Since the GMAC TX descriptor list has a fixed
size but the number of TX DATA buffers is no longer limited updating the
TX descriptor list has to be guarded by a semaphore.
Tested on Atmel SMART SAM E70 Xplained board
Change-Id: I181e1cdd183e173b85d5d1711b6e78cd5165666d
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
The code used net_nbuf_get_reserve_{rx|tx}_data() function to
allocate a fragment. Instead of that low level function, use
net_nbuf_get_frag() instead. There are few places this is not
possible or is too big change like in few test programs.
Change-Id: Ied7e2b7db352de998b200ffa6ff82471bfa5ebe3
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
If we receive lot of packets, it might happen that we exhaust
all the DATA buffers in the system. This would prevent from
us sending anything to the network.
Change this by splitting the DATA buffer pool into RX and TX
parts. This way RX flooding cannot consume all DATA buffers
that needs to be sent.
Change-Id: I8e8934c6d5fdd47b579ffa6268721b5eb3d64b6d
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
The interface L2 address type is set at the same time as the
L2 address is set to the network interface. This is most
convinient place to set the address type.
Change-Id: I712d7357d075959eb79df3463141cfbc6d163a74
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
This commit changes the net_buf getter functions in nbuf.h
by adding a timeout parameter. These function prototypes
are changed to accept a timeout parameter.
net_nbuf_get_rx()
net_nbuf_get_tx()
net_nbuf_get_data()
net_nbuf_get_reserve_rx()
net_nbuf_get_reserve_tx()
net_nbuf_get_reserve_data()
net_nbuf_copy()
net_nbuf_copy_all()
net_nbuf_push()
net_nbuf_append()
net_nbuf_write()
net_nbuf_insert()
Following convinience functions have not been changed
net_nbuf_append_u8
net_nbuf_append_be16
net_nbuf_append_be32
net_nbuf_insert_u8
net_nbuf_insert_be16
net_nbuf_insert_be32
net_nbuf_write_u8
net_nbuf_write_be16
net_nbuf_write_be32
so they call the base function using K_FOREVER. Use the
base function if you want to have a timeout when net_buf
is allocated.
Change-Id: I20bb602ffb73069e5a02668fce60575141586c0f
Signed-off-by: Jukka Rissanen <jukka.rissanen@linux.intel.com>
This is a zero-copy networking implementation of Ethernet driver.
Limitations:
- one shot PHY setup, no support for PHY disconnect/reconnect
- no support for devices with DCache enabled due to missing
non-cacheable RAM regions in Zephyr.
Tested on Atmel SMART SAM E70 Xplained board
Origin: Original
Jira: ZEP-1492
Change-Id: Ib944f91193efbd12c1142b0bcf1f635388bf1b87
Signed-off-by: Piotr Mienkowski <piotr.mienkowski@gmail.com>
