priv is not instantiated at gfar_of_init() time, when
parsing the DT for info on supported HW queues. Before
the netdev can be allocated, the number of supported
queues must be known. Because the number of supported
queues depends on device type, move the compatibility
checks before netdev allocation. Local vars are used
to hold the operation mode info before netdev allocation.
This fixes the null accesses for priv->.., in gfar_of_init.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For the "fsl,etsec2" compatible models the driver currently
supports 8 Tx and Rx DMA rings (aka HW queues). However, there
are only 2 pairs of Rx/Tx interrupt lines, as these controllers
are integrated in low power SoCs with 2 CPUs at most. As a result,
there are at most 2 NAPI instances that have to service multiple
Tx and Rx queues for these devices. This complicates the NAPI
polling routine having to iterate over the mutiple Rx/Tx queues
hooked to the same interrupt lines. And there's also an overhead
at HW level, as the controller needs to service all the 8 Tx rings
in a round robin manner. The combined overhead shows up for multi
parallel Tx flows transmitted by the kernel stack, when the driver
usually starts returning NETDEV_TX_BUSY leading to NETDEV WATCHDOG
Tx timeout triggering if the Tx path is congested for too long.
As an alternative, this patch makes the driver support only one
Tx/Rx DMA ring per NAPI instance (per interrupt group or pair
of Tx/Rx interrupt lines) by default. The simplified single queue
polling routine (gfar_poll_sq) will be the default napi poll routine
for the etsec2 devices too. Some adjustments needed to be made to
link the Tx/Rx HW queues with each NAPI instance (2 in this case).
The gfar_poll_sq() is already successfully used by older SQ_SG_MODE
(single interrupt group) controllers.
This patch fixes Tx timeout triggering under heavy Tx traffic load
(i.e. iperf -c -P 8) for the "fsl,etsec2" (currently the only
MQ_MG_MODE devices). There's also a significant memory footprint
reduction by supporting 2 Rx/Tx DMA rings (at most), instead of 8,
for these devices.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are some concurrency issues on devices w/ 2 CPUs related
to the handling of Rx and Tx interrupts. eTSEC has separate
interrupt lines for Rx and Tx but a single imask register
to mask these interrupts and a single NAPI instance to handle
both Rx and Tx work. As a result, the Rx and Tx ISRs are
identical, both are invoking gfar_schedule_cleanup(), however
both handlers can be entered at the same time when the Rx and
Tx interrupts are taken by different CPUs. In this case
spurrious interrupts (SPU) show up (in /proc/interrupts)
indicating a concurrency issue. Also, Tx overruns followed
by Tx timeout have been observed under heavy Tx traffic load.
To address these issues, the schedule cleanup ISR part has
been changed to handle the Rx and Tx interrupts independently.
The patch adds a separate NAPI poll routine for Tx cleanup to
be triggerred independently by the Tx confirmation interrupts
only. Existing poll functions are modified to handle only
the Rx path processing. The Tx poll routine does not need a
budget, since Tx processing doesn't consume NAPI budget, and
hence it is registered with minimum NAPI weight.
NAPI scheduling does not require locking since there are
different NAPI instances between the Rx and Tx confirmation
paths now.
So, the patch fixes the occurence of spurrious Rx/Tx interrupts.
Tx overruns also occur less frequently now.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Programming the interrupt coalescing (IC) registers while
the controller/DMA is on may incur the loss of one Tx
confirmation interrupt, under certain conditions. This is
a subtle hw race because it does not occur during a burst
of Tx packets. It has been observed on p2020 devices that,
if just one packet is being xmit'ed, the Tx confirmation
doesn't trigger and BQL evetually blocks the Tx queues,
followed by Tx timeout and an un-responsive device.
This issue was not apparent prior to introducing BQL
support, as a late Tx confirmation was not an issue back then
and the next burst of Tx frames would have triggered the
Tx confirmation/ Tx ring cleanup anyway.
Bottom line, the hw specifications state that the IC registers
should not be programmed while the Rx/Tx blocks (the DMA) are
enabled. Further more, these registers are currently re-written
with the same values on the processing path, over and over again.
To fix this, rewriting the IC registers has been removed from
the processing path (napi poll). A complete MAC reset procedure
has been implemented for the ethtool -c option instead, to
reliably update these registers while the controller is stopped.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The device reset procedure, stop_gfar()/startup_gfar(), has
concurrency issues.
"Kernel access of bad area" oopses show up during Tx timeout
device reset or other reset cases (like changing MTU) that
happen while the interface still has traffic. The oopses
happen in start_xmit and clean_tx_ring when accessing tx_queue->
tx_skbuff which is NULL. The race comes from de-allocating the
tx_skbuff while transmission and napi processing are still
active. Though the Tx queues get temoprarily stopped when Tx
timeout occurs, they get re-enabled as a result of Tx congestion
handling inside the napi context (see clean_tx_ring()). Not
disabling the napi during reset is also a bug, because
clean_tx_ring() will try to access tx_skbuff while it is being
de-alloc'ed and re-alloc'ed.
To fix this, stop_gfar() needs to disable napi processing
after stopping the Tx queues. However, in order to prevent
clean_tx_ring() to re-enable the Tx queue before the napi
gets disabled, the device state DOWN has been introduced.
It prevents the Tx congestion management from re-enabling the
de-congested Tx queue while the device is brought down.
An additional locking state, RESETTING, has been introduced
to prevent simultaneous resets or to prevent configuring the
device while it is resetting.
The bogus 'rxlock's (for each Rx queue) have been removed since
their purpose is not justified, as they don't prevent nor are
suited to prevent device reset/reconfig races (such as this one).
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Resetting the device (stop_gfar()/startup_gfar()) should
be fast and to the point, in order to timely recover
from an error condition (like Tx timeout) or during
device reconfig. The irq free/ request routines are just
redundant here, and they should be part of the device
close/ open routines instead.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The RCTRL and TCTRL registers should not be changed
on-the-fly, while the controller is running, otherwise
unexpected behaviour occurs. But that's exactly what
gfar_vlan_mode() does, updating the VLAN acceleration
bits inside RCTRL/TCTRL. The attempt to lock these
operations doesn't help, but only adds to the confusion.
There's also a dependency for Rx FCB insertion (activating
/de-activating the TOE offload block on Rx) which might
change the required rx buffer size. This makes matters
worse as gfar_vlan_mode() ends up calling gfar_change_mtu(),
though the MTU size remains the same. Note that there are
other situations that may affect the required rx buffer size,
like changing RXCSUM or rx hw timestamping, but errorneously
the rx buffer size is not recomputed/ updated in the process.
To fix this, do the vlan updates properly inside the MAC
reset and reconfiguration procedure, which takes care of
the rx buffer size dependecy and the rx TOE block (PRSDEP)
activation/deactivation as well (in the correct order).
As a consequence, MTU/ rx buff size updates are done now
by the same MAC reset and reconfig procedure, so that out
of context updates to MAXFRM, MRBLR, and MACCFG inside
change_mtu() are no longer needed. The rx buffer size
dependecy to Rx FCB is now handled for the other cases too
(RXCSUM and rx hw timestamping).
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The main MAC config registers like: RCTRL/TCTRL, MRBLR,
MAXFRM, RXIC/TXIC, most fields of MACCFG1/2, should not
be changed on-the-fly, but at least after stopping the
DMA and disabling the Rx/Tx blocks and, for increased
reliability, after a MAC soft reset.
Impelement a complete MAC soft reset and reconfig procedure
following the latest HW advisories - gfar_mac_reset() - to
replace gfar_mac_init() and (the confusing) init_registers()
functions.
Factor out separate config functions for RCTRL and TCTRL,
insure programming order of the relevant config regs after
MAC soft reset.
Split gfar_hw_init() into gfar_mac_reset() and the remaining
global regs that don't need to be reconfigured after MAC soft
reset (FIFOCFG, ATTRELI, HW counters a.s.o).
As gfar_hw_init() now makes all the register writes @probe()
time, based on all the device flags and config options, it
must be moved further down, just before register_netdev(),
as the last config step when the config values are comitted
to HW. Also, move netif_carrier_off() after register_netdev(),
because it has no effect if called before.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
gfar_halt() and gfar_start() are responsible for stopping
and starting the DMA and the Rx/Tx hw rings. They implement
the support for the "graceful Rx/Tx stop/start" hw procedure,
and also disable/enable eTSEC's hw interrupts in the process.
The GRS/GTS procedure requires however to have the RQUEUE/TQUEUE
registers cleared first and to wait for a period of time for the
current frame to pass through the interface (around ~10ms for a
jumbo frame). Only then may the GTS and GRS bits from DMACTRL be
set to shut down the DMA, and finally the Tx_EN and Rx_EN bits in
MACCFG1 may be cleared to disable the Tx/Rx blocks.
The same register programming order applies to start the Rx/Tx:
enabling the RQUEUE/TQUEUE *before* clearing the GRS/GTS bits.
This is a HW recommendation in order to avoid a possible
controller "lock up" during graceful reset.
Cleanup the gfar_halt()/start() prototypes, to take priv instead
of ndev as their purpose is to operate on HW. Enabling the
RQUEUE/TQUEUE in the hw_init() is not needed anymore since
that's the job of gfar_start().
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Throughout the code there are places where the controller's
hw interrupt sources need to get disabled/enabled (masked/
un-masked) all at once. The recommendation for disabling
the interrupts is to clear the ievent first then the imask
register (not the other way around).
Use the gfar_ints_enable/disable() helpers to make these
operations consistent.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The RCTRL updates of the FSL_GIANFAR_DEV_HAS_PADDING device
flag get overriden by the FSL_GIANFAR_DEV_HAS_TIMER flag
settings, which impose a Rx padding alignment of 8 bytes.
As all the eTSEC devices that set HAS_PADDING also set the
HAS_TIMER flag, the HAS_PADDING flag is now obsolete.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Removing the sysfs stubs for the Tx FIFOCFG and ATTRELI
(stashing) config registers, as these registers may only
be configured after a MAC reset, with the controller stopped
(i.e. during hw init, at probe() time). The current sysfs
stubs allow on-the-fly updates of these registers (the locking
measures are useless and only add unecessary code).
Changing these registers is discouraged. Only the default values
will be used instead.
Moreover, the stashing (ATTRELI) configuration options were
effectively disabled (didn't get to the hw anyway if changed)
because the stashing device_flags (HAS_BD_STASHING|HAS_BUF_STASHING)
were "accidentally" cleared during probe().
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Factor out gfar_hw_init() to contain all the controller hw
initialization steps for a better control of register writes,
and to significantly simplify the tangled code from gfar_probe().
This results in code size and stack usage reduction (besides
code readability).
Fix memory leak on device removal, by freeing the rx_/tx_queue
structures.
Replace custom bit swapping function with a library one (bitrev8).
Move allocation of rx_/tx_queue struct arrays before the group
structure init, because in order to assign Rx/Tx queues
to groups we need to have the queues first. This also allows
earlier bail out of gfar_probe(), in case the memory allocation
fails.
The flow control checks for maccfg1 were removed from gfar_probe(),
since flow control is disabled at probe time (priv->rx_/tx_pause_en
are 0). Redundant initializations (by 0) also removed.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
None of these files are actually using any __init type directives
and hence don't need to include <linux/init.h>. Most are just a
left over from __devinit and __cpuinit removal, or simply due to
code getting copied from one driver to the next.
This covers everything under drivers/net except for wireless, which
has been submitted separately.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull networking updates from David Miller:
1) The addition of nftables. No longer will we need protocol aware
firewall filtering modules, it can all live in userspace.
At the core of nftables is a, for lack of a better term, virtual
machine that executes byte codes to inspect packet or metadata
(arriving interface index, etc.) and make verdict decisions.
Besides support for loading packet contents and comparing them, the
interpreter supports lookups in various datastructures as
fundamental operations. For example sets are supports, and
therefore one could create a set of whitelist IP address entries
which have ACCEPT verdicts attached to them, and use the appropriate
byte codes to do such lookups.
Since the interpreted code is composed in userspace, userspace can
do things like optimize things before giving it to the kernel.
Another major improvement is the capability of atomically updating
portions of the ruleset. In the existing netfilter implementation,
one has to update the entire rule set in order to make a change and
this is very expensive.
Userspace tools exist to create nftables rules using existing
netfilter rule sets, but both kernel implementations will need to
co-exist for quite some time as we transition from the old to the
new stuff.
Kudos to Patrick McHardy, Pablo Neira Ayuso, and others who have
worked so hard on this.
2) Daniel Borkmann and Hannes Frederic Sowa made several improvements
to our pseudo-random number generator, mostly used for things like
UDP port randomization and netfitler, amongst other things.
In particular the taus88 generater is updated to taus113, and test
cases are added.
3) Support 64-bit rates in HTB and TBF schedulers, from Eric Dumazet
and Yang Yingliang.
4) Add support for new 577xx tigon3 chips to tg3 driver, from Nithin
Sujir.
5) Fix two fatal flaws in TCP dynamic right sizing, from Eric Dumazet,
Neal Cardwell, and Yuchung Cheng.
6) Allow IP_TOS and IP_TTL to be specified in sendmsg() ancillary
control message data, much like other socket option attributes.
From Francesco Fusco.
7) Allow applications to specify a cap on the rate computed
automatically by the kernel for pacing flows, via a new
SO_MAX_PACING_RATE socket option. From Eric Dumazet.
8) Make the initial autotuned send buffer sizing in TCP more closely
reflect actual needs, from Eric Dumazet.
9) Currently early socket demux only happens for TCP sockets, but we
can do it for connected UDP sockets too. Implementation from Shawn
Bohrer.
10) Refactor inet socket demux with the goal of improving hash demux
performance for listening sockets. With the main goals being able
to use RCU lookups on even request sockets, and eliminating the
listening lock contention. From Eric Dumazet.
11) The bonding layer has many demuxes in it's fast path, and an RCU
conversion was started back in 3.11, several changes here extend the
RCU usage to even more locations. From Ding Tianhong and Wang
Yufen, based upon suggestions by Nikolay Aleksandrov and Veaceslav
Falico.
12) Allow stackability of segmentation offloads to, in particular, allow
segmentation offloading over tunnels. From Eric Dumazet.
13) Significantly improve the handling of secret keys we input into the
various hash functions in the inet hashtables, TCP fast open, as
well as syncookies. From Hannes Frederic Sowa. The key fundamental
operation is "net_get_random_once()" which uses static keys.
Hannes even extended this to ipv4/ipv6 fragmentation handling and
our generic flow dissector.
14) The generic driver layer takes care now to set the driver data to
NULL on device removal, so it's no longer necessary for drivers to
explicitly set it to NULL any more. Many drivers have been cleaned
up in this way, from Jingoo Han.
15) Add a BPF based packet scheduler classifier, from Daniel Borkmann.
16) Improve CRC32 interfaces and generic SKB checksum iterators so that
SCTP's checksumming can more cleanly be handled. Also from Daniel
Borkmann.
17) Add a new PMTU discovery mode, IP_PMTUDISC_INTERFACE, which forces
using the interface MTU value. This helps avoid PMTU attacks,
particularly on DNS servers. From Hannes Frederic Sowa.
18) Use generic XPS for transmit queue steering rather than internal
(re-)implementation in virtio-net. From Jason Wang.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1622 commits)
random32: add test cases for taus113 implementation
random32: upgrade taus88 generator to taus113 from errata paper
random32: move rnd_state to linux/random.h
random32: add prandom_reseed_late() and call when nonblocking pool becomes initialized
random32: add periodic reseeding
random32: fix off-by-one in seeding requirement
PHY: Add RTL8201CP phy_driver to realtek
xtsonic: add missing platform_set_drvdata() in xtsonic_probe()
macmace: add missing platform_set_drvdata() in mace_probe()
ethernet/arc/arc_emac: add missing platform_set_drvdata() in arc_emac_probe()
ipv6: protect for_each_sk_fl_rcu in mem_check with rcu_read_lock_bh
vlan: Implement vlan_dev_get_egress_qos_mask as an inline.
ixgbe: add warning when max_vfs is out of range.
igb: Update link modes display in ethtool
netfilter: push reasm skb through instead of original frag skbs
ip6_output: fragment outgoing reassembled skb properly
MAINTAINERS: mv643xx_eth: take over maintainership from Lennart
net_sched: tbf: support of 64bit rates
ixgbe: deleting dfwd stations out of order can cause null ptr deref
ixgbe: fix build err, num_rx_queues is only available with CONFIG_RPS
...
usual for this cycle with lots of clean-up.
- Cross arch clean-up and consolidation of early DT scanning code.
- Clean-up and removal of arch prom.h headers. Makes arch specific
prom.h optional on all but Sparc.
- Addition of interrupts-extended property for devices connected to
multiple interrupt controllers.
- Refactoring of DT interrupt parsing code in preparation for deferred
probe of interrupts.
- ARM cpu and cpu topology bindings documentation.
- Various DT vendor binding documentation updates.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iQEcBAABAgAGBQJSgPQ4AAoJEMhvYp4jgsXif28H/1WkrXq5+lCFQZF8nbYdE2h0
R8PsfiJJmAl6/wFgQTsRel+ScMk2hiP08uTyqf2RLnB1v87gCF7MKVaLOdONfUDi
huXbcQGWCmZv0tbBIklxJe3+X3FIJch4gnyUvPudD1m8a0R0LxWXH/NhdTSFyB20
PNjhN/IzoN40X1PSAhfB5ndWnoxXBoehV/IVHVDU42vkPVbVTyGAw5qJzHW8CLyN
2oGTOalOO4ffQ7dIkBEQfj0mrgGcODToPdDvUQyyGZjYK2FY2sGrjyquir6SDcNa
Q4gwatHTu0ygXpyphjtQf5tc3ZCejJ/F0s3olOAS1ahKGfe01fehtwPRROQnCK8=
=GCbY
-----END PGP SIGNATURE-----
Merge tag 'devicetree-for-3.13' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux
Pull devicetree updates from Rob Herring:
"DeviceTree updates for 3.13. This is a bit larger pull request than
usual for this cycle with lots of clean-up.
- Cross arch clean-up and consolidation of early DT scanning code.
- Clean-up and removal of arch prom.h headers. Makes arch specific
prom.h optional on all but Sparc.
- Addition of interrupts-extended property for devices connected to
multiple interrupt controllers.
- Refactoring of DT interrupt parsing code in preparation for
deferred probe of interrupts.
- ARM cpu and cpu topology bindings documentation.
- Various DT vendor binding documentation updates"
* tag 'devicetree-for-3.13' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux: (82 commits)
powerpc: add missing explicit OF includes for ppc
dt/irq: add empty of_irq_count for !OF_IRQ
dt: disable self-tests for !OF_IRQ
of: irq: Fix interrupt-map entry matching
MIPS: Netlogic: replace early_init_devtree() call
of: Add Panasonic Corporation vendor prefix
of: Add Chunghwa Picture Tubes Ltd. vendor prefix
of: Add AU Optronics Corporation vendor prefix
of/irq: Fix potential buffer overflow
of/irq: Fix bug in interrupt parsing refactor.
of: set dma_mask to point to coherent_dma_mask
of: add vendor prefix for PHYTEC Messtechnik GmbH
DT: sort vendor-prefixes.txt
of: Add vendor prefix for Cadence
of: Add empty for_each_available_child_of_node() macro definition
arm/versatile: Fix versatile irq specifications.
of/irq: create interrupts-extended property
microblaze/pci: Drop PowerPC-ism from irq parsing
of/irq: Create of_irq_parse_and_map_pci() to consolidate arch code.
of/irq: Use irq_of_parse_and_map()
...
Conflicts:
drivers/net/usb/qmi_wwan.c
include/net/dst.h
Trivial merge conflicts, both were overlapping changes.
Signed-off-by: David S. Miller <davem@davemloft.net>
Under certain low traffic conditions, the single core
devices with multiple Rx/Tx queues (MQ mode) may reach
soft lockup due to gfar_poll not returning in proper time.
The following exception was obtained using iperf on a 100Mbit
half-duplex link, for a p1010 single core device:
BUG: soft lockup - CPU#0 stuck for 23s! [iperf:2847]
Modules linked in:
CPU: 0 PID: 2847 Comm: iperf Not tainted 3.12.0-rc3 #16
task: e8bf8000 ti: eeb16000 task.ti: ee646000
NIP: c0255b6c LR: c0367ae8 CTR: c0461c18
REGS: eeb17e70 TRAP: 0901 Not tainted (3.12.0-rc3)
MSR: 00029000 <CE,EE,ME> CR: 44228428 XER: 20000000
GPR00: c0367ad4 eeb17f20 e8bf8000 ee01f4b4 00000008 ffffffff ffffffff
00000000
GPR08: 000000c0 00000008 000000ff ffffffc0 000193fe
NIP [c0255b6c] find_next_bit+0xb8/0xc4
LR [c0367ae8] gfar_poll+0xc8/0x1d8
Call Trace:
[eeb17f20] [c0367ad4] gfar_poll+0xb4/0x1d8 (unreliable)
[eeb17f70] [c0422100] net_rx_action+0xa4/0x158
[eeb17fa0] [c003ec6c] __do_softirq+0xcc/0x17c
[eeb17ff0] [c000c28c] call_do_softirq+0x24/0x3c
[ee647cc0] [c0004660] do_softirq+0x6c/0x94
[ee647ce0] [c003eb9c] local_bh_enable+0x9c/0xa0
[ee647cf0] [c0454fe8] tcp_prequeue_process+0xa4/0xdc
[ee647d10] [c0457e44] tcp_recvmsg+0x498/0x96c
[ee647d80] [c047b630] inet_recvmsg+0x40/0x64
[ee647da0] [c040ca8c] sock_recvmsg+0x90/0xc0
[ee647e30] [c040edb8] SyS_recvfrom+0x98/0xfc
To prevent this, the outer while() loop has been removed
allowing gfar_poll() to return faster even if there's
still budget left. Also, there's no need to recompute
the budget per Rx queue anymore.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Powerpc is a mess of implicit includes by prom.h. Add the necessary
explicit includes to drivers in preparation of prom.h cleanup.
Signed-off-by: Rob Herring <rob.herring@calxeda.com>
Acked-by: Grant Likely <grant.likely@linaro.org>
Enable workaround for P2020/P2010 erratum eTSEC 20,
"Excess delays when transmitting TOE=1 large frames".
The impact is that frames lager than 2500-bytes for which
TOE (i.e. TCP/IP hw accelerations like Tx csum) is enabled
may see excess delay before start of transmission.
This erratum was fixed in Rev 2.0.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use the macros and defines from mpc85xx.h to simplify
and prevent errors in identifying a mpc85xx based SoC
for errata detection.
This should help enabling (and identifying) workarounds
for various mpc85xx based chips and revisions.
For instance, express MPC8548 Rev.2 as:
(SVR_SOC_VER(svr) == SVR_8548) && (SVR_REV(svr) == 0x20)
instead of:
(pvr == 0x80210020 && mod == 0x8030 && rev == 0x0020)
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A002 is still in "no plans to fix" state, and applies to all
the current P1/P2 parts as well, so it's resonable to enable
its workaround by default, for all the soc's with etsec.
The impact of not enabling this workaround for affected parts
is that under certain conditons (runt frames or even frames
with RX error detected at PHY level) during controller reset,
the controller might fail to indicate Rx reset (GRS) completion.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix the amount of sent bytes reported to BQL by reporting the
number of bytes on wire in the xmit routine, and recording that
value for each skb in order to be correctly confirmed on Tx
confirmation cleanup.
Reporting skb->len to BQL just before exiting xmit is not correct
due to possible insertions of TOE block and alignment bytes in the
skb->data, which are being stripped off by the controller before
transmission on wire. This led to mismatch of (incorrectly)
reported bytes to BQL b/w xmit and Tx confirmation, resulting in
Tx timeout firing, for the h/w tx timestamping acceleration case.
There's no easy way to obtain the number of bytes on wire in the Tx
confirmation routine, so skb->cb is used to convey that information
from xmit to Tx confirmation, for now (as proposed by Eric). Revived
the currently unused GFAR_CB() construct for that purpose.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
eTSEC has Rx and Tx flow control capabilities that may be enabled
through MACCFG1[Rx_Flow, Tx_Flow] bits. These bits must not be set
however when eTSEC is operated in Half-Duplex mode. Unfortunately,
the driver currently sets these bits unconditionally.
This patch adds the proper handling of the PAUSE frame capability
register bits by implementing the ethtool -A interface. When pause
autoneg is enabled, the controller uses the phy's capability to
negotiate PAUSE frame settings with the link partner and reconfigures
its Rx_Flow and Tx_Flow settings to match the capabilities of the
link partner. If pause autoneg is off, the PAUSE frame generation
may be forced manually (ethtool -A). Flow control is disabled by
default now.
This implementation is inspired by the tg3 driver.
Signed-off-by: Lutz Jaenicke <ljaenicke@innominate.com>
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cleanup gfar_start_xmit()'s fast path by factoring out "redundant"
FCB insertion code (repeated gfar_add_fcb() calls and related)
and by reducing the number of if() clauses (i.e. if(fcb) checks).
Improve maintainability (e.g. there's less code and easier to read)
also by introducing do_csum and do_vlan to mark the other 2 Tx TOE
functionalities, following the same model as do_tstamp.
fcb_len may also be 0 now, to mark that Tx FCB insertion conditions
(do_csum, do_vlan, do_tstamp) have not been met.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Both [eTSEC76] and [eTSEC12] errata relate to Tx checksum generation
(for some MPC83xx and MCP8548 older revisions). They require the same
workaround: manual checksum computation and insertion, and disabling
the H/W Tx csum acceleration feature (per frame) through Tx FCB
(Frame Control Block) csum offload settings.
The workaround for [eTSEC76] needs to be fixed because it currently
fails to disable H/W Tx csum insertion via FCB. This patch fixes it
and provides a common workaround implementation for both Tx csum errata.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
grp->grp_id is obsolete. It has no use in the current driver.
Remove it from gfar_priv_grp and put the 'rstat' member
in its place, in the 2nd cache line, as rstat needs fast access.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Older Single Queue (SQ_SG_MODE) devices like TSEC (i.e. mpc83xx)
don't feature the frame receive indication bits (RXF) in RSTAT.
For these and for the rest of the SQ_SG_MODE devices, provide the
appropiate polling routine that handles a single pair of Rx/Tx
BD rings, removing the overhead incurred by the multiple queues/
multiple interrupt group devices (veTSEC/ eTSEC2.0 devices).
So this is primarily a fix for the TSEC devices.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use the wrapper functions for getting and setting the driver data using
platform_device instead of using dev_{get,set}_drvdata() with &pdev->dev,
so we can directly pass a struct platform_device.
Also, unnecessary dev_set_drvdata() is removed, because the driver core
clears the driver data to NULL after device_release or on probe failure.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Several call sites were missed when the protocol argument was added to
__vlan_hwaccel_put_tag() in commit
86a9bad3ab ("net: vlan: add protocol
argument to packet tagging functions").
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Rename the hardware VLAN acceleration features to include "CTAG" to indicate
that they only support CTAGs. Follow up patches will introduce 802.1ad
server provider tagging (STAGs) and require the distinction for hardware not
supporting acclerating both.
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The GRO_DROP return code is handled by the core network layer.
The current kernel approach is to factorize this kind of statistics into
the upper layers, instead of having all the drivers maintaining them.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Warning message:
warning: 'budget_per_q' may be used uninitialized in this function
budget_per_q won't be used uninitialized since the only time
it doesn't get initialized is when entering gfar_poll with
num_act_queues == 0, meaning rstat_rxf == 0, in which case
budget_per_q is not utilized (as it has no meaning).
Inititalize budget_per_q to 0 though to suppress this compile
warning.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The only place where gfar_configure_coalescing is called
with an actual bitmask (other than 0xff) is in gfar_poll
(on the hot path). So make gfar_configure_coalescing()
static for the buffer processing path, and export
gfar_configure_coalescing_all() for the remaining cases
that require to set coalescing for all the queues at once
(on the slow path).
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For Multi Q Multi Group (MQ_MG_MODE) mode, the Rx/Tx colescing registers [rt]xic
are aliased with the [rt]xic0 registers (coalescing setting regs for Q0). This
avoids programming twice in a row the coalescing registers for the Rx/Tx hw Q0.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Split the napi budget fairly among the active queues only, instead
of dividing it by the total number of Rx queues assigned to the
given interrupt group.
Use the h/w indication field RXFi in rstat (receive status register)
to identify the active rx queues from the current interrupt group
(i.e. receive event occured on ring i, if ring i is part of the current
interrupt group). This indication field in rstat, RXFi i=0..7,
allows us to find out on which queues of the same interrupt group
do we have incomming traffic once we entered the polling routine for
the given interrupt group. After servicing the ring i, the corresponding
bit RXFi will be written with 1 to clear the active queue indication for
that ring.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There are 2 issues with the current napi poll routine, with regards
to tx ring cleanup:
1) for multi-queue devices (MQ_MG_MODE), should tx_bit_map != rx_bit_map,
which is possible (and supported in h/w) if the DT property "fsl,tx-bit-map"
holds a different value than rx_bit_map, the current polling routine will
service the wrong Tx queues in this case (i.e. the interrupt group will
receive interrupts from tx queues that it will not service)
2) Tx cleanup completion consumes napi budget, whereas the napi budget
should be reserved for Rx work only.
The patch fixes these issues and provides a clean napi polling routine.
Napi poll completion is reached when all the Rx queues have been
serviced and there is no Tx work to do.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I believe these error messages are already logged
on allocation failure by warn_alloc_failed and so
get a dump_stack on OOM.
Remove the unnecessary additional error logging.
Around these deletions:
o Alignment neatening.
o Remove unnecessary casts of dma_alloc_coherent.
o Hoist assigns from ifs.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit ee873fda3b ("gianfar: Pack struct
gfar_priv_grp into three cachelines") moved the irq number and names
off into a separate struct and created accessors for them. However
it was never tested with NET_POLL enabled, and so some conversions
that were simply overlooked went undetected until now.
Make the netpoll ones also use the gfar_irq() accessors.
Reported-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Claudiu Manoil <claudiu.manoil@freescale.com>
Cc: Jianhua Xie <jianhua.xie@freescale.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This fixes a less obvious error on one hand, and prevents futher
similar errors by disambiguating and optimizing RxFCB indication,
on the other hand.
The error consists in NETIF_F_HW_VLAN_TX flag being used as an
indication of Rx FCB insertion. This happened as soon gfar_uses_fcb(),
which despite its name indicates Rx FCB insertion, started
incorporating is_vlan_on().
is_vlan_on(), on the other hand, is also a misleading construct because
we need to differentiate b/w hw VLAN extraction/VLEX (marked by VLAN_RX
flag) and hw VLAN insertion/VLINS (VLAN_TX flag), which are different
mechanisms using different types of FCBs.
The hw spec for the RxFCB feature is as follows:
In the case of RxBD rings, FCBs (Frame Control Block) are inserted by
the eTSEC whenever RCTRL[PRSDEP] is set to a non-zero value. Only one
FCB is inserted per frame (in the buffer pointed to by the RxBD with
bit F set). TOE acceleration for receive is enabled for all rx frames
in this case.
This patch introduces priv->uses_rxfcb field to quickly signal RxFCB
insertion in accordance with the specification above.
The dependency on FSL_GIANFAR_DEV_HAS_TIMER was also eliminated as
another source of confusion. The actual dependency is to priv->hwts_rx_en.
Upon changing priv->hwts_rx_en via IOCTL, the gfar device is being
restarted and on init_mac() the priv->hwts_rx_en flag determines RxFCB
insertion, and rctrl is programmed accordingly. The patch takes care
of this case too.
Though maybe not as self documenting as the inlining version uses_fcb(),
priv->uses_rxfcb has the main purpose to quickly signal, on the hot path,
that the incoming frame has a *Rx* FCB block inserted which needs to be
pulled out before passing the skb to the stack. This is a performance
critical operation, it needs to happen fast, that's why uses_rxfcb is
placed in the first cacheline of gfar_private.
This is also why a cached rctrl cannot be used instead: 1) because
we don't have 32 bits available in the first cacheline of gfar_priv
(but only 16); 2) bit operations are expensive on the hot path.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The controller's ref manual states clearly that when the hw Rx vlan
offload feature is enabled, meaning that the VLEX bit from RCTRL is
correctly enabled, then the hw performs automatic VLAN tag extraction
and deletion from the ethernet frames. So there's no point in trying to
increase the rx buff size when rxvlan is on, as the frame is actually
smaller.
And the Tx vlan hw accel feature (VLINS) has nothing to do with rx buff
size computation.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
No return code is expected from gfar_process_frame(), hence
change it to return void.
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Use device pointer (dev) to simplify the code and to
avoid double indirections, especially on the hot path.
Basically, instead of accessing priv to get the ofdev
reference and then accessing the ofdev structure to
dereference the needed dev pointer, we will get the
dev pointer directly from priv.
The dev pointer is required on the hot path, see gfar_new_rxbdp
or gfar_clean_rx_ring (or xmit), and this patch makes
it available directly from priv's 1st cacheline.
This change is reflected at asm level too, taking (the hot)
gfar_new_rxbdp():
initial version -
18c0: 7c 7e 1b 78 mr r30,r3
18d0: 81 69 04 3c lwz r11,1084(r9)
18d8: 34 6b 00 10 addic. r3,r11,16
18dc: 41 82 00 08 beq- 18e4
patched version -
18d0: 80 69 04 38 lwz r3,1080(r9)
18d8: 2f 83 00 00 cmpwi cr7,r3,0
18dc: 41 9e 00 08 beq- cr7,18e4
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While looking at some asm dump for an unrelated change, Eric
noticed in the following stats count increment code:
50b8: 81 3c 01 f8 lwz r9,504(r28)
50bc: 81 5c 01 fc lwz r10,508(r28)
50c0: 31 4a 00 01 addic r10,r10,1
50c4: 7d 29 01 94 addze r9,r9
50c8: 91 3c 01 f8 stw r9,504(r28)
50cc: 91 5c 01 fc stw r10,508(r28)
that a 64 bit counter was used on ppc-32 without sync
and hence the "ethtool -S" output was racy.
Here we convert all the values to use atomic64_t so that
the output will always be consistent.
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
alloc failures already get standardized OOM
messages and a dump_stack.
For the affected mallocs around these OOM messages:
Converted kmallocs with multiplies to kmalloc_array.
Converted a kmalloc/memcpy to kmemdup.
Removed now unused stack variables.
Removed unnecessary parentheses.
Neatened alignment.
Signed-off-by: Joe Perches <joe@perches.com>
Acked-by: Arend van Spriel <arend@broadcom.com>
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de>
Acked-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit ee873fda3b
"gianfar: Pack struct gfar_priv_grp into three cachelines"
causes the following null dereference at driver init on sbc8548:
libphy: Freescale PowerQUICC MII Bus: probed
Unable to handle kernel paging request for data at address 0x00000000
Faulting instruction address: 0xc01d6a38
Oops: Kernel access of bad area, sig: 11 [#1]
[...]
NIP [c01d6a38] gfar_parse_group+0x228/0x280
LR [c01d6a34] gfar_parse_group+0x224/0x280
Call Trace:
[ef82dd60] [c01d6a34] gfar_parse_group+0x224/0x280 (unreliable)
[ef82dd90] [c01d73a4] gfar_probe+0x284/0xfe0
The reason is that the commit also changed the allocation of the
Rx and error handling irq structs to be skipped for !MQ_MG_MODE.
In the !MQ_MG_MODE case, only the Tx irq struct is allocated.
Digging further, we see that MQ_MG_MODE is set only if we find
the OF compatible string "fsl,etsec2".
A quick grep in the dts directory shows lots of boards that support
Rx/Tx/Err, but without this specific compat string. And hence they
go after the unallocated Rx/Error structs and cause the above oops.
Hence such a change can not be deployed until all the dts files
are updated and sufficiently deployed. Further, the optimization
is of limited value, since the kmalloc'd struct in question has only
a single unsigned int, and an (IFNAMSIZ + 6) sized string.
Note that no changes to the freeing code are needed here, as it
already did an unconditional free of Rx/Tx/Error gfar_irqinfo.
Cc: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
