Create devices for each USB4 port. This is needed when we add retimer
access when there is no device connected but may be useful for other
purposes too following what USB subsystem does. This exports a single
attribute "link" that shows the type of the USB4 link (or "none" if
there is no cable connected).
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The upstream port can be connected to any previous generation
Thunderbolt port so logging as "TBT" is more accurate than "TBT3.
No functional changes.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Intel Goshen Ridge reports wrong DP main credits in NVM 27 and earlier,
so add a quirk that fixes it. We also need to expand the quirk table to
match on hardware vendor/device IDs too.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The USB4 Connection Manager guide provides detailed information how the
USB4 router buffer (credit) allocation information should be used by the
connection manager when it allocates buffers for different paths. This
patch implements it for Linux. For USB 3.x and DisplayPort we use
directly the router preferences. The rest of the buffer space is then
used for PCIe and DMA (peer-to-peer, XDomain) traffic. DMA tunnels
require at least one buffer and PCIe six, so if there is not enough
buffers we fail the tunnel creation.
For the legacy Thunderbolt 1-3 devices we use the existing hard-coded
scheme except for DMA where we use the values suggested by the USB4 spec
chapter 13.
Co-developed-by: Gil Fine <gil.fine@intel.com>
Signed-off-by: Gil Fine <gil.fine@intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Once lane bonding has been enabled (or disabled) both lane adapters may
update their total credits accordingly. For this reason re-read the port
credits after lane bonding has been enabled or disabled.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
USB4 routers must expose their preferred credit (buffer) allocation
information through router operation. This information tells the
connection manager how the router prefers its buffers to be allocated to
get the expected bandwidth for the supported protocols.
Read this information and store it as part of struct tb_switch for each
USB4 router.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
It may take some time until the two lanes enter bonded state so poll for
the link width to match what is expected before going forward. This ensures
the link is in expected state before we start establishing paths through
it.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
With the USB4 buffer allocation the number of credits (and non-flow
credits) may be different depending on the router buffer allocation
preferences. To allow this move the nfc_credits field to struct
tb_path_hop.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
The function does not modify the object in any way so make the parameter
const to reflect this.
No functional changes intended.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Configure qspi's mdma from buffer transfer (max 128 bytes) to
block transfer (max 64K bytes).
mtd_speedtest shows that write throughtput increases :
- from 734 to 782 KiB/s (~6.5%) with s25fl512s SPI-NOR.
- from 4848 to 5319 KiB/s (~9.72%) with Micron SPI-NAND.
Signed-off-by: Christophe Kerello <christophe.kerello@foss.st.com>
Signed-off-by: Patrice Chotard <patrice.chotard@foss.st.com>
Signed-off-by: Alexandre Torgue <alexandre.torgue@foss.st.com>
Fix make dtbs_check warning:
arch/arm/boot/dts/stm32mp157c-dhcom-pdk2.dt.yaml:0:0: /soc/i2c@40015000/polytouch@38: failed to match any schema with compatible: ['edt,edt-ft5x06']
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: linux-stm32@st-md-mailman.stormreply.com
To: linux-arm-kernel@lists.infradead.org
Signed-off-by: Alexandre Torgue <alexandre.torgue@foss.st.com>
Fix make dtbs_check warning:
arch/arm/boot/dts/stm32mp157c-dhcom-pdk2.dt.yaml: gpio-keys-polled: '#address-cells' is a dependency of '#size-cells'
arch/arm/boot/dts/stm32mp157c-dhcom-pdk2.dt.yaml: gpio-keys: '#address-cells' is a dependency of '#size-cells'
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Alexandre Torgue <alexandre.torgue@st.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: linux-stm32@st-md-mailman.stormreply.com
To: linux-arm-kernel@lists.infradead.org
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Alexandre Torgue <alexandre.torgue@foss.st.com>
For the HP Pavilion 15-CK0xx, with audio subsystem ID 0x103c:0x841c,
adding a line in patch_realtek.c to apply the ALC269_FIXUP_HP_MUTE_LED_MIC3
fix activates the mute key LED.
Signed-off-by: Carlos M <carlos.marr.pz@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20210531202026.35427-1-carlos.marr.pz@gmail.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In commit 68dc022d04 ("xfrm: BEET mode doesn't support fragments
for inner packets"), it tried to fix the issue that in TX side the
packet is fragmented before the ESP encapping while in the RX side
the fragments always get reassembled before decapping with ESP.
This is not true for IPv6. IPv6 is different, and it's using exthdr
to save fragment info, as well as the ESP info. Exthdrs are added
in TX and processed in RX both in order. So in the above case, the
ESP decapping will be done earlier than the fragment reassembling
in TX side.
Here just remove the fragment check for the IPv6 inner packets to
recover the fragments support for BEET mode.
Fixes: 68dc022d04 ("xfrm: BEET mode doesn't support fragments for inner packets")
Reported-by: Xiumei Mu <xmu@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Previously this fix was applied only to Bullseye variant laptops,
and should be applied to Cyborg and Warlock variants.
Fixes: 45b14fe200 ("ALSA: hda/cirrus: Use CS8409 filter to fix abnormal sounds on Bullseye")
Signed-off-by: Stefan Binding <sbinding@opensource.cirrus.com>
Signed-off-by: Vitaly Rodionov <vitalyr@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20210531163754.136736-1-vitalyr@opensource.cirrus.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit takes ALSA fireface driver to perform sequence replay for
media clock recovery.
The protocol specific to RME Fireface series is not compliant to
IEC 61883-1/6 since it has no CIP header, therefore presentation time
is not used for media clock recovery. The sequence of the number of data
blocks per packet is important.
I note that the device skips an isochronous cycle corresponding to an
empty packet or a NODATA packet in blocking transmission method of
IEC 61883-1/6. For sequence replay, the cycle is handled as receiving an
empty packet. Furthermore, it doesn't start packet transmission till
receiving any packet.
The sequence replay is tested with below models:
* Fireface 400
* Fireface 800
* Fireface 802
I note that it is better to initialize Fireface 400 in advance by
initialization transaction implemented in snd-fireface-ctl-service of
snd-firewire-ctl-services project. You can see whether initialized or
not by HOST LED on the device. Unless, the device often stops packet
transmission even if session starts.
I guess the sequence replay also works well with below models:
* Fireface UFX
* Fireface UCX
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-7-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit takes ALSA firewire-tascam driver to perform sequence replay
for media clock recovery.
The protocol specific to Tascam FireWire series is not compliant to
IEC 61883-1/6 in terms of syt field of CIP. The protocol doesn't use
presentation time in received CIP for playback timing. The sequence of
the number of data blocks per packet is important for media clock
recovery.
Although the devices in Tascam FireWire series transfer packets
regardless of receiving packets, the tx packets includes no events
in the beginning of streaming. It takes so long to multiplex any event
into the packet after receiving the sequence of packets. As long as I
experienced, it takes several thousands of isochronous cycle. Furthermore,
just after changing sampling transmission frequency, it stops multiplexing
event at once, then starts multiplexing again.
The sequence replay is tested with below models:
* FW-1884
* FW-1804
* FW-1082
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-6-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit takes ALSA firewire-digi00x driver to perform sequence replay
for media clock recovery.
All of models in Digidesign digi00x family don't transfer isochronous
packets till receiving isochronous packets. The on-the-fly mode is used
for the purpose. They don't interpret presentation time expressed in syt
field of received CIP, therefore the sequence of the number of data blocks
per packet is important for media clock recovery.
The sequence replay is tested with below models:
* Digidesign Digi 002
* Digidesign Digi 002 Rack
* Digidesign Digi 003
* Digidesign Digi 003 Rack
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-5-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit takes ALSA oxfw driver to perform sequence replay for media
clock recovery. Unfortunately, OXFW970 ASIC and its firmware has a quirk
called jumbo payload which skips several isochronous cycles for packet
transmission, thus the sequence replay is just adopted to OXFW971 ASIC.
As well as Fireworks, OXFW ASICs also ignores presentation time against
the way in IEC 61883-1/6.
The sequence replay is tested with below models:
* Tascam FireOne
* Stanton Magnetics SCS.1m
* Apogee Duet FireWire
For below models, the sequence replay is tested to be disabled:
* Griffin FireWave
* Behringer F-Control Audio 202
* Loud Technology Tapco Link.FireWire 4x6
* Loud Technology Mackie Onyx Satellite
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-4-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Echo Digital Audio Corporation had US patent US7599388B2 titled as
'System and method for high-bandwidth serial bus data transfer'. In the
patent, dual-banked shared memory is used to deliver data between
serial bus transmission and processor in FIFO way. The patent seems to be
used for Fireworks board module. The mechanism is not compliant to
synchronization based on presentation time expressed in syt field
of CIP header. Fireworks board module takes care of the sequence of
the number of data blocks per packet and just ignores the value of syt
field.
This commit takes fireworks driver to performs sequence replay for media
clock recovery. As long as I tested, Audiofire 2 and 4 have a quirk to
skip an isochronous cycle several thousands after starting packet
transmission.
The sequence replay is tested with below models:
* Loud Technology Mackie 400f
* Echo Audio Audiofire 12 (DSP model)
* Echo Audio Audiofire 12 (FPGA model)
* Echo Audio Audiofire 8 (DSP model)
* Echo Audio Audiofire 8 (FPGA model)
* Echo Audio Audiofire Pre8
* Echo Audio Audiofire 4
* Echo Audio Audiofire 2
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-3-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In the design of Fireworks board module, the device does't adjust its
media clock voluntarily by the sequence of presentation time expressed in
syt field of CIP header of received packet.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Link: https://lore.kernel.org/r/20210531025103.17880-2-o-takashi@sakamocchi.jp
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Add debugfs support for vlan configuraion. create a single file
"vlan_config" for it, and query it by command "cat vlan_config",
return the result to userspace.
The new display style is below:
$ cat vlan_config
I_PORT_VLAN_FILTER: on
E_PORT_VLAN_FILTER: off
FUNC_ID I_VF_VLAN_FILTER E_VF_VLAN_FILTER PORT_VLAN_FILTER_BYPASS
pf off on off
vf0 off on off
FUNC_ID PVID ACCEPT_TAG1 ACCEPT_TAG2 ACCEPT_UNTAG1 ACCEPT_UNTAG2
pf 0 on on on on
vf0 0 on on on on
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previously, there is hardware limitation for VF to modify
the VLAN filter state, and the VLAN filter state is default
enabled. Now the limitation has been removed in some device,
so add capability flag to check whether the device supports
modify VLAN filter state. If flag on, user will be able to
modify the VLAN filter state by ethtool -K.
VF needs to send mailbox to request the PF to modify the VLAN
filter state for it.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
There are some features of VF depend on PF, so it's necessary
for VF to know whether PF supports. For compatibility, modify
the mailbox HCLGE_MBX_GET_TCINFO, extend its function, use to
get the basic information of PF, including mailbox api version
and PF capabilities.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Previously, with hardware limitation, the port VLAN filter are
effective for both PF and its VFs simultaneously, so a single
function is not able to enable/disable separately, and the VLAN
filter state is default enabled. Now some device supports each
function to bypass port VLAN filter, then each function can
switch VLAN filter separately. Add capability flag to check
whether the device supports modify VLAN filter state. If flag
on, user will be able to modify the VLAN filter state by ethtool
-K.
Furtherly, the default VLAN filter state is also changed
according to whether non-zero VLAN used. Then the device can
receive packet with any VLAN tag if only VLAN 0 used.
The function hclge_need_enable_vport_vlan_filter() is used to
help implement above changes. And the VLAN filter handle for
promisc mode can also be simplified by this function.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The struct hclge_vf_vlan_cfg is firstly designed for setting
VLAN filter tag. And it's reused for enable RX VLAN offload
later. It's strange to use member "is_kill" to indicate "enable".
So redefine the struct hclge_vf_vlan_cfg to adapt it.
For there are already 3 subcodes being used in function
hclge_set_vf_vlan_cfg(), use "switch-case" style for each
branch, rather than "if-else". Also simplify the assignment for
each branch to make it more clearly.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For the PF have called hclge_update_port_base_vlan_cfg() already
before notify VF, it's unnecessary to update port based VLAN again
when received mailbox request from VF.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use struct "hclge_vlan_info" instead of separately parameters
for function hclge_push_vf_port_base_vlan_info(), to make it
more concise.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, option "qos" is igored by HNS3 driver for command
"ip link set ethx vf <vf id> vlan <vlan id> qos <qos value>".
Add support for it.
Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The driver is actually OF independent and doesn't need anything from OF (*).
Replace OF headers with mod_devicetable.h and property.h.
*) use of of_match_ptr() is actually wrong here, since it may provoke
a compiler to warn about unused variable.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20210528151113.85943-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Validation script is not happy:
resistive-adc-touch.c:53: warning: Function parameter or member 'x_plate_ohms' not described in 'grts_state'
resistive-adc-touch.c:53: warning: Function parameter or member 'ch' not described in 'grts_state'
Describe parameters in kernel doc to make it happy.
Fixes: fb082cd59a ("Input: resistive-adc-touch - add support for z1 and z2 channels")
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20210528151113.85943-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
xfrm_policy_lookup_bytype loops on seqcount mutex xfrm_policy_hash_generation
within an RCU read side critical section. Although ill advised, this is fine if
the loop is bounded.
xfrm_policy_hash_generation wraps mutex hash_resize_mutex, which is used to
serialize writers (xfrm_hash_resize, xfrm_hash_rebuild). This is fine too.
On PREEMPT_RT=y, the read_seqcount_begin call within xfrm_policy_lookup_bytype
emits a mutex lock/unlock for hash_resize_mutex. Mutex locking is fine, since
RCU read side critical sections are allowed to sleep with PREEMPT_RT.
xfrm_hash_resize can, however, block on synchronize_rcu while holding
hash_resize_mutex.
This leads to the following situation on PREEMPT_RT, where the writer is
blocked on RCU grace period expiry, while the reader is blocked on a lock held
by the writer:
Thead 1 (xfrm_hash_resize) Thread 2 (xfrm_policy_lookup_bytype)
rcu_read_lock();
mutex_lock(&hash_resize_mutex);
read_seqcount_begin(&xfrm_policy_hash_generation);
mutex_lock(&hash_resize_mutex); // block
xfrm_bydst_resize();
synchronize_rcu(); // block
<RCU stalls in xfrm_policy_lookup_bytype>
Move the read_seqcount_begin call outside of the RCU read side critical section,
and do an rcu_read_unlock/retry if we got stale data within the critical section.
On non-PREEMPT_RT, this shortens the time spent within RCU read side critical
section in case the seqcount needs a retry, and avoids unbounded looping.
Fixes: 77cc278f7b ("xfrm: policy: Use sequence counters with associated lock")
Signed-off-by: Varad Gautam <varad.gautam@suse.com>
Cc: linux-rt-users <linux-rt-users@vger.kernel.org>
Cc: netdev@vger.kernel.org
Cc: stable@vger.kernel.org # v4.9
Cc: Steffen Klassert <steffen.klassert@secunet.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Florian Westphal <fw@strlen.de>
Cc: "Ahmed S. Darwish" <a.darwish@linutronix.de>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Acked-by: Ahmed S. Darwish <a.darwish@linutronix.de>
It's possible to trigger NULL pointer dereference by local unprivileged
user, when calling getsockname() after failed bind() (e.g. the bind
fails because LLCP_SAP_MAX used as SAP):
BUG: kernel NULL pointer dereference, address: 0000000000000000
CPU: 1 PID: 426 Comm: llcp_sock_getna Not tainted 5.13.0-rc2-next-20210521+ #9
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-1 04/01/2014
Call Trace:
llcp_sock_getname+0xb1/0xe0
__sys_getpeername+0x95/0xc0
? lockdep_hardirqs_on_prepare+0xd5/0x180
? syscall_enter_from_user_mode+0x1c/0x40
__x64_sys_getpeername+0x11/0x20
do_syscall_64+0x36/0x70
entry_SYSCALL_64_after_hwframe+0x44/0xae
This can be reproduced with Syzkaller C repro (bind followed by
getpeername):
https://syzkaller.appspot.com/x/repro.c?x=14def446e00000
Cc: <stable@vger.kernel.org>
Fixes: d646960f79 ("NFC: Initial LLCP support")
Reported-by: syzbot+80fb126e7f7d8b1a5914@syzkaller.appspotmail.com
Reported-by: butt3rflyh4ck <butterflyhuangxx@gmail.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20210531072138.5219-1-krzysztof.kozlowski@canonical.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Vladimir Oltean says:
====================
Part 2 of SJA1105 DSA driver preparation for new switch introduction (SJA1110)
This series is a continuation of:
https://patchwork.kernel.org/project/netdevbpf/cover/20210524131421.1030789-1-olteanv@gmail.com/
even though it isn't the first time these patches are submitted (they
were part of the group previously called "Add NXP SJA1110 support to the
sja1105 DSA driver"):
https://patchwork.kernel.org/project/netdevbpf/cover/20210526135535.2515123-1-vladimir.oltean@nxp.com/
but I broke that up again since these patches are already reviewed, for
the most part. There are no changes compared to v2 and v1.
This series of patches contains:
- an adaptation of the driver to the new "ethernet-ports" OF node name
- an adaptation of the driver to support more than 1 SGMII port
- a generalization of the supported phy_interface_t values per port
- an adaptation to encode SPEED_10, SPEED_100, SPEED_1000 into the
hardware registers differently depending on switch revision
- a consolidation of the PHY interface type used for RGMII and another
one for the API exposed for sja1105_dynamic_config_read()
====================
Link: https://lore.kernel.org/r/20210530225939.772553-1-olteanv@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The SJA1105 has a static configuration comprised of a number of tables
with entries. Some of these can be read and modified at runtime as well,
through the dynamic configuration interface.
As a careful reader can notice from the comments in this file, the
software interface for accessing a table entry through the dynamic
reconfiguration is a bit of a no man's land, and varies wildly across
switch generations and even from one kind of table to another.
I have tried my best to come up with a software representation of a
'common denominator' SPI command to access a table entry through the
dynamic configuration interface:
struct sja1105_dyn_cmd {
bool search;
u64 valid; /* must be set to 1 */
u64 rdwrset; /* 0 to read, 1 to write */
u64 errors;
u64 valident; /* 0 if entry is invalid, 1 if valid */
u64 index;
};
Relevant to this patch is the VALIDENT bit, which for READ commands is
populated by the switch and lets us know if we're looking at junk or at
a real table entry.
In SJA1105, the dynamic reconfiguration interface for management routes
has notably not implemented the VALIDENT bit, leading to a workaround to
ignore this field in sja1105_dynamic_config_read(), as it will be set to
zero, but the data is valid nonetheless.
In SJA1110, this pattern has sadly been abused to death, and while there
are many more tables which can be read back over the dynamic config
interface compared to SJA1105, their handling isn't in any way more
uniform. Generally speaking, if there is a single possible entry in a
given table, and loading that table in the static config is mandatory as
per the documentation, then the VALIDENT bit is deemed as redundant and
more than likely not implemented.
So it is time to make the workaround more official, and add a bit to the
flags implemented by dynamic config tables. It will be used by more
tables when SJA1110 support arrives.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In SJA1105, the xMII Mode Parameters Table field called PHY_MAC denotes
the 'role' of the port, be it a PHY or a MAC. This makes a difference in
the MII and RMII protocols, but RGMII is symmetric, so either PHY or MAC
settings result in the same hardware behavior.
The SJA1110 is different, and the RGMII ports only work when configured
in MAC mode, so keep the port roles in MAC mode unconditionally.
Why we had an RGMII port in the PHY role in the first place was because
we wanted to have a way in the driver to denote whether RGMII delays
should be applied based on the phy-mode property or not. This is already
done in sja1105_parse_rgmii_delays() based on an intermediary
struct sja1105_dt_port (which contains the port role). So it is a
logical fallacy to use the hardware configuration as a scratchpad for
driver data, it isn't necessary.
We can also remove the gating condition for applying RGMII delays only
for ports in the PHY role. The .setup_rgmii_delay() method looks at
the priv->rgmii_rx_delay[port] and priv->rgmii_tx_delay[port] properties
which are already populated properly (in the case of a port in the MAC
role they are false). Removing this condition generates a few more SPI
writes for these ports (clearing the RGMII delays) which are perhaps
useless for SJA1105P/Q/R/S, where we know that the delays are disabled
by default. But for SJA1110, the firmware on the embedded microcontroller
might have done something funny, so it's always a good idea to clear the
RGMII delays if that's what Linux expects.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In order to support the new speed of 2500Mbps, the SJA1110 has achieved
the great performance of changing the encoding in the MAC Configuration
Table for the port speeds of 10, 100, 1000 compared to SJA1105.
Because this is a common driver, we need a layer of indirection in order
to program the hardware with the right values irrespective of switch
generation.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On the SJA1105, all ports support the parallel "xMII" protocols (MII,
RMII, RGMII) except for port 4 on SJA1105R/S which supports only SGMII.
This was relatively easy to model, by special-casing the SGMII port.
On the SJA1110, certain ports can be pinmuxed between SGMII and xMII, or
between SGMII and an internal 100base-TX PHY. This creates problems,
because the driver's assumption so far was that if a port supports
SGMII, it uses SGMII.
We allow the device tree to tell us how the port pinmuxing is done, and
check that against a PHY interface type compatibility matrix for
plausibility.
The other big change is that instead of doing SGMII configuration based
on what the port supports, we do it based on what is the configured
phy_mode of the port.
The 2500base-x support added in this patch is not complete.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
So far we've succeeded in operating without keeping a copy of the
phy-mode in the driver, since we already have the static config and we
can look at the xMII Mode Parameters Table which already holds that
information.
But with the SJA1110, we cannot make the distinction between sgmii and
2500base-x, because to the hardware's static config, it's all SGMII.
So add a phy_mode property per port inside struct sja1105_private.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Looking at the SGMII PCS from SJA1110, which is accessed indirectly
through a different base address as can be seen in the next patch, it
appears odd that the address accessed through indirection still
references the base address from the SJA1105S register map (first MDIO
register is at 0x1f0000), when it could index the SGMII registers
starting from zero.
Except that the 0x1f0000 is not a base address at all, it seems. It is
0x1f << 16 | 0x0000, and 0x1f is coding for the vendor-specific MMD2.
So, it turns out, the Synopsys PCS implements all its registers inside
the vendor-specific MMDs 1 and 2 (0x1e and 0x1f). This explains why the
PCS has no overlaps (for the other MMDs) with other register regions of
the switch (because no other MMDs are implemented).
Change the code to remove the SGMII "base address" and explicitly encode
the MMD for reads/writes. This will become necessary for SJA1110 support.
Cc: Russell King <linux@armlinux.org.uk>
Cc: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The SJA1105 R and S switches have 1 SGMII port (port 4). Because there
is only one such port, there is no "port" parameter in the configuration
code for the SGMII PCS.
However, the SJA1110 can have up to 4 SGMII ports, each with its own
SGMII register map. So we need to generalize the logic.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Since commit f2f3e09396be ("net: dsa: sja1105: be compatible with
"ethernet-ports" OF node name"), DSA supports the "ethernet-ports" name
for the container node of the ports, but the sja1105 driver doesn't,
because it handles some device tree parsing of its own.
Add the second node name as a fallback.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>