linux-toradex.git/include/linux/stmmac.h, branch v6.3-rc5 Linux kernel for Apalis and Colibri modules net: stmmac: Fix for mismatched host/device DMA address width 2023-03-19T10:56:03+00:00 Jochen Henneberg jh@henneberg-systemdesign.com 2023-03-17T08:08:17+00:00 070246e4674b125860d311c18ce2623e73e2bd51 Currently DMA address width is either read from a RO device register or force set from the platform data. This breaks DMA when the host DMA address width is <=32it but the device is >32bit. Right now the driver may decide to use a 2nd DMA descriptor for another buffer (happens in case of TSO xmit) assuming that 32bit addressing is used due to platform configuration but the device will still use both descriptor addresses as one address. This can be observed with the Intel EHL platform driver that sets 32bit for addr64 but the MAC reports 40bit. The TX queue gets stuck in case of TCP with iptables NAT configuration on TSO packets. The logic should be like this: Whatever we do on the host side (memory allocation GFP flags) should happen with the host DMA width, whenever we decide how to set addresses on the device registers we must use the device DMA address width. This patch renames the platform address width field from addr64 (term used in device datasheet) to host_addr and uses this value exclusively for host side operations while all chip operations consider the device DMA width as read from the device register. Fixes: 7cfc4486e7ea ("stmmac: intel: Configure EHL PSE0 GbE and PSE1 GbE to 32 bits DMA addressing") Signed-off-by: Jochen Henneberg <jh@henneberg-systemdesign.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
Currently DMA address width is either read from a RO device register
or force set from the platform data. This breaks DMA when the host DMA
address width is <=32it but the device is >32bit.

Right now the driver may decide to use a 2nd DMA descriptor for
another buffer (happens in case of TSO xmit) assuming that 32bit
addressing is used due to platform configuration but the device will
still use both descriptor addresses as one address.

This can be observed with the Intel EHL platform driver that sets
32bit for addr64 but the MAC reports 40bit. The TX queue gets stuck in
case of TCP with iptables NAT configuration on TSO packets.

The logic should be like this: Whatever we do on the host side (memory
allocation GFP flags) should happen with the host DMA width, whenever
we decide how to set addresses on the device registers we must use the
device DMA address width.

This patch renames the platform address width field from addr64 (term
used in device datasheet) to host_addr and uses this value exclusively
for host side operations while all chip operations consider the device
DMA width as read from the device register.

Fixes: 7cfc4486e7ea ("stmmac: intel: Configure EHL PSE0 GbE and PSE1 GbE to 32 bits DMA addressing")
Signed-off-by: Jochen Henneberg <jh@henneberg-systemdesign.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: do not stop RX_CLK in Rx LPI state for qcs404 SoC 2023-01-30T07:21:25+00:00 Andrey Konovalov andrey.konovalov@linaro.org 2023-01-26T21:35:39+00:00 54aa39a513dbf2164ca462a19f04519b2407a224 Currently in phy_init_eee() the driver unconditionally configures the PHY to stop RX_CLK after entering Rx LPI state. This causes an LPI interrupt storm on my qcs404-base board. Change the PHY initialization so that for "qcom,qcs404-ethqos" compatible device RX_CLK continues to run even in Rx LPI state. Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net> 
Currently in phy_init_eee() the driver unconditionally configures the PHY
to stop RX_CLK after entering Rx LPI state. This causes an LPI interrupt
storm on my qcs404-base board.

Change the PHY initialization so that for "qcom,qcs404-ethqos" compatible
device RX_CLK continues to run even in Rx LPI state.

Signed-off-by: Andrey Konovalov <andrey.konovalov@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: Power up SERDES after the PHY link 2022-12-05T09:44:42+00:00 Revanth Kumar Uppala ruppala@nvidia.com 2022-12-01T15:58:43+00:00 a46e9010124256f5bf5fc2c241a45cf1944b768e The Tegra MGBE ethernet controller requires that the SERDES link is powered-up after the PHY link is up, otherwise the link fails to become ready following a resume from suspend. Add a variable to indicate that the SERDES link must be powered-up after the PHY link. Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com> Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
The Tegra MGBE ethernet controller requires that the SERDES link is
powered-up after the PHY link is up, otherwise the link fails to
become ready following a resume from suspend. Add a variable to indicate
that the SERDES link must be powered-up after the PHY link.

Signed-off-by: Revanth Kumar Uppala <ruppala@nvidia.com>
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
stmmac: intel: remove unused 'has_crossts' flag 2022-08-18T03:40:52+00:00 Wong Vee Khee veekhee@apple.com 2022-08-17T06:43:24+00:00 e34cfee65ec891a319ce79797dda18083af33a76 The 'has_crossts' flag was not used anywhere in the stmmac driver, removing it from both header file and dwmac-intel driver. Signed-off-by: Wong Vee Khee <veekhee@apple.com> Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de> Link: https://lore.kernel.org/r/20220817064324.10025-1-veekhee@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
The 'has_crossts' flag was not used anywhere in the stmmac driver,
removing it from both header file and dwmac-intel driver.

Signed-off-by: Wong Vee Khee <veekhee@apple.com>
Reviewed-by: Kurt Kanzenbach <kurt@linutronix.de>
Link: https://lore.kernel.org/r/20220817064324.10025-1-veekhee@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: stmmac: switch to use interrupt for hw crosstimestamping 2022-07-18T10:14:35+00:00 Wong Vee Khee vee.khee.wong@linux.intel.com 2022-07-14T07:54:27+00:00 76c16d3e19446deea98b7883f261758b96b8781a Using current implementation of polling mode, there is high chances we will hit into timeout error when running phc2sys. Hence, update the implementation of hardware crosstimestamping to use the MAC interrupt service routine instead of polling for TSIS bit in the MAC Timestamp Interrupt Status register to be set. Cc: Richard Cochran <richardcochran@gmail.com> Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
Using current implementation of polling mode, there is high chances we
will hit into timeout error when running phc2sys. Hence, update the
implementation of hardware crosstimestamping to use the MAC interrupt
service routine instead of polling for TSIS bit in the MAC Timestamp
Interrupt Status register to be set.

Cc: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: disable Split Header (SPH) for Intel platforms 2022-05-01T12:20:03+00:00 Tan Tee Min tee.min.tan@linux.intel.com 2022-04-29T11:58:07+00:00 47f753c1108e287edb3e27fad8a7511a9d55578e Based on DesignWare Ethernet QoS datasheet, we are seeing the limitation of Split Header (SPH) feature is not supported for Ipv4 fragmented packet. This SPH limitation will cause ping failure when the packets size exceed the MTU size. For example, the issue happens once the basic ping packet size is larger than the configured MTU size and the data is lost inside the fragmented packet, replaced by zeros/corrupted values, and leads to ping fail. So, disable the Split Header for Intel platforms. v2: Add fixes tag in commit message. Fixes: 67afd6d1cfdf("net: stmmac: Add Split Header support and enable it in XGMAC cores") Cc: <stable@vger.kernel.org> # 5.10.x Suggested-by: Ong, Boon Leong <boon.leong.ong@intel.com> Signed-off-by: Mohammad Athari Bin Ismail <mohammad.athari.ismail@intel.com> Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com> Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
Based on DesignWare Ethernet QoS datasheet, we are seeing the limitation
of Split Header (SPH) feature is not supported for Ipv4 fragmented packet.
This SPH limitation will cause ping failure when the packets size exceed
the MTU size. For example, the issue happens once the basic ping packet
size is larger than the configured MTU size and the data is lost inside
the fragmented packet, replaced by zeros/corrupted values, and leads to
ping fail.

So, disable the Split Header for Intel platforms.

v2: Add fixes tag in commit message.

Fixes: 67afd6d1cfdf("net: stmmac: Add Split Header support and enable it in XGMAC cores")
Cc: <stable@vger.kernel.org> # 5.10.x
Suggested-by: Ong, Boon Leong <boon.leong.ong@intel.com>
Signed-off-by: Mohammad Athari Bin Ismail <mohammad.athari.ismail@intel.com>
Signed-off-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: Add platform level debug register dump feature 2021-11-30T11:57:58+00:00 Bhupesh Sharma bhupesh.sharma@linaro.org 2021-11-28T19:58:54+00:00 4047b9db1aa7512a10ba3560a3f63821c8c40235 dwmac-qcom-ethqos currently exposes a mechanism to dump rgmii registers after the 'stmmac_dvr_probe()' returns. However with commit 5ec55823438e ("net: stmmac: add clocks management for gmac driver"), we now let 'pm_runtime_put()' disable the clocks before returning from 'stmmac_dvr_probe()'. This causes a crash when 'rgmii_dump()' register dumps are enabled, as the clocks are already off. Since other dwmac drivers (possible future users as well) might require a similar register dump feature, introduce a platform level callback to allow the same. This fixes the crash noticed while enabling rgmii_dump() dumps in dwmac-qcom-ethqos driver as well. It also allows future changes to keep a invoking the register dump callback from the correct place inside 'stmmac_dvr_probe()'. Fixes: 5ec55823438e ("net: stmmac: add clocks management for gmac driver") Cc: Joakim Zhang <qiangqing.zhang@nxp.com> Cc: David S. Miller <davem@davemloft.net> Signed-off-by: Bhupesh Sharma <bhupesh.sharma@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net> 
dwmac-qcom-ethqos currently exposes a mechanism to dump rgmii registers
after the 'stmmac_dvr_probe()' returns. However with commit
5ec55823438e ("net: stmmac: add clocks management for gmac driver"),
we now let 'pm_runtime_put()' disable the clocks before returning from
'stmmac_dvr_probe()'.

This causes a crash when 'rgmii_dump()' register dumps are enabled,
as the clocks are already off.

Since other dwmac drivers (possible future users as well) might
require a similar register dump feature, introduce a platform level
callback to allow the same.

This fixes the crash noticed while enabling rgmii_dump() dumps in
dwmac-qcom-ethqos driver as well. It also allows future changes
to keep a invoking the register dump callback from the correct
place inside 'stmmac_dvr_probe()'.

Fixes: 5ec55823438e ("net: stmmac: add clocks management for gmac driver")
Cc: Joakim Zhang <qiangqing.zhang@nxp.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: Bhupesh Sharma <bhupesh.sharma@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: Calculate CDC error only once 2021-11-24T04:11:05+00:00 Kurt Kanzenbach kurt@linutronix.de 2021-11-22T11:19:31+00:00 c6d5f1933085f9a92ed5c256a859ab31c7a35f88 The clock domain crossing error (CDC) is calculated at every fetch of Tx or Rx timestamps. It includes a division. Especially on arm32 based systems it is expensive. It also requires two conditionals in the hotpath. Add a compensation value cache to struct plat_stmmacenet_data and subtract it unconditionally in the RX/TX functions which spares the conditionals. The value is initialized to 0 and if supported calculated in the PTP initialization code. Suggested-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de> Link: https://lore.kernel.org/r/20211122111931.135135-1-kurt@linutronix.de Signed-off-by: Jakub Kicinski <kuba@kernel.org> 
The clock domain crossing error (CDC) is calculated at every fetch of Tx or Rx
timestamps. It includes a division. Especially on arm32 based systems it is
expensive. It also requires two conditionals in the hotpath.

Add a compensation value cache to struct plat_stmmacenet_data and subtract it
unconditionally in the RX/TX functions which spares the conditionals.

The value is initialized to 0 and if supported calculated in the PTP
initialization code.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Kurt Kanzenbach <kurt@linutronix.de>
Link: https://lore.kernel.org/r/20211122111931.135135-1-kurt@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net: stmmac: ptp: update tas basetime after ptp adjust 2021-07-05T17:16:17+00:00 Xiaoliang Yang xiaoliang.yang_1@nxp.com 2021-07-05T10:26:55+00:00 e9e3720002f61cd637a49ecafae77cac230eefae After adjusting the ptp time, the Qbv base time may be the past time of the new current time. dwmac5 hardware limited the base time cannot be set as past time. This patch add a btr_reserve to store the base time get from qopt, then calculate the base time and reset the Qbv configuration after ptp time adjust. Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
After adjusting the ptp time, the Qbv base time may be the past time
of the new current time. dwmac5 hardware limited the base time cannot
be set as past time. This patch add a btr_reserve to store the base
time get from qopt, then calculate the base time and reset the Qbv
configuration after ptp time adjust.

Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net: stmmac: add mutex lock to protect est parameters 2021-07-05T17:16:17+00:00 Xiaoliang Yang xiaoliang.yang_1@nxp.com 2021-07-05T10:26:54+00:00 b2aae654a4794ef898ad33a179f341eb610f6b85 Add a mutex lock to protect est structure parameters so that the EST parameters can be updated by other threads. Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> 
Add a mutex lock to protect est structure parameters so that the
EST parameters can be updated by other threads.

Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>