- Oct 13, 2021
-
-
Robert Nelson authored
5.10.65-ti-arm64-r14 bb.org_defconfig 5.10 TI Delta: https://github.com/RobertCNelson/ti-linux-kernel/compare/5d12f5b8e60868c69a39c555789fe4824fc774dd...f25fed53436a83d746c481ba852b3d4c10d0bab0 BBDTBS: https://github.com/beagleboard/BeagleBoard-DeviceTrees/commit/c7d854f34548a7d5aec9cf716daa878c83b8d99f WPANUSB: https://github.com/statropy/wpanusb/commit/251f0167545bf2dcaa3cad991a59dbf5ab05490a WIRELESS_REGDB: https://git.kernel.org/pub/scm/linux/kernel/git/sforshee/wireless-regdb.git/commit/?id=47007d0169c4c6c855a3fc5333c142489a43e89e Signed-off-by:
Robert Nelson <robertcnelson@gmail.com>
-
Robert Nelson authored
-
Robert Nelson authored
Signed-off-by: -
Robert Nelson authored
Signed-off-by: -
Robert Nelson authored
Reference: v5.14.12 Signed-off-by: -
Robert Nelson authored
This reverts commit 956b200a . Signed-off-by:
-
Robert Nelson authored
Reference: v5.14.12 Signed-off-by: -
Robert Nelson authored
Reference: v5.10.73 Signed-off-by: -
Robert Nelson authored
Signed-off-by: -
-
-
Jason Kridner authoredFrom https://github.com/statropy/wpanusb
-
-
- Oct 09, 2021
-
-
LCPD Auto Merger authored
TI-Feature: connectivity TI-Branch: connectivity-ti-linux-5.10.y * 'connectivity-ti-linux-5.10.y' of ssh://bitbucket.itg.ti.com/lcpdpublicdom/connectivity : net: ethernet: ti: icssg_prueth: declare NETIF_F_SG support arm64: dts: ti: k3-am642-evm-icssg1-dualemac: add overlay for mii mode net: phy: dp83869: fix mii mode when rgmii strap cfg is used net: ethernet: ti: prueth: fix build warn Signed-off-by:
LCPD Auto Merger <lcpd_integration@list.ti.com>
-
- Oct 07, 2021
-
-
The ICSSG driver can accept fragmented skbs, but it doesn't declare NETIF_F_SG as result Net core will linear all skbs during xmit. It also disables GSO which depends on SG. Hence declare NETIF_F_SG feature support. Signed-off-by:
Grygorii Strashko <grygorii.strashko@ti.com> Signed-off-by:
Vignesh Raghavendra <vigneshr@ti.com>
-
Add overlay to enable 2 ICSSG1 ports in MII mode. Signed-off-by:
-
The DP83869 PHY on TI k3-am642-evm is supported both MII and RGMII interfaces and configured by default to use RGMII interface (strap), but the board design is also allowed to switch dynamically to MII interface for testing purposes by applying different set of pinmuxes. To support such switch to MII interface update the DP83869 PHY driver to configure OP_MODE_DECODE.RGMII_MII_SEL(bit 5) properly when MII PHY interface mode is requested. Signed-off-by:
-
Remove debug leftovers which causes warning: ../drivers/net/ethernet/ti/prueth_core.c: In function ‘prueth_hsr_ptp_ct_tx_ts_enqueue’: ../drivers/net/ethernet/ti/prueth_core.c:1066:96: warning: cast from pointer to integer of different size [-Wpointer-to-int-cast] netdev_warn(other_emac->ndev, "Dropped cut through event waiting for tx ts. %d %x\n", event, (u32)skb); ^ Fixes: 0dc2edeb ("net: ti: prueth: hsr: add support for packets timestamping") Signed-off-by:
-
- Sep 30, 2021
-
-
LCPD Auto Merger authored
TI-Feature: connectivity TI-Branch: connectivity-ti-linux-5.10.y * 'connectivity-ti-linux-5.10.y' of ssh://bitbucket.itg.ti.com/lcpdpublicdom/connectivity : ARM: dts: am57xx-idk: add interrupts for hsr/prp tx ts net: ti: prueth: hsr: add support for packets timestamping net: socket: fix __sock_recv_redinfo_timestamp() net: ethernet: ti: fix ethtool regs dump net: ethernet: ti: prueth_lre: clean up rpmsg-kdrv: ethswitch: update num remote devices to 4 Signed-off-by:
-
Add hsr/prp TX timedstamp IRQs. Signed-off-by:
-
When PTP support is enabled in hsr firmware, the firmware timestamps the ptp packets that are going out via prueth ports. The timestamp value is copied to a shared memory location. After timestamping, a host irq is raised by the firmware as a notification to read the timestamp. When a HSR non-link-local PTP event is cut-through forward by the HSR firmware, in addition to the rx timestamp, a tx timestamp is also provided by the firmware. The tx timestamp is saved in the shared redundant info area of the received skb before the skb is passed to upper layer. Currently cut-through tx timestamp is handled only for HSR PTP over L2 non-link-local event. Add support for handling this ptp tx irq, read the tx timestamp and pass it to userspace layer. In case, of cut through - on RX get RX timestamp and delay PTP pcaket delivery until TX timestamp is ready. Signed-off-by:
Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by:
-
The LRE timestamp passed to user space using struct scm_timestamping_internal in SCM_RED_TIMESTAMPING msg (control), which is wrong as struct scm_timestamping_internal is internal to Kernel and doesn't correspond to UAPI, and can have different size depending on platform. The struct scm_timestamping has to be used instead. Fixes: b98f7ec0 ("net: socket: Forward extra timestamp of received skb to user space") Signed-off-by:
-
The HSR FW has MC and VLAN data placed at different offsets, hence fix it. Fixes: d2e8eb5a ("net: ti: prueth_core: hsr/prp: add HSR/PRP driver") Signed-off-by:
-
- mark some func as static - clean up prueth_lre_emac_rx_packets() Signed-off-by:
-
The EthFW declares 4 rpmsg-kdrv devices hence update RPMSG_KDRV_TP_MAX_DEVICES to 4. Without this change rpmsg-kdrv device info is not passed to kernel (fw version, etc). Fixes: f4454648 ("rpmsg-kdrv: ethswitch: add support for mac-only ports") Signed-off-by:
-
Vignesh Raghavendra authored
Merge tag 'v5.10.65' of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux into ti-linux-5.10.y This is the 5.10.65 stable release * tag 'v5.10.65' of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux : (597 commits) Linux 5.10.65 clk: kirkwood: Fix a clocking boot regression backlight: pwm_bl: Improve bootloader/kernel device handover fbmem: don't allow too huge resolutions IMA: remove the dependency on CRYPTO_MD5 IMA: remove -Wmissing-prototypes warning fuse: flush extending writes fuse: truncate pagecache on atomic_o_trunc ARM: dts: at91: add pinctrl-{names, 0} for all gpios KVM: nVMX: Unconditionally clear nested.pi_pending on nested VM-Enter KVM: VMX: avoid running vmx_handle_exit_irqoff in case of emulation KVM: x86: Update vCPU's hv_clock before back to guest when tsc_offset is adjusted KVM: s390: index kvm->arch.idle_mask by vcpu_idx Revert "KVM: x86: mmu: Add guest physical address check in translate_gpa()" x86/resctrl: Fix a maybe-uninitialized build warning treated as error perf/x86/amd/ibs: Extend PERF_PMU_CAP_NO_EXCLUDE to IBS Op tty: Fix data race between tiocsti() and flush_to_ldisc() bio: fix page leak bio_add_hw_page failure io_uring: IORING_OP_WRITE needs hash_reg_file set time: Handle negative seconds correctly in timespec64_to_ns() ... Signed-off-by:
-
Vignesh Raghavendra authored
Reverting above commit in favor of better fix from upstream stable tree. Signed-off-by:
-
- Sep 27, 2021
-
-
LCPD Auto Merger authored
TI-Feature: connectivity TI-Branch: connectivity-ti-linux-5.10.y * 'connectivity-ti-linux-5.10.y' of ssh://bitbucket.itg.ti.com/lcpdpublicdom/connectivity : net: ethernet: ti: icss_iep: fix phc2sys sync net: ti: prueth_core: always initialize firmware offsets field Signed-off-by:
-
- Sep 24, 2021
-
-
Now when phc2sys is running to sync sys-clock to PHC the sync doesn't happen: phc2sys[2046.273]: CLOCK_REALTIME phc offset -1588667462018990140 s0 freq +100000000 delay 1048389136 phc2sys[2047.273]: clockcheck: clock jumped backward or running slower than expected! phc2sys[2047.273]: CLOCK_REALTIME phc offset -1588667463942743044 s0 freq +100000000 delay 0 phc2sys[2048.273]: CLOCK_REALTIME phc offset -864962291571182424 s2 freq +100000000 delay 1447410346965662944 phc2sys[2049.273]: clockcheck: clock jumped backward or running slower than expected! phc2sys[2049.273]: CLOCK_REALTIME phc offset -1588667465352817468 s0 freq +100000000 delay 1048389136 cmd master: phc2sys -a -rr -m -q -l5& ptp4l -2 -i eth0 -l5 -m -q --priority1 127& cmd slave: phc2sys -a -r -q -m -l6& ptp4l -2 -s -i eth3 -m -q -l5& The issue is due PTP IEP PHC clock was converted to use .gettimex64() callback and code path for AM65x SR1.0 IEP was updated properly to fill struct ptp_system_timestamp *sts parameter, but the same is missing for AM65x SR2.0 IEP code path where gettime() is implemented in icssg_prueth driver through .gettime() callback. Hence fix AM65x SR2.0 IEP to pass struct ptp_system_timestamp to icss_iep_clockops.gettime() callback and fill it properly. After this change: phc2sys -a -r -q -m -l6 phc2sys[177.659]: reconfiguring after port state change phc2sys[177.659]: selecting CLOCK_REALTIME for synchronization phc2sys[177.659]: selecting eth3 as the master clock phc2sys[177.659]: CLOCK_REALTIME phc offset 43656345542569315 s0 freq +0 delay 330 phc2sys[178.659]: CLOCK_REALTIME phc offset 43656345542534414 s1 freq -34897 delay 330 phc2sys[179.659]: CLOCK_REALTIME phc offset -2 s2 freq -34899 delay 330 phc2sys[180.660]: CLOCK_REALTIME phc offset -19 s2 freq -34917 delay 325 phc2sys[181.660]: CLOCK_REALTIME phc offset -18 s2 freq -34921 delay 330 phc2sys[182.660]: CLOCK_REALTIME phc offset -48 s2 freq -34957 delay 330 phc2sys[183.660]: CLOCK_REALTIME phc offset -34 s2 freq -34957 delay 330 phc2sys[184.660]: CLOCK_REALTIME phc offset -32 s2 freq -34965 delay 325 phc2sys[185.660]: CLOCK_REALTIME phc offset -12 s2 freq -34955 delay 330 phc2sys[186.660]: CLOCK_REALTIME phc offset 1 s2 freq -34946 delay 330 phc2sys[187.660]: CLOCK_REALTIME phc offset 2 s2 freq -34944 delay 330 Signed-off-by:
-
prueth_set_fw_offsets() is always called by prueth_probe(), so make sure that prueth->fw_offsets is never NULL. Signed-off-by:
Adrien Ricciardi <aricciardi@baylibre.com> Signed-off-by:
Neil Armstrong <narmstrong@baylibre.com>
-
- Sep 23, 2021
-
-
LCPD Auto Merger authored
TI-Feature: connectivity TI-Branch: connectivity-ti-linux-5.10.y * 'connectivity-ti-linux-5.10.y' of ssh://bitbucket.itg.ti.com/lcpdpublicdom/connectivity : arm64: dts: ti: k3-j721e-eaik: Add DT overlay for RPi camera V2 arm64: dts: ti: k3-j721e-eaik: Add overlay for CSI2 interfaced with OV5640 camera arm64: dts: ti: k3-j721e-main: Add nodes to enable CSI2-1 and DPHY-1 dt-bindings: i2c: i2c-mux-pca954x: Convert to DT schema Signed-off-by:
-
Two RPi camera headers present on EAIK are connected to each port of two CSI2RX instances 0 and 1. Sensors are configured through I2C interface via an I2C switch. CSI2RX-0 ====> \\ ==>RPi header==>RPi-Cam0(IMX219) // MAIN_I2C ==> I2C_switch(TCA954) ===> \\ ==>RPi header==>RPi-Cam1(IMX219) CSI2RX-1 ====> // Both MIPI and RPi connectors are connected to the CSI2 interface. A mux select should be set to high to select the RPi connector. So add a node to set the mux select pin to high. Add node for the I2C switch through which the camera sensors shall be configured. Add IMX219 camera nodes as a child node to the I2C switch. Add endpoint nodes between RPi cameras and CSI2RX. See RPi Camera module V2.1 schematics here: https://datasheets.raspberrypi.org/camera/camera-v2-schematic.pdf Signed-off-by:Sinthu Raja <sinthu.raja@ti.com> Signed-off-by:
-
The OV5640 camera sensor is interfaced with CSI2RX through a MIPI header. The sensor shall be configured through the I2C which is interfaced with the camera sensor via an I2C switch. MAIN_I2C ===> I2C_switch(TCA954) ====> \\ ==>MIPI header ===> OV5640 CSI2RX0 ====> // Add node for the I2C switch through which the camera sensor is connected. Add node for OV5640 camera connected to port 0 of the CSI bridge. Add pinmux for the reset gpio and CSI2 Reference clock to the camera sensor. Signed-off-by: -
Add node for the second instance of Cadence CSI2RX-1, DPHY-1, and TI's CSI2RX wrapper. This enables CSI2 capture from the second instance of the J721E camera subsystem. Signed-off-by:
Pratyush Yadav <p.yadav@ti.com>
-
Convert the i2c-mux-pca954x binding to DT schema format. Add the missing compatible for 'nxp,pca9645' which is already in use. commit 86335b51 upstream Signed-off-by:
Rob Herring <robh@kernel.org> Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by:
Wolfram Sang <wsa@kernel.org> Acked-by:
Peter Rosin <peda@axentia.se> Link: https://lore.kernel.org/r/20210526184839.2937899-6-robh@kernel.org
-
LCPD Auto Merger authored
TI-Feature: connectivity TI-Branch: connectivity-ti-linux-5.10.y * 'connectivity-ti-linux-5.10.y' of ssh://bitbucket.itg.ti.com/lcpdpublicdom/connectivity : (41 commits) arm64: dts: ti: k3-j721e: Update compatible for DPHY phy: cdns-dphy: Add common module reset to DPHY RX phy: dt-bindings: cdns-dphy: Add j721e-dphy compatible media: i2c: ov5647: use pm_runtime_resume_and_get() media: ov5647: Remove 640x480 SBGGR8 mode media: ov5647: Support VIDIOC_SUBSCRIBE_EVENT media: ov5647: Constify oe_enable/disable reglist media: ov5647: Apply controls only when powered media: ov5647: Rework s_stream() operation media: ov5647: Use pm_runtime infrastructure media: ov5647: Advertise the correct exposure range media: ov5647: Support V4L2_CID_VBLANK control media: ov5647: Support V4L2_CID_HBLANK control media: ov5647: Support V4L2_CID_PIXEL_RATE media: ov5647: Set V4L2_SUBDEV_FL_HAS_EVENTS flag media: ov5647: Implement set_fmt pad operation media: ov5647: Use SBGGR10_1X10 640x480 as default media: ov5647: Add SGGBR10_1X10 modes media: ov5647: Rename SBGGR8 VGA mode media: ov5647: Add support for get_selection() ... Signed-off-by:
-
Add the j721e specific compatible to DPHY node. Signed-off-by:
-
DPHY RX module has a common module reset (RSTB_CMN) among four different other resets. The module expects all four resets to be released in a specific order. In TI J721E SR1.0 the RSTB_CMN is internally tied to CSI_RX_RST which is hardware controlled, so no explicit control from software is required to reset the DPHY RX module. With J721E SR2.0 the RSTB_CMN is separated out of the hardware controlled and tied to a MMR bit of the DPHY RX module. An explicit write to control the common module reset from software is required. Update DPHY_RX configure function to release the common module reset. Currently this feature is specific to TI J721E SR2.0 therefore add a compatible string and introduce a driver quirk to release the reset and set it for J721E SR2.0. Signed-off-by:
-
Document the j721e compatible for cdns-dphy. Signed-off-by:
-
commit 5187df40 upstream. Commit dd8088d5 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors. Use the new API, in order to cleanup the error check logic. (cherry picked from commit 5187df40 ) Signed-off-by:
Mauro Carvalho Chehab <mchehab+huawei@kernel.org> Signed-off-by:
Nishanth Menon <nm@ti.com> Signed-off-by:
Jacopo Mondi <jacopo@jmondi.org> Acked-by:
-
