Ipq807x 5.15 pr #1
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
All of the ath11k supported AHB SoC-s as well as PCI cards require firmware binaries. So provide them from the QUIC repos which look like the new upstream for ath11k related firmware images. Signed-off-by: Robert Marko <[email protected]>
QRTR is Qualcomm IPC router protocol and ath11k requires it for both AHB and PCI support, so package it as a kernel module so it can be added as a dependency to the ath11k module. Only kernel 5.15 is currently supported due to various things missing in 5.10 whose backporting is out of scope for this patch. SMD, TUN and MHI variants are packaged. SMD variant depends on the ipq807x target as it has dependency on the RPMSG drivers which are Qualcomm and SoC specific anyway. Signed-off-by: Robert Marko <[email protected]>
ath11k is the upstream driver for Qualcomm 802.11ax radios, both for the internal AHB and PCI based cards. It differs a bit from ath10k as it requires stuff like QRTR, QMI helpers so those had to be packaged before this patch. Since AHB is SoC specific, it currently depeneds on ipq807x target. Signed-off-by: Robert Marko <[email protected]>
NSS packages required for wired networking and NSS ofloading are located in a external package feed. Include the feed by default in order to provide wired networking on IPQ807x. Signed-off-by: Robert Marko <[email protected]>
Qualcomm Atheros IPQ807x is a modern WiSoC featuring: * Quad Core ARMv8 Cortex A-53 * @ 2.2 GHz (IPQ8072A/4A/6A/8A) Codename Hawkeye * @ 1.4 GHz (IPQ8070A/1A) Codename Acorn * Dual Band simultaneaous IEEE 802.11ax * 5G: 8x8/80 or 4x4/160MHz (IPQ8074A/8A) * 5G: 4x4/80 or 2x2/160MHz (IPQ8071A/2A/6A) * 5G: 2x2/80MHz (IPQ8070A) * 2G: 4x4/40MHz (IPQ8072A/4A/6A/8A) * 2G: 2x2/40MHz (IPQ8070A/1A) * 1x PSGMII via QCA8072/5 (Max 5x 1GbE ports) * 2x SGMII/USXGMII (1/2.5/5/10 GbE) on Hawkeye * 2x SGMII/USXGMII (1/2.5/5 GbE) on Acorn * DDR3L/4 32/16 bit up to 2400MT/s * SDIO 3.0/SD card 3.0/eMMC 5.1 * Dual USB 3.0 * One PCIe Gen2.1 and one PCIe Gen3.0 port (Single lane) * Parallel NAND (ONFI)/LCD * 6x QUP BLSP SPI/I2C/UART * I2S, PCM, and TDMA * HW PWM * 1.8V configurable GPIO * Companion PMP8074 PMIC via SPMI (GPIOS, RTC etc) Note that only v2 SOC models aka the ones ending with A suffix are supported, v1 models do not comply to the final 802.11ax and have lower clocks, lack the Gen3 PCIe etc. Signed-off-by: Robert Marko <[email protected]>
Xiaomi AX3600 is a budget 802.11ax dual-band router/AP. Specifications: * CPU: Qualcomm IPQ8071A Quad core Cortex-A53 1.4GHz * RAM: 512MB of DDR3 * Storage: 256MB of parallel NAND * Ethernet: 4x1G RJ45 ports (QCA8075) with 1x status LED per port * WLAN: * PCI based Qualcomm QCA9889 1x1 802.11ac Wawe 2 for IoT * 2.4GHz: Qualcomm QCN5024 2x2 802.11b/g/n/ax 574 Mbps PHY rate * 5GHz: Qualcomm QCN5054 4x4@80MHz or 2x2@160MHz 802.11a/b/g/n/ac/ax 2402 PHY rate * LED-s: * System (Blue and Yellow) * IoT (Blue) * Network (Blue and Yellow) * Buttons: 1x Soft reset * Power: 12V DC Jack Installation instructions: Obtaining SSH access is mandatory https://openwrt.org/inbox/toh/xiaomi/xiaomi_ax3600#obtain_ssh_access For first installation, using ubiformat is mandatory: * Copy openwrt-ipq807x-generic-xiaomi_ax3600-squashfs-nand-factory.ubi to /tmp * cat /proc/mtd and find the rootfs and rootfs_1 mtd index * Use ubiformat to flash the UBI image: ubiformat /dev/mtd12 -y -f /tmp/openwrt-ipq807x-generic-xiaomi_ax3600-squashfs-nand-factory.ubi ubiformat /dev/mtd13 -y -f /tmp/openwrt-ipq807x-generic-xiaomi_ax3600-squashfs-nand-factory.ubi Please confirm that mtd12 and mtd13 are the correct indexes. * reboot After that regular sysupgrade can be used. Signed-off-by: Robert Marko <[email protected]> Signed-off-by: Ansuel Smith <[email protected]>
Redmi AX6 is a budget 802.11ax dual-band router/AP Specifications: * CPU: Qualcomm IPQ8071A Quad core Cortex-A53 1.4GHz * RAM: 512MB of DDR3 * Storage: 128MB NAND * Ethernet: 4x1G RJ45 ports (QCA8075) * WLAN: * 2.4GHz: Qualcomm QCN5024 2x2 802.11b/g/n/ax 574 Mbps PHY rate * 5GHz: Qualcomm QCN5054 4x4@80MHz or 2x2@160MHz 802.11a/b/g/n/ac/ax 2402 PHY rate * LEDs: * System (Blue/Yellow) * Network (Blue/Yellow) *Buttons: 1x soft reset *Power: 12V DC jack Installation instructions: Must obtain SSH access first. For first installlation, using ubiformat is mandatory: * Copy ubi file to the router's /tmp * cat /proc/mtd and find the rootfs and rootfs_1 mtd index * Use ubiformat to flash the UBI image: * if you are on rootfs (mtd12) ubiformat /dev/mtd13 -y -f /tmp/openwrt-ipq807x-generic-redmi_ax6-squashfs-nand-factory.ubi nvram set flag_last_success=1 nvram set flag_boot_rootfs=1 nvram commit * if you are on rootfs_1 (mtd13) ubiformat /dev/mtd12 -y -f /tmp/openwrt-ipq807x-generic-redmi_ax6-squashfs-nand-factory.ubi nvram set flag_last_success=0 nvram set flag_boot_rootfs=0 nvram commit * reboot * ubiformat the other rootfs * Now the regualr sysupgrade can be used Signed-off-by: Zhijun You <[email protected]>
Edimax CAX1800 is a 802.11 ax dual-band AP
with PoE. AP can be ceiling or wall mount.
Specifications:
• CPU: Qualcomm IPQ8070A Quad core Cortex-A53 1.4GHz
• RAM: 512MB of DDR3
• Storage: 128MB NAND (contains rootfs) / 8MB NOR (contains art and uboot-env)
• Ethernet: 1x 1G RJ45 port (QCA8072) PoE
• WLAN:
2.4GHz: Qualcomm QCN5024 2x2 802.11b/g/n/ax 574 Mbps PHY rate
5GHz: Qualcomm QCN5054 2x2 802.11a/b/g/n/ac/ax 1201 PHY rate
• LEDs:
3 x GPIO-controlled System-LEDs
(form one virtual RGB System-LED)
black_small_square Buttons: 1x soft reset
black_small_square Power: 12V DC jack or PoE (802.3af )
An unpopulated serial header is onboard.
RX/TX is working, bootwait is active, secure boot is not enabled.
SSH can be activated in the stock firmware, but it drops only
to a limited shell .
Installation Instructions:
black_small_square obtain serial access
black_small_square stop auto boot
black_small_square tftpboot the initramfs image (serverip is set to 192.168.99.8 in uboot)
black_small_square bootm
black_small_square copy openwrt-ipq807x-generic-edimax_cax1800-squashfs-nand-factory.ubi
to the device
black_small_square write the image to the NAND:
black_small_square cat /proc/mtd and look for rootfs partition (should be mtd0)
black_small_square ubiformat /dev/mtd0 -f -y openwrt-ipq807x-generic-edimax_cax1800-squashfs-
nand-factory.ubi
black_small_square reboot
Note: Device is not using dual partitioning (NAND contains other partitions
with different manufacture data etc.)
Draytek VigorAP 960C and Lancom LW-600 both look similar, but I haven't checked them.
Signed-off-by: Dirk Buchwalder <[email protected]>
QNAP 301w is a AX WIFI router with 4 1G and 2 10G ports.
Specifications:
• CPU: Qualcomm IPQ8072A Quad core Cortex-A53 2.2GHz
• RAM: 1024MB of DDR3
• Storage: 4GB eMMC (contains kernel and rootfs) / 8MB NOR
(contains art and u-boot-env)
• Ethernet: 4x 1G RJ45 ports + 2 10G ports (Aquantia AQR113C)
• WLAN:
2.4GHz: Qualcomm QCN5024 2x2 802.11b/g/n/ax 1174 Mbps PHY rate
5GHz: Qualcomm QCN5054 4x4 802.11a/b/g/n/ac/ax 2402 PHY rate
• LEDs:
7 x GPIO-controlled dual color LEDs + 2 GPIO-controlled single color LEDs
• Buttons: 1x soft reset / 1x WPS
• Power: 12V DC jack
A poulated serial header is onboard.
RX/TX is working, bootwait is active, secure boot is not enabled.
SSH can be activated in the stock firmware, hold WPS button til the second beep
(yes the router has a buzzer)
SSH is available on port 22200, login with user admin and
password "mac address of the router".
Installation Instructions:
• obtain serial access (https://openwrt.org/inbox/toh/qnap/301w#serial)
• stop auto boot
• setenv serverip 192.168.10.1
• setenv ipaddr 192.168.10.10
• tftpboot the initramfs image
(openwrt-ipq807x-generic-qnap_301w-initramfs-fit-uImage.itb)
• bootm
• make sure that current_entry is set to "0":
"fw_printenv -n current_entry" should be print "0". If not,
do "fw_setenv current_entry 0"
• copy openwrt-ipq807x-generic-qnap_301w-squashfs-sysupgrade.bin
to the device to /tmp folder
• sysupgrade -n /tmp/openwrt-ipq807x-generic-qnap_301w-squashfs-sysupgrade.bin
this flashes openwrt to the first kernel and rootfs partition (mmcblk0p1 / mmcblk0p4)
• reboot
Note: this leaves the second kernel / rootfs parition untouched. So if you want
to go back to stock, stop u-boot autoboot, "setenv current_entry 1" ,
"saveenv", "bootipq".
Stock firmware should start from the second partition.
Then do a firmwareupgrade in the stock gui, that should overwrite the openwrt
in the first partitions
Make 10G Aquantia phy's work:
The aquantia phy's need a firmware to work. This can either be loaded
in linux with a userspace tool or in u-boot.
I was not successfull to load the firmware in linux (aq-fw-download) but luckily there is
aq_load_fw available in u-boot. But first the right firmware needs to write
to the 0:ETHPHYFW mtd partition (it is empty on my device)
Grab the ethphy firmware image from:
https://github.com/kirdesde/nbg7815_gpl/blob/master/target/linux/ipq/ipq807x_64/prebuilt_images/AQR_ethphyfw.mbn
and scp that to openwrt.
Check the 0:ETHPHYFW partition number:
cat /proc/mtd|grep "0:ETHPHYFW", should be mtd10.
Backup the 0:ETHPHYFW partition:
dd if=/dev/mtd10 of=/tmp/ethphyfw.backup, scp ethphyfw.backup to a save place.
Write the new firmware image to the 0:ETHPHYFW partition:
"mtd erase /dev/mtd10", "mtd -n write AQR_ethphyfw.mbn /dev/mtd10".
Reboot to u-boot.
Check if aq_load_fw is working:
"aq_load_fw 0", that checks the firmware and if successfull,
loads iram and dram to one of the aquantia phy's.
If that worked, add the aq_load_fw to the bootcmd:
setenv bootcmd "aq_load_fw 0 && aq_load_fw 8 && bootipq"
"saveenv"
"reset"
Board reboots and the firmware load to both phy's should start and
then openwrt boots.
Check if the 10G ports work.
Note: lan port labeled "10G-2" is configured as WAN port as per default.
All other port are in the br-lan. This can be changed in the network config.
Signed-off-by: Dirk Buchwalder <[email protected]>
Xiaomi AX9000 is a premium 802.11ax "tri"-band router/AP. Specifications: * CPU: Qualcomm IPQ8072A Quad core Cortex-A53 2.2GHz * RAM: 1024MB of DDR3 * Storage: 256MB of parallel NAND * Ethernet: * 4x1G RJ45 ports (QCA8075) with 1x status LED per port * 1x2.5G RJ45 port (QCA8081) with 1x status LED * WLAN: * PCI based Qualcomm QCA9889 1x1 802.11ac Wawe 2 for IoT * 2.4GHz: Qualcomm QCN5024 4x4@40MHz 802.11b/g/n/ax 1147 Mbps PHY rate * 5.8GHz: Qualcomm QCN5054 4x4@80MHz or 2x2@160MHz 802.11a/b/g/n/ac/ax 2402Mbps PHY rate * 5GHz: PCI based Qualcomm QCN9024 4x4@160MHz 802.11a/b/g/n/ac/ax 4804Mbps PHY rate * USB: 1x USB3.0 Type-A port * LED-s: * System (Blue and Yellow) * Network (Blue and Yellow) * RGB light bar on top in X shape * Buttons: * 1x Power switch * 1x Soft reset * 1x Mesh button * Power: 12V DC Jack Installation instructions: Obtaining SSH access is mandatory https://openwrt.org/inbox/toh/xiaomi/ax9000#obtain_ssh_access For first installation, using ubiformat is mandatory: * Copy openwrt-ipq807x-generic-xiaomi_ax9000-squashfs-nand-factory.ubi to /tmp * cat /proc/mtd and find the rootfs and rootfs_1 mtd index * Use ubiformat to flash the UBI image: ubiformat /dev/mtd21 -y -f /tmp/openwrt-ipq807x-generic-xiaomi_ax9000-squashfs-nand-factory.ubi ubiformat /dev/mtd22 -y -f /tmp/openwrt-ipq807x-generic-xiaomi_ax9000-squashfs-nand-factory.ubi Please confirm that mtd21 and mtd22 are the correct indexes. * reboot After that regular sysupgrade can be used. Signed-off-by: Robert Marko <[email protected]>
The Edgecore EAP102 is a wall/ceiling mountable AP. The AP can be powered by either PoE or AC adapter. Device info: - IPQ8071-A SoC - 1GiB RAM - 256MiB NAND flash - 32MiB SPI NOR - 2 Ethernet ports - 1 Console port - 2GHz/5GHz AX WLAN - 2 USB 2.0 ports Install instructions: Prerequistes - TFTP server, preferrably within 192.168.1.0/24 Console cable plugged in (115200 8N1 no flow control) 1. Power on device and interrupt u-boot to obtain u-boot CLI 2. set serverip to IP address of the TFTP server: `setenv serverip 192.168.1.250` 3. Download image from TFTP server: `tftpboot 0x44000000 openwrt-ipq807x-generic-edgecore_eap102-squashfs-nand-factory.ubi` 4. Flash ubi image to both partitions and reset: `sf probe imxtract 0x44000000 ubi nand device 0 nand erase 0x0 0x3400000 nand erase 0x3c00000 0x3400000 nand write $fileaddr 0x0 $filesize nand write $fileaddr 0x3c00000 $filesize reset` Signed-off-by: Matthew Hagan <[email protected]>
Add workflow that will build the branch that is aimed at finally making a PR to get the target into OpenWrt. Signed-off-by: Robert Marko <[email protected]>
Backport and ship QMI helpers as part of backports. This way we dont have to patch the kernel to make the tristate symbol user selectable and we dont have to package them. Signed-off-by: Robert Marko <[email protected]>
Utilize the backported QMI helpers instead of the OpenWrt packaged ones. Signed-off-by: Robert Marko <[email protected]>
This should provide a much better changelog. Signed-off-by: Robert Marko <[email protected]>
Include the Ansuels RFC peer fix that is pending upstream. Signed-off-by: Robert Marko <[email protected]>
Enable ath11k ethernet encapsulation/decapsulation offload by default. This improves wireless performance without requiring user configuration. Signed-off-by: Tiago Gaspar <[email protected]>
|
pull from robi/ipq807x-pr |
rmandrad
pushed a commit
that referenced
this pull request
Oct 25, 2022
Turris MOX randomly crashes up, when there is connected miniPCIe card MediaTek MT7915 with the following output: [ 71.457007] Internal error: synchronous external abort: 96000210 [#1] SMP [ 71.464021] Modules linked in: xt_connlimit pppoe ppp_async nf_conncount iptable_nat ath9k xt_state xt_nat xt_helper xt_conntrack xt_connmark xt_connbytes xt_REDIREl [ 71.464187] btintel br_netfilter bnep bluetooth ath9k_hw ath10k_pci ath10k_core ath sch_tbf sch_ingress sch_htb sch_hfsc em_u32 cls_u32 cls_tcindex cls_route cls_mg [ 71.629589] CPU: 0 PID: 1298 Comm: kworker/u5:3 Not tainted 5.4.114 #0 [ 71.636319] Hardware name: CZ.NIC Turris Mox Board (DT) [ 71.641725] Workqueue: napi_workq napi_workfn [ 71.646221] pstate: 80400085 (Nzcv daIf +PAN -UAO) [ 71.651169] pc : mt76_set_irq_mask+0x118/0x150 [mt76] [ 71.656385] lr : mt7915_init_debugfs+0x358/0x368 [mt7915e] [ 71.662038] sp : ffffffc010003cd0 [ 71.665451] x29: ffffffc010003cd0 x28: 0000000000000060 [ 71.670929] x27: ffffffc010a56f98 x26: ffffffc010c0fa9a [ 71.676407] x25: ffffffc010ba8788 x24: ffffff803e01fe00 [ 71.681885] x23: 0000000000000030 x22: ffffffc010003dc4 [ 71.687361] x21: 0000000000000000 x20: ffffff803e01fea4 [ 71.692839] x19: ffffff803cb725c0 x18: 000000002d660780 [ 71.698317] x17: 0000000000000000 x16: 0000000000000001 [ 71.703795] x15: 0000000000005ee0 x14: ffffffc010d1d000 [ 71.709272] x13: 0000000000002f70 x12: 0000000000000000 [ 71.714749] x11: 0000000000000000 x10: 0000000000000040 [ 71.720226] x9 : ffffffc010bbe980 x8 : ffffffc010bbe978 [ 71.725704] x7 : ffffff803e4003f0 x6 : 0000000000000000 [ 71.731181] x5 : ffffffc02f240000 x4 : ffffffc010003e00 [ 71.736658] x3 : 0000000000000000 x2 : ffffffc008e3f230 [ 71.742135] x1 : 00000000000d7010 x0 : ffffffc0114d7010 [ 71.747613] Call trace: [ 71.750137] mt76_set_irq_mask+0x118/0x150 [mt76] [ 71.754990] mt7915_dual_hif_set_irq_mask+0x108/0xdc0 [mt7915e] [ 71.761098] __handle_irq_event_percpu+0x6c/0x170 [ 71.765950] handle_irq_event_percpu+0x34/0x88 [ 71.770531] handle_irq_event+0x40/0xb0 [ 71.774486] handle_level_irq+0xe0/0x170 [ 71.778530] generic_handle_irq+0x24/0x38 [ 71.782667] advk_pcie_irq_handler+0x11c/0x238 [ 71.787249] __handle_irq_event_percpu+0x6c/0x170 [ 71.792099] handle_irq_event_percpu+0x34/0x88 [ 71.796680] handle_irq_event+0x40/0xb0 [ 71.800633] handle_fasteoi_irq+0xdc/0x190 [ 71.804855] generic_handle_irq+0x24/0x38 [ 71.808988] __handle_domain_irq+0x60/0xb8 [ 71.813213] gic_handle_irq+0x8c/0x198 [ 71.817077] el1_irq+0xf0/0x1c0 [ 71.820314] el1_da+0xc/0xc0 [ 71.823288] mt76_set_irq_mask+0x118/0x150 [mt76] [ 71.828141] mt7915_mac_tx_free+0x4c4/0x828 [mt7915e] [ 71.833352] mt7915_queue_rx_skb+0x5c/0xa8 [mt7915e] [ 71.838473] mt76_dma_cleanup+0x89c/0x1248 [mt76] [ 71.843329] __napi_poll+0x38/0xf8 [ 71.846835] napi_workfn+0x58/0xb0 [ 71.850342] process_one_work+0x1fc/0x390 [ 71.854475] worker_thread+0x48/0x4d0 [ 71.858252] kthread+0x120/0x128 [ 71.861581] ret_from_fork+0x10/0x1c [ 71.865273] Code: 52800000 d65f03c0 f9562c00 8b214000 (b9400000) [ 71.871560] ---[ end trace 1d4e29987011411b ]--- [ 71.876320] Kernel panic - not syncing: Fatal exception in interrupt [ 71.882875] SMP: stopping secondary CPUs [ 71.886923] Kernel Offset: disabled [ 71.890519] CPU features: 0x0002,00002008 [ 71.894649] Memory Limit: none [ 71.897799] Rebooting in 3 seconds.. Patch is awaiting upstream merge: https://lore.kernel.org/linux-pci/[email protected]/T/#u There was also discussion about it in the linux-pci mailing list, where can be found response from Marvell's employee regarding A3720 PCIe erratum 3.12, which seems to provide further details which help this issue: https://lore.kernel.org/linux-pci/BN9PR18MB425154FE5019DCAF2028A1D5DB8D9@BN9PR18MB4251.namprd18.prod.outlook.com/t/#u Reported-by: Ondřej Caletka <[email protected]> [Turris MOX] Signed-off-by: Josef Schlehofer <[email protected]> Reviewed-by: Robert Marko <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Dec 16, 2022
This showed up in the log: |Hunk #1 succeeded at 555 (offset -83 lines). Signed-off-by: Christian Lamparter <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 4, 2023
Removed upstreamed: pending-5.15/101-Use-stddefs.h-instead-of-compiler.h.patch[1] ipq806x/patches-5.15/122-01-clk-qcom-clk-krait-fix-wrong-div2-functions.patch[2] bcm27xx/patches-5.15/950-0198-drm-fourcc-Add-packed-10bit-YUV-4-2-0-format.patch[3] Manually rebased: ramips/patches-5.15/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch[4] Added patch/backported: ramips/patches-5.15/107-PCI-mt7621-Add-sentinel-to-quirks-table.patch[5] All other patches automatically rebased. 1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=c160505c9b574b346031fdf2c649d19e7939ca11 2. Cannot find in the stable tree but it is here: torvalds/linux@a051e10 3. gregkh/linux@ec1727f 4. Quilt gave this output when I applied the patch to rebase it: % quilt push -f Applying patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch patching file arch/mips/ralink/Kconfig patching file drivers/pci/controller/Kconfig patching file drivers/pci/controller/Makefile patching file drivers/staging/Kconfig patching file drivers/staging/Makefile patching file drivers/staging/mt7621-pci/Kconfig patching file drivers/staging/mt7621-pci/Makefile patching file drivers/staging/mt7621-pci/TODO patching file drivers/staging/mt7621-pci/mediatek,mt7621-pci.txt patching file drivers/staging/mt7621-pci/pci-mt7621.c Hunk #1 FAILED at 1. Not deleting file drivers/staging/mt7621-pci/pci-mt7621.c as content differs from patch 1 out of 1 hunk FAILED -- saving rejects to file drivers/staging/mt7621-pci/pci-mt7621.c.rej patching file drivers/pci/controller/pcie-mt7621.c Applied patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch (forced; needs refresh) Upon inspecting drivers/staging/mt7621-pci/pci-mt7621.c.rej, it seems that the original patch wants to delete drivers/staging/mt7621-pci/pci-mt7621.c but upstream's version was not an exact match. I opted to delete that file and need some feedback. Was that the correct course of action? 5. Suggestion by hauke: torvalds/linux@1909893 "This patch is in upstream kernel, but it was backported to the old staging driver in kernel 5.15." Build system: x86_64 Build-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod Run-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod Signed-off-by: John Audia <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 4, 2023
Removed upstreamed: pending-5.15/101-Use-stddefs.h-instead-of-compiler.h.patch[1] ipq806x/patches-5.15/122-01-clk-qcom-clk-krait-fix-wrong-div2-functions.patch[2] bcm27xx/patches-5.15/950-0198-drm-fourcc-Add-packed-10bit-YUV-4-2-0-format.patch[3] Manually rebased: ramips/patches-5.15/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch[4] Added patch/backported: ramips/patches-5.15/107-PCI-mt7621-Add-sentinel-to-quirks-table.patch[5] All other patches automatically rebased. 1. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=c160505c9b574b346031fdf2c649d19e7939ca11 2. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=a051e10bfc6906d29dae7a31f0773f2702edfe1b 3. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v5.15.86&id=ec1727f89ecd6f2252c0c75e200058819f7ce47a 4. Quilt gave this output when I applied the patch to rebase it: % quilt push -f Applying patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch patching file arch/mips/ralink/Kconfig patching file drivers/pci/controller/Kconfig patching file drivers/pci/controller/Makefile patching file drivers/staging/Kconfig patching file drivers/staging/Makefile patching file drivers/staging/mt7621-pci/Kconfig patching file drivers/staging/mt7621-pci/Makefile patching file drivers/staging/mt7621-pci/TODO patching file drivers/staging/mt7621-pci/mediatek,mt7621-pci.txt patching file drivers/staging/mt7621-pci/pci-mt7621.c Hunk #1 FAILED at 1. Not deleting file drivers/staging/mt7621-pci/pci-mt7621.c as content differs from patch 1 out of 1 hunk FAILED -- saving rejects to file drivers/staging/mt7621-pci/pci-mt7621.c.rej patching file drivers/pci/controller/pcie-mt7621.c Applied patch platform/100-PCI-mt7621-Add-MediaTek-MT7621-PCIe-host-controller-.patch (forced; needs refresh) Upon inspecting drivers/staging/mt7621-pci/pci-mt7621.c.rej, it seems that the original patch wants to delete drivers/staging/mt7621-pci/pci-mt7621.c but upstream's version was not an exact match. I opted to delete that file. 5. Suggestion by hauke: torvalds/linux@1909893 "This patch is in upstream kernel, but it was backported to the old staging driver in kernel 5.15." Build system: x86_64 Build-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod Run-tested: bcm2711/RPi4B, filogic/xiaomi_redmi-router-ax6000-ubootmod Signed-off-by: John Audia <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
May 20, 2023
Iomega Storcenter ix4-200d is a four-bay SATA NAS powered by a Marvell Kirkwood SoC clocked at 1.2GHz. It has 512MB of RAM and 32MB of flash memory, 3x USB 2.0 and 2x 1Gbit/s NIC Specification: - SoC: Marvell Kirkwood 88F6281 - CPU/Speed: 1200Mhz - Flash size: 32 MiB - RAM: 512MB - LAN: 2x 1Gbit/s - 3x USB 2.0 Notes: - The blue drive LED is triggered by HDD activity, it can not be controlled via GPIO. - The LCD screen requires proprietary code and does not function at this time. - Due to a kernel-related issue with the Marvell 88SE6121 SATA controller, currently only trays numbered #3 and #4 work, #1 and #2 do not. [1] Serial pinout: CN4 -------------- | 10 8 6 4 2 | | 9 7 5 3 1 | -------------- PIN 1 Mark (fat line) 1 = RXD 4 = TXD 6 = GND 9 = 3.3V (not necessary to connect) Installation instructions: 1. download initramfs-uImage and copy into tftp server 2. connect the tftp server to network port #1 3. access uboot environment with serial cable and run setenv mainlineLinux yes setenv arcNumber 1682 setenv console 'console=ttyS0,115200n8' setenv mtdparts 'mtdparts=orion_nand:0x100000@0x000000(u-boot)ro,0x20000@0xA0000(u-boot environment)ro,0x300000@0x100000(kernel),0x1C00000@0x400000(ubi)' setenv bootargs_root 'root=' setenv bootcmd 'setenv bootargs ${console} ${mtdparts} ${bootargs_root}; nand read.e 0x800000 0x100000 0x300000; bootm 0x00800000' saveenv setenv serverip 192.168.1.1 setenv ipaddr 192.168.1.2 tftpboot 0x00800000 [initramfs-uImage filename] bootm 0x00800000 4. connect to LAN on network port #2, log into openwrt and sysupgrade to install into flash [1] https://bugzilla.kernel.org/show_bug.cgi?id=216094 Signed-off-by: Sander van Deijck <[email protected]> (aligned FROM from signed-off. LED+key rename, whitespace removal) Signed-off-by: Christian Lamparter <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 13, 2024
Hardware specification: SoC: MediaTek MT7981B 2x A53 Flash: 16MB NOR RAM: 256MB Ethernet: 2x 10/100/1000 Mbps Switch: MediaTek MT7531AE WiFi: MediaTek MT7976C Button: Reset Power: DC 12V 1A, PoE 802.3af 48V Flash instructions: Option #1 - SSH I was able to SSH into the stock firmware of my device. 1. Attach the router to the network 2. Use scp (-O) to copy the sysupgrade image 3. Connect using SSH and run `sysupgrade -n` Option #2 - U-Boot One way to use the bootloader for flashing is using TFTP: 1. Connect to the router using an ethernet cable 2 Spin up a TFTP server serving the sysupgrade file 3. Open the case and attach a UART 4. Attach power to the router and interrupt the countdown by pressing any key 5. Select option #2 (Upgrade firmware) 6. Enter IP address information and image name 7. Wait patiently Co-Authored-By: Enrique Rodríguez Valencia <[email protected]> Co-Authored-By: Hauke Mehrtens <[email protected]> Signed-off-by: Leon M. Busch-George <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Feb 27, 2024
Recent commit `d08d53346b993431e309797d5e6de8da20147c3c` qca-ssdk: support selecting PCS channel for PORT3 on IPQ6018 Ends up breaking patch. Updated to address the following: Applying patch 102-qca-ssdk-support-selecting-PCS-channel-for-PORT3-on-.patch patching file include/init/ssdk_dts.h Hunk #1 succeeded at 99 (offset -2 lines). Hunk #2 succeeded at 161 (offset -3 lines). patching file src/adpt/cppe/adpt_cppe_portctrl.c patching file src/adpt/hppe/adpt_hppe_uniphy.c Hunk #1 succeeded at 1152 (offset 30 lines). Hunk #2 succeeded at 1161 (offset 30 lines). patching file src/init/ssdk_dts.c Hunk #1 FAILED at 279. Hunk #2 succeeded at 305 (offset -8 lines). Hunk #3 succeeded at 1310 (offset -16 lines). 1 out of 3 hunks FAILED -- rejects in file src/init/ssdk_dts.c Patch 102-qca-ssdk-support-selecting-PCS-channel-for-PORT3-on-.patch does not apply (enforce with -f)
rmandrad
pushed a commit
that referenced
this pull request
Mar 12, 2024
Currently, the existing uncompressed kallsym support is causing qualcommax boards to hang on boot, and only after earlycon and verbose BUG() prints are enabled the trace is visible: [ 0.000000] ------------[ cut here ]------------ [ 0.000000] kernel BUG at kernel/kallsyms.c:340! [ 0.000000] Internal error: Oops - BUG: 00000000f2000800 [#1] SMP Felix has fixed up the uncompressed kallsyms support so modify the current patch with the fix. All credits for the code go to Felix. Signed-off-by: Robert Marko <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Mar 13, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
rmandrad
pushed a commit
that referenced
this pull request
Mar 29, 2024
Don't add KERNEL_TESTING_PATCHVER yet as there are some issues with it. On TP-Link Archer C2300 serial console seems to stop working after preinit: > Press the [f] key and hit [enter] to enter failsafe mode > Press the [1], [2], [3] or [4] key and hit [enter] to select the de� On Netgear R8000P XHCI causes external abort: [ 2.139586] Internal error: synchronous external abort: 0000000096000210 [#1] SMP [ 2.147212] Modules linked in: xhci_plat_hcd(+) xhci_hcd ohci_platform ohci_hcd fsl_mph_dr_of ehci_platform ehci_fsl ehci_hcd gpio_button_hotplug(O) usbcore nls_base usb_common crc32c_generic [ 2.164774] CPU: 0 PID: 358 Comm: kmodloader Tainted: G O 6.6.22 #0 [ 2.172658] Hardware name: Netgear R8000P (DT) [ 2.177229] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 2.184395] pc : xhci_gen_setup+0x80/0x34c [xhci_hcd] [ 2.189591] lr : xhci_gen_setup+0x74/0x34c [xhci_hcd] [ 2.194788] sp : ffffffc0815d37b0 [ 2.198194] x29: ffffffc0815d37b0 x28: ffffff8002000000 x27: ffffff8001172d88 [ 2.205540] x26: ffffff8002000000 x25: 0000000000000000 x24: ffffffc078b603c0 [ 2.212888] x23: ffffffc078b2a008 x22: ffffff8001172c10 x21: ffffff8002000000 [ 2.220235] x20: ffffff8002000000 x19: ffffff8002000250 x18: 0000000000000000 [ 2.227582] x17: 626d756e20737562 x16: 2064656e67697373 x15: ffffffffffffffff [ 2.234929] x14: ffffff80019e9915 x13: ffffff80019e9913 x12: 00000000ffffffea [ 2.242276] x11: 00000000ffffefff x10: 0000000000000062 x9 : 00000000ffffffd0 [ 2.246760] bcm63138_nand ff801800.nand-controller: timeout waiting for command 0x4 [ 2.249623] x8 : 0000000000000073 x7 : ffffffc0815d37c0 x6 : 0000000000000075 [ 2.257513] bcm63138_nand ff801800.nand-controller: intfc status c00000e0 [ 2.264855] x5 : 0000000000000081 x4 : 0000000000000000 x3 : ffffff8001b61800 [ 2.279193] x2 : ffffffc080b5d000 x1 : ffffff80020003a8 x0 : ffffff8002000398 [ 2.286540] Call trace: [ 2.289048] xhci_gen_setup+0x80/0x34c [xhci_hcd] Signed-off-by: Rafał Miłecki <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
May 6, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jun 21, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jun 23, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jun 24, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 3, 2024
If this function is marked as __init, kernel will splat when driver is (re)bind echo "c080000.ethernet">/sys/bus/platform/drivers/ipqess-edma/unbind echo "c080000.ethernet">/sys/bus/platform/drivers/ipqess-edma/bind Example with additional print messages, functions at boot: [ 2.039468] ipqess-edma c080000.ethernet: ipqess_axi_probe pre register_netdev [ 2.039530] ipqess-edma c080000.ethernet: *netdev: c27d2000 [ 2.045609] ipqess-edma c080000.ethernet: &ipqess_init: c0d1e28c [ 2.051122] ipqess_init [ 2.057338] netdev: c27d2000 [ 2.059492] ess = netdev_priv: c27d2500 [ 2.062615] ess->pdev: c2138c00 [ 2.066174] ess->pdev->dev: c2138c10 [ 2.069314] ess->pdev->dev.of_node: ef6f6368 [ 2.073120] ess->pdev->dev.of_node: /soc/ethernet@c080000 fails (bind) after unbind: [ 34.987948] ipqess-edma c080000.ethernet: ipqess_axi_probe pre register_netdev [ 34.988012] ipqess-edma c080000.ethernet: *netdev: c27d6000 [ 34.994088] ipqess-edma c080000.ethernet: &ipqess_init: c0d1e28c [ 34.999652] 8<--- cut here --- [ 35.005802] Unable to handle kernel paging request at virtual address c0d1e28c when execute [ 35.008676] [c0d1e28c] *pgd=80c1941e(bad) [ 35.016918] Internal error: Oops: 8000000d [#1] SMP ARM Signed-off-by: John Thomson <[email protected]> Link: openwrt#15831 Signed-off-by: Robert Marko <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jul 3, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 4, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 11, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 14, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 17, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 23, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 26, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Jul 28, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Aug 1, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Aug 9, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Sep 11, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Sep 16, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Sep 21, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory.
On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in
memory usage was observed as well consistent sustained RX/TX speeds.
BEFORE:
echo 3 >! /proc/sys/vm/drop_caches
free -m
total used free shared buff/cache available
Mem: 867 338 547 90 101 528
Swap: 0 0 0
AFTER:
total used free shared buff/cache available
Mem: 867 242 594 1 81 624
Swap: 0 0 0
NOTE:
For 512MB platforms, users need to test with the following scenarios,
as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is
really only testable on platforms with 512M or less RAM.
1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system
crashes on boot.
2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G'
3.) Remove patches
999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch
999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch
And re-test with #1 and #2
It was incorrectly assumed that setting a 512M for 1G platforms would save
memory, instead it needs to be explicitly set to know proper memory
regions, otherwise it would cause fw crash.
ath11k_nss: fix typo in 512M memory profile
ath11k_nss: remove SFE patch 718-e-mac80211-Deliver-the-frame-to-driver-tx-ops-directly
It is not relevant to NSS builds and only meant for SFE.
ath11k_nss: remove unecessary patches
Color collision should be left on by default, as it's a primary feature
of 802.11AX.
ath11k_nss: fix spacing
ath11k_nss: Remove unnecessary TKIP bloat
Remove TKIP patches that are not being used as 99% of folks are running
modern encryption (AES-CCMP,SAE,etc).
ath11k_nss: parameterize DP_RXDMA_REFILL_RING_SIZE memory profile
ath11k_nss: Remove SFE related code
Cleanup SFE (shortcut fe) related code as we're not using it on NSS
ath11k_nss: idr, ampdu, and skb headroom check optimizations
ath11k_nss: get valid last_rate for rx_bitrate from cpu stats
ath11k_nss: Fix BCCA counter for EMA
Currently BCCA counter is updated to FW via csa counter offs and
beacon with new countdown is updated for every beacon tx completion event.
For EMA, all EMA beacons are updated in one shot, and counter update for
every tx event will mess up the actual sequence of countdown sent over the air.
Allow FW to update the countdown till 1 and finalize the color
change.
ath11k_nss: Fix compile for TRACE feature
rmandrad
pushed a commit
that referenced
this pull request
Oct 1, 2024
Currently, when removing ath11k or any other driver that uses dummy netdev kernel will crash with: [ 365.004961] ------------[ cut here ]------------ [ 365.004992] kernel BUG at net/core/dev.c:10979! [ 365.008642] Internal error: Oops - BUG: 00000000f2000800 [#1] SMP [ 365.012898] Modules linked in: pppoe ppp_async nft_fib_inet nf_flow_table_inet ath11k_ahb(O) ath11k(O) pppox ppp_generic nft_reject_ipv6 nft_reject_ipv4 nft_reject_inet nft_reject nft_redir nft_quota nft_numgen nft_nat nft_masq nft_log nft_limit nft_hash nft_flow_offload nft_fib_c [ 365.064794] CPU: 3 PID: 3896 Comm: procd Tainted: G O 6.6.52 #0 [ 365.087031] Hardware name: QNAP 301w (DT) [ 365.094058] pstate: 20400005 (nzCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 365.098229] pc : free_netdev+0x164/0x1a0 [ 365.104994] lr : free_netdev+0xec/0x1a0 [ 365.109159] sp : ffffffc081d33b90 [ 365.112718] x29: ffffffc081d33b90 x28: ffffff80039d3c00 x27: ffffff800307f000 [ 365.116199] x26: ffffff800538a000 x25: ffffff8005388000 x24: ffffff800538a4c0 [ 365.123317] x23: 00000000000023a4 x22: ffffff8005388f68 x21: ffffff8004e37050 [ 365.130434] x20: ffffff8004e37000 x19: ffffff8004e36ee8 x18: 0000000000000152 [ 365.137552] x17: 63775f357636712e x16: 3030303030646320 x15: ffffffc081436e20 [ 365.144670] x14: ffffff80033f261c x13: 00000000000000d9 x12: 0000000000000002 [ 365.151789] x11: 0000000000000000 x10: 00000000000008a0 x9 : ffffffc081d33980 [ 365.158906] x8 : ffffff80039d4500 x7 : ffffff803fdda6c0 x6 : 0000000000000007 [ 365.166025] x5 : 0000000000000000 x4 : 00000000000000c1 x3 : 00000000000001f4 [ 365.173143] x2 : ffffff803fdd4b78 x1 : ffffff8004e37050 x0 : 0000000000000005 [ 365.180261] Call trace: [ 365.187370] free_netdev+0x164/0x1a0 [ 365.189630] 0xffffffc079b373f0 [ 365.193447] 0xffffffc079b374c0 [ 365.196311] platform_shutdown+0x24/0x34 [ 365.199438] device_shutdown+0x160/0x268 [ 365.203605] kernel_restart+0x40/0xc0 [ 365.207510] __do_sys_reboot+0x104/0x220 [ 365.211070] __arm64_sys_reboot+0x24/0x30 [ 365.215063] invoke_syscall.constprop.0+0x5c/0x108 [ 365.218971] do_el0_svc+0x40/0xc8 [ 365.223655] el0_svc+0x30/0xb8 [ 365.227041] el0t_64_sync_handler+0x120/0x12c [ 365.229995] el0t_64_sync+0x178/0x17c [ 365.234424] Code: f94013f5 a8c37bfd d50323bf d65f03c0 (d4210000) [ 365.238072] ---[ end trace 0000000000000000 ]--- Wireless backports include alloc_netdev_dummy() backport but they dont have the required free_netdev change first, so backport it. Fixes: openwrt#16531 Fixes: 384d079 ("mac80211: update to version 6.11") Link: openwrt#16549 Signed-off-by: Robert Marko <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Oct 26, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Nov 17, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Nov 21, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Nov 29, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Dec 9, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Dec 16, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Dec 28, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Dec 30, 2024
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 4, 2025
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 5, 2025
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 7, 2025
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
rmandrad
pushed a commit
that referenced
this pull request
Jan 10, 2025
QSDK NSS builds utilize skbuff recycling for better handling of memory. On a Dynalink DL-WRX36 (pbuf script should be set to 'auto') a significant drop in memory usage was observed as well consistent sustained RX/TX speeds. BEFORE: echo 3 >! /proc/sys/vm/drop_caches free -m total used free shared buff/cache available Mem: 867 338 547 90 101 528 Swap: 0 0 0 AFTER: total used free shared buff/cache available Mem: 867 242 594 1 81 624 Swap: 0 0 0 NOTE: For 512MB platforms, users need to test with the following scenarios, as the patch `999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch` is really only testable on platforms with 512M or less RAM. 1.) Explicitly setting 'ATH11K_MEM_PROFILE_512M' on and see if system crashes on boot. 2.) Explicitly setting 'ATH11K_MEM_PROFILE_1G' 3.) Remove patches 999-233-ath11k-Disable-rx_header-tlv-for-2K-SKB.patch 999-311-ath11k-configure-nss-thread-priority-during-pdev_ini.patch And re-test with #1 and #2 It was incorrectly assumed that setting a 512M for 1G platforms would save memory, instead it needs to be explicitly set to know proper memory regions, otherwise it would cause fw crash. Signed-off-by: Sean Khan <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Thanks for your contribution to OpenWrt!
To help keep the codebase consistent and readable,
and to help people review your contribution,
we ask you to follow the rules you find in the wiki at this link
https://openwrt.org/submitting-patches
Please remove this message before posting the pull request.