[OpenWrt-Devel] [PATCH] ath79: fix merge/rebase omission for GL-AR300M16 and GL-AR300M-Lite

2019-11-15 Thread Jeff Kletsky
From: Jeff Kletsky Introduction of the ath79/nand images for the GL-AR300M series resulted in a non-bootable set of partition names for the ath79/generic kernels for the GL-AR300M16 and GL-AR300M-Lite. Signed-off-by: Jeff Kletsky --- target/linux/ath79/dts/qca9531_glinet_gl-ar300m-lite.dts

Re: [OpenWrt-Devel] [PATCH 1/2] build: sysupgrade-tar alt-board= for legacy upgrades

2019-11-06 Thread Jeff Kletsky
On 11/6/19 2:47 PM, Daniel Golle wrote: Hi Jeff, On Thu, Oct 24, 2019 at 08:57:52PM -0700, Jeff Kletsky wrote: From: Jeff Kletsky Targets that use nand_do_platform_check() can't use SUPPORTED_DEVICES as the check requires ./sysupgrade-legacy_boardname/CONTROL to be non-zero length

[OpenWrt-Devel] [PATCH] kernel: mtd: spinand: Backport chip definitions

2019-10-24 Thread Jeff Kletsky
From: Jeff Kletsky generic: Add/rename patches for upstream consistency ipq40xx: generic-level patch replaces same-source patches-4.19/ 082-v4.20-mtd-spinand-winbond-Add-support-for-W25N01GV.patch The SPI-NAND framework from Linux uses common driver code that is then "tuned"

[OpenWrt-Devel] [PATCH 0/2] ath79: Prepare NAND subtarget for upstream support of SPI

2019-10-24 Thread Jeff Kletsky
This submission is part of a series of commits that prepares for use of SPI-NAND with the upstream Linux MTD framework for the ath79 target. The entire series of pending changes may be seen on GitHub[1]. The series also addresses upgrade issues associated with tar-style files and metadata,

[OpenWrt-Devel] [PATCH 1/2] ath79: Remove legacy GL.iNet GL-AR300M NAND target

2019-10-24 Thread Jeff Kletsky
From: Jeff Kletsky Remove the sole board on the legacy ath79 NAND target to clarify patches that will reintroduce support on the MTD SPI-NAND framework present in Linux 4.19 Cc: Marty E. Plummer Signed-off-by: Jeff Kletsky --- package/boot/uboot-envtools/files/ath79 | 1 - .../ath79

[OpenWrt-Devel] [PATCH 2/2] ath79: Prepare NAND subtarget for upstream support of SPI NAND

2019-10-24 Thread Jeff Kletsky
From: Jeff Kletsky Linux 4.19 supplies the upstream spi-nand framework, permitting porting and support of boards with SPI NAND. * Adjusted nand/target.mk to provide FEATURES += squashfs nand * Updated config-default to provide current MTD and UBI support Defaults selected

[OpenWrt-Devel] [PATCH 2/2] sysupgrade: NAND: Prefer CONTROL for board, improve robustness

2019-10-24 Thread Jeff Kletsky
From: Jeff Kletsky Improve NAND-based sysupgrade for current and future upgrades nand_upgrade_tar(): * Use the CONTROL file specific to the running board to determine asset directory name in the tar file * Check for string `BOARD=board_name` directly rather than source-ing

[OpenWrt-Devel] [PATCH 1/2] build: sysupgrade-tar alt-board= for legacy upgrades

2019-10-24 Thread Jeff Kletsky
From: Jeff Kletsky Targets that use nand_do_platform_check() can't use SUPPORTED_DEVICES as the check requires ./sysupgrade-legacy_boardname/CONTROL to be non-zero length as extracted from the tar file. Previously, only ./sysupgrade-new_boardname/CONTROL was present. This prevents upgrade

Re: [OpenWrt-Devel] [PATCH v2] ath79: fix several issues for ZyXEL NBG6716

2019-10-29 Thread Jeff Kletsky
a result of a call to mktemp(1), or generically named. Jeff Kletsky ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel

[OpenWrt-Devel] [PATCH v2, 0/1] ath79: Prepare NAND subtarget for upstream support of SPI

2019-10-29 Thread Jeff Kletsky
Supersedes https://patchwork.ozlabs.org/project/openwrt/list/?series=138553 due to need to rebase With commit 758a4d1766 ath79: add AR934x NAND Flash Controller driver cleaning up the ath79 NAND target and the problems with NAND-based sysupgrade, including that it doesn't respect

[OpenWrt-Devel] [PATCH v2 1/1] ath79: Prepare NAND subtarget for upstream support of SPI NAND

2019-10-29 Thread Jeff Kletsky
From: Jeff Kletsky Linux 4.19 supplies the upstream spi-nand framework, permitting porting and support of boards with SPI NAND. * Adjusted nand/target.mk to provide FEATURES += squashfs nand * Updated config-default to provide current MTD and UBI support Defaults selected

[OpenWrt-Devel] build: "bash: GZ_SUFFIX: command not found" -- Likely From Device/Build/image

2019-10-22 Thread Jeff Kletsky
As discovered by "wulfy23"[1], a verbose build (V=s) off `master` indicates     GZ_SUFFIX :=     bash: GZ_SUFFIX: command not found This appears to be "cosmetic", as the images are created, can be flashed, and run as expected. Observed with: * ath79, GL-AR300M (NAND) (1x) * ath79, Archer

[OpenWrt-Devel] [PATCH] ath79: uboot-envtools: Add GL-AR300M-Lite

2019-10-23 Thread Jeff Kletsky
From: Jeff Kletsky Add the GL.iNet GL-AR300M-Lite to the list of supported boards. Signed-off-by: Jeff Kletsky --- package/boot/uboot-envtools/files/ath79 | 1 + 1 file changed, 1 insertion(+) diff --git a/package/boot/uboot-envtools/files/ath79 b/package/boot/uboot-envtools/files/ath79

Re: [OpenWrt-Devel] [PATCH] ath79: split base-files into subtargets

2019-10-23 Thread Jeff Kletsky
and subtarget Note that even a few kB of increased file size can increase the size of the file system by a full JFFS erase block, pushing a device from "supported" to not supportable. Jeff Kletsky [1] https://github.com/openwrt/openwrt/pull/2184 ___

[OpenWrt-Devel] [PATCH] ath79: Refactor GL.iNet GL-AR300M-series generic.mk

2019-10-23 Thread Jeff Kletsky
From: Jeff Kletsky Rework DEVICE_VENDOR, DEVICE_MODEL, and DEVICE_VARIANT for the GL-AR300M series on the ath79-generic target. Changes GL-AR300M-Lite to the current form with DEVICE_VARIANT := Lite (board name is unchanged) Signed-off-by: Jeff Kletsky --- target/linux/ath79/image/generic.mk

[OpenWrt-Devel] [PATCH 2/2] ath79: GL-AR750S: Add I2C Support

2019-10-23 Thread Jeff Kletsky
From: Jeff Kletsky The GL-AR750S has an internal header for I2C. Provide DTS definitions for the i2c-gpio driver. The I2C drivers; kmod-i2c-core, kmod-i2c-gpio consume ~20 kB of flash and can be loaded as modules, Default clock measured ~11.2 ms period, ~89 kHz The board has well-labeled

[OpenWrt-Devel] [PATCH 1/2] ath79: GL-AR300M series: Add I2C Support

2019-10-23 Thread Jeff Kletsky
From: Jeff Kletsky The GL-AR300M series have an internal header for I2C. Provide DTS definitions for the i2c-gpio driver. The I2C drivers; kmod-i2c-core, kmod-i2c-gpio consume ~20 kB of flash and can be loaded as modules, Default clock measured ~11.4 ms period, ~88 kHz The board has two sets

Re: [OpenWrt-Devel] [PATCH] ath79: remove redundant mtd-mac-address for wmac

2019-10-25 Thread Jeff Kletsky
0x1000. Thus, remove the redundant mtd-mac-address for those devices. Signed-off-by: Adrian Schmutzler Tested-by: Jeff Kletsky GL-AR300M-NAND --- Not tested on the devices affected by this patch, but obviously tested quite often with other device support PRs prior to merge. --- target/linux

Re: [OpenWrt-Devel] v5.4 as next kernel / ipq806x

2019-11-01 Thread Jeff Kletsky
On 11/1/19 2:12 AM, Andre Valentin wrote: Hello, I also did several tests on the 4.19 ipq806x (NBG6817). I noticed that VPN throuput (IPsec Performance) dropped to 30% with exactly the same config (kernel and openwrt), kernel crypto and arm crypto stuff tested. Also the whole system feels a

<    1   2