oweals/openwrt.git
5 years agoramips: Derive DTS name from device name in Makefile
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:22 +0000 (23:22 +0200)]
ramips: Derive DTS name from device name in Makefile

This will automatically derive the DTS name as in ath79 and thus
makes specifying DTS for every device obsolete.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
---

This patch only introduces the mechanism and is then followed by
commits with renames and Makefile adjustments per subtarget.

Eventually, those can be all squashed into a single commit or left
as they are to enhance overview.

5 years agoramips/rt305x: Synchronize Makefiles with DTS compatible
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:19 +0000 (23:22 +0200)]
ramips/rt305x: Synchronize Makefiles with DTS compatible

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoramips/rt288x,rt3883: Synchronize Makefiles with DTS compatible
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:16 +0000 (23:22 +0200)]
ramips/rt288x,rt3883: Synchronize Makefiles with DTS compatible

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoramips/mt76x8: Synchronize Makefiles with DTS compatible
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:13 +0000 (23:22 +0200)]
ramips/mt76x8: Synchronize Makefiles with DTS compatible

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoramips/mt7621: Synchronize Makefiles with DTS compatible
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:08 +0000 (23:22 +0200)]
ramips/mt7621: Synchronize Makefiles with DTS compatible

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoramips/mt7620: Synchronize Makefiles with DTS compatible
Adrian Schmutzler [Wed, 3 Jul 2019 21:22:02 +0000 (23:22 +0200)]
ramips/mt7620: Synchronize Makefiles with DTS compatible

This will "rename" devices in Makefiles to the pattern used in
DTS compatible. This will systematize naming of devices
enormously.

As device names are used to for default SUPPORTED_DEVICES entries,
we need to adjust the source for /tmp/sysinfo/board_name, too.
So remove relevant entries from base-files/lib/ramips.sh and
use device compatible for that.

Despite that, base-files are updated, too.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoramips: Improve compatible for TP-Link Archer devices
Adrian Schmutzler [Wed, 3 Jul 2019 21:21:54 +0000 (23:21 +0200)]
ramips: Improve compatible for TP-Link Archer devices

Include "Archer" in compatible as it is part of the device name.

Update Makefile device names where necessary to match compatible.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoath79: fix missing DEVICE_VENDOR for some devices
Chen Minqiang [Sun, 7 Jul 2019 08:02:49 +0000 (16:02 +0800)]
ath79: fix missing DEVICE_VENDOR for some devices

Signed-off-by: Chen Minqiang <ptpt52@gmail.com>
5 years agoramips: add missing DEVICE_VENDOR
Daniel Golle [Wed, 10 Jul 2019 15:15:49 +0000 (17:15 +0200)]
ramips: add missing DEVICE_VENDOR

Fixes commit ee66fe4ea9a6f5e3f47a063f850bac57e1fb411a
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
5 years agosunxi: convert to multi-variable device title
Luis Araneda [Sat, 6 Jul 2019 02:26:00 +0000 (22:26 -0400)]
sunxi: convert to multi-variable device title

Use data from the linux-sunxi community wiki to
fill-in missing device information

Signed-off-by: Luis Araneda <luaraneda@gmail.com>
5 years agoipq40xx: convert to multi-variable device title
Luis Araneda [Sat, 6 Jul 2019 04:50:32 +0000 (00:50 -0400)]
ipq40xx: convert to multi-variable device title

Additionally, fix some vendor and board names

Signed-off-by: Luis Araneda <luaraneda@gmail.com>
5 years agokirkwood: split up DEVICE_TITLE
Moritz Warning [Sun, 7 Jul 2019 13:17:31 +0000 (15:17 +0200)]
kirkwood: split up DEVICE_TITLE

DEVICE_TITLE is split up into DEVICE_VENDOR, DEVICE_MODEL and DEVICE_VARIANT

Signed-off-by: Moritz Warning <moritzwarning@web.de>
5 years agoath25: split up DEVICE_TITLE
Moritz Warning [Sun, 7 Jul 2019 13:17:11 +0000 (15:17 +0200)]
ath25: split up DEVICE_TITLE

DEVICE_TITLE is split up into DEVICE_VENDOR, DEVICE_MODEL and DEVICE_VARIANT

Signed-off-by: Moritz Warning <moritzwarning@web.de>
5 years agooxnas: convert DEVICE_TITLE to new variables
Daniel Golle [Sat, 6 Jul 2019 19:21:34 +0000 (21:21 +0200)]
oxnas: convert DEVICE_TITLE to new variables

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
5 years agokernel: cherry pick patch removing __linux__ check
Fredrik Olofsson [Thu, 4 Jul 2019 11:58:15 +0000 (13:58 +0200)]
kernel: cherry pick patch removing __linux__ check

This is already included in newer upstream. Needed to build BPF programs
using the MIPS kernel include files.

Without this patch, clang fails with "#error Use a Linux compiler or
give up." in sgidefs.h when building BPF programs.

Signed-off-by: Fredrik Olofsson <fredrik.olofsson@anyfinetworks.com>
5 years agoinclude/package.mk: Add support for src-checkout/ folder
John Crispin [Wed, 10 Jul 2019 11:32:24 +0000 (13:32 +0200)]
include/package.mk: Add support for src-checkout/ folder

This feature is similar to the SRC_TREE_OVERRIDE. However instead of having
to manually create a symlink inside the package folder, the buildsystem
will check if $(TOP_DIR)/src-checkout/$(PKG_NAME)/.git is present and use
this instead.

Signed-off-by: John Crispin <john@phrozen.org>
5 years agoath79: Fix whitespace errors in 11-ath10k-caldata
Adrian Schmutzler [Tue, 9 Jul 2019 11:05:46 +0000 (13:05 +0200)]
ath79: Fix whitespace errors in 11-ath10k-caldata

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agobase-files: Fix path check in get_mac_binary
Adrian Schmutzler [Mon, 8 Jul 2019 08:17:28 +0000 (10:17 +0200)]
base-files: Fix path check in get_mac_binary

Logic was inverted when changing from string check to file check.
Fix it.

Fixes: 8592602d0a88 ("base-files: Really check path in get_mac_binary")
Reported-by: Matthias Schiffer <mschiffer@universe-factory.net>
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agorb532: Fix missing DEVICE_TITLE
Adrian Schmutzler [Mon, 8 Jul 2019 21:14:03 +0000 (23:14 +0200)]
rb532: Fix missing DEVICE_TITLE

This fixes the collection of target info during build, where the
device title for rb532/nand is just empty:
tmp/.config-target.in:44925:warning: leading whitespace ignored
tmp/.config-target.in:123738:warning: leading whitespace ignored
tmp/.config-target.in:123786:warning: leading whitespace ignored

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agokernel: bump 4.19 to 4.19.57
Koen Vandeputte [Tue, 9 Jul 2019 10:43:23 +0000 (12:43 +0200)]
kernel: bump 4.19 to 4.19.57

Refreshed all patches.

Compile-tested on: cns3xxx
Runtime-tested on: cns3xxx

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
5 years agokernel: bump 4.14 to 4.14.132
Koen Vandeputte [Tue, 9 Jul 2019 10:18:49 +0000 (12:18 +0200)]
kernel: bump 4.14 to 4.14.132

Refreshed all patches.

Compile-tested on: cns3xxx
Runtime-tested on: cns3xxx

Signed-off-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
5 years agonetsupport: move out sch_cake from kmod-sched
Kevin Darbyshire-Bryant [Tue, 9 Jul 2019 12:40:32 +0000 (13:40 +0100)]
netsupport: move out sch_cake from kmod-sched

Fix file installation clash between kmod-sched & kmod-sched-cake as both
try to install sch_cake.ko

Remove cake from kmod-sched package as cake is supposed to be the
optional qdisc.

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
5 years agokernel: x86: enable NVME support for 64 bit target
Lucian Cristian [Sun, 30 Jun 2019 06:31:15 +0000 (09:31 +0300)]
kernel: x86: enable NVME support for 64 bit target

Newer motherboards support NVMe U.2 connectors but also NVMe storage can
be used on add-on pcie adapters.
Enable kernel driver on x86 for now so the NVMe flash storage can be used

Signed-off-by: Lucian Cristian <lucian.cristian@gmail.com>
5 years agoramips: use upstream RAW_APPENDED_DTB instead of our OWRTDTB
Chuanhong Guo [Thu, 27 Jun 2019 12:44:27 +0000 (20:44 +0800)]
ramips: use upstream RAW_APPENDED_DTB instead of our OWRTDTB

Upstream kernel added support for RAW_APPENDED_DTB on ralink arch
in the following commit:
02564fc89d3d ("ralink: Introduce fw_passed_dtb to arch/mips/ralink")

Use upstream solution and get rid of our OWRTDTB hack.
This commit set DEVICE_DTS to $$(DTS) instead of replacing DTS with
DEVICE_DTS in device profile because DTS variable will be dropped
in later commits.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
[Tested on mt7621/mt76x8]
Tested-by: Chuanhong Guo <gch981213@gmail.com>
[Tested on rt305x/mt7620]
Tested-by: INAGAKI Hiroshi <musashino.open@gmail.com>
5 years agousbreset: Add missing header
Rosen Penev [Sun, 7 Jul 2019 21:30:17 +0000 (14:30 -0700)]
usbreset: Add missing header

Fixes undefined reference to strtoul

Signed-off-by: Rosen Penev <rosenp@gmail.com>
5 years agonetsupport: move out mqprio from kmod-sched
Konstantin Demin [Fri, 28 Jun 2019 18:30:50 +0000 (21:30 +0300)]
netsupport: move out mqprio from kmod-sched

Currently, there's unable to install "kmod-sched-mqprio" after
"kmod-sched" (or vice versa), because "sch_mqprio.ko" is
shipped in both packages.

Fixes: f83522fa6361 ("linux: Add kmod-sched-mqprio")
Fixes: 6af639e0bf78 ("linux: Add kmod-sched-act-vlan")
Fixes: 72c7e2dc467c ("linux: Add kmod-sched-flower")
Signed-off-by: Konstantin Demin <rockdrilla@gmail.com>
[Add cls_flower and act_vlan]
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5 years agoutil-linux: Update to 2.34
Daniel Engberg [Sat, 15 Jun 2019 10:20:44 +0000 (10:20 +0000)]
util-linux: Update to 2.34

Update util-linux to 2.34
Refresh patch

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
5 years agotools/gengetopt: Update to 2.23
Daniel Engberg [Sat, 15 Jun 2019 09:07:53 +0000 (09:07 +0000)]
tools/gengetopt: Update to 2.23

Update gengetopt to 2.23
Remove 100-dependency_fix.patch, fixed upstream
Rework no tests and docs patch
Fix typo in copyright section

Signed-off-by: Daniel Engberg <daniel.engberg.lists@pyret.net>
5 years agoath79: Add missing IMAGE_SIZE for Etactica EG200
Adrian Schmutzler [Thu, 4 Jul 2019 11:54:18 +0000 (13:54 +0200)]
ath79: Add missing IMAGE_SIZE for Etactica EG200

The Etactica EG200 is the only device in ath79 despite nand
target that lacks IMAGE_SIZE.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agozynq: convert to multi-variable device title
Luis Araneda [Thu, 4 Jul 2019 23:35:38 +0000 (19:35 -0400)]
zynq: convert to multi-variable device title

Use the newly introduced variables to better differentiate
between vendor, model and revision

Signed-off-by: Luis Araneda <luaraneda@gmail.com>
5 years agoiperf3: update to 3.7
DENG Qingfang [Fri, 5 Jul 2019 18:42:23 +0000 (02:42 +0800)]
iperf3: update to 3.7

Signed-off-by: DENG Qingfang <dengqf6@mail2.sysu.edu.cn>
5 years agokmod-sched-cake: drop out of tree package, use kernel version
Kevin Darbyshire-Bryant [Wed, 3 Jul 2019 08:22:55 +0000 (09:22 +0100)]
kmod-sched-cake: drop out of tree package, use kernel version

CAKE made it to kernel 4.19 and since OpenWrt now at kernel 4.19 we can
drop the out of tree cake package in base repository.

Add kmod-sched-cake to netsupport so package dependencies are still met.
Similarly CAKE is retained as an optional qdisc module to avoid base
scheduler package size implications.

Backport upstream patches from k5.1 to address some small bugs and
support fwmark usage.

Signed-off-by: Kevin Darbyshire-Bryant <ldir@darbyshire-bryant.me.uk>
5 years agobase-files: drop support for NAND upgrade in platform_pre_upgrade()
Rafał Miłecki [Tue, 2 Jul 2019 14:10:57 +0000 (16:10 +0200)]
base-files: drop support for NAND upgrade in platform_pre_upgrade()

With bcm53xx switched to the new procedure there is no more need for
keeping that backward compatibility code.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
5 years agobcm53xx: sysupgrade: call nand_do_upgrade() from platform_do_upgrade()
Rafał Miłecki [Tue, 2 Jul 2019 14:10:56 +0000 (16:10 +0200)]
bcm53xx: sysupgrade: call nand_do_upgrade() from platform_do_upgrade()

Calling nand_do_upgrade() from platform_pre_upgrade() was obsoleted in
the commit 30f61a34b4cf ("base-files: always use staged sysupgrade").

This change brings bcm53xx in sync with other targets and allows
dropping backward compatibility code. It shouldn't change any user
visible behavior.

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
5 years agobase-files: drop unused jffs2_copy_config()
Rafał Miłecki [Tue, 2 Jul 2019 11:16:08 +0000 (13:16 +0200)]
base-files: drop unused jffs2_copy_config()

Its last usage was dropped back in 2013 in the commit b95bdc8ab56e
("kernel/base-files: clean up old code related to refreshing mtd
partitions, it is no longer used anywhere").

Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
5 years agokernel: Add missing includes mtdsplit_*.c
Hauke Mehrtens [Fri, 28 Jun 2019 19:52:41 +0000 (21:52 +0200)]
kernel: Add missing includes mtdsplit_*.c

This fixes the following compile problem with kernel 4.9 on lantiq:
drivers/mtd/mtdsplit/mtdsplit_uimage.c:244:34: error: array type has incomplete element type 'struct of_device_id'
 static const struct of_device_id mtdsplit_uimage_of_match_table[] = {
                                  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/mtd/mtdsplit/mtdsplit_uimage.c:245:4: error: field name not in record or union initializer
  { .compatible = "denx,uimage" },

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5 years agoath10k-firmware: Fix mirror hash
Hauke Mehrtens [Sun, 7 Jul 2019 11:30:48 +0000 (13:30 +0200)]
ath10k-firmware: Fix mirror hash

Fixes: 7f79882d44ba ("ath10k-firmware: update board-2.bin for community firmwares")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5 years agokernel: Fix MIPS bounds check virt_addr_valid
Hauke Mehrtens [Mon, 17 Jun 2019 18:25:56 +0000 (20:25 +0200)]
kernel: Fix MIPS bounds check virt_addr_valid

This is pending to get into the upstream kernel.
This fixes a bug in the upstream kernel which was added to stable some
time ago.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5 years agoath79: increase kernel partition size for CPE610v1
Andrew Cameron [Sun, 30 Jun 2019 02:45:09 +0000 (21:45 -0500)]
ath79: increase kernel partition size for CPE610v1

Increase the Kernel partition to address the issue discussed here
https://forum.openwrt.org/t/cpe610-v1-sysupgrade-bin-missing-too-big/39637/5
Switch Back to the okli Loader to support increased partition size

Signed-off-by: Andrew Cameron <apcameron@softhome.net>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> [add <> for e-mail]
5 years agozram-swap: Add extra commands for status/compaction
Emil Muratov [Sun, 24 Feb 2019 07:00:03 +0000 (10:00 +0300)]
zram-swap: Add extra commands for status/compaction

This patch adds two new commands:
  zram status - shows memory stats for all zram swaps
  zram compaction - trigger compaction for all zram swaps

Signed-off-by: Emil Muratov <gpm@hotplug.ru>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
5 years agobusybox: enable swapon/off by default to make it consistent with mkswap
Emil Muratov [Thu, 1 Nov 2018 18:53:43 +0000 (21:53 +0300)]
busybox: enable swapon/off by default to make it consistent with mkswap

  No size increase on busybox binary.
  Since busybox mkswap is already enabled by default it seems reasonable
  to enable swapon/off too. For ex. this obsoletes installing block-mount
  dependency for zram-swap.

Signed-off-by: Emil Muratov <gpm@hotplug.ru>
5 years agozram-swap: Add zram compaction and statistics info output
Emil Muratov [Wed, 31 Oct 2018 21:16:58 +0000 (00:16 +0300)]
zram-swap: Add zram compaction and statistics info output

Executing '/etc/init.d/zram start' during runtime (with a swap being already
mounted) triggers zram device compaction and prints out nice stats info about
zram memory usage

Signed-off-by: Emil Muratov <gpm@hotplug.ru>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> [use IEC's MiB unit]
5 years agozram-swap: Fix busybox dependency check
Emil Muratov [Wed, 31 Oct 2018 08:43:49 +0000 (11:43 +0300)]
zram-swap: Fix busybox dependency check

   - fix dependency on BUSYBOX_CONFIG_SWAPONOFF (removed in 84da2a6)
   - add busybox defaults checking (fix zram-swap always installs swap-utils
     and libblkid as dependency, even if busybox includes mkswap by default)

Signed-off-by: Emil Muratov <gpm@hotplug.ru>
5 years agogemini: Add SL93512r missing ImageInfo file
Linus Walleij [Sun, 7 Jul 2019 00:10:42 +0000 (02:10 +0200)]
gemini: Add SL93512r missing ImageInfo file

Somehow I missed to add the magical image cooking file
for the SL93512r. Fix it.

Fixes: 18e2053becb8 ("gemini: Add StorLink SL93512r images")
Reported-by: Christian Lamparter <chunkeey@gmail.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> [added Fixes,
changed subject]

5 years agogemini: Fix device name for StorLink SL93512r
Adrian Schmutzler [Fri, 5 Jul 2019 16:56:06 +0000 (18:56 +0200)]
gemini: Fix device name for StorLink SL93512r

This has been reported by Chen Minqiang (@ptpt52).

Reported-by: Chen Minqiang <ptpt52@gmail.com>
Fixes: 18e2053becb8 ("gemini: Add StorLink SL93512r images")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com> [Added tags]
5 years agolibnftnl: bump to version 1.1.3
Konstantin Demin [Mon, 1 Jul 2019 14:15:50 +0000 (17:15 +0300)]
libnftnl: bump to version 1.1.3

bump ABI version accordingly (thanks to Jo-Philipp Wich).

Signed-off-by: Konstantin Demin <rockdrilla@gmail.com>
5 years agoramips: Fix compatible for YUKAI Engineering BOCCO
Adrian Schmutzler [Thu, 27 Jun 2019 17:27:10 +0000 (19:27 +0200)]
ramips: Fix compatible for YUKAI Engineering BOCCO

Looks like an undetected copy/paste error.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agobase-files: Really check path in get_mac_binary
Adrian Schmutzler [Thu, 4 Jul 2019 21:28:44 +0000 (23:28 +0200)]
base-files: Really check path in get_mac_binary

Currently, path argument is only checked for being not empty.

This changes behavior to actually check whether path exists.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agompc85xx: Use gzip compressed kernel on HiveAP-330
Pawel Dembicki [Thu, 4 Jul 2019 11:17:02 +0000 (13:17 +0200)]
mpc85xx: Use gzip compressed kernel on HiveAP-330

After commit 1e41de2f48 ("mpc85xx: convert TL-WDR4900 v1 to simpleImage")
XZ compression of zImage was enabled. This change exposed a problem with
the HiveAP-330 images, which was fixed by foregoing the compression on
the kernel altogether with commit 98089bb8ba8
("mpc85xx: Use uncompressed kernel on the HiveAP-330").

This patch adds back the gzip compression of the kernel image by
utilizing the generic OpenWRT uImage method instead of relying on
the PowerPC bootwrapper script that did it previously.

Compile-tested: p1020/hiveap-330

Tested-by: Chris Blake <chrisrblake93@gmail.com> [run-tested]
Signed-off-by: Pawel Dembicki <paweldembicki@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[filled in even more text]

5 years agohostapd: adjust removed wolfssl options
Eneas U de Queiroz [Mon, 1 Jul 2019 16:40:01 +0000 (13:40 -0300)]
hostapd: adjust removed wolfssl options

This edjusts the selection of recently removed wolfssl options which
have always been built into the library even in their abscence.
Also remove the selection of libwolfssl itself, allowing the library to
be built as a module.

Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
5 years agowolfssl: update to 4.0.0-stable
Eneas U de Queiroz [Mon, 1 Jul 2019 16:40:00 +0000 (13:40 -0300)]
wolfssl: update to 4.0.0-stable

Removed options that can't be turned off because we're building with
--enable-stunnel, some of which affect hostapd's Config.in.
Adjusted the title of OCSP option, as OCSP itself can't be turned off,
only the stapling part is selectable.
Mark options turned on when wpad support is selected.
Add building options for TLS 1.0, and TLS 1.3.
Add hardware crypto support, which due to a bug, only works when CCM
support is turned off.
Reorganized option conditionals in Makefile.
Add Eneas U de Queiroz as maintainer.

Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
5 years agowolfssl: update to 3.15.7, fix Makefile
Eneas U de Queiroz [Mon, 1 Jul 2019 16:39:59 +0000 (13:39 -0300)]
wolfssl: update to 3.15.7, fix Makefile

This includes a fix for a medium-level potential cache attack with a
variant of Bleichenbacher’s attack.  Patches were refreshed.
Increased FP_MAX_BITS to allow 4096-bit RSA keys.
Fixed poly1305 build option, and some Makefile updates.

Signed-off-by: Eneas U de Queiroz <cotequeiroz@gmail.com>
5 years agoath10k-firmware: update board-2.bin for community firmwares
Christian Lamparter [Sat, 29 Jun 2019 14:56:31 +0000 (16:56 +0200)]
ath10k-firmware: update board-2.bin for community firmwares

This patch updates the board-2.bin for the default
IPQ4019, QCA9984 and QCA9888 ath10k-firmware-xyz-ct
and -ct-htt firmwares.

Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
5 years agotoolchain: Don't force GCC8 on ARC
Rosen Penev [Thu, 20 Jun 2019 19:33:04 +0000 (12:33 -0700)]
toolchain: Don't force GCC8 on ARC

This prevents overriding it to use GCC9.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[changed logic to not fall back on 7.4]

5 years agogdb-arc: Remove
Rosen Penev [Mon, 17 Jun 2019 23:44:12 +0000 (16:44 -0700)]
gdb-arc: Remove

Normal GDB has supported ARC since 8.0

Signed-off-by: Rosen Penev <rosenp@gmail.com>
5 years agogdb: Remove !arc dependency
Rosen Penev [Mon, 17 Jun 2019 23:43:51 +0000 (16:43 -0700)]
gdb: Remove !arc dependency

Supported since 8.0.

Added uClibc-ng patch.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
5 years agokernel: 4.19: Backport usb dwc2 lpm mode hw check
Martin Schiller [Fri, 5 Jul 2019 12:10:07 +0000 (14:10 +0200)]
kernel: 4.19: Backport usb dwc2 lpm mode hw check

This patch backports verbatim the commit from Linux 5.2-rc7 that fixes
the warnings about invalid lpm related parameters on hardware which
don't that.

This is the case for e.g. lantiq xrx200 targets.

Supported only in Linux 4.17 an later.

Signed-off-by: Martin Schiller <ms@dev.tdt.de>
[refresh patches, fix commit title]
Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: fix etop compiler warnings
Mathias Kresin [Thu, 4 Jul 2019 20:43:00 +0000 (22:43 +0200)]
lantiq: fix etop compiler warnings

The following warnings are introduced by our own patches:

 lantiq_etop.c:173:1: warning: no semicolon at end of struct or union
 };
 ^
 lantiq_etop.c: In function 'ltq_etop_change_mtu':
 lantiq_etop.c:725:2: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]
  int max = ETH_HLEN + VLAN_HLEN + new_mtu + ETH_FCS_LEN;
  ^~~

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: use generic node names
Petr Štetiar [Sun, 30 Dec 2018 11:17:25 +0000 (12:17 +0100)]
lantiq: dts: use generic node names

In DTS Checklist[1] we're now demanding proper generic node names, as
the name of a node should reflect the function of the device and use
generic name for that[2]. Everybody seems to be copy&pasting from DTS
files available in the repository today, so let's unify that naming
there as well and provide proper examples.

1. https://openwrt.org/submitting-patches#dts_checklist
2. https://github.com/devicetree-org/devicetree-specification/blob/master/source/devicetree-basics.rst#generic-names-recommendation

Signed-off-by: Petr Štetiar <ynezz@true.cz>
[split out treewide patch, integrated P2812HNUF1 rename patch]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
[merged key and led changes, use generic node name for flash, wifi and
gpio]
Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: drop falcon clock node
Mathias Kresin [Mon, 28 Jan 2019 22:30:10 +0000 (23:30 +0100)]
lantiq: dts: drop falcon clock node

It is pretty ovbious a misuse of the simple-bus binding as no child has
a address.

Luckly the clock node isn't used at all so it is safe to get rid of it.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: drop empty falcon devicetree node
Mathias Kresin [Mon, 28 Jan 2019 22:31:54 +0000 (23:31 +0100)]
lantiq: dts: drop empty falcon devicetree node

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: fix unit address mismatch warning
Mathias Kresin [Sun, 27 Jan 2019 15:59:08 +0000 (16:59 +0100)]
lantiq: dts: fix unit address mismatch warning

Set a unit address matching the reg property for the EASY98000 ethernet
node.

Fixes a unit mismatch devicetree compiler warning.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: move localbus childs to dts
Mathias Kresin [Sun, 27 Jan 2019 15:34:33 +0000 (16:34 +0100)]
lantiq: dts: move localbus childs to dts

The localbus childs are highly board specific and need to be part of the
board dts. Otherwise we run into duplicate unit address issues if a
board has something else (NAND) connected to the bus id.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: fix vmmc unit and register address
Mathias Kresin [Sun, 27 Jan 2019 15:03:00 +0000 (16:03 +0100)]
lantiq: dts: fix vmmc unit and register address

The register address is the same for danube und vr9. The register size
is 0x400 byte for danube and 0x300 byte for vr9.

The change doesn't have an impact as the vmmc driver doesn't use the
devicetree and has hardcoded register offsets.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: use lower case for hex values
Mathias Kresin [Sun, 27 Jan 2019 13:35:28 +0000 (14:35 +0100)]
lantiq: dts: use lower case for hex values

Use only lower case for hex values to keep it consistent.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: drop superfluous address and size cells
Mathias Kresin [Sun, 27 Jan 2019 13:06:06 +0000 (14:06 +0100)]
lantiq: dts: drop superfluous address and size cells

Fixes the following devicetree compiler warning:

  unnecessary #address-cells/#size-cells without "ranges" or child "reg" property

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: move partitions into partition node
Mathias Kresin [Sun, 27 Jan 2019 13:05:38 +0000 (14:05 +0100)]
lantiq: dts: move partitions into partition node

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: add missing size and address cells
Mathias Kresin [Sun, 27 Jan 2019 12:48:03 +0000 (13:48 +0100)]
lantiq: dts: add missing size and address cells

Add size and address cells where missing to fix the following devicetree
compiler warning:

  Warning (avoid_default_addr_size): Relying on default #address-cells value
  Warning (avoid_default_addr_size): Relying on default #size-cells value

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: fix size cells
Mathias Kresin [Sun, 27 Jan 2019 12:42:51 +0000 (13:42 +0100)]
lantiq: dts: fix size cells

SPI nodes only need an address and no size. Drop the size everywhere.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: fix malformed unit addresses
Mathias Kresin [Sun, 27 Jan 2019 12:32:38 +0000 (13:32 +0100)]
lantiq: dts: fix malformed unit addresses

Fixes:

   Warning (unit_address_format): unit name should not have leading "0x"

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: drop superfluous unit address
Mathias Kresin [Sun, 27 Jan 2019 12:21:06 +0000 (13:21 +0100)]
lantiq: dts: drop superfluous unit address

The cpu temperature driver has the register offset hardcode and doesn't
need a reg.

The mdio bus node is only used as a well known name and doesn't need a
reg or unit address.

syscon-reboot doesn't have a unit address or a reg either. The unit name
collides with reset-controller@10 anyway.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: add reg if unit address is set
Mathias Kresin [Sun, 27 Jan 2019 11:23:03 +0000 (12:23 +0100)]
lantiq: dts: add reg if unit address is set

Add the reg property if the node has an unit address. Fixes the
following device tree compiler warning:

  Warning (unit_address_vs_reg): node has a unit name, but no reg property

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: dts: move memory node to board dts
Mathias Kresin [Sun, 27 Jan 2019 11:20:21 +0000 (12:20 +0100)]
lantiq: dts: move memory node to board dts

The memory node is highly board specific and should be only defined in
the board dts.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agoramips: convert DEVICE_TITLE to new variables
Daniel Golle [Sat, 6 Jul 2019 12:51:09 +0000 (14:51 +0200)]
ramips: convert DEVICE_TITLE to new variables

Also unify naming and add details where they were missing.

Signed-off-by: Daniel Golle <daniel@makrotopia.org>
5 years agoath79: Add DEVICE_VENDOR for Ubiquiti Routerstation
Adrian Schmutzler [Fri, 5 Jul 2019 17:03:48 +0000 (19:03 +0200)]
ath79: Add DEVICE_VENDOR for Ubiquiti Routerstation

This has been overlooked.

Reported by: Chen Minqiang (@ptpt52)

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoiproute2: update to 5.1.0
Deng Qingfang [Wed, 3 Jul 2019 05:32:07 +0000 (13:32 +0800)]
iproute2: update to 5.1.0

Update iproute2 to 5.1.0
Remove upstream patch 010-cake-fwmark.patch
Backport a patch to fix struct sysinfo redefinition error

Signed-off-by: Deng Qingfang <dengqf6@mail2.sysu.edu.cn>
5 years agomt76: update to the latest version
Felix Fietkau [Mon, 1 Jul 2019 12:20:17 +0000 (14:20 +0200)]
mt76: update to the latest version

6cafaca mt7603: use READ_ONCE instead of ACCESS_ONCE
9e2e0b8 mt76: round up length on mt76_wr_copy
e378ef1 mt76: mt7615: fix sparse warnings: warning: restricted __le16 degrades to integer
7991dd7 mt76: mt7615: introduce mt7615_regd_notifier
901a4c7 mt76: mt7615: add hw dfs pattern detector support
57c600e mt76: mt7615: do not perform txcalibration before cac is complited
6afc952 mt76: mt7615: add csa support
8919516 mt76: mt7615: add radar pattern test knob to debugfs
3be723c mt76: mt7615: clean up FWDL TXQ during/after firmware upload
47fe37e mt76: mt7615: fall back to sw encryption for unsupported ciphers
bc5e041 mt76: mt7603: enable hardware rate up/down selection
ae760db mt76: mt7615: move mt7615_mcu_set_rates to mac.c
2ae01f7 mt76: mt7615: reset rate index/counters on rate table update
6f98378 mt76: mt7615: sync with mt7603 rate control changes
edbe88e mt76: usb: fix endian in mt76u_copy
f43b622 mt76: usb: remove unneeded {put,get}_unaligned
5e1e5b7 mt76: usb: use full intermediate buffer in mt76u_copy
017d0ff mt76: mt76u: fix typo in mt76u_fill_rx_sg
2c0ccf1 mt76: mt7615: always release sem in mt7615_load_patch
0c6f1a2 mt76: mt7615: introduce mt7615_mcu_send_ram_firmware routine
3dfc1ee mt76: mt7615: fix sparse warnings: incorrect type in assignment (different base types)
9475320 mt76: mt7603: fix sparse warnings: warning: incorrect type in assignment (different base types)
e07451d mt76: mt7615: fix sparse warnings: warning: cast from restricted __le16
b973bef mt7603: do not use tssi-off power value for mt7628

Signed-off-by: Felix Fietkau <nbd@nbd.name>
5 years agoath79: Correctly use DEVICE_VARIANT for NETGEAR WNDR3700
Adrian Schmutzler [Thu, 4 Jul 2019 14:11:24 +0000 (16:11 +0200)]
ath79: Correctly use DEVICE_VARIANT for NETGEAR WNDR3700

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoath79: Move DEVICE_VENDOR for NETGEAR to parent definition
Adrian Schmutzler [Thu, 4 Jul 2019 14:09:15 +0000 (16:09 +0200)]
ath79: Move DEVICE_VENDOR for NETGEAR to parent definition

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoath79: split DEVICE_TITLE in multiple variables
Paul Spooren [Sun, 16 Jun 2019 19:34:46 +0000 (21:34 +0200)]
ath79: split DEVICE_TITLE in multiple variables

This allows the definition of DEVICE_VENDOR, DEVICE_MODEL and
DEVICE_VARIANT. All three are merged together to look the same as the
current DEVICE_TITLE. Also, if DEVICE_TITLE is set it's directly used as
a *fallback* for devices which weren't upgraded yet.

This is based on the work of @sudhanshu16.

The motivation is to create JSON files based on provided metadata,
needing clear differentiation between vendor, model and revision.

Signed-off-by: Paul Spooren <mail@aparcar.org>
5 years agolantiq: drop kernel 4.14 support
Mathias Kresin [Thu, 4 Jul 2019 06:24:00 +0000 (08:24 +0200)]
lantiq: drop kernel 4.14 support

The current state of the kernel 4.14 support is in the openwrt-19.07
branch. No need to keep a not default used kernel in this branch.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: switch to kernel 4.19
Martin Blumenstingl [Sat, 22 Jun 2019 09:45:23 +0000 (11:45 +0200)]
lantiq: switch to kernel 4.19

Use kernel 4.19 by default.

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
5 years agolantiq: 4.19: fix panic when using the DMA API
Martin Blumenstingl [Tue, 2 Jul 2019 05:25:58 +0000 (07:25 +0200)]
lantiq: 4.19: fix panic when using the DMA API

With kernel 4.19 the DMA API relies on a struct device to work properly.
Backport upstream patches which passed a struct device instead of NULL
to the DMA API.

Fixes kernel panics during boot.

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: 4.19: load PCIe driver after PCI driver
Mathias Kresin [Sun, 23 Jun 2019 18:17:08 +0000 (20:17 +0200)]
lantiq: 4.19: load PCIe driver after PCI driver

In case a PCI device is physical present, the Lantiq PCI driver need to
be loaded prior to the Lantiq PCIe driver. Otherwise none of them will
work because updating the BAR 0 of a device attached to the PCIe Bus
will fail.

Till kernel commit 9667bb039bf6 ("MIPS: lantiq: remove unnecessary
of_platform_default_populate call"), the Lantiq PCI driver was probed
early in the boot process via plat_of_setup().

Since the mentioned commit, both drivers are probed via
of_platform_default_populate_init() and they are probed according to
their order in the devicetree source file.

If the Lantiq PCI driver is enabled in the devicetree, defer the probing
og the Lantiq PCIe driver till the Lantiq PCI driver is probed. Use the
presence of a PCU bus as indicater for a loaded Lantiq PCI driver.

This way we don't need to shuffle the PCI/PCIe nodes in the devicetree
source file, as they are already (correctly) ordered based on the unit
address.

Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: 4.19: increase usb reset timeouts
Martin Blumenstingl [Sat, 22 Jun 2019 08:37:45 +0000 (10:37 +0200)]
lantiq: 4.19: increase usb reset timeouts

With kernel 4.19 dwc2 would not want to initialize due to reset
timeouts, while it worked fine with 4.14.

Increase the reset timeouts to 1 second, as it was used by the old
lantiq ifxhcd usb driver.

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: make patches apply
Hauke Mehrtens [Sun, 25 Nov 2018 19:06:39 +0000 (20:06 +0100)]
lantiq: make patches apply

The following patches are dropped because they are now upstreamed:
- 0002-gpio-stp-xway-Implement-get-callback.patch
  upstreamed with commit 5b9b2b5284f819 ("gpio: stp-xway: Implement get
  callback")
- 0027-01-net-phy-intel-xway-add-VR9-version-number.patch
  upstreamed with commit 5b73d9955fb4b0 ("net: phy: intel-xway: add VR9
  version number")
- 0027-02-net-phy-intel-xway-add-VR9-v1.1-phy-ids.patch
  upstreamed with commit f452518c982e57 ("net: phy: intel-xway: add VR9
  v1.1 phy ids")

The following patches were updated:
- 0018-MTD-nand-lots-of-xrx200-fixes.patch
  the mainline driver now resides in drivers/mtd/nand/raw/xway_nand.c
  (instead of drivers/mtd/nand/xway_nand.c)
- 0025-NET-MIPS-lantiq-adds-xrx200-net.patch
  the DMA API now requires a valid device to be passed to all operations
- 0028-NET-lantiq-various-etop-fixes.patch
  the DMA API now requires a valid device to be passed to all operations

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
5 years agolantiq: copy target to kernel 4.19
Hauke Mehrtens [Sun, 25 Nov 2018 18:35:16 +0000 (19:35 +0100)]
lantiq: copy target to kernel 4.19

This just copies the files from the kernel 4.14 specific folders into
the kernel 4.19 specific folder, no changes are done to the files in
this commit.

Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
5 years agolantiq: ltq-tapi: fix compatibility with Linux 4.15+
Martin Blumenstingl [Wed, 3 Jul 2019 18:10:33 +0000 (20:10 +0200)]
lantiq: ltq-tapi: fix compatibility with Linux 4.15+

Linux 4.15 removes the init_timer() API. It's replaced by two functions:
- timer_setup() is used instead of init_timer() and also replaces the
  timer "function" (callback) setup.
- from_timer() has to be used to obtain the use-case specific data from
  a struct timer_list, which is now passed to the timer callback.

Update the timer API to be compatible with Linux 4.15+ so it compiles
with the upcoming Linux 4.19 kernel update.

Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
5 years agobuild: skip kernel stack validation when building on macOS
Felix Fietkau [Wed, 3 Jul 2019 11:29:13 +0000 (13:29 +0200)]
build: skip kernel stack validation when building on macOS

Since we switched to 4.19, the kernel build checks for libelf to decide if
it should build tools for stack validation.

On macOS, this check fails during target/compile, but succeeds during package
build (because of the pkg-config path picking up target libraries).

Signed-off-by: Felix Fietkau <nbd@nbd.name>
5 years agolantiq: enable STP where referenced
Mathias Kresin [Wed, 3 Jul 2019 07:23:35 +0000 (09:23 +0200)]
lantiq: enable STP where referenced

While moving common used parts to dtsi files, the was disabled by
default but not enabled for all boards using the STP.

Fixes: f519fea4c6db ("lantiq: kernel 4.14: cleanup dts files")

Reported-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Mathias Kresin <dev@kresin.me>
Reviewed-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
5 years agoopenvpn: fix handling of list options
Matt Merhar [Sat, 22 Jun 2019 07:05:14 +0000 (03:05 -0400)]
openvpn: fix handling of list options

This addresses an issue where the list option specified in
/etc/config/openvpn i.e. 'tls_cipher' would instead show up in the
generated openvpn-<name>.conf as 'ncp-ciphers'. For context,
'ncp_ciphers' appears after 'tls_cipher' in OPENVPN_LIST from
openvpn.options.

Also, the ordering of the options in the UCI config file is now
preserved when generating the OpenVPN config. The two currently
supported list options deal with cipher preferences.

Signed-off-by: Matt Merhar <mattmerhar@protonmail.com>
5 years agocomgt: add delay option for 3g proto
Florian Eckert [Wed, 26 Jun 2019 10:36:01 +0000 (12:36 +0200)]
comgt: add delay option for 3g proto

All protos for wwan (ncm,qmi,mbim) do have a delay option.
To standardize that add also the missing delay option to the 3g proto.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
5 years agobase-files: supress service restart of umount
Karel Kočí [Sat, 29 Jun 2019 09:10:19 +0000 (11:10 +0200)]
base-files: supress service restart of umount

Restart is in default implemented so it calls stop and start. This is
pretty unsafe to call on umount service. This service should not do
anything on restart the same way as on start. Only use of this service
is on stop.

Signed-off-by: Karel Kočí <cynerd@email.cz>
5 years agouqmi: fix indentation style and boundary
Florian Eckert [Mon, 1 Jul 2019 07:30:26 +0000 (09:30 +0200)]
uqmi: fix indentation style and boundary

Fix indentation style and boundary.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
5 years agouqmi: add mtu config option possibility
Florian Eckert [Thu, 27 Jun 2019 12:42:48 +0000 (14:42 +0200)]
uqmi: add mtu config option possibility

There are mobile carrier who have different MTU size in their network.
With this change it is now possible to configure this with the qmi
proto handler.

Signed-off-by: Florian Eckert <fe@dev.tdt.de>
5 years agoar71xx: fix ath79/rb4xx IRQ initialization on kernel 4.14
W. Michael Petullo [Sun, 30 Jun 2019 00:06:20 +0000 (20:06 -0400)]
ar71xx: fix ath79/rb4xx IRQ initialization on kernel 4.14

Apply the same approach as in commit 3b53d6fdbc24 ("ar71xx: fix pci irq
init on kernel 4.14") to fix IRQ initialization for ath79-based chipsets
on rb4xx.

Ref: PR#2182
Acked-by: Koen Vandeputte <koen.vandeputte@ncentric.com>
Signed-off-by: W. Michael Petullo <mike@flyn.org>
[commit ref fix]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
5 years agoath79: Use -v1 suffix for TP-Link WDR3600/4300
Adrian Schmutzler [Sun, 16 Jun 2019 10:51:57 +0000 (12:51 +0200)]
ath79: Use -v1 suffix for TP-Link WDR3600/4300

In ath79, identifiers tplink_tl-wdr3600 and tplink_tl-wdr4300 have
been used while most other TP-Link devices include the revision.

Although there actually is only one major revision of these
devices, they bear the revision on their bottom (v1.x). TP-Link
also refers to the devices as V1 on its web page.

This patch thus adds -v1 to both so it is more consistent
with other devices and with what you would expect from reading
the on-device sticker and the support pages.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years agoscripts: time.pl: Don't print the time on stderr
Alban Bedel [Sat, 15 Jun 2019 10:18:01 +0000 (12:18 +0200)]
scripts: time.pl: Don't print the time on stderr

Having the build time written on stderr make it appear with V=w
although it is not an error or warning. Just write the time on stdout
to have it part of the build log like all the rest, but not clutter
the output when only warnings and errors should be shown.

Signed-off-by: Alban Bedel <albeu@free.fr>
5 years agolua5.3: stage Lua headers in proper location
Jo-Philipp Wich [Wed, 3 Jul 2019 05:19:41 +0000 (07:19 +0200)]
lua5.3: stage Lua headers in proper location

Fix wrong paths in InstallDev which cause Lua 5.3 headers to be staged
in /usr/include/, overwriting Lua 5.1 headers and leading to widespread
build failures in all Lua related packages.

Fixes: FS#2348
Fixes: 6b161bb8d5 ("lua5.3: package Lua 5.3 version")
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
5 years agowireguard: bump to 0.0.20190702
Jason A. Donenfeld [Tue, 2 Jul 2019 12:42:50 +0000 (14:42 +0200)]
wireguard: bump to 0.0.20190702

* curve25519: not all linkers support bmi2 and adx

This should allow WireGuard to build on older toolchains.

* global: switch to coarse ktime

Our prior use of fast ktime before meant that sometimes, depending on how
broken the motherboard was, we'd wind up calling into the HPET slow path. Here
we move to coarse ktime which is always super speedy. In the process we had to
fix the resolution of the clock, as well as introduce a new interface for it,
landing in 5.3. Older kernels fall back to a fast-enough mechanism based on
jiffies.

https://lore.kernel.org/lkml/tip-e3ff9c3678b4d80e22d2557b68726174578eaf52@git.kernel.org/
https://lore.kernel.org/lkml/20190621203249.3909-3-Jason@zx2c4.com/

* netlink: cast struct over cb->args for type safety

This follow recent upstream changes such as:

https://lore.kernel.org/lkml/20190628144022.31376-1-Jason@zx2c4.com/

* peer: use LIST_HEAD macro

Style nit.

* receive: queue dead packets to napi queue instead of empty rx_queue

This mitigates a WARN_ON being triggered by the workqueue code. It was quite
hard to trigger, except sporadically, or reliably with a PC Engines ALIX, an
extremely slow board with an AMD LX800 that Ryan Whelan of Axatrax was kind
enough to mail me.

Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>