kernel: bump 5.15 to 5.15.116

[graysky2]

All patches rebased automatically.

Build system: x86_64
Build-tested: ramips/tplink_archer-a6-v3, filogic/xiaomi_redmi-router-ax6000-ubootmod
Run-tested: ramips/tplink_archer-a6-v3, filogic/xiaomi_redmi-router-ax6000-ubootmod

[graysky2]

I am getting a build error with ramips/tplink_archer-a6-v3:

make[8]: Entering directory '/scratch/union/build_dir/target-mipsel_24kc_musl/linux-ramips_mt7621/linux-5.15.115'
  CC      drivers/net/dsa/mt7530.o
drivers/net/dsa/mt7530.c: In function 'mt7531_setup':
drivers/net/dsa/mt7530.c:2453:14: error: 'i' undeclared (first use in this function); did you mean 'id'?
 2453 |         for (i = 0; i < MT7530_NUM_PORTS; i++)
      |              ^
      |              id
drivers/net/dsa/mt7530.c:2453:14: note: each undeclared identifier is reported only once for each function it appears in
make[8]: *** [scripts/Makefile.build:289: drivers/net/dsa/mt7530.o] Error 1

[DragonBluep]

https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit?h=v5.15.115&id=5a7266feaa6d708fc6880a161786eaa884ef3c8e
In this patch, upstream version removes the variable i, but our local version still retains it.

[graysky2]

In this patch, upstream version removes the variable i, but our local version still retains it.

@DragonBluep - I don't understand, our local version was removed in this commit.
EDIT: maybe you are saying that I need to grep for this file in all other patches and modify the ones causing this change?

I don't have much time today, so any pointer is appreciated.

[DragonBluep]

What I mean is the upstream merged patch is not exactly the same as our local patch so we can't fully remove it. There is a tiny difference about variable i. We need to add a new patch to redefine variable i.

upstream version:

 mt7531_setup(struct dsa_switch *ds)
 {
 	struct mt7530_priv *priv = ds->priv;
 	struct mt7530_dummy_poll p;
-	struct dsa_port *cpu_dp;
 	u32 val, id;
-	int ret, i;
+	int ret;

local version:

openwrt/target/linux/generic/backport-5.15/790-v6.4-0010-net-dsa-mt7530-split-off-common-parts-from-mt7531_se.patch

Lines 78 to 85 in 0a4b309

	mt7531_setup(struct dsa_switch *ds)
	{
	struct mt7530_priv *priv = ds->priv;
	struct mt7530_dummy_poll p;
	- struct dsa_port *cpu_dp;
	u32 val, id;
	int ret, i;

edit:
or we can modify and define i in this patch:
https://github.com/openwrt/openwrt/blob/main/target/linux/generic/pending-5.15/723-net-mt7531-ensure-all-MACs-are-powered-down-before-r.patch

[graysky2]

Thanks for the suggestions. Built fine and seems to be running fine on ramips/tplink_archer-a6-v3 for me. Can you please review?

I was a bit confused because there are two sections that could fit the lines starting with /* Reset whole chip through gpio pin or memory-mapped registers for ... I placed the block in the one closest to the definition for i as you can see.

[DragonBluep]

Tested on ath79, everything is fine. Most of the changes are related to the mt7531 switch. If mt7621 runs well, there's nothing to worry about.

[Borromini]

After a full day of homeworking, it looks like MT7621 is okay. Was tested on top of 23.05 HEAD but shouldn't matter much yet.

Compile-tested: ath79/generic, ipq40xx, octeon, ramips/mt76{20,21,x8}, realtek/rtl{838,930}x, mvebu/cortexa{9,72}, x86/64.
Run-tested: mt7621 (EAP615-Wall, WAX202), mvebu (RB5009UG+S+IN).

[CharlesMengCA]

Linksys E8450 (UBI) Tested. Working fine.

[graysky2]

@hauke - not sure why the error
https://github.com/openwrt/openwrt/actions/runs/5231935142/jobs/9446653258?pr=12823#step:29:32

This is already set:
https://github.com/openwrt/openwrt/blob/main/target/linux/generic/config-5.15#L1636

[Ansuel]

@graysky2 maybe some kernel modules to update?

[DragonBluep]

@hauke - not sure why the error https://github.com/openwrt/openwrt/actions/runs/5231935142/jobs/9446653258?pr=12823#step:29:32

The missing symbol is CONFIG_DRM_RCAR_USE_LVDS

This is already set: https://github.com/openwrt/openwrt/blob/main/target/linux/generic/config-5.15#L1636
CONFIG_DRM_RCAR_LVDS

The existing one is CONFIG_DRM_RCAR_LVDS

They are different!

[zxlhhyccc]

@hauke- 不知道为什么错误 https://github.com/openwrt/openwrt/actions/runs/5231935142/jobs/9446653258?pr=12823#step:29:32

这已经设置好了： https://github.com/openwrt/openwrt/blob/main/target/linux/generic/config-5.15#L1636

DRM_RCAR_USE_LVDS
CONFIG_DRM_RCAR_LVDS