From: Tomasz Maciej Nowak <tmn...@gmail.com> Because some still unresolved bugs in this driver, which sprout occasional questions what this patch works around, point to the issue which started this. Being here, fill headers required by git am.
Signed-off-by: Tomasz Maciej Nowak <tmn...@gmail.com> --- .../mvebu/patches-5.15/700-mvneta-tx-queue-workaround.patch | 5 +++++ .../mvebu/patches-6.1/700-mvneta-tx-queue-workaround.patch | 5 +++++ 2 files changed, 10 insertions(+) diff --git a/target/linux/mvebu/patches-5.15/700-mvneta-tx-queue-workaround.patch b/target/linux/mvebu/patches-5.15/700-mvneta-tx-queue-workaround.patch index 32e8ef4b7d87..039dbc06dd04 100644 --- a/target/linux/mvebu/patches-5.15/700-mvneta-tx-queue-workaround.patch +++ b/target/linux/mvebu/patches-5.15/700-mvneta-tx-queue-workaround.patch @@ -1,3 +1,6 @@ +From: Felix Fietkau <n...@nbd.name> +Subject: mvneta: tx queue workaround + The hardware queue scheduling is apparently configured with fixed priorities, which creates a nasty fairness issue where traffic from one CPU can starve traffic from all other CPUs. @@ -5,6 +8,8 @@ CPU can starve traffic from all other CPUs. Work around this issue by forcing all tx packets to go through one CPU, until this issue is fixed properly. +Ref: https://github.com/openwrt/openwrt/issues/5411 + Signed-off-by: Felix Fietkau <n...@nbd.name> --- --- a/drivers/net/ethernet/marvell/mvneta.c diff --git a/target/linux/mvebu/patches-6.1/700-mvneta-tx-queue-workaround.patch b/target/linux/mvebu/patches-6.1/700-mvneta-tx-queue-workaround.patch index 15762be81d40..f80d4e301ada 100644 --- a/target/linux/mvebu/patches-6.1/700-mvneta-tx-queue-workaround.patch +++ b/target/linux/mvebu/patches-6.1/700-mvneta-tx-queue-workaround.patch @@ -1,3 +1,6 @@ +From: Felix Fietkau <n...@nbd.name> +Subject: mvneta: tx queue workaround + The hardware queue scheduling is apparently configured with fixed priorities, which creates a nasty fairness issue where traffic from one CPU can starve traffic from all other CPUs. @@ -5,6 +8,8 @@ CPU can starve traffic from all other CPUs. Work around this issue by forcing all tx packets to go through one CPU, until this issue is fixed properly. +Ref: https://github.com/openwrt/openwrt/issues/5411 + Signed-off-by: Felix Fietkau <n...@nbd.name> --- --- a/drivers/net/ethernet/marvell/mvneta.c -- 2.43.0 _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
