openwrt/target/linux/mvebu/patches-4.14/300-mvneta-tx-queue-workaround.patch
Stijn Segers 1199a91095 kernel: bump 4.14 to 4.14.48 for 18.06
Refreshed patches. The following patches were upstreamed and have been deleted:

* target/linux/lantiq/patches-4.14/0025-MIPS-lantiq-gphy-Remove-reboot-remove-reset-asserts.patch
* target/linux/generic/pending-4.14/101-clocksource-mips-gic-timer-fix-clocksource-counter-w.patch
* target/linux/generic/pending-4.14/103-MIPS-c-r4k-fix-data-corruption-related-to-cache-coherence.patch
* target/linux/generic/pending-4.14/181-net-usb-add-lte-modem-wistron-neweb-d18q1.patch

Compile-tested: ramips/mt7621, x86/64
Run-tested: ramips/mt7621

Signed-off-by: Stijn Segers <foss@volatilesystems.org>
2018-06-07 09:03:24 +02:00

36 lines
1.2 KiB
Diff

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.
Work around this issue by forcing all tx packets to go through one CPU,
until this issue is fixed properly.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/drivers/net/ethernet/marvell/mvneta.c
+++ b/drivers/net/ethernet/marvell/mvneta.c
@@ -3962,6 +3962,15 @@ static int mvneta_ethtool_set_wol(struct
return ret;
}
+static u16 mvneta_select_queue(struct net_device *dev, struct sk_buff *skb,
+ void *accel_priv,
+ select_queue_fallback_t fallback)
+{
+ /* XXX: hardware queue scheduling is broken,
+ * use only one queue until it is fixed */
+ return 0;
+}
+
static const struct net_device_ops mvneta_netdev_ops = {
.ndo_open = mvneta_open,
.ndo_stop = mvneta_stop,
@@ -3972,6 +3981,7 @@ static const struct net_device_ops mvnet
.ndo_fix_features = mvneta_fix_features,
.ndo_get_stats64 = mvneta_get_stats64,
.ndo_do_ioctl = mvneta_ioctl,
+ .ndo_select_queue = mvneta_select_queue,
};
static const struct ethtool_ops mvneta_eth_tool_ops = {