Real-time setups try hard to ensure proper isolation between time critical applications and e.g. network processing performed by the network stack in softirq and RPS is used to move the softirq activity away from the isolated core.
If the network configuration is dynamic, with netns and devices routinely created at run-time, enforcing the correct RPS setting on each newly created device allowing to transient bad configuration became complex.
These series try to address the above, introducing a new sysctl knob: rps_default_mask. The new sysctl entry allows configuring a systemwide RPS mask, to be enforced since receive queue creation time without any fourther per device configuration required.
Additionally, a simple self-test is introduced to check the rps_default_mask behavior.
v1 -> v2: - fix sparse warning in patch 2/3
Paolo Abeni (3): net/sysctl: factor-out netdev_rx_queue_set_rps_mask() helper net/core: introduce default_rps_mask netns attribute self-tests: introduce self-tests for RPS default mask
Documentation/admin-guide/sysctl/net.rst | 6 ++ include/linux/netdevice.h | 1 + net/core/net-sysfs.c | 73 +++++++++++-------- net/core/sysctl_net_core.c | 58 +++++++++++++++ tools/testing/selftests/net/Makefile | 1 + tools/testing/selftests/net/config | 3 + .../testing/selftests/net/rps_default_mask.sh | 57 +++++++++++++++ 7 files changed, 169 insertions(+), 30 deletions(-) create mode 100755 tools/testing/selftests/net/rps_default_mask.sh
Will simplify the following patch. No functional change intended.
Signed-off-by: Paolo Abeni pabeni@redhat.com --- net/core/net-sysfs.c | 66 ++++++++++++++++++++++++-------------------- 1 file changed, 36 insertions(+), 30 deletions(-)
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 94fff0700bdd..b57426707216 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -737,42 +737,18 @@ static ssize_t show_rps_map(struct netdev_rx_queue *queue, char *buf) return len < PAGE_SIZE ? len : -EINVAL; }
-static ssize_t store_rps_map(struct netdev_rx_queue *queue, - const char *buf, size_t len) +static int netdev_rx_queue_set_rps_mask(struct netdev_rx_queue *queue, + cpumask_var_t mask) { - struct rps_map *old_map, *map; - cpumask_var_t mask; - int err, cpu, i, hk_flags; static DEFINE_MUTEX(rps_map_mutex); - - if (!capable(CAP_NET_ADMIN)) - return -EPERM; - - if (!alloc_cpumask_var(&mask, GFP_KERNEL)) - return -ENOMEM; - - err = bitmap_parse(buf, len, cpumask_bits(mask), nr_cpumask_bits); - if (err) { - free_cpumask_var(mask); - return err; - } - - if (!cpumask_empty(mask)) { - hk_flags = HK_FLAG_DOMAIN | HK_FLAG_WQ; - cpumask_and(mask, mask, housekeeping_cpumask(hk_flags)); - if (cpumask_empty(mask)) { - free_cpumask_var(mask); - return -EINVAL; - } - } + struct rps_map *old_map, *map; + int cpu, i;
map = kzalloc(max_t(unsigned int, RPS_MAP_SIZE(cpumask_weight(mask)), L1_CACHE_BYTES), GFP_KERNEL); - if (!map) { - free_cpumask_var(mask); + if (!map) return -ENOMEM; - }
i = 0; for_each_cpu_and(cpu, mask, cpu_online_mask) @@ -799,9 +775,39 @@ static ssize_t store_rps_map(struct netdev_rx_queue *queue,
if (old_map) kfree_rcu(old_map, rcu); + return 0; +} + +static ssize_t store_rps_map(struct netdev_rx_queue *queue, + const char *buf, size_t len) +{ + cpumask_var_t mask; + int err, hk_flags; + + if (!capable(CAP_NET_ADMIN)) + return -EPERM; + + if (!alloc_cpumask_var(&mask, GFP_KERNEL)) + return -ENOMEM; + + err = bitmap_parse(buf, len, cpumask_bits(mask), nr_cpumask_bits); + if (err) + goto out;
+ if (!cpumask_empty(mask)) { + hk_flags = HK_FLAG_DOMAIN | HK_FLAG_WQ; + cpumask_and(mask, mask, housekeeping_cpumask(hk_flags)); + if (cpumask_empty(mask)) { + err = -EINVAL; + goto out; + } + } + + err = netdev_rx_queue_set_rps_mask(queue, mask); + +out: free_cpumask_var(mask); - return len; + return err ? : len; }
static ssize_t show_rps_dev_flow_table_cnt(struct netdev_rx_queue *queue,
If RPS is enabled, this allows configuring a default rps mask, which is effective since receive queue creation time.
A default RPS mask allows the system admin to ensure proper isolation, avoiding races at network namespace or device creation time.
The default RPS mask is initially empty, and can be modified via a newly added sysctl entry.
v1 -> v2: - declare rps_default_mask in netdevice.h to avoid a sparse warning - Jakub
Signed-off-by: Paolo Abeni pabeni@redhat.com --- Documentation/admin-guide/sysctl/net.rst | 6 +++ include/linux/netdevice.h | 1 + net/core/net-sysfs.c | 7 +++ net/core/sysctl_net_core.c | 58 ++++++++++++++++++++++++ 4 files changed, 72 insertions(+)
diff --git a/Documentation/admin-guide/sysctl/net.rst b/Documentation/admin-guide/sysctl/net.rst index 57fd6ce68fe0..818cb2030a8b 100644 --- a/Documentation/admin-guide/sysctl/net.rst +++ b/Documentation/admin-guide/sysctl/net.rst @@ -211,6 +211,12 @@ rmem_max
The maximum receive socket buffer size in bytes.
+rps_default_mask +---------------- + +The default RPS CPU mask used on newly created network devices. An empty +mask means RPS disabled by default. + tstamp_allow_data ----------------- Allow processes to receive tx timestamps looped together with the original diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 964b494b0e8d..2593689648d3 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -200,6 +200,7 @@ struct net_device_stats { #include <linux/static_key.h> extern struct static_key_false rps_needed; extern struct static_key_false rfs_needed; +extern struct cpumask rps_default_mask; #endif
struct neighbour; diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index b57426707216..3f3d1d467fe0 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -983,6 +983,13 @@ static int rx_queue_add_kobject(struct net_device *dev, int index) goto err; }
+#if IS_ENABLED(CONFIG_RPS) && IS_ENABLED(CONFIG_SYSCTL) + if (!cpumask_empty(&rps_default_mask)) { + error = netdev_rx_queue_set_rps_mask(queue, &rps_default_mask); + if (error) + goto err; + } +#endif kobject_uevent(kobj, KOBJ_ADD);
return error; diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index d86d8d11cfe4..13451ac88a74 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -15,6 +15,7 @@ #include <linux/vmalloc.h> #include <linux/init.h> #include <linux/slab.h> +#include <linux/sched/isolation.h>
#include <net/ip.h> #include <net/sock.h> @@ -46,6 +47,54 @@ int sysctl_devconf_inherit_init_net __read_mostly; EXPORT_SYMBOL(sysctl_devconf_inherit_init_net);
#ifdef CONFIG_RPS +struct cpumask rps_default_mask; + +static int rps_default_mask_sysctl(struct ctl_table *table, int write, + void *buffer, size_t *lenp, loff_t *ppos) +{ + int len, err = 0; + + rtnl_lock(); + if (write) { + err = cpumask_parse(buffer, &rps_default_mask); + if (err) + goto done; + + if (!cpumask_empty(&rps_default_mask)) { + int hk_flags = HK_FLAG_DOMAIN | HK_FLAG_WQ; + cpumask_and(&rps_default_mask, &rps_default_mask, + housekeeping_cpumask(hk_flags)); + if (cpumask_empty(&rps_default_mask)) { + err = -EINVAL; + goto done; + } + } + } else { + char kbuf[128]; + + if (*ppos || !*lenp) { + *lenp = 0; + goto done; + } + + len = min(sizeof(kbuf) - 1, *lenp); + len = scnprintf(kbuf, len, "%*pb", cpumask_pr_args(&rps_default_mask)); + if (!len) { + *lenp = 0; + goto done; + } + if (len < *lenp) + kbuf[len++] = '\n'; + memcpy(buffer, kbuf, len); + *lenp = len; + *ppos += len; + } + +done: + rtnl_unlock(); + return err; +} + static int rps_sock_flow_sysctl(struct ctl_table *table, int write, void *buffer, size_t *lenp, loff_t *ppos) { @@ -466,6 +515,11 @@ static struct ctl_table net_core_table[] = { .mode = 0644, .proc_handler = rps_sock_flow_sysctl }, + { + .procname = "rps_default_mask", + .mode = 0644, + .proc_handler = rps_default_mask_sysctl + }, #endif #ifdef CONFIG_NET_FLOW_LIMIT { @@ -648,6 +702,10 @@ static __net_initdata struct pernet_operations sysctl_core_ops = {
static __init int sysctl_core_init(void) { +#if IS_ENABLED(CONFIG_RPS) + cpumask_copy(&rps_default_mask, cpu_none_mask); +#endif + register_net_sysctl(&init_net, "net/core", net_core_table); return register_pernet_subsys(&sysctl_core_ops); }
Ensure that RPS default mask changes take place on all newly created netns/devices and don't affect existing ones.
Signed-off-by: Paolo Abeni pabeni@redhat.com --- tools/testing/selftests/net/Makefile | 1 + tools/testing/selftests/net/config | 3 + .../testing/selftests/net/rps_default_mask.sh | 57 +++++++++++++++++++ 3 files changed, 61 insertions(+) create mode 100755 tools/testing/selftests/net/rps_default_mask.sh
diff --git a/tools/testing/selftests/net/Makefile b/tools/testing/selftests/net/Makefile index ef352477cac6..2531ec3e5027 100644 --- a/tools/testing/selftests/net/Makefile +++ b/tools/testing/selftests/net/Makefile @@ -21,6 +21,7 @@ TEST_PROGS += rxtimestamp.sh TEST_PROGS += devlink_port_split.py TEST_PROGS += drop_monitor_tests.sh TEST_PROGS += vrf_route_leaking.sh +TEST_PROGS += rps_default_mask.sh TEST_PROGS_EXTENDED := in_netns.sh TEST_GEN_FILES = socket nettest TEST_GEN_FILES += psock_fanout psock_tpacket msg_zerocopy reuseport_addr_any diff --git a/tools/testing/selftests/net/config b/tools/testing/selftests/net/config index 4d5df8e1eee7..5d467364f082 100644 --- a/tools/testing/selftests/net/config +++ b/tools/testing/selftests/net/config @@ -34,3 +34,6 @@ CONFIG_TRACEPOINTS=y CONFIG_NET_DROP_MONITOR=m CONFIG_NETDEVSIM=m CONFIG_NET_FOU=m +CONFIG_RPS=y +CONFIG_SYSFS=y +CONFIG_PROC_SYSCTL=y diff --git a/tools/testing/selftests/net/rps_default_mask.sh b/tools/testing/selftests/net/rps_default_mask.sh new file mode 100755 index 000000000000..c81c0ac7ddfe --- /dev/null +++ b/tools/testing/selftests/net/rps_default_mask.sh @@ -0,0 +1,57 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 + +readonly ksft_skip=4 +readonly cpus=$(nproc) +ret=0 + +[ $cpus -gt 2 ] || exit $ksft_skip + +readonly INITIAL_RPS_DEFAULT_MASK=$(cat /proc/sys/net/core/rps_default_mask) +readonly NETNS="ns-$(mktemp -u XXXXXX)" + +setup() { + ip netns add "${NETNS}" + ip -netns "${NETNS}" link set lo up +} + +cleanup() { + echo $INITIAL_RPS_DEFAULT_MASK > /proc/sys/net/core/rps_default_mask + ip netns del $NETNS +} + +chk_rps() { + local rps_mask expected_rps_mask=$3 + local dev_name=$2 + local msg=$1 + + rps_mask=$(ip netns exec $NETNS cat /sys/class/net/$dev_name/queues/rx-0/rps_cpus) + printf "%-60s" "$msg" + if [ $rps_mask -eq $expected_rps_mask ]; then + echo "[ ok ]" + else + echo "[fail] expected $expected_rps_mask found $rps_mask" + ret=1 + fi +} + +trap cleanup EXIT + +echo 0 > /proc/sys/net/core/rps_default_mask +setup +chk_rps "empty rps_default_mask" lo 0 +cleanup + +echo 1 > /proc/sys/net/core/rps_default_mask +setup +chk_rps "non zero rps_default_mask" lo 1 + +echo 3 > /proc/sys/net/core/rps_default_mask +chk_rps "changing rps_default_mask dont affect existing netns" lo 1 + +ip -n $NETNS link add type veth +ip -n $NETNS link set dev veth0 up +ip -n $NETNS link set dev veth1 up +chk_rps "changing rps_default_mask affect newly created devices" veth0 3 +chk_rps "changing rps_default_mask affect newly created devices[II]" veth1 3 +exit $ret
On Fri, 30 Oct 2020 12:16:00 +0100 Paolo Abeni wrote:
Real-time setups try hard to ensure proper isolation between time critical applications and e.g. network processing performed by the network stack in softirq and RPS is used to move the softirq activity away from the isolated core.
If the network configuration is dynamic, with netns and devices routinely created at run-time, enforcing the correct RPS setting on each newly created device allowing to transient bad configuration became complex.
These series try to address the above, introducing a new sysctl knob: rps_default_mask. The new sysctl entry allows configuring a systemwide RPS mask, to be enforced since receive queue creation time without any fourther per device configuration required.
Additionally, a simple self-test is introduced to check the rps_default_mask behavior.
RPS is disabled by default, the processing is going to happen wherever the IRQ is mapped, and one would hope that the IRQ is not mapped to the core where the critical processing runs.
Would you mind elaborating further on the use case?
On Mon, 2020-11-02 at 14:54 -0800, Jakub Kicinski wrote:
On Fri, 30 Oct 2020 12:16:00 +0100 Paolo Abeni wrote:
Real-time setups try hard to ensure proper isolation between time critical applications and e.g. network processing performed by the network stack in softirq and RPS is used to move the softirq activity away from the isolated core.
If the network configuration is dynamic, with netns and devices routinely created at run-time, enforcing the correct RPS setting on each newly created device allowing to transient bad configuration became complex.
These series try to address the above, introducing a new sysctl knob: rps_default_mask. The new sysctl entry allows configuring a systemwide RPS mask, to be enforced since receive queue creation time without any fourther per device configuration required.
The whole thing can be replaced with a user daemon scripts that monitors all newly created devices and assign to them whatever rps mask (call it default).
So why do we need this special logic in kernel ?
I am not sure about this, but if rps queues sysfs are available before the netdev is up, then you can also use udevd to assign the rps masks before such devices are even brought up, so you would avoid the race conditions that you described, which are not really clear to me to be honest.
Additionally, a simple self-test is introduced to check the rps_default_mask behavior.
RPS is disabled by default, the processing is going to happen wherever the IRQ is mapped, and one would hope that the IRQ is not mapped to the core where the critical processing runs.
Would you mind elaborating further on the use case?
On Mon, 2020-11-02 at 14:54 -0800, Jakub Kicinski wrote:
On Fri, 30 Oct 2020 12:16:00 +0100 Paolo Abeni wrote:
Real-time setups try hard to ensure proper isolation between time critical applications and e.g. network processing performed by the network stack in softirq and RPS is used to move the softirq activity away from the isolated core.
If the network configuration is dynamic, with netns and devices routinely created at run-time, enforcing the correct RPS setting on each newly created device allowing to transient bad configuration became complex.
These series try to address the above, introducing a new sysctl knob: rps_default_mask. The new sysctl entry allows configuring a systemwide RPS mask, to be enforced since receive queue creation time without any fourther per device configuration required.
Additionally, a simple self-test is introduced to check the rps_default_mask behavior.
RPS is disabled by default, the processing is going to happen wherever the IRQ is mapped, and one would hope that the IRQ is not mapped to the core where the critical processing runs.
Would you mind elaborating further on the use case?
On Mon, 2020-11-02 at 15:27 -0800, Saeed Mahameed wrote:
The whole thing can be replaced with a user daemon scripts that monitors all newly created devices and assign to them whatever rps mask (call it default).
So why do we need this special logic in kernel ?
I am not sure about this, but if rps queues sysfs are available before the netdev is up, then you can also use udevd to assign the rps masks before such devices are even brought up, so you would avoid the race conditions that you described, which are not really clear to me to be honest.
Thank you for the feedback.
Please allow me to answer you both here, as your questions are related.
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Paolo
On Tue, 03 Nov 2020 16:22:07 +0100 Paolo Abeni wrote:
On Mon, 2020-11-02 at 14:54 -0800, Jakub Kicinski wrote:
On Fri, 30 Oct 2020 12:16:00 +0100 Paolo Abeni wrote:
Real-time setups try hard to ensure proper isolation between time critical applications and e.g. network processing performed by the network stack in softirq and RPS is used to move the softirq activity away from the isolated core.
If the network configuration is dynamic, with netns and devices routinely created at run-time, enforcing the correct RPS setting on each newly created device allowing to transient bad configuration became complex.
These series try to address the above, introducing a new sysctl knob: rps_default_mask. The new sysctl entry allows configuring a systemwide RPS mask, to be enforced since receive queue creation time without any fourther per device configuration required.
Additionally, a simple self-test is introduced to check the rps_default_mask behavior.
RPS is disabled by default, the processing is going to happen wherever the IRQ is mapped, and one would hope that the IRQ is not mapped to the core where the critical processing runs.
Would you mind elaborating further on the use case?
On Mon, 2020-11-02 at 15:27 -0800, Saeed Mahameed wrote:
The whole thing can be replaced with a user daemon scripts that monitors all newly created devices and assign to them whatever rps mask (call it default).
So why do we need this special logic in kernel ?
I am not sure about this, but if rps queues sysfs are available before the netdev is up, then you can also use udevd to assign the rps masks before such devices are even brought up, so you would avoid the race conditions that you described, which are not really clear to me to be honest.
Thank you for the feedback.
Please allow me to answer you both here, as your questions are related.
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Thanks, that makes it clearer now.
Depending on how RT-aware your container management is it may or may not be the right place to configure this, as it creates the veth interface. Presumably it's the container management which does the placement of the tasks to cores, why is it not setting other attributes, like RPS?
Also I wonder if it would make sense to turn this knob into something more generic. When we arrive at the threaded NAPIs - could it make sense for the threads to inherit your mask as the CPUs they are allowed to run on?
On Tue, 2020-11-03 at 08:52 -0800, Jakub Kicinski wrote:
On Tue, 03 Nov 2020 16:22:07 +0100 Paolo Abeni wrote:
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Thanks, that makes it clearer now.
Depending on how RT-aware your container management is it may or may not be the right place to configure this, as it creates the veth interface. Presumably it's the container management which does the placement of the tasks to cores, why is it not setting other attributes, like RPS?
The container orchestration is quite complex, and I'm unsure isolation and networking configuration are performed (or can be performed) by the same precess (without an heavy refactor).
On the flip hand, the global rps mask knob looked quite straightforward to me.
Possibly I can reduce the amount of new code introduced by this patchset removing some code duplication between rps_default_mask_sysctl() and flow_limit_cpu_sysctl(). Would that make this change more acceptable? Or should I drop this altogether?
Also I wonder if it would make sense to turn this knob into something more generic. When we arrive at the threaded NAPIs - could it make sense for the threads to inherit your mask as the CPUs they are allowed to run on?
I personally *think* this would be fine - and good. But isn't a bit premature discussing the integration of 2 missing pieces ? :)
Thanks,
Paolo
On Wed, Nov 04, 2020 at 06:36:08PM +0100, Paolo Abeni wrote:
On Tue, 2020-11-03 at 08:52 -0800, Jakub Kicinski wrote:
On Tue, 03 Nov 2020 16:22:07 +0100 Paolo Abeni wrote:
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Thanks, that makes it clearer now.
Depending on how RT-aware your container management is it may or may not be the right place to configure this, as it creates the veth interface. Presumably it's the container management which does the placement of the tasks to cores, why is it not setting other attributes, like RPS?
The container orchestration is quite complex, and I'm unsure isolation and networking configuration are performed (or can be performed) by the same precess (without an heavy refactor).
Also for the host side (no containers) the same issue will have to be handled for PCI hotplug for example. So this fix will have to be performed in every tool that decides to create a network device (while a kernel solution is global).
On the flip hand, the global rps mask knob looked quite straightforward to me.
Possibly I can reduce the amount of new code introduced by this patchset removing some code duplication between rps_default_mask_sysctl() and flow_limit_cpu_sysctl(). Would that make this change more acceptable? Or should I drop this altogether?
Also I wonder if it would make sense to turn this knob into something more generic. When we arrive at the threaded NAPIs - could it make sense for the threads to inherit your mask as the CPUs they are allowed to run on?
I personally *think* this would be fine - and good. But isn't a bit premature discussing the integration of 2 missing pieces ? :)
Thanks,
Paolo
About the potential race:
0) network device creation starts, inherits old default_rps_mask, network device init sleeps 1) set default_rps_mask (new) 2) change all devices across all network namespaces (walk /sys) 3) network device init wakes up, new device shows up in /sys/ using old default_rps_mask
Why this can't happen?
On Wed, 04 Nov 2020 18:36:08 +0100 Paolo Abeni wrote:
On Tue, 2020-11-03 at 08:52 -0800, Jakub Kicinski wrote:
On Tue, 03 Nov 2020 16:22:07 +0100 Paolo Abeni wrote:
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Thanks, that makes it clearer now.
Depending on how RT-aware your container management is it may or may not be the right place to configure this, as it creates the veth interface. Presumably it's the container management which does the placement of the tasks to cores, why is it not setting other attributes, like RPS?
The container orchestration is quite complex, and I'm unsure isolation and networking configuration are performed (or can be performed) by the same precess (without an heavy refactor).
On the flip hand, the global rps mask knob looked quite straightforward to me.
I understand, but I can't shake the feeling this is a hack.
Whatever sets the CPU isolation should take care of the RPS settings.
Possibly I can reduce the amount of new code introduced by this patchset removing some code duplication between rps_default_mask_sysctl() and flow_limit_cpu_sysctl(). Would that make this change more acceptable? Or should I drop this altogether?
I'm leaning towards drop altogether, unless you can get some support/review tags from other netdev developers. So far it appears we only got a down vote from Saeed.
Also I wonder if it would make sense to turn this knob into something more generic. When we arrive at the threaded NAPIs - could it make sense for the threads to inherit your mask as the CPUs they are allowed to run on?
I personally *think* this would be fine - and good. But isn't a bit premature discussing the integration of 2 missing pieces ? :)
Hi all,
On Wed, 2020-11-04 at 12:42 -0700, Jakub Kicinski wrote:
On Wed, 04 Nov 2020 18:36:08 +0100 Paolo Abeni wrote:
On Tue, 2020-11-03 at 08:52 -0800, Jakub Kicinski wrote:
On Tue, 03 Nov 2020 16:22:07 +0100 Paolo Abeni wrote:
The relevant use case is an host running containers (with the related orchestration tools) in a RT environment. Virtual devices (veths, ovs ports, etc.) are created by the orchestration tools at run-time. Critical processes are allowed to send packets/generate outgoing network traffic - but any interrupt is moved away from the related cores, so that usual incoming network traffic processing does not happen there.
Still an xmit operation on a virtual devices may be transmitted via ovs or veth, with the relevant forwarding operation happening in a softirq on the same CPU originating the packet.
RPS is configured (even) on such virtual devices to move away the forwarding from the relevant CPUs.
As Saeed noted, such configuration could be possibly performed via some user-space daemon monitoring network devices and network namespaces creation. That will be anyway prone to some race: the orchestation tool may create and enable the netns and virtual devices before the daemon has properly set the RPS mask.
In the latter scenario some packet forwarding could still slip in the relevant CPU, causing measurable latency. In all non RT scenarios the above will be likely irrelevant, but in the RT context that is not acceptable - e.g. it causes in real environments latency above the defined limits, while the proposed patches avoid the issue.
Do you see any other simple way to avoid the above race?
Please let me know if the above answers your doubts,
Thanks, that makes it clearer now.
Depending on how RT-aware your container management is it may or may not be the right place to configure this, as it creates the veth interface. Presumably it's the container management which does the placement of the tasks to cores, why is it not setting other attributes, like RPS?
The container orchestration is quite complex, and I'm unsure isolation and networking configuration are performed (or can be performed) by the same precess (without an heavy refactor).
On the flip hand, the global rps mask knob looked quite straightforward to me.
I understand, but I can't shake the feeling this is a hack.
Whatever sets the CPU isolation should take care of the RPS settings.
Let me try for a moment to revive this old thread.
Tha series proposed a new sysctl know to implement a global/default rps mask applying to all the network devices as a way to simplify some RT setups. It has been rejected as the required task is doable in user- space.
Currently the orchestration infrastructure does that, setting the per device, per queue rps mask and CPU isolation.
The above leads to a side problem: when there are lot of netns/devices with several queues, even a reasonably optimized user-space solution takes a relevant amount of time to traverse the relevant sysfs dirs and do I/O on them. Overall the additional time required is very measurable, easily ranging in seconds.
The default_rps_mask would basically kill that overhead.
Is the above a suitable use case?
Thanks,
Paolo
On Mon, 30 Jan 2023 10:25:34 +0100 Paolo Abeni wrote:
Let me try for a moment to revive this old thread.
Tha series proposed a new sysctl know to implement a global/default rps mask applying to all the network devices as a way to simplify some RT setups. It has been rejected as the required task is doable in user- space.
Currently the orchestration infrastructure does that, setting the per device, per queue rps mask and CPU isolation.
The above leads to a side problem: when there are lot of netns/devices with several queues, even a reasonably optimized user-space solution takes a relevant amount of time to traverse the relevant sysfs dirs and do I/O on them. Overall the additional time required is very measurable, easily ranging in seconds.
The default_rps_mask would basically kill that overhead.
Is the above a suitable use case?
Alright, thanks for trying the user space fix.
linux-kselftest-mirror@lists.linaro.org
-
Jakub Kicinski -
Marcelo Tosatti -
Paolo Abeni -
Saeed Mahameed