This patchset introduces target resume capability to netconsole allowing it to recover targets when underlying low-level interface comes back online.
The patchset starts by refactoring netconsole state representation in order to allow representing deactivated targets (targets that are disabled due to interfaces going down).
It then modifies netconsole to handle NETDEV_UP events for such targets and setups netpoll. Targets are matched with incoming interfaces depending on how they were initially bound in netconsole (by mac or interface name).
The patchset includes a selftest that validates netconsole target state transitions and that target is functional after resumed.
Signed-off-by: Andre Carvalho asantostc@gmail.com --- Changes in v7: - selftest: use ${EXIT_STATUS} instead of ${ksft_pass} to avoid shellcheck warning - Link to v6: https://lore.kernel.org/r/20251121-netcons-retrigger-v6-0-9c03f5a2bd6f@gmail...
Changes in v6: - Rebase on top of net-next to resolve conflicts, no functional changes. - Link to v5: https://lore.kernel.org/r/20251119-netcons-retrigger-v5-0-2c7dda6055d6@gmail...
Changes in v5: - patch 3: Set (de)enslaved target as DISABLED instead of DEACTIVATED to prevent resuming it. - selftest: Fix test cleanup by moving trap line to outside of loop and remove unneeded 'local' keyword - Rename maybe_resume_target to resume_target, add netconsole_ prefix to process_resumable_targets. - Hold device reference before calling __netpoll_setup. - Link to v4: https://lore.kernel.org/r/20251116-netcons-retrigger-v4-0-5290b5f140c2@gmail...
Changes in v4: - Simplify selftest cleanup, removing trap setup in loop. - Drop netpoll helper (__setup_netpoll_hold) and manage reference inside netconsole. - Move resume_list processing logic to separate function. - Link to v3: https://lore.kernel.org/r/20251109-netcons-retrigger-v3-0-1654c280bbe6@gmail...
Changes in v3: - Resume by mac or interface name depending on how target was created. - Attempt to resume target without holding target list lock, by moving the target to a temporary list. This is required as netpoll may attempt to allocate memory. - Link to v2: https://lore.kernel.org/r/20250921-netcons-retrigger-v2-0-a0e84006237f@gmail...
Changes in v2: - Attempt to resume target in the same thread, instead of using workqueue . - Add wrapper around __netpoll_setup (patch 4). - Renamed resume_target to maybe_resume_target and moved conditionals to inside its implementation, keeping code more clear. - Verify that device addr matches target mac address when target was setup using mac. - Update selftest to cover targets bound by mac and interface name. - Fix typo in selftest comment and sort tests alphabetically in Makefile. - Link to v1: https://lore.kernel.org/r/20250909-netcons-retrigger-v1-0-3aea904926cf@gmail...
--- Andre Carvalho (3): netconsole: convert 'enabled' flag to enum for clearer state management netconsole: resume previously deactivated target selftests: netconsole: validate target resume
Breno Leitao (2): netconsole: add target_state enum netconsole: add STATE_DEACTIVATED to track targets disabled by low level
drivers/net/netconsole.c | 155 +++++++++++++++++---- tools/testing/selftests/drivers/net/Makefile | 1 + .../selftests/drivers/net/lib/sh/lib_netcons.sh | 35 ++++- .../selftests/drivers/net/netcons_resume.sh | 97 +++++++++++++ 4 files changed, 254 insertions(+), 34 deletions(-) --- base-commit: ab084f0b8d6d2ee4b1c6a28f39a2a7430bdfa7f0 change-id: 20250816-netcons-retrigger-a4f547bfc867
Best regards,
From: Breno Leitao leitao@debian.org
Introduces a enum to track netconsole target state which is going to replace the enabled boolean.
Signed-off-by: Breno Leitao leitao@debian.org Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 5 +++++ 1 file changed, 5 insertions(+)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 9cb4dfc242f5..e2ec09f238a0 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -119,6 +119,11 @@ enum sysdata_feature { MAX_SYSDATA_ITEMS = 4, };
+enum target_state { + STATE_DISABLED, + STATE_ENABLED, +}; + /** * struct netconsole_target - Represents a configured netconsole target. * @list: Links this target into the target_list.
This patch refactors the netconsole driver's target enabled state from a simple boolean to an explicit enum (`target_state`).
This allow the states to be expanded to a new state in the upcoming change.
Co-developed-by: Breno Leitao leitao@debian.org Signed-off-by: Breno Leitao leitao@debian.org Reviewed-by: Breno Leitao leitao@debian.org Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 52 ++++++++++++++++++++++++++---------------------- 1 file changed, 28 insertions(+), 24 deletions(-)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index e2ec09f238a0..b21ecea60d52 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -135,12 +135,12 @@ enum target_state { * @sysdata_fields: Sysdata features enabled. * @msgcounter: Message sent counter. * @stats: Packet send stats for the target. Used for debugging. - * @enabled: On / off knob to enable / disable target. + * @state: State of the target. * Visible from userspace (read-write). * We maintain a strict 1:1 correspondence between this and * whether the corresponding netpoll is active or inactive. * Also, other parameters of a target may be modified at - * runtime only when it is disabled (enabled == 0). + * runtime only when it is disabled (state == STATE_DISABLED). * @extended: Denotes whether console is extended or not. * @release: Denotes whether kernel release version should be prepended * to the message. Depends on extended console. @@ -170,7 +170,7 @@ struct netconsole_target { u32 msgcounter; #endif struct netconsole_target_stats stats; - bool enabled; + enum target_state state; bool extended; bool release; struct netpoll np; @@ -262,6 +262,7 @@ static struct netconsole_target *alloc_and_init(void) nt->np.local_port = 6665; nt->np.remote_port = 6666; eth_broadcast_addr(nt->np.remote_mac); + nt->state = STATE_DISABLED;
return nt; } @@ -280,7 +281,7 @@ static void netconsole_process_cleanups_core(void) mutex_lock(&target_cleanup_list_lock); list_for_each_entry_safe(nt, tmp, &target_cleanup_list, list) { /* all entries in the cleanup_list needs to be disabled */ - WARN_ON_ONCE(nt->enabled); + WARN_ON_ONCE(nt->state == STATE_ENABLED); do_netpoll_cleanup(&nt->np); /* moved the cleaned target to target_list. Need to hold both * locks @@ -403,7 +404,7 @@ static void trim_newline(char *s, size_t maxlen)
static ssize_t enabled_show(struct config_item *item, char *buf) { - return sysfs_emit(buf, "%d\n", to_target(item)->enabled); + return sysfs_emit(buf, "%d\n", to_target(item)->state == STATE_ENABLED); }
static ssize_t extended_show(struct config_item *item, char *buf) @@ -570,8 +571,8 @@ static ssize_t enabled_store(struct config_item *item, const char *buf, size_t count) { struct netconsole_target *nt = to_target(item); + bool enabled, current_enabled; unsigned long flags; - bool enabled; ssize_t ret;
mutex_lock(&dynamic_netconsole_mutex); @@ -580,9 +581,10 @@ static ssize_t enabled_store(struct config_item *item, goto out_unlock;
ret = -EINVAL; - if (enabled == nt->enabled) { + current_enabled = nt->state == STATE_ENABLED; + if (enabled == current_enabled) { pr_info("network logging has already %s\n", - nt->enabled ? "started" : "stopped"); + current_enabled ? "started" : "stopped"); goto out_unlock; }
@@ -615,16 +617,16 @@ static ssize_t enabled_store(struct config_item *item, if (ret) goto out_unlock;
- nt->enabled = true; + nt->state = STATE_ENABLED; pr_info("network logging started\n"); } else { /* false */ /* We need to disable the netconsole before cleaning it up * otherwise we might end up in write_msg() with - * nt->np.dev == NULL and nt->enabled == true + * nt->np.dev == NULL and nt->state == STATE_ENABLED */ mutex_lock(&target_cleanup_list_lock); spin_lock_irqsave(&target_list_lock, flags); - nt->enabled = false; + nt->state = STATE_DISABLED; /* Remove the target from the list, while holding * target_list_lock */ @@ -653,7 +655,7 @@ static ssize_t release_store(struct config_item *item, const char *buf, ssize_t ret;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); ret = -EINVAL; @@ -680,7 +682,7 @@ static ssize_t extended_store(struct config_item *item, const char *buf, ssize_t ret;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); ret = -EINVAL; @@ -704,7 +706,7 @@ static ssize_t dev_name_store(struct config_item *item, const char *buf, struct netconsole_target *nt = to_target(item);
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); mutex_unlock(&dynamic_netconsole_mutex); @@ -725,7 +727,7 @@ static ssize_t local_port_store(struct config_item *item, const char *buf, ssize_t ret = -EINVAL;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); goto out_unlock; @@ -747,7 +749,7 @@ static ssize_t remote_port_store(struct config_item *item, ssize_t ret = -EINVAL;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); goto out_unlock; @@ -770,7 +772,7 @@ static ssize_t local_ip_store(struct config_item *item, const char *buf, int ipv6;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); goto out_unlock; @@ -795,7 +797,7 @@ static ssize_t remote_ip_store(struct config_item *item, const char *buf, int ipv6;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); goto out_unlock; @@ -830,7 +832,7 @@ static ssize_t remote_mac_store(struct config_item *item, const char *buf, ssize_t ret = -EINVAL;
mutex_lock(&dynamic_netconsole_mutex); - if (nt->enabled) { + if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); goto out_unlock; @@ -1326,7 +1328,7 @@ static void drop_netconsole_target(struct config_group *group, * The target may have never been enabled, or was manually disabled * before being removed so netpoll may have already been cleaned up. */ - if (nt->enabled) + if (nt->state == STATE_ENABLED) netpoll_cleanup(&nt->np);
config_item_put(&nt->group.cg_item); @@ -1444,7 +1446,7 @@ static int netconsole_netdev_event(struct notifier_block *this, case NETDEV_RELEASE: case NETDEV_JOIN: case NETDEV_UNREGISTER: - nt->enabled = false; + nt->state = STATE_DISABLED; list_move(&nt->list, &target_cleanup_list); stopped = true; } @@ -1725,7 +1727,8 @@ static void write_ext_msg(struct console *con, const char *msg,
spin_lock_irqsave(&target_list_lock, flags); list_for_each_entry(nt, &target_list, list) - if (nt->extended && nt->enabled && netif_running(nt->np.dev)) + if (nt->extended && nt->state == STATE_ENABLED && + netif_running(nt->np.dev)) send_ext_msg_udp(nt, msg, len); spin_unlock_irqrestore(&target_list_lock, flags); } @@ -1745,7 +1748,8 @@ static void write_msg(struct console *con, const char *msg, unsigned int len)
spin_lock_irqsave(&target_list_lock, flags); list_for_each_entry(nt, &target_list, list) { - if (!nt->extended && nt->enabled && netif_running(nt->np.dev)) { + if (!nt->extended && nt->state == STATE_ENABLED && + netif_running(nt->np.dev)) { /* * We nest this inside the for-each-target loop above * so that we're able to get as much logging out to @@ -1901,7 +1905,7 @@ static struct netconsole_target *alloc_param_target(char *target_config, */ goto fail; } else { - nt->enabled = true; + nt->state = STATE_ENABLED; } populate_configfs_item(nt, cmdline_count);
From: Breno Leitao leitao@debian.org
When the low level interface brings a netconsole target down, record this using a new STATE_DEACTIVATED state. This allows netconsole to distinguish between targets explicitly disabled by users and those deactivated due to interface state changes.
It also enables automatic recovery and re-enabling of targets if the underlying low-level interfaces come back online.
From a code perspective, anything that is not STATE_ENABLED is disabled.
Devices (de)enslaving are marked STATE_DISABLED to prevent automatically resuming as enslaved interfaces cannot have netconsole enabled.
Signed-off-by: Breno Leitao leitao@debian.org Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index b21ecea60d52..7a1e5559fc0d 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -122,6 +122,7 @@ enum sysdata_feature { enum target_state { STATE_DISABLED, STATE_ENABLED, + STATE_DEACTIVATED, };
/** @@ -580,6 +581,14 @@ static ssize_t enabled_store(struct config_item *item, if (ret) goto out_unlock;
+ /* When the user explicitly enables or disables a target that is + * currently deactivated, reset its state to disabled. The DEACTIVATED + * state only tracks interface-driven deactivation and should _not_ + * persist when the user manually changes the target's enabled state. + */ + if (nt->state == STATE_DEACTIVATED) + nt->state = STATE_DISABLED; + ret = -EINVAL; current_enabled = nt->state == STATE_ENABLED; if (enabled == current_enabled) { @@ -1445,10 +1454,19 @@ static int netconsole_netdev_event(struct notifier_block *this, break; case NETDEV_RELEASE: case NETDEV_JOIN: - case NETDEV_UNREGISTER: + /* transition target to DISABLED instead of + * DEACTIVATED when (de)enslaving devices as + * their targets should not be automatically + * resumed when the interface is brought up. + */ nt->state = STATE_DISABLED; list_move(&nt->list, &target_cleanup_list); stopped = true; + break; + case NETDEV_UNREGISTER: + nt->state = STATE_DEACTIVATED; + list_move(&nt->list, &target_cleanup_list); + stopped = true; } } netconsole_target_put(nt);
Attempt to resume a previously deactivated target when the associated interface comes back (NETDEV_UP event is received) by calling __netpoll_setup on the device.
Depending on how the target was setup (by mac or interface name), the corresponding field is compared with the device being brought up.
Targets that are candidates for resuming are removed from the target list and added to a temp list, as __netpoll_setup is IRQ unsafe. __netpoll_setup assumes RTNL is held (which is guaranteed to be the case when handling the event). In case of success, hold a reference to the device which will be removed upon target (or netconsole) removal by netpoll_cleanup.
Target transitions to STATE_DISABLED in case of failures resuming it to avoid retrying the same target indefinitely.
Reviewed-by: Breno Leitao leitao@debian.org Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 82 ++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 76 insertions(+), 6 deletions(-)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 7a1e5559fc0d..2dcdfb44a9e0 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -138,10 +138,14 @@ enum target_state { * @stats: Packet send stats for the target. Used for debugging. * @state: State of the target. * Visible from userspace (read-write). - * We maintain a strict 1:1 correspondence between this and - * whether the corresponding netpoll is active or inactive. + * From a userspace perspective, the target is either enabled or + * disabled. Internally, although both STATE_DISABLED and + * STATE_DEACTIVATED correspond to inactive targets, the latter is + * due to automatic interface state changes and will try + * recover automatically, if the interface comes back + * online. * Also, other parameters of a target may be modified at - * runtime only when it is disabled (state == STATE_DISABLED). + * runtime only when it is disabled (state != STATE_ENABLED). * @extended: Denotes whether console is extended or not. * @release: Denotes whether kernel release version should be prepended * to the message. Depends on extended console. @@ -1430,17 +1434,76 @@ static int prepare_sysdata(struct netconsole_target *nt) } #endif /* CONFIG_NETCONSOLE_DYNAMIC */
+/* Attempts to resume logging to a deactivated target. */ +static void resume_target(struct netconsole_target *nt, struct net_device *ndev) +{ + int ret; + + netdev_hold(ndev, &nt->np.dev_tracker, GFP_KERNEL); + + ret = __netpoll_setup(&nt->np, ndev); + if (ret) { + /* netpoll fails setup once, do not try again. */ + nt->state = STATE_DISABLED; + netdev_put(ndev, &nt->np.dev_tracker); + return; + } + + nt->state = STATE_ENABLED; + pr_info("network logging resumed on interface %s\n", nt->np.dev_name); +} + +/* Check if the target was bound by mac address. */ +static bool bound_by_mac(struct netconsole_target *nt) +{ + return is_valid_ether_addr(nt->np.dev_mac); +} + +/* Checks if a deactivated target matches a device. */ +static bool deactivated_target_match(struct netconsole_target *nt, + struct net_device *ndev) +{ + if (nt->state != STATE_DEACTIVATED) + return false; + + if (bound_by_mac(nt)) + return !memcmp(nt->np.dev_mac, ndev->dev_addr, ETH_ALEN); + return !strncmp(nt->np.dev_name, ndev->name, IFNAMSIZ); +} + +/* Process targets in resume_list and returns them to target_list */ +static void netconsole_process_resumable_targets(struct list_head *resume_list, + struct net_device *ndev) +{ + struct netconsole_target *nt, *tmp; + unsigned long flags; + + list_for_each_entry_safe(nt, tmp, resume_list, list) { + resume_target(nt, ndev); + + /* At this point the target is either enabled or disabled and + * was cleaned up before getting deactivated. Either way, add it + * back to target list. + */ + spin_lock_irqsave(&target_list_lock, flags); + list_move(&nt->list, &target_list); + spin_unlock_irqrestore(&target_list_lock, flags); + } +} + /* Handle network interface device notifications */ static int netconsole_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) { - unsigned long flags; - struct netconsole_target *nt, *tmp; struct net_device *dev = netdev_notifier_info_to_dev(ptr); + struct netconsole_target *nt, *tmp; + LIST_HEAD(resume_list); bool stopped = false; + unsigned long flags;
if (!(event == NETDEV_CHANGENAME || event == NETDEV_UNREGISTER || - event == NETDEV_RELEASE || event == NETDEV_JOIN)) + event == NETDEV_RELEASE || event == NETDEV_JOIN || + event == NETDEV_UP)) goto done;
mutex_lock(&target_cleanup_list_lock); @@ -1469,6 +1532,11 @@ static int netconsole_netdev_event(struct notifier_block *this, stopped = true; } } + if (event == NETDEV_UP && deactivated_target_match(nt, dev)) + /* resume_target is IRQ unsafe, remove target from + * target_list in order to resume it with IRQ enabled. + */ + list_move(&nt->list, &resume_list); netconsole_target_put(nt); } spin_unlock_irqrestore(&target_list_lock, flags); @@ -1492,6 +1560,8 @@ static int netconsole_netdev_event(struct notifier_block *this, dev->name, msg); }
+ netconsole_process_resumable_targets(&resume_list, dev); + /* Process target_cleanup_list entries. By the end, target_cleanup_list * should be empty */
Introduce a new netconsole selftest to validate that netconsole is able to resume a deactivated target when the low level interface comes back.
The test setups the network using netdevsim, creates a netconsole target and then remove/add netdevsim in order to bring the same interfaces back. Afterwards, the test validates that the target works as expected.
Targets are created via cmdline parameters to the module to ensure that we are able to resume targets that were bound by mac and interface name.
Reviewed-by: Breno Leitao leitao@debian.org Signed-off-by: Andre Carvalho asantostc@gmail.com --- tools/testing/selftests/drivers/net/Makefile | 1 + .../selftests/drivers/net/lib/sh/lib_netcons.sh | 35 ++++++-- .../selftests/drivers/net/netcons_resume.sh | 97 ++++++++++++++++++++++ 3 files changed, 128 insertions(+), 5 deletions(-)
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile index f5c71d993750..3eba569b3366 100644 --- a/tools/testing/selftests/drivers/net/Makefile +++ b/tools/testing/selftests/drivers/net/Makefile @@ -19,6 +19,7 @@ TEST_PROGS := \ netcons_cmdline.sh \ netcons_fragmented_msg.sh \ netcons_overflow.sh \ + netcons_resume.sh \ netcons_sysdata.sh \ netcons_torture.sh \ netpoll_basic.py \ diff --git a/tools/testing/selftests/drivers/net/lib/sh/lib_netcons.sh b/tools/testing/selftests/drivers/net/lib/sh/lib_netcons.sh index 87f89fd92f8c..239f44d4a45d 100644 --- a/tools/testing/selftests/drivers/net/lib/sh/lib_netcons.sh +++ b/tools/testing/selftests/drivers/net/lib/sh/lib_netcons.sh @@ -203,19 +203,21 @@ function do_cleanup() { function cleanup_netcons() { # delete netconsole dynamic reconfiguration # do not fail if the target is already disabled - if [[ ! -d "${NETCONS_PATH}" ]] + local TARGET_PATH=${1:-${NETCONS_PATH}} + + if [[ ! -d "${TARGET_PATH}" ]] then # in some cases this is called before netcons path is created return fi - if [[ $(cat "${NETCONS_PATH}"/enabled) != 0 ]] + if [[ $(cat "${TARGET_PATH}"/enabled) != 0 ]] then - echo 0 > "${NETCONS_PATH}"/enabled || true + echo 0 > "${TARGET_PATH}"/enabled || true fi # Remove all the keys that got created during the selftest - find "${NETCONS_PATH}/userdata/" -mindepth 1 -type d -delete + find "${TARGET_PATH}/userdata/" -mindepth 1 -type d -delete # Remove the configfs entry - rmdir "${NETCONS_PATH}" + rmdir "${TARGET_PATH}" }
function cleanup() { @@ -377,6 +379,29 @@ function check_netconsole_module() { fi }
+function wait_target_state() { + local TARGET=${1} + local STATE=${2} + local TARGET_PATH="${NETCONS_CONFIGFS}"/"${TARGET}" + local ENABLED=0 + + if [ "${STATE}" == "enabled" ] + then + ENABLED=1 + fi + + if [ ! -d "$TARGET_PATH" ]; then + echo "FAIL: Target does not exist." >&2 + exit "${ksft_fail}" + fi + + local CHECK_CMD="grep "$ENABLED" "$TARGET_PATH/enabled"" + slowwait 2 sh -c "test -n "$($CHECK_CMD)"" || { + echo "FAIL: ${TARGET} is not ${STATE}." >&2 + exit "${ksft_fail}" + } +} + # A wrapper to translate protocol version to udp version function wait_for_port() { local NAMESPACE=${1} diff --git a/tools/testing/selftests/drivers/net/netcons_resume.sh b/tools/testing/selftests/drivers/net/netcons_resume.sh new file mode 100755 index 000000000000..383ad1149271 --- /dev/null +++ b/tools/testing/selftests/drivers/net/netcons_resume.sh @@ -0,0 +1,97 @@ +#!/usr/bin/env bash +# SPDX-License-Identifier: GPL-2.0 + +# This test validates that netconsole is able to resume a target that was +# deactivated when its interface was removed when the interface is brought +# back up. +# +# The test configures a netconsole target and then removes netdevsim module to +# cause the interface to disappear. Targets are configured via cmdline to ensure +# targets bound by interface name and mac address can be resumed. +# The test verifies that the target moved to disabled state before adding +# netdevsim and the interface back. +# +# Finally, the test verifies that the target is re-enabled automatically and +# the message is received on the destination interface. +# +# Author: Andre Carvalho asantostc@gmail.com + +set -euo pipefail + +SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")") + +source "${SCRIPTDIR}"/lib/sh/lib_netcons.sh + +modprobe netdevsim 2> /dev/null || true +rmmod netconsole 2> /dev/null || true + +check_netconsole_module + +function cleanup() { + cleanup_netcons "${NETCONS_CONFIGFS}/cmdline0" + do_cleanup + rmmod netconsole +} + +trap cleanup EXIT + +# Run the test twice, with different cmdline parameters +for BINDMODE in "ifname" "mac" +do + echo "Running with bind mode: ${BINDMODE}" >&2 + # Set current loglevel to KERN_INFO(6), and default to KERN_NOTICE(5) + echo "6 5" > /proc/sys/kernel/printk + + # Create one namespace and two interfaces + set_network + + # Create the command line for netconsole, with the configuration from + # the function above + CMDLINE=$(create_cmdline_str "${BINDMODE}") + + # The content of kmsg will be save to the following file + OUTPUT_FILE="/tmp/${TARGET}-${BINDMODE}" + + # Load the module, with the cmdline set + modprobe netconsole "${CMDLINE}" + # Expose cmdline target in configfs + mkdir "${NETCONS_CONFIGFS}/cmdline0" + + # Target should be enabled + wait_target_state "cmdline0" "enabled" + + # Remove low level module + rmmod netdevsim + # Target should be disabled + wait_target_state "cmdline0" "disabled" + + # Add back low level module + modprobe netdevsim + # Recreate namespace and two interfaces + set_network + # Target should be enabled again + wait_target_state "cmdline0" "enabled" + + # Listen for netconsole port inside the namespace and destination + # interface + listen_port_and_save_to "${OUTPUT_FILE}" & + # Wait for socat to start and listen to the port. + wait_local_port_listen "${NAMESPACE}" "${PORT}" udp + # Send the message + echo "${MSG}: ${TARGET}" > /dev/kmsg + # Wait until socat saves the file to disk + busywait "${BUSYWAIT_TIMEOUT}" test -s "${OUTPUT_FILE}" + # Make sure the message was received in the dst part + # and exit + validate_msg "${OUTPUT_FILE}" + + # kill socat in case it is still running + pkill_socat + # Cleanup & unload the module + cleanup + + echo "${BINDMODE} : Test passed" >&2 +done + +trap - EXIT +exit "${EXIT_STATUS}"
On Wed, 26 Nov 2025 20:22:52 +0000 Andre Carvalho wrote:
This patchset introduces target resume capability to netconsole allowing it to recover targets when underlying low-level interface comes back online.
The patchset starts by refactoring netconsole state representation in order to allow representing deactivated targets (targets that are disabled due to interfaces going down).
It then modifies netconsole to handle NETDEV_UP events for such targets and setups netpoll. Targets are matched with incoming interfaces depending on how they were initially bound in netconsole (by mac or interface name).
Netpoll does not seem to handle DOWN events, so I'm guessing your primary use case is that the device had a HW fault and netdev was recreated after device reset?
Should we not be listening for the REGISTER event then? On boot we force UP the device if we find it, theoretically there may be a case where user space is not configured to UP the device, and then we'd never resume the target?
Hi Jakub!
On Wed, Nov 26, 2025 at 05:36:46PM -0800, Jakub Kicinski wrote:
Netpoll does not seem to handle DOWN events, so I'm guessing your primary use case is that the device had a HW fault and netdev was recreated after device reset?
Correct, this is the intended use case for this series. Handling cases where the device was unregistered and then brought back up.
Should we not be listening for the REGISTER event then? On boot we force UP the device if we find it, theoretically there may be a case where user space is not configured to UP the device, and then we'd never resume the target?
This is indeed a limitation on the current implementation. Based on your feedback, I'm working on a new version of this series handling REGISTER instead of UP and ensuring we force UP the device. This will make it consistent with the boot behavior you described.
Based on my tests, I can't force the device UP while handling the REGISTER event. I believe this is due to dev_open attempting to lock the device which is already held. For this reason, I'm resorting to defering this to a workqueue, similar to my approach on v1 [1] (but correctly handling target_list lock).
Let me know if this approach makes sense or if I'm missing something.
Thanks for the review!
[1] https://lore.kernel.org/all/20250909-netcons-retrigger-v1-4-3aea904926cf@gma...
On Thu, 27 Nov 2025 23:07:02 +0000 Andre Carvalho wrote:
Should we not be listening for the REGISTER event then? On boot we force UP the device if we find it, theoretically there may be a case where user space is not configured to UP the device, and then we'd never resume the target?
This is indeed a limitation on the current implementation. Based on your feedback, I'm working on a new version of this series handling REGISTER instead of UP and ensuring we force UP the device. This will make it consistent with the boot behavior you described.
Based on my tests, I can't force the device UP while handling the REGISTER event. I believe this is due to dev_open attempting to lock the device which is already held. For this reason, I'm resorting to defering this to a workqueue, similar to my approach on v1 [1] (but correctly handling target_list lock).
Let me know if this approach makes sense or if I'm missing something.
SG, that's probably the most resilient solution.
linux-kselftest-mirror@lists.linaro.org
-
Andre Carvalho -
Jakub Kicinski