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 unregister).
It then modifies netconsole to handle NETDEV_REGISTER events for such targets, setups netpoll and forces the device UP. Targets are matched with incoming interfaces depending on how they were bound in netconsole (by mac or interface name). For these reasons, we also attempt resuming on NETDEV_CHANGENAME.
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 v10: - Define wrappers around dynamic_netconsole_mutex lock/unlock and use them on process_resume_target to avoid build failures and #ifdefs inside callsite (suggested by Breno). - Refactored other dynamic_netconsole_mutex to use the wrappers for consistency. - Ensure we cancel pending working during removal of dynamic targets, which requires also holding dynamic_netconsole_mutex. - Introduce standalone workqueue to avoid potential leaks during module cleanup, flushing all pending resume events before removing all targets. - Link to v9: https://lore.kernel.org/r/20260104-netcons-retrigger-v9-0-38aa643d2283@gmail...
Changes in v9: - Hold dynamic_netconsole_mutex on process_resume_target. - Cleanup dev_name as part of netconsole_process_cleanups_core to ensure we correctly resume by mac (for targets bound by mac) - Link to v8: https://lore.kernel.org/r/20251128-netcons-retrigger-v8-0-0bccbf4c6385@gmail...
Changes in v8: - Handle NETDEV_REGISTER/CHANGENAME instead of NETDEV_UP (and force the device UP), to increase the chances of succesfully resuming a target. This requires using a workqueue instead of inline in the event notifier as we can't UP the device otherwise. - Link to v7: https://lore.kernel.org/r/20251126-netcons-retrigger-v7-0-1d86dba83b1c@gmail...
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 (5): netconsole: convert 'enabled' flag to enum for clearer state management netconsole: clear dev_name for devices bound by mac netconsole: introduce helpers for dynamic_netconsole_mutex lock/unlock 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 | 305 ++++++++++++++++----- 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, 364 insertions(+), 74 deletions(-) --- base-commit: 60d8484c4cec811f5ceb6550655df74490d1a165 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);
This patch makes sure netconsole clears dev_name for devices bound by mac in order to allow calling setup_netpoll on targets that have previously been cleaned up (in order to support resuming deactivated targets).
This is required as netpoll_setup populates dev_name even when devices are matched via mac address. The cleanup is done inside netconsole as bound by mac is a netconsole concept.
Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 8 ++++++++ 1 file changed, 8 insertions(+)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 7a1e5559fc0d..02a3463e8d24 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -242,6 +242,12 @@ static void populate_configfs_item(struct netconsole_target *nt, } #endif /* CONFIG_NETCONSOLE_DYNAMIC */
+/* 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); +} + /* Allocate and initialize with defaults. * Note that these targets get their config_item fields zeroed-out. */ @@ -284,6 +290,8 @@ static void netconsole_process_cleanups_core(void) /* all entries in the cleanup_list needs to be disabled */ WARN_ON_ONCE(nt->state == STATE_ENABLED); do_netpoll_cleanup(&nt->np); + if (bound_by_mac(nt)) + memset(&nt->np.dev_name, 0, IFNAMSIZ); /* moved the cleaned target to target_list. Need to hold both * locks */
This commit introduces two helper functions to perform lock/unlock on dynamic_netconsole_mutex providing no-op stub versions when compiled without CONFIG_NETCONSOLE_DYNAMIC and refactors existing call sites to use the new helpers.
This is done following kernel coding style guidelines, in preparation for an upcoming change. It avoids the need for preprocessor conditionals in the call site and keeps the logic easier to follow.
Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 97 +++++++++++++++++++++++++++++------------------- 1 file changed, 58 insertions(+), 39 deletions(-)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 02a3463e8d24..46d990f3d904 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -213,6 +213,16 @@ static void netconsole_target_put(struct netconsole_target *nt) config_group_put(&nt->group); }
+static void dynamic_netconsole_mutex_lock(void) +{ + mutex_lock(&dynamic_netconsole_mutex); +} + +static void dynamic_netconsole_mutex_unlock(void) +{ + mutex_unlock(&dynamic_netconsole_mutex); +} + #else /* !CONFIG_NETCONSOLE_DYNAMIC */
static int __init dynamic_netconsole_init(void) @@ -240,6 +250,15 @@ static void populate_configfs_item(struct netconsole_target *nt, int cmdline_count) { } + +static void dynamic_netconsole_mutex_lock(void) +{ +} + +static void dynamic_netconsole_mutex_unlock(void) +{ +} + #endif /* CONFIG_NETCONSOLE_DYNAMIC */
/* Check if the target was bound by mac address. */ @@ -495,9 +514,9 @@ static ssize_t sysdata_cpu_nr_enabled_show(struct config_item *item, char *buf) struct netconsole_target *nt = to_target(item->ci_parent); bool cpu_nr_enabled;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); cpu_nr_enabled = !!(nt->sysdata_fields & SYSDATA_CPU_NR); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock();
return sysfs_emit(buf, "%d\n", cpu_nr_enabled); } @@ -509,9 +528,9 @@ static ssize_t sysdata_taskname_enabled_show(struct config_item *item, struct netconsole_target *nt = to_target(item->ci_parent); bool taskname_enabled;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); taskname_enabled = !!(nt->sysdata_fields & SYSDATA_TASKNAME); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock();
return sysfs_emit(buf, "%d\n", taskname_enabled); } @@ -522,9 +541,9 @@ static ssize_t sysdata_release_enabled_show(struct config_item *item, struct netconsole_target *nt = to_target(item->ci_parent); bool release_enabled;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); release_enabled = !!(nt->sysdata_fields & SYSDATA_TASKNAME); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock();
return sysfs_emit(buf, "%d\n", release_enabled); } @@ -562,9 +581,9 @@ static ssize_t sysdata_msgid_enabled_show(struct config_item *item, struct netconsole_target *nt = to_target(item->ci_parent); bool msgid_enabled;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); msgid_enabled = !!(nt->sysdata_fields & SYSDATA_MSGID); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock();
return sysfs_emit(buf, "%d\n", msgid_enabled); } @@ -584,7 +603,7 @@ static ssize_t enabled_store(struct config_item *item, unsigned long flags; ssize_t ret;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); ret = kstrtobool(buf, &enabled); if (ret) goto out_unlock; @@ -660,7 +679,7 @@ static ssize_t enabled_store(struct config_item *item, /* Deferred cleanup */ netconsole_process_cleanups(); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -671,7 +690,7 @@ static ssize_t release_store(struct config_item *item, const char *buf, bool release; ssize_t ret;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -687,7 +706,7 @@ static ssize_t release_store(struct config_item *item, const char *buf,
ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -698,7 +717,7 @@ static ssize_t extended_store(struct config_item *item, const char *buf, bool extended; ssize_t ret;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -713,7 +732,7 @@ static ssize_t extended_store(struct config_item *item, const char *buf, nt->extended = extended; ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -722,18 +741,18 @@ 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); + dynamic_netconsole_mutex_lock(); 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); + dynamic_netconsole_mutex_unlock(); return -EINVAL; }
strscpy(nt->np.dev_name, buf, IFNAMSIZ); trim_newline(nt->np.dev_name, IFNAMSIZ);
- mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return strnlen(buf, count); }
@@ -743,7 +762,7 @@ static ssize_t local_port_store(struct config_item *item, const char *buf, struct netconsole_target *nt = to_target(item); ssize_t ret = -EINVAL;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -755,7 +774,7 @@ static ssize_t local_port_store(struct config_item *item, const char *buf, goto out_unlock; ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -765,7 +784,7 @@ static ssize_t remote_port_store(struct config_item *item, struct netconsole_target *nt = to_target(item); ssize_t ret = -EINVAL;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -777,7 +796,7 @@ static ssize_t remote_port_store(struct config_item *item, goto out_unlock; ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -788,7 +807,7 @@ static ssize_t local_ip_store(struct config_item *item, const char *buf, ssize_t ret = -EINVAL; int ipv6;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -802,7 +821,7 @@ static ssize_t local_ip_store(struct config_item *item, const char *buf,
ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -813,7 +832,7 @@ static ssize_t remote_ip_store(struct config_item *item, const char *buf, ssize_t ret = -EINVAL; int ipv6;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -827,7 +846,7 @@ static ssize_t remote_ip_store(struct config_item *item, const char *buf,
ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -848,7 +867,7 @@ static ssize_t remote_mac_store(struct config_item *item, const char *buf, u8 remote_mac[ETH_ALEN]; ssize_t ret = -EINVAL;
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); if (nt->state == STATE_ENABLED) { pr_err("target (%s) is enabled, disable to update parameters\n", config_item_name(&nt->group.cg_item)); @@ -863,7 +882,7 @@ static ssize_t remote_mac_store(struct config_item *item, const char *buf,
ret = strnlen(buf, count); out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); return ret; }
@@ -984,7 +1003,7 @@ static ssize_t userdatum_value_store(struct config_item *item, const char *buf, return -EMSGSIZE;
mutex_lock(&netconsole_subsys.su_mutex); - mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock();
ret = strscpy(udm->value, buf, sizeof(udm->value)); if (ret < 0) @@ -998,7 +1017,7 @@ static ssize_t userdatum_value_store(struct config_item *item, const char *buf, goto out_unlock; ret = count; out_unlock: - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); mutex_unlock(&netconsole_subsys.su_mutex); return ret; } @@ -1026,7 +1045,7 @@ static ssize_t sysdata_msgid_enabled_store(struct config_item *item, return ret;
mutex_lock(&netconsole_subsys.su_mutex); - mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); curr = !!(nt->sysdata_fields & SYSDATA_MSGID); if (msgid_enabled == curr) goto unlock_ok; @@ -1038,7 +1057,7 @@ static ssize_t sysdata_msgid_enabled_store(struct config_item *item,
unlock_ok: ret = strnlen(buf, count); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); mutex_unlock(&netconsole_subsys.su_mutex); return ret; } @@ -1055,7 +1074,7 @@ static ssize_t sysdata_release_enabled_store(struct config_item *item, return ret;
mutex_lock(&netconsole_subsys.su_mutex); - mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); curr = !!(nt->sysdata_fields & SYSDATA_RELEASE); if (release_enabled == curr) goto unlock_ok; @@ -1067,7 +1086,7 @@ static ssize_t sysdata_release_enabled_store(struct config_item *item,
unlock_ok: ret = strnlen(buf, count); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); mutex_unlock(&netconsole_subsys.su_mutex); return ret; } @@ -1084,7 +1103,7 @@ static ssize_t sysdata_taskname_enabled_store(struct config_item *item, return ret;
mutex_lock(&netconsole_subsys.su_mutex); - mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); curr = !!(nt->sysdata_fields & SYSDATA_TASKNAME); if (taskname_enabled == curr) goto unlock_ok; @@ -1096,7 +1115,7 @@ static ssize_t sysdata_taskname_enabled_store(struct config_item *item,
unlock_ok: ret = strnlen(buf, count); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); mutex_unlock(&netconsole_subsys.su_mutex); return ret; } @@ -1114,7 +1133,7 @@ static ssize_t sysdata_cpu_nr_enabled_store(struct config_item *item, return ret;
mutex_lock(&netconsole_subsys.su_mutex); - mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); curr = !!(nt->sysdata_fields & SYSDATA_CPU_NR); if (cpu_nr_enabled == curr) /* no change requested */ @@ -1130,7 +1149,7 @@ static ssize_t sysdata_cpu_nr_enabled_store(struct config_item *item,
unlock_ok: ret = strnlen(buf, count); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); mutex_unlock(&netconsole_subsys.su_mutex); return ret; } @@ -1192,10 +1211,10 @@ static void userdatum_drop(struct config_group *group, struct config_item *item) ud = to_userdata(&group->cg_item); nt = userdata_to_target(ud);
- mutex_lock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_lock(); update_userdata(nt); config_item_put(item); - mutex_unlock(&dynamic_netconsole_mutex); + dynamic_netconsole_mutex_unlock(); }
static struct configfs_attribute *userdata_attrs[] = {
Attempt to resume a previously deactivated target when the associated interface comes back (NETDEV_REGISTER) or when it changes name (NETDEV_CHANGENAME) 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 match the incoming device, are scheduled for resume on a workqueue.
Resuming happens on a workqueue as we can't execute netpoll_setup in the context of the netdev event. A standalone workqueue (as opposed to the global one) is used to allow for proper cleanup process during netconsole module cleanup as we need to be able to flush all pending work before traversing the target list given that targets are temporarily removed from the list during resume_target.
Target transitions to STATE_DISABLED in case of failures resuming it to avoid retrying the same target indefinitely.
Signed-off-by: Andre Carvalho asantostc@gmail.com --- drivers/net/netconsole.c | 127 ++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 120 insertions(+), 7 deletions(-)
diff --git a/drivers/net/netconsole.c b/drivers/net/netconsole.c index 46d990f3d904..82c232f9ede2 100644 --- a/drivers/net/netconsole.c +++ b/drivers/net/netconsole.c @@ -39,6 +39,7 @@ #include <linux/u64_stats_sync.h> #include <linux/utsname.h> #include <linux/rtnetlink.h> +#include <linux/workqueue.h>
MODULE_AUTHOR("Matt Mackall mpm@selenic.com"); MODULE_DESCRIPTION("Console driver for network interfaces"); @@ -85,6 +86,8 @@ static DEFINE_SPINLOCK(target_list_lock); /* This needs to be a mutex because netpoll_cleanup might sleep */ static DEFINE_MUTEX(target_cleanup_list_lock);
+static struct workqueue_struct *netconsole_wq; + /* * Console driver for netconsoles. Register only consoles that have * an associated target of the same type. @@ -138,10 +141,14 @@ 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. @@ -155,6 +162,7 @@ enum target_state { * local_mac (read-only) * remote_mac (read-write) * @buf: The buffer used to send the full msg to the network stack + * @resume_wq: Workqueue to resume deactivated target */ struct netconsole_target { struct list_head list; @@ -177,6 +185,7 @@ struct netconsole_target { struct netpoll np; /* protected by target_list_lock */ char buf[MAX_PRINT_CHUNK]; + struct work_struct resume_wq; };
#ifdef CONFIG_NETCONSOLE_DYNAMIC @@ -267,6 +276,70 @@ static bool bound_by_mac(struct netconsole_target *nt) return is_valid_ether_addr(nt->np.dev_mac); }
+/* Attempts to resume logging to a deactivated target. */ +static void resume_target(struct netconsole_target *nt) +{ + if (netpoll_setup(&nt->np)) { + /* netpoll fails setup once, do not try again. */ + nt->state = STATE_DISABLED; + return; + } + + nt->state = STATE_ENABLED; + pr_info("network logging resumed on interface %s\n", nt->np.dev_name); +} + +/* 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 work scheduled for target resume. */ +static void process_resume_target(struct work_struct *work) +{ + struct netconsole_target *nt; + unsigned long flags; + + nt = container_of(work, struct netconsole_target, resume_wq); + + dynamic_netconsole_mutex_lock(); + + spin_lock_irqsave(&target_list_lock, flags); + /* Check if target is still deactivated as it may have been disabled + * while resume was being scheduled. + */ + if (nt->state != STATE_DEACTIVATED) { + spin_unlock_irqrestore(&target_list_lock, flags); + goto out_unlock; + } + + /* resume_target is IRQ unsafe, remove target from + * target_list in order to resume it with IRQ enabled. + */ + list_del_init(&nt->list); + spin_unlock_irqrestore(&target_list_lock, flags); + + resume_target(nt); + + /* 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_add(&nt->list, &target_list); + spin_unlock_irqrestore(&target_list_lock, flags); + +out_unlock: + dynamic_netconsole_mutex_unlock(); +} + /* Allocate and initialize with defaults. * Note that these targets get their config_item fields zeroed-out. */ @@ -289,6 +362,7 @@ static struct netconsole_target *alloc_and_init(void) nt->np.remote_port = 6666; eth_broadcast_addr(nt->np.remote_mac); nt->state = STATE_DISABLED; + INIT_WORK(&nt->resume_wq, process_resume_target);
return nt; } @@ -1353,13 +1427,29 @@ static struct config_group *make_netconsole_target(struct config_group *group, static void drop_netconsole_target(struct config_group *group, struct config_item *item) { - unsigned long flags; struct netconsole_target *nt = to_target(item); + unsigned long flags; + + dynamic_netconsole_mutex_lock();
spin_lock_irqsave(&target_list_lock, flags); + /* Disable deactivated target to prevent races between resume attempt + * and target removal. + */ + if (nt->state == STATE_DEACTIVATED) + nt->state = STATE_DISABLED; list_del(&nt->list); spin_unlock_irqrestore(&target_list_lock, flags);
+ dynamic_netconsole_mutex_unlock(); + + /* Now that the target has been marked disabled no further work + * can be scheduled. Existing work will skip as targets are not + * deactivated anymore. Cancel any scheduled resume and wait for + * completion. + */ + cancel_work_sync(&nt->resume_wq); + /* * The target may have never been enabled, or was manually disabled * before being removed so netpoll may have already been cleaned up. @@ -1461,13 +1551,14 @@ static int prepare_sysdata(struct netconsole_target *nt) 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; 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_REGISTER)) goto done;
mutex_lock(&target_cleanup_list_lock); @@ -1496,6 +1587,13 @@ static int netconsole_netdev_event(struct notifier_block *this, stopped = true; } } + if ((event == NETDEV_REGISTER || event == NETDEV_CHANGENAME) && + deactivated_target_match(nt, dev)) + /* Schedule resume on a workqueue as it will attempt + * to UP the device, which can't be done as part of this + * notifier. + */ + queue_work(netconsole_wq, &nt->resume_wq); netconsole_target_put(nt); } spin_unlock_irqrestore(&target_list_lock, flags); @@ -1964,6 +2062,7 @@ static struct netconsole_target *alloc_param_target(char *target_config, /* Cleanup netpoll for given target (from boot/module param) and free it */ static void free_param_target(struct netconsole_target *nt) { + cancel_work_sync(&nt->resume_wq); netpoll_cleanup(&nt->np); #ifdef CONFIG_NETCONSOLE_DYNAMIC kfree(nt->userdata); @@ -2018,6 +2117,12 @@ static int __init init_netconsole(void) } }
+ netconsole_wq = alloc_workqueue("netconsole", WQ_UNBOUND, 0); + if (!netconsole_wq) { + err = -ENOMEM; + goto fail; + } + err = register_netdevice_notifier(&netconsole_netdev_notifier); if (err) goto fail; @@ -2040,6 +2145,8 @@ static int __init init_netconsole(void) fail: pr_err("cleaning up\n");
+ if (netconsole_wq) + flush_workqueue(netconsole_wq); /* * Remove all targets and destroy them (only targets created * from the boot/module option exist here). Skipping the list @@ -2050,6 +2157,9 @@ static int __init init_netconsole(void) free_param_target(nt); }
+ if (netconsole_wq) + destroy_workqueue(netconsole_wq); + return err; }
@@ -2063,6 +2173,7 @@ static void __exit cleanup_netconsole(void) unregister_console(&netconsole); dynamic_netconsole_exit(); unregister_netdevice_notifier(&netconsole_netdev_notifier); + flush_workqueue(netconsole_wq);
/* * Targets created via configfs pin references on our module @@ -2076,6 +2187,8 @@ static void __exit cleanup_netconsole(void) list_del(&nt->list); free_param_target(nt); } + + destroy_workqueue(netconsole_wq); }
/*
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 ae8abff4be40..b6093bcf2b06 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}"
linux-kselftest-mirror@lists.linaro.org
-
Andre Carvalho