diff options
| author | Tim Chen <tim.c.chen@linux.intel.com> | 2026-09-21 17:37:22 -0700 |
|---|---|---|
| committer | Ingo Molnar <mingo@kernel.org> | 2026-09-22 10:49:53 +0200 |
| commit | 0d6526f82c3cdefcca47f73f5fc08dc6f335eac6 (patch) | |
| tree | 176dc2d685debcfb082c151ee94f2425ac96b364 /include/net/selftests.h | |
| download | linux-stable-0d6526f82c3cdefcca47f73f5fc08dc6f335eac6.tar.gz linux-stable-0d6526f82c3cdefcca47f73f5fc08dc6f335eac6.zip | |
sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling buggrafted
alb_break_llc() decides whether to break LLC preference during active
load balance. It does so by testing that every runnable fair task on the
source rq prefers its LLC:
env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable
But the two counters cover different sets. nr_pref_llc_running is updated
in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued,
so it follows queued tasks. h_nr_runnable is updated in set_delayed()/
clear_delayed() and drops delay-dequeued tasks.
So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted
in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks,
alb_break_llc() returns false, and active balance is free to pull a task
off its preferred LLC. Active balance only moves runnable tasks, and this
is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB
skips the per-task test in can_migrate_task(). The runnable set is the one
we want.
Fix it on the counter side. A task should be counted in
nr_pref_llc_running exactly while it is both queued on its preferred LLC
(pref_llc_queued) and runnable (!sched_delayed). Define that membership
once in task_pref_llc_runnable(), and adjust the counter only through
pref_llc_running_inc()/pref_llc_running_dec() from the four sites that
change either input: account_llc_enqueue(), account_llc_dequeue(),
set_delayed() and clear_delayed(). Gating every update on the same
predicate keeps the delay, wake and dequeue paths from double-counting
or underflowing; see the comments at those sites for the ordering.
nr_llc_running and sd->llc_counts are not touched and stay on queued
semantics.
Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC")
Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/
Reported-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Suggested-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Kayra Cizmeci <kayracizmeci@gmail.com>
Cc: <stable@kernel.org> # v7.2.x
Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com
Diffstat (limited to 'include/net/selftests.h')
| -rw-r--r-- | include/net/selftests.h | 76 |
1 files changed, 76 insertions, 0 deletions
diff --git a/include/net/selftests.h b/include/net/selftests.h new file mode 100644 index 000000000..c36e07406 --- /dev/null +++ b/include/net/selftests.h @@ -0,0 +1,76 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _NET_SELFTESTS +#define _NET_SELFTESTS + +#include <linux/ethtool.h> +#include <linux/netdevice.h> + +struct net_packet_attrs { + const unsigned char *src; + const unsigned char *dst; + u32 ip_src; + u32 ip_dst; + bool tcp; + u16 sport; + u16 dport; + int timeout; + int size; + int max_size; + u8 id; + u16 queue_mapping; + bool bad_csum; +}; + +struct net_test_priv { + struct net_packet_attrs *packet; + struct packet_type pt; + struct completion comp; + int double_vlan; + int vlan_id; + int ok; +}; + +struct netsfhdr { + __be32 version; + __be64 magic; + u8 id; +} __packed; + +#define NET_TEST_PKT_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr) + \ + sizeof(struct netsfhdr)) +#define NET_TEST_PKT_MAGIC 0xdeadcafecafedeadULL +#define NET_LB_TIMEOUT msecs_to_jiffies(200) + +#if IS_ENABLED(CONFIG_NET_SELFTESTS) + +struct sk_buff *net_test_get_skb(struct net_device *ndev, u8 id, + struct net_packet_attrs *attr); +void net_selftest(struct net_device *ndev, struct ethtool_test *etest, + u64 *buf); +int net_selftest_get_count(void); +void net_selftest_get_strings(u8 *data); + +#else + +static inline struct sk_buff *net_test_get_skb(struct net_device *ndev, u8 id, + struct net_packet_attrs *attr) +{ + return NULL; +} + +static inline void net_selftest(struct net_device *ndev, struct ethtool_test *etest, + u64 *buf) +{ +} + +static inline int net_selftest_get_count(void) +{ + return 0; +} + +static inline void net_selftest_get_strings(u8 *data) +{ +} + +#endif +#endif /* _NET_SELFTESTS */ |
