From 0d6526f82c3cdefcca47f73f5fc08dc6f335eac6 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Mon, 21 Sep 2026 17:37:22 -0700 Subject: sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling bug 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 Suggested-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Kayra Cizmeci Cc: # v7.2.x Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com --- crypto/crc32c.c | 149 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 149 insertions(+) create mode 100644 crypto/crc32c.c (limited to 'crypto/crc32c.c') diff --git a/crypto/crc32c.c b/crypto/crc32c.c new file mode 100644 index 000000000..9c9d0ae21 --- /dev/null +++ b/crypto/crc32c.c @@ -0,0 +1,149 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * crypto_shash support for CRC-32C + * + *@Article{castagnoli-crc, + * author = { Guy Castagnoli and Stefan Braeuer and Martin Herrman}, + * title = {{Optimization of Cyclic Redundancy-Check Codes with 24 + * and 32 Parity Bits}}, + * journal = IEEE Transactions on Communication, + * year = {1993}, + * volume = {41}, + * number = {6}, + * pages = {}, + * month = {June}, + *} + * + * Copyright (c) 2004 Cisco Systems, Inc. + * Copyright (c) 2008 Herbert Xu + */ + +#include +#include +#include +#include +#include +#include +#include + +#define CHKSUM_BLOCK_SIZE 1 +#define CHKSUM_DIGEST_SIZE 4 + +struct chksum_ctx { + u32 key; +}; + +struct chksum_desc_ctx { + u32 crc; +}; + +static int chksum_init(struct shash_desc *desc) +{ + struct chksum_ctx *mctx = crypto_shash_ctx(desc->tfm); + struct chksum_desc_ctx *ctx = shash_desc_ctx(desc); + + ctx->crc = mctx->key; + + return 0; +} + +/* + * Setting the seed allows arbitrary accumulators and flexible XOR policy + * If your algorithm starts with ~0, then XOR with ~0 before you set + * the seed. + */ +static int chksum_setkey(struct crypto_shash *tfm, const u8 *key, + unsigned int keylen) +{ + struct chksum_ctx *mctx = crypto_shash_ctx(tfm); + + if (keylen != sizeof(mctx->key)) + return -EINVAL; + mctx->key = get_unaligned_le32(key); + return 0; +} + +static int chksum_update(struct shash_desc *desc, const u8 *data, + unsigned int length) +{ + struct chksum_desc_ctx *ctx = shash_desc_ctx(desc); + + ctx->crc = crc32c(ctx->crc, data, length); + return 0; +} + +static int chksum_final(struct shash_desc *desc, u8 *out) +{ + struct chksum_desc_ctx *ctx = shash_desc_ctx(desc); + + put_unaligned_le32(~ctx->crc, out); + return 0; +} + +static int __chksum_finup(u32 *crcp, const u8 *data, unsigned int len, u8 *out) +{ + put_unaligned_le32(~crc32c(*crcp, data, len), out); + return 0; +} + +static int chksum_finup(struct shash_desc *desc, const u8 *data, + unsigned int len, u8 *out) +{ + struct chksum_desc_ctx *ctx = shash_desc_ctx(desc); + + return __chksum_finup(&ctx->crc, data, len, out); +} + +static int chksum_digest(struct shash_desc *desc, const u8 *data, + unsigned int length, u8 *out) +{ + struct chksum_ctx *mctx = crypto_shash_ctx(desc->tfm); + + return __chksum_finup(&mctx->key, data, length, out); +} + +static int crc32c_cra_init(struct crypto_tfm *tfm) +{ + struct chksum_ctx *mctx = crypto_tfm_ctx(tfm); + + mctx->key = ~0; + return 0; +} + +static struct shash_alg alg = { + .digestsize = CHKSUM_DIGEST_SIZE, + .setkey = chksum_setkey, + .init = chksum_init, + .update = chksum_update, + .final = chksum_final, + .finup = chksum_finup, + .digest = chksum_digest, + .descsize = sizeof(struct chksum_desc_ctx), + + .base.cra_name = "crc32c", + .base.cra_driver_name = "crc32c-lib", + .base.cra_priority = 100, + .base.cra_flags = CRYPTO_ALG_OPTIONAL_KEY, + .base.cra_blocksize = CHKSUM_BLOCK_SIZE, + .base.cra_ctxsize = sizeof(struct chksum_ctx), + .base.cra_module = THIS_MODULE, + .base.cra_init = crc32c_cra_init, +}; + +static int __init crc32c_mod_init(void) +{ + return crypto_register_shash(&alg); +} + +static void __exit crc32c_mod_fini(void) +{ + crypto_unregister_shash(&alg); +} + +module_init(crc32c_mod_init); +module_exit(crc32c_mod_fini); + +MODULE_AUTHOR("Clay Haapala "); +MODULE_DESCRIPTION("CRC32c (Castagnoli) calculations wrapper for lib/crc32c"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_CRYPTO("crc32c"); -- cgit v1.3.1