- Kernel - mailweb.openeuler.org

[PATCH openEuler-25.03 00/10] Implementation of the uncore frequency modulation feature
by Xiangwei Li 18 Dec '24

18 Dec '24

Devfreq is a general DVFS framework for non-CPU devices in the kernel. It is generally used for bus devices. Implementation of uncore dynamic frequency modulation based on this framework. Jie Zhan (1): PM / devfreq: Add HiSilicon uncore frequency scaling driver Xiangwei Li (9): PM / devfreq: Adjust default HiSilicon uncore devfreq polling interval and workqueue PM / devfreq: event: Add devfreq_event_get_edev_by_dev function PM / devfreq: event: Supplement the name content of the event in devfreq-event PM / devfreq: Add related-package parameter in HiSilicon uncore devfreq PM / devfreq: Add associated events configuration in HiSilicon uncore frequency PM / devfreq: event: Add L3C event for HiSilicon uncore devfreq PM / devfreq: Add until governor PM / devfreq: Add bandwidth verification mechanism PM / devfreq: event: Add DDRC event for HiSilicon uncore devfreq drivers/devfreq/Kconfig | 12 + drivers/devfreq/Makefile | 2 + drivers/devfreq/devfreq-event.c | 35 +- drivers/devfreq/event/Kconfig | 7 + drivers/devfreq/event/Makefile | 2 + drivers/devfreq/event/hisi-uncore-ddrc.c | 160 ++++++ drivers/devfreq/event/hisi-uncore-l3c.c | 160 ++++++ drivers/devfreq/event/hisi-uncore.c | 252 +++++++++ drivers/devfreq/event/hisi-uncore.h | 77 +++ drivers/devfreq/governor_util.c | 127 +++++ drivers/devfreq/hisi_uncore_freq.c | 680 +++++++++++++++++++++++ include/linux/devfreq-event.h | 8 + include/linux/devfreq.h | 15 + 13 files changed, 1536 insertions(+), 1 deletion(-) create mode 100644 drivers/devfreq/event/hisi-uncore-ddrc.c create mode 100644 drivers/devfreq/event/hisi-uncore-l3c.c create mode 100644 drivers/devfreq/event/hisi-uncore.c create mode 100644 drivers/devfreq/event/hisi-uncore.h create mode 100644 drivers/devfreq/governor_util.c create mode 100644 drivers/devfreq/hisi_uncore_freq.c -- 2.25.1

1 10

[PATCH OLK-6.6 00/19] mm: page_alloc: freelist migratetype hygiene
by Liu Shixin 18 Dec '24

18 Dec '24

Huan Yang (1): mm: page_alloc: simpify page del and expand Johannes Weiner (11): mm: page_alloc: remove pcppage migratetype caching mm: page_alloc: optimize free_unref_folios() mm: page_alloc: fix up block types when merging compatible blocks mm: page_alloc: move free pages when converting block during isolation mm: page_alloc: fix move_freepages_block() range error mm: page_alloc: fix freelist movement during block conversion mm: page_alloc: close migratetype race between freeing and stealing mm: page_isolation: prepare for hygienic freelists mm: page_alloc: consolidate free page accounting mm: page_alloc: batch vmstat updates in expand() mm: page_alloc: fix highatomic typing in multi-block buddies Kefeng Wang (1): mm: remove migration for HugePage in isolate_single_pageblock() Kemeng Shi (2): mm/page_alloc: remove unnecessary check in break_down_buddy_pages mm/page_alloc: remove unnecessary next_page in break_down_buddy_pages Vlastimil Babka (1): mm: page_alloc: change move_freepages() to __move_freepages_block() Yajun Deng (1): mm: page_alloc: simplify __free_pages_ok() Yu Zhao (1): mm/page_alloc: keep track of free highatomic Zi Yan (1): mm: page_alloc: set migratetype inside move_freepages() include/linux/mm.h | 18 +- include/linux/mmzone.h | 1 + include/linux/page-isolation.h | 5 +- include/linux/vmstat.h | 8 - mm/debug_page_alloc.c | 12 +- mm/internal.h | 9 - mm/page_alloc.c | 735 +++++++++++++++++++-------------- mm/page_isolation.c | 135 ++---- 8 files changed, 470 insertions(+), 453 deletions(-) -- 2.34.1

2 20

[PATCH OLK-6.6 0/3] kabi OLK-6.6 reverve space
by Luo Gengkun 18 Dec '24

18 Dec '24

Luo Gengkun (3): kabi: reserve space for perf_event.h kabi: reserve space for internal.h kabi: reserve space for uprobes.h include/linux/perf_event.h | 16 ++++++++++++++++ include/linux/uprobes.h | 9 +++++++++ kernel/events/internal.h | 10 ++++++++++ 3 files changed, 35 insertions(+) -- 2.34.1

2 4

[PATCH OLK-6.6] net/kabi: Reserve space for net structures
by Dong Chenchen 18 Dec '24

18 Dec '24

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/IBC1RH -------------------------------- Reserve some fields beforehand for net subsystem related structures prone to change. Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- include/net/flow.h | 2 ++ include/net/netns/netfilter.h | 2 ++ include/net/netns/xfrm.h | 2 ++ include/net/xfrm.h | 4 ++++ 4 files changed, 10 insertions(+) diff --git a/include/net/flow.h b/include/net/flow.h index 0cc5f2ef1000..72d2ea2374ba 100644 --- a/include/net/flow.h +++ b/include/net/flow.h @@ -46,6 +46,8 @@ struct flowi_common { KABI_RESERVE(1) KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; union flowi_uli { diff --git a/include/net/netns/netfilter.h b/include/net/netns/netfilter.h index 4b77a9b031b6..963588269637 100644 --- a/include/net/netns/netfilter.h +++ b/include/net/netns/netfilter.h @@ -34,5 +34,7 @@ struct netns_nf { KABI_RESERVE(1) KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #endif diff --git a/include/net/netns/xfrm.h b/include/net/netns/xfrm.h index a0c1359cc7eb..af7f20ef4823 100644 --- a/include/net/netns/xfrm.h +++ b/include/net/netns/xfrm.h @@ -87,6 +87,8 @@ struct netns_xfrm { KABI_RESERVE(1) KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; #endif diff --git a/include/net/xfrm.h b/include/net/xfrm.h index c875faf98492..b9dec5f9c973 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -294,6 +294,8 @@ struct xfrm_state { KABI_RESERVE(1) KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static inline struct net *xs_net(struct xfrm_state *x) @@ -562,6 +564,8 @@ struct xfrm_policy { KABI_RESERVE(1) KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static inline struct net *xp_net(const struct xfrm_policy *xp) -- 2.25.1

2 1

[PATCH OLK-6.6 v2] kabi: net: reserve space for xdp subsystem related structure
by Zhang Changzhong 18 Dec '24

18 Dec '24

hulk inclusion category: other bugzilla: https://gitee.com/openeuler/kernel/issues/I8OWRC ---------------------------------------------------- Reserve some fields beforehand for xdp framework related structures prone to change. Signed-off-by: Zhang Changzhong <zhangchangzhong(a)huawei.com> --- include/net/xdp.h | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/include/net/xdp.h b/include/net/xdp.h index c283668..9b9c7dc 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -54,6 +54,9 @@ enum xdp_mem_type { struct xdp_mem_info { u32 type; /* enum xdp_mem_type, but known size type */ u32 id; + + KABI_RESERVE(1) + KABI_RESERVE(2) }; struct page_pool; @@ -74,6 +77,9 @@ struct xdp_rxq_info { struct xdp_txq_info { struct net_device *dev; + + KABI_RESERVE(1) + KABI_RESERVE(2) }; enum xdp_buff_flags { @@ -92,6 +98,11 @@ struct xdp_buff { struct xdp_txq_info *txq; u32 frame_sz; /* frame size to deduce data_hard_end/reserved tailroom*/ u32 flags; /* supported values defined in xdp_buff_flags */ + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static __always_inline bool xdp_buff_has_frags(struct xdp_buff *xdp) @@ -181,6 +192,11 @@ struct xdp_frame { struct net_device *dev_rx; /* used by cpumap */ u32 frame_sz; u32 flags; /* supported values defined in xdp_buff_flags */ + + KABI_RESERVE(1) + KABI_RESERVE(2) + KABI_RESERVE(3) + KABI_RESERVE(4) }; static __always_inline bool xdp_frame_has_frags(struct xdp_frame *frame) @@ -198,6 +214,9 @@ struct xdp_frame_bulk { int count; void *xa; void *q[XDP_BULK_QUEUE_SIZE]; + + KABI_RESERVE(1) + KABI_RESERVE(2) }; static __always_inline void xdp_frame_bulk_init(struct xdp_frame_bulk *bq) -- 2.9.5

2 1

[PATCH OLK-6.6 0/3] kabi reserve space for OLK-6.6
by Luo Gengkun 18 Dec '24

18 Dec '24

*** BLURB HERE *** Luo Gengkun (3): kabi: reserve space for perf_event.h kabi: reserve space for internal.h kabi: reserve space for uprobes.h include/linux/perf_event.h | 16 ++++++++++++++++ include/linux/uprobes.h | 9 +++++++++ kernel/events/internal.h | 10 ++++++++++ 3 files changed, 35 insertions(+) -- 2.34.1

2 4

[openeuler:OLK-5.10 2588/2588] mm/damon/core-test.h:284:9: sparse: sparse: incompatible types in comparison expression (different signedness):
by kernel test robot 18 Dec '24

18 Dec '24

tree: https://gitee.com/openeuler/kernel.git OLK-5.10 head: 6ad884d53c04817627bd02b871dd31359ce2d05c commit: 83b931be40b2829e20f38356509d8706ea6b6238 [2588/2588] mm/damon/core-test: test damon_set_regions config: x86_64-randconfig-122-20241218 (https://download.01.org/0day-ci/archive/20241218/202412181634.jaSfAZYT-lkp@…) compiler: clang version 19.1.3 (https://github.com/llvm/llvm-project ab51eccf88f5321e7c60591c5546b254b6afab99) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20241218/202412181634.jaSfAZYT-lkp@…) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp(a)intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202412181634.jaSfAZYT-lkp@intel.com/ sparse warnings: (new ones prefixed by >>) mm/damon/core.c: note: in included file (through include/linux/kernel.h, arch/x86/include/asm/percpu.h, arch/x86/include/asm/current.h, ...): /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: Expected ) in function call /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: got < /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:43: sparse: sparse: not a function <noident> /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:22: sparse: sparse: bad constant expression type mm/damon/core.c: note: in included file (through include/linux/printk.h, include/linux/kernel.h, arch/x86/include/asm/percpu.h, ...): /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: Expected ) in function call /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: got < /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:43: sparse: sparse: not a function <noident> /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:22: sparse: sparse: bad constant expression type mm/damon/core.c: note: in included file (through include/linux/string.h, include/linux/bitmap.h, include/linux/cpumask.h, ...): /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: Expected ) in function call /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:44: sparse: sparse: got < /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:43: sparse: sparse: not a function <noident> /opt/cross/clang-ab51eccf88/lib/clang/19/include/stdarg.h:22:22: sparse: sparse: bad constant expression type mm/damon/core.c: note: in included file: >> mm/damon/core-test.h:284:9: sparse: sparse: incompatible types in comparison expression (different signedness): mm/damon/core-test.h:284:9: sparse: unsigned int * mm/damon/core-test.h:284:9: sparse: int * vim +284 mm/damon/core-test.h 269 270 static void damon_test_set_regions(struct kunit *test) 271 { 272 struct damon_target *t = damon_new_target(); 273 struct damon_region *r1 = damon_new_region(4, 16); 274 struct damon_region *r2 = damon_new_region(24, 32); 275 struct damon_addr_range range = {.start = 8, .end = 28}; 276 unsigned long expects[] = {8, 16, 16, 24, 24, 28}; 277 int expect_idx = 0; 278 struct damon_region *r; 279 280 damon_add_region(r1, t); 281 damon_add_region(r2, t); 282 damon_set_regions(t, &range, 1); 283 > 284 KUNIT_EXPECT_EQ(test, damon_nr_regions(t), 3); 285 damon_for_each_region(r, t) { 286 KUNIT_EXPECT_EQ(test, r->ar.start, expects[expect_idx++]); 287 KUNIT_EXPECT_EQ(test, r->ar.end, expects[expect_idx++]); 288 } 289 damon_destroy_target(t); 290 } 291 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH OLK-6.6 0/5] KABI for cgroup
by Chen Ridong 18 Dec '24

18 Dec '24

KABI for cgroup Chen Ridong (3): Revert "cgroup: fix uaf when proc_cpuset_show" cgroup/cpuset: Prevent UAF in proc_cpuset_show() cgroup: add more reserve kabi Waiman Long (1): cgroup: Move rcu_head up near the top of cgroup_root Yafang Shao (1): cgroup: Make operations on the cgroup root_list RCU safe include/linux/cgroup-defs.h | 14 +++++++++--- include/linux/memcontrol.h | 8 +++++++ kernel/cgroup/cgroup-internal.h | 3 ++- kernel/cgroup/cgroup.c | 23 ++++++++++++++------ kernel/cgroup/cpuset.c | 38 ++++++--------------------------- 5 files changed, 44 insertions(+), 42 deletions(-) -- 2.34.1

2 6

[openeuler:OLK-6.6 1613/1613] include/linux/if_caqm.h:391:23: sparse: sparse: incorrect type in assignment (different base types)
by kernel test robot 18 Dec '24

18 Dec '24

tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: 46064348cc445799cf9e8fd030323a0426fbb5e4 commit: 946747d4a52c34156031e41f9cb5cfb292747dcf [1613/1613] net: tcp: Add CAQM4TCP in the openeuler 6.6 kernel config: loongarch-randconfig-r111-20241210 (https://download.01.org/0day-ci/archive/20241218/202412181628.dTVd4p1C-lkp@…) compiler: loongarch64-linux-gcc (GCC) 14.2.0 reproduce: (https://download.01.org/0day-ci/archive/20241218/202412181628.dTVd4p1C-lkp@…) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp(a)intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202412181628.dTVd4p1C-lkp@intel.com/ sparse warnings: (new ones prefixed by >>) net/core/gro.c: note: in included file (through include/linux/mmzone.h, include/linux/gfp.h, include/linux/xarray.h, ...): include/linux/page-flags.h:244:46: sparse: sparse: self-comparison always evaluates to false net/core/gro.c: note: in included file (through include/linux/ip.h, include/net/gro.h): include/linux/skbuff.h:2721:28: sparse: sparse: self-comparison always evaluates to false net/core/gro.c: note: in included file: >> include/linux/if_caqm.h:391:23: sparse: sparse: incorrect type in assignment (different base types) @@ expected unsigned short [usertype] caqm_hdr_info @@ got restricted __be16 [usertype] h_caqm_info @@ include/linux/if_caqm.h:391:23: sparse: expected unsigned short [usertype] caqm_hdr_info include/linux/if_caqm.h:391:23: sparse: got restricted __be16 [usertype] h_caqm_info -- net/ethernet/eth.c: note: in included file (through include/linux/mmzone.h, include/linux/gfp.h, include/linux/umh.h, include/linux/kmod.h, ...): include/linux/page-flags.h:244:46: sparse: sparse: self-comparison always evaluates to false net/ethernet/eth.c: note: in included file (through include/net/net_namespace.h, include/linux/inet.h): include/linux/skbuff.h:2721:28: sparse: sparse: self-comparison always evaluates to false net/ethernet/eth.c: note: in included file: >> include/linux/if_caqm.h:470:25: sparse: sparse: incorrect type in assignment (different base types) @@ expected unsigned short [usertype] @@ got restricted __be16 [usertype] @@ include/linux/if_caqm.h:470:25: sparse: expected unsigned short [usertype] include/linux/if_caqm.h:470:25: sparse: got restricted __be16 [usertype] vim +391 include/linux/if_caqm.h 359 360 static inline void skb_gro_caqm_untag(struct sk_buff *skb) 361 { 362 #ifdef CONFIG_ETH_CAQM 363 struct caqm_hdr *chdr; 364 struct vlan_hdr *vlanhdr; 365 u16 caqm_hdr_info; 366 __be16 proto = skb->protocol; 367 u8 skip_time = 0; 368 int mac_len, meta_len; 369 void *meta; 370 struct caqm_hdr tmp_caqm_data; 371 372 if (!static_branch_unlikely(&sysctl_caqm_enable) || skb_shared(skb)) 373 return; 374 375 /* We may access the two bytes after caqm_hdr in caqm_set_encap_proto(). */ 376 if (unlikely(!pskb_may_pull(skb, VLAN_HLEN * 2 + CAQM_HLEN + sizeof(unsigned short)))) 377 return; 378 379 vlanhdr = (struct vlan_hdr *)(skb->data - VLAN_HLEN); 380 // Skip at most 2 vlan hdr 381 while (skip_time < 2 && eth_type_vlan(proto)) { 382 vlanhdr++; // chdr move a VLAN len 383 skip_time += 1; 384 proto = vlanhdr->h_vlan_encapsulated_proto; 385 } 386 // Look for vlan header, 2 times 387 if (!eth_type_caqm(proto) || skb_cow(skb, skb_headroom(skb)) < 0) 388 return; 389 390 chdr = (struct caqm_hdr *)(((void *)vlanhdr) + VLAN_HLEN); > 391 caqm_hdr_info = chdr->h_caqm_info; 392 __caqm_put_tag(skb, caqm_hdr_info); 393 394 if ((unsigned char *) chdr != skb->data) { 395 // adjust the caqm header before vlan header 396 // 1. Set the vlanhdr->proto as the next 397 vlanhdr->h_vlan_encapsulated_proto = chdr->h_caqm_encapsulated_proto; 398 // 2. Exchange caqm and vlan hdr 399 tmp_caqm_data = *chdr; 400 memmove(skb->data + CAQM_HLEN, skb->data, skip_time * VLAN_HLEN); 401 *(struct caqm_hdr *)(skb->data) = tmp_caqm_data; 402 } else 403 caqm_set_encap_proto(skb, chdr); 404 405 skb_pull_rcsum(skb, CAQM_HLEN); 406 407 mac_len = skb->data - skb_mac_header(skb); 408 if (likely(mac_len > CAQM_HLEN + ETH_TLEN)) { 409 memmove(skb_mac_header(skb) + CAQM_HLEN, skb_mac_header(skb), 410 mac_len - CAQM_HLEN - ETH_TLEN); 411 } 412 413 meta_len = skb_metadata_len(skb); 414 if (meta_len) { 415 meta = skb_metadata_end(skb) - meta_len; 416 memmove(meta + CAQM_HLEN, meta, meta_len); 417 } 418 419 skb->mac_header += CAQM_HLEN; 420 421 skb_reset_network_header(skb); 422 if (!skb_transport_header_was_set(skb)) 423 skb_reset_transport_header(skb); 424 skb_reset_mac_len(skb); 425 426 #endif 427 } 428 429 #ifdef CONFIG_ETH_CAQM 430 static inline __be16 caqm_get_protocol_and_depth(struct sk_buff *skb, 431 __be16 type, int *depth) 432 { 433 type = vlan_get_protocol_and_depth(skb, type, depth); 434 if (static_branch_unlikely(&sysctl_caqm_enable) && eth_type_caqm(type)) 435 return caqm_get_protocol_and_depth_after_vlan(skb, type, depth); 436 else 437 return type; 438 } 439 #endif 440 441 static inline void caqm_update_hint_in_gro(struct sk_buff *skb, struct sk_buff *p) 442 { 443 #ifdef CONFIG_ETH_CAQM 444 struct skb_caqm_info *cinfo_p = get_skb_caqm_info(p); 445 struct skb_caqm_info *cinfo_skb = get_skb_caqm_info(skb); 446 447 if (static_branch_unlikely(&sysctl_caqm_enable) && cinfo_p->recv_en && cinfo_skb->recv_en) { 448 cinfo_p->recv_hint += cinfo_skb->recv_hint; 449 cinfo_skb->recv_en = 0; 450 } 451 #endif 452 } 453 454 static inline int caqm_add_eth_header(struct sk_buff *skb, unsigned short *type, 455 struct net_device *dev) 456 { 457 #ifdef CONFIG_ETH_CAQM 458 struct skb_caqm_info *cinfo = get_skb_caqm_info(skb); 459 460 if (static_branch_unlikely(&sysctl_caqm_enable) && 461 (sysctl_caqm_filter_nics & (1UL << dev->ifindex)) == 0) 462 cinfo->send_en = 0; 463 if (static_branch_unlikely(&sysctl_caqm_enable) && cinfo->send_en) { 464 cinfo->send_en = 0; 465 if (unlikely(skb_headroom(skb) < ETH_HLEN + CAQM_HLEN)) 466 return -ETH_HLEN; 467 u16 *chdr = skb_push(skb, CAQM_HLEN); 468 469 chdr[0] = (cinfo->send_hdr); > 470 chdr[1] = htons(*type); 471 *type = (unsigned short)sysctl_caqm_tpid; 472 } 473 #endif 474 return 0; 475 } 476 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH OLK-6.6] fs: Allow fine-grained control of folio sizes
by Long Li 18 Dec '24

18 Dec '24

From: "Matthew Wilcox (Oracle)" <willy(a)infradead.org> mainline inclusion from mainline-v6.10-rc2 commit 84429b675bcfd2a518ae167ee4661cdf7539aa7d category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/IBC20Q Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- We need filesystems to be able to communicate acceptable folio sizes to the pagecache for a variety of uses (e.g. large block sizes). Support a range of folio sizes between order-0 and order-31. Signed-off-by: Matthew Wilcox (Oracle) <willy(a)infradead.org> Co-developed-by: Pankaj Raghav <p.raghav(a)samsung.com> Signed-off-by: Pankaj Raghav <p.raghav(a)samsung.com> Link: https://lore.kernel.org/r/20240822135018.1931258-2-kernel@pankajraghav.com Tested-by: David Howells <dhowells(a)redhat.com> Reviewed-by: Hannes Reinecke <hare(a)suse.de> Reviewed-by: Darrick J. Wong <djwong(a)kernel.org> Reviewed-by: Daniel Gomez <da.gomez(a)samsung.com> Signed-off-by: Christian Brauner <brauner(a)kernel.org> Conflicts: include/linux/pagemap.h mm/filemap.c [Conflicts due to not merged 83ee0e20fd9f ("filemap: support disable large folios on active inode")] Signed-off-by: Long Li <leo.lilong(a)huawei.com> --- include/linux/pagemap.h | 90 +++++++++++++++++++++++++++++++++++------ mm/readahead.c | 4 +- 2 files changed, 79 insertions(+), 15 deletions(-) diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index 429627abfef4..e44e377661f2 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -203,12 +203,21 @@ enum mapping_flags { AS_EXITING = 4, /* final truncate in progress */ /* writeback related tags are not used */ AS_NO_WRITEBACK_TAGS = 5, - AS_LARGE_FOLIO_SUPPORT = 6, - AS_RELEASE_ALWAYS, /* Call ->release_folio(), even if no private data */ - AS_STABLE_WRITES, /* must wait for writeback before modifying + AS_RELEASE_ALWAYS = 6, /* Call ->release_folio(), even if no private data */ + AS_STABLE_WRITES = 7, /* must wait for writeback before modifying folio contents */ + AS_INACCESSIBLE = 8, /* Do not attempt direct R/W access to the mapping */ + /* Bits 16-25 are used for FOLIO_ORDER */ + AS_FOLIO_ORDER_BITS = 5, + AS_FOLIO_ORDER_MIN = 16, + AS_FOLIO_ORDER_MAX = AS_FOLIO_ORDER_MIN + AS_FOLIO_ORDER_BITS, }; +#define AS_FOLIO_ORDER_BITS_MASK ((1u << AS_FOLIO_ORDER_BITS) - 1) +#define AS_FOLIO_ORDER_MIN_MASK (AS_FOLIO_ORDER_BITS_MASK << AS_FOLIO_ORDER_MIN) +#define AS_FOLIO_ORDER_MAX_MASK (AS_FOLIO_ORDER_BITS_MASK << AS_FOLIO_ORDER_MAX) +#define AS_FOLIO_ORDER_MASK (AS_FOLIO_ORDER_MIN_MASK | AS_FOLIO_ORDER_MAX_MASK) + /** * mapping_set_error - record a writeback error in the address_space * @mapping: the mapping in which an error should be set @@ -348,9 +357,51 @@ static inline void mapping_set_gfp_mask(struct address_space *m, gfp_t mask) #define MAX_XAS_ORDER (XA_CHUNK_SHIFT * 2 - 1) #define MAX_PAGECACHE_ORDER min(MAX_XAS_ORDER, PREFERRED_MAX_PAGECACHE_ORDER) +/* + * mapping_set_folio_order_range() - Set the orders supported by a file. + * @mapping: The address space of the file. + * @min: Minimum folio order (between 0-MAX_PAGECACHE_ORDER inclusive). + * @max: Maximum folio order (between @min-MAX_PAGECACHE_ORDER inclusive). + * + * The filesystem should call this function in its inode constructor to + * indicate which base size (min) and maximum size (max) of folio the VFS + * can use to cache the contents of the file. This should only be used + * if the filesystem needs special handling of folio sizes (ie there is + * something the core cannot know). + * Do not tune it based on, eg, i_size. + * + * Context: This should not be called while the inode is active as it + * is non-atomic. + */ +static inline void mapping_set_folio_order_range(struct address_space *mapping, + unsigned int min, + unsigned int max) +{ + if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) + return; + + if (min > MAX_PAGECACHE_ORDER) + min = MAX_PAGECACHE_ORDER; + + if (max > MAX_PAGECACHE_ORDER) + max = MAX_PAGECACHE_ORDER; + + if (max < min) + max = min; + + mapping->flags = (mapping->flags & ~AS_FOLIO_ORDER_MASK) | + (min << AS_FOLIO_ORDER_MIN) | (max << AS_FOLIO_ORDER_MAX); +} + +static inline void mapping_set_folio_min_order(struct address_space *mapping, + unsigned int min) +{ + mapping_set_folio_order_range(mapping, min, MAX_PAGECACHE_ORDER); +} + /** * mapping_set_large_folios() - Indicate the file supports large folios. - * @mapping: The file. + * @mapping: The address space of the file. * * The filesystem should call this function in its inode constructor to * indicate that the VFS can use large folios to cache the contents of @@ -361,7 +412,23 @@ static inline void mapping_set_gfp_mask(struct address_space *m, gfp_t mask) */ static inline void mapping_set_large_folios(struct address_space *mapping) { - __set_bit(AS_LARGE_FOLIO_SUPPORT, &mapping->flags); + mapping_set_folio_order_range(mapping, 0, MAX_PAGECACHE_ORDER); +} + +static inline unsigned int +mapping_max_folio_order(const struct address_space *mapping) +{ + if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) + return 0; + return (mapping->flags & AS_FOLIO_ORDER_MAX_MASK) >> AS_FOLIO_ORDER_MAX; +} + +static inline unsigned int +mapping_min_folio_order(const struct address_space *mapping) +{ + if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE)) + return 0; + return (mapping->flags & AS_FOLIO_ORDER_MIN_MASK) >> AS_FOLIO_ORDER_MIN; } /** @@ -375,7 +442,7 @@ static inline void mapping_set_large_folios(struct address_space *mapping) static inline void mapping_clear_large_folios(struct address_space *mapping) { WARN_ON_ONCE(!rwsem_is_locked(&mapping->invalidate_lock)); - __clear_bit(AS_LARGE_FOLIO_SUPPORT, &mapping->flags); + mapping_set_folio_order_range(mapping, 0, 0); } /* @@ -384,20 +451,17 @@ static inline void mapping_clear_large_folios(struct address_space *mapping) */ static inline bool mapping_large_folio_support(struct address_space *mapping) { - /* AS_LARGE_FOLIO_SUPPORT is only reasonable for pagecache folios */ + /* AS_FOLIO_ORDER is only reasonable for pagecache folios */ VM_WARN_ONCE((unsigned long)mapping & PAGE_MAPPING_ANON, "Anonymous mapping always supports large folio"); - return IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE) && - test_bit(AS_LARGE_FOLIO_SUPPORT, &mapping->flags); + return mapping_max_folio_order(mapping) > 0; } /* Return the maximum folio size for this pagecache mapping, in bytes. */ -static inline size_t mapping_max_folio_size(struct address_space *mapping) +static inline size_t mapping_max_folio_size(const struct address_space *mapping) { - if (mapping_large_folio_support(mapping)) - return PAGE_SIZE << MAX_PAGECACHE_ORDER; - return PAGE_SIZE; + return PAGE_SIZE << mapping_max_folio_order(mapping); } static inline int filemap_nr_thps(struct address_space *mapping) diff --git a/mm/readahead.c b/mm/readahead.c index 438f142a3e74..c13c130efcca 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -513,10 +513,10 @@ void page_cache_ra_order(struct readahead_control *ractl, limit = min(limit, index + ra->size - 1); - if (new_order < MAX_PAGECACHE_ORDER) + if (new_order < mapping_max_folio_order(mapping)) new_order += 2; - new_order = min_t(unsigned int, MAX_PAGECACHE_ORDER, new_order); + new_order = min(mapping_max_folio_order(mapping), new_order); new_order = min_t(unsigned int, new_order, ilog2(ra->size)); /* See comment in page_cache_ra_unbounded() */ -- 2.39.2

2 1

2025

2024

2023

2022

2021

2020

2019

Kernel