- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] scsi: ses: Fix slab-out-of-bounds in ses_enclosure_data_process()
by Li Lingfeng 10 Dec '25

10 Dec '25

From: Tomas Henzl <thenzl(a)redhat.com> stable inclusion from stable-v4.19.276 commit 9e5c7d52085b8c84bc82a261580f0eb170039325 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IDBMX7 CVE: CVE-2023-53803 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 9b4f5028e493cb353a5c8f5c45073eeea0303abd upstream. A fix for: BUG: KASAN: slab-out-of-bounds in ses_enclosure_data_process+0x949/0xe30 [ses] Read of size 1 at addr ffff88a1b043a451 by task systemd-udevd/3271 Checking after (and before in next loop) addl_desc_ptr[1] is sufficient, we expect the size to be sanitized before first access to addl_desc_ptr[1]. Make sure we don't walk beyond end of page. Link: https://lore.kernel.org/r/20230202162451.15346-2-thenzl@redhat.com Cc: stable(a)vger.kernel.org Signed-off-by: Tomas Henzl <thenzl(a)redhat.com> Signed-off-by: Martin K. Petersen <martin.petersen(a)oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/scsi/ses.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/ses.c b/drivers/scsi/ses.c index b42c0bb8caa9..3a0cb1fa6767 100644 --- a/drivers/scsi/ses.c +++ b/drivers/scsi/ses.c @@ -625,9 +625,11 @@ static void ses_enclosure_data_process(struct enclosure_device *edev, /* these elements are optional */ type_ptr[0] == ENCLOSURE_COMPONENT_SCSI_TARGET_PORT || type_ptr[0] == ENCLOSURE_COMPONENT_SCSI_INITIATOR_PORT || - type_ptr[0] == ENCLOSURE_COMPONENT_CONTROLLER_ELECTRONICS)) + type_ptr[0] == ENCLOSURE_COMPONENT_CONTROLLER_ELECTRONICS)) { addl_desc_ptr += addl_desc_ptr[1] + 2; - + if (addl_desc_ptr + 1 >= ses_dev->page10 + ses_dev->page10_len) + addl_desc_ptr = NULL; + } } } kfree(buf); -- 2.31.1

2 1

[PATCH OLK-6.6 0/2] Handling Kunpeng Chip ARM Spectre-BHB (CVE-2022-23960) Issue
by Jinqian Yang 10 Dec '25

10 Dec '25

Previously, the mitigation for HIP09 had been incorporated into 6.6. Now, the mitigation for TSV100 is being incorporated into 6.6. And non-privileged eBPF is disabled by default. In addition, the GRUB in the distribution needs to include 'nospectre-bhb'. Jinqian Yang (2): arm64: Add support for TSV110 Spectre-BHB mitigation config: Disable unprivileged BPF syscall by default arch/arm64/configs/openeuler_defconfig | 2 +- arch/arm64/kernel/proton-pack.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) -- 2.33.0

2 3

[openeuler:OLK-6.6 3508/3508] mm/memblock.c:1409:20: warning: no previous prototype for 'memblock_alloc_range_nid_flags'
by kernel test robot 10 Dec '25

10 Dec '25

Hi Ma, FYI, the error/warning still remains. tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: d7d7f131569485a9b6737b85cd8144196135b1a1 commit: 64018b291c1f49622c4b23b303364d760306d662 [3508/3508] mm/memblock: Introduce ability to alloc memory from specify memory region config: x86_64-allnoconfig-bpf (https://download.01.org/0day-ci/archive/20251210/202512100803.QFVpHYI3-lkp@…) compiler: gcc-14 (Debian 14.2.0-19) 14.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251210/202512100803.QFVpHYI3-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/202512100803.QFVpHYI3-lkp@intel.com/ All warnings (new ones prefixed by >>): >> mm/memblock.c:1409:20: warning: no previous prototype for 'memblock_alloc_range_nid_flags' [-Wmissing-prototypes] 1409 | phys_addr_t __init memblock_alloc_range_nid_flags(phys_addr_t size, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ vim +/memblock_alloc_range_nid_flags +1409 mm/memblock.c 1386 1387 /** 1388 * memblock_alloc_range_nid_flags - allocate boot memory block with specify flag 1389 * @size: size of memory block to be allocated in bytes 1390 * @align: alignment of the region and block's size 1391 * @start: the lower bound of the memory region to allocate (phys address) 1392 * @end: the upper bound of the memory region to allocate (phys address) 1393 * @nid: nid of the free area to find, %NUMA_NO_NODE for any node 1394 * @exact_nid: control the allocation fall back to other nodes 1395 * @flags: alloc memory from specify memblock flag 1396 * 1397 * The allocation is performed from memory region limited by 1398 * memblock.current_limit if @end == %MEMBLOCK_ALLOC_ACCESSIBLE. 1399 * 1400 * If the specified node can not hold the requested memory and @exact_nid 1401 * is false, the allocation falls back to any node in the system. 1402 * 1403 * In addition, function sets the min_count to 0 using kmemleak_alloc_phys for 1404 * allocated boot memory block, so that it is never reported as leaks. 1405 * 1406 * Return: 1407 * Physical address of allocated memory block on success, %0 on failure. 1408 */ > 1409 phys_addr_t __init memblock_alloc_range_nid_flags(phys_addr_t size, 1410 phys_addr_t align, phys_addr_t start, 1411 phys_addr_t end, int nid, 1412 bool exact_nid, 1413 enum memblock_flags flags) 1414 { 1415 phys_addr_t found; 1416 1417 if (WARN_ONCE(nid == MAX_NUMNODES, "Usage of MAX_NUMNODES is deprecated. Use NUMA_NO_NODE instead\n")) 1418 nid = NUMA_NO_NODE; 1419 1420 if (!align) { 1421 /* Can't use WARNs this early in boot on powerpc */ 1422 dump_stack(); 1423 align = SMP_CACHE_BYTES; 1424 } 1425 1426 again: 1427 found = memblock_find_in_range_node(size, align, start, end, nid, 1428 flags); 1429 if (found && !memblock_reserve(found, size)) 1430 goto done; 1431 1432 if (nid != NUMA_NO_NODE && !exact_nid) { 1433 found = memblock_find_in_range_node(size, align, start, 1434 end, NUMA_NO_NODE, 1435 flags); 1436 if (found && !memblock_reserve(found, size)) 1437 goto done; 1438 } 1439 1440 if (flags & MEMBLOCK_MIRROR) { 1441 flags &= ~MEMBLOCK_MIRROR; 1442 pr_warn_ratelimited("Could not allocate %pap bytes of mirrored memory\n", 1443 &size); 1444 goto again; 1445 } 1446 1447 return 0; 1448 1449 done: 1450 /* 1451 * Skip kmemleak for those places like kasan_init() and 1452 * early_pgtable_alloc() due to high volume. 1453 */ 1454 if (end != MEMBLOCK_ALLOC_NOLEAKTRACE) 1455 /* 1456 * Memblock allocated blocks are never reported as 1457 * leaks. This is because many of these blocks are 1458 * only referred via the physical address which is 1459 * not looked up by kmemleak. 1460 */ 1461 kmemleak_alloc_phys(found, size, 0); 1462 1463 /* 1464 * Some Virtual Machine platforms, such as Intel TDX or AMD SEV-SNP, 1465 * require memory to be accepted before it can be used by the 1466 * guest. 1467 * 1468 * Accept the memory of the allocated buffer. 1469 */ 1470 accept_memory(found, found + size); 1471 1472 return found; 1473 } 1474 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[openeuler:OLK-6.6 3508/3508] arch/arm64/kvm/virtcca_cvm.c:992:5: warning: no previous prototype for 'kvm_cvm_vgic_nr_lr'
by kernel test robot 10 Dec '25

10 Dec '25

tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: d7d7f131569485a9b6737b85cd8144196135b1a1 commit: befbde867291b5bb068a97b17086ad20484b6068 [3508/3508] VirtCCA: Adapt virtcca operations by cca_base. config: arm64-randconfig-r112-20251210 (https://download.01.org/0day-ci/archive/20251210/202512101452.vWsWIlEF-lkp@…) compiler: aarch64-linux-gcc (GCC) 13.4.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251210/202512101452.vWsWIlEF-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/202512101452.vWsWIlEF-lkp@intel.com/ All warnings (new ones prefixed by >>): arch/arm64/kvm/virtcca_cvm.c:825:5: warning: no previous prototype for 'virtcca_get_tmi_version' [-Wmissing-prototypes] 825 | u64 virtcca_get_tmi_version(void) | ^~~~~~~~~~~~~~~~~~~~~~~ arch/arm64/kvm/virtcca_cvm.c: In function 'kvm_load_user_data': arch/arm64/kvm/virtcca_cvm.c:885:35: error: 'struct kvm_user_data' has no member named 'data_start' 885 | (user_data.data_start + user_data.data_size < user_data.data_start)) | ^ arch/arm64/kvm/virtcca_cvm.c:885:59: error: 'struct kvm_user_data' has no member named 'data_size'; did you mean 'ram_size'? 885 | (user_data.data_start + user_data.data_size < user_data.data_start)) | ^~~~~~~~~ | ram_size arch/arm64/kvm/virtcca_cvm.c:885:80: error: 'struct kvm_user_data' has no member named 'data_start' 885 | (user_data.data_start + user_data.data_size < user_data.data_start)) | ^ arch/arm64/kvm/virtcca_cvm.c:888:58: error: 'struct kvm_user_data' has no member named 'dtb_info' 888 | if (is_dtb_info_has_extend_data(user_data.dtb_info)) { | ^ arch/arm64/kvm/virtcca_cvm.c:890:38: error: 'struct kvm_user_data' has no member named 'data_start' 890 | if (user_data.data_start + user_data.data_size > ipa_end) | ^ arch/arm64/kvm/virtcca_cvm.c:890:62: error: 'struct kvm_user_data' has no member named 'data_size'; did you mean 'ram_size'? 890 | if (user_data.data_start + user_data.data_size > ipa_end) | ^~~~~~~~~ | ram_size arch/arm64/kvm/virtcca_cvm.c:897:38: error: 'struct kvm_user_data' has no member named 'data_start' 897 | if (user_data.data_start + user_data.data_size > user_data.loader_start) | ^ arch/arm64/kvm/virtcca_cvm.c:897:62: error: 'struct kvm_user_data' has no member named 'data_size'; did you mean 'ram_size'? 897 | if (user_data.data_start + user_data.data_size > user_data.loader_start) | ^~~~~~~~~ | ram_size arch/arm64/kvm/virtcca_cvm.c:907:50: error: 'struct kvm_user_data' has no member named 'dtb_info' 907 | if (is_dtb_info_has_extend_data(user_data.dtb_info)) | ^ arch/arm64/kvm/virtcca_cvm.c:908:41: error: 'struct kvm_user_data' has no member named 'data_start' 908 | cvm->dtb_end = user_data.data_start + user_data.data_size; | ^ arch/arm64/kvm/virtcca_cvm.c:908:65: error: 'struct kvm_user_data' has no member named 'data_size'; did you mean 'ram_size'? 908 | cvm->dtb_end = user_data.data_start + user_data.data_size; | ^~~~~~~~~ | ram_size arch/arm64/kvm/virtcca_cvm.c:910:66: error: 'struct kvm_user_data' has no member named 'dtb_info' 910 | cvm->dtb_end = user_data.loader_start + user_data.dtb_info; | ^ arch/arm64/kvm/virtcca_cvm.c:911:44: error: 'struct kvm_user_data' has no member named 'data_start' 911 | cvm->mmio_start = user_data.data_start; | ^ arch/arm64/kvm/virtcca_cvm.c:912:42: error: 'struct kvm_user_data' has no member named 'data_start' 912 | cvm->mmio_end = user_data.data_start + user_data.data_size; | ^ arch/arm64/kvm/virtcca_cvm.c:912:66: error: 'struct kvm_user_data' has no member named 'data_size'; did you mean 'ram_size'? 912 | cvm->mmio_end = user_data.data_start + user_data.data_size; | ^~~~~~~~~ | ram_size arch/arm64/kvm/virtcca_cvm.c: At top level: >> arch/arm64/kvm/virtcca_cvm.c:992:5: warning: no previous prototype for 'kvm_cvm_vgic_nr_lr' [-Wmissing-prototypes] 992 | u32 kvm_cvm_vgic_nr_lr(void) | ^~~~~~~~~~~~~~~~~~ Kconfig warnings: (for reference only) WARNING: unmet direct dependencies detected for HALTPOLL_CPUIDLE Depends on [n]: CPU_IDLE [=n] && ARCH_CPUIDLE_HALTPOLL [=y] && ARCH_HAS_OPTIMIZED_POLL [=y] Selected by [y]: - ARM64 [=y] vim +/kvm_cvm_vgic_nr_lr +992 arch/arm64/kvm/virtcca_cvm.c 991 > 992 u32 kvm_cvm_vgic_nr_lr(void) 993 { 994 return kvm_vgic_global_state.nr_lr; 995 } 996 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH openEuler-1.0-LTS] NFSD: Fix crash in nfsd4_read_release()
by Li Lingfeng 10 Dec '25

10 Dec '25

From: Chuck Lever <chuck.lever(a)oracle.com> mainline inclusion from mainline-v6.18-rc4 commit abb1f08a2121dd270193746e43b2a9373db9ad84 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IDBAJB CVE: CVE-2025-40324 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When tracing is enabled, the trace_nfsd_read_done trace point crashes during the pynfs read.testNoFh test. Fixes: 15a8b55dbb1b ("nfsd: call op_release, even when op_func returns an error") Cc: stable(a)vger.kernel.org Reviewed-by: Jeff Layton <jlayton(a)kernel.org> Signed-off-by: Chuck Lever <chuck.lever(a)oracle.com> Conflicts: fs/nfsd/nfs4proc.c [Commit 5c4583b2b78e ("nfsd: hook up nfs4_preprocess_stateid_op to the nfsd_file cache") replace rd_filp by rd_nf.] Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/nfsd/nfs4proc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 1c3e6de6bcba..ea3db7d8b567 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -796,10 +796,11 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, static void nfsd4_read_release(union nfsd4_op_u *u) { - if (u->read.rd_filp) + if (u->read.rd_filp) { + trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, + u->read.rd_offset, u->read.rd_length); fput(u->read.rd_filp); - trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, - u->read.rd_offset, u->read.rd_length); + } } static __be32 -- 2.31.1

2 1

[PATCH OLK-5.10] NFSD: Fix crash in nfsd4_read_release()
by Li Lingfeng 10 Dec '25

10 Dec '25

From: Chuck Lever <chuck.lever(a)oracle.com> stable inclusion from stable-v5.10.247 commit 930cb4fe3ab4061be31f20ee30bb72a66f7bb6d1 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IDBAJB CVE: CVE-2025-40324 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit abb1f08a2121dd270193746e43b2a9373db9ad84 upstream. When tracing is enabled, the trace_nfsd_read_done trace point crashes during the pynfs read.testNoFh test. Fixes: 15a8b55dbb1b ("nfsd: call op_release, even when op_func returns an error") Cc: stable(a)vger.kernel.org Reviewed-by: Jeff Layton <jlayton(a)kernel.org> Signed-off-by: Chuck Lever <chuck.lever(a)oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/nfsd/nfs4proc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 96c4e2616b51..61c87c54cd7b 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -812,10 +812,11 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, static void nfsd4_read_release(union nfsd4_op_u *u) { - if (u->read.rd_nf) + if (u->read.rd_nf) { + trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, + u->read.rd_offset, u->read.rd_length); nfsd_file_put(u->read.rd_nf); - trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, - u->read.rd_offset, u->read.rd_length); + } } static __be32 -- 2.46.1

2 1

[PATCH OLK-6.6] NFSD: Fix crash in nfsd4_read_release()
by Li Lingfeng 10 Dec '25

10 Dec '25

From: Chuck Lever <chuck.lever(a)oracle.com> stable inclusion from stable-v6.6.117 commit 03524ccff698d4a77d096ed529073d91f5edee5d category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IDBAJB CVE: CVE-2025-40324 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit abb1f08a2121dd270193746e43b2a9373db9ad84 upstream. When tracing is enabled, the trace_nfsd_read_done trace point crashes during the pynfs read.testNoFh test. Fixes: 15a8b55dbb1b ("nfsd: call op_release, even when op_func returns an error") Cc: stable(a)vger.kernel.org Reviewed-by: Jeff Layton <jlayton(a)kernel.org> Signed-off-by: Chuck Lever <chuck.lever(a)oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/nfsd/nfs4proc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index b2bbf3d6d177..04a2c44e0f95 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -978,10 +978,11 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, static void nfsd4_read_release(union nfsd4_op_u *u) { - if (u->read.rd_nf) + if (u->read.rd_nf) { + trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, + u->read.rd_offset, u->read.rd_length); nfsd_file_put(u->read.rd_nf); - trace_nfsd_read_done(u->read.rd_rqstp, u->read.rd_fhp, - u->read.rd_offset, u->read.rd_length); + } } static __be32 -- 2.46.1

2 1

[PATCH OLK-6.6 0/8] xsched: XCU Partition
by Zicheng Qu 10 Dec '25

10 Dec '25

Zicheng Qu (8): xsched: refactor CFS per-XCU rq init/deinit paths xsched: fix NULL pointer risk on attach entry allocation xsched: fix concurrent modification NULL pointer dereference in xcu.sched_class xsched: fix hung_task caused by cgroup_file_show and xcg_mutex deadlock xsched: prevent NULL deref by refcounting css and tracking offline state xsched: fix divide-by-zero caused by u64 overflow in CFS shares weight calculation xsched: DEBUG log cleanup xsched: prevent race between enqueue_ctx and xsched_task_free include/linux/xsched.h | 36 +++++- kernel/xsched/cfs.c | 31 +++++- kernel/xsched/cfs_quota.c | 9 ++ kernel/xsched/cgroup.c | 224 +++++++++++++++++++++----------------- kernel/xsched/core.c | 9 +- kernel/xsched/rt.c | 19 +--- kernel/xsched/vstream.c | 3 + 7 files changed, 200 insertions(+), 131 deletions(-) -- 2.34.1

2 9

[openeuler:OLK-5.10 3399/3399] block/bio.c:412: warning: Function parameter or member 'nr_iovecs_int' not described in 'bio_alloc_bioset'
by kernel test robot 10 Dec '25

10 Dec '25

Hi Li, FYI, the error/warning still remains. tree: https://gitee.com/openeuler/kernel.git OLK-5.10 head: 7579fd89cb4a201d22bd5b78f48df5ce339c0b2d commit: 6381672514c4d94637cd4db830b8644440abe99a [3399/3399] block: fix kabi broken config: x86_64-allnoconfig-bpf (https://download.01.org/0day-ci/archive/20251210/202512100653.S0mIExGw-lkp@…) compiler: gcc-14 (Debian 14.2.0-19) 14.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251210/202512100653.S0mIExGw-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/202512100653.S0mIExGw-lkp@intel.com/ All warnings (new ones prefixed by >>): >> block/bio.c:412: warning: Function parameter or member 'nr_iovecs_int' not described in 'bio_alloc_bioset' >> block/bio.c:412: warning: Excess function parameter 'nr_iovecs' description in 'bio_alloc_bioset' vim +412 block/bio.c df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 374 ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 375 /** ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 376 * bio_alloc_bioset - allocate a bio for I/O 519c8e9ffd8614 block/bio.c Randy Dunlap 2017-10-16 377 * @gfp_mask: the GFP_* mask given to the slab allocator ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 378 * @nr_iovecs: number of iovecs to pre-allocate db18efac0bd644 fs/bio.c Jaak Ristioja 2010-01-15 379 * @bs: the bio_set to allocate from. ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 380 * ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 381 * Description: 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 382 * If @bs is NULL, uses kmalloc() to allocate the bio; else the allocation is 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 383 * backed by the @bs's mempool. 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 384 * d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 385 * When @bs is not NULL, if %__GFP_DIRECT_RECLAIM is set then bio_alloc will d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 386 * always be able to allocate a bio. This is due to the mempool guarantees. d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 387 * To make this work, callers must never allocate more than 1 bio at a time d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 388 * from this pool. Callers that need to allocate more than 1 bio must always d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 389 * submit the previously allocated bio for IO before attempting to allocate d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 390 * a new one. Failure to do so can cause deadlocks under memory pressure. 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 391 * ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 392 * Note that when running under submit_bio_noacct() (i.e. any block df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 393 * driver), bios are not submitted until after you return - see the code in ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 394 * submit_bio_noacct() that converts recursion into iteration, to prevent df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 395 * stack overflows. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 396 * df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 397 * This would normally mean allocating multiple bios under ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 398 * submit_bio_noacct() would be susceptible to deadlocks, but we have df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 399 * deadlock avoidance code that resubmits any blocked bios from a rescuer df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 400 * thread. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 401 * df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 402 * However, we do not guarantee forward progress for allocations from other df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 403 * mempools. Doing multiple allocations from the same mempool under ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 404 * submit_bio_noacct() should be avoided - instead, use bio_set's front_pad df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 405 * for per bio allocations. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 406 * 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 407 * RETURNS: 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 408 * Pointer to new bio on success, NULL on failure. 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 409 */ 6381672514c4d9 block/bio.c Li Lingfeng 2025-02-10 410 struct bio *bio_alloc_bioset(gfp_t gfp_mask, unsigned int nr_iovecs_int, 7a88fa19194458 block/bio.c Dan Carpenter 2017-03-23 411 struct bio_set *bs) ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 @412 { df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 413 gfp_t saved_gfp = gfp_mask; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 414 unsigned front_pad; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 415 unsigned inline_vecs; 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 416 struct bio_vec *bvl = NULL; 451a9ebf653d28 fs/bio.c Tejun Heo 2009-04-15 417 struct bio *bio; 451a9ebf653d28 fs/bio.c Tejun Heo 2009-04-15 418 void *p; 6381672514c4d9 block/bio.c Li Lingfeng 2025-02-10 419 unsigned short nr_iovecs = (unsigned short)nr_iovecs_int; 0a0d96b03a1f3b fs/bio.c Jens Axboe 2008-09-11 420 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 421 if (!bs) { 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 422 if (nr_iovecs > UIO_MAXIOV) 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 423 return NULL; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 424 1f4fe21cf45c79 block/bio.c Gustavo A. R. Silva 2020-06-19 425 p = kmalloc(struct_size(bio, bi_inline_vecs, nr_iovecs), gfp_mask); 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 426 front_pad = 0; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 427 inline_vecs = nr_iovecs; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 428 } else { d8f429e1669b97 block/bio.c Junichi Nomura 2014-10-03 429 /* should not use nobvec bioset for nr_iovecs > 0 */ 8aa6ba2f6e3dea block/bio.c Kent Overstreet 2018-05-08 430 if (WARN_ON_ONCE(!mempool_initialized(&bs->bvec_pool) && 8aa6ba2f6e3dea block/bio.c Kent Overstreet 2018-05-08 431 nr_iovecs > 0)) d8f429e1669b97 block/bio.c Junichi Nomura 2014-10-03 432 return NULL; df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 433 /* ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 434 * submit_bio_noacct() converts recursion to iteration; this df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 435 * means if we're running beneath it, any bios we allocate and df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 436 * submit will not be submitted (and thus freed) until after we df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 437 * return. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 438 * df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 439 * This exposes us to a potential deadlock if we allocate df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 440 * multiple bios from the same bio_set() while running ed00aabd5eb9fb block/bio.c Christoph Hellwig 2020-07-01 441 * underneath submit_bio_noacct(). If we were to allocate df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 442 * multiple bios (say a stacking block driver that was splitting df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 443 * bios), we would deadlock if we exhausted the mempool's df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 444 * reserve. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 445 * df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 446 * We solve this, and guarantee forward progress, with a rescuer df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 447 * workqueue per bio_set. If we go to allocate and there are df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 448 * bios on current->bio_list, we first try the allocation d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 449 * without __GFP_DIRECT_RECLAIM; if that fails, we punt those d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 450 * bios we would be blocking to the rescuer workqueue before d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 451 * we retry with the original gfp_flags. df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 452 */ df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 453 f5fe1b51905df7 block/bio.c NeilBrown 2017-03-10 454 if (current->bio_list && f5fe1b51905df7 block/bio.c NeilBrown 2017-03-10 455 (!bio_list_empty(&current->bio_list[0]) || 47e0fb461fca1a block/bio.c NeilBrown 2017-06-18 456 !bio_list_empty(&current->bio_list[1])) && 47e0fb461fca1a block/bio.c NeilBrown 2017-06-18 457 bs->rescue_workqueue) d0164adc89f6bb block/bio.c Mel Gorman 2015-11-06 458 gfp_mask &= ~__GFP_DIRECT_RECLAIM; df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 459 8aa6ba2f6e3dea block/bio.c Kent Overstreet 2018-05-08 460 p = mempool_alloc(&bs->bio_pool, gfp_mask); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 461 if (!p && gfp_mask != saved_gfp) { df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 462 punt_bios_to_rescuer(bs); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 463 gfp_mask = saved_gfp; 8aa6ba2f6e3dea block/bio.c Kent Overstreet 2018-05-08 464 p = mempool_alloc(&bs->bio_pool, gfp_mask); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 465 } df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 466 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 467 front_pad = bs->front_pad; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 468 inline_vecs = BIO_INLINE_VECS; 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 469 } 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 470 451a9ebf653d28 fs/bio.c Tejun Heo 2009-04-15 471 if (unlikely(!p)) 451a9ebf653d28 fs/bio.c Tejun Heo 2009-04-15 472 return NULL; ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 473 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 474 bio = p + front_pad; 3a83f4677539bc block/bio.c Ming Lei 2016-11-22 475 bio_init(bio, NULL, 0); 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 476 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 477 if (nr_iovecs > inline_vecs) { ed996a52c868b6 block/bio.c Christoph Hellwig 2016-07-19 478 958384da6a734e block/bio.c Christoph Hellwig 2025-02-10 479 bvl = bvec_alloc(&bs->bvec_pool, &nr_iovecs, gfp_mask); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 480 if (!bvl && gfp_mask != saved_gfp) { df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 481 punt_bios_to_rescuer(bs); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 482 gfp_mask = saved_gfp; 958384da6a734e block/bio.c Christoph Hellwig 2025-02-10 483 bvl = bvec_alloc(&bs->bvec_pool, &nr_iovecs, gfp_mask); df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 484 } df2cb6daa4cbc3 fs/bio.c Kent Overstreet 2012-09-10 485 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 486 if (unlikely(!bvl)) 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 487 goto err_free; a38352e0ac02db fs/bio.c Kent Overstreet 2012-05-25 488 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 489 } else if (nr_iovecs) { 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 490 bvl = bio->bi_inline_vecs; 392ddc32982a5c fs/bio.c Jens Axboe 2008-12-23 491 } 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 492 3f86a82aeb03e6 fs/bio.c Kent Overstreet 2012-09-06 493 bio->bi_pool = bs; 392ddc32982a5c fs/bio.c Jens Axboe 2008-12-23 494 bio->bi_max_vecs = nr_iovecs; ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 495 bio->bi_io_vec = bvl; ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 496 return bio; 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 497 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 498 err_free: 8aa6ba2f6e3dea block/bio.c Kent Overstreet 2018-05-08 499 mempool_free(p, &bs->bio_pool); 34053979fb1d92 fs/bio.c Ingo Molnar 2009-02-21 500 return NULL; ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 501 } a112a71d45b5e4 fs/bio.c H Hartley Sweeten 2009-09-26 502 EXPORT_SYMBOL(bio_alloc_bioset); ^1da177e4c3f41 fs/bio.c Linus Torvalds 2005-04-16 503 :::::: The code at line 412 was first introduced by commit :::::: 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2 Linux-2.6.12-rc2 :::::: TO: Linus Torvalds <torvalds(a)ppc970.osdl.org> :::::: CC: Linus Torvalds <torvalds(a)ppc970.osdl.org> -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[openeuler:OLK-6.6 3508/3508] mm/madvise.c:297:6: warning: no previous prototype for 'force_swapin_vma'
by kernel test robot 10 Dec '25

10 Dec '25

Hi Liu, FYI, the error/warning still remains. tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: d7d7f131569485a9b6737b85cd8144196135b1a1 commit: 92a0eb9bde6c03412b39f9f3d20968c091ea3b46 [3508/3508] memcg: introduce per-memcg swapin interface config: x86_64-allnoconfig-bpf (https://download.01.org/0day-ci/archive/20251210/202512100526.hg7amXND-lkp@…) compiler: gcc-14 (Debian 14.2.0-19) 14.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251210/202512100526.hg7amXND-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/202512100526.hg7amXND-lkp@intel.com/ All warnings (new ones prefixed by >>): >> mm/madvise.c:297:6: warning: no previous prototype for 'force_swapin_vma' [-Wmissing-prototypes] 297 | void force_swapin_vma(struct vm_area_struct *vma) | ^~~~~~~~~~~~~~~~ vim +/force_swapin_vma +297 mm/madvise.c 280 281 #ifdef CONFIG_MEMCG_SWAP_QOS 282 void force_swapin_vma(struct vm_area_struct *vma) 283 { 284 struct file *file = vma->vm_file; 285 286 if (!can_madv_lru_vma(vma)) 287 return; 288 289 if (!file) { 290 walk_page_vma(vma, &swapin_walk_ops, vma); 291 lru_add_drain(); 292 } else if (shmem_mapping(file->f_mapping)) 293 shmem_swapin_range(vma, vma->vm_start, 294 vma->vm_end, file->f_mapping); 295 } 296 #else > 297 void force_swapin_vma(struct vm_area_struct *vma) 298 { 299 } 300 #endif 301 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0