- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS 01/45] af_unix: fix garbage collect vs MSG_PEEK
by Yang Yingliang 25 Oct '21

25 Oct '21

From: Miklos Szeredi <mszeredi(a)redhat.com> stable inclusion from linux-4.19.200 commit 1dabafa9f61118b1377fde424d9a94bf8dbf2813 -------------------------------- commit cbcf01128d0a92e131bd09f1688fe032480b65ca upstream. unix_gc() assumes that candidate sockets can never gain an external reference (i.e. be installed into an fd) while the unix_gc_lock is held. Except for MSG_PEEK this is guaranteed by modifying inflight count under the unix_gc_lock. MSG_PEEK does not touch any variable protected by unix_gc_lock (file count is not), yet it needs to be serialized with garbage collection. Do this by locking/unlocking unix_gc_lock: 1) increment file count 2) lock/unlock barrier to make sure incremented file count is visible to garbage collection 3) install file into fd This is a lock barrier (unlike smp_mb()) that ensures that garbage collection is run completely before or completely after the barrier. Cc: <stable(a)vger.kernel.org> Signed-off-by: Miklos Szeredi <mszeredi(a)redhat.com> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- net/unix/af_unix.c | 51 ++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 49 insertions(+), 2 deletions(-) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 337c4797ab167..98c253afa0db2 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -1517,6 +1517,53 @@ static int unix_getname(struct socket *sock, struct sockaddr *uaddr, int peer) return err; } +static void unix_peek_fds(struct scm_cookie *scm, struct sk_buff *skb) +{ + scm->fp = scm_fp_dup(UNIXCB(skb).fp); + + /* + * Garbage collection of unix sockets starts by selecting a set of + * candidate sockets which have reference only from being in flight + * (total_refs == inflight_refs). This condition is checked once during + * the candidate collection phase, and candidates are marked as such, so + * that non-candidates can later be ignored. While inflight_refs is + * protected by unix_gc_lock, total_refs (file count) is not, hence this + * is an instantaneous decision. + * + * Once a candidate, however, the socket must not be reinstalled into a + * file descriptor while the garbage collection is in progress. + * + * If the above conditions are met, then the directed graph of + * candidates (*) does not change while unix_gc_lock is held. + * + * Any operations that changes the file count through file descriptors + * (dup, close, sendmsg) does not change the graph since candidates are + * not installed in fds. + * + * Dequeing a candidate via recvmsg would install it into an fd, but + * that takes unix_gc_lock to decrement the inflight count, so it's + * serialized with garbage collection. + * + * MSG_PEEK is special in that it does not change the inflight count, + * yet does install the socket into an fd. The following lock/unlock + * pair is to ensure serialization with garbage collection. It must be + * done between incrementing the file count and installing the file into + * an fd. + * + * If garbage collection starts after the barrier provided by the + * lock/unlock, then it will see the elevated refcount and not mark this + * as a candidate. If a garbage collection is already in progress + * before the file count was incremented, then the lock/unlock pair will + * ensure that garbage collection is finished before progressing to + * installing the fd. + * + * (*) A -> B where B is on the queue of A or B is on the queue of C + * which is on the queue of listening socket A. + */ + spin_lock(&unix_gc_lock); + spin_unlock(&unix_gc_lock); +} + static int unix_scm_to_skb(struct scm_cookie *scm, struct sk_buff *skb, bool send_fds) { int err = 0; @@ -2142,7 +2189,7 @@ static int unix_dgram_recvmsg(struct socket *sock, struct msghdr *msg, sk_peek_offset_fwd(sk, size); if (UNIXCB(skb).fp) - scm.fp = scm_fp_dup(UNIXCB(skb).fp); + unix_peek_fds(&scm, skb); } err = (flags & MSG_TRUNC) ? skb->len - skip : size; @@ -2383,7 +2430,7 @@ static int unix_stream_read_generic(struct unix_stream_read_state *state, /* It is questionable, see note in unix_dgram_recvmsg. */ if (UNIXCB(skb).fp) - scm.fp = scm_fp_dup(UNIXCB(skb).fp); + unix_peek_fds(&scm, skb); sk_peek_offset_fwd(sk, chunk); -- 2.25.1

1 44

[PATCH kernel-4.19 1/9] iommu/arm-smmu-v3: Add support to configure mpam in STE/CD context
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Xingang Wang <wangxingang5(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I49RB2 CVE: NA ------------------------------------------------- To support limiting qos of device, the partid and pmg need to be set into the SMMU STE/CD context. This introduce support of SMMU mpam feature and add interface to set mpam configuration in STE/CD. Signed-off-by: Xingang Wang <wangxingang5(a)huawei.com> Signed-off-by: Rui Zhu <zhurui3(a)huawei.com> Reviewed-by: Yingtai Xie <xieyingtai(a)huawei.com> Reviewed-by: Zhen Lei <thunder.leizhen(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/iommu/arm-smmu-v3-context.c | 25 +++++ drivers/iommu/arm-smmu-v3.c | 151 ++++++++++++++++++++++++++++ include/linux/ascend_smmu.h | 9 ++ 3 files changed, 185 insertions(+) create mode 100644 include/linux/ascend_smmu.h diff --git a/drivers/iommu/arm-smmu-v3-context.c b/drivers/iommu/arm-smmu-v3-context.c index 2351de86d31f1..b30a93e076608 100644 --- a/drivers/iommu/arm-smmu-v3-context.c +++ b/drivers/iommu/arm-smmu-v3-context.c @@ -66,6 +66,9 @@ #define CTXDESC_CD_1_TTB0_MASK GENMASK_ULL(51, 4) +#define CTXDESC_CD_5_PARTID_MASK GENMASK_ULL(47, 32) +#define CTXDESC_CD_5_PMG_MASK GENMASK_ULL(55, 48) + /* Convert between AArch64 (CPU) TCR format and SMMU CD format */ #define ARM_SMMU_TCR2CD(tcr, fld) FIELD_PREP(CTXDESC_CD_0_TCR_##fld, \ FIELD_GET(ARM64_TCR_##fld, tcr)) @@ -563,6 +566,28 @@ static int arm_smmu_set_cd(struct iommu_pasid_table_ops *ops, int pasid, return arm_smmu_write_ctx_desc(tbl, pasid, cd); } +int arm_smmu_set_cd_mpam(struct iommu_pasid_table_ops *ops, + int ssid, int partid, int pmg) +{ + struct arm_smmu_cd_tables *tbl = pasid_ops_to_tables(ops); + u64 val; + __le64 *cdptr = arm_smmu_get_cd_ptr(tbl, ssid); + + if (!cdptr) + return -ENOMEM; + + val = le64_to_cpu(cdptr[5]); + val &= ~CTXDESC_CD_5_PARTID_MASK; + val |= FIELD_PREP(CTXDESC_CD_5_PARTID_MASK, partid); + val &= ~CTXDESC_CD_5_PMG_MASK; + val |= FIELD_PREP(CTXDESC_CD_5_PMG_MASK, pmg); + WRITE_ONCE(cdptr[5], cpu_to_le64(val)); + + iommu_pasid_flush(&tbl->pasid, ssid, true); + + return 0; +} + static void arm_smmu_clear_cd(struct iommu_pasid_table_ops *ops, int pasid, struct iommu_pasid_entry *entry) { diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c index 8b5083c3e0a16..6634d596f7efc 100644 --- a/drivers/iommu/arm-smmu-v3.c +++ b/drivers/iommu/arm-smmu-v3.c @@ -88,6 +88,10 @@ #define IDR1_SSIDSIZE GENMASK(10, 6) #define IDR1_SIDSIZE GENMASK(5, 0) +#define ARM_SMMU_IDR3 0xc +#define IDR3_MPAM (1 << 7) +#define ARM_SMMU_IDR3_CFG 0x140C + #define ARM_SMMU_IDR5 0x14 #define IDR5_STALL_MAX GENMASK(31, 16) #define IDR5_GRAN64K (1 << 6) @@ -186,6 +190,10 @@ #define ARM_SMMU_PRIQ_IRQ_CFG1 0xd8 #define ARM_SMMU_PRIQ_IRQ_CFG2 0xdc +#define ARM_SMMU_MPAMIDR 0x130 +#define MPAMIDR_PMG_MAX GENMASK(23, 16) +#define MPAMIDR_PARTID_MAX GENMASK(15, 0) + /* Common MSI config fields */ #define MSI_CFG0_ADDR_MASK GENMASK_ULL(51, 2) #define MSI_CFG2_SH GENMASK(5, 4) @@ -250,6 +258,7 @@ #define STRTAB_STE_1_S1COR GENMASK_ULL(5, 4) #define STRTAB_STE_1_S1CSH GENMASK_ULL(7, 6) +#define STRTAB_STE_1_S1MPAM (1UL << 26) #define STRTAB_STE_1_S1STALLD (1UL << 27) #define STRTAB_STE_1_EATS GENMASK_ULL(29, 28) @@ -273,6 +282,11 @@ #define STRTAB_STE_3_S2TTB_MASK GENMASK_ULL(51, 4) +#define STRTAB_STE_4_PARTID_MASK GENMASK_ULL(31, 16) + +#define STRTAB_STE_5_MPAM_NS (1UL << 8) +#define STRTAB_STE_5_PMG_MASK GENMASK_ULL(7, 0) + /* Command queue */ #define CMDQ_ENT_SZ_SHIFT 4 #define CMDQ_ENT_DWORDS ((1 << CMDQ_ENT_SZ_SHIFT) >> 3) @@ -634,6 +648,7 @@ struct arm_smmu_device { #define ARM_SMMU_FEAT_SVA (1 << 17) #define ARM_SMMU_FEAT_HA (1 << 18) #define ARM_SMMU_FEAT_HD (1 << 19) +#define ARM_SMMU_FEAT_MPAM (1 << 20) u32 features; #define ARM_SMMU_OPT_SKIP_PREFETCH (1 << 0) @@ -672,6 +687,9 @@ struct arm_smmu_device { struct mutex streams_mutex; struct iopf_queue *iopf_queue; + + unsigned int mpam_partid_max; + unsigned int mpam_pmg_max; }; struct arm_smmu_stream { @@ -3980,6 +3998,25 @@ static int arm_smmu_device_hw_probe(struct arm_smmu_device *smmu) if (smmu->sid_bits <= STRTAB_SPLIT) smmu->features &= ~ARM_SMMU_FEAT_2_LVL_STRTAB; + /* IDR3 */ + reg = readl_relaxed(smmu->base + ARM_SMMU_IDR3); + + if (!(reg & IDR3_MPAM)) { + reg |= FIELD_PREP(IDR3_MPAM, 1); + writel(reg, smmu->base + ARM_SMMU_IDR3_CFG); + reg = readl_relaxed(smmu->base + ARM_SMMU_IDR3); + if (!(reg & IDR3_MPAM)) + dev_warn(smmu->dev, "enable smmu mpam failed\n"); + } + + if (reg & IDR3_MPAM) { + reg = readl_relaxed(smmu->base + ARM_SMMU_MPAMIDR); + smmu->mpam_partid_max = FIELD_GET(MPAMIDR_PARTID_MAX, reg); + smmu->mpam_pmg_max = FIELD_GET(MPAMIDR_PMG_MAX, reg); + if (smmu->mpam_partid_max || smmu->mpam_pmg_max) + smmu->features |= ARM_SMMU_FEAT_MPAM; + } + /* IDR5 */ reg = readl_relaxed(smmu->base + ARM_SMMU_IDR5); @@ -4127,6 +4164,120 @@ static unsigned long arm_smmu_resource_size(struct arm_smmu_device *smmu) return SZ_128K; } +static int arm_smmu_set_ste_mpam(struct arm_smmu_device *smmu, + int sid, int partid, int pmg, int s1mpam) +{ + u64 val; + __le64 *ste; + + if (!arm_smmu_sid_in_range(smmu, sid)) + return -ERANGE; + + /* get ste ptr */ + ste = arm_smmu_get_step_for_sid(smmu, sid); + + /* write s1mpam to ste */ + val = le64_to_cpu(ste[1]); + val &= ~STRTAB_STE_1_S1MPAM; + val |= FIELD_PREP(STRTAB_STE_1_S1MPAM, s1mpam); + WRITE_ONCE(ste[1], cpu_to_le64(val)); + + val = le64_to_cpu(ste[4]); + val &= ~STRTAB_STE_4_PARTID_MASK; + val |= FIELD_PREP(STRTAB_STE_4_PARTID_MASK, partid); + WRITE_ONCE(ste[4], cpu_to_le64(val)); + + val = le64_to_cpu(ste[5]); + val &= ~STRTAB_STE_5_PMG_MASK; + val |= FIELD_PREP(STRTAB_STE_5_PMG_MASK, pmg); + WRITE_ONCE(ste[5], cpu_to_le64(val)); + + arm_smmu_sync_ste_for_sid(smmu, sid); + + return 0; +} + +int arm_smmu_set_cd_mpam(struct iommu_pasid_table_ops *ops, + int ssid, int partid, int pmg); + +static int arm_smmu_set_mpam(struct arm_smmu_device *smmu, + int sid, int ssid, int partid, int pmg, int s1mpam) +{ + struct arm_smmu_master_data *master = arm_smmu_find_master(smmu, sid); + struct arm_smmu_s1_cfg *cfg = master ? master->ste.s1_cfg : NULL; + struct arm_smmu_domain *domain = master ? master->domain : NULL; + int ret; + + struct arm_smmu_cmdq_ent prefetch_cmd = { + .opcode = CMDQ_OP_PREFETCH_CFG, + .prefetch = { + .sid = sid, + }, + }; + + if (!(smmu->features & ARM_SMMU_FEAT_MPAM)) + return -ENODEV; + + if (WARN_ON(!domain)) + return -EINVAL; + + if (WARN_ON(!cfg)) + return -EINVAL; + + if (WARN_ON(ssid >= (1 << master->ssid_bits))) + return -E2BIG; + + if (partid > smmu->mpam_partid_max || pmg > smmu->mpam_pmg_max) { + dev_err(smmu->dev, + "mpam rmid out of range: partid[0, %d] pmg[0, %d]\n", + smmu->mpam_partid_max, smmu->mpam_pmg_max); + return -ERANGE; + } + + ret = arm_smmu_set_ste_mpam(smmu, sid, partid, pmg, s1mpam); + if (ret < 0) { + dev_err(smmu->dev, "set ste mpam configuration error %d\n", + ret); + return ret; + } + + /* do not modify cd table which owned by guest */ + if (domain->stage == ARM_SMMU_DOMAIN_NESTED) { + dev_err(smmu->dev, + "mpam: smmu cd is owned by guest, not modified\n"); + return 0; + } + + ret = arm_smmu_set_cd_mpam(cfg->ops, ssid, partid, pmg); + if (s1mpam && ret < 0) { + dev_err(smmu->dev, "set cd mpam configuration error %d\n", + ret); + return ret; + } + + /* It's likely that we'll want to use the new STE soon */ + if (!(smmu->options & ARM_SMMU_OPT_SKIP_PREFETCH)) + arm_smmu_cmdq_issue_cmd(smmu, &prefetch_cmd); + + dev_info(smmu->dev, "partid %d, pmg %d\n", partid, pmg); + + return 0; +} + +/** + * arm_smmu_set_dev_mpam() - Set mpam configuration to SMMU STE/CD + */ +int arm_smmu_set_dev_mpam(struct device *dev, int ssid, int partid, int pmg, + int s1mpam) +{ + struct arm_smmu_master_data *master = dev->iommu_fwspec->iommu_priv; + struct arm_smmu_device *smmu = master->domain->smmu; + int sid = master->streams->id; + + return arm_smmu_set_mpam(smmu, sid, ssid, partid, pmg, s1mpam); +} +EXPORT_SYMBOL(arm_smmu_set_dev_mpam); + static int arm_smmu_device_probe(struct platform_device *pdev) { int irq, ret; diff --git a/include/linux/ascend_smmu.h b/include/linux/ascend_smmu.h new file mode 100644 index 0000000000000..bd0edd25057e1 --- /dev/null +++ b/include/linux/ascend_smmu.h @@ -0,0 +1,9 @@ +#ifndef __LINUX_ASCEND_SMMU_H +#define __LINUX_ASCEND_SMMU_H + +#include <linux/device.h> + +extern int arm_smmu_set_dev_mpam(struct device *dev, int ssid, int partid, + int pmg, int s1mpam); + +#endif /* __LINUX_ASCEND_SMMU_H */ -- 2.25.1

1 8

[PATCH kernel-4.19] efi: Change down_interruptible() in virt_efi_reset_system() to down_trylock()
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Zhang Jianhua <chris.zjh(a)huawei.com> hulk inclusion category: bugfix bugzilla: 182892, https://gitee.com/openeuler/kernel/issues/I4F04R -------- While reboot the system by sysrq, the following bug will be occur. BUG: sleeping function called from invalid context at kernel/locking/semaphore.c:90 in_atomic(): 0, irqs_disabled(): 128, non_block: 0, pid: 10052, name: rc.shutdown CPU: 3 PID: 10052 Comm: rc.shutdown Tainted: G W O 5.10.0 #1 Call trace: dump_backtrace+0x0/0x1c8 show_stack+0x18/0x28 dump_stack+0xd0/0x110 ___might_sleep+0x14c/0x160 __might_sleep+0x74/0x88 down_interruptible+0x40/0x118 virt_efi_reset_system+0x3c/0xd0 efi_reboot+0xd4/0x11c machine_restart+0x60/0x9c emergency_restart+0x1c/0x2c sysrq_handle_reboot+0x1c/0x2c __handle_sysrq+0xd0/0x194 write_sysrq_trigger+0xbc/0xe4 proc_reg_write+0xd4/0xf0 vfs_write+0xa8/0x148 ksys_write+0x6c/0xd8 __arm64_sys_write+0x18/0x28 el0_svc_common.constprop.3+0xe4/0x16c do_el0_svc+0x1c/0x2c el0_svc+0x20/0x30 el0_sync_handler+0x80/0x17c el0_sync+0x158/0x180 The reason for this problem is that irq has been disabled in machine_restart() and then it calls down_interruptible() in virt_efi_reset_system(), which would occur sleep in irq context, it is dangerous! Commit 99409b935c9a("locking/semaphore: Add might_sleep() to down_*() family") add might_sleep() in down_interruptible(), so the bug info is here. down_trylock() can solve this problem, cause there is no might_sleep. Signed-off-by: Zhang Jianhua <chris.zjh(a)huawei.com> Reviewed-by: Chen Lifu <chenlifu(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Reviewed-by: Liao Chang <liaochang1(a)huawei.com> Reviewed-by: He Ying <heying24(a)huawei.com> Signed-off-by: Chen Jun <chenjun102(a)huawei.com> Reviewed-by: Yihang Xu <xuyihang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/firmware/efi/runtime-wrappers.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/firmware/efi/runtime-wrappers.c b/drivers/firmware/efi/runtime-wrappers.c index 04288719721fc..ecee810f7a6fa 100644 --- a/drivers/firmware/efi/runtime-wrappers.c +++ b/drivers/firmware/efi/runtime-wrappers.c @@ -408,7 +408,7 @@ static void virt_efi_reset_system(int reset_type, unsigned long data_size, efi_char16_t *data) { - if (down_interruptible(&efi_runtime_lock)) { + if (down_trylock(&efi_runtime_lock)) { pr_warn("failed to invoke the reset_system() runtime service:\n" "could not get exclusive access to the firmware\n"); return; -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] svm: Use vma->vm_pgoff for the nid
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Lijun Fang <fanglijun3(a)huawei.com> ascend inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I4DZYE CVE: NA --------------------------- svm_mmap use the pgoff flag for the security requirement nid, which used as cdm node. Signed-off-by: Lijun Fang <fanglijun3(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/char/svm.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/char/svm.c b/drivers/char/svm.c index f312b7e02263c..62092fbaa0104 100644 --- a/drivers/char/svm.c +++ b/drivers/char/svm.c @@ -1820,9 +1820,17 @@ static int svm_mmap(struct file *file, struct vm_area_struct *vma) if ((vma->vm_end < vma->vm_start) || (vm_size > MMAP_PHY32_MAX)) return -EINVAL; - page = alloc_pages(GFP_KERNEL | GFP_DMA32, get_order(vm_size)); + /* vma->vm_pgoff transfer the nid */ + if (vma->vm_pgoff == 0) + page = alloc_pages(GFP_KERNEL | GFP_DMA32, + get_order(vm_size)); + else + page = alloc_pages_node((int)vma->vm_pgoff, + GFP_KERNEL | __GFP_THISNODE, + get_order(vm_size)); if (!page) { - dev_err(sdev->dev, "fail to alloc page\n"); + dev_err(sdev->dev, "fail to alloc page on node 0x%lx\n", + vma->vm_pgoff); return -ENOMEM; } -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] Ascend/hugetlb:support alloc normal and buddy hugepage
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Zhou Guanghui <zhouguanghui1(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4D63I CVE: NA ---------------------------------------------------- The current function hugetlb_alloc_hugepage implements the allocation from static hugepages first. When the static hugepage is used up, it attempts to apply for hugepages from buddy system. Two additional modes are supported: static hugepages only and buddy hugepages only. Signed-off-by: Zhou Guanghui <zhouguanghui1(a)huawei.com> Signed-off-by: Guo Mengqi <guomengqi3(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/hugetlb.h | 11 +++++++++-- mm/hugetlb.c | 35 +++++++++++++++++++++++++++++++++-- 2 files changed, 42 insertions(+), 4 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 830c41a5ca70e..de6cdfa51694c 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -374,8 +374,15 @@ int huge_add_to_page_cache(struct page *page, struct address_space *mapping, pgoff_t idx); #ifdef CONFIG_ASCEND_FEATURES +#define HUGETLB_ALLOC_NONE 0x00 +#define HUGETLB_ALLOC_NORMAL 0x01 /* normal hugepage */ +#define HUGETLB_ALLOC_BUDDY 0x02 /* buddy hugepage */ +#define HUGETLB_ALLOC_MASK (HUGETLB_ALLOC_NONE | \ + HUGETLB_ALLOC_NORMAL | \ + HUGETLB_ALLOC_BUDDY) + const struct hstate *hugetlb_get_hstate(void); -struct page *hugetlb_alloc_hugepage(int nid); +struct page *hugetlb_alloc_hugepage(int nid, int flag); int hugetlb_insert_hugepage_pte(struct mm_struct *mm, unsigned long addr, pgprot_t prot, struct page *hpage); #else @@ -384,7 +391,7 @@ static inline const struct hstate *hugetlb_get_hstate(void) return NULL; } -static inline struct page *hugetlb_alloc_hugepage(int nid) +static inline struct page *hugetlb_alloc_hugepage(int nid, int flag) { return NULL; } diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 3091c61bb63e3..907b1351b0f5f 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5233,17 +5233,48 @@ const struct hstate *hugetlb_get_hstate(void) } EXPORT_SYMBOL_GPL(hugetlb_get_hstate); +static struct page *hugetlb_alloc_hugepage_normal(struct hstate *h, + gfp_t gfp_mask, int nid) +{ + struct page *page = NULL; + + spin_lock(&hugetlb_lock); + if (h->free_huge_pages - h->resv_huge_pages > 0) + page = dequeue_huge_page_nodemask(h, gfp_mask, nid, NULL, NULL); + spin_unlock(&hugetlb_lock); + + return page; +} + /* * Allocate hugepage without reserve */ -struct page *hugetlb_alloc_hugepage(int nid) +struct page *hugetlb_alloc_hugepage(int nid, int flag) { + struct hstate *h = &default_hstate; + gfp_t gfp_mask = htlb_alloc_mask(h); + struct page *page = NULL; + VM_WARN_ON(nid < 0 || nid >= MAX_NUMNODES); + if (flag & ~HUGETLB_ALLOC_MASK) + return NULL; + if (nid == NUMA_NO_NODE) nid = numa_mem_id(); - return alloc_huge_page_node(&default_hstate, nid); + gfp_mask |= __GFP_THISNODE; + + if (flag & HUGETLB_ALLOC_NORMAL) + page = hugetlb_alloc_hugepage_normal(h, gfp_mask, nid); + else if (flag & HUGETLB_ALLOC_BUDDY) { + if (enable_charge_mighp) + gfp_mask |= __GFP_ACCOUNT; + page = alloc_migrate_huge_page(h, gfp_mask, nid, NULL); + } else + page = alloc_huge_page_node(h, nid); + + return page; } EXPORT_SYMBOL_GPL(hugetlb_alloc_hugepage); -- 2.25.1

1 0

[PATCH openEuler-1.0-LTS] Ascend/memcg: Use CONFIG_ASCEND_FEATURES for customized interfaces
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Zhou Guanghui <zhouguanghui1(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I4D63I CVE: NA ------------------------------------------------------------- The following functions are used only in the ascend scenario: hugetlb_get_hstate, hugetlb_alloc_hugepage, hugetlb_insert_hugepage_pte, hugetlb_insert_hugepage_pte_by_pa Remove unused interface hugetlb_insert_hugepage Signed-off-by: Zhou Guanghui <zhouguanghui1(a)huawei.com> Signed-off-by: Guo Mengqi <guomengqi3(a)huawei.com> Reviewed-by: Weilong Chen <chenweilong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/linux/hugetlb.h | 33 ++++++++++++++++++++++++++------- mm/hugetlb.c | 2 +- 2 files changed, 27 insertions(+), 8 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 6831b24defd64..830c41a5ca70e 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -373,11 +373,27 @@ struct page *alloc_huge_page_vma(struct hstate *h, struct vm_area_struct *vma, int huge_add_to_page_cache(struct page *page, struct address_space *mapping, pgoff_t idx); -#ifdef CONFIG_ARM64 +#ifdef CONFIG_ASCEND_FEATURES const struct hstate *hugetlb_get_hstate(void); struct page *hugetlb_alloc_hugepage(int nid); int hugetlb_insert_hugepage_pte(struct mm_struct *mm, unsigned long addr, pgprot_t prot, struct page *hpage); +#else +static inline const struct hstate *hugetlb_get_hstate(void) +{ + return NULL; +} + +static inline struct page *hugetlb_alloc_hugepage(int nid) +{ + return NULL; +} + +static inline int hugetlb_insert_hugepage_pte(struct mm_struct *mm, + unsigned long addr, pgprot_t prot, struct page *hpage) +{ + return -EPERM; +} #endif /* arch callback */ @@ -614,12 +630,6 @@ static inline void set_huge_swap_pte_at(struct mm_struct *mm, unsigned long addr { } -static inline int hugetlb_insert_hugepage_pte_by_pa(struct mm_struct *mm, - unsigned long vir_addr, - pgprot_t prot, unsigned long phy_addr) -{ - return 0; -} #endif /* CONFIG_HUGETLB_PAGE */ static inline spinlock_t *huge_pte_lock(struct hstate *h, @@ -632,4 +642,13 @@ static inline spinlock_t *huge_pte_lock(struct hstate *h, return ptl; } +#ifndef CONFIG_ASCEND_FEATURES +static inline int hugetlb_insert__hugepage_pte_by_pa(struct mm_struct *mm, + unsigned long vir_addr, + pgprot_t prot, unsigned long phy_addr) +{ + return -EPERM; +} +#endif + #endif /* _LINUX_HUGETLB_H */ diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 5132fefb60c86..3091c61bb63e3 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5226,7 +5226,7 @@ void move_hugetlb_state(struct page *oldpage, struct page *newpage, int reason) } } -#ifdef CONFIG_ARM64 +#ifdef CONFIG_ASCEND_FEATURES const struct hstate *hugetlb_get_hstate(void) { return &default_hstate; -- 2.25.1

1 0

[PATCH kernel-4.19 1/2] fuse: truncate pagecache on atomic_o_trunc
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Miklos Szeredi <mszeredi(a)redhat.com> mainline inclusion from mainline-5.14 commit 76224355db7570cbe6b6f75c8929a1558828dd55 category: bugfix bugzilla: 181107 CVE: NA --------------------------- fuse_finish_open() will be called with FUSE_NOWRITE in case of atomic O_TRUNC. This can deadlock with fuse_wait_on_page_writeback() in fuse_launder_page() triggered by invalidate_inode_pages2(). Fix by replacing invalidate_inode_pages2() in fuse_finish_open() with a truncate_pagecache() call. This makes sense regardless of FOPEN_KEEP_CACHE or fc->writeback cache, so do it unconditionally. Reported-by: Xie Yongji <xieyongji(a)bytedance.com> Reported-and-tested-by: syzbot+bea44a5189836d956894(a)syzkaller.appspotmail.com Fixes: e4648309b85a ("fuse: truncate pending writes on O_TRUNC") Cc: <stable(a)vger.kernel.org> Signed-off-by: Miklos Szeredi <mszeredi(a)redhat.com> Conflict: fs/fuse/file.c Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/fuse/file.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 41e2a7b567d7f..5ecabacabd477 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -178,8 +178,6 @@ void fuse_finish_open(struct inode *inode, struct file *file) if (ff->open_flags & FOPEN_DIRECT_IO) file->f_op = &fuse_direct_io_file_operations; - if (!(ff->open_flags & FOPEN_KEEP_CACHE)) - invalidate_inode_pages2(inode->i_mapping); if (ff->open_flags & FOPEN_STREAM) stream_open(inode, file); else if (ff->open_flags & FOPEN_NONSEEKABLE) @@ -191,10 +189,13 @@ void fuse_finish_open(struct inode *inode, struct file *file) fi->attr_version = ++fc->attr_version; i_size_write(inode, 0); spin_unlock(&fc->lock); + truncate_pagecache(inode, 0); fuse_invalidate_attr(inode); if (fc->writeback_cache) file_update_time(file); - } + } else if (!(ff->open_flags & FOPEN_KEEP_CACHE)) + invalidate_inode_pages2(inode->i_mapping); + if ((file->f_mode & FMODE_WRITE) && fc->writeback_cache) fuse_link_write_file(file); } -- 2.25.1

1 1

unsubscribe
by 🐟๛ก(ｰ̀ωｰ́ก) 22 Oct '21

22 Oct '21

unsubscribe

1 0

(no subject)
by 🐟๛ก(ｰ̀ωｰ́ก) 22 Oct '21

22 Oct '21

help

1 0

[PATCH kernel-4.19 1/4] ext4: check and update i_disksize properly
by Yang Yingliang 22 Oct '21

22 Oct '21

From: Zhang Yi <yi.zhang(a)huawei.com> mainline inclusion from mainline-5.15-rc4 commit 4df031ff5876d94b48dd9ee486ba5522382a06b2 category: perf bugzilla: NA CVE: NA --------------------------- After commit 3da40c7b0898 ("ext4: only call ext4_truncate when size <= isize"), i_disksize could always be updated to i_size in ext4_setattr(), and we could sure that i_disksize <= i_size since holding inode lock and if i_disksize < i_size there are delalloc writes pending in the range upto i_size. If the end of the current write is <= i_size, there's no need to touch i_disksize since writeback will push i_disksize upto i_size eventually. So we can switch to check i_size instead of i_disksize in ext4_da_write_end() when write to the end of the file. we also could remove ext4_mark_inode_dirty() together because we defer inode dirtying to generic_write_end() or ext4_da_write_inline_data_end(). Signed-off-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: Jan Kara <jack(a)suse.cz> Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Link: https://lore.kernel.org/r/20210716122024.1105856-2-yi.zhang@huawei.com Conflicts: fs/ext4/inode.c Reviewed-by: Yang Erkun <yangerkun(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/ext4/inode.c | 34 ++++++++++++++++++---------------- 1 file changed, 18 insertions(+), 16 deletions(-) diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b1bede3c489e7..9bb38c936e020 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3190,35 +3190,37 @@ static int ext4_da_write_end(struct file *file, end = start + copied - 1; /* - * generic_write_end() will run mark_inode_dirty() if i_size - * changes. So let's piggyback the i_disksize mark_inode_dirty - * into that. + * Since we are holding inode lock, we are sure i_disksize <= + * i_size. We also know that if i_disksize < i_size, there are + * delalloc writes pending in the range upto i_size. If the end of + * the current write is <= i_size, there's no need to touch + * i_disksize since writeback will push i_disksize upto i_size + * eventually. If the end of the current write is > i_size and + * inside an allocated block (ext4_da_should_update_i_disksize() + * check), we need to update i_disksize here as neither + * ext4_writepage() nor certain ext4_writepages() paths not + * allocating blocks update i_disksize. + * + * Note that we defer inode dirtying to generic_write_end() / + * ext4_da_write_inline_data_end(). */ new_i_size = pos + copied; - if (copied && new_i_size > EXT4_I(inode)->i_disksize) { + if (copied && new_i_size > inode->i_size) { if (ext4_has_inline_data(inode) || - ext4_da_should_update_i_disksize(page, end)) { + ext4_da_should_update_i_disksize(page, end)) ext4_update_i_disksize(inode, new_i_size); - /* We need to mark inode dirty even if - * new_i_size is less that inode->i_size - * bu greater than i_disksize.(hint delalloc) - */ - ext4_mark_inode_dirty(handle, inode); - } } if (write_mode != CONVERT_INLINE_DATA && ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA) && ext4_has_inline_data(inode)) - ret2 = ext4_da_write_inline_data_end(inode, pos, len, copied, + ret = ext4_da_write_inline_data_end(inode, pos, len, copied, page); else - ret2 = generic_write_end(file, mapping, pos, len, copied, + ret = generic_write_end(file, mapping, pos, len, copied, page, fsdata); - copied = ret2; - if (ret2 < 0) - ret = ret2; + copied = ret; ret2 = ext4_journal_stop(handle); if (!ret) ret = ret2; -- 2.25.1

1 3