diff options
-rw-r--r-- | arch/s390/Kconfig | 1 | ||||
-rw-r--r-- | arch/s390/defconfig | 37 | ||||
-rw-r--r-- | arch/s390/include/asm/facility.h | 3 | ||||
-rw-r--r-- | arch/s390/include/asm/pgalloc.h | 3 | ||||
-rw-r--r-- | arch/s390/include/asm/swab.h | 2 | ||||
-rw-r--r-- | arch/s390/include/asm/tlb.h | 22 | ||||
-rw-r--r-- | arch/s390/kernel/head.S | 2 | ||||
-rw-r--r-- | arch/s390/kernel/irq.c | 9 | ||||
-rw-r--r-- | arch/s390/kernel/perf_cpum_cf.c | 4 | ||||
-rw-r--r-- | arch/s390/mm/maccess.c | 27 | ||||
-rw-r--r-- | arch/s390/mm/pgtable.c | 63 | ||||
-rw-r--r-- | drivers/s390/block/dasd_eckd.c | 24 | ||||
-rw-r--r-- | drivers/s390/char/vmur.c | 2 | ||||
-rw-r--r-- | fs/fuse/dir.c | 25 | ||||
-rw-r--r-- | fs/fuse/file.c | 129 | ||||
-rw-r--r-- | fs/fuse/inode.c | 1 | ||||
-rw-r--r-- | include/linux/fuse.h | 2 |
17 files changed, 254 insertions, 102 deletions
diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index 2b7c0fbe578e..9015060919a0 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -90,7 +90,6 @@ config S390 select HAVE_KERNEL_XZ select HAVE_ARCH_MUTEX_CPU_RELAX select HAVE_ARCH_JUMP_LABEL if !MARCH_G5 - select HAVE_RCU_TABLE_FREE if SMP select ARCH_SAVE_PAGE_KEYS if HIBERNATION select HAVE_MEMBLOCK select HAVE_MEMBLOCK_NODE_MAP diff --git a/arch/s390/defconfig b/arch/s390/defconfig index 6cf8e26b3137..1957a9dd256d 100644 --- a/arch/s390/defconfig +++ b/arch/s390/defconfig @@ -1,8 +1,12 @@ CONFIG_EXPERIMENTAL=y CONFIG_SYSVIPC=y CONFIG_POSIX_MQUEUE=y +CONFIG_FHANDLE=y +CONFIG_TASKSTATS=y +CONFIG_TASK_DELAY_ACCT=y +CONFIG_TASK_XACCT=y +CONFIG_TASK_IO_ACCOUNTING=y CONFIG_AUDIT=y -CONFIG_RCU_TRACE=y CONFIG_IKCONFIG=y CONFIG_IKCONFIG_PROC=y CONFIG_CGROUPS=y @@ -14,16 +18,22 @@ CONFIG_CGROUP_MEM_RES_CTLR_SWAP=y CONFIG_CGROUP_SCHED=y CONFIG_RT_GROUP_SCHED=y CONFIG_BLK_CGROUP=y +CONFIG_NAMESPACES=y CONFIG_BLK_DEV_INITRD=y -# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set +CONFIG_RD_BZIP2=y +CONFIG_RD_LZMA=y +CONFIG_RD_XZ=y +CONFIG_RD_LZO=y +CONFIG_EXPERT=y # CONFIG_COMPAT_BRK is not set -CONFIG_SLAB=y CONFIG_PROFILING=y CONFIG_OPROFILE=y CONFIG_KPROBES=y CONFIG_MODULES=y CONFIG_MODULE_UNLOAD=y CONFIG_MODVERSIONS=y +CONFIG_PARTITION_ADVANCED=y +CONFIG_IBM_PARTITION=y CONFIG_DEFAULT_DEADLINE=y CONFIG_NO_HZ=y CONFIG_HIGH_RES_TIMERS=y @@ -34,18 +44,15 @@ CONFIG_KSM=y CONFIG_BINFMT_MISC=m CONFIG_CMM=m CONFIG_HZ_100=y -CONFIG_KEXEC=y -CONFIG_PM=y +CONFIG_CRASH_DUMP=y CONFIG_HIBERNATION=y CONFIG_PACKET=y CONFIG_UNIX=y CONFIG_NET_KEY=y -CONFIG_AFIUCV=m CONFIG_INET=y CONFIG_IP_MULTICAST=y # CONFIG_INET_LRO is not set CONFIG_IPV6=y -CONFIG_NET_SCTPPROBE=m CONFIG_L2TP=m CONFIG_L2TP_DEBUGFS=m CONFIG_VLAN_8021Q=y @@ -84,15 +91,14 @@ CONFIG_SCSI_CONSTANTS=y CONFIG_SCSI_LOGGING=y CONFIG_SCSI_SCAN_ASYNC=y CONFIG_ZFCP=y -CONFIG_ZFCP_DIF=y CONFIG_NETDEVICES=y -CONFIG_DUMMY=m CONFIG_BONDING=m +CONFIG_DUMMY=m CONFIG_EQUALIZER=m CONFIG_TUN=m -CONFIG_NET_ETHERNET=y CONFIG_VIRTIO_NET=y CONFIG_RAW_DRIVER=m +CONFIG_VIRTIO_BALLOON=y CONFIG_EXT2_FS=y CONFIG_EXT3_FS=y # CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set @@ -103,27 +109,21 @@ CONFIG_PROC_KCORE=y CONFIG_TMPFS=y CONFIG_TMPFS_POSIX_ACL=y # CONFIG_NETWORK_FILESYSTEMS is not set -CONFIG_PARTITION_ADVANCED=y -CONFIG_IBM_PARTITION=y -CONFIG_DLM=m CONFIG_MAGIC_SYSRQ=y -CONFIG_DEBUG_KERNEL=y CONFIG_TIMER_STATS=y CONFIG_PROVE_LOCKING=y CONFIG_PROVE_RCU=y CONFIG_LOCK_STAT=y CONFIG_DEBUG_LOCKDEP=y -CONFIG_DEBUG_SPINLOCK_SLEEP=y CONFIG_DEBUG_LIST=y CONFIG_DEBUG_NOTIFIERS=y -# CONFIG_RCU_CPU_STALL_DETECTOR is not set +CONFIG_RCU_TRACE=y CONFIG_KPROBES_SANITY_TEST=y CONFIG_DEBUG_FORCE_WEAK_PER_CPU=y CONFIG_CPU_NOTIFIER_ERROR_INJECT=m CONFIG_LATENCYTOP=y -CONFIG_SYSCTL_SYSCALL_CHECK=y CONFIG_DEBUG_PAGEALLOC=y -# CONFIG_FTRACE is not set +CONFIG_BLK_DEV_IO_TRACE=y # CONFIG_STRICT_DEVMEM is not set CONFIG_CRYPTO_NULL=m CONFIG_CRYPTO_CRYPTD=m @@ -173,4 +173,3 @@ CONFIG_CRYPTO_SHA512_S390=m CONFIG_CRYPTO_DES_S390=m CONFIG_CRYPTO_AES_S390=m CONFIG_CRC7=m -CONFIG_VIRTIO_BALLOON=y diff --git a/arch/s390/include/asm/facility.h b/arch/s390/include/asm/facility.h index 1e5b27edc0c9..2ee66a65f2d4 100644 --- a/arch/s390/include/asm/facility.h +++ b/arch/s390/include/asm/facility.h @@ -38,12 +38,11 @@ static inline void stfle(u64 *stfle_fac_list, int size) unsigned long nr; preempt_disable(); - S390_lowcore.stfl_fac_list = 0; asm volatile( " .insn s,0xb2b10000,0(0)\n" /* stfl */ "0:\n" EX_TABLE(0b, 0b) - : "=m" (S390_lowcore.stfl_fac_list)); + : "+m" (S390_lowcore.stfl_fac_list)); nr = 4; /* bytes stored by stfl */ memcpy(stfle_fac_list, &S390_lowcore.stfl_fac_list, 4); if (S390_lowcore.stfl_fac_list & 0x01000000) { diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h index 8eef9b5b3cf4..78e3041919de 100644 --- a/arch/s390/include/asm/pgalloc.h +++ b/arch/s390/include/asm/pgalloc.h @@ -22,10 +22,7 @@ void crst_table_free(struct mm_struct *, unsigned long *); unsigned long *page_table_alloc(struct mm_struct *, unsigned long); void page_table_free(struct mm_struct *, unsigned long *); -#ifdef CONFIG_HAVE_RCU_TABLE_FREE void page_table_free_rcu(struct mmu_gather *, unsigned long *); -void __tlb_remove_table(void *_table); -#endif static inline void clear_table(unsigned long *s, unsigned long val, size_t n) { diff --git a/arch/s390/include/asm/swab.h b/arch/s390/include/asm/swab.h index 6bdee21c077e..a3e4ebb32090 100644 --- a/arch/s390/include/asm/swab.h +++ b/arch/s390/include/asm/swab.h @@ -77,7 +77,7 @@ static inline __u16 __arch_swab16p(const __u16 *x) asm volatile( #ifndef __s390x__ - " icm %0,2,%O+1(%R1)\n" + " icm %0,2,%O1+1(%R1)\n" " ic %0,%1\n" : "=&d" (result) : "Q" (*x) : "cc"); #else /* __s390x__ */ diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h index c687a2c83462..775a5eea8f9e 100644 --- a/arch/s390/include/asm/tlb.h +++ b/arch/s390/include/asm/tlb.h @@ -30,14 +30,10 @@ struct mmu_gather { struct mm_struct *mm; -#ifdef CONFIG_HAVE_RCU_TABLE_FREE struct mmu_table_batch *batch; -#endif unsigned int fullmm; - unsigned int need_flush; }; -#ifdef CONFIG_HAVE_RCU_TABLE_FREE struct mmu_table_batch { struct rcu_head rcu; unsigned int nr; @@ -49,7 +45,6 @@ struct mmu_table_batch { extern void tlb_table_flush(struct mmu_gather *tlb); extern void tlb_remove_table(struct mmu_gather *tlb, void *table); -#endif static inline void tlb_gather_mmu(struct mmu_gather *tlb, struct mm_struct *mm, @@ -57,29 +52,20 @@ static inline void tlb_gather_mmu(struct mmu_gather *tlb, { tlb->mm = mm; tlb->fullmm = full_mm_flush; - tlb->need_flush = 0; -#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb->batch = NULL; -#endif if (tlb->fullmm) __tlb_flush_mm(mm); } static inline void tlb_flush_mmu(struct mmu_gather *tlb) { - if (!tlb->need_flush) - return; - tlb->need_flush = 0; - __tlb_flush_mm(tlb->mm); -#ifdef CONFIG_HAVE_RCU_TABLE_FREE tlb_table_flush(tlb); -#endif } static inline void tlb_finish_mmu(struct mmu_gather *tlb, unsigned long start, unsigned long end) { - tlb_flush_mmu(tlb); + tlb_table_flush(tlb); } /* @@ -105,10 +91,8 @@ static inline void tlb_remove_page(struct mmu_gather *tlb, struct page *page) static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte, unsigned long address) { -#ifdef CONFIG_HAVE_RCU_TABLE_FREE if (!tlb->fullmm) return page_table_free_rcu(tlb, (unsigned long *) pte); -#endif page_table_free(tlb->mm, (unsigned long *) pte); } @@ -125,10 +109,8 @@ static inline void pmd_free_tlb(struct mmu_gather *tlb, pmd_t *pmd, #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 31)) return; -#ifdef CONFIG_HAVE_RCU_TABLE_FREE if (!tlb->fullmm) return tlb_remove_table(tlb, pmd); -#endif crst_table_free(tlb->mm, (unsigned long *) pmd); #endif } @@ -146,10 +128,8 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud, #ifdef __s390x__ if (tlb->mm->context.asce_limit <= (1UL << 42)) return; -#ifdef CONFIG_HAVE_RCU_TABLE_FREE if (!tlb->fullmm) return tlb_remove_table(tlb, pud); -#endif crst_table_free(tlb->mm, (unsigned long *) pud); #endif } diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index c27a0727f930..adccd908ebc7 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -474,9 +474,9 @@ ENTRY(startup_kdump) stck __LC_LAST_UPDATE_CLOCK spt 5f-.LPG0(%r13) mvc __LC_LAST_UPDATE_TIMER(8),5f-.LPG0(%r13) + xc __LC_STFL_FAC_LIST(8),__LC_STFL_FAC_LIST #ifndef CONFIG_MARCH_G5 # check capabilities against MARCH_{G5,Z900,Z990,Z9_109,Z10} - xc __LC_STFL_FAC_LIST(8),__LC_STFL_FAC_LIST .insn s,0xb2b10000,__LC_STFL_FAC_LIST # store facility list tm __LC_STFL_FAC_LIST,0x01 # stfle available ? jz 0f diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index 1c2cdd59ccd0..8a22c27219dd 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -118,9 +118,10 @@ asmlinkage void do_softirq(void) "a" (__do_softirq) : "0", "1", "2", "3", "4", "5", "14", "cc", "memory" ); - } else + } else { /* We are already on the async stack. */ __do_softirq(); + } } local_irq_restore(flags); @@ -192,11 +193,12 @@ int unregister_external_interrupt(u16 code, ext_int_handler_t handler) int index = ext_hash(code); spin_lock_irqsave(&ext_int_hash_lock, flags); - list_for_each_entry_rcu(p, &ext_int_hash[index], entry) + list_for_each_entry_rcu(p, &ext_int_hash[index], entry) { if (p->code == code && p->handler == handler) { list_del_rcu(&p->entry); kfree_rcu(p, rcu); } + } spin_unlock_irqrestore(&ext_int_hash_lock, flags); return 0; } @@ -211,9 +213,10 @@ void __irq_entry do_extint(struct pt_regs *regs, struct ext_code ext_code, old_regs = set_irq_regs(regs); irq_enter(); - if (S390_lowcore.int_clock >= S390_lowcore.clock_comparator) + if (S390_lowcore.int_clock >= S390_lowcore.clock_comparator) { /* Serve timer interrupts first. */ clock_comparator_work(); + } kstat_cpu(smp_processor_id()).irqs[EXTERNAL_INTERRUPT]++; if (ext_code.code != 0x1004) __get_cpu_var(s390_idle).nohz_delay = 1; diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c index 46405086479c..cb019f429e88 100644 --- a/arch/s390/kernel/perf_cpum_cf.c +++ b/arch/s390/kernel/perf_cpum_cf.c @@ -178,7 +178,7 @@ static void cpumf_pmu_enable(struct pmu *pmu) err = lcctl(cpuhw->state); if (err) { pr_err("Enabling the performance measuring unit " - "failed with rc=%lx\n", err); + "failed with rc=%x\n", err); return; } @@ -203,7 +203,7 @@ static void cpumf_pmu_disable(struct pmu *pmu) err = lcctl(inactive); if (err) { pr_err("Disabling the performance measuring unit " - "failed with rc=%lx\n", err); + "failed with rc=%x\n", err); return; } diff --git a/arch/s390/mm/maccess.c b/arch/s390/mm/maccess.c index 7bb15fcca75e..e1335dc2b1b7 100644 --- a/arch/s390/mm/maccess.c +++ b/arch/s390/mm/maccess.c @@ -61,21 +61,14 @@ long probe_kernel_write(void *dst, const void *src, size_t size) return copied < 0 ? -EFAULT : 0; } -/* - * Copy memory in real mode (kernel to kernel) - */ -int memcpy_real(void *dest, void *src, size_t count) +static int __memcpy_real(void *dest, void *src, size_t count) { register unsigned long _dest asm("2") = (unsigned long) dest; register unsigned long _len1 asm("3") = (unsigned long) count; register unsigned long _src asm("4") = (unsigned long) src; register unsigned long _len2 asm("5") = (unsigned long) count; - unsigned long flags; int rc = -EFAULT; - if (!count) - return 0; - flags = __arch_local_irq_stnsm(0xf8UL); asm volatile ( "0: mvcle %1,%2,0x0\n" "1: jo 0b\n" @@ -86,7 +79,23 @@ int memcpy_real(void *dest, void *src, size_t count) "+d" (_len2), "=m" (*((long *) dest)) : "m" (*((long *) src)) : "cc", "memory"); - arch_local_irq_restore(flags); + return rc; +} + +/* + * Copy memory in real mode (kernel to kernel) + */ +int memcpy_real(void *dest, void *src, size_t count) +{ + unsigned long flags; + int rc; + + if (!count) + return 0; + local_irq_save(flags); + __arch_local_irq_stnsm(0xfbUL); + rc = __memcpy_real(dest, src, count); + local_irq_restore(flags); return rc; } diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 373adf69b01c..6e765bf00670 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -678,8 +678,6 @@ void page_table_free(struct mm_struct *mm, unsigned long *table) } } -#ifdef CONFIG_HAVE_RCU_TABLE_FREE - static void __page_table_free_rcu(void *table, unsigned bit) { struct page *page; @@ -733,7 +731,66 @@ void __tlb_remove_table(void *_table) free_pages((unsigned long) table, ALLOC_ORDER); } -#endif +static void tlb_remove_table_smp_sync(void *arg) +{ + /* Simply deliver the interrupt */ +} + +static void tlb_remove_table_one(void *table) +{ + /* + * This isn't an RCU grace period and hence the page-tables cannot be + * assumed to be actually RCU-freed. + * + * It is however sufficient for software page-table walkers that rely + * on IRQ disabling. See the comment near struct mmu_table_batch. + */ + smp_call_function(tlb_remove_table_smp_sync, NULL, 1); + __tlb_remove_table(table); +} + +static void tlb_remove_table_rcu(struct rcu_head *head) +{ + struct mmu_table_batch *batch; + int i; + + batch = container_of(head, struct mmu_table_batch, rcu); + + for (i = 0; i < batch->nr; i++) + __tlb_remove_table(batch->tables[i]); + + free_page((unsigned long)batch); +} + +void tlb_table_flush(struct mmu_gather *tlb) +{ + struct mmu_table_batch **batch = &tlb->batch; + + if (*batch) { + __tlb_flush_mm(tlb->mm); + call_rcu_sched(&(*batch)->rcu, tlb_remove_table_rcu); + *batch = NULL; + } +} + +void tlb_remove_table(struct mmu_gather *tlb, void *table) +{ + struct mmu_table_batch **batch = &tlb->batch; + + if (*batch == NULL) { + *batch = (struct mmu_table_batch *) + __get_free_page(GFP_NOWAIT | __GFP_NOWARN); + if (*batch == NULL) { + __tlb_flush_mm(tlb->mm); + tlb_remove_table_one(table); + return; + } + (*batch)->nr = 0; + } + (*batch)->tables[(*batch)->nr++] = table; + if ((*batch)->nr == MAX_TABLE_BATCH) + tlb_table_flush(tlb); +} /* * switch on pgstes for its userspace process (for kvm) diff --git a/drivers/s390/block/dasd_eckd.c b/drivers/s390/block/dasd_eckd.c index c21871a4e73d..bc2e8a7c265b 100644 --- a/drivers/s390/block/dasd_eckd.c +++ b/drivers/s390/block/dasd_eckd.c @@ -2844,6 +2844,7 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( sector_t recid, trkid; unsigned int offs; unsigned int count, count_to_trk_end; + int ret; basedev = block->base; if (rq_data_dir(req) == READ) { @@ -2884,8 +2885,8 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( itcw = itcw_init(cqr->data, itcw_size, itcw_op, 0, ctidaw, 0); if (IS_ERR(itcw)) { - dasd_sfree_request(cqr, startdev); - return ERR_PTR(-EINVAL); + ret = -EINVAL; + goto out_error; } cqr->cpaddr = itcw_get_tcw(itcw); if (prepare_itcw(itcw, first_trk, last_trk, @@ -2897,8 +2898,8 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( /* Clock not in sync and XRC is enabled. * Try again later. */ - dasd_sfree_request(cqr, startdev); - return ERR_PTR(-EAGAIN); + ret = -EAGAIN; + goto out_error; } len_to_track_end = 0; /* @@ -2937,8 +2938,10 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( tidaw_flags = 0; last_tidaw = itcw_add_tidaw(itcw, tidaw_flags, dst, part_len); - if (IS_ERR(last_tidaw)) - return ERR_PTR(-EINVAL); + if (IS_ERR(last_tidaw)) { + ret = -EINVAL; + goto out_error; + } dst += part_len; } } @@ -2947,8 +2950,10 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( dst = page_address(bv->bv_page) + bv->bv_offset; last_tidaw = itcw_add_tidaw(itcw, 0x00, dst, bv->bv_len); - if (IS_ERR(last_tidaw)) - return ERR_PTR(-EINVAL); + if (IS_ERR(last_tidaw)) { + ret = -EINVAL; + goto out_error; + } } } last_tidaw->flags |= TIDAW_FLAGS_LAST; @@ -2968,6 +2973,9 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_tpm_track( cqr->buildclk = get_clock(); cqr->status = DASD_CQR_FILLED; return cqr; +out_error: + dasd_sfree_request(cqr, startdev); + return ERR_PTR(ret); } static struct dasd_ccw_req *dasd_eckd_build_cp(struct dasd_device *startdev, diff --git a/drivers/s390/char/vmur.c b/drivers/s390/char/vmur.c index 85f4a9a5d12e..73bef0bd394c 100644 --- a/drivers/s390/char/vmur.c +++ b/drivers/s390/char/vmur.c @@ -903,7 +903,7 @@ static int ur_set_online(struct ccw_device *cdev) goto fail_urdev_put; } - cdev_init(urd->char_device, &ur_fops); + urd->char_device->ops = &ur_fops; urd->char_device->dev = MKDEV(major, minor); urd->char_device->owner = ur_fops.owner; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 206632887bb4..df5ac048dc74 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -387,9 +387,6 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, if (fc->no_create) return -ENOSYS; - if (flags & O_DIRECT) - return -EINVAL; - forget = fuse_alloc_forget(); if (!forget) return -ENOMEM; @@ -644,13 +641,12 @@ static int fuse_unlink(struct inode *dir, struct dentry *entry) fuse_put_request(fc, req); if (!err) { struct inode *inode = entry->d_inode; + struct fuse_inode *fi = get_fuse_inode(inode); - /* - * Set nlink to zero so the inode can be cleared, if the inode - * does have more links this will be discovered at the next - * lookup/getattr. - */ - clear_nlink(inode); + spin_lock(&fc->lock); + fi->attr_version = ++fc->attr_version; + drop_nlink(inode); + spin_unlock(&fc->lock); fuse_invalidate_attr(inode); fuse_invalidate_attr(dir); fuse_invalidate_entry_cache(entry); @@ -762,8 +758,17 @@ static int fuse_link(struct dentry *entry, struct inode *newdir, will reflect changes in the backing inode (link count, etc.) */ - if (!err || err == -EINTR) + if (!err) { + struct fuse_inode *fi = get_fuse_inode(inode); + + spin_lock(&fc->lock); + fi->attr_version = ++fc->attr_version; + inc_nlink(inode); + spin_unlock(&fc->lock); + fuse_invalidate_attr(inode); + } else if (err == -EINTR) { fuse_invalidate_attr(inode); + } return err; } diff --git a/fs/fuse/file.c b/fs/fuse/file.c index a841868bf9ce..504e61b7fd75 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -194,10 +194,6 @@ int fuse_open_common(struct inode *inode, struct file *file, bool isdir) struct fuse_conn *fc = get_fuse_conn(inode); int err; - /* VFS checks this, but only _after_ ->open() */ - if (file->f_flags & O_DIRECT) - return -EINVAL; - err = generic_file_open(inode, file); if (err) return err; @@ -932,17 +928,23 @@ static ssize_t fuse_file_aio_write(struct kiocb *iocb, const struct iovec *iov, struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; size_t count = 0; + size_t ocount = 0; ssize_t written = 0; + ssize_t written_buffered = 0; struct inode *inode = mapping->host; ssize_t err; struct iov_iter i; + loff_t endbyte = 0; WARN_ON(iocb->ki_pos != pos); - err = generic_segment_checks(iov, &nr_segs, &count, VERIFY_READ); + ocount = 0; + err = generic_segment_checks(iov, &nr_segs, &ocount, VERIFY_READ); if (err) return err; + count = ocount; + mutex_lock(&inode->i_mutex); vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE); @@ -962,11 +964,41 @@ static ssize_t fuse_file_aio_write(struct kiocb *iocb, const struct iovec *iov, file_update_time(file); - iov_iter_init(&i, iov, nr_segs, count, 0); - written = fuse_perform_write(file, mapping, &i, pos); - if (written >= 0) - iocb->ki_pos = pos + written; + if (file->f_flags & O_DIRECT) { + written = generic_file_direct_write(iocb, iov, &nr_segs, + pos, &iocb->ki_pos, + count, ocount); + if (written < 0 || written == count) + goto out; + + pos += written; + count -= written; + iov_iter_init(&i, iov, nr_segs, count, written); + written_buffered = fuse_perform_write(file, mapping, &i, pos); + if (written_buffered < 0) { + err = written_buffered; + goto out; + } + endbyte = pos + written_buffered - 1; + + err = filemap_write_and_wait_range(file->f_mapping, pos, + endbyte); + if (err) + goto out; + + invalidate_mapping_pages(file->f_mapping, + pos >> PAGE_CACHE_SHIFT, + endbyte >> PAGE_CACHE_SHIFT); + + written += written_buffered; + iocb->ki_pos = pos + written_buffered; + } else { + iov_iter_init(&i, iov, nr_segs, count, 0); + written = fuse_perform_write(file, mapping, &i, pos); + if (written >= 0) + iocb->ki_pos = pos + written; + } out: current->backing_dev_info = NULL; mutex_unlock(&inode->i_mutex); @@ -1101,30 +1133,41 @@ static ssize_t fuse_direct_read(struct file *file, char __user *buf, return res; } -static ssize_t fuse_direct_write(struct file *file, const char __user *buf, - size_t count, loff_t *ppos) +static ssize_t __fuse_direct_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) { struct inode *inode = file->f_path.dentry->d_inode; ssize_t res; - if (is_bad_inode(inode)) - return -EIO; - - /* Don't allow parallel writes to the same file */ - mutex_lock(&inode->i_mutex); res = generic_write_checks(file, ppos, &count, 0); if (!res) { res = fuse_direct_io(file, buf, count, ppos, 1); if (res > 0) fuse_write_update_size(inode, *ppos); } - mutex_unlock(&inode->i_mutex); fuse_invalidate_attr(inode); return res; } +static ssize_t fuse_direct_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct inode *inode = file->f_path.dentry->d_inode; + ssize_t res; + + if (is_bad_inode(inode)) + return -EIO; + + /* Don't allow parallel writes to the same file */ + mutex_lock(&inode->i_mutex); + res = __fuse_direct_write(file, buf, count, ppos); + mutex_unlock(&inode->i_mutex); + + return res; +} + static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req) { __free_page(req->pages[0]); @@ -2077,6 +2120,57 @@ int fuse_notify_poll_wakeup(struct fuse_conn *fc, return 0; } +static ssize_t fuse_loop_dio(struct file *filp, const struct iovec *iov, + unsigned long nr_segs, loff_t *ppos, int rw) +{ + const struct iovec *vector = iov; + ssize_t ret = 0; + + while (nr_segs > 0) { + void __user *base; + size_t len; + ssize_t nr; + + base = vector->iov_base; + len = vector->iov_len; + vector++; + nr_segs--; + + if (rw == WRITE) + nr = __fuse_direct_write(filp, base, len, ppos); + else + nr = fuse_direct_read(filp, base, len, ppos); + + if (nr < 0) { + if (!ret) + ret = nr; + break; + } + ret += nr; + if (nr != len) + break; + } + + return ret; +} + + +static ssize_t +fuse_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, + loff_t offset, unsigned long nr_segs) +{ + ssize_t ret = 0; + struct file *file = NULL; + loff_t pos = 0; + + file = iocb->ki_filp; + pos = offset; + + ret = fuse_loop_dio(file, iov, nr_segs, &pos, rw); + + return ret; +} + static const struct file_operations fuse_file_operations = { .llseek = fuse_file_llseek, .read = do_sync_read, @@ -2120,6 +2214,7 @@ static const struct address_space_operations fuse_file_aops = { .readpages = fuse_readpages, .set_page_dirty = __set_page_dirty_nobuffers, .bmap = fuse_bmap, + .direct_IO = fuse_direct_IO, }; void fuse_init_file_inode(struct inode *inode) diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 4aec5995867e..26783eb2b1fc 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -947,6 +947,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) sb->s_magic = FUSE_SUPER_MAGIC; sb->s_op = &fuse_super_operations; sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_time_gran = 1; sb->s_export_op = &fuse_export_operations; file = fget(d.fd); diff --git a/include/linux/fuse.h b/include/linux/fuse.h index 8ba2c9460b28..8f2ab8fef929 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -593,7 +593,7 @@ struct fuse_dirent { __u64 off; __u32 namelen; __u32 type; - char name[0]; + char name[]; }; #define FUSE_NAME_OFFSET offsetof(struct fuse_dirent, name) |