diff options
author | 2016-03-26 12:59:04 -0700 | |
---|---|---|
committer | 2016-03-26 12:59:04 -0700 | |
commit | 698f415cf5756e320623bdb015a600945743377c (patch) | |
tree | cf39b9233a9aea178156e876843fb742fa5ed070 | |
parent | Merge tag 'ntb-4.6' of git://github.com/jonmason/ntb (diff) | |
parent | orangefs: fix orangefs_superblock locking (diff) | |
download | linux-698f415cf5756e320623bdb015a600945743377c.tar.gz linux-698f415cf5756e320623bdb015a600945743377c.tar.bz2 linux-698f415cf5756e320623bdb015a600945743377c.zip |
Merge tag 'ofs-pull-tag-1' of git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux
Pull orangefs filesystem from Mike Marshall.
This finally merges the long-pending orangefs filesystem, which has been
much cleaned up with input from Al Viro over the last six months. From
the documentation file:
"OrangeFS is an LGPL userspace scale-out parallel storage system. It
is ideal for large storage problems faced by HPC, BigData, Streaming
Video, Genomics, Bioinformatics.
Orangefs, originally called PVFS, was first developed in 1993 by Walt
Ligon and Eric Blumer as a parallel file system for Parallel Virtual
Machine (PVM) as part of a NASA grant to study the I/O patterns of
parallel programs.
Orangefs features include:
- Distributes file data among multiple file servers
- Supports simultaneous access by multiple clients
- Stores file data and metadata on servers using local file system
and access methods
- Userspace implementation is easy to install and maintain
- Direct MPI support
- Stateless"
see Documentation/filesystems/orangefs.txt for more in-depth details.
* tag 'ofs-pull-tag-1' of git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux: (174 commits)
orangefs: fix orangefs_superblock locking
orangefs: fix do_readv_writev() handling of error halfway through
orangefs: have ->kill_sb() evict the VFS side of things first
orangefs: sanitize ->llseek()
orangefs-bufmap.h: trim unused junk
orangefs: saner calling conventions for getting a slot
orangefs_copy_{to,from}_bufmap(): don't pass bufmap pointer
orangefs: get rid of readdir_handle_s
ornagefs: ensure that truncate has an up to date inode size
orangefs: move code which sets i_link to orangefs_inode_getattr
orangefs: remove needless wrapper around GFP_KERNEL
orangefs: remove wrapper around mutex_lock(&inode->i_mutex)
orangefs: refactor inode type or link_target change detection
orangefs: use new getattr for revalidate and remove old getattr
orangefs: use new getattr in inode getattr and permission
orangefs: use new orangefs_inode_getattr to get size in write and llseek
orangefs: use new orangefs_inode_getattr to create new inodes
orangefs: rename orangefs_inode_getattr to orangefs_inode_old_getattr
orangefs: remove inode->i_lock wrapper
orangefs: put register_chrdev immediately before register_filesystem
...
33 files changed, 11243 insertions, 0 deletions
diff --git a/Documentation/ABI/stable/sysfs-fs-orangefs b/Documentation/ABI/stable/sysfs-fs-orangefs new file mode 100644 index 000000000000..affdb114bd33 --- /dev/null +++ b/Documentation/ABI/stable/sysfs-fs-orangefs @@ -0,0 +1,87 @@ +What: /sys/fs/orangefs/perf_counters/* +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Counters and settings for various caches. + Read only. + + +What: /sys/fs/orangefs/perf_counter_reset +Date: June 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + echo a 0 or a 1 into perf_counter_reset to + reset all the counters in + /sys/fs/orangefs/perf_counters + except ones with PINT_PERF_PRESERVE set. + + +What: /sys/fs/orangefs/perf_time_interval_secs +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Length of perf counter intervals in + seconds. + + +What: /sys/fs/orangefs/perf_history_size +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + The perf_counters cache statistics have N, or + perf_history_size, samples. The default is + one. + + Every perf_time_interval_secs the (first) + samples are reset. + + If N is greater than one, the "current" set + of samples is reset, and the samples from the + other N-1 intervals remain available. + + +What: /sys/fs/orangefs/op_timeout_secs +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Service operation timeout in seconds. + + +What: /sys/fs/orangefs/slot_timeout_secs +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + "Slot" timeout in seconds. A "slot" + is an indexed buffer in the shared + memory segment used for communication + between the kernel module and userspace. + Slots are requested and waited for, + the wait times out after slot_timeout_secs. + + +What: /sys/fs/orangefs/acache/* +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Attribute cache configurable settings. + + +What: /sys/fs/orangefs/ncache/* +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Name cache configurable settings. + + +What: /sys/fs/orangefs/capcache/* +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Capability cache configurable settings. + + +What: /sys/fs/orangefs/ccache/* +Date: Jun 2015 +Contact: Mike Marshall <hubcap@omnibond.com> +Description: + Credential cache configurable settings. diff --git a/Documentation/filesystems/orangefs.txt b/Documentation/filesystems/orangefs.txt new file mode 100644 index 000000000000..e1a0056a365f --- /dev/null +++ b/Documentation/filesystems/orangefs.txt @@ -0,0 +1,406 @@ +ORANGEFS +======== + +OrangeFS is an LGPL userspace scale-out parallel storage system. It is ideal +for large storage problems faced by HPC, BigData, Streaming Video, +Genomics, Bioinformatics. + +Orangefs, originally called PVFS, was first developed in 1993 by +Walt Ligon and Eric Blumer as a parallel file system for Parallel +Virtual Machine (PVM) as part of a NASA grant to study the I/O patterns +of parallel programs. + +Orangefs features include: + + * Distributes file data among multiple file servers + * Supports simultaneous access by multiple clients + * Stores file data and metadata on servers using local file system + and access methods + * Userspace implementation is easy to install and maintain + * Direct MPI support + * Stateless + + +MAILING LIST +============ + +http://beowulf-underground.org/mailman/listinfo/pvfs2-users + + +DOCUMENTATION +============= + +http://www.orangefs.org/documentation/ + + +USERSPACE FILESYSTEM SOURCE +=========================== + +http://www.orangefs.org/download + +Orangefs versions prior to 2.9.3 would not be compatible with the +upstream version of the kernel client. + + +BUILDING THE USERSPACE FILESYSTEM ON A SINGLE SERVER +==================================================== + +When Orangefs is upstream, "--with-kernel" shouldn't be needed, but +until then the path to where the kernel with the Orangefs kernel client +patch was built is needed to ensure that pvfs2-client-core (the bridge +between kernel space and user space) will build properly. You can omit +--prefix if you don't care that things are sprinkled around in +/usr/local. + +./configure --prefix=/opt/ofs --with-kernel=/path/to/orangefs/kernel + +make + +make install + +Create an orangefs config file: +/opt/ofs/bin/pvfs2-genconfig /etc/pvfs2.conf + + for "Enter hostnames", use the hostname, don't let it default to + localhost. + +create a pvfs2tab file in /etc: +cat /etc/pvfs2tab +tcp://myhostname:3334/orangefs /mymountpoint pvfs2 defaults,noauto 0 0 + +create the mount point you specified in the tab file if needed: +mkdir /mymountpoint + +bootstrap the server: +/opt/ofs/sbin/pvfs2-server /etc/pvfs2.conf -f + +start the server: +/opt/osf/sbin/pvfs2-server /etc/pvfs2.conf + +Now the server is running. At this point you might like to +prove things are working with: + +/opt/osf/bin/pvfs2-ls /mymountpoint + +You might not want to enforce selinux, it doesn't seem to matter by +linux 3.11... + +If stuff seems to be working, turn on the client core: +/opt/osf/sbin/pvfs2-client -p /opt/osf/sbin/pvfs2-client-core + +Mount your filesystem. +mount -t pvfs2 tcp://myhostname:3334/orangefs /mymountpoint + + +OPTIONS +======= + +The following mount options are accepted: + + acl + Allow the use of Access Control Lists on files and directories. + + intr + Some operations between the kernel client and the user space + filesystem can be interruptible, such as changes in debug levels + and the setting of tunable parameters. + + local_lock + Enable posix locking from the perspective of "this" kernel. The + default file_operations lock action is to return ENOSYS. Posix + locking kicks in if the filesystem is mounted with -o local_lock. + Distributed locking is being worked on for the future. + + +DEBUGGING +========= + +If you want the debug (GOSSIP) statements in a particular +source file (inode.c for example) go to syslog: + + echo inode > /sys/kernel/debug/orangefs/kernel-debug + +No debugging (the default): + + echo none > /sys/kernel/debug/orangefs/kernel-debug + +Debugging from several source files: + + echo inode,dir > /sys/kernel/debug/orangefs/kernel-debug + +All debugging: + + echo all > /sys/kernel/debug/orangefs/kernel-debug + +Get a list of all debugging keywords: + + cat /sys/kernel/debug/orangefs/debug-help + + +PROTOCOL BETWEEN KERNEL MODULE AND USERSPACE +============================================ + +Orangefs is a user space filesystem and an associated kernel module. +We'll just refer to the user space part of Orangefs as "userspace" +from here on out. Orangefs descends from PVFS, and userspace code +still uses PVFS for function and variable names. Userspace typedefs +many of the important structures. Function and variable names in +the kernel module have been transitioned to "orangefs", and The Linux +Coding Style avoids typedefs, so kernel module structures that +correspond to userspace structures are not typedefed. + +The kernel module implements a pseudo device that userspace +can read from and write to. Userspace can also manipulate the +kernel module through the pseudo device with ioctl. + +THE BUFMAP: + +At startup userspace allocates two page-size-aligned (posix_memalign) +mlocked memory buffers, one is used for IO and one is used for readdir +operations. The IO buffer is 41943040 bytes and the readdir buffer is +4194304 bytes. Each buffer contains logical chunks, or partitions, and +a pointer to each buffer is added to its own PVFS_dev_map_desc structure +which also describes its total size, as well as the size and number of +the partitions. + +A pointer to the IO buffer's PVFS_dev_map_desc structure is sent to a +mapping routine in the kernel module with an ioctl. The structure is +copied from user space to kernel space with copy_from_user and is used +to initialize the kernel module's "bufmap" (struct orangefs_bufmap), which +then contains: + + * refcnt - a reference counter + * desc_size - PVFS2_BUFMAP_DEFAULT_DESC_SIZE (4194304) - the IO buffer's + partition size, which represents the filesystem's block size and + is used for s_blocksize in super blocks. + * desc_count - PVFS2_BUFMAP_DEFAULT_DESC_COUNT (10) - the number of + partitions in the IO buffer. + * desc_shift - log2(desc_size), used for s_blocksize_bits in super blocks. + * total_size - the total size of the IO buffer. + * page_count - the number of 4096 byte pages in the IO buffer. + * page_array - a pointer to page_count * (sizeof(struct page*)) bytes + of kcalloced memory. This memory is used as an array of pointers + to each of the pages in the IO buffer through a call to get_user_pages. + * desc_array - a pointer to desc_count * (sizeof(struct orangefs_bufmap_desc)) + bytes of kcalloced memory. This memory is further intialized: + + user_desc is the kernel's copy of the IO buffer's ORANGEFS_dev_map_desc + structure. user_desc->ptr points to the IO buffer. + + pages_per_desc = bufmap->desc_size / PAGE_SIZE + offset = 0 + + bufmap->desc_array[0].page_array = &bufmap->page_array[offset] + bufmap->desc_array[0].array_count = pages_per_desc = 1024 + bufmap->desc_array[0].uaddr = (user_desc->ptr) + (0 * 1024 * 4096) + offset += 1024 + . + . + . + bufmap->desc_array[9].page_array = &bufmap->page_array[offset] + bufmap->desc_array[9].array_count = pages_per_desc = 1024 + bufmap->desc_array[9].uaddr = (user_desc->ptr) + + (9 * 1024 * 4096) + offset += 1024 + + * buffer_index_array - a desc_count sized array of ints, used to + indicate which of the IO buffer's partitions are available to use. + * buffer_index_lock - a spinlock to protect buffer_index_array during update. + * readdir_index_array - a five (ORANGEFS_READDIR_DEFAULT_DESC_COUNT) element + int array used to indicate which of the readdir buffer's partitions are + available to use. + * readdir_index_lock - a spinlock to protect readdir_index_array during + update. + +OPERATIONS: + +The kernel module builds an "op" (struct orangefs_kernel_op_s) when it +needs to communicate with userspace. Part of the op contains the "upcall" +which expresses the request to userspace. Part of the op eventually +contains the "downcall" which expresses the results of the request. + +The slab allocator is used to keep a cache of op structures handy. + +At init time the kernel module defines and initializes a request list +and an in_progress hash table to keep track of all the ops that are +in flight at any given time. + +Ops are stateful: + + * unknown - op was just initialized + * waiting - op is on request_list (upward bound) + * inprogr - op is in progress (waiting for downcall) + * serviced - op has matching downcall; ok + * purged - op has to start a timer since client-core + exited uncleanly before servicing op + * given up - submitter has given up waiting for it + +When some arbitrary userspace program needs to perform a +filesystem operation on Orangefs (readdir, I/O, create, whatever) +an op structure is initialized and tagged with a distinguishing ID +number. The upcall part of the op is filled out, and the op is +passed to the "service_operation" function. + +Service_operation changes the op's state to "waiting", puts +it on the request list, and signals the Orangefs file_operations.poll +function through a wait queue. Userspace is polling the pseudo-device +and thus becomes aware of the upcall request that needs to be read. + +When the Orangefs file_operations.read function is triggered, the +request list is searched for an op that seems ready-to-process. +The op is removed from the request list. The tag from the op and +the filled-out upcall struct are copy_to_user'ed back to userspace. + +If any of these (and some additional protocol) copy_to_users fail, +the op's state is set to "waiting" and the op is added back to +the request list. Otherwise, the op's state is changed to "in progress", +and the op is hashed on its tag and put onto the end of a list in the +in_progress hash table at the index the tag hashed to. + +When userspace has assembled the response to the upcall, it +writes the response, which includes the distinguishing tag, back to +the pseudo device in a series of io_vecs. This triggers the Orangefs +file_operations.write_iter function to find the op with the associated +tag and remove it from the in_progress hash table. As long as the op's +state is not "canceled" or "given up", its state is set to "serviced". +The file_operations.write_iter function returns to the waiting vfs, +and back to service_operation through wait_for_matching_downcall. + +Service operation returns to its caller with the op's downcall +part (the response to the upcall) filled out. + +The "client-core" is the bridge between the kernel module and +userspace. The client-core is a daemon. The client-core has an +associated watchdog daemon. If the client-core is ever signaled +to die, the watchdog daemon restarts the client-core. Even though +the client-core is restarted "right away", there is a period of +time during such an event that the client-core is dead. A dead client-core +can't be triggered by the Orangefs file_operations.poll function. +Ops that pass through service_operation during a "dead spell" can timeout +on the wait queue and one attempt is made to recycle them. Obviously, +if the client-core stays dead too long, the arbitrary userspace processes +trying to use Orangefs will be negatively affected. Waiting ops +that can't be serviced will be removed from the request list and +have their states set to "given up". In-progress ops that can't +be serviced will be removed from the in_progress hash table and +have their states set to "given up". + +Readdir and I/O ops are atypical with respect to their payloads. + + - readdir ops use the smaller of the two pre-allocated pre-partitioned + memory buffers. The readdir buffer is only available to userspace. + The kernel module obtains an index to a free partition before launching + a readdir op. Userspace deposits the results into the indexed partition + and then writes them to back to the pvfs device. + + - io (read and write) ops use the larger of the two pre-allocated + pre-partitioned memory buffers. The IO buffer is accessible from + both userspace and the kernel module. The kernel module obtains an + index to a free partition before launching an io op. The kernel module + deposits write data into the indexed partition, to be consumed + directly by userspace. Userspace deposits the results of read + requests into the indexed partition, to be consumed directly + by the kernel module. + +Responses to kernel requests are all packaged in pvfs2_downcall_t +structs. Besides a few other members, pvfs2_downcall_t contains a +union of structs, each of which is associated with a particular +response type. + +The several members outside of the union are: + - int32_t type - type of operation. + - int32_t status - return code for the operation. + - int64_t trailer_size - 0 unless readdir operation. + - char *trailer_buf - initialized to NULL, used during readdir operations. + +The appropriate member inside the union is filled out for any +particular response. + + PVFS2_VFS_OP_FILE_IO + fill a pvfs2_io_response_t + + PVFS2_VFS_OP_LOOKUP + fill a PVFS_object_kref + + PVFS2_VFS_OP_CREATE + fill a PVFS_object_kref + + PVFS2_VFS_OP_SYMLINK + fill a PVFS_object_kref + + PVFS2_VFS_OP_GETATTR + fill in a PVFS_sys_attr_s (tons of stuff the kernel doesn't need) + fill in a string with the link target when the object is a symlink. + + PVFS2_VFS_OP_MKDIR + fill a PVFS_object_kref + + PVFS2_VFS_OP_STATFS + fill a pvfs2_statfs_response_t with useless info <g>. It is hard for + us to know, in a timely fashion, these statistics about our + distributed network filesystem. + + PVFS2_VFS_OP_FS_MOUNT + fill a pvfs2_fs_mount_response_t which is just like a PVFS_object_kref + except its members are in a different order and "__pad1" is replaced + with "id". + + PVFS2_VFS_OP_GETXATTR + fill a pvfs2_getxattr_response_t + + PVFS2_VFS_OP_LISTXATTR + fill a pvfs2_listxattr_response_t + + PVFS2_VFS_OP_PARAM + fill a pvfs2_param_response_t + + PVFS2_VFS_OP_PERF_COUNT + fill a pvfs2_perf_count_response_t + + PVFS2_VFS_OP_FSKEY + file a pvfs2_fs_key_response_t + + PVFS2_VFS_OP_READDIR + jamb everything needed to represent a pvfs2_readdir_response_t into + the readdir buffer descriptor specified in the upcall. + +Userspace uses writev() on /dev/pvfs2-req to pass responses to the requests +made by the kernel side. + +A buffer_list containing: + - a pointer to the prepared response to the request from the + kernel (struct pvfs2_downcall_t). + - and also, in the case of a readdir request, a pointer to a + buffer containing descriptors for the objects in the target + directory. +... is sent to the function (PINT_dev_write_list) which performs +the writev. + +PINT_dev_write_list has a local iovec array: struct iovec io_array[10]; + +The first four elements of io_array are initialized like this for all +responses: + + io_array[0].iov_base = address of local variable "proto_ver" (int32_t) + io_array[0].iov_len = sizeof(int32_t) + + io_array[1].iov_base = address of global variable "pdev_magic" (int32_t) + io_array[1].iov_len = sizeof(int32_t) + + io_array[2].iov_base = address of parameter "tag" (PVFS_id_gen_t) + io_array[2].iov_len = sizeof(int64_t) + + io_array[3].iov_base = address of out_downcall member (pvfs2_downcall_t) + of global variable vfs_request (vfs_request_t) + io_array[3].iov_len = sizeof(pvfs2_downcall_t) + +Readdir responses initialize the fifth element io_array like this: + + io_array[4].iov_base = contents of member trailer_buf (char *) + from out_downcall member of global variable + vfs_request + io_array[4].iov_len = contents of member trailer_size (PVFS_size) + from out_downcall member of global variable + vfs_request + + diff --git a/MAINTAINERS b/MAINTAINERS index df8cf6b924c6..03e00c7c88eb 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8251,6 +8251,14 @@ S: Supported F: fs/overlayfs/ F: Documentation/filesystems/overlayfs.txt +ORANGEFS FILESYSTEM +M: Mike Marshall <hubcap@omnibond.com> +L: pvfs2-developers@beowulf-underground.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux.git +S: Supported +F: fs/orangefs/ +F: Documentation/filesystems/orangefs.txt + P54 WIRELESS DRIVER M: Christian Lamparter <chunkeey@googlemail.com> L: linux-wireless@vger.kernel.org diff --git a/fs/Kconfig b/fs/Kconfig index 9d757673bf40..6725f59c18e6 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -209,6 +209,7 @@ menuconfig MISC_FILESYSTEMS if MISC_FILESYSTEMS +source "fs/orangefs/Kconfig" source "fs/adfs/Kconfig" source "fs/affs/Kconfig" source "fs/ecryptfs/Kconfig" diff --git a/fs/Makefile b/fs/Makefile index 252c96898a43..85b6e13b62d3 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -106,6 +106,7 @@ obj-$(CONFIG_AUTOFS4_FS) += autofs4/ obj-$(CONFIG_ADFS_FS) += adfs/ obj-$(CONFIG_FUSE_FS) += fuse/ obj-$(CONFIG_OVERLAY_FS) += overlayfs/ +obj-$(CONFIG_ORANGEFS_FS) += orangefs/ obj-$(CONFIG_UDF_FS) += udf/ obj-$(CONFIG_SUN_OPENPROMFS) += openpromfs/ obj-$(CONFIG_OMFS_FS) += omfs/ diff --git a/fs/orangefs/Kconfig b/fs/orangefs/Kconfig new file mode 100644 index 000000000000..1554c02489de --- /dev/null +++ b/fs/orangefs/Kconfig @@ -0,0 +1,6 @@ +config ORANGEFS_FS + tristate "ORANGEFS (Powered by PVFS) support" + select FS_POSIX_ACL + help + Orange is a parallel file system designed for use on high end + computing (HEC) systems. diff --git a/fs/orangefs/Makefile b/fs/orangefs/Makefile new file mode 100644 index 000000000000..a9d6a968fe6d --- /dev/null +++ b/fs/orangefs/Makefile @@ -0,0 +1,10 @@ +# +# Makefile for the ORANGEFS filesystem. +# + +obj-$(CONFIG_ORANGEFS_FS) += orangefs.o + +orangefs-objs := acl.o file.o orangefs-cache.o orangefs-utils.o xattr.o \ + dcache.o inode.o orangefs-sysfs.o orangefs-mod.o super.o \ + devorangefs-req.o namei.o symlink.o dir.o orangefs-bufmap.o \ + orangefs-debugfs.o waitqueue.o diff --git a/fs/orangefs/acl.c b/fs/orangefs/acl.c new file mode 100644 index 000000000000..03f89dbb2512 --- /dev/null +++ b/fs/orangefs/acl.c @@ -0,0 +1,175 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/posix_acl_xattr.h> +#include <linux/fs_struct.h> + +struct posix_acl *orangefs_get_acl(struct inode *inode, int type) +{ + struct posix_acl *acl; + int ret; + char *key = NULL, *value = NULL; + + switch (type) { + case ACL_TYPE_ACCESS: + key = ORANGEFS_XATTR_NAME_ACL_ACCESS; + break; + case ACL_TYPE_DEFAULT: + key = ORANGEFS_XATTR_NAME_ACL_DEFAULT; + break; + default: + gossip_err("orangefs_get_acl: bogus value of type %d\n", type); + return ERR_PTR(-EINVAL); + } + /* + * Rather than incurring a network call just to determine the exact + * length of the attribute, I just allocate a max length to save on + * the network call. Conceivably, we could pass NULL to + * orangefs_inode_getxattr() to probe the length of the value, but + * I don't do that for now. + */ + value = kmalloc(ORANGEFS_MAX_XATTR_VALUELEN, GFP_KERNEL); + if (value == NULL) + return ERR_PTR(-ENOMEM); + + gossip_debug(GOSSIP_ACL_DEBUG, + "inode %pU, key %s, type %d\n", + get_khandle_from_ino(inode), + key, + type); + ret = orangefs_inode_getxattr(inode, + "", + key, + value, + ORANGEFS_MAX_XATTR_VALUELEN); + /* if the key exists, convert it to an in-memory rep */ + if (ret > 0) { + acl = posix_acl_from_xattr(&init_user_ns, value, ret); + } else if (ret == -ENODATA || ret == -ENOSYS) { + acl = NULL; + } else { + gossip_err("inode %pU retrieving acl's failed with error %d\n", + get_khandle_from_ino(inode), + ret); + acl = ERR_PTR(ret); + } + /* kfree(NULL) is safe, so don't worry if value ever got used */ + kfree(value); + return acl; +} + +int orangefs_set_acl(struct inode *inode, struct posix_acl *acl, int type) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + int error = 0; + void *value = NULL; + size_t size = 0; + const char *name = NULL; + + switch (type) { + case ACL_TYPE_ACCESS: + name = ORANGEFS_XATTR_NAME_ACL_ACCESS; + if (acl) { + umode_t mode = inode->i_mode; + /* + * can we represent this with the traditional file + * mode permission bits? + */ + error = posix_acl_equiv_mode(acl, &mode); + if (error < 0) { + gossip_err("%s: posix_acl_equiv_mode err: %d\n", + __func__, + error); + return error; + } + + if (inode->i_mode != mode) + SetModeFlag(orangefs_inode); + inode->i_mode = mode; + mark_inode_dirty_sync(inode); + if (error == 0) + acl = NULL; + } + break; + case ACL_TYPE_DEFAULT: + name = ORANGEFS_XATTR_NAME_ACL_DEFAULT; + break; + default: + gossip_err("%s: invalid type %d!\n", __func__, type); + return -EINVAL; + } + + gossip_debug(GOSSIP_ACL_DEBUG, + "%s: inode %pU, key %s type %d\n", + __func__, get_khandle_from_ino(inode), + name, + type); + + if (acl) { + size = posix_acl_xattr_size(acl->a_count); + value = kmalloc(size, GFP_KERNEL); + if (!value) + return -ENOMEM; + + error = posix_acl_to_xattr(&init_user_ns, acl, value, size); + if (error < 0) + goto out; + } + + gossip_debug(GOSSIP_ACL_DEBUG, + "%s: name %s, value %p, size %zd, acl %p\n", + __func__, name, value, size, acl); + /* + * Go ahead and set the extended attribute now. NOTE: Suppose acl + * was NULL, then value will be NULL and size will be 0 and that + * will xlate to a removexattr. However, we don't want removexattr + * complain if attributes does not exist. + */ + error = orangefs_inode_setxattr(inode, "", name, value, size, 0); + +out: + kfree(value); + if (!error) + set_cached_acl(inode, type, acl); + return error; +} + +int orangefs_init_acl(struct inode *inode, struct inode *dir) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct posix_acl *default_acl, *acl; + umode_t mode = inode->i_mode; + int error = 0; + + ClearModeFlag(orangefs_inode); + + error = posix_acl_create(dir, &mode, &default_acl, &acl); + if (error) + return error; + + if (default_acl) { + error = orangefs_set_acl(inode, default_acl, ACL_TYPE_DEFAULT); + posix_acl_release(default_acl); + } + + if (acl) { + if (!error) + error = orangefs_set_acl(inode, acl, ACL_TYPE_ACCESS); + posix_acl_release(acl); + } + + /* If mode of the inode was changed, then do a forcible ->setattr */ + if (mode != inode->i_mode) { + SetModeFlag(orangefs_inode); + inode->i_mode = mode; + orangefs_flush_inode(inode); + } + + return error; +} diff --git a/fs/orangefs/dcache.c b/fs/orangefs/dcache.c new file mode 100644 index 000000000000..5dfc4f3cfe68 --- /dev/null +++ b/fs/orangefs/dcache.c @@ -0,0 +1,138 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Implementation of dentry (directory cache) functions. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* Returns 1 if dentry can still be trusted, else 0. */ +static int orangefs_revalidate_lookup(struct dentry *dentry) +{ + struct dentry *parent_dentry = dget_parent(dentry); + struct inode *parent_inode = parent_dentry->d_inode; + struct orangefs_inode_s *parent = ORANGEFS_I(parent_inode); + struct inode *inode = dentry->d_inode; + struct orangefs_kernel_op_s *new_op; + int ret = 0; + int err = 0; + + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: attempting lookup.\n", __func__); + + new_op = op_alloc(ORANGEFS_VFS_OP_LOOKUP); + if (!new_op) + goto out_put_parent; + + new_op->upcall.req.lookup.sym_follow = ORANGEFS_LOOKUP_LINK_NO_FOLLOW; + new_op->upcall.req.lookup.parent_refn = parent->refn; + strncpy(new_op->upcall.req.lookup.d_name, + dentry->d_name.name, + ORANGEFS_NAME_MAX); + + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d interrupt flag [%d]\n", + __FILE__, + __func__, + __LINE__, + get_interruptible_flag(parent_inode)); + + err = service_operation(new_op, "orangefs_lookup", + get_interruptible_flag(parent_inode)); + + /* Positive dentry: reject if error or not the same inode. */ + if (inode) { + if (err) { + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d lookup failure.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + if (!match_handle(new_op->downcall.resp.lookup.refn.khandle, + inode)) { + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d no match.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + + /* Negative dentry: reject if success or error other than ENOENT. */ + } else { + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: negative dentry.\n", + __func__); + if (!err || err != -ENOENT) { + if (new_op->downcall.status != 0) + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d lookup failure.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + } + + ret = 1; +out_release_op: + op_release(new_op); +out_put_parent: + dput(parent_dentry); + return ret; +out_drop: + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s:%s:%d revalidate failed\n", + __FILE__, __func__, __LINE__); + goto out_release_op; +} + +/* + * Verify that dentry is valid. + * + * Should return 1 if dentry can still be trusted, else 0. + */ +static int orangefs_d_revalidate(struct dentry *dentry, unsigned int flags) +{ + int ret; + + if (flags & LOOKUP_RCU) + return -ECHILD; + + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: called on dentry %p.\n", + __func__, dentry); + + /* skip root handle lookups. */ + if (dentry->d_inode && is_root_handle(dentry->d_inode)) + return 1; + + /* + * If this passes, the positive dentry still exists or the negative + * dentry still does not exist. + */ + if (!orangefs_revalidate_lookup(dentry)) + return 0; + + /* We do not need to continue with negative dentries. */ + if (!dentry->d_inode) + goto out; + + /* Now we must perform a getattr to validate the inode contents. */ + + ret = orangefs_inode_check_changed(dentry->d_inode); + if (ret < 0) { + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s:%s:%d getattr failure.\n", + __FILE__, __func__, __LINE__); + return 0; + } + if (ret == 0) + return 0; + +out: + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: negative dentry or positive dentry and inode valid.\n", + __func__); + return 1; +} + +const struct dentry_operations orangefs_dentry_operations = { + .d_revalidate = orangefs_d_revalidate, +}; diff --git a/fs/orangefs/devorangefs-req.c b/fs/orangefs/devorangefs-req.c new file mode 100644 index 000000000000..db170beba797 --- /dev/null +++ b/fs/orangefs/devorangefs-req.c @@ -0,0 +1,943 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * Changes by Acxiom Corporation to add protocol version to kernel + * communication, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-dev-proto.h" +#include "orangefs-bufmap.h" + +#include <linux/debugfs.h> +#include <linux/slab.h> + +/* this file implements the /dev/pvfs2-req device node */ + +static int open_access_count; + +#define DUMP_DEVICE_ERROR() \ +do { \ + gossip_err("*****************************************************\n");\ + gossip_err("ORANGEFS Device Error: You cannot open the device file "); \ + gossip_err("\n/dev/%s more than once. Please make sure that\nthere " \ + "are no ", ORANGEFS_REQDEVICE_NAME); \ + gossip_err("instances of a program using this device\ncurrently " \ + "running. (You must verify this!)\n"); \ + gossip_err("For example, you can use the lsof program as follows:\n");\ + gossip_err("'lsof | grep %s' (run this as root)\n", \ + ORANGEFS_REQDEVICE_NAME); \ + gossip_err(" open_access_count = %d\n", open_access_count); \ + gossip_err("*****************************************************\n");\ +} while (0) + +static int hash_func(__u64 tag, int table_size) +{ + return do_div(tag, (unsigned int)table_size); +} + +static void orangefs_devreq_add_op(struct orangefs_kernel_op_s *op) +{ + int index = hash_func(op->tag, hash_table_size); + + list_add_tail(&op->list, &htable_ops_in_progress[index]); +} + +/* + * find the op with this tag and remove it from the in progress + * hash table. + */ +static struct orangefs_kernel_op_s *orangefs_devreq_remove_op(__u64 tag) +{ + struct orangefs_kernel_op_s *op, *next; + int index; + + index = hash_func(tag, hash_table_size); + + spin_lock(&htable_ops_in_progress_lock); + list_for_each_entry_safe(op, + next, + &htable_ops_in_progress[index], + list) { + if (op->tag == tag && !op_state_purged(op) && + !op_state_given_up(op)) { + list_del_init(&op->list); + spin_unlock(&htable_ops_in_progress_lock); + return op; + } + } + + spin_unlock(&htable_ops_in_progress_lock); + return NULL; +} + +/* Returns whether any FS are still pending remounted */ +static int mark_all_pending_mounts(void) +{ + int unmounted = 1; + struct orangefs_sb_info_s *orangefs_sb = NULL; + + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + /* All of these file system require a remount */ + orangefs_sb->mount_pending = 1; + unmounted = 0; + } + spin_unlock(&orangefs_superblocks_lock); + return unmounted; +} + +/* + * Determine if a given file system needs to be remounted or not + * Returns -1 on error + * 0 if already mounted + * 1 if needs remount + */ +static int fs_mount_pending(__s32 fsid) +{ + int mount_pending = -1; + struct orangefs_sb_info_s *orangefs_sb = NULL; + + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + if (orangefs_sb->fs_id == fsid) { + mount_pending = orangefs_sb->mount_pending; + break; + } + } + spin_unlock(&orangefs_superblocks_lock); + return mount_pending; +} + +static int orangefs_devreq_open(struct inode *inode, struct file *file) +{ + int ret = -EINVAL; + + if (!(file->f_flags & O_NONBLOCK)) { + gossip_err("%s: device cannot be opened in blocking mode\n", + __func__); + goto out; + } + ret = -EACCES; + gossip_debug(GOSSIP_DEV_DEBUG, "client-core: opening device\n"); + mutex_lock(&devreq_mutex); + + if (open_access_count == 0) { + open_access_count = 1; + ret = 0; + } else { + DUMP_DEVICE_ERROR(); + } + mutex_unlock(&devreq_mutex); + +out: + + gossip_debug(GOSSIP_DEV_DEBUG, + "pvfs2-client-core: open device complete (ret = %d)\n", + ret); + return ret; +} + +/* Function for read() callers into the device */ +static ssize_t orangefs_devreq_read(struct file *file, + char __user *buf, + size_t count, loff_t *offset) +{ + struct orangefs_kernel_op_s *op, *temp; + __s32 proto_ver = ORANGEFS_KERNEL_PROTO_VERSION; + static __s32 magic = ORANGEFS_DEVREQ_MAGIC; + struct orangefs_kernel_op_s *cur_op = NULL; + unsigned long ret; + + /* We do not support blocking IO. */ + if (!(file->f_flags & O_NONBLOCK)) { + gossip_err("%s: blocking read from client-core.\n", + __func__); + return -EINVAL; + } + + /* + * The client will do an ioctl to find MAX_DEV_REQ_UPSIZE, then + * always read with that size buffer. + */ + if (count != MAX_DEV_REQ_UPSIZE) { + gossip_err("orangefs: client-core tried to read wrong size\n"); + return -EINVAL; + } + +restart: + /* Get next op (if any) from top of list. */ + spin_lock(&orangefs_request_list_lock); + list_for_each_entry_safe(op, temp, &orangefs_request_list, list) { + __s32 fsid; + /* This lock is held past the end of the loop when we break. */ + spin_lock(&op->lock); + if (unlikely(op_state_purged(op) || op_state_given_up(op))) { + spin_unlock(&op->lock); + continue; + } + + fsid = fsid_of_op(op); + if (fsid != ORANGEFS_FS_ID_NULL) { + int ret; + /* Skip ops whose filesystem needs to be mounted. */ + ret = fs_mount_pending(fsid); + if (ret == 1) { + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: mount pending, skipping op tag " + "%llu %s\n", + __func__, + llu(op->tag), + get_opname_string(op)); + spin_unlock(&op->lock); + continue; + /* + * Skip ops whose filesystem we don't know about unless + * it is being mounted. + */ + /* XXX: is there a better way to detect this? */ + } else if (ret == -1 && + !(op->upcall.type == + ORANGEFS_VFS_OP_FS_MOUNT || + op->upcall.type == + ORANGEFS_VFS_OP_GETATTR)) { + gossip_debug(GOSSIP_DEV_DEBUG, + "orangefs: skipping op tag %llu %s\n", + llu(op->tag), get_opname_string(op)); + gossip_err( + "orangefs: ERROR: fs_mount_pending %d\n", + fsid); + spin_unlock(&op->lock); + continue; + } + } + /* + * Either this op does not pertain to a filesystem, is mounting + * a filesystem, or pertains to a mounted filesystem. Let it + * through. + */ + cur_op = op; + break; + } + + /* + * At this point we either have a valid op and can continue or have not + * found an op and must ask the client to try again later. + */ + if (!cur_op) { + spin_unlock(&orangefs_request_list_lock); + return -EAGAIN; + } + + gossip_debug(GOSSIP_DEV_DEBUG, "%s: reading op tag %llu %s\n", + __func__, + llu(cur_op->tag), + get_opname_string(cur_op)); + + /* + * Such an op should never be on the list in the first place. If so, we + * will abort. + */ + if (op_state_in_progress(cur_op) || op_state_serviced(cur_op)) { + gossip_err("orangefs: ERROR: Current op already queued.\n"); + list_del_init(&cur_op->list); + spin_unlock(&cur_op->lock); + spin_unlock(&orangefs_request_list_lock); + return -EAGAIN; + } + + list_del_init(&cur_op->list); + spin_unlock(&orangefs_request_list_lock); + + spin_unlock(&cur_op->lock); + + /* Push the upcall out. */ + ret = copy_to_user(buf, &proto_ver, sizeof(__s32)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+sizeof(__s32), &magic, sizeof(__s32)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+2 * sizeof(__s32), &cur_op->tag, sizeof(__u64)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+2*sizeof(__s32)+sizeof(__u64), &cur_op->upcall, + sizeof(struct orangefs_upcall_s)); + if (ret != 0) + goto error; + + spin_lock(&htable_ops_in_progress_lock); + spin_lock(&cur_op->lock); + if (unlikely(op_state_given_up(cur_op))) { + spin_unlock(&cur_op->lock); + spin_unlock(&htable_ops_in_progress_lock); + complete(&cur_op->waitq); + goto restart; + } + + /* + * Set the operation to be in progress and move it between lists since + * it has been sent to the client. + */ + set_op_state_inprogress(cur_op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: 1 op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(cur_op), + cur_op->op_state, + current->comm); + orangefs_devreq_add_op(cur_op); + spin_unlock(&cur_op->lock); + spin_unlock(&htable_ops_in_progress_lock); + + /* The client only asks to read one size buffer. */ + return MAX_DEV_REQ_UPSIZE; +error: + /* + * We were unable to copy the op data to the client. Put the op back in + * list. If client has crashed, the op will be purged later when the + * device is released. + */ + gossip_err("orangefs: Failed to copy data to user space\n"); + spin_lock(&orangefs_request_list_lock); + spin_lock(&cur_op->lock); + if (likely(!op_state_given_up(cur_op))) { + set_op_state_waiting(cur_op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: 2 op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(cur_op), + cur_op->op_state, + current->comm); + list_add(&cur_op->list, &orangefs_request_list); + spin_unlock(&cur_op->lock); + } else { + spin_unlock(&cur_op->lock); + complete(&cur_op->waitq); + } + spin_unlock(&orangefs_request_list_lock); + return -EFAULT; +} + +/* + * Function for writev() callers into the device. + * + * Userspace should have written: + * - __u32 version + * - __u32 magic + * - __u64 tag + * - struct orangefs_downcall_s + * - trailer buffer (in the case of READDIR operations) + */ +static ssize_t orangefs_devreq_write_iter(struct kiocb *iocb, + struct iov_iter *iter) +{ + ssize_t ret; + struct orangefs_kernel_op_s *op = NULL; + struct { + __u32 version; + __u32 magic; + __u64 tag; + } head; + int total = ret = iov_iter_count(iter); + int n; + int downcall_size = sizeof(struct orangefs_downcall_s); + int head_size = sizeof(head); + + gossip_debug(GOSSIP_DEV_DEBUG, "%s: total:%d: ret:%zd:\n", + __func__, + total, + ret); + + if (total < MAX_DEV_REQ_DOWNSIZE) { + gossip_err("%s: total:%d: must be at least:%u:\n", + __func__, + total, + (unsigned int) MAX_DEV_REQ_DOWNSIZE); + return -EFAULT; + } + + n = copy_from_iter(&head, head_size, iter); + if (n < head_size) { + gossip_err("%s: failed to copy head.\n", __func__); + return -EFAULT; + } + + if (head.version < ORANGEFS_MINIMUM_USERSPACE_VERSION) { + gossip_err("%s: userspace claims version" + "%d, minimum version required: %d.\n", + __func__, + head.version, + ORANGEFS_MINIMUM_USERSPACE_VERSION); + return -EPROTO; + } + + if (head.magic != ORANGEFS_DEVREQ_MAGIC) { + gossip_err("Error: Device magic number does not match.\n"); + return -EPROTO; + } + + /* remove the op from the in progress hash table */ + op = orangefs_devreq_remove_op(head.tag); + if (!op) { + gossip_err("WARNING: No one's waiting for tag %llu\n", + llu(head.tag)); + return ret; + } + + n = copy_from_iter(&op->downcall, downcall_size, iter); + if (n != downcall_size) { + gossip_err("%s: failed to copy downcall.\n", __func__); + goto Efault; + } + + if (op->downcall.status) + goto wakeup; + + /* + * We've successfully peeled off the head and the downcall. + * Something has gone awry if total doesn't equal the + * sum of head_size, downcall_size and trailer_size. + */ + if ((head_size + downcall_size + op->downcall.trailer_size) != total) { + gossip_err("%s: funky write, head_size:%d" + ": downcall_size:%d: trailer_size:%lld" + ": total size:%d:\n", + __func__, + head_size, + downcall_size, + op->downcall.trailer_size, + total); + goto Efault; + } + + /* Only READDIR operations should have trailers. */ + if ((op->downcall.type != ORANGEFS_VFS_OP_READDIR) && + (op->downcall.trailer_size != 0)) { + gossip_err("%s: %x operation with trailer.", + __func__, + op->downcall.type); + goto Efault; + } + + /* READDIR operations should always have trailers. */ + if ((op->downcall.type == ORANGEFS_VFS_OP_READDIR) && + (op->downcall.trailer_size == 0)) { + gossip_err("%s: %x operation with no trailer.", + __func__, + op->downcall.type); + goto Efault; + } + + if (op->downcall.type != ORANGEFS_VFS_OP_READDIR) + goto wakeup; + + op->downcall.trailer_buf = + vmalloc(op->downcall.trailer_size); + if (op->downcall.trailer_buf == NULL) { + gossip_err("%s: failed trailer vmalloc.\n", + __func__); + goto Enomem; + } + memset(op->downcall.trailer_buf, 0, op->downcall.trailer_size); + n = copy_from_iter(op->downcall.trailer_buf, + op->downcall.trailer_size, + iter); + if (n != op->downcall.trailer_size) { + gossip_err("%s: failed to copy trailer.\n", __func__); + vfree(op->downcall.trailer_buf); + goto Efault; + } + +wakeup: + /* + * Return to vfs waitqueue, and back to service_operation + * through wait_for_matching_downcall. + */ + spin_lock(&op->lock); + if (unlikely(op_is_cancel(op))) { + spin_unlock(&op->lock); + put_cancel(op); + } else if (unlikely(op_state_given_up(op))) { + spin_unlock(&op->lock); + complete(&op->waitq); + } else { + set_op_state_serviced(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + spin_unlock(&op->lock); + } + return ret; + +Efault: + op->downcall.status = -(ORANGEFS_ERROR_BIT | 9); + ret = -EFAULT; + goto wakeup; + +Enomem: + op->downcall.status = -(ORANGEFS_ERROR_BIT | 8); + ret = -ENOMEM; + goto wakeup; +} + +/* + * NOTE: gets called when the last reference to this device is dropped. + * Using the open_access_count variable, we enforce a reference count + * on this file so that it can be opened by only one process at a time. + * the devreq_mutex is used to make sure all i/o has completed + * before we call orangefs_bufmap_finalize, and similar such tricky + * situations + */ +static int orangefs_devreq_release(struct inode *inode, struct file *file) +{ + int unmounted = 0; + + gossip_debug(GOSSIP_DEV_DEBUG, + "%s:pvfs2-client-core: exiting, closing device\n", + __func__); + + mutex_lock(&devreq_mutex); + orangefs_bufmap_finalize(); + + open_access_count = -1; + + unmounted = mark_all_pending_mounts(); + gossip_debug(GOSSIP_DEV_DEBUG, "ORANGEFS Device Close: Filesystem(s) %s\n", + (unmounted ? "UNMOUNTED" : "MOUNTED")); + + purge_waiting_ops(); + purge_inprogress_ops(); + + orangefs_bufmap_run_down(); + + gossip_debug(GOSSIP_DEV_DEBUG, + "pvfs2-client-core: device close complete\n"); + open_access_count = 0; + mutex_unlock(&devreq_mutex); + return 0; +} + +int is_daemon_in_service(void) +{ + int in_service; + + /* + * What this function does is checks if client-core is alive + * based on the access count we maintain on the device. + */ + mutex_lock(&devreq_mutex); + in_service = open_access_count == 1 ? 0 : -EIO; + mutex_unlock(&devreq_mutex); + return in_service; +} + +bool __is_daemon_in_service(void) +{ + return open_access_count == 1; +} + +static inline long check_ioctl_command(unsigned int command) +{ + /* Check for valid ioctl codes */ + if (_IOC_TYPE(command) != ORANGEFS_DEV_MAGIC) { + gossip_err("device ioctl magic numbers don't match! Did you rebuild pvfs2-client-core/libpvfs2? [cmd %x, magic %x != %x]\n", + command, + _IOC_TYPE(command), + ORANGEFS_DEV_MAGIC); + return -EINVAL; + } + /* and valid ioctl commands */ + if (_IOC_NR(command) >= ORANGEFS_DEV_MAXNR || _IOC_NR(command) <= 0) { + gossip_err("Invalid ioctl command number [%d >= %d]\n", + _IOC_NR(command), ORANGEFS_DEV_MAXNR); + return -ENOIOCTLCMD; + } + return 0; +} + +static long dispatch_ioctl_command(unsigned int command, unsigned long arg) +{ + static __s32 magic = ORANGEFS_DEVREQ_MAGIC; + static __s32 max_up_size = MAX_DEV_REQ_UPSIZE; + static __s32 max_down_size = MAX_DEV_REQ_DOWNSIZE; + struct ORANGEFS_dev_map_desc user_desc; + int ret = 0; + struct dev_mask_info_s mask_info = { 0 }; + struct dev_mask2_info_s mask2_info = { 0, 0 }; + int upstream_kmod = 1; + struct orangefs_sb_info_s *orangefs_sb; + + /* mtmoore: add locking here */ + + switch (command) { + case ORANGEFS_DEV_GET_MAGIC: + return ((put_user(magic, (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_GET_MAX_UPSIZE: + return ((put_user(max_up_size, + (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_GET_MAX_DOWNSIZE: + return ((put_user(max_down_size, + (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_MAP: + ret = copy_from_user(&user_desc, + (struct ORANGEFS_dev_map_desc __user *) + arg, + sizeof(struct ORANGEFS_dev_map_desc)); + /* WTF -EIO and not -EFAULT? */ + return ret ? -EIO : orangefs_bufmap_initialize(&user_desc); + case ORANGEFS_DEV_REMOUNT_ALL: + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: got ORANGEFS_DEV_REMOUNT_ALL\n", + __func__); + + /* + * remount all mounted orangefs volumes to regain the lost + * dynamic mount tables (if any) -- NOTE: this is done + * without keeping the superblock list locked due to the + * upcall/downcall waiting. also, the request mutex is + * used to ensure that no operations will be serviced until + * all of the remounts are serviced (to avoid ops between + * mounts to fail) + */ + ret = mutex_lock_interruptible(&request_mutex); + if (ret < 0) + return ret; + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: priority remount in progress\n", + __func__); + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + /* + * We have to drop the spinlock, so entries can be + * removed. They can't be freed, though, so we just + * keep the forward pointers and zero the back ones - + * that way we can get to the rest of the list. + */ + if (!orangefs_sb->list.prev) + continue; + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: Remounting SB %p\n", + __func__, + orangefs_sb); + + spin_unlock(&orangefs_superblocks_lock); + ret = orangefs_remount(orangefs_sb); + spin_lock(&orangefs_superblocks_lock); + if (ret) { + gossip_debug(GOSSIP_DEV_DEBUG, + "SB %p remount failed\n", + orangefs_sb); + break; + } + } + spin_unlock(&orangefs_superblocks_lock); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: priority remount complete\n", + __func__); + mutex_unlock(&request_mutex); + return ret; + + case ORANGEFS_DEV_UPSTREAM: + ret = copy_to_user((void __user *)arg, + &upstream_kmod, + sizeof(upstream_kmod)); + + if (ret != 0) + return -EIO; + else + return ret; + + case ORANGEFS_DEV_CLIENT_MASK: + ret = copy_from_user(&mask2_info, + (void __user *)arg, + sizeof(struct dev_mask2_info_s)); + + if (ret != 0) + return -EIO; + + client_debug_mask.mask1 = mask2_info.mask1_value; + client_debug_mask.mask2 = mask2_info.mask2_value; + + pr_info("%s: client debug mask has been been received " + ":%llx: :%llx:\n", + __func__, + (unsigned long long)client_debug_mask.mask1, + (unsigned long long)client_debug_mask.mask2); + + return ret; + + case ORANGEFS_DEV_CLIENT_STRING: + ret = copy_from_user(&client_debug_array_string, + (void __user *)arg, + ORANGEFS_MAX_DEBUG_STRING_LEN); + /* + * The real client-core makes an effort to ensure + * that actual strings that aren't too long to fit in + * this buffer is what we get here. We're going to use + * string functions on the stuff we got, so we'll make + * this extra effort to try and keep from + * flowing out of this buffer when we use the string + * functions, even if somehow the stuff we end up + * with here is garbage. + */ + client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN - 1] = + '\0'; + + if (ret != 0) { + pr_info("%s: CLIENT_STRING: copy_from_user failed\n", + __func__); + return -EIO; + } + + pr_info("%s: client debug array string has been received.\n", + __func__); + + if (!help_string_initialized) { + + /* Free the "we don't know yet" default string... */ + kfree(debug_help_string); + + /* build a proper debug help string */ + if (orangefs_prepare_debugfs_help_string(0)) { + gossip_err("%s: no debug help string \n", + __func__); + return -EIO; + } + + /* Replace the boilerplate boot-time debug-help file. */ + debugfs_remove(help_file_dentry); + + help_file_dentry = + debugfs_create_file( + ORANGEFS_KMOD_DEBUG_HELP_FILE, + 0444, + debug_dir, + debug_help_string, + &debug_help_fops); + + if (!help_file_dentry) { + gossip_err("%s: debugfs_create_file failed for" + " :%s:!\n", + __func__, + ORANGEFS_KMOD_DEBUG_HELP_FILE); + return -EIO; + } + } + + debug_mask_to_string(&client_debug_mask, 1); + + debugfs_remove(client_debug_dentry); + + orangefs_client_debug_init(); + + help_string_initialized++; + + return ret; + + case ORANGEFS_DEV_DEBUG: + ret = copy_from_user(&mask_info, + (void __user *)arg, + sizeof(mask_info)); + + if (ret != 0) + return -EIO; + + if (mask_info.mask_type == KERNEL_MASK) { + if ((mask_info.mask_value == 0) + && (kernel_mask_set_mod_init)) { + /* + * the kernel debug mask was set when the + * kernel module was loaded; don't override + * it if the client-core was started without + * a value for ORANGEFS_KMODMASK. + */ + return 0; + } + debug_mask_to_string(&mask_info.mask_value, + mask_info.mask_type); + gossip_debug_mask = mask_info.mask_value; + pr_info("%s: kernel debug mask has been modified to " + ":%s: :%llx:\n", + __func__, + kernel_debug_string, + (unsigned long long)gossip_debug_mask); + } else if (mask_info.mask_type == CLIENT_MASK) { + debug_mask_to_string(&mask_info.mask_value, + mask_info.mask_type); + pr_info("%s: client debug mask has been modified to" + ":%s: :%llx:\n", + __func__, + client_debug_string, + llu(mask_info.mask_value)); + } else { + gossip_lerr("Invalid mask type....\n"); + return -EINVAL; + } + + return ret; + + default: + return -ENOIOCTLCMD; + } + return -ENOIOCTLCMD; +} + +static long orangefs_devreq_ioctl(struct file *file, + unsigned int command, unsigned long arg) +{ + long ret; + + /* Check for properly constructed commands */ + ret = check_ioctl_command(command); + if (ret < 0) + return (int)ret; + + return (int)dispatch_ioctl_command(command, arg); +} + +#ifdef CONFIG_COMPAT /* CONFIG_COMPAT is in .config */ + +/* Compat structure for the ORANGEFS_DEV_MAP ioctl */ +struct ORANGEFS_dev_map_desc32 { + compat_uptr_t ptr; + __s32 total_size; + __s32 size; + __s32 count; +}; + +static unsigned long translate_dev_map26(unsigned long args, long *error) +{ + struct ORANGEFS_dev_map_desc32 __user *p32 = (void __user *)args; + /* + * Depending on the architecture, allocate some space on the + * user-call-stack based on our expected layout. + */ + struct ORANGEFS_dev_map_desc __user *p = + compat_alloc_user_space(sizeof(*p)); + compat_uptr_t addr; + + *error = 0; + /* get the ptr from the 32 bit user-space */ + if (get_user(addr, &p32->ptr)) + goto err; + /* try to put that into a 64-bit layout */ + if (put_user(compat_ptr(addr), &p->ptr)) + goto err; + /* copy the remaining fields */ + if (copy_in_user(&p->total_size, &p32->total_size, sizeof(__s32))) + goto err; + if (copy_in_user(&p->size, &p32->size, sizeof(__s32))) + goto err; + if (copy_in_user(&p->count, &p32->count, sizeof(__s32))) + goto err; + return (unsigned long)p; +err: + *error = -EFAULT; + return 0; +} + +/* + * 32 bit user-space apps' ioctl handlers when kernel modules + * is compiled as a 64 bit one + */ +static long orangefs_devreq_compat_ioctl(struct file *filp, unsigned int cmd, + unsigned long args) +{ + long ret; + unsigned long arg = args; + + /* Check for properly constructed commands */ + ret = check_ioctl_command(cmd); + if (ret < 0) + return ret; + if (cmd == ORANGEFS_DEV_MAP) { + /* + * convert the arguments to what we expect internally + * in kernel space + */ + arg = translate_dev_map26(args, &ret); + if (ret < 0) { + gossip_err("Could not translate dev map\n"); + return ret; + } + } + /* no other ioctl requires translation */ + return dispatch_ioctl_command(cmd, arg); +} + +#endif /* CONFIG_COMPAT is in .config */ + +/* the assigned character device major number */ +static int orangefs_dev_major; + +/* + * Initialize orangefs device specific state: + * Must be called at module load time only + */ +int orangefs_dev_init(void) +{ + /* register orangefs-req device */ + orangefs_dev_major = register_chrdev(0, + ORANGEFS_REQDEVICE_NAME, + &orangefs_devreq_file_operations); + if (orangefs_dev_major < 0) { + gossip_debug(GOSSIP_DEV_DEBUG, + "Failed to register /dev/%s (error %d)\n", + ORANGEFS_REQDEVICE_NAME, orangefs_dev_major); + return orangefs_dev_major; + } + + gossip_debug(GOSSIP_DEV_DEBUG, + "*** /dev/%s character device registered ***\n", + ORANGEFS_REQDEVICE_NAME); + gossip_debug(GOSSIP_DEV_DEBUG, "'mknod /dev/%s c %d 0'.\n", + ORANGEFS_REQDEVICE_NAME, orangefs_dev_major); + return 0; +} + +void orangefs_dev_cleanup(void) +{ + unregister_chrdev(orangefs_dev_major, ORANGEFS_REQDEVICE_NAME); + gossip_debug(GOSSIP_DEV_DEBUG, + "*** /dev/%s character device unregistered ***\n", + ORANGEFS_REQDEVICE_NAME); +} + +static unsigned int orangefs_devreq_poll(struct file *file, + struct poll_table_struct *poll_table) +{ + int poll_revent_mask = 0; + + poll_wait(file, &orangefs_request_list_waitq, poll_table); + + if (!list_empty(&orangefs_request_list)) + poll_revent_mask |= POLL_IN; + return poll_revent_mask; +} + +const struct file_operations orangefs_devreq_file_operations = { + .owner = THIS_MODULE, + .read = orangefs_devreq_read, + .write_iter = orangefs_devreq_write_iter, + .open = orangefs_devreq_open, + .release = orangefs_devreq_release, + .unlocked_ioctl = orangefs_devreq_ioctl, + +#ifdef CONFIG_COMPAT /* CONFIG_COMPAT is in .config */ + .compat_ioctl = orangefs_devreq_compat_ioctl, +#endif + .poll = orangefs_devreq_poll +}; diff --git a/fs/orangefs/dir.c b/fs/orangefs/dir.c new file mode 100644 index 000000000000..f30b6ecacdd1 --- /dev/null +++ b/fs/orangefs/dir.c @@ -0,0 +1,400 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +/* + * decode routine used by kmod to deal with the blob sent from + * userspace for readdirs. The blob contains zero or more of these + * sub-blobs: + * __u32 - represents length of the character string that follows. + * string - between 1 and ORANGEFS_NAME_MAX bytes long. + * padding - (if needed) to cause the __u32 plus the string to be + * eight byte aligned. + * khandle - sizeof(khandle) bytes. + */ +static long decode_dirents(char *ptr, size_t size, + struct orangefs_readdir_response_s *readdir) +{ + int i; + struct orangefs_readdir_response_s *rd = + (struct orangefs_readdir_response_s *) ptr; + char *buf = ptr; + int khandle_size = sizeof(struct orangefs_khandle); + size_t offset = offsetof(struct orangefs_readdir_response_s, + dirent_array); + /* 8 reflects eight byte alignment */ + int smallest_blob = khandle_size + 8; + __u32 len; + int aligned_len; + int sizeof_u32 = sizeof(__u32); + long ret; + + gossip_debug(GOSSIP_DIR_DEBUG, "%s: size:%zu:\n", __func__, size); + + /* size is = offset on empty dirs, > offset on non-empty dirs... */ + if (size < offset) { + gossip_err("%s: size:%zu: offset:%zu:\n", + __func__, + size, + offset); + ret = -EINVAL; + goto out; + } + + if ((size == offset) && (readdir->orangefs_dirent_outcount != 0)) { + gossip_err("%s: size:%zu: dirent_outcount:%d:\n", + __func__, + size, + readdir->orangefs_dirent_outcount); + ret = -EINVAL; + goto out; + } + + readdir->token = rd->token; + readdir->orangefs_dirent_outcount = rd->orangefs_dirent_outcount; + readdir->dirent_array = kcalloc(readdir->orangefs_dirent_outcount, + sizeof(*readdir->dirent_array), + GFP_KERNEL); + if (readdir->dirent_array == NULL) { + gossip_err("%s: kcalloc failed.\n", __func__); + ret = -ENOMEM; + goto out; + } + + buf += offset; + size -= offset; + + for (i = 0; i < readdir->orangefs_dirent_outcount; i++) { + if (size < smallest_blob) { + gossip_err("%s: size:%zu: smallest_blob:%d:\n", + __func__, + size, + smallest_blob); + ret = -EINVAL; + goto free; + } + + len = *(__u32 *)buf; + if ((len < 1) || (len > ORANGEFS_NAME_MAX)) { + gossip_err("%s: len:%d:\n", __func__, len); + ret = -EINVAL; + goto free; + } + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: size:%zu: len:%d:\n", + __func__, + size, + len); + + readdir->dirent_array[i].d_name = buf + sizeof_u32; + readdir->dirent_array[i].d_length = len; + + /* + * Calculate "aligned" length of this string and its + * associated __u32 descriptor. + */ + aligned_len = ((sizeof_u32 + len + 1) + 7) & ~7; + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: aligned_len:%d:\n", + __func__, + aligned_len); + + /* + * The end of the blob should coincide with the end + * of the last sub-blob. + */ + if (size < aligned_len + khandle_size) { + gossip_err("%s: ran off the end of the blob.\n", + __func__); + ret = -EINVAL; + goto free; + } + size -= aligned_len + khandle_size; + + buf += aligned_len; + + readdir->dirent_array[i].khandle = + *(struct orangefs_khandle *) buf; + buf += khandle_size; + } + ret = buf - ptr; + gossip_debug(GOSSIP_DIR_DEBUG, "%s: returning:%ld:\n", __func__, ret); + goto out; + +free: + kfree(readdir->dirent_array); + readdir->dirent_array = NULL; + +out: + return ret; +} + +/* + * Read directory entries from an instance of an open directory. + */ +static int orangefs_readdir(struct file *file, struct dir_context *ctx) +{ + int ret = 0; + int buffer_index; + /* + * ptoken supports Orangefs' distributed directory logic, added + * in 2.9.2. + */ + __u64 *ptoken = file->private_data; + __u64 pos = 0; + ino_t ino = 0; + struct dentry *dentry = file->f_path.dentry; + struct orangefs_kernel_op_s *new_op = NULL; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(dentry->d_inode); + int buffer_full = 0; + struct orangefs_readdir_response_s readdir_response; + void *dents_buf; + int i = 0; + int len = 0; + ino_t current_ino = 0; + char *current_entry = NULL; + long bytes_decoded; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: ctx->pos:%lld, ptoken = %llu\n", + __func__, + lld(ctx->pos), + llu(*ptoken)); + + pos = (__u64) ctx->pos; + + /* are we done? */ + if (pos == ORANGEFS_READDIR_END) { + gossip_debug(GOSSIP_DIR_DEBUG, + "Skipping to termination path\n"); + return 0; + } + + gossip_debug(GOSSIP_DIR_DEBUG, + "orangefs_readdir called on %s (pos=%llu)\n", + dentry->d_name.name, llu(pos)); + + memset(&readdir_response, 0, sizeof(readdir_response)); + + new_op = op_alloc(ORANGEFS_VFS_OP_READDIR); + if (!new_op) + return -ENOMEM; + + /* + * Only the indices are shared. No memory is actually shared, but the + * mechanism is used. + */ + new_op->uses_shared_memory = 1; + new_op->upcall.req.readdir.refn = orangefs_inode->refn; + new_op->upcall.req.readdir.max_dirent_count = + ORANGEFS_MAX_DIRENT_COUNT_READDIR; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: upcall.req.readdir.refn.khandle: %pU\n", + __func__, + &new_op->upcall.req.readdir.refn.khandle); + + new_op->upcall.req.readdir.token = *ptoken; + +get_new_buffer_index: + buffer_index = orangefs_readdir_index_get(); + if (buffer_index < 0) { + ret = buffer_index; + gossip_lerr("orangefs_readdir: orangefs_readdir_index_get() failure (%d)\n", + ret); + goto out_free_op; + } + new_op->upcall.req.readdir.buf_index = buffer_index; + + ret = service_operation(new_op, + "orangefs_readdir", + get_interruptible_flag(dentry->d_inode)); + + gossip_debug(GOSSIP_DIR_DEBUG, + "Readdir downcall status is %d. ret:%d\n", + new_op->downcall.status, + ret); + + orangefs_readdir_index_put(buffer_index); + + if (ret == -EAGAIN && op_state_purged(new_op)) { + /* Client-core indices are invalid after it restarted. */ + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: Getting new buffer_index for retry of readdir..\n", + __func__); + goto get_new_buffer_index; + } + + if (ret == -EIO && op_state_purged(new_op)) { + gossip_err("%s: Client is down. Aborting readdir call.\n", + __func__); + goto out_slot; + } + + if (ret < 0 || new_op->downcall.status != 0) { + gossip_debug(GOSSIP_DIR_DEBUG, + "Readdir request failed. Status:%d\n", + new_op->downcall.status); + if (ret >= 0) + ret = new_op->downcall.status; + goto out_slot; + } + + dents_buf = new_op->downcall.trailer_buf; + if (dents_buf == NULL) { + gossip_err("Invalid NULL buffer in readdir response\n"); + ret = -ENOMEM; + goto out_slot; + } + + bytes_decoded = decode_dirents(dents_buf, new_op->downcall.trailer_size, + &readdir_response); + if (bytes_decoded < 0) { + ret = bytes_decoded; + gossip_err("Could not decode readdir from buffer %d\n", ret); + goto out_vfree; + } + + if (bytes_decoded != new_op->downcall.trailer_size) { + gossip_err("orangefs_readdir: # bytes decoded (%ld) " + "!= trailer size (%ld)\n", + bytes_decoded, + (long)new_op->downcall.trailer_size); + ret = -EINVAL; + goto out_destroy_handle; + } + + /* + * orangefs doesn't actually store dot and dot-dot, but + * we need to have them represented. + */ + if (pos == 0) { + ino = get_ino_from_khandle(dentry->d_inode); + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: calling dir_emit of \".\" with pos = %llu\n", + __func__, + llu(pos)); + ret = dir_emit(ctx, ".", 1, ino, DT_DIR); + pos += 1; + } + + if (pos == 1) { + ino = get_parent_ino_from_dentry(dentry); + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: calling dir_emit of \"..\" with pos = %llu\n", + __func__, + llu(pos)); + ret = dir_emit(ctx, "..", 2, ino, DT_DIR); + pos += 1; + } + + /* + * we stored ORANGEFS_ITERATE_NEXT in ctx->pos last time around + * to prevent "finding" dot and dot-dot on any iteration + * other than the first. + */ + if (ctx->pos == ORANGEFS_ITERATE_NEXT) + ctx->pos = 0; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: dirent_outcount:%d:\n", + __func__, + readdir_response.orangefs_dirent_outcount); + for (i = ctx->pos; + i < readdir_response.orangefs_dirent_outcount; + i++) { + len = readdir_response.dirent_array[i].d_length; + current_entry = readdir_response.dirent_array[i].d_name; + current_ino = orangefs_khandle_to_ino( + &readdir_response.dirent_array[i].khandle); + + gossip_debug(GOSSIP_DIR_DEBUG, + "calling dir_emit for %s with len %d" + ", ctx->pos %ld\n", + current_entry, + len, + (unsigned long)ctx->pos); + /* + * type is unknown. We don't return object type + * in the dirent_array. This leaves getdents + * clueless about type. + */ + ret = + dir_emit(ctx, current_entry, len, current_ino, DT_UNKNOWN); + if (!ret) + break; + ctx->pos++; + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: ctx->pos:%lld\n", + __func__, + lld(ctx->pos)); + + } + + /* + * we ran all the way through the last batch, set up for + * getting another batch... + */ + if (ret) { + *ptoken = readdir_response.token; + ctx->pos = ORANGEFS_ITERATE_NEXT; + } + + /* + * Did we hit the end of the directory? + */ + if (readdir_response.token == ORANGEFS_READDIR_END && + !buffer_full) { + gossip_debug(GOSSIP_DIR_DEBUG, + "End of dir detected; setting ctx->pos to ORANGEFS_READDIR_END.\n"); + ctx->pos = ORANGEFS_READDIR_END; + } + +out_destroy_handle: + /* kfree(NULL) is safe */ + kfree(readdir_response.dirent_array); +out_vfree: + gossip_debug(GOSSIP_DIR_DEBUG, "vfree %p\n", dents_buf); + vfree(dents_buf); +out_slot: + orangefs_readdir_index_put(buffer_index); +out_free_op: + op_release(new_op); + gossip_debug(GOSSIP_DIR_DEBUG, "orangefs_readdir returning %d\n", ret); + return ret; +} + +static int orangefs_dir_open(struct inode *inode, struct file *file) +{ + __u64 *ptoken; + + file->private_data = kmalloc(sizeof(__u64), GFP_KERNEL); + if (!file->private_data) + return -ENOMEM; + + ptoken = file->private_data; + *ptoken = ORANGEFS_READDIR_START; + return 0; +} + +static int orangefs_dir_release(struct inode *inode, struct file *file) +{ + orangefs_flush_inode(inode); + kfree(file->private_data); + return 0; +} + +/** ORANGEFS implementation of VFS directory operations */ +const struct file_operations orangefs_dir_operations = { + .read = generic_read_dir, + .iterate = orangefs_readdir, + .open = orangefs_dir_open, + .release = orangefs_dir_release, +}; diff --git a/fs/orangefs/downcall.h b/fs/orangefs/downcall.h new file mode 100644 index 000000000000..66b99210f1f9 --- /dev/null +++ b/fs/orangefs/downcall.h @@ -0,0 +1,133 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Definitions of downcalls used in Linux kernel module. + */ + +#ifndef __DOWNCALL_H +#define __DOWNCALL_H + +/* + * Sanitized the device-client core interaction + * for clean 32-64 bit usage + */ +struct orangefs_io_response { + __s64 amt_complete; +}; + +struct orangefs_lookup_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_create_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_symlink_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_getattr_response { + struct ORANGEFS_sys_attr_s attributes; + char link_target[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_mkdir_response { + struct orangefs_object_kref refn; +}; + +/* + * duplication of some system interface structures so that I don't have + * to allocate extra memory + */ +struct orangefs_dirent { + char *d_name; + int d_length; + struct orangefs_khandle khandle; +}; + +struct orangefs_statfs_response { + __s64 block_size; + __s64 blocks_total; + __s64 blocks_avail; + __s64 files_total; + __s64 files_avail; +}; + +struct orangefs_fs_mount_response { + __s32 fs_id; + __s32 id; + struct orangefs_khandle root_khandle; +}; + +/* the getxattr response is the attribute value */ +struct orangefs_getxattr_response { + __s32 val_sz; + __s32 __pad1; + char val[ORANGEFS_MAX_XATTR_VALUELEN]; +}; + +/* the listxattr response is an array of attribute names */ +struct orangefs_listxattr_response { + __s32 returned_count; + __s32 __pad1; + __u64 token; + char key[ORANGEFS_MAX_XATTR_LISTLEN * ORANGEFS_MAX_XATTR_NAMELEN]; + __s32 keylen; + __s32 __pad2; + __s32 lengths[ORANGEFS_MAX_XATTR_LISTLEN]; +}; + +struct orangefs_param_response { + __s64 value; +}; + +#define PERF_COUNT_BUF_SIZE 4096 +struct orangefs_perf_count_response { + char buffer[PERF_COUNT_BUF_SIZE]; +}; + +#define FS_KEY_BUF_SIZE 4096 +struct orangefs_fs_key_response { + __s32 fs_keylen; + __s32 __pad1; + char fs_key[FS_KEY_BUF_SIZE]; +}; + +struct orangefs_downcall_s { + __s32 type; + __s32 status; + /* currently trailer is used only by readdir */ + __s64 trailer_size; + char *trailer_buf; + + union { + struct orangefs_io_response io; + struct orangefs_lookup_response lookup; + struct orangefs_create_response create; + struct orangefs_symlink_response sym; + struct orangefs_getattr_response getattr; + struct orangefs_mkdir_response mkdir; + struct orangefs_statfs_response statfs; + struct orangefs_fs_mount_response fs_mount; + struct orangefs_getxattr_response getxattr; + struct orangefs_listxattr_response listxattr; + struct orangefs_param_response param; + struct orangefs_perf_count_response perf_count; + struct orangefs_fs_key_response fs_key; + } resp; +}; + +struct orangefs_readdir_response_s { + __u64 token; + __u64 directory_version; + __u32 __pad2; + __u32 orangefs_dirent_outcount; + struct orangefs_dirent *dirent_array; +}; + +#endif /* __DOWNCALL_H */ diff --git a/fs/orangefs/file.c b/fs/orangefs/file.c new file mode 100644 index 000000000000..ae92795ed965 --- /dev/null +++ b/fs/orangefs/file.c @@ -0,0 +1,717 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS file operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/fs.h> +#include <linux/pagemap.h> + +/* + * Copy to client-core's address space from the buffers specified + * by the iovec upto total_size bytes. + * NOTE: the iovector can either contain addresses which + * can futher be kernel-space or user-space addresses. + * or it can pointers to struct page's + */ +static int precopy_buffers(int buffer_index, + struct iov_iter *iter, + size_t total_size) +{ + int ret = 0; + /* + * copy data from application/kernel by pulling it out + * of the iovec. + */ + + + if (total_size) { + ret = orangefs_bufmap_copy_from_iovec(iter, + buffer_index, + total_size); + if (ret < 0) + gossip_err("%s: Failed to copy-in buffers. Please make sure that the pvfs2-client is running. %ld\n", + __func__, + (long)ret); + } + + if (ret < 0) + gossip_err("%s: Failed to copy-in buffers. Please make sure that the pvfs2-client is running. %ld\n", + __func__, + (long)ret); + return ret; +} + +/* + * Copy from client-core's address space to the buffers specified + * by the iovec upto total_size bytes. + * NOTE: the iovector can either contain addresses which + * can futher be kernel-space or user-space addresses. + * or it can pointers to struct page's + */ +static int postcopy_buffers(int buffer_index, + struct iov_iter *iter, + size_t total_size) +{ + int ret = 0; + /* + * copy data to application/kernel by pushing it out to + * the iovec. NOTE; target buffers can be addresses or + * struct page pointers. + */ + if (total_size) { + ret = orangefs_bufmap_copy_to_iovec(iter, + buffer_index, + total_size); + if (ret < 0) + gossip_err("%s: Failed to copy-out buffers. Please make sure that the pvfs2-client is running (%ld)\n", + __func__, + (long)ret); + } + return ret; +} + +/* + * Post and wait for the I/O upcall to finish + */ +static ssize_t wait_for_direct_io(enum ORANGEFS_io_type type, struct inode *inode, + loff_t *offset, struct iov_iter *iter, + size_t total_size, loff_t readahead_size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_khandle *handle = &orangefs_inode->refn.khandle; + struct orangefs_kernel_op_s *new_op = NULL; + struct iov_iter saved = *iter; + int buffer_index = -1; + ssize_t ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_FILE_IO); + if (!new_op) + return -ENOMEM; + + /* synchronous I/O */ + new_op->upcall.req.io.readahead_size = readahead_size; + new_op->upcall.req.io.io_type = type; + new_op->upcall.req.io.refn = orangefs_inode->refn; + +populate_shared_memory: + /* get a shared buffer index */ + buffer_index = orangefs_bufmap_get(); + if (buffer_index < 0) { + ret = buffer_index; + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: orangefs_bufmap_get failure (%zd)\n", + __func__, ret); + goto out; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): GET op %p -> buffer_index %d\n", + __func__, + handle, + new_op, + buffer_index); + + new_op->uses_shared_memory = 1; + new_op->upcall.req.io.buf_index = buffer_index; + new_op->upcall.req.io.count = total_size; + new_op->upcall.req.io.offset = *offset; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): offset: %llu total_size: %zd\n", + __func__, + handle, + llu(*offset), + total_size); + /* + * Stage 1: copy the buffers into client-core's address space + * precopy_buffers only pertains to writes. + */ + if (type == ORANGEFS_IO_WRITE) { + ret = precopy_buffers(buffer_index, + iter, + total_size); + if (ret < 0) + goto out; + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Calling post_io_request with tag (%llu)\n", + __func__, + handle, + llu(new_op->tag)); + + /* Stage 2: Service the I/O operation */ + ret = service_operation(new_op, + type == ORANGEFS_IO_WRITE ? + "file_write" : + "file_read", + get_interruptible_flag(inode)); + + /* + * If service_operation() returns -EAGAIN #and# the operation was + * purged from orangefs_request_list or htable_ops_in_progress, then + * we know that the client was restarted, causing the shared memory + * area to be wiped clean. To restart a write operation in this + * case, we must re-copy the data from the user's iovec to a NEW + * shared memory location. To restart a read operation, we must get + * a new shared memory location. + */ + if (ret == -EAGAIN && op_state_purged(new_op)) { + orangefs_bufmap_put(buffer_index); + buffer_index = -1; + if (type == ORANGEFS_IO_WRITE) + *iter = saved; + gossip_debug(GOSSIP_FILE_DEBUG, + "%s:going to repopulate_shared_memory.\n", + __func__); + goto populate_shared_memory; + } + + if (ret < 0) { + if (ret == -EINTR) { + /* + * We can't return EINTR if any data was written, + * it's not POSIX. It is minimally acceptable + * to give a partial write, the way NFS does. + * + * It would be optimal to return all or nothing, + * but if a userspace write is bigger than + * an IO buffer, and the interrupt occurs + * between buffer writes, that would not be + * possible. + */ + switch (new_op->op_state - OP_VFS_STATE_GIVEN_UP) { + /* + * If the op was waiting when the interrupt + * occurred, then the client-core did not + * trigger the write. + */ + case OP_VFS_STATE_WAITING: + if (*offset == 0) + ret = -EINTR; + else + ret = 0; + break; + /* + * If the op was in progress when the interrupt + * occurred, then the client-core was able to + * trigger the write. + */ + case OP_VFS_STATE_INPROGR: + ret = total_size; + break; + default: + gossip_err("%s: unexpected op state :%d:.\n", + __func__, + new_op->op_state); + ret = 0; + break; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: got EINTR, state:%d: %p\n", + __func__, + new_op->op_state, + new_op); + } else { + gossip_err("%s: error in %s handle %pU, returning %zd\n", + __func__, + type == ORANGEFS_IO_READ ? + "read from" : "write to", + handle, ret); + } + if (orangefs_cancel_op_in_progress(new_op)) + return ret; + + goto out; + } + + /* + * Stage 3: Post copy buffers from client-core's address space + * postcopy_buffers only pertains to reads. + */ + if (type == ORANGEFS_IO_READ) { + ret = postcopy_buffers(buffer_index, + iter, + new_op->downcall.resp.io.amt_complete); + if (ret < 0) + goto out; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Amount %s, returned by the sys-io call:%d\n", + __func__, + handle, + type == ORANGEFS_IO_READ ? "read" : "written", + (int)new_op->downcall.resp.io.amt_complete); + + ret = new_op->downcall.resp.io.amt_complete; + +out: + if (buffer_index >= 0) { + orangefs_bufmap_put(buffer_index); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): PUT buffer_index %d\n", + __func__, handle, buffer_index); + buffer_index = -1; + } + op_release(new_op); + return ret; +} + +/* + * Common entry point for read/write/readv/writev + * This function will dispatch it to either the direct I/O + * or buffered I/O path depending on the mount options and/or + * augmented/extended metadata attached to the file. + * Note: File extended attributes override any mount options. + */ +static ssize_t do_readv_writev(enum ORANGEFS_io_type type, struct file *file, + loff_t *offset, struct iov_iter *iter) +{ + struct inode *inode = file->f_mapping->host; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_khandle *handle = &orangefs_inode->refn.khandle; + size_t count = iov_iter_count(iter); + ssize_t total_count = 0; + ssize_t ret = -EINVAL; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s-BEGIN(%pU): count(%d) after estimate_max_iovecs.\n", + __func__, + handle, + (int)count); + + if (type == ORANGEFS_IO_WRITE) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): proceeding with offset : %llu, " + "size %d\n", + __func__, + handle, + llu(*offset), + (int)count); + } + + if (count == 0) { + ret = 0; + goto out; + } + + while (iov_iter_count(iter)) { + size_t each_count = iov_iter_count(iter); + size_t amt_complete; + + /* how much to transfer in this loop iteration */ + if (each_count > orangefs_bufmap_size_query()) + each_count = orangefs_bufmap_size_query(); + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): size of each_count(%d)\n", + __func__, + handle, + (int)each_count); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): BEFORE wait_for_io: offset is %d\n", + __func__, + handle, + (int)*offset); + + ret = wait_for_direct_io(type, inode, offset, iter, + each_count, 0); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): return from wait_for_io:%d\n", + __func__, + handle, + (int)ret); + + if (ret < 0) + goto out; + + *offset += ret; + total_count += ret; + amt_complete = ret; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): AFTER wait_for_io: offset is %d\n", + __func__, + handle, + (int)*offset); + + /* + * if we got a short I/O operations, + * fall out and return what we got so far + */ + if (amt_complete < each_count) + break; + } /*end while */ + +out: + if (total_count > 0) + ret = total_count; + if (ret > 0) { + if (type == ORANGEFS_IO_READ) { + file_accessed(file); + } else { + SetMtimeFlag(orangefs_inode); + inode->i_mtime = CURRENT_TIME; + mark_inode_dirty_sync(inode); + } + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Value(%d) returned.\n", + __func__, + handle, + (int)ret); + + return ret; +} + +/* + * Read data from a specified offset in a file (referenced by inode). + * Data may be placed either in a user or kernel buffer. + */ +ssize_t orangefs_inode_read(struct inode *inode, + struct iov_iter *iter, + loff_t *offset, + loff_t readahead_size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + size_t count = iov_iter_count(iter); + size_t bufmap_size; + ssize_t ret = -EINVAL; + + g_orangefs_stats.reads++; + + bufmap_size = orangefs_bufmap_size_query(); + if (count > bufmap_size) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: count is too large (%zd/%zd)!\n", + __func__, count, bufmap_size); + return -EINVAL; + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU) %zd@%llu\n", + __func__, + &orangefs_inode->refn.khandle, + count, + llu(*offset)); + + ret = wait_for_direct_io(ORANGEFS_IO_READ, inode, offset, iter, + count, readahead_size); + if (ret > 0) + *offset += ret; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Value(%zd) returned.\n", + __func__, + &orangefs_inode->refn.khandle, + ret); + + return ret; +} + +static ssize_t orangefs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) +{ + struct file *file = iocb->ki_filp; + loff_t pos = *(&iocb->ki_pos); + ssize_t rc = 0; + + BUG_ON(iocb->private); + + gossip_debug(GOSSIP_FILE_DEBUG, "orangefs_file_read_iter\n"); + + g_orangefs_stats.reads++; + + rc = do_readv_writev(ORANGEFS_IO_READ, file, &pos, iter); + iocb->ki_pos = pos; + + return rc; +} + +static ssize_t orangefs_file_write_iter(struct kiocb *iocb, struct iov_iter *iter) +{ + struct file *file = iocb->ki_filp; + loff_t pos; + ssize_t rc; + + BUG_ON(iocb->private); + + gossip_debug(GOSSIP_FILE_DEBUG, "orangefs_file_write_iter\n"); + + mutex_lock(&file->f_mapping->host->i_mutex); + + /* Make sure generic_write_checks sees an up to date inode size. */ + if (file->f_flags & O_APPEND) { + rc = orangefs_inode_getattr(file->f_mapping->host, 0, 1); + if (rc == -ESTALE) + rc = -EIO; + if (rc) { + gossip_err("%s: orangefs_inode_getattr failed, " + "rc:%zd:.\n", __func__, rc); + goto out; + } + } + + if (file->f_pos > i_size_read(file->f_mapping->host)) + orangefs_i_size_write(file->f_mapping->host, file->f_pos); + + rc = generic_write_checks(iocb, iter); + + if (rc <= 0) { + gossip_err("%s: generic_write_checks failed, rc:%zd:.\n", + __func__, rc); + goto out; + } + + /* + * if we are appending, generic_write_checks would have updated + * pos to the end of the file, so we will wait till now to set + * pos... + */ + pos = *(&iocb->ki_pos); + + rc = do_readv_writev(ORANGEFS_IO_WRITE, + file, + &pos, + iter); + if (rc < 0) { + gossip_err("%s: do_readv_writev failed, rc:%zd:.\n", + __func__, rc); + goto out; + } + + iocb->ki_pos = pos; + g_orangefs_stats.writes++; + +out: + + mutex_unlock(&file->f_mapping->host->i_mutex); + return rc; +} + +/* + * Perform a miscellaneous operation on a file. + */ +static long orangefs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + int ret = -ENOTTY; + __u64 val = 0; + unsigned long uval; + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: called with cmd %d\n", + cmd); + + /* + * we understand some general ioctls on files, such as the immutable + * and append flags + */ + if (cmd == FS_IOC_GETFLAGS) { + val = 0; + ret = orangefs_inode_getxattr(file_inode(file), + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + "user.pvfs2.meta_hint", + &val, sizeof(val)); + if (ret < 0 && ret != -ENODATA) + return ret; + else if (ret == -ENODATA) + val = 0; + uval = val; + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: FS_IOC_GETFLAGS: %llu\n", + (unsigned long long)uval); + return put_user(uval, (int __user *)arg); + } else if (cmd == FS_IOC_SETFLAGS) { + ret = 0; + if (get_user(uval, (int __user *)arg)) + return -EFAULT; + /* + * ORANGEFS_MIRROR_FL is set internally when the mirroring mode + * is turned on for a file. The user is not allowed to turn + * on this bit, but the bit is present if the user first gets + * the flags and then updates the flags with some new + * settings. So, we ignore it in the following edit. bligon. + */ + if ((uval & ~ORANGEFS_MIRROR_FL) & + (~(FS_IMMUTABLE_FL | FS_APPEND_FL | FS_NOATIME_FL))) { + gossip_err("orangefs_ioctl: the FS_IOC_SETFLAGS only supports setting one of FS_IMMUTABLE_FL|FS_APPEND_FL|FS_NOATIME_FL\n"); + return -EINVAL; + } + val = uval; + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: FS_IOC_SETFLAGS: %llu\n", + (unsigned long long)val); + ret = orangefs_inode_setxattr(file_inode(file), + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + "user.pvfs2.meta_hint", + &val, sizeof(val), 0); + } + + return ret; +} + +/* + * Memory map a region of a file. + */ +static int orangefs_file_mmap(struct file *file, struct vm_area_struct *vma) +{ + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_mmap: called on %s\n", + (file ? + (char *)file->f_path.dentry->d_name.name : + (char *)"Unknown")); + + /* set the sequential readahead hint */ + vma->vm_flags |= VM_SEQ_READ; + vma->vm_flags &= ~VM_RAND_READ; + + /* Use readonly mmap since we cannot support writable maps. */ + return generic_file_readonly_mmap(file, vma); +} + +#define mapping_nrpages(idata) ((idata)->nrpages) + +/* + * Called to notify the module that there are no more references to + * this file (i.e. no processes have it open). + * + * \note Not called when each file is closed. + */ +static int orangefs_file_release(struct inode *inode, struct file *file) +{ + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_release: called on %s\n", + file->f_path.dentry->d_name.name); + + orangefs_flush_inode(inode); + + /* + * remove all associated inode pages from the page cache and mmap + * readahead cache (if any); this forces an expensive refresh of + * data for the next caller of mmap (or 'get_block' accesses) + */ + if (file->f_path.dentry->d_inode && + file->f_path.dentry->d_inode->i_mapping && + mapping_nrpages(&file->f_path.dentry->d_inode->i_data)) + truncate_inode_pages(file->f_path.dentry->d_inode->i_mapping, + 0); + return 0; +} + +/* + * Push all data for a specific file onto permanent storage. + */ +static int orangefs_fsync(struct file *file, + loff_t start, + loff_t end, + int datasync) +{ + int ret = -EINVAL; + struct orangefs_inode_s *orangefs_inode = + ORANGEFS_I(file->f_path.dentry->d_inode); + struct orangefs_kernel_op_s *new_op = NULL; + + /* required call */ + filemap_write_and_wait_range(file->f_mapping, start, end); + + new_op = op_alloc(ORANGEFS_VFS_OP_FSYNC); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.fsync.refn = orangefs_inode->refn; + + ret = service_operation(new_op, + "orangefs_fsync", + get_interruptible_flag(file->f_path.dentry->d_inode)); + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_fsync got return value of %d\n", + ret); + + op_release(new_op); + + orangefs_flush_inode(file->f_path.dentry->d_inode); + return ret; +} + +/* + * Change the file pointer position for an instance of an open file. + * + * \note If .llseek is overriden, we must acquire lock as described in + * Documentation/filesystems/Locking. + * + * Future upgrade could support SEEK_DATA and SEEK_HOLE but would + * require much changes to the FS + */ +static loff_t orangefs_file_llseek(struct file *file, loff_t offset, int origin) +{ + int ret = -EINVAL; + struct inode *inode = file_inode(file); + + if (origin == SEEK_END) { + /* + * revalidate the inode's file size. + * NOTE: We are only interested in file size here, + * so we set mask accordingly. + */ + ret = orangefs_inode_getattr(file->f_mapping->host, 0, 1); + if (ret == -ESTALE) + ret = -EIO; + if (ret) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s:%s:%d calling make bad inode\n", + __FILE__, + __func__, + __LINE__); + return ret; + } + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_llseek: offset is %ld | origin is %d" + " | inode size is %lu\n", + (long)offset, + origin, + (unsigned long)i_size_read(inode)); + + return generic_file_llseek(file, offset, origin); +} + +/* + * Support local locks (locks that only this kernel knows about) + * if Orangefs was mounted -o local_lock. + */ +static int orangefs_lock(struct file *filp, int cmd, struct file_lock *fl) +{ + int rc = -EINVAL; + + if (ORANGEFS_SB(filp->f_inode->i_sb)->flags & ORANGEFS_OPT_LOCAL_LOCK) { + if (cmd == F_GETLK) { + rc = 0; + posix_test_lock(filp, fl); + } else { + rc = posix_lock_file(filp, fl, NULL); + } + } + + return rc; +} + +/** ORANGEFS implementation of VFS file operations */ +const struct file_operations orangefs_file_operations = { + .llseek = orangefs_file_llseek, + .read_iter = orangefs_file_read_iter, + .write_iter = orangefs_file_write_iter, + .lock = orangefs_lock, + .unlocked_ioctl = orangefs_ioctl, + .mmap = orangefs_file_mmap, + .open = generic_file_open, + .release = orangefs_file_release, + .fsync = orangefs_fsync, +}; diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c new file mode 100644 index 000000000000..2382e267b49e --- /dev/null +++ b/fs/orangefs/inode.c @@ -0,0 +1,475 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS inode operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +static int read_one_page(struct page *page) +{ + int ret; + int max_block; + ssize_t bytes_read = 0; + struct inode *inode = page->mapping->host; + const __u32 blocksize = PAGE_CACHE_SIZE; /* inode->i_blksize */ + const __u32 blockbits = PAGE_CACHE_SHIFT; /* inode->i_blkbits */ + struct iov_iter to; + struct bio_vec bv = {.bv_page = page, .bv_len = PAGE_SIZE}; + + iov_iter_bvec(&to, ITER_BVEC | READ, &bv, 1, PAGE_SIZE); + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_readpage called with page %p\n", + page); + + max_block = ((inode->i_size / blocksize) + 1); + + if (page->index < max_block) { + loff_t blockptr_offset = (((loff_t) page->index) << blockbits); + + bytes_read = orangefs_inode_read(inode, + &to, + &blockptr_offset, + inode->i_size); + } + /* this will only zero remaining unread portions of the page data */ + iov_iter_zero(~0U, &to); + /* takes care of potential aliasing */ + flush_dcache_page(page); + if (bytes_read < 0) { + ret = bytes_read; + SetPageError(page); + } else { + SetPageUptodate(page); + if (PageError(page)) + ClearPageError(page); + ret = 0; + } + /* unlock the page after the ->readpage() routine completes */ + unlock_page(page); + return ret; +} + +static int orangefs_readpage(struct file *file, struct page *page) +{ + return read_one_page(page); +} + +static int orangefs_readpages(struct file *file, + struct address_space *mapping, + struct list_head *pages, + unsigned nr_pages) +{ + int page_idx; + int ret; + + gossip_debug(GOSSIP_INODE_DEBUG, "orangefs_readpages called\n"); + + for (page_idx = 0; page_idx < nr_pages; page_idx++) { + struct page *page; + + page = list_entry(pages->prev, struct page, lru); + list_del(&page->lru); + if (!add_to_page_cache(page, + mapping, + page->index, + GFP_KERNEL)) { + ret = read_one_page(page); + gossip_debug(GOSSIP_INODE_DEBUG, + "failure adding page to cache, read_one_page returned: %d\n", + ret); + } else { + page_cache_release(page); + } + } + BUG_ON(!list_empty(pages)); + return 0; +} + +static void orangefs_invalidatepage(struct page *page, + unsigned int offset, + unsigned int length) +{ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_invalidatepage called on page %p " + "(offset is %u)\n", + page, + offset); + + ClearPageUptodate(page); + ClearPageMappedToDisk(page); + return; + +} + +static int orangefs_releasepage(struct page *page, gfp_t foo) +{ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_releasepage called on page %p\n", + page); + return 0; +} + +/* + * Having a direct_IO entry point in the address_space_operations + * struct causes the kernel to allows us to use O_DIRECT on + * open. Nothing will ever call this thing, but in the future we + * will need to be able to use O_DIRECT on open in order to support + * AIO. Modeled after NFS, they do this too. + */ +/* + * static ssize_t orangefs_direct_IO(int rw, + * struct kiocb *iocb, + * struct iov_iter *iter, + * loff_t offset) + *{ + * gossip_debug(GOSSIP_INODE_DEBUG, + * "orangefs_direct_IO: %s\n", + * iocb->ki_filp->f_path.dentry->d_name.name); + * + * return -EINVAL; + *} + */ + +struct backing_dev_info orangefs_backing_dev_info = { + .name = "orangefs", + .ra_pages = 0, + .capabilities = BDI_CAP_NO_ACCT_DIRTY | BDI_CAP_NO_WRITEBACK, +}; + +/** ORANGEFS2 implementation of address space operations */ +const struct address_space_operations orangefs_address_operations = { + .readpage = orangefs_readpage, + .readpages = orangefs_readpages, + .invalidatepage = orangefs_invalidatepage, + .releasepage = orangefs_releasepage, +/* .direct_IO = orangefs_direct_IO */ +}; + +static int orangefs_setattr_size(struct inode *inode, struct iattr *iattr) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + loff_t orig_size; + int ret = -EINVAL; + + gossip_debug(GOSSIP_INODE_DEBUG, + "%s: %pU: Handle is %pU | fs_id %d | size is %llu\n", + __func__, + get_khandle_from_ino(inode), + &orangefs_inode->refn.khandle, + orangefs_inode->refn.fs_id, + iattr->ia_size); + + /* Ensure that we have a up to date size, so we know if it changed. */ + ret = orangefs_inode_getattr(inode, 0, 1); + if (ret == -ESTALE) + ret = -EIO; + if (ret) { + gossip_err("%s: orangefs_inode_getattr failed, ret:%d:.\n", + __func__, ret); + return ret; + } + orig_size = i_size_read(inode); + + truncate_setsize(inode, iattr->ia_size); + + new_op = op_alloc(ORANGEFS_VFS_OP_TRUNCATE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.truncate.refn = orangefs_inode->refn; + new_op->upcall.req.truncate.size = (__s64) iattr->ia_size; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + + /* + * the truncate has no downcall members to retrieve, but + * the status value tells us if it went through ok or not + */ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs: orangefs_truncate got return value of %d\n", + ret); + + op_release(new_op); + + if (ret != 0) + return ret; + + /* + * Only change the c/mtime if we are changing the size or we are + * explicitly asked to change it. This handles the semantic difference + * between truncate() and ftruncate() as implemented in the VFS. + * + * The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a + * special case where we need to update the times despite not having + * these flags set. For all other operations the VFS set these flags + * explicitly if it wants a timestamp update. + */ + if (orig_size != i_size_read(inode) && + !(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) { + iattr->ia_ctime = iattr->ia_mtime = + current_fs_time(inode->i_sb); + iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME; + } + + return ret; +} + +/* + * Change attributes of an object referenced by dentry. + */ +int orangefs_setattr(struct dentry *dentry, struct iattr *iattr) +{ + int ret = -EINVAL; + struct inode *inode = dentry->d_inode; + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_setattr: called on %s\n", + dentry->d_name.name); + + ret = inode_change_ok(inode, iattr); + if (ret) + goto out; + + if ((iattr->ia_valid & ATTR_SIZE) && + iattr->ia_size != i_size_read(inode)) { + ret = orangefs_setattr_size(inode, iattr); + if (ret) + goto out; + } + + setattr_copy(inode, iattr); + mark_inode_dirty(inode); + + ret = orangefs_inode_setattr(inode, iattr); + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_setattr: inode_setattr returned %d\n", + ret); + + if (!ret && (iattr->ia_valid & ATTR_MODE)) + /* change mod on a file that has ACLs */ + ret = posix_acl_chmod(inode, inode->i_mode); + +out: + gossip_debug(GOSSIP_INODE_DEBUG, "orangefs_setattr: returning %d\n", ret); + return ret; +} + +/* + * Obtain attributes of an object given a dentry + */ +int orangefs_getattr(struct vfsmount *mnt, + struct dentry *dentry, + struct kstat *kstat) +{ + int ret = -ENOENT; + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *orangefs_inode = NULL; + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_getattr: called on %s\n", + dentry->d_name.name); + + ret = orangefs_inode_getattr(inode, 0, 1); + if (ret == 0) { + generic_fillattr(inode, kstat); + + /* override block size reported to stat */ + orangefs_inode = ORANGEFS_I(inode); + kstat->blksize = orangefs_inode->blksize; + } + return ret; +} + +int orangefs_permission(struct inode *inode, int mask) +{ + int ret; + + if (mask & MAY_NOT_BLOCK) + return -ECHILD; + + gossip_debug(GOSSIP_INODE_DEBUG, "%s: refreshing\n", __func__); + + /* Make sure the permission (and other common attrs) are up to date. */ + ret = orangefs_inode_getattr(inode, 0, 0); + if (ret < 0) + return ret; + + return generic_permission(inode, mask); +} + +/* ORANGEDS2 implementation of VFS inode operations for files */ +struct inode_operations orangefs_file_inode_operations = { + .get_acl = orangefs_get_acl, + .set_acl = orangefs_set_acl, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .listxattr = orangefs_listxattr, + .removexattr = generic_removexattr, + .permission = orangefs_permission, +}; + +static int orangefs_init_iops(struct inode *inode) +{ + inode->i_mapping->a_ops = &orangefs_address_operations; + + switch (inode->i_mode & S_IFMT) { + case S_IFREG: + inode->i_op = &orangefs_file_inode_operations; + inode->i_fop = &orangefs_file_operations; + inode->i_blkbits = PAGE_CACHE_SHIFT; + break; + case S_IFLNK: + inode->i_op = &orangefs_symlink_inode_operations; + break; + case S_IFDIR: + inode->i_op = &orangefs_dir_inode_operations; + inode->i_fop = &orangefs_dir_operations; + break; + default: + gossip_debug(GOSSIP_INODE_DEBUG, + "%s: unsupported mode\n", + __func__); + return -EINVAL; + } + + return 0; +} + +/* + * Given a ORANGEFS object identifier (fsid, handle), convert it into a ino_t type + * that will be used as a hash-index from where the handle will + * be searched for in the VFS hash table of inodes. + */ +static inline ino_t orangefs_handle_hash(struct orangefs_object_kref *ref) +{ + if (!ref) + return 0; + return orangefs_khandle_to_ino(&(ref->khandle)); +} + +/* + * Called to set up an inode from iget5_locked. + */ +static int orangefs_set_inode(struct inode *inode, void *data) +{ + struct orangefs_object_kref *ref = (struct orangefs_object_kref *) data; + ORANGEFS_I(inode)->refn.fs_id = ref->fs_id; + ORANGEFS_I(inode)->refn.khandle = ref->khandle; + return 0; +} + +/* + * Called to determine if handles match. + */ +static int orangefs_test_inode(struct inode *inode, void *data) +{ + struct orangefs_object_kref *ref = (struct orangefs_object_kref *) data; + struct orangefs_inode_s *orangefs_inode = NULL; + + orangefs_inode = ORANGEFS_I(inode); + return (!ORANGEFS_khandle_cmp(&(orangefs_inode->refn.khandle), &(ref->khandle)) + && orangefs_inode->refn.fs_id == ref->fs_id); +} + +/* + * Front-end to lookup the inode-cache maintained by the VFS using the ORANGEFS + * file handle. + * + * @sb: the file system super block instance. + * @ref: The ORANGEFS object for which we are trying to locate an inode structure. + */ +struct inode *orangefs_iget(struct super_block *sb, struct orangefs_object_kref *ref) +{ + struct inode *inode = NULL; + unsigned long hash; + int error; + + hash = orangefs_handle_hash(ref); + inode = iget5_locked(sb, hash, orangefs_test_inode, orangefs_set_inode, ref); + if (!inode || !(inode->i_state & I_NEW)) + return inode; + + error = orangefs_inode_getattr(inode, 1, 0); + if (error) { + iget_failed(inode); + return ERR_PTR(error); + } + + inode->i_ino = hash; /* needed for stat etc */ + orangefs_init_iops(inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_INODE_DEBUG, + "iget handle %pU, fsid %d hash %ld i_ino %lu\n", + &ref->khandle, + ref->fs_id, + hash, + inode->i_ino); + + return inode; +} + +/* + * Allocate an inode for a newly created file and insert it into the inode hash. + */ +struct inode *orangefs_new_inode(struct super_block *sb, struct inode *dir, + int mode, dev_t dev, struct orangefs_object_kref *ref) +{ + unsigned long hash = orangefs_handle_hash(ref); + struct inode *inode; + int error; + + gossip_debug(GOSSIP_INODE_DEBUG, + "%s:(sb is %p | MAJOR(dev)=%u | MINOR(dev)=%u mode=%o)\n", + __func__, + sb, + MAJOR(dev), + MINOR(dev), + mode); + + inode = new_inode(sb); + if (!inode) + return NULL; + + orangefs_set_inode(inode, ref); + inode->i_ino = hash; /* needed for stat etc */ + + error = orangefs_inode_getattr(inode, 1, 0); + if (error) + goto out_iput; + + orangefs_init_iops(inode); + + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + inode->i_size = PAGE_CACHE_SIZE; + inode->i_rdev = dev; + + error = insert_inode_locked4(inode, hash, orangefs_test_inode, ref); + if (error < 0) + goto out_iput; + + gossip_debug(GOSSIP_INODE_DEBUG, + "Initializing ACL's for inode %pU\n", + get_khandle_from_ino(inode)); + orangefs_init_acl(inode, dir); + return inode; + +out_iput: + iput(inode); + return ERR_PTR(error); +} diff --git a/fs/orangefs/namei.c b/fs/orangefs/namei.c new file mode 100644 index 000000000000..5a60c508af4e --- /dev/null +++ b/fs/orangefs/namei.c @@ -0,0 +1,462 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS namei operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* + * Get a newly allocated inode to go with a negative dentry. + */ +static int orangefs_create(struct inode *dir, + struct dentry *dentry, + umode_t mode, + bool exclusive) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s: %s\n", + __func__, + dentry->d_name.name); + + new_op = op_alloc(ORANGEFS_VFS_OP_CREATE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.create.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.create.attributes, + ORANGEFS_TYPE_METAFILE, mode); + + strncpy(new_op->upcall.req.create.d_name, + dentry->d_name.name, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: %s: handle:%pU: fsid:%d: new_op:%p: ret:%d:\n", + __func__, + dentry->d_name.name, + &new_op->downcall.resp.create.refn.khandle, + new_op->downcall.resp.create.refn.fs_id, + new_op, + ret); + + if (ret < 0) + goto out; + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFREG | mode, 0, + &new_op->downcall.resp.create.refn); + if (IS_ERR(inode)) { + gossip_err("%s: Failed to allocate inode for file :%s:\n", + __func__, + dentry->d_name.name); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: Assigned inode :%pU: for file :%s:\n", + __func__, + get_khandle_from_ino(inode), + dentry->d_name.name); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: dentry instantiated for %s\n", + __func__, + dentry->d_name.name); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + ret = 0; +out: + op_release(new_op); + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: %s: returning %d\n", + __func__, + dentry->d_name.name, + ret); + return ret; +} + +/* + * Attempt to resolve an object name (dentry->d_name), parent handle, and + * fsid into a handle for the object. + */ +static struct dentry *orangefs_lookup(struct inode *dir, struct dentry *dentry, + unsigned int flags) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + struct dentry *res; + int ret = -EINVAL; + + /* + * in theory we could skip a lookup here (if the intent is to + * create) in order to avoid a potentially failed lookup, but + * leaving it in can skip a valid lookup and try to create a file + * that already exists (e.g. the vfs already handles checking for + * -EEXIST on O_EXCL opens, which is broken if we skip this lookup + * in the create path) + */ + gossip_debug(GOSSIP_NAME_DEBUG, "%s called on %s\n", + __func__, dentry->d_name.name); + + if (dentry->d_name.len > (ORANGEFS_NAME_MAX - 1)) + return ERR_PTR(-ENAMETOOLONG); + + new_op = op_alloc(ORANGEFS_VFS_OP_LOOKUP); + if (!new_op) + return ERR_PTR(-ENOMEM); + + new_op->upcall.req.lookup.sym_follow = ORANGEFS_LOOKUP_LINK_NO_FOLLOW; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d using parent %pU\n", + __FILE__, + __func__, + __LINE__, + &parent->refn.khandle); + new_op->upcall.req.lookup.parent_refn = parent->refn; + + strncpy(new_op->upcall.req.lookup.d_name, dentry->d_name.name, + ORANGEFS_NAME_MAX); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: doing lookup on %s under %pU,%d\n", + __func__, + new_op->upcall.req.lookup.d_name, + &new_op->upcall.req.lookup.parent_refn.khandle, + new_op->upcall.req.lookup.parent_refn.fs_id); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Lookup Got %pU, fsid %d (ret=%d)\n", + &new_op->downcall.resp.lookup.refn.khandle, + new_op->downcall.resp.lookup.refn.fs_id, + ret); + + if (ret < 0) { + if (ret == -ENOENT) { + /* + * if no inode was found, add a negative dentry to + * dcache anyway; if we don't, we don't hold expected + * lookup semantics and we most noticeably break + * during directory renames. + * + * however, if the operation failed or exited, do not + * add the dentry (e.g. in the case that a touch is + * issued on a file that already exists that was + * interrupted during this lookup -- no need to add + * another negative dentry for an existing file) + */ + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_lookup: Adding *negative* dentry " + "%p for %s\n", + dentry, + dentry->d_name.name); + + d_add(dentry, NULL); + res = NULL; + goto out; + } + + /* must be a non-recoverable error */ + res = ERR_PTR(ret); + goto out; + } + + inode = orangefs_iget(dir->i_sb, &new_op->downcall.resp.lookup.refn); + if (IS_ERR(inode)) { + gossip_debug(GOSSIP_NAME_DEBUG, + "error %ld from iget\n", PTR_ERR(inode)); + res = ERR_CAST(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s:%s:%d " + "Found good inode [%lu] with count [%d]\n", + __FILE__, + __func__, + __LINE__, + inode->i_ino, + (int)atomic_read(&inode->i_count)); + + /* update dentry/inode pair into dcache */ + res = d_splice_alias(inode, dentry); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Lookup success (inode ct = %d)\n", + (int)atomic_read(&inode->i_count)); +out: + op_release(new_op); + return res; +} + +/* return 0 on success; non-zero otherwise */ +static int orangefs_unlink(struct inode *dir, struct dentry *dentry) +{ + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: called on %s\n" + " (inode %pU): Parent is %pU | fs_id %d\n", + __func__, + dentry->d_name.name, + get_khandle_from_ino(inode), + &parent->refn.khandle, + parent->refn.fs_id); + + new_op = op_alloc(ORANGEFS_VFS_OP_REMOVE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.remove.parent_refn = parent->refn; + strncpy(new_op->upcall.req.remove.d_name, dentry->d_name.name, + ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, "orangefs_unlink", + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: service_operation returned:%d:\n", + __func__, + ret); + + op_release(new_op); + + if (!ret) { + drop_nlink(inode); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + } + return ret; +} + +static int orangefs_symlink(struct inode *dir, + struct dentry *dentry, + const char *symname) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int mode = 755; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s: called\n", __func__); + + if (!symname) + return -EINVAL; + + if (strlen(symname)+1 > ORANGEFS_NAME_MAX) + return -ENAMETOOLONG; + + new_op = op_alloc(ORANGEFS_VFS_OP_SYMLINK); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.sym.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.sym.attributes, + ORANGEFS_TYPE_SYMLINK, + mode); + + strncpy(new_op->upcall.req.sym.entry_name, + dentry->d_name.name, + ORANGEFS_NAME_MAX); + strncpy(new_op->upcall.req.sym.target, symname, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Symlink Got ORANGEFS handle %pU on fsid %d (ret=%d)\n", + &new_op->downcall.resp.sym.refn.khandle, + new_op->downcall.resp.sym.refn.fs_id, ret); + + if (ret < 0) { + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: failed with error code %d\n", + __func__, ret); + goto out; + } + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFLNK | mode, 0, + &new_op->downcall.resp.sym.refn); + if (IS_ERR(inode)) { + gossip_err + ("*** Failed to allocate orangefs symlink inode\n"); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "Assigned symlink inode new number of %pU\n", + get_khandle_from_ino(inode)); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Inode (Symlink) %pU -> %s\n", + get_khandle_from_ino(inode), + dentry->d_name.name); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + ret = 0; +out: + op_release(new_op); + return ret; +} + +static int orangefs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_MKDIR); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.mkdir.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.mkdir.attributes, + ORANGEFS_TYPE_DIRECTORY, mode); + + strncpy(new_op->upcall.req.mkdir.d_name, + dentry->d_name.name, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Mkdir Got ORANGEFS handle %pU on fsid %d\n", + &new_op->downcall.resp.mkdir.refn.khandle, + new_op->downcall.resp.mkdir.refn.fs_id); + + if (ret < 0) { + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: failed with error code %d\n", + __func__, ret); + goto out; + } + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFDIR | mode, 0, + &new_op->downcall.resp.mkdir.refn); + if (IS_ERR(inode)) { + gossip_err("*** Failed to allocate orangefs dir inode\n"); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "Assigned dir inode new number of %pU\n", + get_khandle_from_ino(inode)); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Inode (Directory) %pU -> %s\n", + get_khandle_from_ino(inode), + dentry->d_name.name); + + /* + * NOTE: we have no good way to keep nlink consistent for directories + * across clients; keep constant at 1. + */ + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); +out: + op_release(new_op); + return ret; +} + +static int orangefs_rename(struct inode *old_dir, + struct dentry *old_dentry, + struct inode *new_dir, + struct dentry *new_dentry) +{ + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_rename: called (%s/%s => %s/%s) ct=%d\n", + old_dentry->d_parent->d_name.name, + old_dentry->d_name.name, + new_dentry->d_parent->d_name.name, + new_dentry->d_name.name, + d_count(new_dentry)); + + new_op = op_alloc(ORANGEFS_VFS_OP_RENAME); + if (!new_op) + return -EINVAL; + + new_op->upcall.req.rename.old_parent_refn = ORANGEFS_I(old_dir)->refn; + new_op->upcall.req.rename.new_parent_refn = ORANGEFS_I(new_dir)->refn; + + strncpy(new_op->upcall.req.rename.d_old_name, + old_dentry->d_name.name, + ORANGEFS_NAME_MAX); + strncpy(new_op->upcall.req.rename.d_new_name, + new_dentry->d_name.name, + ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, + "orangefs_rename", + get_interruptible_flag(old_dentry->d_inode)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_rename: got downcall status %d\n", + ret); + + if (new_dentry->d_inode) + new_dentry->d_inode->i_ctime = CURRENT_TIME; + + op_release(new_op); + return ret; +} + +/* ORANGEFS implementation of VFS inode operations for directories */ +struct inode_operations orangefs_dir_inode_operations = { + .lookup = orangefs_lookup, + .get_acl = orangefs_get_acl, + .set_acl = orangefs_set_acl, + .create = orangefs_create, + .unlink = orangefs_unlink, + .symlink = orangefs_symlink, + .mkdir = orangefs_mkdir, + .rmdir = orangefs_unlink, + .rename = orangefs_rename, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = orangefs_listxattr, + .permission = orangefs_permission, +}; diff --git a/fs/orangefs/orangefs-bufmap.c b/fs/orangefs/orangefs-bufmap.c new file mode 100644 index 000000000000..1f8acc9f9a88 --- /dev/null +++ b/fs/orangefs/orangefs-bufmap.c @@ -0,0 +1,556 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +struct slot_map { + int c; + wait_queue_head_t q; + int count; + unsigned long *map; +}; + +static struct slot_map rw_map = { + .c = -1, + .q = __WAIT_QUEUE_HEAD_INITIALIZER(rw_map.q) +}; +static struct slot_map readdir_map = { + .c = -1, + .q = __WAIT_QUEUE_HEAD_INITIALIZER(readdir_map.q) +}; + + +static void install(struct slot_map *m, int count, unsigned long *map) +{ + spin_lock(&m->q.lock); + m->c = m->count = count; + m->map = map; + wake_up_all_locked(&m->q); + spin_unlock(&m->q.lock); +} + +static void mark_killed(struct slot_map *m) +{ + spin_lock(&m->q.lock); + m->c -= m->count + 1; + spin_unlock(&m->q.lock); +} + +static void run_down(struct slot_map *m) +{ + DEFINE_WAIT(wait); + spin_lock(&m->q.lock); + if (m->c != -1) { + for (;;) { + if (likely(list_empty(&wait.task_list))) + __add_wait_queue_tail(&m->q, &wait); + set_current_state(TASK_UNINTERRUPTIBLE); + + if (m->c == -1) + break; + + spin_unlock(&m->q.lock); + schedule(); + spin_lock(&m->q.lock); + } + __remove_wait_queue(&m->q, &wait); + __set_current_state(TASK_RUNNING); + } + m->map = NULL; + spin_unlock(&m->q.lock); +} + +static void put(struct slot_map *m, int slot) +{ + int v; + spin_lock(&m->q.lock); + __clear_bit(slot, m->map); + v = ++m->c; + if (unlikely(v == 1)) /* no free slots -> one free slot */ + wake_up_locked(&m->q); + else if (unlikely(v == -1)) /* finished dying */ + wake_up_all_locked(&m->q); + spin_unlock(&m->q.lock); +} + +static int wait_for_free(struct slot_map *m) +{ + long left = slot_timeout_secs * HZ; + DEFINE_WAIT(wait); + + do { + long n = left, t; + if (likely(list_empty(&wait.task_list))) + __add_wait_queue_tail_exclusive(&m->q, &wait); + set_current_state(TASK_INTERRUPTIBLE); + + if (m->c > 0) + break; + + if (m->c < 0) { + /* we are waiting for map to be installed */ + /* it would better be there soon, or we go away */ + if (n > ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ) + n = ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ; + } + spin_unlock(&m->q.lock); + t = schedule_timeout(n); + spin_lock(&m->q.lock); + if (unlikely(!t) && n != left && m->c < 0) + left = t; + else + left = t + (left - n); + if (unlikely(signal_pending(current))) + left = -EINTR; + } while (left > 0); + + if (!list_empty(&wait.task_list)) + list_del(&wait.task_list); + else if (left <= 0 && waitqueue_active(&m->q)) + __wake_up_locked_key(&m->q, TASK_INTERRUPTIBLE, NULL); + __set_current_state(TASK_RUNNING); + + if (likely(left > 0)) + return 0; + + return left < 0 ? -EINTR : -ETIMEDOUT; +} + +static int get(struct slot_map *m) +{ + int res = 0; + spin_lock(&m->q.lock); + if (unlikely(m->c <= 0)) + res = wait_for_free(m); + if (likely(!res)) { + m->c--; + res = find_first_zero_bit(m->map, m->count); + __set_bit(res, m->map); + } + spin_unlock(&m->q.lock); + return res; +} + +/* used to describe mapped buffers */ +struct orangefs_bufmap_desc { + void *uaddr; /* user space address pointer */ + struct page **page_array; /* array of mapped pages */ + int array_count; /* size of above arrays */ + struct list_head list_link; +}; + +static struct orangefs_bufmap { + int desc_size; + int desc_shift; + int desc_count; + int total_size; + int page_count; + + struct page **page_array; + struct orangefs_bufmap_desc *desc_array; + + /* array to track usage of buffer descriptors */ + unsigned long *buffer_index_array; + + /* array to track usage of buffer descriptors for readdir */ +#define N DIV_ROUND_UP(ORANGEFS_READDIR_DEFAULT_DESC_COUNT, BITS_PER_LONG) + unsigned long readdir_index_array[N]; +#undef N +} *__orangefs_bufmap; + +static DEFINE_SPINLOCK(orangefs_bufmap_lock); + +static void +orangefs_bufmap_unmap(struct orangefs_bufmap *bufmap) +{ + int i; + + for (i = 0; i < bufmap->page_count; i++) + page_cache_release(bufmap->page_array[i]); +} + +static void +orangefs_bufmap_free(struct orangefs_bufmap *bufmap) +{ + kfree(bufmap->page_array); + kfree(bufmap->desc_array); + kfree(bufmap->buffer_index_array); + kfree(bufmap); +} + +/* + * XXX: Can the size and shift change while the caller gives up the + * XXX: lock between calling this and doing something useful? + */ + +int orangefs_bufmap_size_query(void) +{ + struct orangefs_bufmap *bufmap; + int size = 0; + spin_lock(&orangefs_bufmap_lock); + bufmap = __orangefs_bufmap; + if (bufmap) + size = bufmap->desc_size; + spin_unlock(&orangefs_bufmap_lock); + return size; +} + +int orangefs_bufmap_shift_query(void) +{ + struct orangefs_bufmap *bufmap; + int shift = 0; + spin_lock(&orangefs_bufmap_lock); + bufmap = __orangefs_bufmap; + if (bufmap) + shift = bufmap->desc_shift; + spin_unlock(&orangefs_bufmap_lock); + return shift; +} + +static DECLARE_WAIT_QUEUE_HEAD(bufmap_waitq); +static DECLARE_WAIT_QUEUE_HEAD(readdir_waitq); + +/* + * orangefs_get_bufmap_init + * + * If bufmap_init is 1, then the shared memory system, including the + * buffer_index_array, is available. Otherwise, it is not. + * + * returns the value of bufmap_init + */ +int orangefs_get_bufmap_init(void) +{ + return __orangefs_bufmap ? 1 : 0; +} + + +static struct orangefs_bufmap * +orangefs_bufmap_alloc(struct ORANGEFS_dev_map_desc *user_desc) +{ + struct orangefs_bufmap *bufmap; + + bufmap = kzalloc(sizeof(*bufmap), GFP_KERNEL); + if (!bufmap) + goto out; + + bufmap->total_size = user_desc->total_size; + bufmap->desc_count = user_desc->count; + bufmap->desc_size = user_desc->size; + bufmap->desc_shift = ilog2(bufmap->desc_size); + + bufmap->buffer_index_array = + kzalloc(DIV_ROUND_UP(bufmap->desc_count, BITS_PER_LONG), GFP_KERNEL); + if (!bufmap->buffer_index_array) { + gossip_err("orangefs: could not allocate %d buffer indices\n", + bufmap->desc_count); + goto out_free_bufmap; + } + + bufmap->desc_array = + kcalloc(bufmap->desc_count, sizeof(struct orangefs_bufmap_desc), + GFP_KERNEL); + if (!bufmap->desc_array) { + gossip_err("orangefs: could not allocate %d descriptors\n", + bufmap->desc_count); + goto out_free_index_array; + } + + bufmap->page_count = bufmap->total_size / PAGE_SIZE; + + /* allocate storage to track our page mappings */ + bufmap->page_array = + kcalloc(bufmap->page_count, sizeof(struct page *), GFP_KERNEL); + if (!bufmap->page_array) + goto out_free_desc_array; + + return bufmap; + +out_free_desc_array: + kfree(bufmap->desc_array); +out_free_index_array: + kfree(bufmap->buffer_index_array); +out_free_bufmap: + kfree(bufmap); +out: + return NULL; +} + +static int +orangefs_bufmap_map(struct orangefs_bufmap *bufmap, + struct ORANGEFS_dev_map_desc *user_desc) +{ + int pages_per_desc = bufmap->desc_size / PAGE_SIZE; + int offset = 0, ret, i; + + /* map the pages */ + ret = get_user_pages_fast((unsigned long)user_desc->ptr, + bufmap->page_count, 1, bufmap->page_array); + + if (ret < 0) + return ret; + + if (ret != bufmap->page_count) { + gossip_err("orangefs error: asked for %d pages, only got %d.\n", + bufmap->page_count, ret); + + for (i = 0; i < ret; i++) { + SetPageError(bufmap->page_array[i]); + page_cache_release(bufmap->page_array[i]); + } + return -ENOMEM; + } + + /* + * ideally we want to get kernel space pointers for each page, but + * we can't kmap that many pages at once if highmem is being used. + * so instead, we just kmap/kunmap the page address each time the + * kaddr is needed. + */ + for (i = 0; i < bufmap->page_count; i++) + flush_dcache_page(bufmap->page_array[i]); + + /* build a list of available descriptors */ + for (offset = 0, i = 0; i < bufmap->desc_count; i++) { + bufmap->desc_array[i].page_array = &bufmap->page_array[offset]; + bufmap->desc_array[i].array_count = pages_per_desc; + bufmap->desc_array[i].uaddr = + (user_desc->ptr + (i * pages_per_desc * PAGE_SIZE)); + offset += pages_per_desc; + } + + return 0; +} + +/* + * orangefs_bufmap_initialize() + * + * initializes the mapped buffer interface + * + * returns 0 on success, -errno on failure + */ +int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc) +{ + struct orangefs_bufmap *bufmap; + int ret = -EINVAL; + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_initialize: called (ptr (" + "%p) sz (%d) cnt(%d).\n", + user_desc->ptr, + user_desc->size, + user_desc->count); + + /* + * sanity check alignment and size of buffer that caller wants to + * work with + */ + if (PAGE_ALIGN((unsigned long)user_desc->ptr) != + (unsigned long)user_desc->ptr) { + gossip_err("orangefs error: memory alignment (front). %p\n", + user_desc->ptr); + goto out; + } + + if (PAGE_ALIGN(((unsigned long)user_desc->ptr + user_desc->total_size)) + != (unsigned long)(user_desc->ptr + user_desc->total_size)) { + gossip_err("orangefs error: memory alignment (back).(%p + %d)\n", + user_desc->ptr, + user_desc->total_size); + goto out; + } + + if (user_desc->total_size != (user_desc->size * user_desc->count)) { + gossip_err("orangefs error: user provided an oddly sized buffer: (%d, %d, %d)\n", + user_desc->total_size, + user_desc->size, + user_desc->count); + goto out; + } + + if ((user_desc->size % PAGE_SIZE) != 0) { + gossip_err("orangefs error: bufmap size not page size divisible (%d).\n", + user_desc->size); + goto out; + } + + ret = -ENOMEM; + bufmap = orangefs_bufmap_alloc(user_desc); + if (!bufmap) + goto out; + + ret = orangefs_bufmap_map(bufmap, user_desc); + if (ret) + goto out_free_bufmap; + + + spin_lock(&orangefs_bufmap_lock); + if (__orangefs_bufmap) { + spin_unlock(&orangefs_bufmap_lock); + gossip_err("orangefs: error: bufmap already initialized.\n"); + ret = -EINVAL; + goto out_unmap_bufmap; + } + __orangefs_bufmap = bufmap; + install(&rw_map, + bufmap->desc_count, + bufmap->buffer_index_array); + install(&readdir_map, + ORANGEFS_READDIR_DEFAULT_DESC_COUNT, + bufmap->readdir_index_array); + spin_unlock(&orangefs_bufmap_lock); + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_initialize: exiting normally\n"); + return 0; + +out_unmap_bufmap: + orangefs_bufmap_unmap(bufmap); +out_free_bufmap: + orangefs_bufmap_free(bufmap); +out: + return ret; +} + +/* + * orangefs_bufmap_finalize() + * + * shuts down the mapped buffer interface and releases any resources + * associated with it + * + * no return value + */ +void orangefs_bufmap_finalize(void) +{ + struct orangefs_bufmap *bufmap = __orangefs_bufmap; + if (!bufmap) + return; + gossip_debug(GOSSIP_BUFMAP_DEBUG, "orangefs_bufmap_finalize: called\n"); + mark_killed(&rw_map); + mark_killed(&readdir_map); + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_finalize: exiting normally\n"); +} + +void orangefs_bufmap_run_down(void) +{ + struct orangefs_bufmap *bufmap = __orangefs_bufmap; + if (!bufmap) + return; + run_down(&rw_map); + run_down(&readdir_map); + spin_lock(&orangefs_bufmap_lock); + __orangefs_bufmap = NULL; + spin_unlock(&orangefs_bufmap_lock); + orangefs_bufmap_unmap(bufmap); + orangefs_bufmap_free(bufmap); +} + +/* + * orangefs_bufmap_get() + * + * gets a free mapped buffer descriptor, will sleep until one becomes + * available if necessary + * + * returns slot on success, -errno on failure + */ +int orangefs_bufmap_get(void) +{ + return get(&rw_map); +} + +/* + * orangefs_bufmap_put() + * + * returns a mapped buffer descriptor to the collection + * + * no return value + */ +void orangefs_bufmap_put(int buffer_index) +{ + put(&rw_map, buffer_index); +} + +/* + * orangefs_readdir_index_get() + * + * gets a free descriptor, will sleep until one becomes + * available if necessary. + * Although the readdir buffers are not mapped into kernel space + * we could do that at a later point of time. Regardless, these + * indices are used by the client-core. + * + * returns slot on success, -errno on failure + */ +int orangefs_readdir_index_get(void) +{ + return get(&readdir_map); +} + +void orangefs_readdir_index_put(int buffer_index) +{ + put(&readdir_map, buffer_index); +} + +/* + * we've been handed an iovec, we need to copy it to + * the shared memory descriptor at "buffer_index". + */ +int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter, + int buffer_index, + size_t size) +{ + struct orangefs_bufmap_desc *to; + int i; + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "%s: buffer_index:%d: size:%zu:\n", + __func__, buffer_index, size); + + to = &__orangefs_bufmap->desc_array[buffer_index]; + for (i = 0; size; i++) { + struct page *page = to->page_array[i]; + size_t n = size; + if (n > PAGE_SIZE) + n = PAGE_SIZE; + n = copy_page_from_iter(page, 0, n, iter); + if (!n) + return -EFAULT; + size -= n; + } + return 0; + +} + +/* + * we've been handed an iovec, we need to fill it from + * the shared memory descriptor at "buffer_index". + */ +int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter, + int buffer_index, + size_t size) +{ + struct orangefs_bufmap_desc *from; + int i; + + from = &__orangefs_bufmap->desc_array[buffer_index]; + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "%s: buffer_index:%d: size:%zu:\n", + __func__, buffer_index, size); + + + for (i = 0; size; i++) { + struct page *page = from->page_array[i]; + size_t n = size; + if (n > PAGE_SIZE) + n = PAGE_SIZE; + n = copy_page_to_iter(page, 0, n, iter); + if (!n) + return -EFAULT; + size -= n; + } + return 0; +} diff --git a/fs/orangefs/orangefs-bufmap.h b/fs/orangefs/orangefs-bufmap.h new file mode 100644 index 000000000000..71f64f4057b5 --- /dev/null +++ b/fs/orangefs/orangefs-bufmap.h @@ -0,0 +1,36 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef __ORANGEFS_BUFMAP_H +#define __ORANGEFS_BUFMAP_H + +int orangefs_bufmap_size_query(void); + +int orangefs_bufmap_shift_query(void); + +int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc); + +void orangefs_bufmap_finalize(void); + +void orangefs_bufmap_run_down(void); + +int orangefs_bufmap_get(void); + +void orangefs_bufmap_put(int buffer_index); + +int orangefs_readdir_index_get(void); + +void orangefs_readdir_index_put(int buffer_index); + +int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter, + int buffer_index, + size_t size); + +int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter, + int buffer_index, + size_t size); + +#endif /* __ORANGEFS_BUFMAP_H */ diff --git a/fs/orangefs/orangefs-cache.c b/fs/orangefs/orangefs-cache.c new file mode 100644 index 000000000000..900a2e38e11b --- /dev/null +++ b/fs/orangefs/orangefs-cache.c @@ -0,0 +1,161 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* tags assigned to kernel upcall operations */ +static __u64 next_tag_value; +static DEFINE_SPINLOCK(next_tag_value_lock); + +/* the orangefs memory caches */ + +/* a cache for orangefs upcall/downcall operations */ +static struct kmem_cache *op_cache; + +int op_cache_initialize(void) +{ + op_cache = kmem_cache_create("orangefs_op_cache", + sizeof(struct orangefs_kernel_op_s), + 0, + ORANGEFS_CACHE_CREATE_FLAGS, + NULL); + + if (!op_cache) { + gossip_err("Cannot create orangefs_op_cache\n"); + return -ENOMEM; + } + + /* initialize our atomic tag counter */ + spin_lock(&next_tag_value_lock); + next_tag_value = 100; + spin_unlock(&next_tag_value_lock); + return 0; +} + +int op_cache_finalize(void) +{ + kmem_cache_destroy(op_cache); + return 0; +} + +char *get_opname_string(struct orangefs_kernel_op_s *new_op) +{ + if (new_op) { + __s32 type = new_op->upcall.type; + + if (type == ORANGEFS_VFS_OP_FILE_IO) + return "OP_FILE_IO"; + else if (type == ORANGEFS_VFS_OP_LOOKUP) + return "OP_LOOKUP"; + else if (type == ORANGEFS_VFS_OP_CREATE) + return "OP_CREATE"; + else if (type == ORANGEFS_VFS_OP_GETATTR) + return "OP_GETATTR"; + else if (type == ORANGEFS_VFS_OP_REMOVE) + return "OP_REMOVE"; + else if (type == ORANGEFS_VFS_OP_MKDIR) + return "OP_MKDIR"; + else if (type == ORANGEFS_VFS_OP_READDIR) + return "OP_READDIR"; + else if (type == ORANGEFS_VFS_OP_READDIRPLUS) + return "OP_READDIRPLUS"; + else if (type == ORANGEFS_VFS_OP_SETATTR) + return "OP_SETATTR"; + else if (type == ORANGEFS_VFS_OP_SYMLINK) + return "OP_SYMLINK"; + else if (type == ORANGEFS_VFS_OP_RENAME) + return "OP_RENAME"; + else if (type == ORANGEFS_VFS_OP_STATFS) + return "OP_STATFS"; + else if (type == ORANGEFS_VFS_OP_TRUNCATE) + return "OP_TRUNCATE"; + else if (type == ORANGEFS_VFS_OP_MMAP_RA_FLUSH) + return "OP_MMAP_RA_FLUSH"; + else if (type == ORANGEFS_VFS_OP_FS_MOUNT) + return "OP_FS_MOUNT"; + else if (type == ORANGEFS_VFS_OP_FS_UMOUNT) + return "OP_FS_UMOUNT"; + else if (type == ORANGEFS_VFS_OP_GETXATTR) + return "OP_GETXATTR"; + else if (type == ORANGEFS_VFS_OP_SETXATTR) + return "OP_SETXATTR"; + else if (type == ORANGEFS_VFS_OP_LISTXATTR) + return "OP_LISTXATTR"; + else if (type == ORANGEFS_VFS_OP_REMOVEXATTR) + return "OP_REMOVEXATTR"; + else if (type == ORANGEFS_VFS_OP_PARAM) + return "OP_PARAM"; + else if (type == ORANGEFS_VFS_OP_PERF_COUNT) + return "OP_PERF_COUNT"; + else if (type == ORANGEFS_VFS_OP_CANCEL) + return "OP_CANCEL"; + else if (type == ORANGEFS_VFS_OP_FSYNC) + return "OP_FSYNC"; + else if (type == ORANGEFS_VFS_OP_FSKEY) + return "OP_FSKEY"; + } + return "OP_UNKNOWN?"; +} + +void orangefs_new_tag(struct orangefs_kernel_op_s *op) +{ + spin_lock(&next_tag_value_lock); + op->tag = next_tag_value++; + if (next_tag_value == 0) + next_tag_value = 100; + spin_unlock(&next_tag_value_lock); +} + +struct orangefs_kernel_op_s *op_alloc(__s32 type) +{ + struct orangefs_kernel_op_s *new_op = NULL; + + new_op = kmem_cache_zalloc(op_cache, GFP_KERNEL); + if (new_op) { + INIT_LIST_HEAD(&new_op->list); + spin_lock_init(&new_op->lock); + init_completion(&new_op->waitq); + + new_op->upcall.type = ORANGEFS_VFS_OP_INVALID; + new_op->downcall.type = ORANGEFS_VFS_OP_INVALID; + new_op->downcall.status = -1; + + new_op->op_state = OP_VFS_STATE_UNKNOWN; + + /* initialize the op specific tag and upcall credentials */ + orangefs_new_tag(new_op); + new_op->upcall.type = type; + new_op->attempts = 0; + gossip_debug(GOSSIP_CACHE_DEBUG, + "Alloced OP (%p: %llu %s)\n", + new_op, + llu(new_op->tag), + get_opname_string(new_op)); + + new_op->upcall.uid = from_kuid(current_user_ns(), + current_fsuid()); + + new_op->upcall.gid = from_kgid(current_user_ns(), + current_fsgid()); + } else { + gossip_err("op_alloc: kmem_cache_zalloc failed!\n"); + } + return new_op; +} + +void op_release(struct orangefs_kernel_op_s *orangefs_op) +{ + if (orangefs_op) { + gossip_debug(GOSSIP_CACHE_DEBUG, + "Releasing OP (%p: %llu)\n", + orangefs_op, + llu(orangefs_op->tag)); + kmem_cache_free(op_cache, orangefs_op); + } else { + gossip_err("NULL pointer in op_release\n"); + } +} diff --git a/fs/orangefs/orangefs-debug.h b/fs/orangefs/orangefs-debug.h new file mode 100644 index 000000000000..387db17cde2b --- /dev/null +++ b/fs/orangefs/orangefs-debug.h @@ -0,0 +1,92 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* This file just defines debugging masks to be used with the gossip + * logging utility. All debugging masks for ORANGEFS are kept here to make + * sure we don't have collisions. + */ + +#ifndef __ORANGEFS_DEBUG_H +#define __ORANGEFS_DEBUG_H + +#ifdef __KERNEL__ +#include <linux/types.h> +#else +#include <stdint.h> +#endif + +#define GOSSIP_NO_DEBUG (__u64)0 + +#define GOSSIP_SUPER_DEBUG ((__u64)1 << 0) +#define GOSSIP_INODE_DEBUG ((__u64)1 << 1) +#define GOSSIP_FILE_DEBUG ((__u64)1 << 2) +#define GOSSIP_DIR_DEBUG ((__u64)1 << 3) +#define GOSSIP_UTILS_DEBUG ((__u64)1 << 4) +#define GOSSIP_WAIT_DEBUG ((__u64)1 << 5) +#define GOSSIP_ACL_DEBUG ((__u64)1 << 6) +#define GOSSIP_DCACHE_DEBUG ((__u64)1 << 7) +#define GOSSIP_DEV_DEBUG ((__u64)1 << 8) +#define GOSSIP_NAME_DEBUG ((__u64)1 << 9) +#define GOSSIP_BUFMAP_DEBUG ((__u64)1 << 10) +#define GOSSIP_CACHE_DEBUG ((__u64)1 << 11) +#define GOSSIP_DEBUGFS_DEBUG ((__u64)1 << 12) +#define GOSSIP_XATTR_DEBUG ((__u64)1 << 13) +#define GOSSIP_INIT_DEBUG ((__u64)1 << 14) +#define GOSSIP_SYSFS_DEBUG ((__u64)1 << 15) + +#define GOSSIP_MAX_NR 16 +#define GOSSIP_MAX_DEBUG (((__u64)1 << GOSSIP_MAX_NR) - 1) + +/*function prototypes*/ +__u64 ORANGEFS_kmod_eventlog_to_mask(const char *event_logging); +__u64 ORANGEFS_debug_eventlog_to_mask(const char *event_logging); +char *ORANGEFS_debug_mask_to_eventlog(__u64 mask); +char *ORANGEFS_kmod_mask_to_eventlog(__u64 mask); + +/* a private internal type */ +struct __keyword_mask_s { + const char *keyword; + __u64 mask_val; +}; + +/* + * Map all kmod keywords to kmod debug masks here. Keep this + * structure "packed": + * + * "all" is always last... + * + * keyword mask_val index + * foo 1 0 + * bar 2 1 + * baz 4 2 + * qux 8 3 + * . . . + */ +static struct __keyword_mask_s s_kmod_keyword_mask_map[] = { + {"super", GOSSIP_SUPER_DEBUG}, + {"inode", GOSSIP_INODE_DEBUG}, + {"file", GOSSIP_FILE_DEBUG}, + {"dir", GOSSIP_DIR_DEBUG}, + {"utils", GOSSIP_UTILS_DEBUG}, + {"wait", GOSSIP_WAIT_DEBUG}, + {"acl", GOSSIP_ACL_DEBUG}, + {"dcache", GOSSIP_DCACHE_DEBUG}, + {"dev", GOSSIP_DEV_DEBUG}, + {"name", GOSSIP_NAME_DEBUG}, + {"bufmap", GOSSIP_BUFMAP_DEBUG}, + {"cache", GOSSIP_CACHE_DEBUG}, + {"debugfs", GOSSIP_DEBUGFS_DEBUG}, + {"xattr", GOSSIP_XATTR_DEBUG}, + {"init", GOSSIP_INIT_DEBUG}, + {"sysfs", GOSSIP_SYSFS_DEBUG}, + {"none", GOSSIP_NO_DEBUG}, + {"all", GOSSIP_MAX_DEBUG} +}; + +static const int num_kmod_keyword_mask_map = (int) + (sizeof(s_kmod_keyword_mask_map) / sizeof(struct __keyword_mask_s)); + +#endif /* __ORANGEFS_DEBUG_H */ diff --git a/fs/orangefs/orangefs-debugfs.c b/fs/orangefs/orangefs-debugfs.c new file mode 100644 index 000000000000..19670b8b4053 --- /dev/null +++ b/fs/orangefs/orangefs-debugfs.c @@ -0,0 +1,455 @@ +/* + * What: /sys/kernel/debug/orangefs/debug-help + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * List of client and kernel debug keywords. + * + * + * What: /sys/kernel/debug/orangefs/client-debug + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Debug setting for "the client", the userspace + * helper for the kernel module. + * + * + * What: /sys/kernel/debug/orangefs/kernel-debug + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Debug setting for the orangefs kernel module. + * + * Any of the keywords, or comma-separated lists + * of keywords, from debug-help can be catted to + * client-debug or kernel-debug. + * + * "none", "all" and "verbose" are special keywords + * for client-debug. Setting client-debug to "all" + * is kind of like trying to drink water from a + * fire hose, "verbose" triggers most of the same + * output except for the constant flow of output + * from the main wait loop. + * + * "none" and "all" are similar settings for kernel-debug + * no need for a "verbose". + */ +#include <linux/debugfs.h> +#include <linux/slab.h> + +#include <linux/uaccess.h> + +#include "orangefs-debugfs.h" +#include "protocol.h" +#include "orangefs-kernel.h" + +static int orangefs_debug_disabled = 1; + +static int orangefs_debug_help_open(struct inode *, struct file *); + +const struct file_operations debug_help_fops = { + .open = orangefs_debug_help_open, + .read = seq_read, + .release = seq_release, + .llseek = seq_lseek, +}; + +static void *help_start(struct seq_file *, loff_t *); +static void *help_next(struct seq_file *, void *, loff_t *); +static void help_stop(struct seq_file *, void *); +static int help_show(struct seq_file *, void *); + +static const struct seq_operations help_debug_ops = { + .start = help_start, + .next = help_next, + .stop = help_stop, + .show = help_show, +}; + +/* + * Used to protect data in ORANGEFS_KMOD_DEBUG_FILE and + * ORANGEFS_KMOD_DEBUG_FILE. + */ +static DEFINE_MUTEX(orangefs_debug_lock); + +int orangefs_debug_open(struct inode *, struct file *); + +static ssize_t orangefs_debug_read(struct file *, + char __user *, + size_t, + loff_t *); + +static ssize_t orangefs_debug_write(struct file *, + const char __user *, + size_t, + loff_t *); + +static const struct file_operations kernel_debug_fops = { + .open = orangefs_debug_open, + .read = orangefs_debug_read, + .write = orangefs_debug_write, + .llseek = generic_file_llseek, +}; + +/* + * initialize kmod debug operations, create orangefs debugfs dir and + * ORANGEFS_KMOD_DEBUG_HELP_FILE. + */ +int orangefs_debugfs_init(void) +{ + + int rc = -ENOMEM; + + debug_dir = debugfs_create_dir("orangefs", NULL); + if (!debug_dir) { + pr_info("%s: debugfs_create_dir failed.\n", __func__); + goto out; + } + + help_file_dentry = debugfs_create_file(ORANGEFS_KMOD_DEBUG_HELP_FILE, + 0444, + debug_dir, + debug_help_string, + &debug_help_fops); + if (!help_file_dentry) { + pr_info("%s: debugfs_create_file failed.\n", __func__); + goto out; + } + + orangefs_debug_disabled = 0; + rc = 0; + +out: + + return rc; +} + +void orangefs_debugfs_cleanup(void) +{ + if (debug_dir) + debugfs_remove_recursive(debug_dir); +} + +/* open ORANGEFS_KMOD_DEBUG_HELP_FILE */ +static int orangefs_debug_help_open(struct inode *inode, struct file *file) +{ + int rc = -ENODEV; + int ret; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_help_open: start\n"); + + if (orangefs_debug_disabled) + goto out; + + ret = seq_open(file, &help_debug_ops); + if (ret) + goto out; + + ((struct seq_file *)(file->private_data))->private = inode->i_private; + + rc = 0; + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_help_open: rc:%d:\n", + rc); + return rc; +} + +/* + * I think start always gets called again after stop. Start + * needs to return NULL when it is done. The whole "payload" + * in this case is a single (long) string, so by the second + * time we get to start (pos = 1), we're done. + */ +static void *help_start(struct seq_file *m, loff_t *pos) +{ + void *payload = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_start: start\n"); + + if (*pos == 0) + payload = m->private; + + return payload; +} + +static void *help_next(struct seq_file *m, void *v, loff_t *pos) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_next: start\n"); + + return NULL; +} + +static void help_stop(struct seq_file *m, void *p) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_stop: start\n"); +} + +static int help_show(struct seq_file *m, void *v) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_show: start\n"); + + seq_puts(m, v); + + return 0; +} + +/* + * initialize the kernel-debug file. + */ +int orangefs_kernel_debug_init(void) +{ + int rc = -ENOMEM; + struct dentry *ret; + char *k_buffer = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: start\n", __func__); + + k_buffer = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!k_buffer) + goto out; + + if (strlen(kernel_debug_string) + 1 < ORANGEFS_MAX_DEBUG_STRING_LEN) { + strcpy(k_buffer, kernel_debug_string); + strcat(k_buffer, "\n"); + } else { + strcpy(k_buffer, "none\n"); + pr_info("%s: overflow 1!\n", __func__); + } + + ret = debugfs_create_file(ORANGEFS_KMOD_DEBUG_FILE, + 0444, + debug_dir, + k_buffer, + &kernel_debug_fops); + if (!ret) { + pr_info("%s: failed to create %s.\n", + __func__, + ORANGEFS_KMOD_DEBUG_FILE); + goto out; + } + + rc = 0; + +out: + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: rc:%d:\n", __func__, rc); + return rc; +} + +/* + * initialize the client-debug file. + */ +int orangefs_client_debug_init(void) +{ + + int rc = -ENOMEM; + char *c_buffer = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: start\n", __func__); + + c_buffer = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!c_buffer) + goto out; + + if (strlen(client_debug_string) + 1 < ORANGEFS_MAX_DEBUG_STRING_LEN) { + strcpy(c_buffer, client_debug_string); + strcat(c_buffer, "\n"); + } else { + strcpy(c_buffer, "none\n"); + pr_info("%s: overflow! 2\n", __func__); + } + + client_debug_dentry = debugfs_create_file(ORANGEFS_CLIENT_DEBUG_FILE, + 0444, + debug_dir, + c_buffer, + &kernel_debug_fops); + if (!client_debug_dentry) { + pr_info("%s: failed to create updated %s.\n", + __func__, + ORANGEFS_CLIENT_DEBUG_FILE); + goto out; + } + + rc = 0; + +out: + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: rc:%d:\n", __func__, rc); + return rc; +} + +/* open ORANGEFS_KMOD_DEBUG_FILE or ORANGEFS_CLIENT_DEBUG_FILE.*/ +int orangefs_debug_open(struct inode *inode, struct file *file) +{ + int rc = -ENODEV; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: orangefs_debug_disabled: %d\n", + __func__, + orangefs_debug_disabled); + + if (orangefs_debug_disabled) + goto out; + + rc = 0; + mutex_lock(&orangefs_debug_lock); + file->private_data = inode->i_private; + mutex_unlock(&orangefs_debug_lock); + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_open: rc: %d\n", + rc); + return rc; +} + +static ssize_t orangefs_debug_read(struct file *file, + char __user *ubuf, + size_t count, + loff_t *ppos) +{ + char *buf; + int sprintf_ret; + ssize_t read_ret = -ENOMEM; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "orangefs_debug_read: start\n"); + + buf = kmalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!buf) + goto out; + + mutex_lock(&orangefs_debug_lock); + sprintf_ret = sprintf(buf, "%s", (char *)file->private_data); + mutex_unlock(&orangefs_debug_lock); + + read_ret = simple_read_from_buffer(ubuf, count, ppos, buf, sprintf_ret); + + kfree(buf); + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_read: ret: %zu\n", + read_ret); + + return read_ret; +} + +static ssize_t orangefs_debug_write(struct file *file, + const char __user *ubuf, + size_t count, + loff_t *ppos) +{ + char *buf; + int rc = -EFAULT; + size_t silly = 0; + char *debug_string; + struct orangefs_kernel_op_s *new_op = NULL; + struct client_debug_mask c_mask = { NULL, 0, 0 }; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_write: %s\n", + file->f_path.dentry->d_name.name); + + /* + * Thwart users who try to jamb a ridiculous number + * of bytes into the debug file... + */ + if (count > ORANGEFS_MAX_DEBUG_STRING_LEN + 1) { + silly = count; + count = ORANGEFS_MAX_DEBUG_STRING_LEN + 1; + } + + buf = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!buf) + goto out; + + if (copy_from_user(buf, ubuf, count - 1)) { + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: copy_from_user failed!\n", + __func__); + goto out; + } + + /* + * Map the keyword string from userspace into a valid debug mask. + * The mapping process involves mapping the human-inputted string + * into a valid mask, and then rebuilding the string from the + * verified valid mask. + * + * A service operation is required to set a new client-side + * debug mask. + */ + if (!strcmp(file->f_path.dentry->d_name.name, + ORANGEFS_KMOD_DEBUG_FILE)) { + debug_string_to_mask(buf, &gossip_debug_mask, 0); + debug_mask_to_string(&gossip_debug_mask, 0); + debug_string = kernel_debug_string; + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "New kernel debug string is %s\n", + kernel_debug_string); + } else { + /* Can't reset client debug mask if client is not running. */ + if (is_daemon_in_service()) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + debug_string_to_mask(buf, &c_mask, 1); + debug_mask_to_string(&c_mask, 1); + debug_string = client_debug_string; + + new_op = op_alloc(ORANGEFS_VFS_OP_PARAM); + if (!new_op) { + pr_info("%s: op_alloc failed!\n", __func__); + goto out; + } + + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_TWO_MASK_VALUES; + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_SET; + memset(new_op->upcall.req.param.s_value, + 0, + ORANGEFS_MAX_DEBUG_STRING_LEN); + sprintf(new_op->upcall.req.param.s_value, + "%llx %llx\n", + c_mask.mask1, + c_mask.mask2); + + /* service_operation returns 0 on success... */ + rc = service_operation(new_op, + "orangefs_param", + ORANGEFS_OP_INTERRUPTIBLE); + + if (rc) + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: service_operation failed! rc:%d:\n", + __func__, + rc); + + op_release(new_op); + } + + mutex_lock(&orangefs_debug_lock); + memset(file->f_inode->i_private, 0, ORANGEFS_MAX_DEBUG_STRING_LEN); + sprintf((char *)file->f_inode->i_private, "%s\n", debug_string); + mutex_unlock(&orangefs_debug_lock); + + *ppos += count; + if (silly) + rc = silly; + else + rc = count; + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_write: rc: %d\n", + rc); + kfree(buf); + return rc; +} diff --git a/fs/orangefs/orangefs-debugfs.h b/fs/orangefs/orangefs-debugfs.h new file mode 100644 index 000000000000..e4828c0e3ef9 --- /dev/null +++ b/fs/orangefs/orangefs-debugfs.h @@ -0,0 +1,3 @@ +int orangefs_debugfs_init(void); +int orangefs_kernel_debug_init(void); +void orangefs_debugfs_cleanup(void); diff --git a/fs/orangefs/orangefs-dev-proto.h b/fs/orangefs/orangefs-dev-proto.h new file mode 100644 index 000000000000..9eac9d9a3f3a --- /dev/null +++ b/fs/orangefs/orangefs-dev-proto.h @@ -0,0 +1,62 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef _ORANGEFS_DEV_PROTO_H +#define _ORANGEFS_DEV_PROTO_H + +/* + * types and constants shared between user space and kernel space for + * device interaction using a common protocol + */ + +/* + * valid orangefs kernel operation types + */ +#define ORANGEFS_VFS_OP_INVALID 0xFF000000 +#define ORANGEFS_VFS_OP_FILE_IO 0xFF000001 +#define ORANGEFS_VFS_OP_LOOKUP 0xFF000002 +#define ORANGEFS_VFS_OP_CREATE 0xFF000003 +#define ORANGEFS_VFS_OP_GETATTR 0xFF000004 +#define ORANGEFS_VFS_OP_REMOVE 0xFF000005 +#define ORANGEFS_VFS_OP_MKDIR 0xFF000006 +#define ORANGEFS_VFS_OP_READDIR 0xFF000007 +#define ORANGEFS_VFS_OP_SETATTR 0xFF000008 +#define ORANGEFS_VFS_OP_SYMLINK 0xFF000009 +#define ORANGEFS_VFS_OP_RENAME 0xFF00000A +#define ORANGEFS_VFS_OP_STATFS 0xFF00000B +#define ORANGEFS_VFS_OP_TRUNCATE 0xFF00000C +#define ORANGEFS_VFS_OP_MMAP_RA_FLUSH 0xFF00000D +#define ORANGEFS_VFS_OP_FS_MOUNT 0xFF00000E +#define ORANGEFS_VFS_OP_FS_UMOUNT 0xFF00000F +#define ORANGEFS_VFS_OP_GETXATTR 0xFF000010 +#define ORANGEFS_VFS_OP_SETXATTR 0xFF000011 +#define ORANGEFS_VFS_OP_LISTXATTR 0xFF000012 +#define ORANGEFS_VFS_OP_REMOVEXATTR 0xFF000013 +#define ORANGEFS_VFS_OP_PARAM 0xFF000014 +#define ORANGEFS_VFS_OP_PERF_COUNT 0xFF000015 +#define ORANGEFS_VFS_OP_CANCEL 0xFF00EE00 +#define ORANGEFS_VFS_OP_FSYNC 0xFF00EE01 +#define ORANGEFS_VFS_OP_FSKEY 0xFF00EE02 +#define ORANGEFS_VFS_OP_READDIRPLUS 0xFF00EE03 + +/* + * Misc constants. Please retain them as multiples of 8! + * Otherwise 32-64 bit interactions will be messed up :) + */ +#define ORANGEFS_MAX_DEBUG_STRING_LEN 0x00000400 +#define ORANGEFS_MAX_DEBUG_ARRAY_LEN 0x00000800 + +/* + * The maximum number of directory entries in a single request is 96. + * XXX: Why can this not be higher. The client-side code can handle up to 512. + * XXX: What happens if we expect more than the client can return? + */ +#define ORANGEFS_MAX_DIRENT_COUNT_READDIR 96 + +#include "upcall.h" +#include "downcall.h" + +#endif diff --git a/fs/orangefs/orangefs-kernel.h b/fs/orangefs/orangefs-kernel.h new file mode 100644 index 000000000000..a9925e296ceb --- /dev/null +++ b/fs/orangefs/orangefs-kernel.h @@ -0,0 +1,623 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * The ORANGEFS Linux kernel support allows ORANGEFS volumes to be mounted and + * accessed through the Linux VFS (i.e. using standard I/O system calls). + * This support is only needed on clients that wish to mount the file system. + * + */ + +/* + * Declarations and macros for the ORANGEFS Linux kernel support. + */ + +#ifndef __ORANGEFSKERNEL_H +#define __ORANGEFSKERNEL_H + +#include <linux/kernel.h> +#include <linux/moduleparam.h> +#include <linux/statfs.h> +#include <linux/backing-dev.h> +#include <linux/device.h> +#include <linux/mpage.h> +#include <linux/namei.h> +#include <linux/errno.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/types.h> +#include <linux/fs.h> +#include <linux/vmalloc.h> + +#include <linux/aio.h> +#include <linux/posix_acl.h> +#include <linux/posix_acl_xattr.h> +#include <linux/compat.h> +#include <linux/mount.h> +#include <linux/uaccess.h> +#include <linux/atomic.h> +#include <linux/uio.h> +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/wait.h> +#include <linux/dcache.h> +#include <linux/pagemap.h> +#include <linux/poll.h> +#include <linux/rwsem.h> +#include <linux/xattr.h> +#include <linux/exportfs.h> + +#include <asm/unaligned.h> + +#include "orangefs-dev-proto.h" + +#ifdef ORANGEFS_KERNEL_DEBUG +#define ORANGEFS_DEFAULT_OP_TIMEOUT_SECS 10 +#else +#define ORANGEFS_DEFAULT_OP_TIMEOUT_SECS 20 +#endif + +#define ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS 30 + +#define ORANGEFS_DEFAULT_SLOT_TIMEOUT_SECS 900 /* 15 minutes */ + +#define ORANGEFS_REQDEVICE_NAME "pvfs2-req" + +#define ORANGEFS_DEVREQ_MAGIC 0x20030529 +#define ORANGEFS_LINK_MAX 0x000000FF +#define ORANGEFS_PURGE_RETRY_COUNT 0x00000005 +#define ORANGEFS_MAX_NUM_OPTIONS 0x00000004 +#define ORANGEFS_MAX_MOUNT_OPT_LEN 0x00000080 +#define ORANGEFS_MAX_FSKEY_LEN 64 + +#define MAX_DEV_REQ_UPSIZE (2 * sizeof(__s32) + \ +sizeof(__u64) + sizeof(struct orangefs_upcall_s)) +#define MAX_DEV_REQ_DOWNSIZE (2 * sizeof(__s32) + \ +sizeof(__u64) + sizeof(struct orangefs_downcall_s)) + +/* + * valid orangefs kernel operation states + * + * unknown - op was just initialized + * waiting - op is on request_list (upward bound) + * inprogr - op is in progress (waiting for downcall) + * serviced - op has matching downcall; ok + * purged - op has to start a timer since client-core + * exited uncleanly before servicing op + * given up - submitter has given up waiting for it + */ +enum orangefs_vfs_op_states { + OP_VFS_STATE_UNKNOWN = 0, + OP_VFS_STATE_WAITING = 1, + OP_VFS_STATE_INPROGR = 2, + OP_VFS_STATE_SERVICED = 4, + OP_VFS_STATE_PURGED = 8, + OP_VFS_STATE_GIVEN_UP = 16, +}; + +/* + * An array of client_debug_mask will be built to hold debug keyword/mask + * values fetched from userspace. + */ +struct client_debug_mask { + char *keyword; + __u64 mask1; + __u64 mask2; +}; + +/* + * orangefs kernel memory related flags + */ + +#if ((defined ORANGEFS_KERNEL_DEBUG) && (defined CONFIG_DEBUG_SLAB)) +#define ORANGEFS_CACHE_CREATE_FLAGS SLAB_RED_ZONE +#else +#define ORANGEFS_CACHE_CREATE_FLAGS 0 +#endif /* ((defined ORANGEFS_KERNEL_DEBUG) && (defined CONFIG_DEBUG_SLAB)) */ + +/* orangefs xattr and acl related defines */ +#define ORANGEFS_XATTR_INDEX_POSIX_ACL_ACCESS 1 +#define ORANGEFS_XATTR_INDEX_POSIX_ACL_DEFAULT 2 +#define ORANGEFS_XATTR_INDEX_TRUSTED 3 +#define ORANGEFS_XATTR_INDEX_DEFAULT 4 + +#define ORANGEFS_XATTR_NAME_ACL_ACCESS XATTR_NAME_POSIX_ACL_ACCESS +#define ORANGEFS_XATTR_NAME_ACL_DEFAULT XATTR_NAME_POSIX_ACL_DEFAULT +#define ORANGEFS_XATTR_NAME_TRUSTED_PREFIX "trusted." +#define ORANGEFS_XATTR_NAME_DEFAULT_PREFIX "" + +/* these functions are defined in orangefs-utils.c */ +int orangefs_prepare_cdm_array(char *debug_array_string); +int orangefs_prepare_debugfs_help_string(int); + +/* defined in orangefs-debugfs.c */ +int orangefs_client_debug_init(void); + +void debug_string_to_mask(char *, void *, int); +void do_c_mask(int, char *, struct client_debug_mask **); +void do_k_mask(int, char *, __u64 **); + +void debug_mask_to_string(void *, int); +void do_k_string(void *, int); +void do_c_string(void *, int); +int check_amalgam_keyword(void *, int); +int keyword_is_amalgam(char *); + +/*these variables are defined in orangefs-mod.c */ +extern char kernel_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern char client_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern char client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern unsigned int kernel_mask_set_mod_init; + +extern int orangefs_init_acl(struct inode *inode, struct inode *dir); +extern const struct xattr_handler *orangefs_xattr_handlers[]; + +extern struct posix_acl *orangefs_get_acl(struct inode *inode, int type); +extern int orangefs_set_acl(struct inode *inode, struct posix_acl *acl, int type); + +/* + * Redefine xtvec structure so that we could move helper functions out of + * the define + */ +struct xtvec { + __kernel_off_t xtv_off; /* must be off_t */ + __kernel_size_t xtv_len; /* must be size_t */ +}; + +/* + * orangefs data structures + */ +struct orangefs_kernel_op_s { + enum orangefs_vfs_op_states op_state; + __u64 tag; + + /* + * Set uses_shared_memory to non zero if this operation uses + * shared memory. If true, then a retry on the op must also + * get a new shared memory buffer and re-populate it. + * Cancels don't care - it only matters for service_operation() + * retry logics and cancels don't go through it anymore. It + * safely stays non-zero when we use it as slot_to_free. + */ + union { + int uses_shared_memory; + int slot_to_free; + }; + + struct orangefs_upcall_s upcall; + struct orangefs_downcall_s downcall; + + struct completion waitq; + spinlock_t lock; + + int attempts; + + struct list_head list; +}; + +#define set_op_state_waiting(op) ((op)->op_state = OP_VFS_STATE_WAITING) +#define set_op_state_inprogress(op) ((op)->op_state = OP_VFS_STATE_INPROGR) +#define set_op_state_given_up(op) ((op)->op_state = OP_VFS_STATE_GIVEN_UP) +static inline void set_op_state_serviced(struct orangefs_kernel_op_s *op) +{ + op->op_state = OP_VFS_STATE_SERVICED; + complete(&op->waitq); +} + +#define op_state_waiting(op) ((op)->op_state & OP_VFS_STATE_WAITING) +#define op_state_in_progress(op) ((op)->op_state & OP_VFS_STATE_INPROGR) +#define op_state_serviced(op) ((op)->op_state & OP_VFS_STATE_SERVICED) +#define op_state_purged(op) ((op)->op_state & OP_VFS_STATE_PURGED) +#define op_state_given_up(op) ((op)->op_state & OP_VFS_STATE_GIVEN_UP) +#define op_is_cancel(op) ((op)->upcall.type == ORANGEFS_VFS_OP_CANCEL) + +void op_release(struct orangefs_kernel_op_s *op); + +extern void orangefs_bufmap_put(int); +static inline void put_cancel(struct orangefs_kernel_op_s *op) +{ + orangefs_bufmap_put(op->slot_to_free); + op_release(op); +} + +static inline void set_op_state_purged(struct orangefs_kernel_op_s *op) +{ + spin_lock(&op->lock); + if (unlikely(op_is_cancel(op))) { + list_del_init(&op->list); + spin_unlock(&op->lock); + put_cancel(op); + } else { + op->op_state |= OP_VFS_STATE_PURGED; + complete(&op->waitq); + spin_unlock(&op->lock); + } +} + +/* per inode private orangefs info */ +struct orangefs_inode_s { + struct orangefs_object_kref refn; + char link_target[ORANGEFS_NAME_MAX]; + __s64 blksize; + /* + * Reading/Writing Extended attributes need to acquire the appropriate + * reader/writer semaphore on the orangefs_inode_s structure. + */ + struct rw_semaphore xattr_sem; + + struct inode vfs_inode; + sector_t last_failed_block_index_read; + + /* + * State of in-memory attributes not yet flushed to disk associated + * with this object + */ + unsigned long pinode_flags; +}; + +#define P_ATIME_FLAG 0 +#define P_MTIME_FLAG 1 +#define P_CTIME_FLAG 2 +#define P_MODE_FLAG 3 + +#define ClearAtimeFlag(pinode) clear_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) +#define SetAtimeFlag(pinode) set_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) +#define AtimeFlag(pinode) test_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) + +#define ClearMtimeFlag(pinode) clear_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) +#define SetMtimeFlag(pinode) set_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) +#define MtimeFlag(pinode) test_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) + +#define ClearCtimeFlag(pinode) clear_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) +#define SetCtimeFlag(pinode) set_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) +#define CtimeFlag(pinode) test_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) + +#define ClearModeFlag(pinode) clear_bit(P_MODE_FLAG, &(pinode)->pinode_flags) +#define SetModeFlag(pinode) set_bit(P_MODE_FLAG, &(pinode)->pinode_flags) +#define ModeFlag(pinode) test_bit(P_MODE_FLAG, &(pinode)->pinode_flags) + +/* per superblock private orangefs info */ +struct orangefs_sb_info_s { + struct orangefs_khandle root_khandle; + __s32 fs_id; + int id; + int flags; +#define ORANGEFS_OPT_INTR 0x01 +#define ORANGEFS_OPT_LOCAL_LOCK 0x02 + char devname[ORANGEFS_MAX_SERVER_ADDR_LEN]; + struct super_block *sb; + int mount_pending; + struct list_head list; +}; + +/* + * structure that holds the state of any async I/O operation issued + * through the VFS. Needed especially to handle cancellation requests + * or even completion notification so that the VFS client-side daemon + * can free up its vfs_request slots. + */ +struct orangefs_kiocb_s { + /* the pointer to the task that initiated the AIO */ + struct task_struct *tsk; + + /* pointer to the kiocb that kicked this operation */ + struct kiocb *kiocb; + + /* buffer index that was used for the I/O */ + struct orangefs_bufmap *bufmap; + int buffer_index; + + /* orangefs kernel operation type */ + struct orangefs_kernel_op_s *op; + + /* The user space buffers from/to which I/O is being staged */ + struct iovec *iov; + + /* number of elements in the iovector */ + unsigned long nr_segs; + + /* set to indicate the type of the operation */ + int rw; + + /* file offset */ + loff_t offset; + + /* and the count in bytes */ + size_t bytes_to_be_copied; + + ssize_t bytes_copied; + int needs_cleanup; +}; + +struct orangefs_stats { + unsigned long cache_hits; + unsigned long cache_misses; + unsigned long reads; + unsigned long writes; +}; + +extern struct orangefs_stats g_orangefs_stats; + +/* + * NOTE: See Documentation/filesystems/porting for information + * on implementing FOO_I and properly accessing fs private data + */ +static inline struct orangefs_inode_s *ORANGEFS_I(struct inode *inode) +{ + return container_of(inode, struct orangefs_inode_s, vfs_inode); +} + +static inline struct orangefs_sb_info_s *ORANGEFS_SB(struct super_block *sb) +{ + return (struct orangefs_sb_info_s *) sb->s_fs_info; +} + +/* ino_t descends from "unsigned long", 8 bytes, 64 bits. */ +static inline ino_t orangefs_khandle_to_ino(struct orangefs_khandle *khandle) +{ + union { + unsigned char u[8]; + __u64 ino; + } ihandle; + + ihandle.u[0] = khandle->u[0] ^ khandle->u[4]; + ihandle.u[1] = khandle->u[1] ^ khandle->u[5]; + ihandle.u[2] = khandle->u[2] ^ khandle->u[6]; + ihandle.u[3] = khandle->u[3] ^ khandle->u[7]; + ihandle.u[4] = khandle->u[12] ^ khandle->u[8]; + ihandle.u[5] = khandle->u[13] ^ khandle->u[9]; + ihandle.u[6] = khandle->u[14] ^ khandle->u[10]; + ihandle.u[7] = khandle->u[15] ^ khandle->u[11]; + + return ihandle.ino; +} + +static inline struct orangefs_khandle *get_khandle_from_ino(struct inode *inode) +{ + return &(ORANGEFS_I(inode)->refn.khandle); +} + +static inline __s32 get_fsid_from_ino(struct inode *inode) +{ + return ORANGEFS_I(inode)->refn.fs_id; +} + +static inline ino_t get_ino_from_khandle(struct inode *inode) +{ + struct orangefs_khandle *khandle; + ino_t ino; + + khandle = get_khandle_from_ino(inode); + ino = orangefs_khandle_to_ino(khandle); + return ino; +} + +static inline ino_t get_parent_ino_from_dentry(struct dentry *dentry) +{ + return get_ino_from_khandle(dentry->d_parent->d_inode); +} + +static inline int is_root_handle(struct inode *inode) +{ + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: root handle: %pU, this handle: %pU:\n", + __func__, + &ORANGEFS_SB(inode->i_sb)->root_khandle, + get_khandle_from_ino(inode)); + + if (ORANGEFS_khandle_cmp(&(ORANGEFS_SB(inode->i_sb)->root_khandle), + get_khandle_from_ino(inode))) + return 0; + else + return 1; +} + +static inline int match_handle(struct orangefs_khandle resp_handle, + struct inode *inode) +{ + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: one handle: %pU, another handle:%pU:\n", + __func__, + &resp_handle, + get_khandle_from_ino(inode)); + + if (ORANGEFS_khandle_cmp(&resp_handle, get_khandle_from_ino(inode))) + return 0; + else + return 1; +} + +/* + * defined in orangefs-cache.c + */ +int op_cache_initialize(void); +int op_cache_finalize(void); +struct orangefs_kernel_op_s *op_alloc(__s32 type); +void orangefs_new_tag(struct orangefs_kernel_op_s *op); +char *get_opname_string(struct orangefs_kernel_op_s *new_op); + +int orangefs_inode_cache_initialize(void); +int orangefs_inode_cache_finalize(void); + +/* + * defined in orangefs-mod.c + */ +void purge_inprogress_ops(void); + +/* + * defined in waitqueue.c + */ +void purge_waiting_ops(void); + +/* + * defined in super.c + */ +struct dentry *orangefs_mount(struct file_system_type *fst, + int flags, + const char *devname, + void *data); + +void orangefs_kill_sb(struct super_block *sb); +int orangefs_remount(struct orangefs_sb_info_s *); + +int fsid_key_table_initialize(void); +void fsid_key_table_finalize(void); + +/* + * defined in inode.c + */ +__u32 convert_to_orangefs_mask(unsigned long lite_mask); +struct inode *orangefs_new_inode(struct super_block *sb, + struct inode *dir, + int mode, + dev_t dev, + struct orangefs_object_kref *ref); + +int orangefs_setattr(struct dentry *dentry, struct iattr *iattr); + +int orangefs_getattr(struct vfsmount *mnt, + struct dentry *dentry, + struct kstat *kstat); + +int orangefs_permission(struct inode *inode, int mask); + +/* + * defined in xattr.c + */ +int orangefs_setxattr(struct dentry *dentry, + const char *name, + const void *value, + size_t size, + int flags); + +ssize_t orangefs_getxattr(struct dentry *dentry, + const char *name, + void *buffer, + size_t size); + +ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size); + +/* + * defined in namei.c + */ +struct inode *orangefs_iget(struct super_block *sb, + struct orangefs_object_kref *ref); + +ssize_t orangefs_inode_read(struct inode *inode, + struct iov_iter *iter, + loff_t *offset, + loff_t readahead_size); + +/* + * defined in devorangefs-req.c + */ +int orangefs_dev_init(void); +void orangefs_dev_cleanup(void); +int is_daemon_in_service(void); +bool __is_daemon_in_service(void); + +/* + * defined in orangefs-utils.c + */ +__s32 fsid_of_op(struct orangefs_kernel_op_s *op); + +int orangefs_flush_inode(struct inode *inode); + +ssize_t orangefs_inode_getxattr(struct inode *inode, + const char *prefix, + const char *name, + void *buffer, + size_t size); + +int orangefs_inode_setxattr(struct inode *inode, + const char *prefix, + const char *name, + const void *value, + size_t size, + int flags); + +int orangefs_inode_getattr(struct inode *inode, int new, int size); + +int orangefs_inode_check_changed(struct inode *inode); + +int orangefs_inode_setattr(struct inode *inode, struct iattr *iattr); + +void orangefs_make_bad_inode(struct inode *inode); + +int orangefs_unmount_sb(struct super_block *sb); + +bool orangefs_cancel_op_in_progress(struct orangefs_kernel_op_s *op); + +int orangefs_normalize_to_errno(__s32 error_code); + +extern struct mutex devreq_mutex; +extern struct mutex request_mutex; +extern int debug; +extern int op_timeout_secs; +extern int slot_timeout_secs; +extern struct list_head orangefs_superblocks; +extern spinlock_t orangefs_superblocks_lock; +extern struct list_head orangefs_request_list; +extern spinlock_t orangefs_request_list_lock; +extern wait_queue_head_t orangefs_request_list_waitq; +extern struct list_head *htable_ops_in_progress; +extern spinlock_t htable_ops_in_progress_lock; +extern int hash_table_size; + +extern const struct address_space_operations orangefs_address_operations; +extern struct backing_dev_info orangefs_backing_dev_info; +extern struct inode_operations orangefs_file_inode_operations; +extern const struct file_operations orangefs_file_operations; +extern struct inode_operations orangefs_symlink_inode_operations; +extern struct inode_operations orangefs_dir_inode_operations; +extern const struct file_operations orangefs_dir_operations; +extern const struct dentry_operations orangefs_dentry_operations; +extern const struct file_operations orangefs_devreq_file_operations; + +extern wait_queue_head_t orangefs_bufmap_init_waitq; + +/* + * misc convenience macros + */ + +#define ORANGEFS_OP_INTERRUPTIBLE 1 /* service_operation() is interruptible */ +#define ORANGEFS_OP_PRIORITY 2 /* service_operation() is high priority */ +#define ORANGEFS_OP_CANCELLATION 4 /* this is a cancellation */ +#define ORANGEFS_OP_NO_MUTEX 8 /* don't acquire request_mutex */ +#define ORANGEFS_OP_ASYNC 16 /* Queue it, but don't wait */ + +int service_operation(struct orangefs_kernel_op_s *op, + const char *op_name, + int flags); + +#define get_interruptible_flag(inode) \ + ((ORANGEFS_SB(inode->i_sb)->flags & ORANGEFS_OPT_INTR) ? \ + ORANGEFS_OP_INTERRUPTIBLE : 0) + +#define fill_default_sys_attrs(sys_attr, type, mode) \ +do { \ + sys_attr.owner = from_kuid(current_user_ns(), current_fsuid()); \ + sys_attr.group = from_kgid(current_user_ns(), current_fsgid()); \ + sys_attr.perms = ORANGEFS_util_translate_mode(mode); \ + sys_attr.mtime = 0; \ + sys_attr.atime = 0; \ + sys_attr.ctime = 0; \ + sys_attr.mask = ORANGEFS_ATTR_SYS_ALL_SETABLE; \ +} while (0) + +static inline void orangefs_i_size_write(struct inode *inode, loff_t i_size) +{ +#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) + mutex_lock(&inode->i_mutex); +#endif + i_size_write(inode, i_size); +#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) + mutex_unlock(&inode->i_mutex); +#endif +} + +#endif /* __ORANGEFSKERNEL_H */ diff --git a/fs/orangefs/orangefs-mod.c b/fs/orangefs/orangefs-mod.c new file mode 100644 index 000000000000..6f072a8c0de1 --- /dev/null +++ b/fs/orangefs/orangefs-mod.c @@ -0,0 +1,293 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * Changes by Acxiom Corporation to add proc file handler for pvfs2 client + * parameters, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-debugfs.h" +#include "orangefs-sysfs.h" + +/* ORANGEFS_VERSION is a ./configure define */ +#ifndef ORANGEFS_VERSION +#define ORANGEFS_VERSION "upstream" +#endif + +/* + * global variables declared here + */ + +/* array of client debug keyword/mask values */ +struct client_debug_mask *cdm_array; +int cdm_element_count; + +char kernel_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN] = "none"; +char client_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +char client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; + +char *debug_help_string; +int help_string_initialized; +struct dentry *help_file_dentry; +struct dentry *client_debug_dentry; +struct dentry *debug_dir; +int client_verbose_index; +int client_all_index; +struct orangefs_stats g_orangefs_stats; + +/* the size of the hash tables for ops in progress */ +int hash_table_size = 509; + +static ulong module_parm_debug_mask; +__u64 gossip_debug_mask; +struct client_debug_mask client_debug_mask = { NULL, 0, 0 }; +unsigned int kernel_mask_set_mod_init; /* implicitly false */ +int op_timeout_secs = ORANGEFS_DEFAULT_OP_TIMEOUT_SECS; +int slot_timeout_secs = ORANGEFS_DEFAULT_SLOT_TIMEOUT_SECS; + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("ORANGEFS Development Team"); +MODULE_DESCRIPTION("The Linux Kernel VFS interface to ORANGEFS"); +MODULE_PARM_DESC(module_parm_debug_mask, "debugging level (see orangefs-debug.h for values)"); +MODULE_PARM_DESC(op_timeout_secs, "Operation timeout in seconds"); +MODULE_PARM_DESC(slot_timeout_secs, "Slot timeout in seconds"); +MODULE_PARM_DESC(hash_table_size, + "size of hash table for operations in progress"); + +static struct file_system_type orangefs_fs_type = { + .name = "pvfs2", + .mount = orangefs_mount, + .kill_sb = orangefs_kill_sb, + .owner = THIS_MODULE, +}; + +module_param(hash_table_size, int, 0); +module_param(module_parm_debug_mask, ulong, 0644); +module_param(op_timeout_secs, int, 0); +module_param(slot_timeout_secs, int, 0); + +/* synchronizes the request device file */ +DEFINE_MUTEX(devreq_mutex); + +/* + * Blocks non-priority requests from being queued for servicing. This + * could be used for protecting the request list data structure, but + * for now it's only being used to stall the op addition to the request + * list + */ +DEFINE_MUTEX(request_mutex); + +/* hash table for storing operations waiting for matching downcall */ +struct list_head *htable_ops_in_progress; +DEFINE_SPINLOCK(htable_ops_in_progress_lock); + +/* list for queueing upcall operations */ +LIST_HEAD(orangefs_request_list); + +/* used to protect the above orangefs_request_list */ +DEFINE_SPINLOCK(orangefs_request_list_lock); + +/* used for incoming request notification */ +DECLARE_WAIT_QUEUE_HEAD(orangefs_request_list_waitq); + +static int __init orangefs_init(void) +{ + int ret = -1; + __u32 i = 0; + + /* convert input debug mask to a 64-bit unsigned integer */ + gossip_debug_mask = (unsigned long long) module_parm_debug_mask; + + /* + * set the kernel's gossip debug string; invalid mask values will + * be ignored. + */ + debug_mask_to_string(&gossip_debug_mask, 0); + + /* remove any invalid values from the mask */ + debug_string_to_mask(kernel_debug_string, &gossip_debug_mask, 0); + + /* + * if the mask has a non-zero value, then indicate that the mask + * was set when the kernel module was loaded. The orangefs dev ioctl + * command will look at this boolean to determine if the kernel's + * debug mask should be overwritten when the client-core is started. + */ + if (gossip_debug_mask != 0) + kernel_mask_set_mod_init = true; + + pr_info("%s: called with debug mask: :%s: :%llx:\n", + __func__, + kernel_debug_string, + (unsigned long long)gossip_debug_mask); + + ret = bdi_init(&orangefs_backing_dev_info); + + if (ret) + return ret; + + if (op_timeout_secs < 0) + op_timeout_secs = 0; + + if (slot_timeout_secs < 0) + slot_timeout_secs = 0; + + /* initialize global book keeping data structures */ + ret = op_cache_initialize(); + if (ret < 0) + goto err; + + ret = orangefs_inode_cache_initialize(); + if (ret < 0) + goto cleanup_op; + + htable_ops_in_progress = + kcalloc(hash_table_size, sizeof(struct list_head), GFP_KERNEL); + if (!htable_ops_in_progress) { + gossip_err("Failed to initialize op hashtable"); + ret = -ENOMEM; + goto cleanup_inode; + } + + /* initialize a doubly linked at each hash table index */ + for (i = 0; i < hash_table_size; i++) + INIT_LIST_HEAD(&htable_ops_in_progress[i]); + + ret = fsid_key_table_initialize(); + if (ret < 0) + goto cleanup_progress_table; + + /* + * Build the contents of /sys/kernel/debug/orangefs/debug-help + * from the keywords in the kernel keyword/mask array. + * + * The keywords in the client keyword/mask array are + * unknown at boot time. + * + * orangefs_prepare_debugfs_help_string will be used again + * later to rebuild the debug-help file after the client starts + * and passes along the needed info. The argument signifies + * which time orangefs_prepare_debugfs_help_string is being + * called. + */ + ret = orangefs_prepare_debugfs_help_string(1); + if (ret) + goto cleanup_key_table; + + ret = orangefs_debugfs_init(); + if (ret) + goto debugfs_init_failed; + + ret = orangefs_kernel_debug_init(); + if (ret) + goto kernel_debug_init_failed; + + ret = orangefs_sysfs_init(); + if (ret) + goto sysfs_init_failed; + + /* Initialize the orangefsdev subsystem. */ + ret = orangefs_dev_init(); + if (ret < 0) { + gossip_err("%s: could not initialize device subsystem %d!\n", + __func__, + ret); + goto cleanup_device; + } + + ret = register_filesystem(&orangefs_fs_type); + if (ret == 0) { + pr_info("orangefs: module version %s loaded\n", ORANGEFS_VERSION); + ret = 0; + goto out; + } + + orangefs_sysfs_exit(); + +cleanup_device: + orangefs_dev_cleanup(); + +sysfs_init_failed: + +kernel_debug_init_failed: + +debugfs_init_failed: + orangefs_debugfs_cleanup(); + +cleanup_key_table: + fsid_key_table_finalize(); + +cleanup_progress_table: + kfree(htable_ops_in_progress); + +cleanup_inode: + orangefs_inode_cache_finalize(); + +cleanup_op: + op_cache_finalize(); + +err: + bdi_destroy(&orangefs_backing_dev_info); + +out: + return ret; +} + +static void __exit orangefs_exit(void) +{ + int i = 0; + gossip_debug(GOSSIP_INIT_DEBUG, "orangefs: orangefs_exit called\n"); + + unregister_filesystem(&orangefs_fs_type); + orangefs_debugfs_cleanup(); + orangefs_sysfs_exit(); + fsid_key_table_finalize(); + orangefs_dev_cleanup(); + BUG_ON(!list_empty(&orangefs_request_list)); + for (i = 0; i < hash_table_size; i++) + BUG_ON(!list_empty(&htable_ops_in_progress[i])); + + orangefs_inode_cache_finalize(); + op_cache_finalize(); + + kfree(htable_ops_in_progress); + + bdi_destroy(&orangefs_backing_dev_info); + + pr_info("orangefs: module version %s unloaded\n", ORANGEFS_VERSION); +} + +/* + * What we do in this function is to walk the list of operations + * that are in progress in the hash table and mark them as purged as well. + */ +void purge_inprogress_ops(void) +{ + int i; + + for (i = 0; i < hash_table_size; i++) { + struct orangefs_kernel_op_s *op; + struct orangefs_kernel_op_s *next; + + spin_lock(&htable_ops_in_progress_lock); + list_for_each_entry_safe(op, + next, + &htable_ops_in_progress[i], + list) { + set_op_state_purged(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + } + spin_unlock(&htable_ops_in_progress_lock); + } +} + +module_init(orangefs_init); +module_exit(orangefs_exit); diff --git a/fs/orangefs/orangefs-sysfs.c b/fs/orangefs/orangefs-sysfs.c new file mode 100644 index 000000000000..5c03113e3ad2 --- /dev/null +++ b/fs/orangefs/orangefs-sysfs.c @@ -0,0 +1,1772 @@ +/* + * Documentation/ABI/stable/orangefs-sysfs: + * + * What: /sys/fs/orangefs/perf_counter_reset + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * echo a 0 or a 1 into perf_counter_reset to + * reset all the counters in + * /sys/fs/orangefs/perf_counters + * except ones with PINT_PERF_PRESERVE set. + * + * + * What: /sys/fs/orangefs/perf_counters/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Counters and settings for various caches. + * Read only. + * + * + * What: /sys/fs/orangefs/perf_time_interval_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Length of perf counter intervals in + * seconds. + * + * + * What: /sys/fs/orangefs/perf_history_size + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * The perf_counters cache statistics have N, or + * perf_history_size, samples. The default is + * one. + * + * Every perf_time_interval_secs the (first) + * samples are reset. + * + * If N is greater than one, the "current" set + * of samples is reset, and the samples from the + * other N-1 intervals remain available. + * + * + * What: /sys/fs/orangefs/op_timeout_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Service operation timeout in seconds. + * + * + * What: /sys/fs/orangefs/slot_timeout_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * "Slot" timeout in seconds. A "slot" + * is an indexed buffer in the shared + * memory segment used for communication + * between the kernel module and userspace. + * Slots are requested and waited for, + * the wait times out after slot_timeout_secs. + * + * + * What: /sys/fs/orangefs/acache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Attribute cache configurable settings. + * + * + * What: /sys/fs/orangefs/ncache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Name cache configurable settings. + * + * + * What: /sys/fs/orangefs/capcache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Capability cache configurable settings. + * + * + * What: /sys/fs/orangefs/ccache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Credential cache configurable settings. + * + */ + +#include <linux/fs.h> +#include <linux/kobject.h> +#include <linux/string.h> +#include <linux/sysfs.h> +#include <linux/module.h> +#include <linux/init.h> + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-sysfs.h" + +#define ORANGEFS_KOBJ_ID "orangefs" +#define ACACHE_KOBJ_ID "acache" +#define CAPCACHE_KOBJ_ID "capcache" +#define CCACHE_KOBJ_ID "ccache" +#define NCACHE_KOBJ_ID "ncache" +#define PC_KOBJ_ID "pc" +#define STATS_KOBJ_ID "stats" + +struct orangefs_obj { + struct kobject kobj; + int op_timeout_secs; + int perf_counter_reset; + int perf_history_size; + int perf_time_interval_secs; + int slot_timeout_secs; +}; + +struct acache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_msecs; +}; + +struct capcache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_secs; +}; + +struct ccache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_secs; +}; + +struct ncache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_msecs; +}; + +struct pc_orangefs_obj { + struct kobject kobj; + char *acache; + char *capcache; + char *ncache; +}; + +struct stats_orangefs_obj { + struct kobject kobj; + int reads; + int writes; +}; + +struct orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct acache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct capcache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct capcache_orangefs_obj *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct capcache_orangefs_obj *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct ccache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct ccache_orangefs_obj *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct ccache_orangefs_obj *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct ncache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct pc_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct stats_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +static ssize_t orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct orangefs_attribute *attribute; + struct orangefs_obj *orangefs_obj; + int rc; + + attribute = container_of(attr, struct orangefs_attribute, attr); + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct orangefs_attribute *attribute; + struct orangefs_obj *orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct orangefs_attribute, attr); + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops orangefs_sysfs_ops = { + .show = orangefs_attr_show, + .store = orangefs_attr_store, +}; + +static ssize_t acache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct acache_orangefs_attribute *attribute; + struct acache_orangefs_obj *acache_orangefs_obj; + int rc; + + attribute = container_of(attr, struct acache_orangefs_attribute, attr); + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(acache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t acache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct acache_orangefs_attribute *attribute; + struct acache_orangefs_obj *acache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "acache_orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct acache_orangefs_attribute, attr); + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(acache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops acache_orangefs_sysfs_ops = { + .show = acache_orangefs_attr_show, + .store = acache_orangefs_attr_store, +}; + +static ssize_t capcache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct capcache_orangefs_attribute *attribute; + struct capcache_orangefs_obj *capcache_orangefs_obj; + int rc; + + attribute = + container_of(attr, struct capcache_orangefs_attribute, attr); + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(capcache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t capcache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct capcache_orangefs_attribute *attribute; + struct capcache_orangefs_obj *capcache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "capcache_orangefs_attr_store: start\n"); + + attribute = + container_of(attr, struct capcache_orangefs_attribute, attr); + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(capcache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops capcache_orangefs_sysfs_ops = { + .show = capcache_orangefs_attr_show, + .store = capcache_orangefs_attr_store, +}; + +static ssize_t ccache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct ccache_orangefs_attribute *attribute; + struct ccache_orangefs_obj *ccache_orangefs_obj; + int rc; + + attribute = + container_of(attr, struct ccache_orangefs_attribute, attr); + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(ccache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t ccache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct ccache_orangefs_attribute *attribute; + struct ccache_orangefs_obj *ccache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "ccache_orangefs_attr_store: start\n"); + + attribute = + container_of(attr, struct ccache_orangefs_attribute, attr); + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(ccache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops ccache_orangefs_sysfs_ops = { + .show = ccache_orangefs_attr_show, + .store = ccache_orangefs_attr_store, +}; + +static ssize_t ncache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct ncache_orangefs_attribute *attribute; + struct ncache_orangefs_obj *ncache_orangefs_obj; + int rc; + + attribute = container_of(attr, struct ncache_orangefs_attribute, attr); + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(ncache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t ncache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct ncache_orangefs_attribute *attribute; + struct ncache_orangefs_obj *ncache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "ncache_orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct ncache_orangefs_attribute, attr); + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(ncache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops ncache_orangefs_sysfs_ops = { + .show = ncache_orangefs_attr_show, + .store = ncache_orangefs_attr_store, +}; + +static ssize_t pc_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct pc_orangefs_attribute *attribute; + struct pc_orangefs_obj *pc_orangefs_obj; + int rc; + + attribute = container_of(attr, struct pc_orangefs_attribute, attr); + pc_orangefs_obj = + container_of(kobj, struct pc_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(pc_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static const struct sysfs_ops pc_orangefs_sysfs_ops = { + .show = pc_orangefs_attr_show, +}; + +static ssize_t stats_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct stats_orangefs_attribute *attribute; + struct stats_orangefs_obj *stats_orangefs_obj; + int rc; + + attribute = container_of(attr, struct stats_orangefs_attribute, attr); + stats_orangefs_obj = + container_of(kobj, struct stats_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(stats_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static const struct sysfs_ops stats_orangefs_sysfs_ops = { + .show = stats_orangefs_attr_show, +}; + +static void orangefs_release(struct kobject *kobj) +{ + struct orangefs_obj *orangefs_obj; + + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + kfree(orangefs_obj); +} + +static void acache_orangefs_release(struct kobject *kobj) +{ + struct acache_orangefs_obj *acache_orangefs_obj; + + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + kfree(acache_orangefs_obj); +} + +static void capcache_orangefs_release(struct kobject *kobj) +{ + struct capcache_orangefs_obj *capcache_orangefs_obj; + + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + kfree(capcache_orangefs_obj); +} + +static void ccache_orangefs_release(struct kobject *kobj) +{ + struct ccache_orangefs_obj *ccache_orangefs_obj; + + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + kfree(ccache_orangefs_obj); +} + +static void ncache_orangefs_release(struct kobject *kobj) +{ + struct ncache_orangefs_obj *ncache_orangefs_obj; + + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + kfree(ncache_orangefs_obj); +} + +static void pc_orangefs_release(struct kobject *kobj) +{ + struct pc_orangefs_obj *pc_orangefs_obj; + + pc_orangefs_obj = + container_of(kobj, struct pc_orangefs_obj, kobj); + kfree(pc_orangefs_obj); +} + +static void stats_orangefs_release(struct kobject *kobj) +{ + struct stats_orangefs_obj *stats_orangefs_obj; + + stats_orangefs_obj = + container_of(kobj, struct stats_orangefs_obj, kobj); + kfree(stats_orangefs_obj); +} + +static ssize_t sysfs_int_show(char *kobj_id, char *buf, void *attr) +{ + int rc = -EIO; + struct orangefs_attribute *orangefs_attr; + struct stats_orangefs_attribute *stats_orangefs_attr; + + gossip_debug(GOSSIP_SYSFS_DEBUG, "sysfs_int_show: id:%s:\n", kobj_id); + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "op_timeout_secs")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + op_timeout_secs); + goto out; + } else if (!strcmp(orangefs_attr->attr.name, + "slot_timeout_secs")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + slot_timeout_secs); + goto out; + } else { + goto out; + } + + } else if (!strcmp(kobj_id, STATS_KOBJ_ID)) { + stats_orangefs_attr = (struct stats_orangefs_attribute *)attr; + + if (!strcmp(stats_orangefs_attr->attr.name, "reads")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%lu\n", + g_orangefs_stats.reads); + goto out; + } else if (!strcmp(stats_orangefs_attr->attr.name, "writes")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%lu\n", + g_orangefs_stats.writes); + goto out; + } else { + goto out; + } + } + +out: + + return rc; +} + +static ssize_t int_orangefs_show(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf) +{ + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_orangefs_show:start attr->attr.name:%s:\n", + attr->attr.name); + + rc = sysfs_int_show(ORANGEFS_KOBJ_ID, buf, (void *) attr); + + return rc; +} + +static ssize_t int_stats_show(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + char *buf) +{ + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_stats_show:start attr->attr.name:%s:\n", + attr->attr.name); + + rc = sysfs_int_show(STATS_KOBJ_ID, buf, (void *) attr); + + return rc; +} + +static ssize_t int_store(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_store: start attr->attr.name:%s: buf:%s:\n", + attr->attr.name, buf); + + if (!strcmp(attr->attr.name, "op_timeout_secs")) { + rc = kstrtoint(buf, 0, &op_timeout_secs); + goto out; + } else if (!strcmp(attr->attr.name, "slot_timeout_secs")) { + rc = kstrtoint(buf, 0, &slot_timeout_secs); + goto out; + } else { + goto out; + } + +out: + if (rc) + rc = -EINVAL; + else + rc = count; + + return rc; +} + +/* + * obtain attribute values from userspace with a service operation. + */ +static int sysfs_service_op_show(char *kobj_id, char *buf, void *attr) +{ + struct orangefs_kernel_op_s *new_op = NULL; + int rc = 0; + char *ser_op_type = NULL; + struct orangefs_attribute *orangefs_attr; + struct acache_orangefs_attribute *acache_attr; + struct capcache_orangefs_attribute *capcache_attr; + struct ccache_orangefs_attribute *ccache_attr; + struct ncache_orangefs_attribute *ncache_attr; + struct pc_orangefs_attribute *pc_attr; + __u32 op_alloc_type; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "sysfs_service_op_show: id:%s:\n", + kobj_id); + + if (strcmp(kobj_id, PC_KOBJ_ID)) + op_alloc_type = ORANGEFS_VFS_OP_PARAM; + else + op_alloc_type = ORANGEFS_VFS_OP_PERF_COUNT; + + new_op = op_alloc(op_alloc_type); + if (!new_op) + return -ENOMEM; + + /* Can't do a service_operation if the client is not running... */ + rc = is_daemon_in_service(); + if (rc) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + if (strcmp(kobj_id, PC_KOBJ_ID)) + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_GET; + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "perf_history_size")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE; + else if (!strcmp(orangefs_attr->attr.name, + "perf_time_interval_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS; + else if (!strcmp(orangefs_attr->attr.name, + "perf_counter_reset")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET; + + } else if (!strcmp(kobj_id, ACACHE_KOBJ_ID)) { + acache_attr = (struct acache_orangefs_attribute *)attr; + + if (!strcmp(acache_attr->attr.name, "timeout_msecs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS; + + if (!strcmp(acache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT; + + if (!strcmp(acache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT; + + if (!strcmp(acache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, CAPCACHE_KOBJ_ID)) { + capcache_attr = (struct capcache_orangefs_attribute *)attr; + + if (!strcmp(capcache_attr->attr.name, "timeout_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS; + + if (!strcmp(capcache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT; + + if (!strcmp(capcache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT; + + if (!strcmp(capcache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, CCACHE_KOBJ_ID)) { + ccache_attr = (struct ccache_orangefs_attribute *)attr; + + if (!strcmp(ccache_attr->attr.name, "timeout_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS; + + if (!strcmp(ccache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT; + + if (!strcmp(ccache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT; + + if (!strcmp(ccache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, NCACHE_KOBJ_ID)) { + ncache_attr = (struct ncache_orangefs_attribute *)attr; + + if (!strcmp(ncache_attr->attr.name, "timeout_msecs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS; + + if (!strcmp(ncache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT; + + if (!strcmp(ncache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT; + + if (!strcmp(ncache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, PC_KOBJ_ID)) { + pc_attr = (struct pc_orangefs_attribute *)attr; + + if (!strcmp(pc_attr->attr.name, ACACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_ACACHE; + + if (!strcmp(pc_attr->attr.name, CAPCACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_CAPCACHE; + + if (!strcmp(pc_attr->attr.name, NCACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_NCACHE; + + } else { + gossip_err("sysfs_service_op_show: unknown kobj_id:%s:\n", + kobj_id); + rc = -EINVAL; + goto out; + } + + + if (strcmp(kobj_id, PC_KOBJ_ID)) + ser_op_type = "orangefs_param"; + else + ser_op_type = "orangefs_perf_count"; + + /* + * The service_operation will return an errno return code on + * error, and zero on success. + */ + rc = service_operation(new_op, ser_op_type, ORANGEFS_OP_INTERRUPTIBLE); + +out: + if (!rc) { + if (strcmp(kobj_id, PC_KOBJ_ID)) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + (int)new_op->downcall.resp.param.value); + } else { + rc = scnprintf( + buf, + PAGE_SIZE, + "%s", + new_op->downcall.resp.perf_count.buffer); + } + } + + op_release(new_op); + + return rc; + +} + +static ssize_t service_orangefs_show(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(ORANGEFS_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_acache_show(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(ACACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t service_capcache_show(struct capcache_orangefs_obj + *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(CAPCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t service_ccache_show(struct ccache_orangefs_obj + *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(CCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_ncache_show(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(NCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_pc_show(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(PC_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +/* + * pass attribute values back to userspace with a service operation. + * + * We have to do a memory allocation, an sscanf and a service operation. + * And we have to evaluate what the user entered, to make sure the + * value is within the range supported by the attribute. So, there's + * a lot of return code checking and mapping going on here. + * + * We want to return 1 if we think everything went OK, and + * EINVAL if not. + */ +static int sysfs_service_op_store(char *kobj_id, const char *buf, void *attr) +{ + struct orangefs_kernel_op_s *new_op = NULL; + int val = 0; + int rc = 0; + struct orangefs_attribute *orangefs_attr; + struct acache_orangefs_attribute *acache_attr; + struct capcache_orangefs_attribute *capcache_attr; + struct ccache_orangefs_attribute *ccache_attr; + struct ncache_orangefs_attribute *ncache_attr; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "sysfs_service_op_store: id:%s:\n", + kobj_id); + + new_op = op_alloc(ORANGEFS_VFS_OP_PARAM); + if (!new_op) + return -EINVAL; /* sic */ + + /* Can't do a service_operation if the client is not running... */ + rc = is_daemon_in_service(); + if (rc) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + /* + * The value we want to send back to userspace is in buf. + */ + rc = kstrtoint(buf, 0, &val); + if (rc) + goto out; + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "perf_history_size")) { + if (val > 0) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(orangefs_attr->attr.name, + "perf_time_interval_secs")) { + if (val > 0) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(orangefs_attr->attr.name, + "perf_counter_reset")) { + if ((val == 0) || (val == 1)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, ACACHE_KOBJ_ID)) { + acache_attr = (struct acache_orangefs_attribute *)attr; + + if (!strcmp(acache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, "timeout_msecs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, CAPCACHE_KOBJ_ID)) { + capcache_attr = (struct capcache_orangefs_attribute *)attr; + + if (!strcmp(capcache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, "timeout_secs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, CCACHE_KOBJ_ID)) { + ccache_attr = (struct ccache_orangefs_attribute *)attr; + + if (!strcmp(ccache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, "timeout_secs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, NCACHE_KOBJ_ID)) { + ncache_attr = (struct ncache_orangefs_attribute *)attr; + + if (!strcmp(ncache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, "timeout_msecs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS; + } else { + rc = 0; + goto out; + } + } + + } else { + gossip_err("sysfs_service_op_store: unknown kobj_id:%s:\n", + kobj_id); + rc = -EINVAL; + goto out; + } + + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_SET; + + new_op->upcall.req.param.value = val; + + /* + * The service_operation will return a errno return code on + * error, and zero on success. + */ + rc = service_operation(new_op, "orangefs_param", ORANGEFS_OP_INTERRUPTIBLE); + + if (rc < 0) { + gossip_err("sysfs_service_op_store: service op returned:%d:\n", + rc); + rc = 0; + } else { + rc = 1; + } + +out: + op_release(new_op); + + if (rc == -ENOMEM || rc == 0) + rc = -EINVAL; + + return rc; +} + +static ssize_t + service_orangefs_store(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(ORANGEFS_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_acache_store(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(ACACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_capcache_store(struct capcache_orangefs_obj + *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(CAPCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t service_ccache_store(struct ccache_orangefs_obj + *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(CCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_ncache_store(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(NCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static struct orangefs_attribute op_timeout_secs_attribute = + __ATTR(op_timeout_secs, 0664, int_orangefs_show, int_store); + +static struct orangefs_attribute slot_timeout_secs_attribute = + __ATTR(slot_timeout_secs, 0664, int_orangefs_show, int_store); + +static struct orangefs_attribute perf_counter_reset_attribute = + __ATTR(perf_counter_reset, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct orangefs_attribute perf_history_size_attribute = + __ATTR(perf_history_size, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct orangefs_attribute perf_time_interval_secs_attribute = + __ATTR(perf_time_interval_secs, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct attribute *orangefs_default_attrs[] = { + &op_timeout_secs_attribute.attr, + &slot_timeout_secs_attribute.attr, + &perf_counter_reset_attribute.attr, + &perf_history_size_attribute.attr, + &perf_time_interval_secs_attribute.attr, + NULL, +}; + +static struct kobj_type orangefs_ktype = { + .sysfs_ops = &orangefs_sysfs_ops, + .release = orangefs_release, + .default_attrs = orangefs_default_attrs, +}; + +static struct acache_orangefs_attribute acache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_timeout_msecs_attribute = + __ATTR(timeout_msecs, + 0664, + service_acache_show, + service_acache_store); + +static struct attribute *acache_orangefs_default_attrs[] = { + &acache_hard_limit_attribute.attr, + &acache_reclaim_percent_attribute.attr, + &acache_soft_limit_attribute.attr, + &acache_timeout_msecs_attribute.attr, + NULL, +}; + +static struct kobj_type acache_orangefs_ktype = { + .sysfs_ops = &acache_orangefs_sysfs_ops, + .release = acache_orangefs_release, + .default_attrs = acache_orangefs_default_attrs, +}; + +static struct capcache_orangefs_attribute capcache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_timeout_secs_attribute = + __ATTR(timeout_secs, + 0664, + service_capcache_show, + service_capcache_store); + +static struct attribute *capcache_orangefs_default_attrs[] = { + &capcache_hard_limit_attribute.attr, + &capcache_reclaim_percent_attribute.attr, + &capcache_soft_limit_attribute.attr, + &capcache_timeout_secs_attribute.attr, + NULL, +}; + +static struct kobj_type capcache_orangefs_ktype = { + .sysfs_ops = &capcache_orangefs_sysfs_ops, + .release = capcache_orangefs_release, + .default_attrs = capcache_orangefs_default_attrs, +}; + +static struct ccache_orangefs_attribute ccache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_timeout_secs_attribute = + __ATTR(timeout_secs, + 0664, + service_ccache_show, + service_ccache_store); + +static struct attribute *ccache_orangefs_default_attrs[] = { + &ccache_hard_limit_attribute.attr, + &ccache_reclaim_percent_attribute.attr, + &ccache_soft_limit_attribute.attr, + &ccache_timeout_secs_attribute.attr, + NULL, +}; + +static struct kobj_type ccache_orangefs_ktype = { + .sysfs_ops = &ccache_orangefs_sysfs_ops, + .release = ccache_orangefs_release, + .default_attrs = ccache_orangefs_default_attrs, +}; + +static struct ncache_orangefs_attribute ncache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_timeout_msecs_attribute = + __ATTR(timeout_msecs, + 0664, + service_ncache_show, + service_ncache_store); + +static struct attribute *ncache_orangefs_default_attrs[] = { + &ncache_hard_limit_attribute.attr, + &ncache_reclaim_percent_attribute.attr, + &ncache_soft_limit_attribute.attr, + &ncache_timeout_msecs_attribute.attr, + NULL, +}; + +static struct kobj_type ncache_orangefs_ktype = { + .sysfs_ops = &ncache_orangefs_sysfs_ops, + .release = ncache_orangefs_release, + .default_attrs = ncache_orangefs_default_attrs, +}; + +static struct pc_orangefs_attribute pc_acache_attribute = + __ATTR(acache, + 0664, + service_pc_show, + NULL); + +static struct pc_orangefs_attribute pc_capcache_attribute = + __ATTR(capcache, + 0664, + service_pc_show, + NULL); + +static struct pc_orangefs_attribute pc_ncache_attribute = + __ATTR(ncache, + 0664, + service_pc_show, + NULL); + +static struct attribute *pc_orangefs_default_attrs[] = { + &pc_acache_attribute.attr, + &pc_capcache_attribute.attr, + &pc_ncache_attribute.attr, + NULL, +}; + +static struct kobj_type pc_orangefs_ktype = { + .sysfs_ops = &pc_orangefs_sysfs_ops, + .release = pc_orangefs_release, + .default_attrs = pc_orangefs_default_attrs, +}; + +static struct stats_orangefs_attribute stats_reads_attribute = + __ATTR(reads, + 0664, + int_stats_show, + NULL); + +static struct stats_orangefs_attribute stats_writes_attribute = + __ATTR(writes, + 0664, + int_stats_show, + NULL); + +static struct attribute *stats_orangefs_default_attrs[] = { + &stats_reads_attribute.attr, + &stats_writes_attribute.attr, + NULL, +}; + +static struct kobj_type stats_orangefs_ktype = { + .sysfs_ops = &stats_orangefs_sysfs_ops, + .release = stats_orangefs_release, + .default_attrs = stats_orangefs_default_attrs, +}; + +static struct orangefs_obj *orangefs_obj; +static struct acache_orangefs_obj *acache_orangefs_obj; +static struct capcache_orangefs_obj *capcache_orangefs_obj; +static struct ccache_orangefs_obj *ccache_orangefs_obj; +static struct ncache_orangefs_obj *ncache_orangefs_obj; +static struct pc_orangefs_obj *pc_orangefs_obj; +static struct stats_orangefs_obj *stats_orangefs_obj; + +int orangefs_sysfs_init(void) +{ + int rc = -EINVAL; + + gossip_debug(GOSSIP_SYSFS_DEBUG, "orangefs_sysfs_init: start\n"); + + /* create /sys/fs/orangefs. */ + orangefs_obj = kzalloc(sizeof(*orangefs_obj), GFP_KERNEL); + if (!orangefs_obj) + goto out; + + rc = kobject_init_and_add(&orangefs_obj->kobj, + &orangefs_ktype, + fs_kobj, + ORANGEFS_KOBJ_ID); + + if (rc) + goto ofs_obj_bail; + + kobject_uevent(&orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/acache. */ + acache_orangefs_obj = kzalloc(sizeof(*acache_orangefs_obj), GFP_KERNEL); + if (!acache_orangefs_obj) { + rc = -EINVAL; + goto ofs_obj_bail; + } + + rc = kobject_init_and_add(&acache_orangefs_obj->kobj, + &acache_orangefs_ktype, + &orangefs_obj->kobj, + ACACHE_KOBJ_ID); + + if (rc) + goto acache_obj_bail; + + kobject_uevent(&acache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/capcache. */ + capcache_orangefs_obj = + kzalloc(sizeof(*capcache_orangefs_obj), GFP_KERNEL); + if (!capcache_orangefs_obj) { + rc = -EINVAL; + goto acache_obj_bail; + } + + rc = kobject_init_and_add(&capcache_orangefs_obj->kobj, + &capcache_orangefs_ktype, + &orangefs_obj->kobj, + CAPCACHE_KOBJ_ID); + if (rc) + goto capcache_obj_bail; + + kobject_uevent(&capcache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/ccache. */ + ccache_orangefs_obj = + kzalloc(sizeof(*ccache_orangefs_obj), GFP_KERNEL); + if (!ccache_orangefs_obj) { + rc = -EINVAL; + goto capcache_obj_bail; + } + + rc = kobject_init_and_add(&ccache_orangefs_obj->kobj, + &ccache_orangefs_ktype, + &orangefs_obj->kobj, + CCACHE_KOBJ_ID); + if (rc) + goto ccache_obj_bail; + + kobject_uevent(&ccache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/ncache. */ + ncache_orangefs_obj = kzalloc(sizeof(*ncache_orangefs_obj), GFP_KERNEL); + if (!ncache_orangefs_obj) { + rc = -EINVAL; + goto ccache_obj_bail; + } + + rc = kobject_init_and_add(&ncache_orangefs_obj->kobj, + &ncache_orangefs_ktype, + &orangefs_obj->kobj, + NCACHE_KOBJ_ID); + + if (rc) + goto ncache_obj_bail; + + kobject_uevent(&ncache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/perf_counters. */ + pc_orangefs_obj = kzalloc(sizeof(*pc_orangefs_obj), GFP_KERNEL); + if (!pc_orangefs_obj) { + rc = -EINVAL; + goto ncache_obj_bail; + } + + rc = kobject_init_and_add(&pc_orangefs_obj->kobj, + &pc_orangefs_ktype, + &orangefs_obj->kobj, + "perf_counters"); + + if (rc) + goto pc_obj_bail; + + kobject_uevent(&pc_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/stats. */ + stats_orangefs_obj = kzalloc(sizeof(*stats_orangefs_obj), GFP_KERNEL); + if (!stats_orangefs_obj) { + rc = -EINVAL; + goto pc_obj_bail; + } + + rc = kobject_init_and_add(&stats_orangefs_obj->kobj, + &stats_orangefs_ktype, + &orangefs_obj->kobj, + STATS_KOBJ_ID); + + if (rc) + goto stats_obj_bail; + + kobject_uevent(&stats_orangefs_obj->kobj, KOBJ_ADD); + goto out; + +stats_obj_bail: + kobject_put(&stats_orangefs_obj->kobj); + +pc_obj_bail: + kobject_put(&pc_orangefs_obj->kobj); + +ncache_obj_bail: + kobject_put(&ncache_orangefs_obj->kobj); + +ccache_obj_bail: + kobject_put(&ccache_orangefs_obj->kobj); + +capcache_obj_bail: + kobject_put(&capcache_orangefs_obj->kobj); + +acache_obj_bail: + kobject_put(&acache_orangefs_obj->kobj); + +ofs_obj_bail: + kobject_put(&orangefs_obj->kobj); +out: + return rc; +} + +void orangefs_sysfs_exit(void) +{ + gossip_debug(GOSSIP_SYSFS_DEBUG, "orangefs_sysfs_exit: start\n"); + + kobject_put(&acache_orangefs_obj->kobj); + kobject_put(&capcache_orangefs_obj->kobj); + kobject_put(&ccache_orangefs_obj->kobj); + kobject_put(&ncache_orangefs_obj->kobj); + kobject_put(&pc_orangefs_obj->kobj); + kobject_put(&stats_orangefs_obj->kobj); + + kobject_put(&orangefs_obj->kobj); +} diff --git a/fs/orangefs/orangefs-sysfs.h b/fs/orangefs/orangefs-sysfs.h new file mode 100644 index 000000000000..f0b76382db02 --- /dev/null +++ b/fs/orangefs/orangefs-sysfs.h @@ -0,0 +1,2 @@ +extern int orangefs_sysfs_init(void); +extern void orangefs_sysfs_exit(void); diff --git a/fs/orangefs/orangefs-utils.c b/fs/orangefs/orangefs-utils.c new file mode 100644 index 000000000000..40f5163b56aa --- /dev/null +++ b/fs/orangefs/orangefs-utils.c @@ -0,0 +1,1048 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-dev-proto.h" +#include "orangefs-bufmap.h" + +__s32 fsid_of_op(struct orangefs_kernel_op_s *op) +{ + __s32 fsid = ORANGEFS_FS_ID_NULL; + + if (op) { + switch (op->upcall.type) { + case ORANGEFS_VFS_OP_FILE_IO: + fsid = op->upcall.req.io.refn.fs_id; + break; + case ORANGEFS_VFS_OP_LOOKUP: + fsid = op->upcall.req.lookup.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_CREATE: + fsid = op->upcall.req.create.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_GETATTR: + fsid = op->upcall.req.getattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_REMOVE: + fsid = op->upcall.req.remove.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_MKDIR: + fsid = op->upcall.req.mkdir.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_READDIR: + fsid = op->upcall.req.readdir.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SETATTR: + fsid = op->upcall.req.setattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SYMLINK: + fsid = op->upcall.req.sym.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_RENAME: + fsid = op->upcall.req.rename.old_parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_STATFS: + fsid = op->upcall.req.statfs.fs_id; + break; + case ORANGEFS_VFS_OP_TRUNCATE: + fsid = op->upcall.req.truncate.refn.fs_id; + break; + case ORANGEFS_VFS_OP_MMAP_RA_FLUSH: + fsid = op->upcall.req.ra_cache_flush.refn.fs_id; + break; + case ORANGEFS_VFS_OP_FS_UMOUNT: + fsid = op->upcall.req.fs_umount.fs_id; + break; + case ORANGEFS_VFS_OP_GETXATTR: + fsid = op->upcall.req.getxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SETXATTR: + fsid = op->upcall.req.setxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_LISTXATTR: + fsid = op->upcall.req.listxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_REMOVEXATTR: + fsid = op->upcall.req.removexattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_FSYNC: + fsid = op->upcall.req.fsync.refn.fs_id; + break; + default: + break; + } + } + return fsid; +} + +static int orangefs_inode_flags(struct ORANGEFS_sys_attr_s *attrs) +{ + int flags = 0; + if (attrs->flags & ORANGEFS_IMMUTABLE_FL) + flags |= S_IMMUTABLE; + else + flags &= ~S_IMMUTABLE; + if (attrs->flags & ORANGEFS_APPEND_FL) + flags |= S_APPEND; + else + flags &= ~S_APPEND; + if (attrs->flags & ORANGEFS_NOATIME_FL) + flags |= S_NOATIME; + else + flags &= ~S_NOATIME; + return flags; +} + +static int orangefs_inode_perms(struct ORANGEFS_sys_attr_s *attrs) +{ + int perm_mode = 0; + + if (attrs->perms & ORANGEFS_O_EXECUTE) + perm_mode |= S_IXOTH; + if (attrs->perms & ORANGEFS_O_WRITE) + perm_mode |= S_IWOTH; + if (attrs->perms & ORANGEFS_O_READ) + perm_mode |= S_IROTH; + + if (attrs->perms & ORANGEFS_G_EXECUTE) + perm_mode |= S_IXGRP; + if (attrs->perms & ORANGEFS_G_WRITE) + perm_mode |= S_IWGRP; + if (attrs->perms & ORANGEFS_G_READ) + perm_mode |= S_IRGRP; + + if (attrs->perms & ORANGEFS_U_EXECUTE) + perm_mode |= S_IXUSR; + if (attrs->perms & ORANGEFS_U_WRITE) + perm_mode |= S_IWUSR; + if (attrs->perms & ORANGEFS_U_READ) + perm_mode |= S_IRUSR; + + if (attrs->perms & ORANGEFS_G_SGID) + perm_mode |= S_ISGID; + if (attrs->perms & ORANGEFS_U_SUID) + perm_mode |= S_ISUID; + + return perm_mode; +} + +/* + * NOTE: in kernel land, we never use the sys_attr->link_target for + * anything, so don't bother copying it into the sys_attr object here. + */ +static inline int copy_attributes_from_inode(struct inode *inode, + struct ORANGEFS_sys_attr_s *attrs, + struct iattr *iattr) +{ + umode_t tmp_mode; + + if (!iattr || !inode || !attrs) { + gossip_err("NULL iattr (%p), inode (%p), attrs (%p) " + "in copy_attributes_from_inode!\n", + iattr, + inode, + attrs); + return -EINVAL; + } + /* + * We need to be careful to only copy the attributes out of the + * iattr object that we know are valid. + */ + attrs->mask = 0; + if (iattr->ia_valid & ATTR_UID) { + attrs->owner = from_kuid(current_user_ns(), iattr->ia_uid); + attrs->mask |= ORANGEFS_ATTR_SYS_UID; + gossip_debug(GOSSIP_UTILS_DEBUG, "(UID) %d\n", attrs->owner); + } + if (iattr->ia_valid & ATTR_GID) { + attrs->group = from_kgid(current_user_ns(), iattr->ia_gid); + attrs->mask |= ORANGEFS_ATTR_SYS_GID; + gossip_debug(GOSSIP_UTILS_DEBUG, "(GID) %d\n", attrs->group); + } + + if (iattr->ia_valid & ATTR_ATIME) { + attrs->mask |= ORANGEFS_ATTR_SYS_ATIME; + if (iattr->ia_valid & ATTR_ATIME_SET) { + attrs->atime = (time64_t)iattr->ia_atime.tv_sec; + attrs->mask |= ORANGEFS_ATTR_SYS_ATIME_SET; + } + } + if (iattr->ia_valid & ATTR_MTIME) { + attrs->mask |= ORANGEFS_ATTR_SYS_MTIME; + if (iattr->ia_valid & ATTR_MTIME_SET) { + attrs->mtime = (time64_t)iattr->ia_mtime.tv_sec; + attrs->mask |= ORANGEFS_ATTR_SYS_MTIME_SET; + } + } + if (iattr->ia_valid & ATTR_CTIME) + attrs->mask |= ORANGEFS_ATTR_SYS_CTIME; + + /* + * ORANGEFS cannot set size with a setattr operation. Probably not likely + * to be requested through the VFS, but just in case, don't worry about + * ATTR_SIZE + */ + + if (iattr->ia_valid & ATTR_MODE) { + tmp_mode = iattr->ia_mode; + if (tmp_mode & (S_ISVTX)) { + if (is_root_handle(inode)) { + /* + * allow sticky bit to be set on root (since + * it shows up that way by default anyhow), + * but don't show it to the server + */ + tmp_mode -= S_ISVTX; + } else { + gossip_debug(GOSSIP_UTILS_DEBUG, + "User attempted to set sticky bit on non-root directory; returning EINVAL.\n"); + return -EINVAL; + } + } + + if (tmp_mode & (S_ISUID)) { + gossip_debug(GOSSIP_UTILS_DEBUG, + "Attempting to set setuid bit (not supported); returning EINVAL.\n"); + return -EINVAL; + } + + attrs->perms = ORANGEFS_util_translate_mode(tmp_mode); + attrs->mask |= ORANGEFS_ATTR_SYS_PERM; + } + + return 0; +} + +static int orangefs_inode_type(enum orangefs_ds_type objtype) +{ + if (objtype == ORANGEFS_TYPE_METAFILE) + return S_IFREG; + else if (objtype == ORANGEFS_TYPE_DIRECTORY) + return S_IFDIR; + else if (objtype == ORANGEFS_TYPE_SYMLINK) + return S_IFLNK; + else + return -1; +} + +static int orangefs_inode_is_stale(struct inode *inode, int new, + struct ORANGEFS_sys_attr_s *attrs, char *link_target) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + int type = orangefs_inode_type(attrs->objtype); + if (!new) { + /* + * If the inode type or symlink target have changed then this + * inode is stale. + */ + if (type == -1 || !(inode->i_mode & type)) { + orangefs_make_bad_inode(inode); + return 1; + } + if (type == S_IFLNK && strncmp(orangefs_inode->link_target, + link_target, ORANGEFS_NAME_MAX)) { + orangefs_make_bad_inode(inode); + return 1; + } + } + return 0; +} + +int orangefs_inode_getattr(struct inode *inode, int new, int size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + loff_t inode_size, rounded_up_size; + int ret, type; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: called on inode %pU\n", __func__, + get_khandle_from_ino(inode)); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETATTR); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.getattr.refn = orangefs_inode->refn; + new_op->upcall.req.getattr.mask = size ? + ORANGEFS_ATTR_SYS_ALL_NOHINT : ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto out; + + type = orangefs_inode_type(new_op-> + downcall.resp.getattr.attributes.objtype); + ret = orangefs_inode_is_stale(inode, new, + &new_op->downcall.resp.getattr.attributes, + new_op->downcall.resp.getattr.link_target); + if (ret) { + ret = -ESTALE; + goto out; + } + + switch (type) { + case S_IFREG: + inode->i_flags = orangefs_inode_flags(&new_op-> + downcall.resp.getattr.attributes); + if (size) { + inode_size = (loff_t)new_op-> + downcall.resp.getattr.attributes.size; + rounded_up_size = + (inode_size + (4096 - (inode_size % 4096))); + inode->i_size = inode_size; + orangefs_inode->blksize = + new_op->downcall.resp.getattr.attributes.blksize; + spin_lock(&inode->i_lock); + inode->i_bytes = inode_size; + inode->i_blocks = + (unsigned long)(rounded_up_size / 512); + spin_unlock(&inode->i_lock); + } + break; + case S_IFDIR: + inode->i_size = PAGE_CACHE_SIZE; + orangefs_inode->blksize = (1 << inode->i_blkbits); + spin_lock(&inode->i_lock); + inode_set_bytes(inode, inode->i_size); + spin_unlock(&inode->i_lock); + set_nlink(inode, 1); + break; + case S_IFLNK: + if (new) { + inode->i_size = (loff_t)strlen(new_op-> + downcall.resp.getattr.link_target); + orangefs_inode->blksize = (1 << inode->i_blkbits); + strlcpy(orangefs_inode->link_target, + new_op->downcall.resp.getattr.link_target, + ORANGEFS_NAME_MAX); + inode->i_link = orangefs_inode->link_target; + } + break; + } + + inode->i_uid = make_kuid(&init_user_ns, new_op-> + downcall.resp.getattr.attributes.owner); + inode->i_gid = make_kgid(&init_user_ns, new_op-> + downcall.resp.getattr.attributes.group); + inode->i_atime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.atime; + inode->i_mtime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.mtime; + inode->i_ctime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.ctime; + inode->i_atime.tv_nsec = 0; + inode->i_mtime.tv_nsec = 0; + inode->i_ctime.tv_nsec = 0; + + /* special case: mark the root inode as sticky */ + inode->i_mode = type | (is_root_handle(inode) ? S_ISVTX : 0) | + orangefs_inode_perms(&new_op->downcall.resp.getattr.attributes); + + ret = 0; +out: + op_release(new_op); + return ret; +} + +int orangefs_inode_check_changed(struct inode *inode) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: called on inode %pU\n", __func__, + get_khandle_from_ino(inode)); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETATTR); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.getattr.refn = orangefs_inode->refn; + new_op->upcall.req.getattr.mask = ORANGEFS_ATTR_SYS_TYPE | + ORANGEFS_ATTR_SYS_LNK_TARGET; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto out; + + ret = orangefs_inode_is_stale(inode, 0, + &new_op->downcall.resp.getattr.attributes, + new_op->downcall.resp.getattr.link_target); +out: + op_release(new_op); + return ret; +} + +/* + * issues a orangefs setattr request to make sure the new attribute values + * take effect if successful. returns 0 on success; -errno otherwise + */ +int orangefs_inode_setattr(struct inode *inode, struct iattr *iattr) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_SETATTR); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.setattr.refn = orangefs_inode->refn; + ret = copy_attributes_from_inode(inode, + &new_op->upcall.req.setattr.attributes, + iattr); + if (ret >= 0) { + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_inode_setattr: returning %d\n", + ret); + } + + op_release(new_op); + + /* + * successful setattr should clear the atime, mtime and + * ctime flags. + */ + if (ret == 0) { + ClearAtimeFlag(orangefs_inode); + ClearMtimeFlag(orangefs_inode); + ClearCtimeFlag(orangefs_inode); + ClearModeFlag(orangefs_inode); + } + + return ret; +} + +int orangefs_flush_inode(struct inode *inode) +{ + /* + * If it is a dirty inode, this function gets called. + * Gather all the information that needs to be setattr'ed + * Right now, this will only be used for mode, atime, mtime + * and/or ctime. + */ + struct iattr wbattr; + int ret; + int mtime_flag; + int ctime_flag; + int atime_flag; + int mode_flag; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + memset(&wbattr, 0, sizeof(wbattr)); + + /* + * check inode flags up front, and clear them if they are set. This + * will prevent multiple processes from all trying to flush the same + * inode if they call close() simultaneously + */ + mtime_flag = MtimeFlag(orangefs_inode); + ClearMtimeFlag(orangefs_inode); + ctime_flag = CtimeFlag(orangefs_inode); + ClearCtimeFlag(orangefs_inode); + atime_flag = AtimeFlag(orangefs_inode); + ClearAtimeFlag(orangefs_inode); + mode_flag = ModeFlag(orangefs_inode); + ClearModeFlag(orangefs_inode); + + /* -- Lazy atime,mtime and ctime update -- + * Note: all times are dictated by server in the new scheme + * and not by the clients + * + * Also mode updates are being handled now.. + */ + + if (mtime_flag) + wbattr.ia_valid |= ATTR_MTIME; + if (ctime_flag) + wbattr.ia_valid |= ATTR_CTIME; + if (atime_flag) + wbattr.ia_valid |= ATTR_ATIME; + + if (mode_flag) { + wbattr.ia_mode = inode->i_mode; + wbattr.ia_valid |= ATTR_MODE; + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "*********** orangefs_flush_inode: %pU " + "(ia_valid %d)\n", + get_khandle_from_ino(inode), + wbattr.ia_valid); + if (wbattr.ia_valid == 0) { + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_flush_inode skipping setattr()\n"); + return 0; + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_flush_inode (%pU) writing mode %o\n", + get_khandle_from_ino(inode), + inode->i_mode); + + ret = orangefs_inode_setattr(inode, &wbattr); + + return ret; +} + +int orangefs_unmount_sb(struct super_block *sb) +{ + int ret = -EINVAL; + struct orangefs_kernel_op_s *new_op = NULL; + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_unmount_sb called on sb %p\n", + sb); + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_UMOUNT); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.fs_umount.id = ORANGEFS_SB(sb)->id; + new_op->upcall.req.fs_umount.fs_id = ORANGEFS_SB(sb)->fs_id; + strncpy(new_op->upcall.req.fs_umount.orangefs_config_server, + ORANGEFS_SB(sb)->devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "Attempting ORANGEFS Unmount via host %s\n", + new_op->upcall.req.fs_umount.orangefs_config_server); + + ret = service_operation(new_op, "orangefs_fs_umount", 0); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_unmount: got return value of %d\n", ret); + if (ret) + sb = ERR_PTR(ret); + else + ORANGEFS_SB(sb)->mount_pending = 1; + + op_release(new_op); + return ret; +} + +void orangefs_make_bad_inode(struct inode *inode) +{ + if (is_root_handle(inode)) { + /* + * if this occurs, the pvfs2-client-core was killed but we + * can't afford to lose the inode operations and such + * associated with the root handle in any case. + */ + gossip_debug(GOSSIP_UTILS_DEBUG, + "*** NOT making bad root inode %pU\n", + get_khandle_from_ino(inode)); + } else { + gossip_debug(GOSSIP_UTILS_DEBUG, + "*** making bad inode %pU\n", + get_khandle_from_ino(inode)); + make_bad_inode(inode); + } +} + +/* + * The following is a very dirty hack that is now a permanent part of the + * ORANGEFS protocol. See protocol.h for more error definitions. + */ + +/* The order matches include/orangefs-types.h in the OrangeFS source. */ +static int PINT_errno_mapping[] = { + 0, EPERM, ENOENT, EINTR, EIO, ENXIO, EBADF, EAGAIN, ENOMEM, + EFAULT, EBUSY, EEXIST, ENODEV, ENOTDIR, EISDIR, EINVAL, EMFILE, + EFBIG, ENOSPC, EROFS, EMLINK, EPIPE, EDEADLK, ENAMETOOLONG, + ENOLCK, ENOSYS, ENOTEMPTY, ELOOP, EWOULDBLOCK, ENOMSG, EUNATCH, + EBADR, EDEADLOCK, ENODATA, ETIME, ENONET, EREMOTE, ECOMM, + EPROTO, EBADMSG, EOVERFLOW, ERESTART, EMSGSIZE, EPROTOTYPE, + ENOPROTOOPT, EPROTONOSUPPORT, EOPNOTSUPP, EADDRINUSE, + EADDRNOTAVAIL, ENETDOWN, ENETUNREACH, ENETRESET, ENOBUFS, + ETIMEDOUT, ECONNREFUSED, EHOSTDOWN, EHOSTUNREACH, EALREADY, + EACCES, ECONNRESET, ERANGE +}; + +int orangefs_normalize_to_errno(__s32 error_code) +{ + __u32 i; + + /* Success */ + if (error_code == 0) { + return 0; + /* + * This shouldn't ever happen. If it does it should be fixed on the + * server. + */ + } else if (error_code > 0) { + gossip_err("orangefs: error status receieved.\n"); + gossip_err("orangefs: assuming error code is inverted.\n"); + error_code = -error_code; + } + + /* + * XXX: This is very bad since error codes from ORANGEFS may not be + * suitable for return into userspace. + */ + + /* + * Convert ORANGEFS error values into errno values suitable for return + * from the kernel. + */ + if ((-error_code) & ORANGEFS_NON_ERRNO_ERROR_BIT) { + if (((-error_code) & + (ORANGEFS_ERROR_NUMBER_BITS|ORANGEFS_NON_ERRNO_ERROR_BIT| + ORANGEFS_ERROR_BIT)) == ORANGEFS_ECANCEL) { + /* + * cancellation error codes generally correspond to + * a timeout from the client's perspective + */ + error_code = -ETIMEDOUT; + } else { + /* assume a default error code */ + gossip_err("orangefs: warning: got error code without errno equivalent: %d.\n", error_code); + error_code = -EINVAL; + } + + /* Convert ORANGEFS encoded errno values into regular errno values. */ + } else if ((-error_code) & ORANGEFS_ERROR_BIT) { + i = (-error_code) & ~(ORANGEFS_ERROR_BIT|ORANGEFS_ERROR_CLASS_BITS); + if (i < sizeof(PINT_errno_mapping)/sizeof(*PINT_errno_mapping)) + error_code = -PINT_errno_mapping[i]; + else + error_code = -EINVAL; + + /* + * Only ORANGEFS protocol error codes should ever come here. Otherwise + * there is a bug somewhere. + */ + } else { + gossip_err("orangefs: orangefs_normalize_to_errno: got error code which is not from ORANGEFS.\n"); + } + return error_code; +} + +#define NUM_MODES 11 +__s32 ORANGEFS_util_translate_mode(int mode) +{ + int ret = 0; + int i = 0; + static int modes[NUM_MODES] = { + S_IXOTH, S_IWOTH, S_IROTH, + S_IXGRP, S_IWGRP, S_IRGRP, + S_IXUSR, S_IWUSR, S_IRUSR, + S_ISGID, S_ISUID + }; + static int orangefs_modes[NUM_MODES] = { + ORANGEFS_O_EXECUTE, ORANGEFS_O_WRITE, ORANGEFS_O_READ, + ORANGEFS_G_EXECUTE, ORANGEFS_G_WRITE, ORANGEFS_G_READ, + ORANGEFS_U_EXECUTE, ORANGEFS_U_WRITE, ORANGEFS_U_READ, + ORANGEFS_G_SGID, ORANGEFS_U_SUID + }; + + for (i = 0; i < NUM_MODES; i++) + if (mode & modes[i]) + ret |= orangefs_modes[i]; + + return ret; +} +#undef NUM_MODES + +/* + * After obtaining a string representation of the client's debug + * keywords and their associated masks, this function is called to build an + * array of these values. + */ +int orangefs_prepare_cdm_array(char *debug_array_string) +{ + int i; + int rc = -EINVAL; + char *cds_head = NULL; + char *cds_delimiter = NULL; + int keyword_len = 0; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + /* + * figure out how many elements the cdm_array needs. + */ + for (i = 0; i < strlen(debug_array_string); i++) + if (debug_array_string[i] == '\n') + cdm_element_count++; + + if (!cdm_element_count) { + pr_info("No elements in client debug array string!\n"); + goto out; + } + + cdm_array = + kzalloc(cdm_element_count * sizeof(struct client_debug_mask), + GFP_KERNEL); + if (!cdm_array) { + pr_info("malloc failed for cdm_array!\n"); + rc = -ENOMEM; + goto out; + } + + cds_head = debug_array_string; + + for (i = 0; i < cdm_element_count; i++) { + cds_delimiter = strchr(cds_head, '\n'); + *cds_delimiter = '\0'; + + keyword_len = strcspn(cds_head, " "); + + cdm_array[i].keyword = kzalloc(keyword_len + 1, GFP_KERNEL); + if (!cdm_array[i].keyword) { + rc = -ENOMEM; + goto out; + } + + sscanf(cds_head, + "%s %llx %llx", + cdm_array[i].keyword, + (unsigned long long *)&(cdm_array[i].mask1), + (unsigned long long *)&(cdm_array[i].mask2)); + + if (!strcmp(cdm_array[i].keyword, ORANGEFS_VERBOSE)) + client_verbose_index = i; + + if (!strcmp(cdm_array[i].keyword, ORANGEFS_ALL)) + client_all_index = i; + + cds_head = cds_delimiter + 1; + } + + rc = cdm_element_count; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: rc:%d:\n", __func__, rc); + +out: + + return rc; + +} + +/* + * /sys/kernel/debug/orangefs/debug-help can be catted to + * see all the available kernel and client debug keywords. + * + * When the kernel boots, we have no idea what keywords the + * client supports, nor their associated masks. + * + * We pass through this function once at boot and stamp a + * boilerplate "we don't know" message for the client in the + * debug-help file. We pass through here again when the client + * starts and then we can fill out the debug-help file fully. + * + * The client might be restarted any number of times between + * reboots, we only build the debug-help file the first time. + */ +int orangefs_prepare_debugfs_help_string(int at_boot) +{ + int rc = -EINVAL; + int i; + int byte_count = 0; + char *client_title = "Client Debug Keywords:\n"; + char *kernel_title = "Kernel Debug Keywords:\n"; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (at_boot) { + byte_count += strlen(HELP_STRING_UNINITIALIZED); + client_title = HELP_STRING_UNINITIALIZED; + } else { + /* + * fill the client keyword/mask array and remember + * how many elements there were. + */ + cdm_element_count = + orangefs_prepare_cdm_array(client_debug_array_string); + if (cdm_element_count <= 0) + goto out; + + /* Count the bytes destined for debug_help_string. */ + byte_count += strlen(client_title); + + for (i = 0; i < cdm_element_count; i++) { + byte_count += strlen(cdm_array[i].keyword + 2); + if (byte_count >= DEBUG_HELP_STRING_SIZE) { + pr_info("%s: overflow 1!\n", __func__); + goto out; + } + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "%s: cdm_element_count:%d:\n", + __func__, + cdm_element_count); + } + + byte_count += strlen(kernel_title); + for (i = 0; i < num_kmod_keyword_mask_map; i++) { + byte_count += + strlen(s_kmod_keyword_mask_map[i].keyword + 2); + if (byte_count >= DEBUG_HELP_STRING_SIZE) { + pr_info("%s: overflow 2!\n", __func__); + goto out; + } + } + + /* build debug_help_string. */ + debug_help_string = kzalloc(DEBUG_HELP_STRING_SIZE, GFP_KERNEL); + if (!debug_help_string) { + rc = -ENOMEM; + goto out; + } + + strcat(debug_help_string, client_title); + + if (!at_boot) { + for (i = 0; i < cdm_element_count; i++) { + strcat(debug_help_string, "\t"); + strcat(debug_help_string, cdm_array[i].keyword); + strcat(debug_help_string, "\n"); + } + } + + strcat(debug_help_string, "\n"); + strcat(debug_help_string, kernel_title); + + for (i = 0; i < num_kmod_keyword_mask_map; i++) { + strcat(debug_help_string, "\t"); + strcat(debug_help_string, s_kmod_keyword_mask_map[i].keyword); + strcat(debug_help_string, "\n"); + } + + rc = 0; + +out: + + return rc; + +} + +/* + * kernel = type 0 + * client = type 1 + */ +void debug_mask_to_string(void *mask, int type) +{ + int i; + int len = 0; + char *debug_string; + int element_count = 0; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (type) { + debug_string = client_debug_string; + element_count = cdm_element_count; + } else { + debug_string = kernel_debug_string; + element_count = num_kmod_keyword_mask_map; + } + + memset(debug_string, 0, ORANGEFS_MAX_DEBUG_STRING_LEN); + + /* + * Some keywords, like "all" or "verbose", are amalgams of + * numerous other keywords. Make a special check for those + * before grinding through the whole mask only to find out + * later... + */ + if (check_amalgam_keyword(mask, type)) + goto out; + + /* Build the debug string. */ + for (i = 0; i < element_count; i++) + if (type) + do_c_string(mask, i); + else + do_k_string(mask, i); + + len = strlen(debug_string); + + if ((len) && (type)) + client_debug_string[len - 1] = '\0'; + else if (len) + kernel_debug_string[len - 1] = '\0'; + else if (type) + strcpy(client_debug_string, "none"); + else + strcpy(kernel_debug_string, "none"); + +out: +gossip_debug(GOSSIP_UTILS_DEBUG, "%s: string:%s:\n", __func__, debug_string); + + return; + +} + +void do_k_string(void *k_mask, int index) +{ + __u64 *mask = (__u64 *) k_mask; + + if (keyword_is_amalgam((char *) s_kmod_keyword_mask_map[index].keyword)) + goto out; + + if (*mask & s_kmod_keyword_mask_map[index].mask_val) { + if ((strlen(kernel_debug_string) + + strlen(s_kmod_keyword_mask_map[index].keyword)) + < ORANGEFS_MAX_DEBUG_STRING_LEN - 1) { + strcat(kernel_debug_string, + s_kmod_keyword_mask_map[index].keyword); + strcat(kernel_debug_string, ","); + } else { + gossip_err("%s: overflow!\n", __func__); + strcpy(kernel_debug_string, ORANGEFS_ALL); + goto out; + } + } + +out: + + return; +} + +void do_c_string(void *c_mask, int index) +{ + struct client_debug_mask *mask = (struct client_debug_mask *) c_mask; + + if (keyword_is_amalgam(cdm_array[index].keyword)) + goto out; + + if ((mask->mask1 & cdm_array[index].mask1) || + (mask->mask2 & cdm_array[index].mask2)) { + if ((strlen(client_debug_string) + + strlen(cdm_array[index].keyword) + 1) + < ORANGEFS_MAX_DEBUG_STRING_LEN - 2) { + strcat(client_debug_string, + cdm_array[index].keyword); + strcat(client_debug_string, ","); + } else { + gossip_err("%s: overflow!\n", __func__); + strcpy(client_debug_string, ORANGEFS_ALL); + goto out; + } + } +out: + return; +} + +int keyword_is_amalgam(char *keyword) +{ + int rc = 0; + + if ((!strcmp(keyword, ORANGEFS_ALL)) || (!strcmp(keyword, ORANGEFS_VERBOSE))) + rc = 1; + + return rc; +} + +/* + * kernel = type 0 + * client = type 1 + * + * return 1 if we found an amalgam. + */ +int check_amalgam_keyword(void *mask, int type) +{ + __u64 *k_mask; + struct client_debug_mask *c_mask; + int k_all_index = num_kmod_keyword_mask_map - 1; + int rc = 0; + + if (type) { + c_mask = (struct client_debug_mask *) mask; + + if ((c_mask->mask1 == cdm_array[client_all_index].mask1) && + (c_mask->mask2 == cdm_array[client_all_index].mask2)) { + strcpy(client_debug_string, ORANGEFS_ALL); + rc = 1; + goto out; + } + + if ((c_mask->mask1 == cdm_array[client_verbose_index].mask1) && + (c_mask->mask2 == cdm_array[client_verbose_index].mask2)) { + strcpy(client_debug_string, ORANGEFS_VERBOSE); + rc = 1; + goto out; + } + + } else { + k_mask = (__u64 *) mask; + + if (*k_mask >= s_kmod_keyword_mask_map[k_all_index].mask_val) { + strcpy(kernel_debug_string, ORANGEFS_ALL); + rc = 1; + goto out; + } + } + +out: + + return rc; +} + +/* + * kernel = type 0 + * client = type 1 + */ +void debug_string_to_mask(char *debug_string, void *mask, int type) +{ + char *unchecked_keyword; + int i; + char *strsep_fodder = kstrdup(debug_string, GFP_KERNEL); + char *original_pointer; + int element_count = 0; + struct client_debug_mask *c_mask; + __u64 *k_mask; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (type) { + c_mask = (struct client_debug_mask *)mask; + element_count = cdm_element_count; + } else { + k_mask = (__u64 *)mask; + *k_mask = 0; + element_count = num_kmod_keyword_mask_map; + } + + original_pointer = strsep_fodder; + while ((unchecked_keyword = strsep(&strsep_fodder, ","))) + if (strlen(unchecked_keyword)) { + for (i = 0; i < element_count; i++) + if (type) + do_c_mask(i, + unchecked_keyword, + &c_mask); + else + do_k_mask(i, + unchecked_keyword, + &k_mask); + } + + kfree(original_pointer); +} + +void do_c_mask(int i, + char *unchecked_keyword, + struct client_debug_mask **sane_mask) +{ + + if (!strcmp(cdm_array[i].keyword, unchecked_keyword)) { + (**sane_mask).mask1 = (**sane_mask).mask1 | cdm_array[i].mask1; + (**sane_mask).mask2 = (**sane_mask).mask2 | cdm_array[i].mask2; + } +} + +void do_k_mask(int i, char *unchecked_keyword, __u64 **sane_mask) +{ + + if (!strcmp(s_kmod_keyword_mask_map[i].keyword, unchecked_keyword)) + **sane_mask = (**sane_mask) | + s_kmod_keyword_mask_map[i].mask_val; +} diff --git a/fs/orangefs/protocol.h b/fs/orangefs/protocol.h new file mode 100644 index 000000000000..45ce4ff4cbc7 --- /dev/null +++ b/fs/orangefs/protocol.h @@ -0,0 +1,452 @@ +#include <linux/types.h> +#include <linux/spinlock_types.h> +#include <linux/slab.h> +#include <linux/ioctl.h> + +extern struct client_debug_mask *cdm_array; +extern char *debug_help_string; +extern int help_string_initialized; +extern struct dentry *debug_dir; +extern struct dentry *help_file_dentry; +extern struct dentry *client_debug_dentry; +extern const struct file_operations debug_help_fops; +extern int client_all_index; +extern int client_verbose_index; +extern int cdm_element_count; +#define DEBUG_HELP_STRING_SIZE 4096 +#define HELP_STRING_UNINITIALIZED \ + "Client Debug Keywords are unknown until the first time\n" \ + "the client is started after boot.\n" +#define ORANGEFS_KMOD_DEBUG_HELP_FILE "debug-help" +#define ORANGEFS_KMOD_DEBUG_FILE "kernel-debug" +#define ORANGEFS_CLIENT_DEBUG_FILE "client-debug" +#define ORANGEFS_VERBOSE "verbose" +#define ORANGEFS_ALL "all" + +/* pvfs2-config.h ***********************************************************/ +#define ORANGEFS_VERSION_MAJOR 2 +#define ORANGEFS_VERSION_MINOR 9 +#define ORANGEFS_VERSION_SUB 0 + +/* khandle stuff ***********************************************************/ + +/* + * The 2.9 core will put 64 bit handles in here like this: + * 1234 0000 0000 5678 + * The 3.0 and beyond cores will put 128 bit handles in here like this: + * 1234 5678 90AB CDEF + * The kernel module will always use the first four bytes and + * the last four bytes as an inum. + */ +struct orangefs_khandle { + unsigned char u[16]; +} __aligned(8); + +/* + * kernel version of an object ref. + */ +struct orangefs_object_kref { + struct orangefs_khandle khandle; + __s32 fs_id; + __s32 __pad1; +}; + +/* + * compare 2 khandles assumes little endian thus from large address to + * small address + */ +static inline int ORANGEFS_khandle_cmp(const struct orangefs_khandle *kh1, + const struct orangefs_khandle *kh2) +{ + int i; + + for (i = 15; i >= 0; i--) { + if (kh1->u[i] > kh2->u[i]) + return 1; + if (kh1->u[i] < kh2->u[i]) + return -1; + } + + return 0; +} + +static inline void ORANGEFS_khandle_to(const struct orangefs_khandle *kh, + void *p, int size) +{ + + memset(p, 0, size); + memcpy(p, kh->u, 16); + +} + +static inline void ORANGEFS_khandle_from(struct orangefs_khandle *kh, + void *p, int size) +{ + memset(kh, 0, 16); + memcpy(kh->u, p, 16); + +} + +/* pvfs2-types.h ************************************************************/ +typedef __u32 ORANGEFS_uid; +typedef __u32 ORANGEFS_gid; +typedef __s32 ORANGEFS_fs_id; +typedef __u32 ORANGEFS_permissions; +typedef __u64 ORANGEFS_time; +typedef __s64 ORANGEFS_size; +typedef __u64 ORANGEFS_flags; +typedef __u64 ORANGEFS_ds_position; +typedef __s32 ORANGEFS_error; +typedef __s64 ORANGEFS_offset; + +#define ORANGEFS_SUPER_MAGIC 0x20030528 + +/* + * ORANGEFS error codes are a signed 32-bit integer. Error codes are negative, but + * the sign is stripped before decoding. + */ + +/* Bit 31 is not used since it is the sign. */ + +/* + * Bit 30 specifies that this is a ORANGEFS error. A ORANGEFS error is either an + * encoded errno value or a ORANGEFS protocol error. + */ +#define ORANGEFS_ERROR_BIT (1 << 30) + +/* + * Bit 29 specifies that this is a ORANGEFS protocol error and not an encoded + * errno value. + */ +#define ORANGEFS_NON_ERRNO_ERROR_BIT (1 << 29) + +/* + * Bits 9, 8, and 7 specify the error class, which encodes the section of + * server code the error originated in for logging purposes. It is not used + * in the kernel except to be masked out. + */ +#define ORANGEFS_ERROR_CLASS_BITS 0x380 + +/* Bits 6 - 0 are reserved for the actual error code. */ +#define ORANGEFS_ERROR_NUMBER_BITS 0x7f + +/* Encoded errno values decoded by PINT_errno_mapping in orangefs-utils.c. */ + +/* Our own ORANGEFS protocol error codes. */ +#define ORANGEFS_ECANCEL (1|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EDEVINIT (2|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EDETAIL (3|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EHOSTNTFD (4|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EADDRNTFD (5|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ENORECVR (6|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ETRYAGAIN (7|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ENOTPVFS (8|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ESECURITY (9|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) + +/* permission bits */ +#define ORANGEFS_O_EXECUTE (1 << 0) +#define ORANGEFS_O_WRITE (1 << 1) +#define ORANGEFS_O_READ (1 << 2) +#define ORANGEFS_G_EXECUTE (1 << 3) +#define ORANGEFS_G_WRITE (1 << 4) +#define ORANGEFS_G_READ (1 << 5) +#define ORANGEFS_U_EXECUTE (1 << 6) +#define ORANGEFS_U_WRITE (1 << 7) +#define ORANGEFS_U_READ (1 << 8) +/* no ORANGEFS_U_VTX (sticky bit) */ +#define ORANGEFS_G_SGID (1 << 10) +#define ORANGEFS_U_SUID (1 << 11) + +/* definition taken from stdint.h */ +#define INT32_MAX (2147483647) +#define ORANGEFS_ITERATE_START (INT32_MAX - 1) +#define ORANGEFS_ITERATE_END (INT32_MAX - 2) +#define ORANGEFS_ITERATE_NEXT (INT32_MAX - 3) +#define ORANGEFS_READDIR_START ORANGEFS_ITERATE_START +#define ORANGEFS_READDIR_END ORANGEFS_ITERATE_END +#define ORANGEFS_IMMUTABLE_FL FS_IMMUTABLE_FL +#define ORANGEFS_APPEND_FL FS_APPEND_FL +#define ORANGEFS_NOATIME_FL FS_NOATIME_FL +#define ORANGEFS_MIRROR_FL 0x01000000ULL +#define ORANGEFS_O_EXECUTE (1 << 0) +#define ORANGEFS_FS_ID_NULL ((__s32)0) +#define ORANGEFS_ATTR_SYS_UID (1 << 0) +#define ORANGEFS_ATTR_SYS_GID (1 << 1) +#define ORANGEFS_ATTR_SYS_PERM (1 << 2) +#define ORANGEFS_ATTR_SYS_ATIME (1 << 3) +#define ORANGEFS_ATTR_SYS_CTIME (1 << 4) +#define ORANGEFS_ATTR_SYS_MTIME (1 << 5) +#define ORANGEFS_ATTR_SYS_TYPE (1 << 6) +#define ORANGEFS_ATTR_SYS_ATIME_SET (1 << 7) +#define ORANGEFS_ATTR_SYS_MTIME_SET (1 << 8) +#define ORANGEFS_ATTR_SYS_SIZE (1 << 20) +#define ORANGEFS_ATTR_SYS_LNK_TARGET (1 << 24) +#define ORANGEFS_ATTR_SYS_DFILE_COUNT (1 << 25) +#define ORANGEFS_ATTR_SYS_DIRENT_COUNT (1 << 26) +#define ORANGEFS_ATTR_SYS_BLKSIZE (1 << 28) +#define ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT (1 << 29) +#define ORANGEFS_ATTR_SYS_COMMON_ALL \ + (ORANGEFS_ATTR_SYS_UID | \ + ORANGEFS_ATTR_SYS_GID | \ + ORANGEFS_ATTR_SYS_PERM | \ + ORANGEFS_ATTR_SYS_ATIME | \ + ORANGEFS_ATTR_SYS_CTIME | \ + ORANGEFS_ATTR_SYS_MTIME | \ + ORANGEFS_ATTR_SYS_TYPE) + +#define ORANGEFS_ATTR_SYS_ALL_SETABLE \ +(ORANGEFS_ATTR_SYS_COMMON_ALL-ORANGEFS_ATTR_SYS_TYPE) + +#define ORANGEFS_ATTR_SYS_ALL_NOHINT \ + (ORANGEFS_ATTR_SYS_COMMON_ALL | \ + ORANGEFS_ATTR_SYS_SIZE | \ + ORANGEFS_ATTR_SYS_LNK_TARGET | \ + ORANGEFS_ATTR_SYS_DFILE_COUNT | \ + ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT | \ + ORANGEFS_ATTR_SYS_DIRENT_COUNT | \ + ORANGEFS_ATTR_SYS_BLKSIZE) + +#define ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE \ + (ORANGEFS_ATTR_SYS_COMMON_ALL | \ + ORANGEFS_ATTR_SYS_LNK_TARGET | \ + ORANGEFS_ATTR_SYS_DFILE_COUNT | \ + ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT | \ + ORANGEFS_ATTR_SYS_DIRENT_COUNT | \ + ORANGEFS_ATTR_SYS_BLKSIZE) + +#define ORANGEFS_XATTR_REPLACE 0x2 +#define ORANGEFS_XATTR_CREATE 0x1 +#define ORANGEFS_MAX_SERVER_ADDR_LEN 256 +#define ORANGEFS_NAME_MAX 256 +/* + * max extended attribute name len as imposed by the VFS and exploited for the + * upcall request types. + * NOTE: Please retain them as multiples of 8 even if you wish to change them + * This is *NECESSARY* for supporting 32 bit user-space binaries on a 64-bit + * kernel. Due to implementation within DBPF, this really needs to be + * ORANGEFS_NAME_MAX, which it was the same value as, but no reason to let it + * break if that changes in the future. + */ +#define ORANGEFS_MAX_XATTR_NAMELEN ORANGEFS_NAME_MAX /* Not the same as + * XATTR_NAME_MAX defined + * by <linux/xattr.h> + */ +#define ORANGEFS_MAX_XATTR_VALUELEN 8192 /* Not the same as XATTR_SIZE_MAX + * defined by <linux/xattr.h> + */ +#define ORANGEFS_MAX_XATTR_LISTLEN 16 /* Not the same as XATTR_LIST_MAX + * defined by <linux/xattr.h> + */ +/* + * ORANGEFS I/O operation types, used in both system and server interfaces. + */ +enum ORANGEFS_io_type { + ORANGEFS_IO_READ = 1, + ORANGEFS_IO_WRITE = 2 +}; + +/* + * If this enum is modified the server parameters related to the precreate pool + * batch and low threshold sizes may need to be modified to reflect this + * change. + */ +enum orangefs_ds_type { + ORANGEFS_TYPE_NONE = 0, + ORANGEFS_TYPE_METAFILE = (1 << 0), + ORANGEFS_TYPE_DATAFILE = (1 << 1), + ORANGEFS_TYPE_DIRECTORY = (1 << 2), + ORANGEFS_TYPE_SYMLINK = (1 << 3), + ORANGEFS_TYPE_DIRDATA = (1 << 4), + ORANGEFS_TYPE_INTERNAL = (1 << 5) /* for the server's private use */ +}; + +/* + * ORANGEFS_certificate simply stores a buffer with the buffer size. + * The buffer can be converted to an OpenSSL X509 struct for use. + */ +struct ORANGEFS_certificate { + __u32 buf_size; + unsigned char *buf; +}; + +/* + * A credential identifies a user and is signed by the client/user + * private key. + */ +struct ORANGEFS_credential { + __u32 userid; /* user id */ + __u32 num_groups; /* length of group_array */ + __u32 *group_array; /* groups for which the user is a member */ + char *issuer; /* alias of the issuing server */ + __u64 timeout; /* seconds after epoch to time out */ + __u32 sig_size; /* length of the signature in bytes */ + unsigned char *signature; /* digital signature */ + struct ORANGEFS_certificate certificate; /* user certificate buffer */ +}; +#define extra_size_ORANGEFS_credential (ORANGEFS_REQ_LIMIT_GROUPS * \ + sizeof(__u32) + \ + ORANGEFS_REQ_LIMIT_ISSUER + \ + ORANGEFS_REQ_LIMIT_SIGNATURE + \ + extra_size_ORANGEFS_certificate) + +/* This structure is used by the VFS-client interaction alone */ +struct ORANGEFS_keyval_pair { + char key[ORANGEFS_MAX_XATTR_NAMELEN]; + __s32 key_sz; /* __s32 for portable, fixed-size structures */ + __s32 val_sz; + char val[ORANGEFS_MAX_XATTR_VALUELEN]; +}; + +/* pvfs2-sysint.h ***********************************************************/ +/* Describes attributes for a file, directory, or symlink. */ +struct ORANGEFS_sys_attr_s { + __u32 owner; + __u32 group; + __u32 perms; + __u64 atime; + __u64 mtime; + __u64 ctime; + __s64 size; + + /* NOTE: caller must free if valid */ + char *link_target; + + /* Changed to __s32 so that size of structure does not change */ + __s32 dfile_count; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_servers_initial; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_servers_max; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_split_size; + + __u32 mirror_copies_count; + + /* NOTE: caller must free if valid */ + char *dist_name; + + /* NOTE: caller must free if valid */ + char *dist_params; + + __s64 dirent_count; + enum orangefs_ds_type objtype; + __u64 flags; + __u32 mask; + __s64 blksize; +}; + +#define ORANGEFS_LOOKUP_LINK_NO_FOLLOW 0 + +/* pint-dev.h ***************************************************************/ + +/* parameter structure used in ORANGEFS_DEV_DEBUG ioctl command */ +struct dev_mask_info_s { + enum { + KERNEL_MASK, + CLIENT_MASK, + } mask_type; + __u64 mask_value; +}; + +struct dev_mask2_info_s { + __u64 mask1_value; + __u64 mask2_value; +}; + +/* pvfs2-util.h *************************************************************/ +__s32 ORANGEFS_util_translate_mode(int mode); + +/* pvfs2-debug.h ************************************************************/ +#include "orangefs-debug.h" + +/* pvfs2-internal.h *********************************************************/ +#define llu(x) (unsigned long long)(x) +#define lld(x) (long long)(x) + +/* pint-dev-shared.h ********************************************************/ +#define ORANGEFS_DEV_MAGIC 'k' + +#define ORANGEFS_READDIR_DEFAULT_DESC_COUNT 5 + +#define DEV_GET_MAGIC 0x1 +#define DEV_GET_MAX_UPSIZE 0x2 +#define DEV_GET_MAX_DOWNSIZE 0x3 +#define DEV_MAP 0x4 +#define DEV_REMOUNT_ALL 0x5 +#define DEV_DEBUG 0x6 +#define DEV_UPSTREAM 0x7 +#define DEV_CLIENT_MASK 0x8 +#define DEV_CLIENT_STRING 0x9 +#define DEV_MAX_NR 0xa + +/* supported ioctls, codes are with respect to user-space */ +enum { + ORANGEFS_DEV_GET_MAGIC = _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAGIC, __s32), + ORANGEFS_DEV_GET_MAX_UPSIZE = + _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAX_UPSIZE, __s32), + ORANGEFS_DEV_GET_MAX_DOWNSIZE = + _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAX_DOWNSIZE, __s32), + ORANGEFS_DEV_MAP = _IO(ORANGEFS_DEV_MAGIC, DEV_MAP), + ORANGEFS_DEV_REMOUNT_ALL = _IO(ORANGEFS_DEV_MAGIC, DEV_REMOUNT_ALL), + ORANGEFS_DEV_DEBUG = _IOR(ORANGEFS_DEV_MAGIC, DEV_DEBUG, __s32), + ORANGEFS_DEV_UPSTREAM = _IOW(ORANGEFS_DEV_MAGIC, DEV_UPSTREAM, int), + ORANGEFS_DEV_CLIENT_MASK = _IOW(ORANGEFS_DEV_MAGIC, + DEV_CLIENT_MASK, + struct dev_mask2_info_s), + ORANGEFS_DEV_CLIENT_STRING = _IOW(ORANGEFS_DEV_MAGIC, + DEV_CLIENT_STRING, + char *), + ORANGEFS_DEV_MAXNR = DEV_MAX_NR, +}; + +/* + * version number for use in communicating between kernel space and user + * space. Zero signifies the upstream version of the kernel module. + */ +#define ORANGEFS_KERNEL_PROTO_VERSION 0 +#define ORANGEFS_MINIMUM_USERSPACE_VERSION 20904 + +/* + * describes memory regions to map in the ORANGEFS_DEV_MAP ioctl. + * NOTE: See devorangefs-req.c for 32 bit compat structure. + * Since this structure has a variable-sized layout that is different + * on 32 and 64 bit platforms, we need to normalize to a 64 bit layout + * on such systems before servicing ioctl calls from user-space binaries + * that may be 32 bit! + */ +struct ORANGEFS_dev_map_desc { + void *ptr; + __s32 total_size; + __s32 size; + __s32 count; +}; + +/* gossip.h *****************************************************************/ + +#ifdef GOSSIP_DISABLE_DEBUG +#define gossip_debug(mask, format, f...) do {} while (0) +#else +extern __u64 gossip_debug_mask; +extern struct client_debug_mask client_debug_mask; + +/* try to avoid function call overhead by checking masks in macro */ +#define gossip_debug(mask, format, f...) \ +do { \ + if (gossip_debug_mask & mask) \ + printk(format, ##f); \ +} while (0) +#endif /* GOSSIP_DISABLE_DEBUG */ + +/* do file and line number printouts w/ the GNU preprocessor */ +#define gossip_ldebug(mask, format, f...) \ + gossip_debug(mask, "%s: " format, __func__, ##f) + +#define gossip_err printk +#define gossip_lerr(format, f...) \ + gossip_err("%s line %d: " format, \ + __FILE__, \ + __LINE__, \ + ##f) diff --git a/fs/orangefs/super.c b/fs/orangefs/super.c new file mode 100644 index 000000000000..b9da9a0281c9 --- /dev/null +++ b/fs/orangefs/super.c @@ -0,0 +1,559 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +#include <linux/parser.h> + +/* a cache for orangefs-inode objects (i.e. orangefs inode private data) */ +static struct kmem_cache *orangefs_inode_cache; + +/* list for storing orangefs specific superblocks in use */ +LIST_HEAD(orangefs_superblocks); + +DEFINE_SPINLOCK(orangefs_superblocks_lock); + +enum { + Opt_intr, + Opt_acl, + Opt_local_lock, + + Opt_err +}; + +static const match_table_t tokens = { + { Opt_acl, "acl" }, + { Opt_intr, "intr" }, + { Opt_local_lock, "local_lock" }, + { Opt_err, NULL } +}; + + +static int parse_mount_options(struct super_block *sb, char *options, + int silent) +{ + struct orangefs_sb_info_s *orangefs_sb = ORANGEFS_SB(sb); + substring_t args[MAX_OPT_ARGS]; + char *p; + + /* + * Force any potential flags that might be set from the mount + * to zero, ie, initialize to unset. + */ + sb->s_flags &= ~MS_POSIXACL; + orangefs_sb->flags &= ~ORANGEFS_OPT_INTR; + orangefs_sb->flags &= ~ORANGEFS_OPT_LOCAL_LOCK; + + while ((p = strsep(&options, ",")) != NULL) { + int token; + + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_acl: + sb->s_flags |= MS_POSIXACL; + break; + case Opt_intr: + orangefs_sb->flags |= ORANGEFS_OPT_INTR; + break; + case Opt_local_lock: + orangefs_sb->flags |= ORANGEFS_OPT_LOCAL_LOCK; + break; + default: + goto fail; + } + } + + return 0; +fail: + if (!silent) + gossip_err("Error: mount option [%s] is not supported.\n", p); + return -EINVAL; +} + +static void orangefs_inode_cache_ctor(void *req) +{ + struct orangefs_inode_s *orangefs_inode = req; + + inode_init_once(&orangefs_inode->vfs_inode); + init_rwsem(&orangefs_inode->xattr_sem); + + orangefs_inode->vfs_inode.i_version = 1; +} + +static struct inode *orangefs_alloc_inode(struct super_block *sb) +{ + struct orangefs_inode_s *orangefs_inode; + + orangefs_inode = kmem_cache_alloc(orangefs_inode_cache, GFP_KERNEL); + if (orangefs_inode == NULL) { + gossip_err("Failed to allocate orangefs_inode\n"); + return NULL; + } + + /* + * We want to clear everything except for rw_semaphore and the + * vfs_inode. + */ + memset(&orangefs_inode->refn.khandle, 0, 16); + orangefs_inode->refn.fs_id = ORANGEFS_FS_ID_NULL; + orangefs_inode->last_failed_block_index_read = 0; + memset(orangefs_inode->link_target, 0, sizeof(orangefs_inode->link_target)); + orangefs_inode->pinode_flags = 0; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_alloc_inode: allocated %p\n", + &orangefs_inode->vfs_inode); + return &orangefs_inode->vfs_inode; +} + +static void orangefs_destroy_inode(struct inode *inode) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "%s: deallocated %p destroying inode %pU\n", + __func__, orangefs_inode, get_khandle_from_ino(inode)); + + kmem_cache_free(orangefs_inode_cache, orangefs_inode); +} + +/* + * NOTE: information filled in here is typically reflected in the + * output of the system command 'df' +*/ +static int orangefs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + int ret = -ENOMEM; + struct orangefs_kernel_op_s *new_op = NULL; + int flags = 0; + struct super_block *sb = NULL; + + sb = dentry->d_sb; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_statfs: called on sb %p (fs_id is %d)\n", + sb, + (int)(ORANGEFS_SB(sb)->fs_id)); + + new_op = op_alloc(ORANGEFS_VFS_OP_STATFS); + if (!new_op) + return ret; + new_op->upcall.req.statfs.fs_id = ORANGEFS_SB(sb)->fs_id; + + if (ORANGEFS_SB(sb)->flags & ORANGEFS_OPT_INTR) + flags = ORANGEFS_OP_INTERRUPTIBLE; + + ret = service_operation(new_op, "orangefs_statfs", flags); + + if (new_op->downcall.status < 0) + goto out_op_release; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "%s: got %ld blocks available | " + "%ld blocks total | %ld block size | " + "%ld files total | %ld files avail\n", + __func__, + (long)new_op->downcall.resp.statfs.blocks_avail, + (long)new_op->downcall.resp.statfs.blocks_total, + (long)new_op->downcall.resp.statfs.block_size, + (long)new_op->downcall.resp.statfs.files_total, + (long)new_op->downcall.resp.statfs.files_avail); + + buf->f_type = sb->s_magic; + memcpy(&buf->f_fsid, &ORANGEFS_SB(sb)->fs_id, sizeof(buf->f_fsid)); + buf->f_bsize = new_op->downcall.resp.statfs.block_size; + buf->f_namelen = ORANGEFS_NAME_MAX; + + buf->f_blocks = (sector_t) new_op->downcall.resp.statfs.blocks_total; + buf->f_bfree = (sector_t) new_op->downcall.resp.statfs.blocks_avail; + buf->f_bavail = (sector_t) new_op->downcall.resp.statfs.blocks_avail; + buf->f_files = (sector_t) new_op->downcall.resp.statfs.files_total; + buf->f_ffree = (sector_t) new_op->downcall.resp.statfs.files_avail; + buf->f_frsize = sb->s_blocksize; + +out_op_release: + op_release(new_op); + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_statfs: returning %d\n", ret); + return ret; +} + +/* + * Remount as initiated by VFS layer. We just need to reparse the mount + * options, no need to signal pvfs2-client-core about it. + */ +static int orangefs_remount_fs(struct super_block *sb, int *flags, char *data) +{ + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_remount_fs: called\n"); + return parse_mount_options(sb, data, 1); +} + +/* + * Remount as initiated by pvfs2-client-core on restart. This is used to + * repopulate mount information left from previous pvfs2-client-core. + * + * the idea here is that given a valid superblock, we're + * re-initializing the user space client with the initial mount + * information specified when the super block was first initialized. + * this is very different than the first initialization/creation of a + * superblock. we use the special service_priority_operation to make + * sure that the mount gets ahead of any other pending operation that + * is waiting for servicing. this means that the pvfs2-client won't + * fail to start several times for all other pending operations before + * the client regains all of the mount information from us. + * NOTE: this function assumes that the request_mutex is already acquired! + */ +int orangefs_remount(struct orangefs_sb_info_s *orangefs_sb) +{ + struct orangefs_kernel_op_s *new_op; + int ret = -EINVAL; + + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_remount: called\n"); + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT); + if (!new_op) + return -ENOMEM; + strncpy(new_op->upcall.req.fs_mount.orangefs_config_server, + orangefs_sb->devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Attempting ORANGEFS Remount via host %s\n", + new_op->upcall.req.fs_mount.orangefs_config_server); + + /* + * we assume that the calling function has already acquired the + * request_mutex to prevent other operations from bypassing + * this one + */ + ret = service_operation(new_op, "orangefs_remount", + ORANGEFS_OP_PRIORITY | ORANGEFS_OP_NO_MUTEX); + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_remount: mount got return value of %d\n", + ret); + if (ret == 0) { + /* + * store the id assigned to this sb -- it's just a + * short-lived mapping that the system interface uses + * to map this superblock to a particular mount entry + */ + orangefs_sb->id = new_op->downcall.resp.fs_mount.id; + orangefs_sb->mount_pending = 0; + } + + op_release(new_op); + return ret; +} + +int fsid_key_table_initialize(void) +{ + return 0; +} + +void fsid_key_table_finalize(void) +{ +} + +/* Called whenever the VFS dirties the inode in response to atime updates */ +static void orangefs_dirty_inode(struct inode *inode, int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_dirty_inode: %pU\n", + get_khandle_from_ino(inode)); + SetAtimeFlag(orangefs_inode); +} + +static const struct super_operations orangefs_s_ops = { + .alloc_inode = orangefs_alloc_inode, + .destroy_inode = orangefs_destroy_inode, + .dirty_inode = orangefs_dirty_inode, + .drop_inode = generic_delete_inode, + .statfs = orangefs_statfs, + .remount_fs = orangefs_remount_fs, + .show_options = generic_show_options, +}; + +static struct dentry *orangefs_fh_to_dentry(struct super_block *sb, + struct fid *fid, + int fh_len, + int fh_type) +{ + struct orangefs_object_kref refn; + + if (fh_len < 5 || fh_type > 2) + return NULL; + + ORANGEFS_khandle_from(&(refn.khandle), fid->raw, 16); + refn.fs_id = (u32) fid->raw[4]; + gossip_debug(GOSSIP_SUPER_DEBUG, + "fh_to_dentry: handle %pU, fs_id %d\n", + &refn.khandle, + refn.fs_id); + + return d_obtain_alias(orangefs_iget(sb, &refn)); +} + +static int orangefs_encode_fh(struct inode *inode, + __u32 *fh, + int *max_len, + struct inode *parent) +{ + int len = parent ? 10 : 5; + int type = 1; + struct orangefs_object_kref refn; + + if (*max_len < len) { + gossip_lerr("fh buffer is too small for encoding\n"); + *max_len = len; + type = 255; + goto out; + } + + refn = ORANGEFS_I(inode)->refn; + ORANGEFS_khandle_to(&refn.khandle, fh, 16); + fh[4] = refn.fs_id; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Encoding fh: handle %pU, fsid %u\n", + &refn.khandle, + refn.fs_id); + + + if (parent) { + refn = ORANGEFS_I(parent)->refn; + ORANGEFS_khandle_to(&refn.khandle, (char *) fh + 20, 16); + fh[9] = refn.fs_id; + + type = 2; + gossip_debug(GOSSIP_SUPER_DEBUG, + "Encoding parent: handle %pU, fsid %u\n", + &refn.khandle, + refn.fs_id); + } + *max_len = len; + +out: + return type; +} + +static const struct export_operations orangefs_export_ops = { + .encode_fh = orangefs_encode_fh, + .fh_to_dentry = orangefs_fh_to_dentry, +}; + +static int orangefs_fill_sb(struct super_block *sb, + struct orangefs_fs_mount_response *fs_mount, + void *data, int silent) +{ + int ret = -EINVAL; + struct inode *root = NULL; + struct dentry *root_dentry = NULL; + struct orangefs_object_kref root_object; + + /* alloc and init our private orangefs sb info */ + sb->s_fs_info = kzalloc(sizeof(struct orangefs_sb_info_s), GFP_KERNEL); + if (!ORANGEFS_SB(sb)) + return -ENOMEM; + ORANGEFS_SB(sb)->sb = sb; + + ORANGEFS_SB(sb)->root_khandle = fs_mount->root_khandle; + ORANGEFS_SB(sb)->fs_id = fs_mount->fs_id; + ORANGEFS_SB(sb)->id = fs_mount->id; + + if (data) { + ret = parse_mount_options(sb, data, silent); + if (ret) + return ret; + } + + /* Hang the xattr handlers off the superblock */ + sb->s_xattr = orangefs_xattr_handlers; + sb->s_magic = ORANGEFS_SUPER_MAGIC; + sb->s_op = &orangefs_s_ops; + sb->s_d_op = &orangefs_dentry_operations; + + sb->s_blocksize = orangefs_bufmap_size_query(); + sb->s_blocksize_bits = orangefs_bufmap_shift_query(); + sb->s_maxbytes = MAX_LFS_FILESIZE; + + root_object.khandle = ORANGEFS_SB(sb)->root_khandle; + root_object.fs_id = ORANGEFS_SB(sb)->fs_id; + gossip_debug(GOSSIP_SUPER_DEBUG, + "get inode %pU, fsid %d\n", + &root_object.khandle, + root_object.fs_id); + + root = orangefs_iget(sb, &root_object); + if (IS_ERR(root)) + return PTR_ERR(root); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Allocated root inode [%p] with mode %x\n", + root, + root->i_mode); + + /* allocates and places root dentry in dcache */ + root_dentry = d_make_root(root); + if (!root_dentry) + return -ENOMEM; + + sb->s_export_op = &orangefs_export_ops; + sb->s_root = root_dentry; + return 0; +} + +struct dentry *orangefs_mount(struct file_system_type *fst, + int flags, + const char *devname, + void *data) +{ + int ret = -EINVAL; + struct super_block *sb = ERR_PTR(-EINVAL); + struct orangefs_kernel_op_s *new_op; + struct dentry *d = ERR_PTR(-EINVAL); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_mount: called with devname %s\n", + devname); + + if (!devname) { + gossip_err("ERROR: device name not specified.\n"); + return ERR_PTR(-EINVAL); + } + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT); + if (!new_op) + return ERR_PTR(-ENOMEM); + + strncpy(new_op->upcall.req.fs_mount.orangefs_config_server, + devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Attempting ORANGEFS Mount via host %s\n", + new_op->upcall.req.fs_mount.orangefs_config_server); + + ret = service_operation(new_op, "orangefs_mount", 0); + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_mount: mount got return value of %d\n", ret); + if (ret) + goto free_op; + + if (new_op->downcall.resp.fs_mount.fs_id == ORANGEFS_FS_ID_NULL) { + gossip_err("ERROR: Retrieved null fs_id\n"); + ret = -EINVAL; + goto free_op; + } + + sb = sget(fst, NULL, set_anon_super, flags, NULL); + + if (IS_ERR(sb)) { + d = ERR_CAST(sb); + goto free_op; + } + + ret = orangefs_fill_sb(sb, + &new_op->downcall.resp.fs_mount, data, + flags & MS_SILENT ? 1 : 0); + + if (ret) { + d = ERR_PTR(ret); + goto free_op; + } + + /* + * on successful mount, store the devname and data + * used + */ + strncpy(ORANGEFS_SB(sb)->devname, + devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + /* mount_pending must be cleared */ + ORANGEFS_SB(sb)->mount_pending = 0; + + /* + * finally, add this sb to our list of known orangefs + * sb's + */ + gossip_debug(GOSSIP_SUPER_DEBUG, + "Adding SB %p to orangefs superblocks\n", + ORANGEFS_SB(sb)); + spin_lock(&orangefs_superblocks_lock); + list_add_tail(&ORANGEFS_SB(sb)->list, &orangefs_superblocks); + spin_unlock(&orangefs_superblocks_lock); + op_release(new_op); + return dget(sb->s_root); + +free_op: + gossip_err("orangefs_mount: mount request failed with %d\n", ret); + if (ret == -EINVAL) { + gossip_err("Ensure that all orangefs-servers have the same FS configuration files\n"); + gossip_err("Look at pvfs2-client-core log file (typically /tmp/pvfs2-client.log) for more details\n"); + } + + op_release(new_op); + + return d; +} + +void orangefs_kill_sb(struct super_block *sb) +{ + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_kill_sb: called\n"); + + /* provided sb cleanup */ + kill_anon_super(sb); + + /* + * issue the unmount to userspace to tell it to remove the + * dynamic mount info it has for this superblock + */ + orangefs_unmount_sb(sb); + + /* remove the sb from our list of orangefs specific sb's */ + + spin_lock(&orangefs_superblocks_lock); + __list_del_entry(&ORANGEFS_SB(sb)->list); /* not list_del_init */ + ORANGEFS_SB(sb)->list.prev = NULL; + spin_unlock(&orangefs_superblocks_lock); + + /* + * make sure that ORANGEFS_DEV_REMOUNT_ALL loop that might've seen us + * gets completed before we free the dang thing. + */ + mutex_lock(&request_mutex); + mutex_unlock(&request_mutex); + + /* free the orangefs superblock private data */ + kfree(ORANGEFS_SB(sb)); +} + +int orangefs_inode_cache_initialize(void) +{ + orangefs_inode_cache = kmem_cache_create("orangefs_inode_cache", + sizeof(struct orangefs_inode_s), + 0, + ORANGEFS_CACHE_CREATE_FLAGS, + orangefs_inode_cache_ctor); + + if (!orangefs_inode_cache) { + gossip_err("Cannot create orangefs_inode_cache\n"); + return -ENOMEM; + } + return 0; +} + +int orangefs_inode_cache_finalize(void) +{ + kmem_cache_destroy(orangefs_inode_cache); + return 0; +} diff --git a/fs/orangefs/symlink.c b/fs/orangefs/symlink.c new file mode 100644 index 000000000000..6418dd638680 --- /dev/null +++ b/fs/orangefs/symlink.c @@ -0,0 +1,19 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +struct inode_operations orangefs_symlink_inode_operations = { + .readlink = generic_readlink, + .get_link = simple_get_link, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .listxattr = orangefs_listxattr, + .setxattr = generic_setxattr, + .permission = orangefs_permission, +}; diff --git a/fs/orangefs/upcall.h b/fs/orangefs/upcall.h new file mode 100644 index 000000000000..001b20239407 --- /dev/null +++ b/fs/orangefs/upcall.h @@ -0,0 +1,246 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef __UPCALL_H +#define __UPCALL_H + +/* + * Sanitized this header file to fix + * 32-64 bit interaction issues between + * client-core and device + */ +struct orangefs_io_request_s { + __s32 __pad1; + __s32 buf_index; + __s32 count; + __s32 __pad2; + __s64 offset; + struct orangefs_object_kref refn; + enum ORANGEFS_io_type io_type; + __s32 readahead_size; +}; + +struct orangefs_lookup_request_s { + __s32 sym_follow; + __s32 __pad1; + struct orangefs_object_kref parent_refn; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_create_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_symlink_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char entry_name[ORANGEFS_NAME_MAX]; + char target[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_getattr_request_s { + struct orangefs_object_kref refn; + __u32 mask; + __u32 __pad1; +}; + +struct orangefs_setattr_request_s { + struct orangefs_object_kref refn; + struct ORANGEFS_sys_attr_s attributes; +}; + +struct orangefs_remove_request_s { + struct orangefs_object_kref parent_refn; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_mkdir_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_readdir_request_s { + struct orangefs_object_kref refn; + __u64 token; + __s32 max_dirent_count; + __s32 buf_index; +}; + +struct orangefs_readdirplus_request_s { + struct orangefs_object_kref refn; + __u64 token; + __s32 max_dirent_count; + __u32 mask; + __s32 buf_index; + __s32 __pad1; +}; + +struct orangefs_rename_request_s { + struct orangefs_object_kref old_parent_refn; + struct orangefs_object_kref new_parent_refn; + char d_old_name[ORANGEFS_NAME_MAX]; + char d_new_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_statfs_request_s { + __s32 fs_id; + __s32 __pad1; +}; + +struct orangefs_truncate_request_s { + struct orangefs_object_kref refn; + __s64 size; +}; + +struct orangefs_mmap_ra_cache_flush_request_s { + struct orangefs_object_kref refn; +}; + +struct orangefs_fs_mount_request_s { + char orangefs_config_server[ORANGEFS_MAX_SERVER_ADDR_LEN]; +}; + +struct orangefs_fs_umount_request_s { + __s32 id; + __s32 fs_id; + char orangefs_config_server[ORANGEFS_MAX_SERVER_ADDR_LEN]; +}; + +struct orangefs_getxattr_request_s { + struct orangefs_object_kref refn; + __s32 key_sz; + __s32 __pad1; + char key[ORANGEFS_MAX_XATTR_NAMELEN]; +}; + +struct orangefs_setxattr_request_s { + struct orangefs_object_kref refn; + struct ORANGEFS_keyval_pair keyval; + __s32 flags; + __s32 __pad1; +}; + +struct orangefs_listxattr_request_s { + struct orangefs_object_kref refn; + __s32 requested_count; + __s32 __pad1; + __u64 token; +}; + +struct orangefs_removexattr_request_s { + struct orangefs_object_kref refn; + __s32 key_sz; + __s32 __pad1; + char key[ORANGEFS_MAX_XATTR_NAMELEN]; +}; + +struct orangefs_op_cancel_s { + __u64 op_tag; +}; + +struct orangefs_fsync_request_s { + struct orangefs_object_kref refn; +}; + +enum orangefs_param_request_type { + ORANGEFS_PARAM_REQUEST_SET = 1, + ORANGEFS_PARAM_REQUEST_GET = 2 +}; + +enum orangefs_param_request_op { + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS = 1, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT = 2, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT = 3, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE = 4, + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS = 5, + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE = 6, + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET = 7, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS = 8, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT = 9, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT = 10, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE = 11, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_TIMEOUT_MSECS = 12, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_HARD_LIMIT = 13, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_SOFT_LIMIT = 14, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_RECLAIM_PERCENTAGE = 15, + ORANGEFS_PARAM_REQUEST_OP_CLIENT_DEBUG = 16, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS = 17, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT = 18, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT = 19, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE = 20, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS = 21, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT = 22, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT = 23, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE = 24, + ORANGEFS_PARAM_REQUEST_OP_TWO_MASK_VALUES = 25, +}; + +struct orangefs_param_request_s { + enum orangefs_param_request_type type; + enum orangefs_param_request_op op; + __s64 value; + char s_value[ORANGEFS_MAX_DEBUG_STRING_LEN]; +}; + +enum orangefs_perf_count_request_type { + ORANGEFS_PERF_COUNT_REQUEST_ACACHE = 1, + ORANGEFS_PERF_COUNT_REQUEST_NCACHE = 2, + ORANGEFS_PERF_COUNT_REQUEST_CAPCACHE = 3, +}; + +struct orangefs_perf_count_request_s { + enum orangefs_perf_count_request_type type; + __s32 __pad1; +}; + +struct orangefs_fs_key_request_s { + __s32 fsid; + __s32 __pad1; +}; + +struct orangefs_upcall_s { + __s32 type; + __u32 uid; + __u32 gid; + int pid; + int tgid; + /* Trailers unused but must be retained for protocol compatibility. */ + __s64 trailer_size; + char *trailer_buf; + + union { + struct orangefs_io_request_s io; + struct orangefs_lookup_request_s lookup; + struct orangefs_create_request_s create; + struct orangefs_symlink_request_s sym; + struct orangefs_getattr_request_s getattr; + struct orangefs_setattr_request_s setattr; + struct orangefs_remove_request_s remove; + struct orangefs_mkdir_request_s mkdir; + struct orangefs_readdir_request_s readdir; + struct orangefs_readdirplus_request_s readdirplus; + struct orangefs_rename_request_s rename; + struct orangefs_statfs_request_s statfs; + struct orangefs_truncate_request_s truncate; + struct orangefs_mmap_ra_cache_flush_request_s ra_cache_flush; + struct orangefs_fs_mount_request_s fs_mount; + struct orangefs_fs_umount_request_s fs_umount; + struct orangefs_getxattr_request_s getxattr; + struct orangefs_setxattr_request_s setxattr; + struct orangefs_listxattr_request_s listxattr; + struct orangefs_removexattr_request_s removexattr; + struct orangefs_op_cancel_s cancel; + struct orangefs_fsync_request_s fsync; + struct orangefs_param_request_s param; + struct orangefs_perf_count_request_s perf_count; + struct orangefs_fs_key_request_s fs_key; + } req; +}; + +#endif /* __UPCALL_H */ diff --git a/fs/orangefs/waitqueue.c b/fs/orangefs/waitqueue.c new file mode 100644 index 000000000000..31635bc303fe --- /dev/null +++ b/fs/orangefs/waitqueue.c @@ -0,0 +1,357 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * (C) 2011 Omnibond Systems + * + * Changes by Acxiom Corporation to implement generic service_operation() + * function, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +/* + * In-kernel waitqueue operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +static int wait_for_matching_downcall(struct orangefs_kernel_op_s *, long, bool); +static void orangefs_clean_up_interrupted_operation(struct orangefs_kernel_op_s *); + +/* + * What we do in this function is to walk the list of operations that are + * present in the request queue and mark them as purged. + * NOTE: This is called from the device close after client-core has + * guaranteed that no new operations could appear on the list since the + * client-core is anyway going to exit. + */ +void purge_waiting_ops(void) +{ + struct orangefs_kernel_op_s *op; + + spin_lock(&orangefs_request_list_lock); + list_for_each_entry(op, &orangefs_request_list, list) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "pvfs2-client-core: purging op tag %llu %s\n", + llu(op->tag), + get_opname_string(op)); + set_op_state_purged(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + } + spin_unlock(&orangefs_request_list_lock); +} + +/* + * submits a ORANGEFS operation and waits for it to complete + * + * Note op->downcall.status will contain the status of the operation (in + * errno format), whether provided by pvfs2-client or a result of failure to + * service the operation. If the caller wishes to distinguish, then + * op->state can be checked to see if it was serviced or not. + * + * Returns contents of op->downcall.status for convenience + */ +int service_operation(struct orangefs_kernel_op_s *op, + const char *op_name, + int flags) +{ + long timeout = MAX_SCHEDULE_TIMEOUT; + int ret = 0; + + DEFINE_WAIT(wait_entry); + + op->upcall.tgid = current->tgid; + op->upcall.pid = current->pid; + +retry_servicing: + op->downcall.status = 0; + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: %s op:%p: process:%s: pid:%d:\n", + __func__, + op_name, + op, + current->comm, + current->pid); + + /* + * If ORANGEFS_OP_NO_MUTEX was set in flags, we need to avoid + * acquiring the request_mutex because we're servicing a + * high priority remount operation and the request_mutex is + * already taken. + */ + if (!(flags & ORANGEFS_OP_NO_MUTEX)) { + if (flags & ORANGEFS_OP_INTERRUPTIBLE) + ret = mutex_lock_interruptible(&request_mutex); + else + ret = mutex_lock_killable(&request_mutex); + /* + * check to see if we were interrupted while waiting for + * mutex + */ + if (ret < 0) { + op->downcall.status = ret; + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: service_operation interrupted.\n", + __func__); + return ret; + } + } + + /* queue up the operation */ + spin_lock(&orangefs_request_list_lock); + spin_lock(&op->lock); + set_op_state_waiting(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + /* add high priority remount op to the front of the line. */ + if (flags & ORANGEFS_OP_PRIORITY) + list_add(&op->list, &orangefs_request_list); + else + list_add_tail(&op->list, &orangefs_request_list); + spin_unlock(&op->lock); + wake_up_interruptible(&orangefs_request_list_waitq); + if (!__is_daemon_in_service()) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s:client core is NOT in service.\n", + __func__); + timeout = op_timeout_secs * HZ; + } + spin_unlock(&orangefs_request_list_lock); + + if (!(flags & ORANGEFS_OP_NO_MUTEX)) + mutex_unlock(&request_mutex); + + ret = wait_for_matching_downcall(op, timeout, + flags & ORANGEFS_OP_INTERRUPTIBLE); + + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: wait_for_matching_downcall returned %d for %p\n", + __func__, + ret, + op); + + /* got matching downcall; make sure status is in errno format */ + if (!ret) { + spin_unlock(&op->lock); + op->downcall.status = + orangefs_normalize_to_errno(op->downcall.status); + ret = op->downcall.status; + goto out; + } + + /* failed to get matching downcall */ + if (ret == -ETIMEDOUT) { + gossip_err("%s: %s -- wait timed out; aborting attempt.\n", + __func__, + op_name); + } + + /* + * remove a waiting op from the request list or + * remove an in-progress op from the in-progress list. + */ + orangefs_clean_up_interrupted_operation(op); + + op->downcall.status = ret; + /* retry if operation has not been serviced and if requested */ + if (ret == -EAGAIN) { + op->attempts++; + timeout = op_timeout_secs * HZ; + gossip_debug(GOSSIP_WAIT_DEBUG, + "orangefs: tag %llu (%s)" + " -- operation to be retried (%d attempt)\n", + llu(op->tag), + op_name, + op->attempts); + + /* + * io ops (ops that use the shared memory buffer) have + * to be returned to their caller for a retry. Other ops + * can just be recycled here. + */ + if (!op->uses_shared_memory) + goto retry_servicing; + } + +out: + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: %s returning: %d for %p.\n", + __func__, + op_name, + ret, + op); + return ret; +} + +/* This can get called on an I/O op if it had a bad service_operation. */ +bool orangefs_cancel_op_in_progress(struct orangefs_kernel_op_s *op) +{ + u64 tag = op->tag; + if (!op_state_in_progress(op)) + return false; + + op->slot_to_free = op->upcall.req.io.buf_index; + memset(&op->upcall, 0, sizeof(op->upcall)); + memset(&op->downcall, 0, sizeof(op->downcall)); + op->upcall.type = ORANGEFS_VFS_OP_CANCEL; + op->upcall.req.cancel.op_tag = tag; + op->downcall.type = ORANGEFS_VFS_OP_INVALID; + op->downcall.status = -1; + orangefs_new_tag(op); + + spin_lock(&orangefs_request_list_lock); + /* orangefs_request_list_lock is enough of a barrier here */ + if (!__is_daemon_in_service()) { + spin_unlock(&orangefs_request_list_lock); + return false; + } + spin_lock(&op->lock); + set_op_state_waiting(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + list_add(&op->list, &orangefs_request_list); + spin_unlock(&op->lock); + spin_unlock(&orangefs_request_list_lock); + + gossip_debug(GOSSIP_WAIT_DEBUG, + "Attempting ORANGEFS operation cancellation of tag %llu\n", + llu(tag)); + return true; +} + +/* + * Change an op to the "given up" state and remove it from its list. + */ +static void + orangefs_clean_up_interrupted_operation(struct orangefs_kernel_op_s *op) +{ + /* + * handle interrupted cases depending on what state we were in when + * the interruption is detected. + * + * Called with op->lock held. + */ + + /* + * List manipulation code elsewhere will ignore ops that + * have been given up upon. + */ + op->op_state |= OP_VFS_STATE_GIVEN_UP; + + if (list_empty(&op->list)) { + /* caught copying to/from daemon */ + BUG_ON(op_state_serviced(op)); + spin_unlock(&op->lock); + wait_for_completion(&op->waitq); + } else if (op_state_waiting(op)) { + /* + * upcall hasn't been read; remove op from upcall request + * list. + */ + spin_unlock(&op->lock); + spin_lock(&orangefs_request_list_lock); + list_del_init(&op->list); + spin_unlock(&orangefs_request_list_lock); + gossip_debug(GOSSIP_WAIT_DEBUG, + "Interrupted: Removed op %p from request_list\n", + op); + } else if (op_state_in_progress(op)) { + /* op must be removed from the in progress htable */ + spin_unlock(&op->lock); + spin_lock(&htable_ops_in_progress_lock); + list_del_init(&op->list); + spin_unlock(&htable_ops_in_progress_lock); + gossip_debug(GOSSIP_WAIT_DEBUG, + "Interrupted: Removed op %p" + " from htable_ops_in_progress\n", + op); + } else { + spin_unlock(&op->lock); + gossip_err("interrupted operation is in a weird state 0x%x\n", + op->op_state); + } + reinit_completion(&op->waitq); +} + +/* + * Sleeps on waitqueue waiting for matching downcall. + * If client-core finishes servicing, then we are good to go. + * else if client-core exits, we get woken up here, and retry with a timeout + * + * When this call returns to the caller, the specified op will no + * longer be in either the in_progress hash table or on the request list. + * + * Returns 0 on success and -errno on failure + * Errors are: + * EAGAIN in case we want the caller to requeue and try again.. + * EINTR/EIO/ETIMEDOUT indicating we are done trying to service this + * operation since client-core seems to be exiting too often + * or if we were interrupted. + * + * Returns with op->lock taken. + */ +static int wait_for_matching_downcall(struct orangefs_kernel_op_s *op, + long timeout, + bool interruptible) +{ + long n; + + /* + * There's a "schedule_timeout" inside of these wait + * primitives, during which the op is out of the hands of the + * user process that needs something done and is being + * manipulated by the client-core process. + */ + if (interruptible) + n = wait_for_completion_interruptible_timeout(&op->waitq, + timeout); + else + n = wait_for_completion_killable_timeout(&op->waitq, timeout); + + spin_lock(&op->lock); + + if (op_state_serviced(op)) + return 0; + + if (unlikely(n < 0)) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation interrupted, tag %llu, %p\n", + __func__, + llu(op->tag), + op); + return -EINTR; + } + if (op_state_purged(op)) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation purged, tag %llu, %p, %d\n", + __func__, + llu(op->tag), + op, + op->attempts); + return (op->attempts < ORANGEFS_PURGE_RETRY_COUNT) ? + -EAGAIN : + -EIO; + } + /* must have timed out, then... */ + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation timed out, tag %llu, %p, %d)\n", + __func__, + llu(op->tag), + op, + op->attempts); + return -ETIMEDOUT; +} diff --git a/fs/orangefs/xattr.c b/fs/orangefs/xattr.c new file mode 100644 index 000000000000..ef5da7538cd5 --- /dev/null +++ b/fs/orangefs/xattr.c @@ -0,0 +1,545 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS extended attribute operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/posix_acl_xattr.h> +#include <linux/xattr.h> + + +#define SYSTEM_ORANGEFS_KEY "system.pvfs2." +#define SYSTEM_ORANGEFS_KEY_LEN 13 + +/* + * this function returns + * 0 if the key corresponding to name is not meant to be printed as part + * of a listxattr. + * 1 if the key corresponding to name is meant to be returned as part of + * a listxattr. + * The ones that start SYSTEM_ORANGEFS_KEY are the ones to avoid printing. + */ +static int is_reserved_key(const char *key, size_t size) +{ + + if (size < SYSTEM_ORANGEFS_KEY_LEN) + return 1; + + return strncmp(key, SYSTEM_ORANGEFS_KEY, SYSTEM_ORANGEFS_KEY_LEN) ? 1 : 0; +} + +static inline int convert_to_internal_xattr_flags(int setxattr_flags) +{ + int internal_flag = 0; + + if (setxattr_flags & XATTR_REPLACE) { + /* Attribute must exist! */ + internal_flag = ORANGEFS_XATTR_REPLACE; + } else if (setxattr_flags & XATTR_CREATE) { + /* Attribute must not exist */ + internal_flag = ORANGEFS_XATTR_CREATE; + } + return internal_flag; +} + + +/* + * Tries to get a specified key's attributes of a given + * file into a user-specified buffer. Note that the getxattr + * interface allows for the users to probe the size of an + * extended attribute by passing in a value of 0 to size. + * Thus our return value is always the size of the attribute + * unless the key does not exist for the file and/or if + * there were errors in fetching the attribute value. + */ +ssize_t orangefs_inode_getxattr(struct inode *inode, const char *prefix, + const char *name, void *buffer, size_t size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op = NULL; + ssize_t ret = -ENOMEM; + ssize_t length = 0; + int fsuid; + int fsgid; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "%s: prefix %s name %s, buffer_size %zd\n", + __func__, prefix, name, size); + + if (name == NULL || (size > 0 && buffer == NULL)) { + gossip_err("orangefs_inode_getxattr: bogus NULL pointers\n"); + return -EINVAL; + } + if ((strlen(name) + strlen(prefix)) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err("Invalid key length (%d)\n", + (int)(strlen(name) + strlen(prefix))); + return -EINVAL; + } + + fsuid = from_kuid(current_user_ns(), current_fsuid()); + fsgid = from_kgid(current_user_ns(), current_fsgid()); + + gossip_debug(GOSSIP_XATTR_DEBUG, + "getxattr on inode %pU, name %s " + "(uid %o, gid %o)\n", + get_khandle_from_ino(inode), + name, + fsuid, + fsgid); + + down_read(&orangefs_inode->xattr_sem); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETXATTR); + if (!new_op) + goto out_unlock; + + new_op->upcall.req.getxattr.refn = orangefs_inode->refn; + ret = snprintf((char *)new_op->upcall.req.getxattr.key, + ORANGEFS_MAX_XATTR_NAMELEN, "%s%s", prefix, name); + + /* + * NOTE: Although keys are meant to be NULL terminated textual + * strings, I am going to explicitly pass the length just in case + * we change this later on... + */ + new_op->upcall.req.getxattr.key_sz = ret + 1; + + ret = service_operation(new_op, "orangefs_inode_getxattr", + get_interruptible_flag(inode)); + if (ret != 0) { + if (ret == -ENOENT) { + ret = -ENODATA; + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_getxattr: inode %pU key %s" + " does not exist!\n", + get_khandle_from_ino(inode), + (char *)new_op->upcall.req.getxattr.key); + } + goto out_release_op; + } + + /* + * Length returned includes null terminator. + */ + length = new_op->downcall.resp.getxattr.val_sz; + + /* + * Just return the length of the queried attribute. + */ + if (size == 0) { + ret = length; + goto out_release_op; + } + + /* + * Check to see if key length is > provided buffer size. + */ + if (length > size) { + ret = -ERANGE; + goto out_release_op; + } + + memset(buffer, 0, size); + memcpy(buffer, new_op->downcall.resp.getxattr.val, length); + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_getxattr: inode %pU " + "key %s key_sz %d, val_len %d\n", + get_khandle_from_ino(inode), + (char *)new_op-> + upcall.req.getxattr.key, + (int)new_op-> + upcall.req.getxattr.key_sz, + (int)ret); + + ret = length; + +out_release_op: + op_release(new_op); +out_unlock: + up_read(&orangefs_inode->xattr_sem); + return ret; +} + +static int orangefs_inode_removexattr(struct inode *inode, + const char *prefix, + const char *name, + int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op = NULL; + int ret = -ENOMEM; + + down_write(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_REMOVEXATTR); + if (!new_op) + goto out_unlock; + + new_op->upcall.req.removexattr.refn = orangefs_inode->refn; + /* + * NOTE: Although keys are meant to be NULL terminated + * textual strings, I am going to explicitly pass the + * length just in case we change this later on... + */ + ret = snprintf((char *)new_op->upcall.req.removexattr.key, + ORANGEFS_MAX_XATTR_NAMELEN, + "%s%s", + (prefix ? prefix : ""), + name); + new_op->upcall.req.removexattr.key_sz = ret + 1; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_removexattr: key %s, key_sz %d\n", + (char *)new_op->upcall.req.removexattr.key, + (int)new_op->upcall.req.removexattr.key_sz); + + ret = service_operation(new_op, + "orangefs_inode_removexattr", + get_interruptible_flag(inode)); + if (ret == -ENOENT) { + /* + * Request to replace a non-existent attribute is an error. + */ + if (flags & XATTR_REPLACE) + ret = -ENODATA; + else + ret = 0; + } + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_removexattr: returning %d\n", ret); + + op_release(new_op); +out_unlock: + up_write(&orangefs_inode->xattr_sem); + return ret; +} + +/* + * Tries to set an attribute for a given key on a file. + * + * Returns a -ve number on error and 0 on success. Key is text, but value + * can be binary! + */ +int orangefs_inode_setxattr(struct inode *inode, const char *prefix, + const char *name, const void *value, size_t size, int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int internal_flag = 0; + int ret = -ENOMEM; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "%s: prefix %s, name %s, buffer_size %zd\n", + __func__, prefix, name, size); + + if (size < 0 || + size >= ORANGEFS_MAX_XATTR_VALUELEN || + flags < 0) { + gossip_err("orangefs_inode_setxattr: bogus values of size(%d), flags(%d)\n", + (int)size, + flags); + return -EINVAL; + } + + if (name == NULL || + (size > 0 && value == NULL)) { + gossip_err("orangefs_inode_setxattr: bogus NULL pointers!\n"); + return -EINVAL; + } + + internal_flag = convert_to_internal_xattr_flags(flags); + + if (prefix) { + if (strlen(name) + strlen(prefix) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err + ("orangefs_inode_setxattr: bogus key size (%d)\n", + (int)(strlen(name) + strlen(prefix))); + return -EINVAL; + } + } else { + if (strlen(name) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err + ("orangefs_inode_setxattr: bogus key size (%d)\n", + (int)(strlen(name))); + return -EINVAL; + } + } + + /* This is equivalent to a removexattr */ + if (size == 0 && value == NULL) { + gossip_debug(GOSSIP_XATTR_DEBUG, + "removing xattr (%s%s)\n", + prefix, + name); + return orangefs_inode_removexattr(inode, prefix, name, flags); + } + + gossip_debug(GOSSIP_XATTR_DEBUG, + "setxattr on inode %pU, name %s\n", + get_khandle_from_ino(inode), + name); + + down_write(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_SETXATTR); + if (!new_op) + goto out_unlock; + + + new_op->upcall.req.setxattr.refn = orangefs_inode->refn; + new_op->upcall.req.setxattr.flags = internal_flag; + /* + * NOTE: Although keys are meant to be NULL terminated textual + * strings, I am going to explicitly pass the length just in + * case we change this later on... + */ + ret = snprintf((char *)new_op->upcall.req.setxattr.keyval.key, + ORANGEFS_MAX_XATTR_NAMELEN, + "%s%s", + prefix, name); + new_op->upcall.req.setxattr.keyval.key_sz = ret + 1; + memcpy(new_op->upcall.req.setxattr.keyval.val, value, size); + new_op->upcall.req.setxattr.keyval.val_sz = size; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_setxattr: key %s, key_sz %d " + " value size %zd\n", + (char *)new_op->upcall.req.setxattr.keyval.key, + (int)new_op->upcall.req.setxattr.keyval.key_sz, + size); + + ret = service_operation(new_op, + "orangefs_inode_setxattr", + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_setxattr: returning %d\n", + ret); + + /* when request is serviced properly, free req op struct */ + op_release(new_op); +out_unlock: + up_write(&orangefs_inode->xattr_sem); + return ret; +} + +/* + * Tries to get a specified object's keys into a user-specified buffer of a + * given size. Note that like the previous instances of xattr routines, this + * also allows you to pass in a NULL pointer and 0 size to probe the size for + * subsequent memory allocations. Thus our return value is always the size of + * all the keys unless there were errors in fetching the keys! + */ +ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size) +{ + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + __u64 token = ORANGEFS_ITERATE_START; + ssize_t ret = -ENOMEM; + ssize_t total = 0; + int count_keys = 0; + int key_size; + int i = 0; + int returned_count = 0; + + if (size > 0 && buffer == NULL) { + gossip_err("%s: bogus NULL pointers\n", __func__); + return -EINVAL; + } + if (size < 0) { + gossip_err("Invalid size (%d)\n", (int)size); + return -EINVAL; + } + + down_read(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_LISTXATTR); + if (!new_op) + goto out_unlock; + + if (buffer && size > 0) + memset(buffer, 0, size); + +try_again: + key_size = 0; + new_op->upcall.req.listxattr.refn = orangefs_inode->refn; + new_op->upcall.req.listxattr.token = token; + new_op->upcall.req.listxattr.requested_count = + (size == 0) ? 0 : ORANGEFS_MAX_XATTR_LISTLEN; + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto done; + + if (size == 0) { + /* + * This is a bit of a big upper limit, but I did not want to + * spend too much time getting this correct, since users end + * up allocating memory rather than us... + */ + total = new_op->downcall.resp.listxattr.returned_count * + ORANGEFS_MAX_XATTR_NAMELEN; + goto done; + } + + returned_count = new_op->downcall.resp.listxattr.returned_count; + if (returned_count < 0 || + returned_count >= ORANGEFS_MAX_XATTR_LISTLEN) { + gossip_err("%s: impossible value for returned_count:%d:\n", + __func__, + returned_count); + ret = -EIO; + goto done; + } + + /* + * Check to see how much can be fit in the buffer. Fit only whole keys. + */ + for (i = 0; i < returned_count; i++) { + if (new_op->downcall.resp.listxattr.lengths[i] < 0 || + new_op->downcall.resp.listxattr.lengths[i] > + ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err("%s: impossible value for lengths[%d]\n", + __func__, + new_op->downcall.resp.listxattr.lengths[i]); + ret = -EIO; + goto done; + } + if (total + new_op->downcall.resp.listxattr.lengths[i] > size) + goto done; + + /* + * Since many dumb programs try to setxattr() on our reserved + * xattrs this is a feeble attempt at defeating those by not + * listing them in the output of listxattr.. sigh + */ + if (is_reserved_key(new_op->downcall.resp.listxattr.key + + key_size, + new_op->downcall.resp. + listxattr.lengths[i])) { + gossip_debug(GOSSIP_XATTR_DEBUG, "Copying key %d -> %s\n", + i, new_op->downcall.resp.listxattr.key + + key_size); + memcpy(buffer + total, + new_op->downcall.resp.listxattr.key + key_size, + new_op->downcall.resp.listxattr.lengths[i]); + total += new_op->downcall.resp.listxattr.lengths[i]; + count_keys++; + } else { + gossip_debug(GOSSIP_XATTR_DEBUG, "[RESERVED] key %d -> %s\n", + i, new_op->downcall.resp.listxattr.key + + key_size); + } + key_size += new_op->downcall.resp.listxattr.lengths[i]; + } + + /* + * Since the buffer was large enough, we might have to continue + * fetching more keys! + */ + token = new_op->downcall.resp.listxattr.token; + if (token != ORANGEFS_ITERATE_END) + goto try_again; + +done: + gossip_debug(GOSSIP_XATTR_DEBUG, "%s: returning %d" + " [size of buffer %ld] (filled in %d keys)\n", + __func__, + ret ? (int)ret : (int)total, + (long)size, + count_keys); + op_release(new_op); + if (ret == 0) + ret = total; +out_unlock: + up_read(&orangefs_inode->xattr_sem); + return ret; +} + +static int orangefs_xattr_set_default(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + const void *buffer, + size_t size, + int flags) +{ + return orangefs_inode_setxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + name, + buffer, + size, + flags); +} + +static int orangefs_xattr_get_default(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + void *buffer, + size_t size) +{ + return orangefs_inode_getxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + name, + buffer, + size); + +} + +static int orangefs_xattr_set_trusted(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + const void *buffer, + size_t size, + int flags) +{ + return orangefs_inode_setxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + name, + buffer, + size, + flags); +} + +static int orangefs_xattr_get_trusted(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + void *buffer, + size_t size) +{ + return orangefs_inode_getxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + name, + buffer, + size); +} + +static struct xattr_handler orangefs_xattr_trusted_handler = { + .prefix = ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + .get = orangefs_xattr_get_trusted, + .set = orangefs_xattr_set_trusted, +}; + +static struct xattr_handler orangefs_xattr_default_handler = { + /* + * NOTE: this is set to be the empty string. + * so that all un-prefixed xattrs keys get caught + * here! + */ + .prefix = ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + .get = orangefs_xattr_get_default, + .set = orangefs_xattr_set_default, +}; + +const struct xattr_handler *orangefs_xattr_handlers[] = { + &posix_acl_access_xattr_handler, + &posix_acl_default_xattr_handler, + &orangefs_xattr_trusted_handler, + &orangefs_xattr_default_handler, + NULL +}; |