Subversion Repositories HelenOS

Rev

Go to most recent revision | Show changed files | Directory listing | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
3386 5873 d 17 h decky /branches/network/ add new network branch (copy of current trunk)  
3352 5885 d 13 h jermar /trunk/uspace/srv/ Add reference to the mounted FS root.

This makes it possible to keep the FS root logically unlinked (i.e. lnkcnt == 0)
because it prevents VFS from attempting to destroy it during closedir().		  
3335 5889 d 12 h jermar /trunk/uspace/srv/fs/fat/ Support for readdir() on FAT.  
3325 5892 d 13 h jermar /trunk/uspace/srv/fs/fat/ Determine the size of a directory node by walking the FAT and counting clusters.
This eliminates the hardcoded limit on the number of directory entries.		  
3314 5893 d 14 h jermar /trunk/uspace/srv/fs/fat/ fat_read() should expect IPC_M_DATA_READ instead of IPC_M_DATA_WRITE.  
3312 5893 d 14 h jermar /trunk/uspace/srv/fs/fat/ When removing a FAT node from the ffn list, use the correct pointer instead of
an uninitialized one.		  
3308 5895 d 12 h jermar /trunk/uspace/srv/fs/fat/ More complete fat_read().  
3307 5895 d 14 h jermar /trunk/uspace/srv/fs/fat/ Dummy implementation of fat_read().  
3282 5897 d 11 h jermar /trunk/uspace/srv/fs/fat/ FAT dentry type used for subdirectories doesn't store the directory size in the 'size' field.
Introduce a simple workaround that will let us have 64 directory entries at maximum.		  
3272 5898 d 0 h jermar /trunk/uspace/srv/fs/fat/ Several fixes for the FAT file system:
- do not copy padding during file name canonization
- properly nul-terminate the FAT file name during canonization
- compare FAT file name and the component name case insensitively
- block_get() takes a block offset, not byte offset as parameter		  
3257 5898 d 23 h jermar /trunk/uspace/srv/fs/fat/ First attempt at delivering the FAT16 read-only support.
Needless to say it doesn't work yet, but now it's only
the lurking bugs that need to be fixed.		  
3253 5900 d 12 h jermar /trunk/uspace/srv/fs/fat/ Pass block size to block_get() in the last argument.  
3252 5900 d 12 h jermar /trunk/uspace/ Add a comment to libfs_blockread() and change type from size_t to off_t at
various places.		  
3251 5900 d 14 h jermar /trunk/uspace/ Move tmpfs_blockread() to libfs and rename it to libfs_blockread().
Add one argument to specify block size.		  
3250 5900 d 15 h jermar /trunk/uspace/srv/fs/tmpfs/ Let tmpfs_blockread() use async_req_2_1() instead of ipc_call_sync_2_1().  
3247 5901 d 15 h jermar /trunk/uspace/srv/fs/tmpfs/ Avoid magic numbers in tmpfs_dump.c.  
3201 5935 d 20 h svoboda /trunk/uspace/ Do not disassemble needlessly.  
3119 5961 d 18 h jermar /trunk/uspace/srv/fs/fat/ Initialize in-core FAT root directory node during mount.  
3111 5962 d 12 h jermar /trunk/uspace/srv/fs/fat/ Put a frequently used construct into a function.  
3110 5962 d 13 h jermar /trunk/uspace/srv/fs/fat/ Initialize global FAT IDX structures when the FAT server starts and initialize
per-file system FAT IDX structures when a file system gets mounted.		  
3109 5962 d 18 h jermar /trunk/uspace/srv/ Split the 'mount another filesystem here' and 'you are being mounted and the
device is this' mount semantics. Add VFS_MOUNTED VFS operation that corresponds
to the latter and reserve VFS_MOUNT only for the former. Because of this
change, the VFS server does not maintain the mr_node VFS node for the name space
root anymore and the VFS_LOOKUP operation is now not meant to be used on
unmounted file system, not even for looking up the root node of unmounted file
systems. In the light of these changes, TMPFS is now initialized from
tmpfs_mounted() function.		  
3103 5963 d 10 h jermar /trunk/uspace/srv/fs/tmpfs/ Evict the code which loads the TMPFS file system from a file system dump to a
dedicated source file.		  
3091 5965 d 14 h jermar /trunk/uspace/srv/fs/tmpfs/ Return value argument to ipc_call_sync_X_Y() may not be int or it will
break on 64-machines that are sensitive to unaligned memory accesses.
Moreover, overcasting int * to sysarg_t * is not a good idea either
even if the int is by chance aligned on a 64-bit boundary as some
unknown memory will get overwritten during the assignment. Change the
type of the retval argument to ipcarg_t.		  
3087 5965 d 18 h decky /trunk/uspace/srv/fs/tmpfs/ initialize tmpfs from RAM disk  
2958 5981 d 11 h jermar /trunk/uspace/srv/ Make VFS_MOUNT call even when mounting the root file system.  
2953 5982 d 12 h jermar /trunk/uspace/srv/fs/fat/ Add synchronization to fat_match().  
2951 5982 d 13 h jermar /trunk/uspace/srv/fs/fat/ Add locks to FAT index structures, FAT in-core node structures. Add futex to
protect the list of free cached FAT in-core nodes. Make fat_node_get() and
fat_node_put() and some other functions aware of these new locks.		  
2945 5983 d 12 h jermar /trunk/uspace/srv/fs/fat/ Introduce a futex protecting the up_hash and ui_hash.  
2944 5983 d 12 h jermar /trunk/uspace/srv/fs/fat/ Introduce futex for protecting the list of unused structures in FAT server.  
2910 5987 d 17 h jermar /trunk/uspace/srv/fs/fat/ Implement fat_node_put() and remove the node from the free node list when
bumping its refernce count from zero in fat_node_get().		  
2893 5988 d 22 h jermar /trunk/uspace/srv/fs/fat/ More bits of fat_node_get().
Still needs to do some locking.		  
2891 5991 d 12 h jermar /trunk/uspace/srv/fs/fat/ New first bits of fat_node_get().  
2890 5991 d 14 h jermar /trunk/uspace/srv/fs/fat/ Need a way to locate the FAT index structure by VFS index. Also fix code that
locates the FAT index structure by position in the file system namespace.		  
2889 5994 d 13 h jermar /trunk/uspace/srv/fs/fat/ Add hash table for used fat_idx_t structures and implement fat_idx_map().  
2884 5995 d 12 h jermar /trunk/uspace/srv/fs/fat/ merge vs. coalesce  
2881 5995 d 17 h jermar /trunk/uspace/srv/fs/fat/ Code for de/allocation of VFS indices for FAT.  
2876 5995 d 20 h jermar /trunk/uspace/srv/fs/fat/ Add a dedicated source file for handling FAT indices.  
2864 6009 d 11 h jermar /trunk/uspace/srv/fs/fat/ Introduce the foundation for unique and stable FAT VFS node indices.

The first two reserved FAT entries are physically present in each FAT, so fix
the code to be aware of this.		  
2863 6009 d 22 h jermar /trunk/uspace/ Revert libfs node_get() API to a model without the parent node pointer
argument.		  
2859 6015 d 3 h jermar /trunk/uspace/srv/fs/fat/ Add fat_block_get().