Age | Commit message (Collapse) | Author | |
---|---|---|---|
1996-04-23 | (diskfs_truncate): Use read/write_disk_entry when reading/writing | Michael I. Bushnell | |
on-disk indirect blocks. (diskfs_grow): Likewise. (indir_release): Likewise. (diskfs_truncate): Use read/write_disk_entry when reading/writing on-disk inode fields. (diskfs_grow): Likewise. | |||
1996-04-18 | (diskfs_grow): New variable `pagerpt'. | Michael I. Bushnell | |
(offer_zeroes, block_extended): New functions. (diskfs_grow): In initializing newly allocated data disk blocks with zeroes, use less aggressive offer_zeroes instead of immediate synchronous writes. After ffs_realloccg succeeds, use block_extended to handle the magic. Get rid of old poke calls. | |||
1996-03-29 | (diskfs_truncate): Cast DI->di_shortlink to correct type before adding | Michael I. Bushnell | |
a character count to it. | |||
1996-01-06 | (indir_release): Use `disk_pager' in place of `diskpager->p'. | Roland McGrath | |
1995-10-19 | (diskfs_lookup, diskfs_dirempty): Give diskfs_get_filemap a protection arg. | Miles Bader | |
1995-10-13 | (diskfs_grow): | Miles Bader | |
Use diskfs_device_{read,write}_synce instead of dev_{read,write}_sync. | |||
1995-08-25 | (indir_release): When freeing direct blocks mentioned in a single | Michael I. Bushnell | |
indirect block, or single indirect blocks mentioned in a double, only call the free routine (ffs_blkfree or indir_release, respectively) if the block is actually allocated. | |||
1995-06-21 | (diskfs_truncate): Use ports reference calls directly instead of pager | Michael I. Bushnell | |
wrappers. | |||
1995-06-20 | (diskfs_grow): Provide initialization of POKE_OFF. | Michael I. Bushnell | |
1995-06-20 | (diskfs_truncate): Remove assignment from if test. | Michael I. Bushnell | |
1995-03-31 | (diskfs_truncate): Don't acquire writer lock on NP->dn->allocptrlock | Michael I. Bushnell | |
until after forcing delayed copies through; otherwise the pageins will deadlock attempting to get a reader lock to service them. This is safe, because we only need NP->allocsize here, and that can't change as long as we hold NP->lock. | |||
1994-10-28 | entered into RCS | Michael I. Bushnell | |
1994-10-04 | Formerly sizes.c.~24~ | Michael I. Bushnell | |
1994-10-03 | Formerly sizes.c.~23~ | Michael I. Bushnell | |
1994-09-27 | Formerly sizes.c.~22~ | Michael I. Bushnell | |
1994-09-27 | Formerly sizes.c.~21~ | Michael I. Bushnell | |
1994-09-23 | Formerly sizes.c.~20~ | Michael I. Bushnell | |
1994-09-22 | Formerly sizes.c.~19~ | Michael I. Bushnell | |
1994-09-21 | Formerly sizes.c.~18~ | Michael I. Bushnell | |
1994-09-21 | Formerly sizes.c.~17~ | Michael I. Bushnell | |
1994-07-15 | Formerly sizes.c.~16~ | Michael I. Bushnell | |
1994-07-14 | Formerly sizes.c.~15~ | Michael I. Bushnell | |
1994-06-13 | Formerly sizes.c.~14~ | Michael I. Bushnell | |
1994-06-09 | Initial revision | Michael I. Bushnell | |