summaryrefslogtreecommitdiff
path: root/hlink.c
AgeCommit message (Collapse)AuthorFilesLines
2008-03-01Include 2008 in the copyright years.Wayne Davison1-1/+1
2008-02-06Fixed return code from hard_link_one() when not verbose.Wayne Davison1-1/+1
2007-12-31Some minor tweaks:Wayne Davison1-5/+5
- Improved some comments in hlink.c. - Changed "the" to "a" in rsyncd.conf.yo. - Improved the PATCH.name filtering loop in patch-update.
2007-11-29Don't try to process hard-link data in list-only mode.Wayne Davison1-12/+15
2007-11-22New logging categories added to allow differentiation betweenWayne Davison1-4/+4
transfer errors, normal errors, and warnings. New messages are translated into old FERROR/FINFO categories for older protocols.
2007-11-06Some changes to allow an unsorted file list even if the iconv optionWayne Davison1-9/+3
was disabled via configure.
2007-10-21Make last fix even better.Wayne Davison1-3/+3
2007-10-21Avoid dropping a hard-linked itemized output with -ii.Wayne Davison1-1/+2
2007-09-23Changed "statx" to "stat_x" to try to work around a build problem on AIX.Wayne Davison1-4/+4
2007-09-10- Simplified the flist-spanning gnum logic in match_gnums().Wayne Davison1-36/+44
- Don't use hlink_flist in hard_link_check(). - Made the prev_name logic in hard_link_check() better. - Added an "ndx" arg to the finish_hard_link() call. Also, set the prev value of any prior entries to the first finished item (which ensures that any item that follows a re-shuffled item (which was waiting for a transferred file to finish) can find the FIRST+DONE item.
2007-09-04If we turn off FLAG_HLINKED, set FLAG_HLINK_DONE so that anyWayne Davison1-0/+3
file-entry values affected by HLINK_BUMP() stay put.
2007-09-03Changes to allow hard-linking to work in inc_recurse mode.Wayne Davison1-56/+139
2007-09-03Removed the old version of the hashtable functions and updatedWayne Davison1-131/+12
the code to use the structures and functions calls.
2007-07-10Tweaking the license text a bit more.Wayne Davison1-2/+3
2007-07-07Changed "count" to "used" in struct file_list since there canWayne Davison1-2/+2
be empty items that make the count inaccurate (and the name somewhat deceiving).
2007-07-07Switching to GPL 3.Wayne Davison1-3/+2
2007-05-22Added a comment to a shared iterator to avoid a warning from IBM's checker.Wayne Davison1-1/+1
2007-04-26The --iconv option has now made it to the trunk.Wayne Davison1-10/+19
2007-03-11The ACL support has arrived! This version has a brand new protocolWayne Davison1-19/+49
that makes it incompatible with all prior versions. A patch will be provided to allow talking with older (patched) rsync versions.
2007-02-04Further modifications to the copyright comment section.Wayne Davison1-4/+3
2006-12-28A few minor tweaks to handle the newest variable names.Wayne Davison1-7/+7
2006-12-17A few more minor tweaks.Wayne Davison1-8/+7
2006-12-17Got rid of the code that sorted based on struct idev objects sinceWayne Davison1-71/+6
we always use a hardlinked group-number now.
2006-12-15Added a simple hashtable routine for hashing st_dev and st_ino infoWayne Davison1-25/+240
on the sending side, and the support routines for the receiving side that handle using a "group number" for each hard-link cluster rather than having to manage a pool of dev+inode data. (For protocol 30)
2006-12-09- Revamped the hard-link algorithm to save memory.Wayne Davison1-189/+193
- Improved the function names. - Improved a few variable names. - Got rid of the "is a hard link" message in favor of the normal status messages for files (e.g. --itemize-changes indicates hard-links). - Fixed a long-standing bug when combining -H with --compare-dest. - Made sure that code compiles when hard-linking is not available.
2006-12-07- Use renamed hard-link defines: F_HL_IDEV() and F_HL_LIST().Wayne Davison1-11/+11
2006-12-06- Make some char* pointers const.Wayne Davison1-26/+15
- Tweaked a couple variable names. - Changed the #ifdefs to not define any hard-link functions when hard-linking is disabled.
2006-12-05Use the latest F_*() accessors.Wayne Davison1-3/+2
2006-12-04Use new send_msg_int() function.Wayne Davison1-12/+6
2006-12-03- Typical tranfers now save 12-20 bytes per file because several varsWayne Davison1-46/+63
were moved out of file_struct into an optional member-var setup. - Renamed and reorganized the FLAG_* defines. - Use NDX_DONE instead of a literal -1 when sending/checking the end-of-phase index value.
2006-11-19Added "const" to appropriate char pointers.Wayne Davison1-2/+1
2006-10-17Make sure the new link_dest_used value gets zeroed on init.Wayne Davison1-0/+2
2006-10-13Use new FNONE value when setting/comparing enum logcode vars.Wayne Davison1-3/+5
2006-10-09- If a finished link came from a link-dest dir and --dry-run is inWayne Davison1-6/+12
effect, we now stat the link-dest file instead of the non-existent destination file. - Use int32 for the index variables.
2006-06-30Decided that I like remove-source-files better than remove-sender-files.Wayne Davison1-3/+3
2006-06-30Handle the --remove-sender-files option by sending MSG_SUCCESS forWayne Davison1-0/+12
any file that is being hard-linked.
2006-05-29Use the renamed stdout_format_has_i variable.Wayne Davison1-3/+3
2006-04-25Updated the FSF's address to an even newer one.Wayne Davison1-3/+3
2006-04-25- Updated the address for the FSF in the opening comment.Wayne Davison1-18/+21
- Standardized the format of the opening comment, including adding a brief description of what's in the file for those that lacked it. - Added some missing copyright lines. - Some minor whitespace tweaks (in a few of the files).
2006-02-24Got rid of unused externs.Wayne Davison1-1/+0
2006-01-14Call f_name() with revised args (which also replaces f_name_to()).Wayne Davison1-5/+7
2006-01-13Got rid of calls to (the soon to vanish) safe_fname() function.Wayne Davison1-9/+5
2005-12-15When compare-dest/link-dest/copy-dest are in effect, we now checkWayne Davison1-1/+28
if the file we're about to hard-link to the leader is present in one of the dest dirs -- if it is, we adjust the text we output accordingly.
2005-07-31There is no longer a need to check file->link_u.links at the startWayne Davison1-2/+0
of hard_link_check() (because the caller has already checked it).
2005-06-09- Never output the '"FOO" is a hard link' message when -i (%i)Wayne Davison1-1/+2
output is enabled. - Go back to outputting '"FOO" is a hard link' message when -i (%i) is not enabled and verbose > 1.
2005-04-23- Added a new function, maybe_hard_link(), which now holds the codeWayne Davison1-32/+69
(moved from hard_link_cluster()) that checks the inode/dev of an existing file and either does nothing (when properly linked), or removes the existing file before calling hard_link_one(). - Changed hard_link_check() so that it checks if the master item is marked with FINISHED_LINK (in its hlindex int), and if so, it finishes the current file by calling maybe_hard_link() (and marking it as finished too). Otherwise it marks the item as skipped by setting hlindex to SKIPPED_LINK. - Fixed the outputting of an error in hard_link_one() so that the message is always output when the caller did not request "terse" processing. - Changed hard_link_cluster() so that it marks the master item with FINISHED_LINK and then links only the skipped items (marking them with FINISHED_LINK too).
2005-04-15- Fixed a potential crash/infinite-loop bug if -H was usedWayne Davison1-22/+19
when sending a single file. - Made a couple variable improvements in link_idev_data(). - Got rid of the non-NULL check of hlink_list in init_hard_links() because is enough to just check file->link_u.links these days.
2005-03-09When itemizing, we now set ITEM_LOCAL_CHANGE and ITEM_XNAME_FOLLOWS.Wayne Davison1-3/+5
2005-03-05- Changed hlink_list[] to store file-list indexes instead ofWayne Davison1-73/+90
pointers. - Made hard_link_one() non-static so that the generator can call it. Improved it to do itemized output. - Replaced do_hard_links() with hard_link_cluster(), which changes the hard-linking from a post-transfer loop into a per-cluster operation that occurs incrementally as the transfer updates (or finds unchanged) one item in the cluster.
2005-02-27Added 2 args to do_hard_links() so that it can call maybe_send_keepalive().Wayne Davison1-1/+3