summaryrefslogtreecommitdiff
path: root/libfshelp/fetch-root.c
AgeCommit message (Collapse)Author
1999-04-30Wed Apr 28 03:06:19 1999 Thomas Bushnell, BSG <tb@mit.edu>Thomas Bushnell
* fetch-root.c (fshelp_fetch_root): Don't deallocate the INIT_PORT_CWDIR after starting the passive translator; we still need to keep our reference around until the getroot call happens. Reported by Marcus Brinkmann (Marcus.Brinkmann@ruhr-uni-bochum.de).
1999-04-21Tue Apr 20 21:43:20 1999 Thomas Bushnell, BSG <tb@mit.edu>Thomas Bushnell
* fetch-root.c (fshelp_fetch_root): When starting a passive translator, set cwd to be DOTDOT instead of our own cwd.
1996-11-18Fri Nov 15 14:43:44 1996 Thomas Bushnell, n/BSG <thomas@gnu.ai.mit.edu>Thomas Bushnell
* fshelp.h (fshelp_access): Uh, shift bits the right direction. Tue Nov 12 22:07:41 1996 Thomas Bushnell, n/BSG <thomas@gnu.ai.mit.edu> * fetch-root.c (fshelp_fetch_root): Name idvec members correctly. * fshelp.h: <iohelp.h> -> <hurd/iohelp.h>. Include <sys/stat.h>. (fshelp_access): Declare `gotit'.
1996-11-13Wed Nov 6 17:49:33 1996 Thomas Bushnell, n/BSG <thomas@gnu.ai.mit.edu>Thomas Bushnell
* fshelp.h (fshelp_fetch_root): Delete args `uids', `gids', `uids_len', and `gids_len'. New arg `user'. * fetch-root.c (fshelp_fetch_root): Ditto. * fshelp.h: Include <iohelp.h>. (fshelp_access, fshelp_isowner, fshelp_checkdirmod): New functions.
1996-07-08(fshelp_fetch_root): Don't use unsafe MOVE_SEND in call toMichael I. Bushnell
fshelp_start_translator_long.
1996-06-24(fshelp_fetch_root): Deal properly with errors from auth_makeauth.Michael I. Bushnell
1996-05-09(fshelp_fetch_root) [reauth]: Use new args for auth_user_authenticate.Michael I. Bushnell
1996-02-21(fshelp_fetch_root): Make sure the returned fsys control port is valid.Miles Bader
1996-01-29(fshelp_fetch_root): Use hurd_condition_wait instead ofMichael I. Bushnell
condition_wait.
1996-01-27(fshelp_fetch_root):Miles Bader
If io_reauthenticate returns an error, just return MACH_PORT_NULL instead of aborting (the server probably died; not a good sign for the health of the translator, but it's better than dying ourselves...).
1995-11-06(fshelp_fetch_root):Miles Bader
When PORT_TYPE is MACH_MSG_TYPE_MAKE_SEND, make the right *before* using it. Don't bother reauthenticating the underlying node returned by CALLBACK2 -- it already has the right ids. This also gets rid of a problem with giving away our auth port prematurely.
1995-11-02(fshelp_fetch_root):Miles Bader
Uses two callbacks now. Pass an appropiate function to fshelp_start_translator_long instead of the actual underlying node.
1995-06-26(fshelp_fetch_root): Wakeup other blocked calls *before* returningMichael I. Bushnell
errors provided by fshelp_start_translator_long or CALLBACK.
1995-06-23(fshelp_fetch_root): Parenthesize assert test correctly.Michael I. Bushnell
1995-06-22(fshelp_fetch_root) [reauth]: If PORT is null, then just return it.Michael I. Bushnell
1995-06-22(fshelp_fetch_root): Pass type and length parameters in the rightMichael I. Bushnell
order in calls to auth_makeauth and fshelp_start_translator_long.
1995-06-21(fshelp_fetch_root): Pass new third arg to fshelp_set_active.Michael I. Bushnell
1995-06-19(fshelp_fetch_root): Reduce levels of loops. Only change ACTIVE whenMichael I. Bushnell
we are holding the lock. Use condition mechanism to serialize attempts to start the passive translator.
1995-06-14(fshelp_fetch_root): Never consume the right on DOTDOT.Michael I. Bushnell
1995-06-14(fshelp_fetch_root): Free ARGZ after we're done with them.Michael I. Bushnell
1995-06-14(fshelp_fetch_root): Take new arg COOKIE and pass it to CALLBACK.Michael I. Bushnell
1995-06-14*** empty log message ***Michael I. Bushnell
1994-09-06entered into RCSMichael I. Bushnell
1994-09-06Initial revisionMichael I. Bushnell