| Commit message (Expand) | Author | Age | Files | Lines |
* | Add g_wither_geom_close() function. | phk | 2004-10-29 | 1 | -0/+17 |
* | Don't call g_waitidle(), it happens automagically now. | phk | 2004-10-23 | 1 | -1/+0 |
* | - Turn KASSERT()s into warning printf()'s in the g_class_load() routine. | arr | 2004-10-22 | 1 | -10/+26 |
* | When loading GEOM modules, we expect the actual load process to be done | green | 2004-10-12 | 1 | -1/+9 |
* | For removable devices without media we set a zero mediasize but a non-zero | phk | 2004-09-05 | 1 | -0/+3 |
* | OK, now check geom class version numbers. | phk | 2004-08-08 | 1 | -1/+1 |
* | OOps, that check was a bit premature. Allow zero versions as well. | phk | 2004-08-08 | 1 | -1/+1 |
* | Give classes a version number and refuse to touch classes which are not | phk | 2004-08-08 | 1 | -0/+14 |
* | Only detach consumers which are attached when we wither stuff away. | phk | 2004-07-09 | 1 | -1/+2 |
* | Make withering water tight. | phk | 2004-07-08 | 1 | -25/+59 |
* | Fail normally rather than KASSERT if attempt to open a spoiled consumer. | phk | 2004-07-08 | 1 | -5/+3 |
* | Move "is consumer attached?" check before G_VALID_PROVIDER() check, | pjd | 2004-03-18 | 1 | -1/+1 |
* | Be more insistent on destroying geoms at unload time. Still not perfect, | phk | 2004-03-11 | 1 | -2/+8 |
* | Rearrange some of the GEOM debugging tools to be more structured. | phk | 2004-03-10 | 1 | -57/+42 |
* | Don't panic on providers already withered when we wither a geom. | phk | 2004-03-07 | 1 | -1/+2 |
* | Add a KASSERT which checks that a class never fails a closing ->access() | phk | 2004-02-14 | 1 | -0/+3 |
* | Remove the absolute count g_access_abs() function since experience has | phk | 2004-02-12 | 1 | -21/+3 |
* | Allow a GEOM class to unload if it has no geoms or a method function to | phk | 2004-02-02 | 1 | -1/+6 |
* | - Use proper names in KASSERTs. | pjd | 2004-02-02 | 1 | -6/+6 |
* | Add KASSERTS. | phk | 2004-01-23 | 1 | -0/+11 |
* | Plug an insignificant memoryleak. | phk | 2004-01-23 | 1 | -0/+3 |
* | Print the correct pointer in a KASSERT. | phk | 2004-01-11 | 1 | -1/+1 |
* | KASSERT against no-op access requests. | phk | 2004-01-09 | 1 | -0/+1 |
* | Call class->init() an class->fini() while the class is hooked up, | phk | 2003-11-18 | 1 | -3/+5 |
* | This is a crude bandaid for 5.2 to protect against providers which disappear | phk | 2003-11-15 | 1 | -0/+38 |
* | Introduce a per provider wither flag | phk | 2003-10-06 | 1 | -0/+2 |
* | Don't panic over the fact that unloading failed if we already knew that. | phk | 2003-08-22 | 1 | -3/+5 |
* | Use __FBSDID(). | obrien | 2003-06-11 | 1 | -2/+2 |
* | Drop a memory-corruption debugging test-tool. | phk | 2003-06-07 | 1 | -13/+0 |
* | Add missing va_end() calls. | phk | 2003-06-07 | 1 | -2/+4 |
* | Introduce g_provider_by_name() function, and use it. | phk | 2003-06-04 | 1 | -0/+17 |
* | Add a KASSERT to prevent the same GEOM class from being processed loaded | phk | 2003-06-04 | 1 | -0/+6 |
* | constify g_sanity() | phk | 2003-06-01 | 1 | -1/+1 |
* | Introduce a init and fini member functions on a class. | phk | 2003-05-31 | 1 | -22/+100 |
* | Back out all the stuff that didn't belong in the last commit. | phk | 2003-05-02 | 1 | -9/+12 |
* | Use g_slice_spoiled() rather than g_std_spoiled(). | phk | 2003-05-02 | 1 | -12/+9 |
* | Rework the "withering" mechanism: | phk | 2003-05-02 | 1 | -17/+46 |
* | Remove the now obsolete geomidorname hack. | phk | 2003-05-01 | 1 | -83/+0 |
* | Remove now unneeded special case for "geom.ctl". | phk | 2003-05-01 | 1 | -2/+0 |
* | Rename g_call_me() to g_post_event(), and give it a flag | phk | 2003-04-23 | 1 | -4/+5 |
* | Turn EV_NEW_PROVIDER into a g_call_me() event. | phk | 2003-04-23 | 1 | -2/+31 |
* | Convert EV_SPOILED event to use g_call_me(). | phk | 2003-04-23 | 1 | -1/+23 |
* | Turn the hardwired NEW_CLASS event into a g_call_me() event. | phk | 2003-04-23 | 1 | -2/+33 |
* | Remove unused event pointers in object structures. | phk | 2003-04-23 | 1 | -3/+0 |
* | Time has run from the "run GEOM in userland" harness, and the new regression | phk | 2003-04-13 | 1 | -11/+0 |
* | If we hit access ahead of a spoil event, we should have negative | phk | 2003-04-12 | 1 | -7/+3 |
* | Change events to have an array of "void *" references, and give the | phk | 2003-04-02 | 1 | -7/+7 |
* | Only orphan things if the open/close actually succeeded. | phk | 2003-04-02 | 1 | -11/+10 |
* | g_class_by_name() was unused too. | phk | 2003-03-25 | 1 | -13/+0 |
* | Remove unuse g_insert_geom(). | phk | 2003-03-25 | 1 | -38/+0 |