summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* Revert rev 1.356 and 1.352 (pmap_mapdev hacks). It wasn't worth thepeter2002-08-052-22/+16
| | | | pain.
* Add a macro for SLIST traversal 'SLIST_FOREACH_PREVPTR',alfred2002-08-051-0/+5
| | | | | this macro keeps a pointer to the previous element's next pointer to allow for search and O(1) removal.
* Cleanup:alfred2002-08-051-28/+24
| | | | | | Fix line wrapping. Remove 'register'. malloc(9) with M_WAITOK can't fail, so remove checks for that.
* o Introduce pmap_page_is_mapped(). Its purpose is to obsoletealc2002-08-052-0/+2
| | | | the PG_MAPPED flag.
* Use m_getcl() to allocate buffers for the receive ring.luigi2002-08-051-11/+2
| | | | MFC after: 3 days
* o Don't set PG_MAPPED or PG_WRITEABLE when a page is mappedalc2002-08-055-39/+10
| | | | | using pmap_kenter() or pmap_qenter(). o Use VM_ALLOC_WIRED in pmap_new_thread().
* Extend the interface to ether_input(): a NULL eh pointer means thatluigi2002-08-041-12/+26
| | | | | | | | | | | | | | | | | | | | the mbuf contains the ethernet header (eh) as well, which ether_input() will strip off as needed. This permits the removal (in a backward compatible way) of the header removal code which right now is replicated in all drivers, sometimes in an inconsistent way. Also, because many functions called after ether_input() require the eh in the mbuf, eventually we can propagate the interface and handle outdated drivers just in ether_input(). Individual driver changes to use the new interface will follow as we have a chance to touch them. NOTE THAT THIS CHANGE IS FULLY BACKWARD COMPATIBLE AND DOES NOT BREAK BINARY COMPATIBILITY FOR DRIVERS. MFC after: 3 days
* Fix handling of Receiver Not Ready errors when doing polling.luigi2002-08-041-50/+82
| | | | | | | | | | Also take this chance to cleanup the code in fxp_intr_body. Add a missing block of code to disable interrupts when reinitializing the interface while doing polling (the RELENG_4 version was correct). MFC after: 3 days
* Use m_getcl() to allocate mbuf+cluster for the receive ring.luigi2002-08-041-25/+9
| | | | | | | Remove the sis_quick variable, as it was there for testing purposes only. MFC after: 3 days
* Temporarily disable polling when no processes are active, while Iluigi2002-08-041-1/+1
| | | | | | | | | investigate the problem described below. I am seeing some strange livelock on recent -current sources with a slow box under heavy load, which disappears with this change. This might suggest some kind of problem (either insufficient locking, or mishandling of priorities) in the poll_idle thread.
* Repond properly to NGM_TEXT_CONFIG messages.phk2002-08-041-0/+1
|
* bugfix: move check for udp_blackhole before the one for icmp_bandlim.luigi2002-08-041-2/+2
| | | | MFC after: 3 days
* Fix typo in the last revision.mike2002-08-042-2/+2
| | | | Noticed by: i386 tinderbox
* se -> sab.jake2002-08-043-2/+3
|
* o Acquire the page queues lock before checking the page's busy statusalc2002-08-041-2/+4
| | | | | in vm_page_grab(). Also, replace the nearby tsleep() with an msleep() on the page queues lock.
* Ported to FreeBSD.jake2002-08-041-582/+487
|
* Add Jason L. Wright's driver for the SAB82532 serial chip, found in manyjake2002-08-042-0/+1684
| | | | | | sun ultras. Obtained from: OpenBSD
* Add device agp to GENERIC, filter it out of floppy buildsanholt2002-08-042-0/+4
| | | | Approved by: des (mentor)
* Remove this abortive attempt.jake2002-08-043-526/+0
|
* Fix a mistake in 1.352 - I was returning a pointer to the rounded downpeter2002-08-042-2/+2
| | | | address. I expect this will fix acpica.
* Remove commented out PCI_ENABLE_IO_MODES. It is gone now.imp2002-08-043-3/+0
|
* Simplify the handling of a fragmented file_id descriptor. Alsoscottl2002-08-042-51/+50
| | | | de-obfuscate the file_char flags.
* - Replace v_flag with v_iflag and v_vflagjeff2002-08-0488-414/+620
| | | | | | | | | | | | | | | - v_vflag is protected by the vnode lock and is used when synchronization with VOP calls is needed. - v_iflag is protected by interlock and is used for dealing with vnode management issues. These flags include X/O LOCK, FREE, DOOMED, etc. - All accesses to v_iflag and v_vflag have either been locked or marked with mp_fixme's. - Many ASSERT_VOP_LOCKED calls have been added where the locking was not clear. - Many functions in vfs_subr.c were restructured to provide for stronger locking. Idea stolen from: BSD/OS
* o Extend the scope of the page queues lock in contigmalloc1().alc2002-08-041-8/+8
| | | | | o Replace vm_page_sleep_busy() with vm_page_sleep_if_busy() in vm_contig_launder().
* o Convert a vm_page_sleep_busy() into a vm_page_sleep_if_busy()alc2002-08-041-1/+3
| | | | with appropriate page queue locking.
* o Request a wired page from vm_page_grab() in _pmap_allocpte().alc2002-08-042-10/+2
|
* Print out the strings in vers.o instead of hardcoding the loader banner.jake2002-08-041-2/+4
|
* Kernel modifications necessary to allow to follow fork()ed children.mdodd2002-08-042-0/+11
| | | | | PR: bin/25587 (in part) MFC after: 3 weeks
* o Ask for a prezeroed page in pmap_pinit() for the page directory page.alc2002-08-032-2/+2
|
* o Convert two instances of vm_page_sleep_busy() to vm_page_sleep_if_busy()alc2002-08-031-7/+6
| | | | with appropriate page queue locking.
* Fix handling of packets which matched an "ipfw fwd" rule on the input side.luigi2002-08-031-0/+13
|
* o Don't set PG_MAPPED on the page allocated and mapped in _pmap_allocpte().alc2002-08-032-2/+0
| | | | | (Only set this flag if the mapping has a corresponding pv list entry, which this mapping doesn't.)
* Ignore memory above 4GB for now due to unpleasant pci issues.peter2002-08-031-0/+15
|
* o Remove the setting of PG_MAPPED from vm_page_wire() andalc2002-08-031-2/+0
| | | | vm_page_alloc(VM_ALLOC_WIRED).
* Take advantage of the fact that there is a small 1MB direct mapped regionpeter2002-08-032-16/+22
| | | | | on x86 in between KERNBASE and the kernel load address. pmap_mapdev() can return pointers to this for devices operating in the isa "hole".
* Take a shot at fixing a nasty bug in the pmap changes that I did. Ipeter2002-08-032-2/+2
| | | | | | | | | missed the pmap_kenter/kremove in this file, which leads to read()/write() of /dev/mem using stale TLB entries. (gah!) Fortunately, mmap of /dev/mem wasn't affected, so it wasn't as bad as it could have been. This throws some light on the 'X server affects stability' thread.... Pointed out by: bde
* We now have opt_wi.h, so go ahead and add it. Default it to 1.imp2002-08-031-1/+4
|
* Add code that will download firmware to a Symbol LA4100-series of CFimp2002-08-034-1/+4537
| | | | | | | | | | | | cards. Since the firmware is hard coded into the kernel, I've made it a kernel option (WI_SYMBOL_FIRMWARE). Note: This only downloads into the RAM of these cards. It doesn't download into FLASH, and is somewhat limited. There needs to be a better way to deal, but this works for now. My Symbol LA4132 CF card works now. Obtained from: NetBSD
* While PCI interrupts are shareable, this should not have been committed justimp2002-08-021-7/+0
| | | | yet.
* Make the consumers of the linker_load_file() function usemux2002-08-024-10/+9
| | | | | | | | | | | | | linker_load_module() instead. This fixes a bug where the kernel was unable to properly locate and load a kernel module in vfs_mount() (and probably in the netgraph code as well since it was using the same function). This is because the linker_load_file() does not properly search the module path. Problem found by: peter Reviewed by: peter Thanks to: peter
* Introduce support for Mandatory Access Control and extensiblerwatson2002-08-021-0/+5
| | | | | | | | | | kernel access control. When generating nd6 output on an interface, label the packet appropriately. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
* When preserving the IP header in extra mbuf in the IP forwardingrwatson2002-08-021-0/+7
| | | | | | | | case, also preserve the MAC label. Note that this mbuf allocation is fairly non-optimal, but not my fault. Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
* Improve ordering of MAC entry points in mac_policy_op structure.rwatson2002-08-022-12/+12
| | | | | Obtained from: TrustedBSD Project Sponsored by: DARPA, NAI Labs
* o Convert two instances of vm_page_sleep_busy() into vm_page_sleep_if_busy()alc2002-08-021-6/+9
| | | | with appropriate page queue locking.
* Move a prototype to the least wrong place.phk2002-08-026-2/+4
| | | | Suggested by: bde
* ptvmmap does not exist on alpha.peter2002-08-021-1/+0
|
* Work to fix LINT build.rwatson2002-08-021-0/+1
| | | | Reported by: phk
* SYSINIT needs to be SI_SUB_PSEUDO. Add a printf to tell we are here.phk2002-08-021-1/+2
|
* typo.phk2002-08-021-1/+1
|
* Add the minimalist elan-mmcr device driver.phk2002-08-023-0/+87
| | | | | This driver allows a userland program to mmap the MMCR of the AMD Elan sc520 CPU.
OpenPOWER on IntegriCloud