NetBSD/sys/uvm/uvm_fault_i.h

229 lines
5.8 KiB
C
Raw Normal View History

a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
2001-09-16 00:36:31 +04:00
/* $NetBSD: uvm_fault_i.h,v 1.15 2001/09/15 20:36:45 chs Exp $ */
/*
*
* Copyright (c) 1997 Charles D. Cranor and Washington University.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Charles D. Cranor and
* Washington University.
* 4. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
1998-02-07 14:07:38 +03:00
*
* from: Id: uvm_fault_i.h,v 1.1.6.1 1997/12/08 16:07:12 chuck Exp
*/
#ifndef _UVM_UVM_FAULT_I_H_
#define _UVM_UVM_FAULT_I_H_
/*
* uvm_fault_i.h: fault inline functions
*/
static boolean_t uvmfault_lookup __P((struct uvm_faultinfo *, boolean_t));
static boolean_t uvmfault_relock __P((struct uvm_faultinfo *));
static void uvmfault_unlockall __P((struct uvm_faultinfo *, struct vm_amap *,
struct uvm_object *, struct vm_anon *));
static void uvmfault_unlockmaps __P((struct uvm_faultinfo *, boolean_t));
/*
* uvmfault_unlockmaps: unlock the maps
*/
1998-03-09 03:58:55 +03:00
static __inline void
uvmfault_unlockmaps(ufi, write_locked)
struct uvm_faultinfo *ufi;
boolean_t write_locked;
{
/*
* ufi can be NULL when this isn't really a fault,
* but merely paging in anon data.
*/
if (ufi == NULL) {
return;
}
1998-03-09 03:58:55 +03:00
if (write_locked) {
vm_map_unlock(ufi->map);
} else {
vm_map_unlock_read(ufi->map);
}
}
/*
* uvmfault_unlockall: unlock everything passed in.
*
* => maps must be read-locked (not write-locked).
*/
1998-03-09 03:58:55 +03:00
static __inline void
uvmfault_unlockall(ufi, amap, uobj, anon)
struct uvm_faultinfo *ufi;
struct vm_amap *amap;
struct uvm_object *uobj;
struct vm_anon *anon;
{
1998-03-09 03:58:55 +03:00
if (anon)
simple_unlock(&anon->an_lock);
if (uobj)
simple_unlock(&uobj->vmobjlock);
if (amap)
amap_unlock(amap);
1998-03-09 03:58:55 +03:00
uvmfault_unlockmaps(ufi, FALSE);
}
/*
* uvmfault_lookup: lookup a virtual address in a map
*
* => caller must provide a uvm_faultinfo structure with the IN
* params properly filled in
* => we will lookup the map entry (handling submaps) as we go
* => if the lookup is a success we will return with the maps locked
* => if "write_lock" is TRUE, we write_lock the map, otherwise we only
* get a read lock.
2001-05-25 08:06:11 +04:00
* => note that submaps can only appear in the kernel and they are
* required to use the same virtual addresses as the map they
* are referenced by (thus address translation between the main
* map and the submap is unnecessary).
*/
1998-03-09 03:58:55 +03:00
static __inline boolean_t
uvmfault_lookup(ufi, write_lock)
struct uvm_faultinfo *ufi;
boolean_t write_lock;
{
struct vm_map *tmpmap;
1998-03-09 03:58:55 +03:00
/*
* init ufi values for lookup.
*/
ufi->map = ufi->orig_map;
ufi->size = ufi->orig_size;
/*
* keep going down levels until we are done. note that there can
* only be two levels so we won't loop very long.
*/
while (1) {
/*
* Make sure this is not an "interrupt safe" map.
* Such maps are never supposed to be involved in
* a fault.
*/
if (ufi->map->flags & VM_MAP_INTRSAFE)
return (FALSE);
1998-03-09 03:58:55 +03:00
/*
* lock map
*/
if (write_lock) {
vm_map_lock(ufi->map);
} else {
vm_map_lock_read(ufi->map);
}
/*
* lookup
*/
2001-05-25 08:06:11 +04:00
if (!uvm_map_lookup_entry(ufi->map, ufi->orig_rvaddr,
&ufi->entry)) {
1998-03-09 03:58:55 +03:00
uvmfault_unlockmaps(ufi, write_lock);
return(FALSE);
}
/*
* reduce size if necessary
*/
if (ufi->entry->end - ufi->orig_rvaddr < ufi->size)
ufi->size = ufi->entry->end - ufi->orig_rvaddr;
1998-03-09 03:58:55 +03:00
/*
* submap? replace map with the submap and lookup again.
* note: VAs in submaps must match VAs in main map.
*/
if (UVM_ET_ISSUBMAP(ufi->entry)) {
tmpmap = ufi->entry->object.sub_map;
if (write_lock) {
vm_map_unlock(ufi->map);
} else {
vm_map_unlock_read(ufi->map);
}
ufi->map = tmpmap;
continue;
}
/*
* got it!
*/
1998-03-09 03:58:55 +03:00
ufi->mapv = ufi->map->timestamp;
return(TRUE);
1998-03-09 03:58:55 +03:00
} /* while loop */
1998-03-09 03:58:55 +03:00
/*NOTREACHED*/
}
/*
* uvmfault_relock: attempt to relock the same version of the map
*
* => fault data structures should be unlocked before calling.
* => if a success (TRUE) maps will be locked after call.
*/
1998-03-09 03:58:55 +03:00
static __inline boolean_t
uvmfault_relock(ufi)
struct uvm_faultinfo *ufi;
{
/*
* ufi can be NULL when this isn't really a fault,
* but merely paging in anon data.
*/
if (ufi == NULL) {
return TRUE;
}
1998-03-09 03:58:55 +03:00
uvmexp.fltrelck++;
1998-03-09 03:58:55 +03:00
/*
2001-05-25 08:06:11 +04:00
* relock map. fail if version mismatch (in which case nothing
* gets locked).
1998-03-09 03:58:55 +03:00
*/
vm_map_lock_read(ufi->map);
if (ufi->mapv != ufi->map->timestamp) {
vm_map_unlock_read(ufi->map);
return(FALSE);
}
uvmexp.fltrelckok++;
a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
2001-09-16 00:36:31 +04:00
return(TRUE);
}
#endif /* _UVM_UVM_FAULT_I_H_ */