Module Name: src
Committed By: pooka
Date: Tue Aug 4 23:58:29 UTC 2009
Modified Files:
src/sys/rump/librump/rumpkern: vm.c
Added Files:
src/sys/rump/librump/rumpkern/opt: opt_uvmhist.h
Log Message:
* support uvm_pagermapin/out(), adapt uvm_pageratop()
Instead of doing actual page remapping, which we can't portably
do in userspace without extensive trickery (read: signals), simply
allocate the kva window with new physical backing, copy page
contents, return, and copy contents back in mapout. Since the
pages are locked during the mapping cycle, we can do this without
hazard.
* add lots of stubbies necessary for new stuff coming soon
To generate a diff of this commit:
cvs rdiff -u -r1.60 -r1.61 src/sys/rump/librump/rumpkern/vm.c
cvs rdiff -u -r0 -r1.1 src/sys/rump/librump/rumpkern/opt/opt_uvmhist.h
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/rump/librump/rumpkern/vm.c
diff -u src/sys/rump/librump/rumpkern/vm.c:1.60 src/sys/rump/librump/rumpkern/vm.c:1.61
--- src/sys/rump/librump/rumpkern/vm.c:1.60 Mon Aug 3 23:32:06 2009
+++ src/sys/rump/librump/rumpkern/vm.c Tue Aug 4 23:58:29 2009
@@ -1,4 +1,4 @@
-/* $NetBSD: vm.c,v 1.60 2009/08/03 23:32:06 pooka Exp $ */
+/* $NetBSD: vm.c,v 1.61 2009/08/04 23:58:29 pooka Exp $ */
/*
* Copyright (c) 2007 Antti Kantee. All Rights Reserved.
@@ -42,7 +42,7 @@
*/
#include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: vm.c,v 1.60 2009/08/03 23:32:06 pooka Exp $");
+__KERNEL_RCSID(0, "$NetBSD: vm.c,v 1.61 2009/08/04 23:58:29 pooka Exp $");
#include <sys/param.h>
#include <sys/atomic.h>
@@ -108,6 +108,18 @@
return pg;
}
+/* these are going away very soon */
+void rumpvm_enterva(vaddr_t addr, struct vm_page *pg) {}
+void rumpvm_flushva(struct uvm_object *uobj) {}
+
+struct vm_page *
+uvm_pagealloc_strat(struct uvm_object *uobj, voff_t off, struct vm_anon *anon,
+ int flags, int strat, int free_list)
+{
+
+ return rumpvm_makepage(uobj, off);
+}
+
/*
* Release a page.
*
@@ -127,43 +139,12 @@
kmem_free(pg, sizeof(*pg));
}
-struct rumpva {
- vaddr_t addr;
- struct vm_page *pg;
-
- LIST_ENTRY(rumpva) entries;
-};
-static LIST_HEAD(, rumpva) rvahead = LIST_HEAD_INITIALIZER(rvahead);
-static kmutex_t rvamtx;
-
void
-rumpvm_enterva(vaddr_t addr, struct vm_page *pg)
+uvm_pagezero(struct vm_page *pg)
{
- struct rumpva *rva;
- rva = kmem_alloc(sizeof(struct rumpva), KM_SLEEP);
- rva->addr = addr;
- rva->pg = pg;
- mutex_enter(&rvamtx);
- LIST_INSERT_HEAD(&rvahead, rva, entries);
- mutex_exit(&rvamtx);
-}
-
-void
-rumpvm_flushva(struct uvm_object *uobj)
-{
- struct rumpva *rva, *rva_next;
-
- mutex_enter(&rvamtx);
- for (rva = LIST_FIRST(&rvahead); rva; rva = rva_next) {
- rva_next = LIST_NEXT(rva, entries);
- if (rva->pg->uobject == uobj) {
- LIST_REMOVE(rva, entries);
- uvm_page_unbusy(&rva->pg, 1);
- kmem_free(rva, sizeof(*rva));
- }
- }
- mutex_exit(&rvamtx);
+ pg->flags &= ~PG_CLEAN;
+ memset((void *)pg->uanon, 0, PAGE_SIZE);
}
/*
@@ -251,6 +232,8 @@
* Misc routines
*/
+static kmutex_t pagermtx;
+
void
rumpvm_init(void)
{
@@ -259,7 +242,7 @@
uvm.pagedaemon_lwp = NULL; /* doesn't match curlwp */
rump_vmspace.vm_map.pmap = pmap_kernel();
- mutex_init(&rvamtx, MUTEX_DEFAULT, 0);
+ mutex_init(&pagermtx, MUTEX_DEFAULT, 0);
mutex_init(&uvm_pageqlock, MUTEX_DEFAULT, 0);
kernel_map->pmap = pmap_kernel();
@@ -268,12 +251,7 @@
callback_head_init(&kmem_map_store.vmk_reclaim_callback, IPL_VM);
}
-void
-uvm_pageactivate(struct vm_page *pg)
-{
- /* nada */
-}
void
uvm_pagewire(struct vm_page *pg)
@@ -297,43 +275,134 @@
panic("%s: unimplemented", __func__);
}
+struct pagerinfo {
+ vaddr_t pgr_kva;
+ int pgr_npages;
+ struct vm_page **pgr_pgs;
+ bool pgr_read;
+
+ LIST_ENTRY(pagerinfo) pgr_entries;
+};
+static LIST_HEAD(, pagerinfo) pagerlist = LIST_HEAD_INITIALIZER(pagerlist);
+
+/*
+ * Pager "map" in routine. Instead of mapping, we allocate memory
+ * and copy page contents there. Not optimal or even strictly
+ * correct (the caller might modify the page contents after mapping
+ * them in), but what the heck. Assumes UVMPAGER_MAPIN_WAITOK.
+ */
vaddr_t
-uvm_pagermapin(struct vm_page **pps, int npages, int flags)
+uvm_pagermapin(struct vm_page **pgs, int npages, int flags)
{
+ struct pagerinfo *pgri;
+ vaddr_t curkva;
+ int i;
- panic("%s: unimplemented", __func__);
+ /* allocate structures */
+ pgri = kmem_alloc(sizeof(*pgri), KM_SLEEP);
+ pgri->pgr_kva = (vaddr_t)kmem_alloc(npages * PAGE_SIZE, KM_SLEEP);
+ pgri->pgr_npages = npages;
+ pgri->pgr_pgs = kmem_alloc(sizeof(struct vm_page *) * npages, KM_SLEEP);
+ pgri->pgr_read = (flags & UVMPAGER_MAPIN_READ) != 0;
+
+ /* copy contents to "mapped" memory */
+ for (i = 0, curkva = pgri->pgr_kva;
+ i < npages;
+ i++, curkva += PAGE_SIZE) {
+ /*
+ * We need to copy the previous contents of the pages to
+ * the window even if we are reading from the
+ * device, since the device might not fill the contents of
+ * the full mapped range and we will end up corrupting
+ * data when we unmap the window.
+ */
+ memcpy((void*)curkva, pgs[i]->uanon, PAGE_SIZE);
+ pgri->pgr_pgs[i] = pgs[i];
+ }
+
+ mutex_enter(&pagermtx);
+ LIST_INSERT_HEAD(&pagerlist, pgri, pgr_entries);
+ mutex_exit(&pagermtx);
+
+ return pgri->pgr_kva;
}
-/* Called with the vm object locked */
-struct vm_page *
-uvm_pagelookup(struct uvm_object *uobj, voff_t off)
+/*
+ * map out the pager window. return contents from VA to page storage
+ * and free structures.
+ *
+ * Note: does not currently support partial frees
+ */
+void
+uvm_pagermapout(vaddr_t kva, int npages)
{
- struct vm_page *pg;
+ struct pagerinfo *pgri;
+ vaddr_t curkva;
+ int i;
- TAILQ_FOREACH(pg, &uobj->memq, listq.queue) {
- if (pg->offset == off) {
- return pg;
+ mutex_enter(&pagermtx);
+ LIST_FOREACH(pgri, &pagerlist, pgr_entries) {
+ if (pgri->pgr_kva == kva)
+ break;
+ }
+ KASSERT(pgri);
+ if (pgri->pgr_npages != npages)
+ panic("uvm_pagermapout: partial unmapping not supported");
+ LIST_REMOVE(pgri, pgr_entries);
+ mutex_exit(&pagermtx);
+
+ if (pgri->pgr_read) {
+ for (i = 0, curkva = pgri->pgr_kva;
+ i < pgri->pgr_npages;
+ i++, curkva += PAGE_SIZE) {
+ memcpy(pgri->pgr_pgs[i]->uanon,(void*)curkva,PAGE_SIZE);
}
}
- return NULL;
+ kmem_free(pgri->pgr_pgs, npages * sizeof(struct vm_page *));
+ kmem_free((void*)pgri->pgr_kva, npages * PAGE_SIZE);
+ kmem_free(pgri, sizeof(*pgri));
}
+/*
+ * convert va in pager window to page structure.
+ * XXX: how expensive is this (global lock, list traversal)?
+ */
struct vm_page *
uvm_pageratop(vaddr_t va)
{
- struct rumpva *rva;
+ struct pagerinfo *pgri;
+ struct vm_page *pg = NULL;
+ int i;
- mutex_enter(&rvamtx);
- LIST_FOREACH(rva, &rvahead, entries)
- if (rva->addr == va)
+ mutex_enter(&pagermtx);
+ LIST_FOREACH(pgri, &pagerlist, pgr_entries) {
+ if (pgri->pgr_kva <= va
+ && va < pgri->pgr_kva + pgri->pgr_npages*PAGE_SIZE)
break;
- mutex_exit(&rvamtx);
+ }
+ if (pgri) {
+ i = (va - pgri->pgr_kva) >> PAGE_SHIFT;
+ pg = pgri->pgr_pgs[i];
+ }
+ mutex_exit(&pagermtx);
- if (rva == NULL)
- panic("%s: va %llu", __func__, (unsigned long long)va);
+ return pg;
+}
- return rva->pg;
+/* Called with the vm object locked */
+struct vm_page *
+uvm_pagelookup(struct uvm_object *uobj, voff_t off)
+{
+ struct vm_page *pg;
+
+ TAILQ_FOREACH(pg, &uobj->memq, listq.queue) {
+ if (pg->offset == off) {
+ return pg;
+ }
+ }
+
+ return NULL;
}
void
@@ -607,3 +676,42 @@
bp->b_data = bp->b_saveaddr;
bp->b_saveaddr = 0;
}
+
+void
+uvm_wait(const char *msg)
+{
+
+ /* nothing to wait for */
+}
+
+/*
+ * page life cycle stuff. it really doesn't exist, so just stubs.
+ */
+
+void
+uvm_pageactivate(struct vm_page *pg)
+{
+
+ /* nada */
+}
+
+void
+uvm_pagedeactivate(struct vm_page *pg)
+{
+
+ /* nada */
+}
+
+void
+uvm_pagedequeue(struct vm_page *pg)
+{
+
+ /* nada*/
+}
+
+void
+uvm_pageenqueue(struct vm_page *pg)
+{
+
+ /* nada */
+}
Added files:
Index: src/sys/rump/librump/rumpkern/opt/opt_uvmhist.h
diff -u /dev/null src/sys/rump/librump/rumpkern/opt/opt_uvmhist.h:1.1
--- /dev/null Tue Aug 4 23:58:29 2009
+++ src/sys/rump/librump/rumpkern/opt/opt_uvmhist.h Tue Aug 4 23:58:29 2009
@@ -0,0 +1 @@
+/* $NetBSD: opt_uvmhist.h,v 1.1 2009/08/04 23:58:29 pooka Exp $ */
