Date: Sat, 5 Dec 2015 08:29:40 +0300 From: Gleb Smirnoff <glebius@FreeBSD.org> To: jeff@FreeBSD.org, alc@FreeBSD.org, kib@FreeBSD.org Cc: scottl@FreeBSD.org, pho@FreeBSD.org, arch@FreeBSD.org Subject: new vm_pager_get_pages() KPI, round 3 Message-ID: <20151205052940.GJ42565@FreeBSD.org>
next in thread | raw e-mail | index | archive | help
--OwLcNYc0lM97+oe1
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Hi,
[first paragraph for arch subscribers, To: recepients may skip]
This patch is kinda a prerequisite for the non-blocking sendfile(2),
that was jointly developed by NGINX and Netflix in 2014 and has been
running in Netflix production for a year, serving 35% of the whole
North America (US, Canada, Mexico) Internet traffic.
Technically, the new sendfile(2) doesn't require the new
vm_pager_get_pages() KPI. We currently run it on the old KPI. However,
kib@ suggested that we are abusing the KPI, carefully using its
edge cases. To address this critic, back in spring, I suggested a KPI,
where vm_pager_get_pages() offers all-or-none approach to the array of
pages. Again, kib@ wasn't satisfied, as for "the main user" of
vm_pager_get_pages, the vm_fault(), all-or-none approach isn't optimal.
The problem was slowly debated through the summer. And then in October
jeff@ suggested yet another extension of the KPI, which I have
implemented and it is described below.
[for those interested in new sendfile(2), skip to the last paragraph,
for those willing to review new pager KPI, read on]
The new KPI offers this prototype for vm_pager_get_pages():
int
vm_pager_get_pages(vm_object_t object, vm_page_t pages[], int count,
int *rbehind, in *rahead);
Where "count" stands for number of pages in the array. The rbehind
and rahead if not NULL specify how many pages the caller is willing to
allow the pager to pre-cache, if the pager can.
Pager doesn't promise to do any read behind or read ahead. If it does,
then only the pager is responsive for grabbing, busying, unbusying and
queueing these pages. It also writes the actual values of completed
read ahead and read behind back to the pointers.
Pager promises to page in "count" pages or fail. Pager expects the
pages to be busied, and returns them busied. For a multi page requests,
the pager demands that the region is a valid region, that exists in
the pager, which can be checked by preceding call to vm_pager_haspage().
For single page requests, there is no such demand.
The net result is a win for both vm_fault() and for new sendfile().
The vm_fault() no longer needs to do prepatory vm_pager_haspage(),
which removes one I/O operation. The logic for read ahead/behind,
which is strongly UFS/EXT-centric, moves into vnode_pager.c. So
we no longer do useless operations when having a fault on ZFS.
The vm_fault() now knows precisely the read ahead that happened,
when updates fs.entry->next_read index. This reduces number of
hardfaults by a tiny fraction (measured building world tree).
The new sendfile() has a stronger KPI, that doesn't unbusy pages,
that sendfile() needs to be kept busied.
Also, the new KPI removes some ugly edges. E.g., since the old
KPI used to unbusy and free pages in the array in case of an
error, the pages could not be wired. However, there are places in
kernel where we want to page in into a wired page. These places
simply violated the assumption, relying on lack of errors in the
pager. Moreover, the swap pager has a special function to skip
wired pages, while doing the freeing sweep, to avoid hitting
assertion. That means passing wired pages to swapper is kinda
OK, while to any other pager it is not. So, we end up with
vm_pager_get_pages() being not pager agnostic, while it is
designed to be so. Now this is fixed.
Peter, if you can, please try the patch in your tests. I already
did that, but you are always better at this :)
[the new sendfile]
As already mentioned, Netflix runs new sendfile(2) in production,
and it is one of key components, that allows us to serve over
80 Gbit/s from a single box. We strongly want to contribute this
code and see it in FreeBSD 11.0-RELEASE. I believe, many FreeBSD
users, who run it as a content server, also want that. Although the
code was production ready back in 2014, it is still not in head.
The reason is the drama with vm_pager_get_pages() KPI. I was very
patient during the whole 2015. Sometimes I was waiting for a feedback
from guys in "To:" for several weeks. I was very gentle to not commit
anything to sys/vm without a review. Now we've got only 2 months left
before the 11.0-RELEASE cycle. And since I want the new sendfile be
there in 11.0, I'm going to push that strongly, putting off all my
patience and gentleness. I won't buy any dislikes on the KPI again,
since this is a third round of compromises from my side. I will wait
only one week for pre-commit reviews, and then all reviews and
asjustments are post-commit.
--
Totus tuus, Glebius.
--OwLcNYc0lM97+oe1
Content-Type: text/x-diff; charset=us-ascii
Content-Disposition: attachment; filename="vm_pager_get_pages.diff"
Index: sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c
===================================================================
--- sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c (revision 291639)
+++ sys/cddl/contrib/opensolaris/uts/common/fs/zfs/zfs_vnops.c (working copy)
@@ -5762,12 +5762,13 @@ ioflags(int ioflags)
}
static int
-zfs_getpages(struct vnode *vp, vm_page_t *m, int count, int reqpage)
+zfs_getpages(struct vnode *vp, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
znode_t *zp = VTOZ(vp);
zfsvfs_t *zfsvfs = zp->z_zfsvfs;
objset_t *os = zp->z_zfsvfs->z_os;
- vm_page_t mfirst, mlast, mreq;
+ vm_page_t mlast;
vm_object_t object;
caddr_t va;
struct sf_buf *sf;
@@ -5776,82 +5777,46 @@ static int
vm_pindex_t reqstart, reqend;
int pcount, lsize, reqsize, size;
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
+
ZFS_ENTER(zfsvfs);
ZFS_VERIFY_ZP(zp);
pcount = OFF_TO_IDX(round_page(count));
- mreq = m[reqpage];
- object = mreq->object;
- error = 0;
- if (pcount > 1 && zp->z_blksz > PAGESIZE) {
- startoff = rounddown(IDX_TO_OFF(mreq->pindex), zp->z_blksz);
- reqstart = OFF_TO_IDX(round_page(startoff));
- if (reqstart < m[0]->pindex)
- reqstart = 0;
- else
- reqstart = reqstart - m[0]->pindex;
- endoff = roundup(IDX_TO_OFF(mreq->pindex) + PAGE_SIZE,
- zp->z_blksz);
- reqend = OFF_TO_IDX(trunc_page(endoff)) - 1;
- if (reqend > m[pcount - 1]->pindex)
- reqend = m[pcount - 1]->pindex;
- reqsize = reqend - m[reqstart]->pindex + 1;
- KASSERT(reqstart <= reqpage && reqpage < reqstart + reqsize,
- ("reqpage beyond [reqstart, reqstart + reqsize[ bounds"));
- } else {
- reqstart = reqpage;
- reqsize = 1;
- }
- mfirst = m[reqstart];
- mlast = m[reqstart + reqsize - 1];
-
zfs_vmobject_wlock(object);
-
- for (i = 0; i < reqstart; i++) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
- }
- for (i = reqstart + reqsize; i < pcount; i++) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
- }
-
- if (mreq->valid && reqsize == 1) {
- if (mreq->valid != VM_PAGE_BITS_ALL)
- vm_page_zero_invalid(mreq, TRUE);
+ if (m[pcount - 1]->valid != 0 && --pcount == 0) {
zfs_vmobject_wunlock(object);
ZFS_EXIT(zfsvfs);
return (zfs_vm_pagerret_ok);
}
- PCPU_INC(cnt.v_vnodein);
- PCPU_ADD(cnt.v_vnodepgsin, reqsize);
+ object = m[0]->object;
+ mlast = m[pcount - 1];
- if (IDX_TO_OFF(mreq->pindex) >= object->un_pager.vnp.vnp_size) {
- for (i = reqstart; i < reqstart + reqsize; i++) {
- if (i != reqpage) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
- }
- }
+ if (IDX_TO_OFF(mlast->pindex) >=
+ object->un_pager.vnp.vnp_size) {
zfs_vmobject_wunlock(object);
ZFS_EXIT(zfsvfs);
return (zfs_vm_pagerret_bad);
}
+ PCPU_INC(cnt.v_vnodein);
+ PCPU_ADD(cnt.v_vnodepgsin, reqsize);
+
lsize = PAGE_SIZE;
if (IDX_TO_OFF(mlast->pindex) + lsize > object->un_pager.vnp.vnp_size)
- lsize = object->un_pager.vnp.vnp_size - IDX_TO_OFF(mlast->pindex);
-
+ lsize = object->un_pager.vnp.vnp_size -
+ IDX_TO_OFF(mlast->pindex);
zfs_vmobject_wunlock(object);
- for (i = reqstart; i < reqstart + reqsize; i++) {
+ error = 0;
+ for (i = 0; i < pcount; i++) {
size = PAGE_SIZE;
- if (i == (reqstart + reqsize - 1))
+ if (i == pcount - 1)
size = lsize;
va = zfs_map_page(m[i], &sf);
error = dmu_read(os, zp->z_id, IDX_TO_OFF(m[i]->pindex),
@@ -5860,21 +5825,15 @@ static int
bzero(va + size, PAGE_SIZE - size);
zfs_unmap_page(sf);
if (error != 0)
- break;
+ goto out;
}
zfs_vmobject_wlock(object);
-
- for (i = reqstart; i < reqstart + reqsize; i++) {
- if (!error)
- m[i]->valid = VM_PAGE_BITS_ALL;
- KASSERT(m[i]->dirty == 0, ("zfs_getpages: page %p is dirty", m[i]));
- if (i != reqpage)
- vm_page_readahead_finish(m[i]);
- }
-
+ for (i = 0; i < pcount; i++)
+ m[i]->valid = VM_PAGE_BITS_ALL;
zfs_vmobject_wunlock(object);
+out:
ZFS_ACCESSTIME_STAMP(zfsvfs, zp);
ZFS_EXIT(zfsvfs);
return (error ? zfs_vm_pagerret_error : zfs_vm_pagerret_ok);
@@ -5886,11 +5845,13 @@ zfs_freebsd_getpages(ap)
struct vnode *a_vp;
vm_page_t *a_m;
int a_count;
- int a_reqpage;
+ int *a_rbehind;
+ int *a_rahead;
} */ *ap;
{
- return (zfs_getpages(ap->a_vp, ap->a_m, ap->a_count, ap->a_reqpage));
+ return (zfs_getpages(ap->a_vp, ap->a_m, ap->a_count, ap->a_rbehind,
+ ap->a_rahead));
}
static int
Index: sys/dev/drm2/i915/i915_gem.c
===================================================================
--- sys/dev/drm2/i915/i915_gem.c (revision 291639)
+++ sys/dev/drm2/i915/i915_gem.c (working copy)
@@ -4338,7 +4338,7 @@ i915_gem_wire_page(vm_object_t object, vm_pindex_t
page = vm_page_grab(object, pindex, VM_ALLOC_NORMAL);
if (page->valid != VM_PAGE_BITS_ALL) {
if (vm_pager_has_page(object, pindex, NULL, NULL)) {
- rv = vm_pager_get_pages(object, &page, 1, 0);
+ rv = vm_pager_get_pages(object, &page, 1, NULL, NULL);
if (rv != VM_PAGER_OK) {
vm_page_lock(page);
vm_page_free(page);
Index: sys/dev/drm2/ttm/ttm_tt.c
===================================================================
--- sys/dev/drm2/ttm/ttm_tt.c (revision 291639)
+++ sys/dev/drm2/ttm/ttm_tt.c (working copy)
@@ -291,7 +291,8 @@ int ttm_tt_swapin(struct ttm_tt *ttm)
from_page = vm_page_grab(obj, i, VM_ALLOC_NORMAL);
if (from_page->valid != VM_PAGE_BITS_ALL) {
if (vm_pager_has_page(obj, i, NULL, NULL)) {
- rv = vm_pager_get_pages(obj, &from_page, 1, 0);
+ rv = vm_pager_get_pages(obj, &from_page, 1,
+ NULL, NULL);
if (rv != VM_PAGER_OK) {
vm_page_lock(from_page);
vm_page_free(from_page);
Index: sys/dev/md/md.c
===================================================================
--- sys/dev/md/md.c (revision 291639)
+++ sys/dev/md/md.c (working copy)
@@ -847,7 +847,8 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
if (m->valid == VM_PAGE_BITS_ALL)
rv = VM_PAGER_OK;
else
- rv = vm_pager_get_pages(sc->object, &m, 1, 0);
+ rv = vm_pager_get_pages(sc->object, &m, 1,
+ NULL, NULL);
if (rv == VM_PAGER_ERROR) {
vm_page_xunbusy(m);
break;
@@ -870,7 +871,8 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
}
} else if (bp->bio_cmd == BIO_WRITE) {
if (len != PAGE_SIZE && m->valid != VM_PAGE_BITS_ALL)
- rv = vm_pager_get_pages(sc->object, &m, 1, 0);
+ rv = vm_pager_get_pages(sc->object, &m, 1,
+ NULL, NULL);
else
rv = VM_PAGER_OK;
if (rv == VM_PAGER_ERROR) {
@@ -886,7 +888,8 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
m->valid = VM_PAGE_BITS_ALL;
} else if (bp->bio_cmd == BIO_DELETE) {
if (len != PAGE_SIZE && m->valid != VM_PAGE_BITS_ALL)
- rv = vm_pager_get_pages(sc->object, &m, 1, 0);
+ rv = vm_pager_get_pages(sc->object, &m, 1,
+ NULL, NULL);
else
rv = VM_PAGER_OK;
if (rv == VM_PAGER_ERROR) {
Index: sys/fs/fuse/fuse_vnops.c
===================================================================
--- sys/fs/fuse/fuse_vnops.c (revision 291639)
+++ sys/fs/fuse/fuse_vnops.c (working copy)
@@ -1753,6 +1753,10 @@ fuse_vnop_getpages(struct vop_getpages_args *ap)
cred = curthread->td_ucred; /* XXX */
pages = ap->a_m;
count = ap->a_count;
+ if (ap->a_rbehind)
+ *ap->a_rbehind = 0;
+ if (ap->a_rahead)
+ *ap->a_rahead = 0;
if (!fsess_opt_mmap(vnode_mount(vp))) {
FS_DEBUG("called on non-cacheable vnode??\n");
@@ -1761,26 +1765,21 @@ fuse_vnop_getpages(struct vop_getpages_args *ap)
npages = btoc(count);
/*
- * If the requested page is partially valid, just return it and
- * allow the pager to zero-out the blanks. Partially valid pages
- * can only occur at the file EOF.
+ * If the last page is partially valid, just return it and allow
+ * the pager to zero-out the blanks. Partially valid pages can
+ * only occur at the file EOF.
+ *
+ * XXXGL: is that true for FUSE, which is a local filesystem,
+ * but still somewhat disconnected from the kernel?
*/
-
VM_OBJECT_WLOCK(vp->v_object);
- fuse_vm_page_lock_queues();
- if (pages[ap->a_reqpage]->valid != 0) {
- for (i = 0; i < npages; ++i) {
- if (i != ap->a_reqpage) {
- fuse_vm_page_lock(pages[i]);
- vm_page_free(pages[i]);
- fuse_vm_page_unlock(pages[i]);
- }
+ if (pages[npages - 1]->valid != 0) {
+ if (--npages == 0) {
+ VM_OBJECT_WUNLOCK(vp->v_object);
+ return (VM_PAGER_OK);
}
- fuse_vm_page_unlock_queues();
- VM_OBJECT_WUNLOCK(vp->v_object);
- return 0;
- }
- fuse_vm_page_unlock_queues();
+ count = npages << PAGE_SHIFT;
+ }
VM_OBJECT_WUNLOCK(vp->v_object);
/*
@@ -1811,17 +1810,6 @@ fuse_vnop_getpages(struct vop_getpages_args *ap)
if (error && (uio.uio_resid == count)) {
FS_DEBUG("error %d\n", error);
- VM_OBJECT_WLOCK(vp->v_object);
- fuse_vm_page_lock_queues();
- for (i = 0; i < npages; ++i) {
- if (i != ap->a_reqpage) {
- fuse_vm_page_lock(pages[i]);
- vm_page_free(pages[i]);
- fuse_vm_page_unlock(pages[i]);
- }
- }
- fuse_vm_page_unlock_queues();
- VM_OBJECT_WUNLOCK(vp->v_object);
return VM_PAGER_ERROR;
}
/*
@@ -1862,8 +1850,6 @@ fuse_vnop_getpages(struct vop_getpages_args *ap)
*/
;
}
- if (i != ap->a_reqpage)
- vm_page_readahead_finish(m);
}
fuse_vm_page_unlock_queues();
VM_OBJECT_WUNLOCK(vp->v_object);
Index: sys/fs/nfsclient/nfs_clbio.c
===================================================================
--- sys/fs/nfsclient/nfs_clbio.c (revision 291639)
+++ sys/fs/nfsclient/nfs_clbio.c (working copy)
@@ -101,6 +101,10 @@ ncl_getpages(struct vop_getpages_args *ap)
nmp = VFSTONFS(vp->v_mount);
pages = ap->a_m;
count = ap->a_count;
+ if (ap->a_rbehind)
+ *ap->a_rbehind = 0;
+ if (ap->a_rahead)
+ *ap->a_rahead = 0;
if ((object = vp->v_object) == NULL) {
ncl_printf("nfs_getpages: called with non-merged cache vnode??\n");
@@ -132,12 +136,18 @@ ncl_getpages(struct vop_getpages_args *ap)
* If the requested page is partially valid, just return it and
* allow the pager to zero-out the blanks. Partially valid pages
* can only occur at the file EOF.
+ *
+ * XXXGL: is that true for NFS, where short read can occur???
*/
- if (pages[ap->a_reqpage]->valid != 0) {
- vm_pager_free_nonreq(object, pages, ap->a_reqpage, npages,
- FALSE);
- return (VM_PAGER_OK);
+ VM_OBJECT_WLOCK(object);
+ if (pages[npages - 1]->valid != 0) {
+ if (--npages == 0) {
+ VM_OBJECT_WUNLOCK(object);
+ return (VM_PAGER_OK);
+ }
+ count = npages << PAGE_SHIFT;
}
+ VM_OBJECT_WUNLOCK(object);
/*
* We use only the kva address for the buffer, but this is extremely
@@ -167,8 +177,6 @@ ncl_getpages(struct vop_getpages_args *ap)
if (error && (uio.uio_resid == count)) {
ncl_printf("nfs_getpages: error %d\n", error);
- vm_pager_free_nonreq(object, pages, ap->a_reqpage, npages,
- FALSE);
return (VM_PAGER_ERROR);
}
@@ -212,8 +220,6 @@ ncl_getpages(struct vop_getpages_args *ap)
*/
;
}
- if (i != ap->a_reqpage)
- vm_page_readahead_finish(m);
}
VM_OBJECT_WUNLOCK(object);
return (0);
Index: sys/fs/smbfs/smbfs_io.c
===================================================================
--- sys/fs/smbfs/smbfs_io.c (revision 291639)
+++ sys/fs/smbfs/smbfs_io.c (working copy)
@@ -424,7 +424,7 @@ smbfs_getpages(ap)
#ifdef SMBFS_RWGENERIC
return vop_stdgetpages(ap);
#else
- int i, error, nextoff, size, toff, npages, count, reqpage;
+ int i, error, nextoff, size, toff, npages, count;
struct uio uio;
struct iovec iov;
vm_offset_t kva;
@@ -436,7 +436,7 @@ smbfs_getpages(ap)
struct smbnode *np;
struct smb_cred *scred;
vm_object_t object;
- vm_page_t *pages, m;
+ vm_page_t *pages;
vp = ap->a_vp;
if ((object = vp->v_object) == NULL) {
@@ -451,26 +451,25 @@ smbfs_getpages(ap)
pages = ap->a_m;
count = ap->a_count;
npages = btoc(count);
- reqpage = ap->a_reqpage;
+ if (ap->a_rbehind)
+ *ap->a_rbehind = 0;
+ if (ap->a_rahead)
+ *ap->a_rahead = 0;
/*
* If the requested page is partially valid, just return it and
* allow the pager to zero-out the blanks. Partially valid pages
* can only occur at the file EOF.
+ *
+ * XXXGL: is that true for SMB filesystem?
*/
- m = pages[reqpage];
-
VM_OBJECT_WLOCK(object);
- if (m->valid != 0) {
- for (i = 0; i < npages; ++i) {
- if (i != reqpage) {
- vm_page_lock(pages[i]);
- vm_page_free(pages[i]);
- vm_page_unlock(pages[i]);
- }
+ if (pages[npages - 1]->valid != 0) {
+ if (--npages == 0) {
+ VM_OBJECT_WUNLOCK(object);
+ return (VM_PAGER_OK);
}
- VM_OBJECT_WUNLOCK(object);
- return 0;
+ count = npages << PAGE_SHIFT;
}
VM_OBJECT_WUNLOCK(object);
@@ -500,22 +499,14 @@ smbfs_getpages(ap)
relpbuf(bp, &smbfs_pbuf_freecnt);
- VM_OBJECT_WLOCK(object);
if (error && (uio.uio_resid == count)) {
printf("smbfs_getpages: error %d\n",error);
- for (i = 0; i < npages; i++) {
- if (reqpage != i) {
- vm_page_lock(pages[i]);
- vm_page_free(pages[i]);
- vm_page_unlock(pages[i]);
- }
- }
- VM_OBJECT_WUNLOCK(object);
return VM_PAGER_ERROR;
}
size = count - uio.uio_resid;
+ VM_OBJECT_WLOCK(object);
for (i = 0, toff = 0; i < npages; i++, toff = nextoff) {
vm_page_t m;
nextoff = toff + PAGE_SIZE;
@@ -544,9 +535,6 @@ smbfs_getpages(ap)
*/
;
}
-
- if (i != reqpage)
- vm_page_readahead_finish(m);
}
VM_OBJECT_WUNLOCK(object);
return 0;
Index: sys/fs/tmpfs/tmpfs_subr.c
===================================================================
--- sys/fs/tmpfs/tmpfs_subr.c (revision 291639)
+++ sys/fs/tmpfs/tmpfs_subr.c (working copy)
@@ -1370,7 +1370,8 @@ retry:
VM_OBJECT_WLOCK(uobj);
goto retry;
} else if (m->valid != VM_PAGE_BITS_ALL)
- rv = vm_pager_get_pages(uobj, &m, 1, 0);
+ rv = vm_pager_get_pages(uobj, &m, 1,
+ NULL, NULL);
else
/* A cached page was reactivated. */
rv = VM_PAGER_OK;
Index: sys/kern/kern_exec.c
===================================================================
--- sys/kern/kern_exec.c (revision 291639)
+++ sys/kern/kern_exec.c (working copy)
@@ -950,8 +950,7 @@ int
exec_map_first_page(imgp)
struct image_params *imgp;
{
- int rv, i;
- int initial_pagein;
+ int rv, i, after, initial_pagein;
vm_page_t ma[VM_INITIAL_PAGEIN];
vm_object_t object;
@@ -967,9 +966,18 @@ exec_map_first_page(imgp)
#endif
ma[0] = vm_page_grab(object, 0, VM_ALLOC_NORMAL);
if (ma[0]->valid != VM_PAGE_BITS_ALL) {
- initial_pagein = VM_INITIAL_PAGEIN;
- if (initial_pagein > object->size)
- initial_pagein = object->size;
+ if (!vm_pager_has_page(object, 0, NULL, &after)) {
+ vm_page_lock(ma[0]);
+ vm_page_free(ma[0]);
+ vm_page_unlock(ma[0]);
+ vm_page_xunbusy(ma[0]);
+ VM_OBJECT_WUNLOCK(object);
+ return (EIO);
+ }
+ initial_pagein = min(after, VM_INITIAL_PAGEIN);
+ KASSERT(initial_pagein <= object->size,
+ ("%s: initial_pagein %d object->size %ju",
+ __func__, initial_pagein, (uintmax_t )object->size));
for (i = 1; i < initial_pagein; i++) {
if ((ma[i] = vm_page_next(ma[i - 1])) != NULL) {
if (ma[i]->valid)
@@ -984,14 +992,19 @@ exec_map_first_page(imgp)
}
}
initial_pagein = i;
- rv = vm_pager_get_pages(object, ma, initial_pagein, 0);
+ rv = vm_pager_get_pages(object, ma, initial_pagein, NULL, NULL);
if (rv != VM_PAGER_OK) {
- vm_page_lock(ma[0]);
- vm_page_free(ma[0]);
- vm_page_unlock(ma[0]);
+ for (i = 0; i < initial_pagein; i++) {
+ vm_page_lock(ma[i]);
+ vm_page_free(ma[i]);
+ vm_page_unlock(ma[i]);
+ vm_page_xunbusy(ma[i]);
+ }
VM_OBJECT_WUNLOCK(object);
return (EIO);
}
+ for (i = 1; i < initial_pagein; i++)
+ vm_page_readahead_finish(ma[i]);
}
vm_page_xunbusy(ma[0]);
vm_page_lock(ma[0]);
Index: sys/kern/uipc_shm.c
===================================================================
--- sys/kern/uipc_shm.c (revision 291639)
+++ sys/kern/uipc_shm.c (working copy)
@@ -189,7 +189,7 @@ uiomove_object_page(vm_object_t obj, size_t len, s
m = vm_page_grab(obj, idx, VM_ALLOC_NORMAL);
if (m->valid != VM_PAGE_BITS_ALL) {
if (vm_pager_has_page(obj, idx, NULL, NULL)) {
- rv = vm_pager_get_pages(obj, &m, 1, 0);
+ rv = vm_pager_get_pages(obj, &m, 1, NULL, NULL);
if (rv != VM_PAGER_OK) {
printf(
"uiomove_object: vm_obj %p idx %jd valid %x pager error %d\n",
@@ -460,7 +460,7 @@ retry:
goto retry;
} else if (m->valid != VM_PAGE_BITS_ALL)
rv = vm_pager_get_pages(object, &m, 1,
- 0);
+ NULL, NULL);
else
/* A cached page was reactivated. */
rv = VM_PAGER_OK;
Index: sys/kern/uipc_syscalls.c
===================================================================
--- sys/kern/uipc_syscalls.c (revision 291639)
+++ sys/kern/uipc_syscalls.c (working copy)
@@ -2033,7 +2033,7 @@ sendfile_readpage(vm_object_t obj, struct vnode *v
VM_OBJECT_WLOCK(obj);
} else {
if (vm_pager_has_page(obj, pindex, NULL, NULL)) {
- rv = vm_pager_get_pages(obj, &m, 1, 0);
+ rv = vm_pager_get_pages(obj, &m, 1, NULL, NULL);
SFSTAT_INC(sf_iocnt);
if (rv != VM_PAGER_OK) {
vm_page_lock(m);
Index: sys/kern/vfs_default.c
===================================================================
--- sys/kern/vfs_default.c (revision 291639)
+++ sys/kern/vfs_default.c (working copy)
@@ -731,12 +731,13 @@ vop_stdgetpages(ap)
struct vnode *a_vp;
vm_page_t *a_m;
int a_count;
- int a_reqpage;
+ int *a_rbehind;
+ int *a_rahead;
} */ *ap;
{
return vnode_pager_generic_getpages(ap->a_vp, ap->a_m,
- ap->a_count, ap->a_reqpage, NULL, NULL);
+ ap->a_count, ap->a_rbehind, ap->a_rahead, NULL, NULL);
}
static int
@@ -744,8 +745,9 @@ vop_stdgetpages_async(struct vop_getpages_async_ar
{
int error;
- error = VOP_GETPAGES(ap->a_vp, ap->a_m, ap->a_count, ap->a_reqpage);
- ap->a_iodone(ap->a_arg, ap->a_m, ap->a_reqpage, error);
+ error = VOP_GETPAGES(ap->a_vp, ap->a_m, ap->a_count, ap->a_rbehind,
+ ap->a_rahead);
+ ap->a_iodone(ap->a_arg, ap->a_m, ap->a_count, error);
return (error);
}
Index: sys/kern/vnode_if.src
===================================================================
--- sys/kern/vnode_if.src (revision 291639)
+++ sys/kern/vnode_if.src (working copy)
@@ -473,7 +473,8 @@ vop_getpages {
IN struct vnode *vp;
IN vm_page_t *m;
IN int count;
- IN int reqpage;
+ IN int *rbehind;
+ IN int *rahead;
};
@@ -483,7 +484,8 @@ vop_getpages_async {
IN struct vnode *vp;
IN vm_page_t *m;
IN int count;
- IN int reqpage;
+ IN int *rbehind;
+ IN int *rahead;
IN vop_getpages_iodone_t *iodone;
IN void *arg;
};
Index: sys/sys/buf.h
===================================================================
--- sys/sys/buf.h (revision 291639)
+++ sys/sys/buf.h (working copy)
@@ -122,14 +122,13 @@ struct buf {
struct ucred *b_rcred; /* Read credentials reference. */
struct ucred *b_wcred; /* Write credentials reference. */
union {
- TAILQ_ENTRY(buf) bu_freelist; /* (Q) */
+ TAILQ_ENTRY(buf) b_freelist; /* (Q) */
struct {
- void (*pg_iodone)(void *, vm_page_t *, int, int);
- int pg_reqpage;
- } bu_pager;
- } b_union;
-#define b_freelist b_union.bu_freelist
-#define b_pager b_union.bu_pager
+ void (*b_pgiodone)(void *, vm_page_t *, int, int);
+ int b_pgbefore;
+ int b_pgafter;
+ };
+ };
union cluster_info {
TAILQ_HEAD(cluster_list_head, buf) cluster_head;
TAILQ_ENTRY(buf) cluster_entry;
Index: sys/vm/default_pager.c
===================================================================
--- sys/vm/default_pager.c (revision 291639)
+++ sys/vm/default_pager.c (working copy)
@@ -56,7 +56,7 @@ __FBSDID("$FreeBSD$");
static vm_object_t default_pager_alloc(void *, vm_ooffset_t, vm_prot_t,
vm_ooffset_t, struct ucred *);
static void default_pager_dealloc(vm_object_t);
-static int default_pager_getpages(vm_object_t, vm_page_t *, int, int);
+static int default_pager_getpages(vm_object_t, vm_page_t *, int, int *, int *);
static void default_pager_putpages(vm_object_t, vm_page_t *, int,
boolean_t, int *);
static boolean_t default_pager_haspage(vm_object_t, vm_pindex_t, int *,
@@ -122,13 +122,16 @@ default_pager_dealloc(object)
* see a vm_page with assigned swap here.
*/
static int
-default_pager_getpages(object, m, count, reqpage)
- vm_object_t object;
- vm_page_t *m;
- int count;
- int reqpage;
+default_pager_getpages(vm_object_t object, vm_page_t *m, int count,
+ int *rbehind, int *rahead)
{
- return VM_PAGER_FAIL;
+
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
+
+ return (VM_PAGER_FAIL);
}
/*
Index: sys/vm/device_pager.c
===================================================================
--- sys/vm/device_pager.c (revision 291639)
+++ sys/vm/device_pager.c (working copy)
@@ -59,7 +59,7 @@ static void dev_pager_init(void);
static vm_object_t dev_pager_alloc(void *, vm_ooffset_t, vm_prot_t,
vm_ooffset_t, struct ucred *);
static void dev_pager_dealloc(vm_object_t);
-static int dev_pager_getpages(vm_object_t, vm_page_t *, int, int);
+static int dev_pager_getpages(vm_object_t, vm_page_t *, int, int *, int *);
static void dev_pager_putpages(vm_object_t, vm_page_t *, int, int, int *);
static boolean_t dev_pager_haspage(vm_object_t, vm_pindex_t, int *, int *);
static void dev_pager_free_page(vm_object_t object, vm_page_t m);
@@ -257,28 +257,33 @@ dev_pager_dealloc(vm_object_t object)
}
static int
-dev_pager_getpages(vm_object_t object, vm_page_t *ma, int count, int reqpage)
+dev_pager_getpages(vm_object_t object, vm_page_t *ma, int count, int *rbehind,
+ int *rahead)
{
int error;
+ /* Since our haspage reports zero after/before, the count is 1. */
+ KASSERT(count == 1, ("%s: count %d", __func__, count));
VM_OBJECT_ASSERT_WLOCKED(object);
error = object->un_pager.devp.ops->cdev_pg_fault(object,
- IDX_TO_OFF(ma[reqpage]->pindex), PROT_READ, &ma[reqpage]);
+ IDX_TO_OFF(ma[0]->pindex), PROT_READ, &ma[0]);
VM_OBJECT_ASSERT_WLOCKED(object);
- vm_pager_free_nonreq(object, ma, reqpage, count, TRUE);
-
if (error == VM_PAGER_OK) {
KASSERT((object->type == OBJT_DEVICE &&
- (ma[reqpage]->oflags & VPO_UNMANAGED) != 0) ||
+ (ma[0]->oflags & VPO_UNMANAGED) != 0) ||
(object->type == OBJT_MGTDEVICE &&
- (ma[reqpage]->oflags & VPO_UNMANAGED) == 0),
- ("Wrong page type %p %p", ma[reqpage], object));
+ (ma[0]->oflags & VPO_UNMANAGED) == 0),
+ ("Wrong page type %p %p", ma[0], object));
if (object->type == OBJT_DEVICE) {
TAILQ_INSERT_TAIL(&object->un_pager.devp.devp_pglist,
- ma[reqpage], plinks.q);
+ ma[0], plinks.q);
}
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
}
return (error);
Index: sys/vm/phys_pager.c
===================================================================
--- sys/vm/phys_pager.c (revision 291639)
+++ sys/vm/phys_pager.c (working copy)
@@ -139,7 +139,8 @@ phys_pager_dealloc(vm_object_t object)
* Fill as many pages as vm_fault has allocated for us.
*/
static int
-phys_pager_getpages(vm_object_t object, vm_page_t *m, int count, int reqpage)
+phys_pager_getpages(vm_object_t object, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
int i;
@@ -154,14 +155,11 @@ static int
("phys_pager_getpages: partially valid page %p", m[i]));
KASSERT(m[i]->dirty == 0,
("phys_pager_getpages: dirty page %p", m[i]));
- /* The requested page must remain busy, the others not. */
- if (i == reqpage) {
- vm_page_lock(m[i]);
- vm_page_flash(m[i]);
- vm_page_unlock(m[i]);
- } else
- vm_page_xunbusy(m[i]);
}
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
return (VM_PAGER_OK);
}
Index: sys/vm/sg_pager.c
===================================================================
--- sys/vm/sg_pager.c (revision 291639)
+++ sys/vm/sg_pager.c (working copy)
@@ -49,7 +49,7 @@ __FBSDID("$FreeBSD$");
static vm_object_t sg_pager_alloc(void *, vm_ooffset_t, vm_prot_t,
vm_ooffset_t, struct ucred *);
static void sg_pager_dealloc(vm_object_t);
-static int sg_pager_getpages(vm_object_t, vm_page_t *, int, int);
+static int sg_pager_getpages(vm_object_t, vm_page_t *, int, int *, int *);
static void sg_pager_putpages(vm_object_t, vm_page_t *, int,
boolean_t, int *);
static boolean_t sg_pager_haspage(vm_object_t, vm_pindex_t, int *,
@@ -135,7 +135,8 @@ sg_pager_dealloc(vm_object_t object)
}
static int
-sg_pager_getpages(vm_object_t object, vm_page_t *m, int count, int reqpage)
+sg_pager_getpages(vm_object_t object, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
struct sglist *sg;
vm_page_t m_paddr, page;
@@ -145,11 +146,13 @@ static int
size_t space;
int i;
+ /* Since our haspage reports zero after/before, the count is 1. */
+ KASSERT(count == 1, ("%s: count %d", __func__, count));
VM_OBJECT_ASSERT_WLOCKED(object);
sg = object->handle;
memattr = object->memattr;
VM_OBJECT_WUNLOCK(object);
- offset = m[reqpage]->pindex;
+ offset = m[0]->pindex;
/*
* Lookup the physical address of the requested page. An initial
@@ -178,7 +181,7 @@ static int
}
/* Return a fake page for the requested page. */
- KASSERT(!(m[reqpage]->flags & PG_FICTITIOUS),
+ KASSERT(!(m[0]->flags & PG_FICTITIOUS),
("backing page for SG is fake"));
/* Construct a new fake page. */
@@ -185,19 +188,16 @@ static int
page = vm_page_getfake(paddr, memattr);
VM_OBJECT_WLOCK(object);
TAILQ_INSERT_TAIL(&object->un_pager.sgp.sgp_pglist, page, plinks.q);
-
- /* Free the original pages and insert this fake page into the object. */
- for (i = 0; i < count; i++) {
- if (i == reqpage &&
- vm_page_replace(page, object, offset) != m[i])
- panic("sg_pager_getpages: invalid place replacement");
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
- }
- m[reqpage] = page;
+ if (vm_page_replace(page, object, offset) != m[0])
+ panic("sg_pager_getpages: invalid place replacement");
+ m[0] = page;
page->valid = VM_PAGE_BITS_ALL;
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
+
return (VM_PAGER_OK);
}
Index: sys/vm/swap_pager.c
===================================================================
--- sys/vm/swap_pager.c (revision 291639)
+++ sys/vm/swap_pager.c (working copy)
@@ -357,9 +357,10 @@ static vm_object_t
swap_pager_alloc(void *handle, vm_ooffset_t size,
vm_prot_t prot, vm_ooffset_t offset, struct ucred *);
static void swap_pager_dealloc(vm_object_t object);
-static int swap_pager_getpages(vm_object_t, vm_page_t *, int, int);
-static int swap_pager_getpages_async(vm_object_t, vm_page_t *, int, int,
- pgo_getpages_iodone_t, void *);
+static int swap_pager_getpages(vm_object_t, vm_page_t *, int, int *,
+ int *);
+static int swap_pager_getpages_async(vm_object_t, vm_page_t *, int, int *,
+ int *, pgo_getpages_iodone_t, void *);
static void swap_pager_putpages(vm_object_t, vm_page_t *, int, boolean_t, int *);
static boolean_t
swap_pager_haspage(vm_object_t object, vm_pindex_t pindex, int *before, int *after);
@@ -413,16 +414,6 @@ static void swp_pager_meta_free(vm_object_t, vm_pi
static void swp_pager_meta_free_all(vm_object_t);
static daddr_t swp_pager_meta_ctl(vm_object_t, vm_pindex_t, int);
-static void
-swp_pager_free_nrpage(vm_page_t m)
-{
-
- vm_page_lock(m);
- if (m->wire_count == 0)
- vm_page_free(m);
- vm_page_unlock(m);
-}
-
/*
* SWP_SIZECHECK() - update swap_pager_full indication
*
@@ -1103,16 +1094,12 @@ swap_pager_unswapped(vm_page_t m)
* left busy, but the others adjusted.
*/
static int
-swap_pager_getpages(vm_object_t object, vm_page_t *m, int count, int reqpage)
+swap_pager_getpages(vm_object_t object, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
struct buf *bp;
- vm_page_t mreq;
- int i;
- int j;
daddr_t blk;
- mreq = m[reqpage];
-
/*
* Calculate range to retrieve. The pages have already been assigned
* their swapblks. We require a *contiguous* range but we know it to
@@ -1122,45 +1109,18 @@ static int
*
* The swp_*() calls must be made with the object locked.
*/
- blk = swp_pager_meta_ctl(mreq->object, mreq->pindex, 0);
+ blk = swp_pager_meta_ctl(m[0]->object, m[0]->pindex, 0);
- for (i = reqpage - 1; i >= 0; --i) {
- daddr_t iblk;
-
- iblk = swp_pager_meta_ctl(m[i]->object, m[i]->pindex, 0);
- if (blk != iblk + (reqpage - i))
- break;
- }
- ++i;
-
- for (j = reqpage + 1; j < count; ++j) {
- daddr_t jblk;
-
- jblk = swp_pager_meta_ctl(m[j]->object, m[j]->pindex, 0);
- if (blk != jblk - (j - reqpage))
- break;
- }
-
- /*
- * free pages outside our collection range. Note: we never free
- * mreq, it must remain busy throughout.
- */
- if (0 < i || j < count) {
- int k;
-
- for (k = 0; k < i; ++k)
- swp_pager_free_nrpage(m[k]);
- for (k = j; k < count; ++k)
- swp_pager_free_nrpage(m[k]);
- }
-
- /*
- * Return VM_PAGER_FAIL if we have nothing to do. Return mreq
- * still busy, but the others unbusied.
- */
if (blk == SWAPBLK_NONE)
return (VM_PAGER_FAIL);
+#ifdef INVARIANTS
+ for (int i = 0; i < count; i++)
+ KASSERT(blk + i ==
+ swp_pager_meta_ctl(m[i]->object, m[i]->pindex, 0),
+ ("%s: range is not contiguous", __func__));
+#endif
+
/*
* Getpbuf() can sleep.
*/
@@ -1175,21 +1135,16 @@ static int
bp->b_iodone = swp_pager_async_iodone;
bp->b_rcred = crhold(thread0.td_ucred);
bp->b_wcred = crhold(thread0.td_ucred);
- bp->b_blkno = blk - (reqpage - i);
- bp->b_bcount = PAGE_SIZE * (j - i);
- bp->b_bufsize = PAGE_SIZE * (j - i);
- bp->b_pager.pg_reqpage = reqpage - i;
+ bp->b_blkno = blk;
+ bp->b_bcount = PAGE_SIZE * count;
+ bp->b_bufsize = PAGE_SIZE * count;
+ bp->b_npages = count;
VM_OBJECT_WLOCK(object);
- {
- int k;
-
- for (k = i; k < j; ++k) {
- bp->b_pages[k - i] = m[k];
- m[k]->oflags |= VPO_SWAPINPROG;
- }
+ for (int i = 0; i < count; i++) {
+ bp->b_pages[i] = m[i];
+ m[i]->oflags |= VPO_SWAPINPROG;
}
- bp->b_npages = j - i;
PCPU_INC(cnt.v_swapin);
PCPU_ADD(cnt.v_swappgsin, bp->b_npages);
@@ -1221,8 +1176,8 @@ static int
* is set in the meta-data.
*/
VM_OBJECT_WLOCK(object);
- while ((mreq->oflags & VPO_SWAPINPROG) != 0) {
- mreq->oflags |= VPO_SWAPSLEEP;
+ while ((m[0]->oflags & VPO_SWAPINPROG) != 0) {
+ m[0]->oflags |= VPO_SWAPSLEEP;
PCPU_INC(cnt.v_intrans);
if (VM_OBJECT_SLEEP(object, &object->paging_in_progress, PSWP,
"swread", hz * 20)) {
@@ -1233,16 +1188,19 @@ static int
}
/*
- * mreq is left busied after completion, but all the other pages
- * are freed. If we had an unrecoverable read error the page will
- * not be valid.
+ * If we had an unrecoverable read error pages will not be valid.
*/
- if (mreq->valid != VM_PAGE_BITS_ALL) {
- return (VM_PAGER_ERROR);
- } else {
- return (VM_PAGER_OK);
- }
+ for (int i = 0; i < count; i++)
+ if (m[i]->valid != VM_PAGE_BITS_ALL)
+ return (VM_PAGER_ERROR);
+ if (rbehind)
+ *rbehind = 0;
+ if (rahead)
+ *rahead = 0;
+
+ return (VM_PAGER_OK);
+
/*
* A final note: in a low swap situation, we cannot deallocate swap
* and mark a page dirty here because the caller is likely to mark
@@ -1259,11 +1217,11 @@ static int
*/
static int
swap_pager_getpages_async(vm_object_t object, vm_page_t *m, int count,
- int reqpage, pgo_getpages_iodone_t iodone, void *arg)
+ int *rbehind, int *rahead, pgo_getpages_iodone_t iodone, void *arg)
{
int r, error;
- r = swap_pager_getpages(object, m, count, reqpage);
+ r = swap_pager_getpages(object, m, count, rbehind, rahead);
VM_OBJECT_WUNLOCK(object);
switch (r) {
case VM_PAGER_OK:
@@ -1527,33 +1485,11 @@ swp_pager_async_iodone(struct buf *bp)
*/
if (bp->b_iocmd == BIO_READ) {
/*
- * When reading, reqpage needs to stay
- * locked for the parent, but all other
- * pages can be freed. We still want to
- * wakeup the parent waiting on the page,
- * though. ( also: pg_reqpage can be -1 and
- * not match anything ).
- *
- * We have to wake specifically requested pages
- * up too because we cleared VPO_SWAPINPROG and
- * someone may be waiting for that.
- *
* NOTE: for reads, m->dirty will probably
* be overridden by the original caller of
* getpages so don't play cute tricks here.
*/
m->valid = 0;
- if (i != bp->b_pager.pg_reqpage)
- swp_pager_free_nrpage(m);
- else {
- vm_page_lock(m);
- vm_page_flash(m);
- vm_page_unlock(m);
- }
- /*
- * If i == bp->b_pager.pg_reqpage, do not wake
- * the page up. The caller needs to.
- */
} else {
/*
* If a write error occurs, reactivate page
@@ -1575,38 +1511,12 @@ swp_pager_async_iodone(struct buf *bp)
* want to do that anyway, but it was an optimization
* that existed in the old swapper for a time before
* it got ripped out due to precisely this problem.
- *
- * If not the requested page then deactivate it.
- *
- * Note that the requested page, reqpage, is left
- * busied, but we still have to wake it up. The
- * other pages are released (unbusied) by
- * vm_page_xunbusy().
*/
KASSERT(!pmap_page_is_mapped(m),
("swp_pager_async_iodone: page %p is mapped", m));
- m->valid = VM_PAGE_BITS_ALL;
KASSERT(m->dirty == 0,
("swp_pager_async_iodone: page %p is dirty", m));
-
- /*
- * We have to wake specifically requested pages
- * up too because we cleared VPO_SWAPINPROG and
- * could be waiting for it in getpages. However,
- * be sure to not unbusy getpages specifically
- * requested page - getpages expects it to be
- * left busy.
- */
- if (i != bp->b_pager.pg_reqpage) {
- vm_page_lock(m);
- vm_page_deactivate(m);
- vm_page_unlock(m);
- vm_page_xunbusy(m);
- } else {
- vm_page_lock(m);
- vm_page_flash(m);
- vm_page_unlock(m);
- }
+ m->valid = VM_PAGE_BITS_ALL;
} else {
/*
* For write success, clear the dirty
@@ -1727,7 +1637,7 @@ swp_pager_force_pagein(vm_object_t object, vm_pind
return;
}
- if (swap_pager_getpages(object, &m, 1, 0) != VM_PAGER_OK)
+ if (swap_pager_getpages(object, &m, 1, NULL, NULL) != VM_PAGER_OK)
panic("swap_pager_force_pagein: read from swap failed");/*XXX*/
vm_object_pip_wakeup(object);
vm_page_dirty(m);
Index: sys/vm/vm_fault.c
===================================================================
--- sys/vm/vm_fault.c (revision 291639)
+++ sys/vm/vm_fault.c (working copy)
@@ -107,13 +107,8 @@ __FBSDID("$FreeBSD$");
#define PFBAK 4
#define PFFOR 4
-static int vm_fault_additional_pages(vm_page_t, int, int, vm_page_t *, int *);
-
-#define VM_FAULT_READ_BEHIND 8
#define VM_FAULT_READ_DEFAULT (1 + VM_FAULT_READ_AHEAD_INIT)
#define VM_FAULT_READ_MAX (1 + VM_FAULT_READ_AHEAD_MAX)
-#define VM_FAULT_NINCR (VM_FAULT_READ_MAX / VM_FAULT_READ_BEHIND)
-#define VM_FAULT_SUM (VM_FAULT_NINCR * (VM_FAULT_NINCR + 1) / 2)
#define VM_FAULT_DONTNEED_MIN 1048576
@@ -133,7 +128,7 @@ struct faultstate {
static void vm_fault_dontneed(const struct faultstate *fs, vm_offset_t vaddr,
int ahead);
static void vm_fault_prefault(const struct faultstate *fs, vm_offset_t addra,
- int faultcount, int reqpage);
+ int backward, int forward);
static inline void
release_page(struct faultstate *fs)
@@ -288,11 +283,10 @@ vm_fault_hold(vm_map_t map, vm_offset_t vaddr, vm_
int fault_flags, vm_page_t *m_hold)
{
vm_prot_t prot;
- int alloc_req, era, faultcount, nera, reqpage, result;
+ int alloc_req, era, faultcount, nera, result;
boolean_t growstack, is_first_object_locked, wired;
int map_generation;
vm_object_t next_object;
- vm_page_t marray[VM_FAULT_READ_MAX];
int hardfault;
struct faultstate fs;
struct vnode *vp;
@@ -303,7 +297,7 @@ vm_fault_hold(vm_map_t map, vm_offset_t vaddr, vm_
growstack = TRUE;
PCPU_INC(cnt.v_vm_faults);
fs.vp = NULL;
- faultcount = reqpage = 0;
+ faultcount = 0;
RetryFault:;
@@ -389,7 +383,7 @@ RetryFault:;
FALSE);
VM_OBJECT_RUNLOCK(fs.first_object);
if (!wired)
- vm_fault_prefault(&fs, vaddr, 0, 0);
+ vm_fault_prefault(&fs, vaddr, PFBAK, PFFOR);
vm_map_lookup_done(fs.map, fs.entry);
curthread->td_ru.ru_minflt++;
return (KERN_SUCCESS);
@@ -652,36 +646,13 @@ vnode_locked:
("vm_fault: vnode-backed object mapped by system map"));
/*
- * now we find out if any other pages should be paged
- * in at this time this routine checks to see if the
- * pages surrounding this fault reside in the same
- * object as the page for this fault. If they do,
- * then they are faulted in also into the object. The
- * array "marray" returned contains an array of
- * vm_page_t structs where one of them is the
- * vm_page_t passed to the routine. The reqpage
- * return value is the index into the marray for the
- * vm_page_t passed to the routine.
- *
- * fs.m plus the additional pages are exclusive busied.
+ * Page in the requested page and hint the pager,
+ * that it may bring up surrounding pages.
*/
- faultcount = vm_fault_additional_pages(
- fs.m, behind, ahead, marray, &reqpage);
-
- rv = faultcount ?
- vm_pager_get_pages(fs.object, marray, faultcount,
- reqpage) : VM_PAGER_FAIL;
-
+ rv = vm_pager_get_pages(fs.object, &fs.m, 1,
+ &behind, &ahead);
if (rv == VM_PAGER_OK) {
- /*
- * Found the page. Leave it busy while we play
- * with it.
- *
- * Pager could have changed the page. Pager
- * is responsible for disposition of old page
- * if moved.
- */
- fs.m = marray[reqpage];
+ faultcount = behind + 1 + ahead;
hardfault++;
break; /* break to PAGE HAS BEEN FOUND */
}
@@ -961,16 +932,13 @@ vnode_locked:
}
/*
* If the page was filled by a pager, update the map entry's
- * last read offset. Since the pager does not return the
- * actual set of pages that it read, this update is based on
- * the requested set. Typically, the requested and actual
- * sets are the same.
+ * last read offset.
*
* XXX The following assignment modifies the map
* without holding a write lock on it.
*/
if (hardfault)
- fs.entry->next_read = fs.pindex + faultcount - reqpage;
+ fs.entry->next_read = fs.pindex + ahead + 1;
vm_fault_dirty(fs.entry, fs.m, prot, fault_type, fault_flags, TRUE);
vm_page_assert_xbusied(fs.m);
@@ -993,7 +961,9 @@ vnode_locked:
fault_type | (wired ? PMAP_ENTER_WIRED : 0), 0);
if (faultcount != 1 && (fault_flags & VM_FAULT_WIRE) == 0 &&
wired == 0)
- vm_fault_prefault(&fs, vaddr, faultcount, reqpage);
+ vm_fault_prefault(&fs, vaddr,
+ faultcount > 0 ? behind : PFBAK,
+ faultcount > 0 ? ahead : PFFOR);
VM_OBJECT_WLOCK(fs.object);
vm_page_lock(fs.m);
@@ -1110,7 +1080,7 @@ vm_fault_dontneed(const struct faultstate *fs, vm_
*/
static void
vm_fault_prefault(const struct faultstate *fs, vm_offset_t addra,
- int faultcount, int reqpage)
+ int backward, int forward)
{
pmap_t pmap;
vm_map_entry_t entry;
@@ -1118,19 +1088,12 @@ vm_fault_prefault(const struct faultstate *fs, vm_
vm_offset_t addr, starta;
vm_pindex_t pindex;
vm_page_t m;
- int backward, forward, i;
+ int i;
pmap = fs->map->pmap;
if (pmap != vmspace_pmap(curthread->td_proc->p_vmspace))
return;
- if (faultcount > 0) {
- backward = reqpage;
- forward = faultcount - reqpage - 1;
- } else {
- backward = PFBAK;
- forward = PFFOR;
- }
entry = fs->entry;
starta = addra - backward * PAGE_SIZE;
@@ -1461,134 +1424,7 @@ again:
}
}
-
/*
- * This routine checks around the requested page for other pages that
- * might be able to be faulted in. This routine brackets the viable
- * pages for the pages to be paged in.
- *
- * Inputs:
- * m, rbehind, rahead
- *
- * Outputs:
- * marray (array of vm_page_t), reqpage (index of requested page)
- *
- * Return value:
- * number of pages in marray
- */
-static int
-vm_fault_additional_pages(m, rbehind, rahead, marray, reqpage)
- vm_page_t m;
- int rbehind;
- int rahead;
- vm_page_t *marray;
- int *reqpage;
-{
- int i,j;
- vm_object_t object;
- vm_pindex_t pindex, startpindex, endpindex, tpindex;
- vm_page_t rtm;
- int cbehind, cahead;
-
- VM_OBJECT_ASSERT_WLOCKED(m->object);
-
- object = m->object;
- pindex = m->pindex;
- cbehind = cahead = 0;
-
- /*
- * if the requested page is not available, then give up now
- */
- if (!vm_pager_has_page(object, pindex, &cbehind, &cahead)) {
- return 0;
- }
-
- if ((cbehind == 0) && (cahead == 0)) {
- *reqpage = 0;
- marray[0] = m;
- return 1;
- }
-
- if (rahead > cahead) {
- rahead = cahead;
- }
-
- if (rbehind > cbehind) {
- rbehind = cbehind;
- }
-
- /*
- * scan backward for the read behind pages -- in memory
- */
- if (pindex > 0) {
- if (rbehind > pindex) {
- rbehind = pindex;
- startpindex = 0;
- } else {
- startpindex = pindex - rbehind;
- }
-
- if ((rtm = TAILQ_PREV(m, pglist, listq)) != NULL &&
- rtm->pindex >= startpindex)
- startpindex = rtm->pindex + 1;
-
- /* tpindex is unsigned; beware of numeric underflow. */
- for (i = 0, tpindex = pindex - 1; tpindex >= startpindex &&
- tpindex < pindex; i++, tpindex--) {
-
- rtm = vm_page_alloc(object, tpindex, VM_ALLOC_NORMAL |
- VM_ALLOC_IFNOTCACHED);
- if (rtm == NULL) {
- /*
- * Shift the allocated pages to the
- * beginning of the array.
- */
- for (j = 0; j < i; j++) {
- marray[j] = marray[j + tpindex + 1 -
- startpindex];
- }
- break;
- }
-
- marray[tpindex - startpindex] = rtm;
- }
- } else {
- startpindex = 0;
- i = 0;
- }
-
- marray[i] = m;
- /* page offset of the required page */
- *reqpage = i;
-
- tpindex = pindex + 1;
- i++;
-
- /*
- * scan forward for the read ahead pages
- */
- endpindex = tpindex + rahead;
- if ((rtm = TAILQ_NEXT(m, listq)) != NULL && rtm->pindex < endpindex)
- endpindex = rtm->pindex;
- if (endpindex > object->size)
- endpindex = object->size;
-
- for (; tpindex < endpindex; i++, tpindex++) {
-
- rtm = vm_page_alloc(object, tpindex, VM_ALLOC_NORMAL |
- VM_ALLOC_IFNOTCACHED);
- if (rtm == NULL) {
- break;
- }
-
- marray[i] = rtm;
- }
-
- /* return number of pages */
- return i;
-}
-
-/*
* Block entry into the machine-independent layer's page fault handler by
* the calling thread. Subsequent calls to vm_fault() by that thread will
* return KERN_PROTECTION_FAILURE. Enable machine-dependent handling of
Index: sys/vm/vm_glue.c
===================================================================
--- sys/vm/vm_glue.c (revision 291639)
+++ sys/vm/vm_glue.c (working copy)
@@ -238,7 +238,7 @@ vm_imgact_hold_page(vm_object_t object, vm_ooffset
pindex = OFF_TO_IDX(offset);
m = vm_page_grab(object, pindex, VM_ALLOC_NORMAL);
if (m->valid != VM_PAGE_BITS_ALL) {
- rv = vm_pager_get_pages(object, &m, 1, 0);
+ rv = vm_pager_get_pages(object, &m, 1, NULL, NULL);
if (rv != VM_PAGER_OK) {
vm_page_lock(m);
vm_page_free(m);
@@ -567,37 +567,37 @@ vm_thread_swapin(struct thread *td)
{
vm_object_t ksobj;
vm_page_t ma[KSTACK_MAX_PAGES];
- int i, j, pages, rv;
+ int pages;
pages = td->td_kstack_pages;
ksobj = td->td_kstack_obj;
VM_OBJECT_WLOCK(ksobj);
- for (i = 0; i < pages; i++)
+ for (int i = 0; i < pages; i++)
ma[i] = vm_page_grab(ksobj, i, VM_ALLOC_NORMAL |
VM_ALLOC_WIRED);
- for (i = 0; i < pages; i++) {
- if (ma[i]->valid != VM_PAGE_BITS_ALL) {
- vm_page_assert_xbusied(ma[i]);
- vm_object_pip_add(ksobj, 1);
- for (j = i + 1; j < pages; j++) {
- if (ma[j]->valid != VM_PAGE_BITS_ALL)
- vm_page_assert_xbusied(ma[j]);
- if (ma[j]->valid == VM_PAGE_BITS_ALL)
- break;
- }
- rv = vm_pager_get_pages(ksobj, ma + i, j - i, 0);
- if (rv != VM_PAGER_OK)
- panic("vm_thread_swapin: cannot get kstack for proc: %d",
- td->td_proc->p_pid);
- /*
- * All pages in the array are in place, due to the
- * pager is always the swap pager, which doesn't
- * free or remove wired non-req pages from object.
- */
- vm_object_pip_wakeup(ksobj);
+ for (int i = 0; i < pages;) {
+ int j, a, count, rv;
+
+ vm_page_assert_xbusied(ma[i]);
+ if (ma[i]->valid == VM_PAGE_BITS_ALL) {
vm_page_xunbusy(ma[i]);
- } else if (vm_page_xbusied(ma[i]))
- vm_page_xunbusy(ma[i]);
+ i++;
+ continue;
+ }
+ vm_object_pip_add(ksobj, 1);
+ for (j = i + 1; j < pages; j++)
+ if (ma[j]->valid == VM_PAGE_BITS_ALL)
+ break;
+ rv = vm_pager_has_page(ksobj, ma[i]->pindex, NULL, &a);
+ KASSERT(rv == 1, ("%s: missing page %p", __func__, ma[i]));
+ count = min(a + 1, j - i);
+ rv = vm_pager_get_pages(ksobj, ma + i, count, NULL, NULL);
+ KASSERT(rv == VM_PAGER_OK, ("%s: cannot get kstack for proc %d",
+ __func__, td->td_proc->p_pid));
+ vm_object_pip_wakeup(ksobj);
+ for (j = i; j < i + count; j++)
+ vm_page_xunbusy(ma[j]);
+ i += count;
}
VM_OBJECT_WUNLOCK(ksobj);
pmap_qenter(td->td_kstack, ma, pages);
Index: sys/vm/vm_object.c
===================================================================
--- sys/vm/vm_object.c (revision 291639)
+++ sys/vm/vm_object.c (working copy)
@@ -2023,7 +2023,7 @@ vm_object_populate(vm_object_t object, vm_pindex_t
for (pindex = start; pindex < end; pindex++) {
m = vm_page_grab(object, pindex, VM_ALLOC_NORMAL);
if (m->valid != VM_PAGE_BITS_ALL) {
- rv = vm_pager_get_pages(object, &m, 1, 0);
+ rv = vm_pager_get_pages(object, &m, 1, NULL, NULL);
if (rv != VM_PAGER_OK) {
vm_page_lock(m);
vm_page_free(m);
Index: sys/vm/vm_object.h
===================================================================
--- sys/vm/vm_object.h (revision 291639)
+++ sys/vm/vm_object.h (working copy)
@@ -243,6 +243,8 @@ extern struct vm_object kmem_object_store;
rw_try_upgrade(&(object)->lock)
#define VM_OBJECT_WLOCK(object) \
rw_wlock(&(object)->lock)
+#define VM_OBJECT_WOWNED(object) \
+ rw_wowned(&(object)->lock)
#define VM_OBJECT_WUNLOCK(object) \
rw_wunlock(&(object)->lock)
Index: sys/vm/vm_page.c
===================================================================
--- sys/vm/vm_page.c (revision 291639)
+++ sys/vm/vm_page.c (working copy)
@@ -979,38 +979,28 @@ vm_page_free_zero(vm_page_t m)
/*
* Unbusy and handle the page queueing for a page from the VOP_GETPAGES()
- * array which is not the request page.
+ * array which was read ahead.
*/
void
vm_page_readahead_finish(vm_page_t m)
{
- if (m->valid != 0) {
- /*
- * Since the page is not the requested page, whether
- * it should be activated or deactivated is not
- * obvious. Empirical results have shown that
- * deactivating the page is usually the best choice,
- * unless the page is wanted by another thread.
- */
- vm_page_lock(m);
- if ((m->busy_lock & VPB_BIT_WAITERS) != 0)
- vm_page_activate(m);
- else
- vm_page_deactivate(m);
- vm_page_unlock(m);
- vm_page_xunbusy(m);
- } else {
- /*
- * Free the completely invalid page. Such page state
- * occurs due to the short read operation which did
- * not covered our page at all, or in case when a read
- * error happens.
- */
- vm_page_lock(m);
- vm_page_free(m);
- vm_page_unlock(m);
- }
+ /* We shouldn't put invalid pages on queues. */
+ KASSERT(m->valid != 0, ("%s: %p is invalid", __func__, m));
+
+ /*
+ * Since the page is not the actually needed one, whether it should
+ * be activated or deactivated is not obvious. Empirical results
+ * have shown that deactivating the page is usually the best choice,
+ * unless the page is wanted by another thread.
+ */
+ vm_page_lock(m);
+ if ((m->busy_lock & VPB_BIT_WAITERS) != 0)
+ vm_page_activate(m);
+ else
+ vm_page_deactivate(m);
+ vm_page_unlock(m);
+ vm_page_xunbusy(m);
}
/*
Index: sys/vm/vm_pager.c
===================================================================
--- sys/vm/vm_pager.c (revision 291639)
+++ sys/vm/vm_pager.c (working copy)
@@ -88,7 +88,7 @@ int cluster_pbuf_freecnt = -1; /* unlimited to beg
struct buf *swbuf;
-static int dead_pager_getpages(vm_object_t, vm_page_t *, int, int);
+static int dead_pager_getpages(vm_object_t, vm_page_t *, int, int *, int *);
static vm_object_t dead_pager_alloc(void *, vm_ooffset_t, vm_prot_t,
vm_ooffset_t, struct ucred *);
static void dead_pager_putpages(vm_object_t, vm_page_t *, int, int, int *);
@@ -96,13 +96,11 @@ static boolean_t dead_pager_haspage(vm_object_t, v
static void dead_pager_dealloc(vm_object_t);
static int
-dead_pager_getpages(obj, ma, count, req)
- vm_object_t obj;
- vm_page_t *ma;
- int count;
- int req;
+dead_pager_getpages(vm_object_t obj, vm_page_t *ma, int count, int *rbehind,
+ int *rahead)
{
- return VM_PAGER_FAIL;
+
+ return (VM_PAGER_FAIL);
}
static vm_object_t
@@ -282,45 +280,47 @@ vm_pager_assert_in(vm_object_t object, vm_page_t *
* The requested page must be fully valid on successful return.
*/
int
-vm_pager_get_pages(vm_object_t object, vm_page_t *m, int count, int reqpage)
+vm_pager_get_pages(vm_object_t object, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
+#ifdef INVARIANTS
+ vm_pindex_t pindex = m[0]->pindex;
+#endif
int r;
vm_pager_assert_in(object, m, count);
- r = (*pagertab[object->type]->pgo_getpages)(object, m, count, reqpage);
+ r = (*pagertab[object->type]->pgo_getpages)(object, m, count, rbehind,
+ rahead);
if (r != VM_PAGER_OK)
return (r);
- /*
- * If pager has replaced the page, assert that it had
- * updated the array. Also assert that page is still
- * busied.
- */
- KASSERT(m[reqpage] == vm_page_lookup(object, m[reqpage]->pindex),
- ("%s: mismatch page %p pindex %ju", __func__,
- m[reqpage], (uintmax_t )m[reqpage]->pindex));
- vm_page_assert_xbusied(m[reqpage]);
-
- /*
- * Pager didn't fill up entire page. Zero out
- * partially filled data.
- */
- if (m[reqpage]->valid != VM_PAGE_BITS_ALL)
- vm_page_zero_invalid(m[reqpage], TRUE);
-
+ for (int i = 0; i < count; i++) {
+ /*
+ * If pager has replaced a page, assert that it had
+ * updated the array.
+ */
+ KASSERT(m[i] == vm_page_lookup(object, pindex++),
+ ("%s: mismatch page %p pindex %ju", __func__,
+ m[i], (uintmax_t )pindex - 1));
+ /*
+ * Zero out partially filled data.
+ */
+ if (m[i]->valid != VM_PAGE_BITS_ALL)
+ vm_page_zero_invalid(m[i], TRUE);
+ }
return (VM_PAGER_OK);
}
int
vm_pager_get_pages_async(vm_object_t object, vm_page_t *m, int count,
- int reqpage, pgo_getpages_iodone_t iodone, void *arg)
+ int *rbehind, int *rahead, pgo_getpages_iodone_t iodone, void *arg)
{
vm_pager_assert_in(object, m, count);
return ((*pagertab[object->type]->pgo_getpages_async)(object, m,
- count, reqpage, iodone, arg));
+ count, rbehind, rahead, iodone, arg));
}
/*
@@ -355,39 +355,6 @@ vm_pager_object_lookup(struct pagerlst *pg_list, v
}
/*
- * Free the non-requested pages from the given array. To remove all pages,
- * caller should provide out of range reqpage number.
- */
-void
-vm_pager_free_nonreq(vm_object_t object, vm_page_t ma[], int reqpage,
- int npages, boolean_t object_locked)
-{
- enum { UNLOCKED, CALLER_LOCKED, INTERNALLY_LOCKED } locked;
- int i;
-
- if (object_locked) {
- VM_OBJECT_ASSERT_WLOCKED(object);
- locked = CALLER_LOCKED;
- } else {
- VM_OBJECT_ASSERT_UNLOCKED(object);
- locked = UNLOCKED;
- }
- for (i = 0; i < npages; ++i) {
- if (i != reqpage) {
- if (locked == UNLOCKED) {
- VM_OBJECT_WLOCK(object);
- locked = INTERNALLY_LOCKED;
- }
- vm_page_lock(ma[i]);
- vm_page_free(ma[i]);
- vm_page_unlock(ma[i]);
- }
- }
- if (locked == INTERNALLY_LOCKED)
- VM_OBJECT_WUNLOCK(object);
-}
-
-/*
* initialize a physical buffer
*/
Index: sys/vm/vm_pager.h
===================================================================
--- sys/vm/vm_pager.h (revision 291639)
+++ sys/vm/vm_pager.h (working copy)
@@ -50,9 +50,9 @@ typedef void pgo_init_t(void);
typedef vm_object_t pgo_alloc_t(void *, vm_ooffset_t, vm_prot_t, vm_ooffset_t,
struct ucred *);
typedef void pgo_dealloc_t(vm_object_t);
-typedef int pgo_getpages_t(vm_object_t, vm_page_t *, int, int);
+typedef int pgo_getpages_t(vm_object_t, vm_page_t *, int, int *, int *);
typedef void pgo_getpages_iodone_t(void *, vm_page_t *, int, int);
-typedef int pgo_getpages_async_t(vm_object_t, vm_page_t *, int, int,
+typedef int pgo_getpages_async_t(vm_object_t, vm_page_t *, int, int *, int *,
pgo_getpages_iodone_t, void *);
typedef void pgo_putpages_t(vm_object_t, vm_page_t *, int, int, int *);
typedef boolean_t pgo_haspage_t(vm_object_t, vm_pindex_t, int *, int *);
@@ -106,14 +106,12 @@ vm_object_t vm_pager_allocate(objtype_t, void *, v
vm_ooffset_t, struct ucred *);
void vm_pager_bufferinit(void);
void vm_pager_deallocate(vm_object_t);
-int vm_pager_get_pages(vm_object_t, vm_page_t *, int, int);
-int vm_pager_get_pages_async(vm_object_t, vm_page_t *, int, int,
+int vm_pager_get_pages(vm_object_t, vm_page_t *, int, int *, int *);
+int vm_pager_get_pages_async(vm_object_t, vm_page_t *, int, int *, int *,
pgo_getpages_iodone_t, void *);
static __inline boolean_t vm_pager_has_page(vm_object_t, vm_pindex_t, int *, int *);
void vm_pager_init(void);
vm_object_t vm_pager_object_lookup(struct pagerlst *, void *);
-void vm_pager_free_nonreq(vm_object_t object, vm_page_t ma[], int reqpage,
- int npages, boolean_t object_locked);
static __inline void
vm_pager_put_pages(
Index: sys/vm/vnode_pager.c
===================================================================
--- sys/vm/vnode_pager.c (revision 291639)
+++ sys/vm/vnode_pager.c (working copy)
@@ -84,11 +84,9 @@ static int vnode_pager_addr(struct vnode *vp, vm_o
static int vnode_pager_input_smlfs(vm_object_t object, vm_page_t m);
static int vnode_pager_input_old(vm_object_t object, vm_page_t m);
static void vnode_pager_dealloc(vm_object_t);
-static int vnode_pager_local_getpages0(struct vnode *, vm_page_t *, int, int,
- vop_getpages_iodone_t, void *);
-static int vnode_pager_getpages(vm_object_t, vm_page_t *, int, int);
-static int vnode_pager_getpages_async(vm_object_t, vm_page_t *, int, int,
- vop_getpages_iodone_t, void *);
+static int vnode_pager_getpages(vm_object_t, vm_page_t *, int, int *, int *);
+static int vnode_pager_getpages_async(vm_object_t, vm_page_t *, int, int *,
+ int *, vop_getpages_iodone_t, void *);
static void vnode_pager_putpages(vm_object_t, vm_page_t *, int, int, int *);
static boolean_t vnode_pager_haspage(vm_object_t, vm_pindex_t, int *, int *);
static vm_object_t vnode_pager_alloc(void *, vm_ooffset_t, vm_prot_t,
@@ -673,15 +671,15 @@ vnode_pager_input_old(vm_object_t object, vm_page_
* backing vp's VOP_GETPAGES.
*/
static int
-vnode_pager_getpages(vm_object_t object, vm_page_t *m, int count, int reqpage)
+vnode_pager_getpages(vm_object_t object, vm_page_t *m, int count, int *rbehind,
+ int *rahead)
{
+ struct vnode *vp;
int rtval;
- struct vnode *vp;
- int bytes = count * PAGE_SIZE;
vp = object->handle;
VM_OBJECT_WUNLOCK(object);
- rtval = VOP_GETPAGES(vp, m, bytes, reqpage);
+ rtval = VOP_GETPAGES(vp, m, count, rbehind, rahead);
KASSERT(rtval != EOPNOTSUPP,
("vnode_pager: FS getpages not implemented\n"));
VM_OBJECT_WLOCK(object);
@@ -690,7 +688,7 @@ static int
static int
vnode_pager_getpages_async(vm_object_t object, vm_page_t *m, int count,
- int reqpage, vop_getpages_iodone_t iodone, void *arg)
+ int *rbehind, int *rahead, vop_getpages_iodone_t iodone, void *arg)
{
struct vnode *vp;
int rtval;
@@ -697,8 +695,7 @@ vnode_pager_getpages_async(vm_object_t object, vm_
vp = object->handle;
VM_OBJECT_WUNLOCK(object);
- rtval = VOP_GETPAGES_ASYNC(vp, m, count * PAGE_SIZE, reqpage,
- iodone, arg);
+ rtval = VOP_GETPAGES_ASYNC(vp, m, count, rbehind, rahead, iodone, arg);
KASSERT(rtval != EOPNOTSUPP,
("vnode_pager: FS getpages_async not implemented\n"));
VM_OBJECT_WLOCK(object);
@@ -714,8 +711,8 @@ int
vnode_pager_local_getpages(struct vop_getpages_args *ap)
{
- return (vnode_pager_local_getpages0(ap->a_vp, ap->a_m, ap->a_count,
- ap->a_reqpage, NULL, NULL));
+ return (vnode_pager_generic_getpages(ap->a_vp, ap->a_m, ap->a_count,
+ ap->a_rbehind, ap->a_rahead, NULL, NULL));
}
int
@@ -722,68 +719,54 @@ int
vnode_pager_local_getpages_async(struct vop_getpages_async_args *ap)
{
- return (vnode_pager_local_getpages0(ap->a_vp, ap->a_m, ap->a_count,
- ap->a_reqpage, ap->a_iodone, ap->a_arg));
+ return (vnode_pager_generic_getpages(ap->a_vp, ap->a_m, ap->a_count,
+ ap->a_rbehind, ap->a_rahead, ap->a_iodone, ap->a_arg));
}
-static int
-vnode_pager_local_getpages0(struct vnode *vp, vm_page_t *m, int bytecount,
- int reqpage, vop_getpages_iodone_t iodone, void *arg)
-{
- vm_page_t mreq;
-
- mreq = m[reqpage];
-
- /*
- * Since the caller has busied the requested page, that page's valid
- * field will not be changed by other threads.
- */
- vm_page_assert_xbusied(mreq);
-
- /*
- * The requested page has valid blocks. Invalid part can only
- * exist at the end of file, and the page is made fully valid
- * by zeroing in vm_pager_get_pages(). Free non-requested
- * pages, since no i/o is done to read its content.
- */
- if (mreq->valid != 0) {
- vm_pager_free_nonreq(mreq->object, m, reqpage,
- round_page(bytecount) / PAGE_SIZE, FALSE);
- if (iodone != NULL)
- iodone(arg, m, reqpage, 0);
- return (VM_PAGER_OK);
- }
-
- return (vnode_pager_generic_getpages(vp, m, bytecount, reqpage,
- iodone, arg));
-}
-
/*
* This is now called from local media FS's to operate against their
* own vnodes if they fail to implement VOP_GETPAGES.
*/
int
-vnode_pager_generic_getpages(struct vnode *vp, vm_page_t *m, int bytecount,
- int reqpage, vop_getpages_iodone_t iodone, void *arg)
+vnode_pager_generic_getpages(struct vnode *vp, vm_page_t *m, int count,
+ int *a_rbehind, int *a_rahead, vop_getpages_iodone_t iodone, void *arg)
{
vm_object_t object;
struct bufobj *bo;
struct buf *bp;
- daddr_t firstaddr, reqblock;
- off_t foff, pib;
- int pbefore, pafter, i, size, bsize, first, last, *freecnt;
- int count, error, before, after, secmask;
+ off_t foff;
+ int bsize, pagesperblock, *freecnt;
+ int error, before, after, rbehind, rahead, poff, i;
+ int bytecount, secmask;
KASSERT(vp->v_type != VCHR && vp->v_type != VBLK,
- ("vnode_pager_generic_getpages does not support devices"));
+ ("%s does not support devices", __func__));
+
if (vp->v_iflag & VI_DOOMED)
return (VM_PAGER_BAD);
object = vp->v_object;
- count = bytecount / PAGE_SIZE;
+ foff = IDX_TO_OFF(m[0]->pindex);
bsize = vp->v_mount->mnt_stat.f_iosize;
+ pagesperblock = bsize / PAGE_SIZE;
+ KASSERT(foff < object->un_pager.vnp.vnp_size,
+ ("%s: page %p offset beyond vp %p size", __func__, m[0], vp));
+ KASSERT(count <= sizeof(bp->b_pages),
+ ("%s: requested %d pages", __func__, count));
+
/*
+ * The last page has valid blocks. Invalid part can only
+ * exist at the end of file, and the page is made fully valid
+ * by zeroing in vm_pager_get_pages().
+ */
+ if (m[count - 1]->valid != 0 && --count == 0) {
+ if (iodone != NULL)
+ iodone(arg, m, 1, 0);
+ return (VM_PAGER_OK);
+ }
+
+ /*
* Synchronous and asynchronous paging operations use different
* free pbuf counters. This is done to avoid asynchronous requests
* to consume all pbufs.
@@ -800,130 +783,182 @@ int
* If the file system doesn't support VOP_BMAP, use old way of
* getting pages via VOP_READ.
*/
- error = VOP_BMAP(vp, IDX_TO_OFF(m[reqpage]->pindex) / bsize, &bo,
- &reqblock, &after, &before);
+ error = VOP_BMAP(vp, foff / bsize, &bo, &bp->b_blkno, &after, &before);
if (error == EOPNOTSUPP) {
relpbuf(bp, freecnt);
VM_OBJECT_WLOCK(object);
- for (i = 0; i < count; i++)
- if (i != reqpage) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
- }
- PCPU_INC(cnt.v_vnodein);
- PCPU_INC(cnt.v_vnodepgsin);
- error = vnode_pager_input_old(object, m[reqpage]);
+ for (i = 0; i < count; i++) {
+ PCPU_INC(cnt.v_vnodein);
+ PCPU_INC(cnt.v_vnodepgsin);
+ error = vnode_pager_input_old(object, m[i]);
+ if (error)
+ break;
+ }
VM_OBJECT_WUNLOCK(object);
return (error);
} else if (error != 0) {
relpbuf(bp, freecnt);
- vm_pager_free_nonreq(object, m, reqpage, count, FALSE);
return (VM_PAGER_ERROR);
-
- /*
- * If the blocksize is smaller than a page size, then use
- * special small filesystem code.
- */
- } else if ((PAGE_SIZE / bsize) > 1) {
- relpbuf(bp, freecnt);
- vm_pager_free_nonreq(object, m, reqpage, count, FALSE);
- PCPU_INC(cnt.v_vnodein);
- PCPU_INC(cnt.v_vnodepgsin);
- return (vnode_pager_input_smlfs(object, m[reqpage]));
}
/*
- * Since the caller has busied the requested page, that page's valid
- * field will not be changed by other threads.
+ * If the file system supports BMAP, but blocksize is smaller
+ * than a page size, then use special small filesystem code.
*/
- vm_page_assert_xbusied(m[reqpage]);
+ if (pagesperblock == 0) {
+ for (i = 0; i < count; i++) {
+ PCPU_INC(cnt.v_vnodein);
+ PCPU_INC(cnt.v_vnodepgsin);
+ error = vnode_pager_input_smlfs(object, m[i]);
+ if (error)
+ break;
+ }
+ return (error);
+ }
/*
- * If we have a completely valid page available to us, we can
- * clean up and return. Otherwise we have to re-read the
- * media.
+ * A sparse file can be encountered only for a single page request,
+ * which may not be preceeded by call to vm_pager_haspage().
*/
- if (m[reqpage]->valid == VM_PAGE_BITS_ALL) {
+ if (bp->b_blkno == -1) {
+ KASSERT(count == 1,
+ ("%s: array[%d] request to a sparse file %p", __func__,
+ count, vp));
relpbuf(bp, freecnt);
- vm_pager_free_nonreq(object, m, reqpage, count, FALSE);
- return (VM_PAGER_OK);
- } else if (reqblock == -1) {
- relpbuf(bp, freecnt);
- pmap_zero_page(m[reqpage]);
- KASSERT(m[reqpage]->dirty == 0,
- ("vnode_pager_generic_getpages: page %p is dirty", m));
+ pmap_zero_page(m[0]);
+ KASSERT(m[0]->dirty == 0, ("%s: page %p is dirty",
+ __func__, m[0]));
VM_OBJECT_WLOCK(object);
- m[reqpage]->valid = VM_PAGE_BITS_ALL;
- vm_pager_free_nonreq(object, m, reqpage, count, TRUE);
+ m[0]->valid = VM_PAGE_BITS_ALL;
VM_OBJECT_WUNLOCK(object);
return (VM_PAGER_OK);
- } else if (m[reqpage]->valid != 0) {
- VM_OBJECT_WLOCK(object);
- m[reqpage]->valid = 0;
- VM_OBJECT_WUNLOCK(object);
}
- pib = IDX_TO_OFF(m[reqpage]->pindex) % bsize;
- pbefore = ((daddr_t)before * bsize + pib) / PAGE_SIZE;
- pafter = ((daddr_t)(after + 1) * bsize - pib) / PAGE_SIZE - 1;
- first = reqpage < pbefore ? 0 : reqpage - pbefore;
- last = reqpage + pafter >= count ? count - 1 : reqpage + pafter;
- if (first > 0 || last + 1 < count) {
+ bp->b_blkno += (foff % bsize) / DEV_BSIZE;
+
+ /* Recalculate blocks available after/before to pages. */
+ poff = (foff % bsize) / PAGE_SIZE;
+ before *= pagesperblock;
+ before += poff;
+ after *= pagesperblock;
+ after += pagesperblock - (poff + 1);
+ if (m[0]->pindex + after >= object->size)
+ after = object->size - 1 - m[0]->pindex;
+ KASSERT(count <= after + 1, ("%s: %d pages asked, can do only %d",
+ __func__, count, after + 1));
+ after -= count - 1;
+
+ /* Trim requested rbehind/rahead to possible values. */
+ rbehind = a_rbehind ? *a_rbehind : 0;
+ rahead = a_rahead ? *a_rahead : 0;
+ rbehind = min(rbehind, before);
+ rbehind = min(rbehind, m[0]->pindex);
+ rahead = min(rahead, after);
+ rahead = min(rahead, object->size - m[count - 1]->pindex);
+ KASSERT(rbehind + rahead + count <= sizeof(bp->b_pages),
+ ("%s: behind %d ahead %d count %d", __func__,
+ rbehind, rahead, count));
+
+ /*
+ * Fill in the bp->b_pages[] array with requested and optional
+ * read behind or read ahead pages. Read behind pages are looked
+ * up in a backward direction, down to a first cached page. Same
+ * for read ahead pages, but there is no need to shift the array
+ * in case of encountering a cached page.
+ */
+ i = bp->b_npages = 0;
+ if (rbehind) {
+ vm_pindex_t startpindex, tpindex;
+ vm_page_t p;
+
VM_OBJECT_WLOCK(object);
- for (i = 0; i < first; i++) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
+ startpindex = m[0]->pindex - rbehind;
+ if ((p = TAILQ_PREV(m[0], pglist, listq)) != NULL &&
+ p->pindex >= startpindex)
+ startpindex = p->pindex + 1;
+
+ /* tpindex is unsigned; beware of numeric underflow. */
+ for (tpindex = m[0]->pindex - 1;
+ tpindex >= startpindex && tpindex < m[0]->pindex;
+ tpindex--, i++) {
+ p = vm_page_alloc(object, tpindex, VM_ALLOC_NORMAL |
+ VM_ALLOC_IFNOTCACHED);
+ if (p == NULL) {
+ /* Shift the array. */
+ for (int j = 0; j < i; j++)
+ bp->b_pages[j] = bp->b_pages[j +
+ tpindex + 1 - startpindex];
+ break;
+ }
+ bp->b_pages[tpindex - startpindex] = p;
}
- for (i = last + 1; i < count; i++) {
- vm_page_lock(m[i]);
- vm_page_free(m[i]);
- vm_page_unlock(m[i]);
+
+ bp->b_pgbefore = i;
+ bp->b_npages += i;
+ bp->b_blkno -= IDX_TO_OFF(i) / DEV_BSIZE;
+ } else
+ bp->b_pgbefore = 0;
+
+ /* Requested pages. */
+ for (int j = 0; j < count; j++, i++)
+ bp->b_pages[i] = m[j];
+ bp->b_npages += count;
+
+ if (rahead) {
+ vm_pindex_t endpindex, tpindex;
+ vm_page_t p;
+
+ if (!VM_OBJECT_WOWNED(object))
+ VM_OBJECT_WLOCK(object);
+ endpindex = m[count - 1]->pindex + rahead + 1;
+ if ((p = TAILQ_NEXT(m[count - 1], listq)) != NULL &&
+ p->pindex < endpindex)
+ endpindex = p->pindex;
+ if (endpindex > object->size)
+ endpindex = object->size;
+
+ for (tpindex = m[count - 1]->pindex + 1;
+ tpindex < endpindex; i++, tpindex++) {
+ p = vm_page_alloc(object, tpindex, VM_ALLOC_NORMAL |
+ VM_ALLOC_IFNOTCACHED);
+ if (p == NULL)
+ break;
+ bp->b_pages[i] = p;
}
- VM_OBJECT_WUNLOCK(object);
- }
- /*
- * here on direct device I/O
- */
- firstaddr = reqblock;
- firstaddr += pib / DEV_BSIZE;
- firstaddr -= IDX_TO_OFF(reqpage - first) / DEV_BSIZE;
+ bp->b_pgafter = i - bp->b_npages;
+ bp->b_npages = i;
+ } else
+ bp->b_pgafter = 0;
- /*
- * The first and last page have been calculated now, move
- * input pages to be zero based, and adjust the count.
- */
- m += first;
- reqpage -= first;
- count = last - first + 1;
+ if (VM_OBJECT_WOWNED(object))
+ VM_OBJECT_WUNLOCK(object);
- /*
- * calculate the file virtual address for the transfer
- */
- foff = IDX_TO_OFF(m[0]->pindex);
+ /* Report back actual behind/ahead read. */
+ if (a_rbehind)
+ *a_rbehind = bp->b_pgbefore;
+ if (a_rahead)
+ *a_rahead = bp->b_pgafter;
- /*
- * calculate the size of the transfer
- */
- size = count * PAGE_SIZE;
- KASSERT(count > 0, ("zero count"));
- if ((foff + size) > object->un_pager.vnp.vnp_size)
- size = object->un_pager.vnp.vnp_size - foff;
- KASSERT(size > 0, ("zero size"));
+ KASSERT(bp->b_npages <= sizeof(bp->b_pages),
+ ("%s: buf %p overflowed", __func__, bp));
/*
- * round up physical size for real devices.
+ * Recalculate first offset and bytecount with regards to read behind.
+ * Truncate bytecount to vnode real size and round up physical size
+ * for real devices.
*/
+ foff = IDX_TO_OFF(bp->b_pages[0]->pindex);
+ bytecount = bp->b_npages << PAGE_SHIFT;
+ if ((foff + bytecount) > object->un_pager.vnp.vnp_size)
+ bytecount = object->un_pager.vnp.vnp_size - foff;
secmask = bo->bo_bsize - 1;
KASSERT(secmask < PAGE_SIZE && secmask > 0,
- ("vnode_pager_generic_getpages: sector size %d too large",
- secmask + 1));
- size = (size + secmask) & ~secmask;
+ ("%s: sector size %d too large", __func__, secmask + 1));
+ bytecount = (bytecount + secmask) & ~secmask;
/*
- * and map the pages to be read into the kva, if the filesystem
+ * And map the pages to be read into the kva, if the filesystem
* requires mapped buffers.
*/
if ((vp->v_mount->mnt_kern_flag & MNTK_UNMAPPED_BUFS) != 0 &&
@@ -932,41 +967,32 @@ int
bp->b_offset = 0;
} else {
bp->b_data = bp->b_kvabase;
- pmap_qenter((vm_offset_t)bp->b_data, m, count);
+ pmap_qenter((vm_offset_t)bp->b_data, bp->b_pages, bp->b_npages);
}
- /* build a minimal buffer header */
+ /* Build a minimal buffer header. */
bp->b_iocmd = BIO_READ;
KASSERT(bp->b_rcred == NOCRED, ("leaking read ucred"));
KASSERT(bp->b_wcred == NOCRED, ("leaking write ucred"));
bp->b_rcred = crhold(curthread->td_ucred);
bp->b_wcred = crhold(curthread->td_ucred);
- bp->b_blkno = firstaddr;
pbgetbo(bo, bp);
bp->b_vp = vp;
- bp->b_bcount = size;
- bp->b_bufsize = size;
- bp->b_runningbufspace = bp->b_bufsize;
- for (i = 0; i < count; i++)
- bp->b_pages[i] = m[i];
- bp->b_npages = count;
- bp->b_pager.pg_reqpage = reqpage;
+ bp->b_bcount = bp->b_bufsize = bp->b_runningbufspace = bytecount;
+ bp->b_iooffset = dbtob(bp->b_blkno);
+
atomic_add_long(&runningbufspace, bp->b_runningbufspace);
-
PCPU_INC(cnt.v_vnodein);
- PCPU_ADD(cnt.v_vnodepgsin, count);
+ PCPU_ADD(cnt.v_vnodepgsin, bp->b_npages);
- /* do the input */
- bp->b_iooffset = dbtob(bp->b_blkno);
-
if (iodone != NULL) { /* async */
- bp->b_pager.pg_iodone = iodone;
+ bp->b_pgiodone = iodone;
bp->b_caller1 = arg;
bp->b_iodone = vnode_pager_generic_getpages_done_async;
bp->b_flags |= B_ASYNC;
BUF_KERNPROC(bp);
bstrategy(bp);
- /* Good bye! */
+ return (VM_PAGER_OK);
} else {
bp->b_iodone = bdone;
bstrategy(bp);
@@ -977,9 +1003,8 @@ int
bp->b_vp = NULL;
pbrelbo(bp);
relpbuf(bp, &vnode_pbuf_freecnt);
+ return (error != 0 ? VM_PAGER_ERROR : VM_PAGER_OK);
}
-
- return (error != 0 ? VM_PAGER_ERROR : VM_PAGER_OK);
}
static void
@@ -988,8 +1013,7 @@ vnode_pager_generic_getpages_done_async(struct buf
int error;
error = vnode_pager_generic_getpages_done(bp);
- bp->b_pager.pg_iodone(bp->b_caller1, bp->b_pages,
- bp->b_pager.pg_reqpage, error);
+ bp->b_pgiodone(bp->b_caller1, bp->b_pages, bp->b_npages, error);
for (int i = 0; i < bp->b_npages; i++)
bp->b_pages[i] = NULL;
bp->b_vp = NULL;
@@ -1052,8 +1076,8 @@ vnode_pager_generic_getpages_done(struct buf *bp)
object->un_pager.vnp.vnp_size - tfoff)) == 0,
("%s: page %p is dirty", __func__, mt));
}
-
- if (i != bp->b_pager.pg_reqpage)
+
+ if (i < bp->b_pgbefore || i >= bp->b_npages - bp->b_pgafter)
vm_page_readahead_finish(mt);
}
VM_OBJECT_WUNLOCK(object);
Index: sys/vm/vnode_pager.h
===================================================================
--- sys/vm/vnode_pager.h (revision 291639)
+++ sys/vm/vnode_pager.h (working copy)
@@ -41,7 +41,8 @@
#ifdef _KERNEL
int vnode_pager_generic_getpages(struct vnode *vp, vm_page_t *m,
- int count, int reqpage, vop_getpages_iodone_t iodone, void *arg);
+ int count, int *rbehind, int *rahead, vop_getpages_iodone_t iodone,
+ void *arg);
int vnode_pager_generic_putpages(struct vnode *vp, vm_page_t *m,
int count, boolean_t sync,
int *rtvals);
--OwLcNYc0lM97+oe1--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20151205052940.GJ42565>