- 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.
adjusted via sysctl. file systems that have hash tables which are
sized based on the value of this variable now resize those hash tables
using the new value. the max number of FFS softdeps is also recalculated.
convert various file systems to use the <sys/queue.h> macros for
their hash tables.
PR4394: be more consistent with other MSDOSFS_DEBUG messages
PR4395: fix generation numbers as in the PR, and fix short name for e.g. x.aaaa
PR4396: easier fix then given in the PR
All PRs by Rick Byers. Thanks Rick for pointing these out
Don't panic if renaming a file to itself
Don't try to keep access times, there is no place for them
While being here, fix some minor bugs with VFAT handling
(Of course, nfs shouldn't rely on the credentials not being referenced)
Don't give directory entry to deget, it could result in a deadlock
Use device blocks, not clusters for logical block numbers
Correct handling of rmdir'ing open directories
Correct implementation of rename (includes renaming of directories)
Handle root directories that are not multiple clusters in size
* Update the `archive' bit any time the file's time stamp is updated.
* Don't set the `archive' bit automatically for directories, and don't
update the time stamp of directories automatically. (There are the DOS
semantics.)
* Use DE_TIMES() to change the time stamp in deupdat(). Remove the extra
time stamp argument, as it's superfluous, and we may want to make changes
without touching the time stamp.
* Allow denode updates iff DE_MODIFIED is set after DE_TIMES() is called.
This allows us to make changes that are totally independent of the time
stamp, and to avoid rewriting the entry if the time stamp was ignored (e.g.
for directories).
* Make time stamp changes in setattr() asynchronous.
* Make some of the code look more like UFS.
* Check permissions in lookup().
* Move the directory size special case into DE_EXTERNALIZE().
* Fix some conditions where lookup() might not release a buffer.
* Remove bogus flag handling in setattr().
* Pass timespec, not timeval, to deupdat().
* Check more error conditions.
* Fix possible panics in rename().
* Simplify readdir().
* General code cleanup; add prototypes, delete unused variables, etc.
<polk@bsdi.com>. His notes are as follows:
------------------------------------------------------------------------------
July 22, 1993
- Changed name of entire package from PCFS to MSDOSFS
- Fixed bugs:
root directory size in clusters instead of bytes
growing directory didn't update in-core size
link, symlink, mknod didn't free locked parent (deadlock)
lookup returned real error on create and rename instead of EJUSTRETURN
rename changed `.' entry in child instead of name entry in parent
rename removed `.' entry in child instead of removing entry in
parent when moving a directory from one dir to another
createde() left new node locked when write of parent failed (deadlock)
removede() decremented refcount even on error (rmdir's which failed
due to write errors left in-core cache entries inconsistent)
changed validation for filesystem to not check for the boot signature
since some disks (e.g., mtools) aren't bootable
directories are always show current time as modify time
(needed for NFS export since DOS never updates dir mod times --
ctime is true create time).
- Added support for cookies changes to the readdir() vnode
interface (#ifdef __bsdi__)
- Punted on the whole problem of inode generation numbers. This means
that there's a chance of using a stale file handle to access a new
file, but it doesn't appear to be the common case, and I don't see
how to generate reasonable generation numbers without changing something
on the disk (which is the way the SVR4 filesystem survival kit guys
did it). I don't think it would be very safe to change the on-disk
format.
Jeff Polk (polk@BSDI.COM)
------------------------------------------------------------------------------
chs