kqueue provides a stateful and efficient event notification framework
currently supported events include socket, file, directory, fifo,
pipe, tty and device changes, and monitoring of processes and signals
kqueue is supported by all writable filesystems in NetBSD tree
(with exception of Coda) and all device drivers supporting poll(2)
based on work done by Jonathan Lemon for FreeBSD
initial NetBSD port done by Luke Mewburn and Jason Thorpe
This merge changes the device switch tables from static array to
dynamically generated by config(8).
- All device switches is defined as a constant structure in device drivers.
- The new grammer ``device-major'' is introduced to ``files''.
device-major <prefix> char <num> [block <num>] [<rules>]
- All device major numbers must be listed up in port dependent majors.<arch>
by using this grammer.
- Added the new naming convention.
The name of the device switch must be <prefix>_[bc]devsw for auto-generation
of device switch tables.
- The backward compatibility of loading block/character device
switch by LKM framework is broken. This is necessary to convert
from block/character device major to device name in runtime and vice versa.
- The restriction to assign device major by LKM is completely removed.
We don't need to reserve LKM entries for dynamic loading of device switch.
- In compile time, device major numbers list is packed into the kernel and
the LKM framework will refer it to assign device major number dynamically.
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.
Previous revision mistakely removed fix for DIAGNOSTIC "getblk: block size
invariant failed" panic, incorporated in rev. 1.5. The problem there
has been that depending on 'off', the code did bread() of 512 and after that
1024 bytes for same 'blkno'. Fix this differently than in rev. 1.5 -
if 'off' would bump the read over to next device block, trim the read to
end on block boundary and read rest in next loop iteration. This seems
to fix the problem, and avoids the performance hit rev. 1.5 incurred.
Fix bug in code handling holes for reads via read(2) - a variable
was incremented instead of decremented.
the final null byte.
This makes programs using fts(3) (like find(1)/du(1)) print correct
name again. Problem reported by Luke Mewburn in private e-mail, thanks.
under heavy vnode usage - vget() must be called with unlocked ntnode!
Actually check the vget() return value and DTRT if it fails (i.e. when
the vnode has been recycled for other use).
to recode Unicode names to other representation, like encoding form
or national character/code sets. This replaces the ugly NTFS_U28() hack.
Use these hooks to encode the filenames to UTF-8.
scan of all directory entries; the mangled file name we return to
userland on getdents(2) may end up being unsearchable otherwise
This change is useful even through we will soon change to encode the
NTFS file names to utf-8 - we want to support recoding of filenames to
national 8bit charsets eventually.
while here, remove some unnecessary casts, change NTFS_U28() macro so that
the mangled filename doesn't contain control characters and use malloc()/free()
instead MALLOC()/FREE() for rdbuf buffer
happens fairly commonly if e.g. some files are still opened
then the unmount is attempted
* if the second vflush() (including system vnodes) fails, we should
probably panic, it's too serious; for now, just add a comment
<vm/pglist.h> -> <uvm/uvm_pglist.h>
<vm/vm_inherit.h> -> <uvm/uvm_inherit.h>
<vm/vm_kern.h> -> into <uvm/uvm_extern.h>
<vm/vm_object.h> -> nothing
<vm/vm_pager.h> -> into <uvm/uvm_pager.h>
also includes a bunch of <vm/vm_page.h> include removals (due to redudancy
with <vm/vm.h>), and a scattering of other similar headers.
in vfs_detach(). vfs_done may free global filesystem's resources,
typically those allocated in respective filesystem's init function.
Needed so those filesystems which went in via LKM have a chance to
clean after themselves before unloading. This fixes random panics
when LKM for filesystem using pools was loaded and unloaded several
times.
For each leaf filesystem, add appropriate vfs_done routine.
proper file attribute (a.k.a. "resource fork") is picked up when
the filehandle is mapped back to filesystem object.
This should make attributes working on NFS-exported NTFS filesystems
or other layered filesystems using filehandles.
lock counts consistent. DIAGNOSTIC checks will barf otherwise.
XXX ntfs_ntput should not free the node, the reclaim vop should do it
XXX why does the ntfs code do its own internal locking?
also fixes bug introduced in previous commit - the check for VBAD device node
added in rev 1.20 was accidentaly run AFTER v_specinfo was touched,
making the check useless
lukem