routines. The descriptor contains parameters which could be used
within those routines (eg. ip_output() ).
On passing, add IPPROTO_PGM entry to netinet/in.h
abuses its argument, which is supposed to be an integer unit number, as
a pointer to the head of the 'struct slot' list. When this code was
hacked^Wupdated for newbus, a new mechanism for registering slot_irq_handler()
was put in place and the significance of the unit number was overlooked.
When registering an interrupt, we have both device_id and unit. The unit
number is passed as 'unit' but /sys/i386/usa/intr_machdep.c:register_intr()
drops unit on the floor and uses device_id instead. Since pccard_alloc_intr()
always sets device_id to 0, this means the unit number is always zero, and
slot_irq_handler() is always called with 0, which becomes a NULL pointer
which slot_irq_handler() tries to dereference and the kernel explodes.
Result: if you assign a PCMCIA driver in the kernel config file like this:
device wi0 at isa? port? irq?
Then the system will panic the moment a PCMCIA device is attached and
an interrupt is triggered.
The quick fix: make pccard_alloc_intr() pass the unit number as both
the device_id and unit arguments to register_pcic_intr(). The correct fix
would be to rewrite /sys/pccard to be less disgusting.
+ plug an mbuf leak when dummynet used with bridging
+ make prototype of dummynet_io consistent with usage
+ code cleanup so that now bandwidth regulation is precise to the
bit/s and not to (8*HZ) bit/s as before.
confusing the directory read cookie cache. The nfs_access implementation
for v2 mounts attempts to read from the directory if root is the user
so that root can't access cached files when the server remaps root
to some other user.
Submitted by: Doug Rabson <dfr@nlsystems.com>
Reviewed by: Matthew Dillon <dillon@apollo.backplane.com>
bug in the stats accounting (nicSendBDs counter was bogus when TX ring was
configured to be in host memory).
Update if_tireg.h to look for new firmware fix level.
* Make the network code in the bootstrap more chatty (helps debugging)
* Add nfs root stuff to cpu_rootconf(). I also added a check to make sure
it really was netbooting which allows the use of the same kernel for local
and network boots.
* Tweak the de driver so that it takes the speed setting from the console
for the alpha (some PWSs have broken de chipsets). This is the same
behaviour as NetBSD/alpha.
Submitted by: Andrew Gallatin <gallatin@cs.duke.edu>
piecemeal, middle-of-file writes for NFS. These hacks have caused no
end of trouble, especially when combined with mmap(). I've removed
them. Instead, NFS will issue a read-before-write to fully
instantiate the struct buf containing the write. NFS does, however,
optimize piecemeal appends to files. For most common file operations,
you will not notice the difference. The sole remaining fragment in
the VFS/BIO system is b_dirtyoff/end, which NFS uses to avoid cache
coherency issues with read-merge-write style operations. NFS also
optimizes the write-covers-entire-buffer case by avoiding the
read-before-write. There is quite a bit of room for further
optimization in these areas.
The VM system marks pages fully-valid (AKA vm_page_t->valid =
VM_PAGE_BITS_ALL) in several places, most noteably in vm_fault. This
is not correct operation. The vm_pager_get_pages() code is now
responsible for marking VM pages all-valid. A number of VM helper
routines have been added to aid in zeroing-out the invalid portions of
a VM page prior to the page being marked all-valid. This operation is
necessary to properly support mmap(). The zeroing occurs most often
when dealing with file-EOF situations. Several bugs have been fixed
in the NFS subsystem, including bits handling file and directory EOF
situations and buf->b_flags consistancy issues relating to clearing
B_ERROR & B_INVAL, and handling B_DONE.
getblk() and allocbuf() have been rewritten. B_CACHE operation is now
formally defined in comments and more straightforward in
implementation. B_CACHE for VMIO buffers is based on the validity of
the backing store. B_CACHE for non-VMIO buffers is based simply on
whether the buffer is B_INVAL or not (B_CACHE set if B_INVAL clear,
and vise-versa). biodone() is now responsible for setting B_CACHE
when a successful read completes. B_CACHE is also set when a bdwrite()
is initiated and when a bwrite() is initiated. VFS VOP_BWRITE
routines (there are only two - nfs_bwrite() and bwrite()) are now
expected to set B_CACHE. This means that bowrite() and bawrite() also
set B_CACHE indirectly.
There are a number of places in the code which were previously using
buf->b_bufsize (which is DEV_BSIZE aligned) when they should have
been using buf->b_bcount. These have been fixed. getblk() now clears
B_DONE on return because the rest of the system is so bad about
dealing with B_DONE.
Major fixes to NFS/TCP have been made. A server-side bug could cause
requests to be lost by the server due to nfs_realign() overwriting
other rpc's in the same TCP mbuf chain. The server's kernel must be
recompiled to get the benefit of the fixes.
Submitted by: Matthew Dillon <dillon@apollo.backplane.com>
Guy Helmer
Tatsumi Hosokawa
Luoqi Chen
Luigi Rizzo
Masafumi Max NAKANE
Bill Paul
KATO Takenori
Jordan K. Hubbard
Joseph Koshy
Brian Somers
Poul-Henning Kamp
Dima Ruban
Mike Smith
Bill Fumerola
Nick Hibma
Alan Cox
Doug Rabson
David E. O'Brien
Gary Palmer
Matt Jacob
Jun Kuriyama
Peter Wemm
Steve Price
Greg Lehey