| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|
|
|
| |
Take a drive down if it's not open
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
this will allow us to manage bloat in the loader by using a bytecoded HLL
rather than lots of C code. It also offers an opportunity for vendors
or others with special applications to significantly customise the boot
process without having to commit to a divergent code branch.
This early commit is to allow others to experiment with the most effective
mechanisms for integrating FICL with the loader as it currently stands.
Ficl is distributed with the following license conditions:
"Ficl is freeware. Use it in any way that you like, with the understanding
that the code is not supported."
All source files contain authorship attributions.
Obtained from: John Sadler (john_sadler@alum.mit.edu)
|
| |
|
| |
|
| |
|
|
|
|
| |
less than accessible places.
|
|
|
|
|
| |
by the user-land program.
PR: i386/8344
|
|
|
|
|
| |
LRU. Use a 31-bit counter instead. If we decide to do heavy I/O through
the bootloader this will have to be revisited.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
and will bypass transfers for more than 8k. Blocks are invalidated after
2 seconds, so removable media should not confuse the cache.
The 8k threshold is a compromise; all UFS transfers performed by
libstand are 8k or less, so large file reads thrash the cache.
However many filesystem metadata operations are also performed using
8k blocks, so using a lower threshold gives poor performance.
Those of you with an eye for cache algorithms are welcome to tell me
how badly this one sucks; you can start with the 'bcachestats' command
which will print the contents of the cache and access statistics.
|
| |
|
|
|
|
| |
common than ','.
|
| |
|
|
|
|
|
|
| |
"Looking up host %s".
Suggested by: msmith
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
the top half to do it.
Put in a dubious check for subdisk integrity when trying to bring
up a plex where others are already up. This particular kludge is
crying out for a rewrite of the whole state code.
Add code to set_plex_state and set_volume_state to defer updates when
called from an interrupt context. This doesn't happen yet, but it
could do.
|
| |
|
|
|
|
|
|
| |
Ensure correct order of requests.
Call launch_requests at splhigh to see if this stops the mysterious
SMP panics
|
|
|
|
| |
Don't save invalid drive config information, remove the drive
|
|
|
|
| |
Get object counts right when removing
|
|
|
|
|
|
|
|
|
|
|
| |
the NFSv3 ACCESS RPC problems a little for busy clients that do a lot of
open/close. The nfs code could probably cache the results, but I'm not
sure whether this would be legal or useful. The problem is that with
a CPU farm, on each open there would be a lookup, getattr then access RPC
then the read/write RPC activity. Caching the access results probably
isn't going to help much if the clients access lots of files. Having the
nfs_access() routine interpret the getattr results is a bit of a hack, but
it's how NFSv2 is done and it might be OK for a mount attribute for v3.
|
|
|
|
|
|
| |
manipulation away from the length comparison. Measurements on beast.cdrom.com
show >3X improvement over the original code on large block sizes, putting the
performance on par with the optimized assembly code in libc.
|
| |
|
| |
|
|
|
|
|
| |
contributed their ideas to this document, among them Terry Lambert and
Bryan Mann, both @whistle.com. Thanks!
|
| |
|
|
|
|
|
|
| |
(official Philips I2C parallel interface)
Add comments for bktr port to the new I2C framework
|
| |
|
|
|
|
|
|
|
| |
most practical purposes, this should be indistinguishable from a
more strictly correct approach).
Feedback and testing: msmith
|
|
|
|
|
| |
instead the public writable directory /tmp
PR: conf/8330
|
|
|
|
|
|
| |
assert()
setjmp()/longjmp()
vsprintf()
|
|
|
|
| |
like others.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
build 2.2-stable worlds on 3.0-current systems again. objformat
calls getobjformat(), which doesn't exist in 2.2's libc.
Technically there should have been a version number bump when it was
added in -current. But it's used in so few places that it hardly
seems worth that. Besides, the objformat program is very heavily
used during a make world; it won't hurt to have it load a little
faster.
|
|
|
|
| |
Submitted by: Mike Spengler <mks@networkcs.com>
|
|
|
|
| |
Submitted by: Mike Spengler <mks@networkcs.com>
|
|
|
|
|
| |
Reviewed by: Mike Spengler <mks@networkcs.com>
Submitted by: phk
|
|
|
|
|
|
| |
response anyway. This is what Win95 does, despite it not
being documented in the cbcp spec....
Reported by: Bill <petheram@acm.org>
|
|
|
|
|
|
| |
line and in the ``load'' & ``dial'' commands. The last
label loaded becomes the current label name.
Only require a label for -auto mode.
|
|
|
|
|
|
|
|
|
| |
and increase the tx interrupt threshold to 4. This fixes performance
problems on slower systems.
Also fix a mind-o in the rx ring init routine: I used the TX
constant instead of the RX. This isn't a problem as long as the
rings are the same size, but if they aren't hijinx will ensue.
|
|
|
|
|
| |
a sysctl hierarchy (vm.stats.*). SYSCTL descriptions are only present
in source, they do not get compiled into the binaries taking up memory.
|
|
|
|
|
|
|
|
|
| |
* Embed the stack into the bss section for loader and netboot. This
is required for netboot since otherwise the stack would be inside our
heap.
* Install loader and netboot in /boot by default.
* Fix getbootfile so that it searches for a ',' instead of a ';'
when terminating the filename.
|
|
|
|
|
|
| |
sure that this is necessary to be a sync write here since a VOP_FSYNC()
follows and it will schedule, sort and complete the writes that the
vm_object_page_clean() started (as I think I understand things).
|
| |
|