me:
1) TLB flush optimization that effectively eliminates half of all of the
TLB flushes. This works by only flushing the TLB when a page is "present"
in memory (i.e. the valid bit is set in the page table entry). See section
5.3.5 of the Intel 386 Programmer's Reference Manual.
2) The handling of "CMAP" has been improved to catch attempts at multiple
simultaneous use.
John:
1) Added pmap_qenter/pmap_qremove functions for fast mapping of pages into
the kernel. This is for future optimizations and support for the upcoming
merged VM/buffer cache.
Reviewed by: John Dyson
the interface output queue and further udp packets would be fragmented
and only partially sent - keeping the output queue full and jamming the
network, but not actually getting any real work done (because you can't
send just 'part' of a udp packet - if you fragment it, you must send
the whole thing). The fix involves adding a check to make sure that the
output queue has sufficient space for all of the fragments.
Imported libmd. This library contains MD2, MD4 and MD5.
These three boggers pop up all over the place all of the time, so I
decided we needed a library with them. In general they are used for
security checks, so if you use them you want to link them static.
kernel a.out file, meaning that this is where the kernel starts.
(not at KERNBASE) - KERNBASE is 0xf0000000, while the kernel loads at
0xf0100000
Reviewed by:
a tty.
Note that this might conflict with the collateral use of TS_WOPEN, but
for the moment I can find no problems associated with this. (TS_WOPEN
will likely go away in the future anyway). This should be looked at
again in the future (the potential problem is that the cblock pool
may either run out or accumulate too many cblocks).
and cleandir: targets, simple use a CLEANFILES+= to handle this very
simple special case.
Add ${COPY} knob to install commands so that files don't disappear out
of the obj tree after a make install.
ready to go deal with just yet.
Disable man for now it will be fixed shortly, just wanted all the man
page converion stuff to be done togeather since that is a major functional
change and really belongs in a seperate commit.
1. Use ${MAKE} everywhere again. Whoops.
2. Replace multiple invocations of gzip ... split ... with one variable.
3. Add src-clean target for making the src tree presentable before
making a src tarball out of it.
upon disk type. In far more cases than not this is the optimal setting
for any disk drive made after 1990.
This now means all installs will have the disks newfs'ed with either:
newfs -b 8192 -f 1024 -d 0 -n 1
or
newfs -n 4096 -f 512 -d 0 -n 1
depending on what the user chooses for the blocking factor.
date!!) and rename them to something more eye-catching so people will read them
again (considering the previous state of affairs, I'm actually rather glad they didn't!).
1. Add to secr and bindists to possibly save the occasional fool who
doesn't RTFM and uses the wrong command to extract this (or even someone
who's legitimately using this to extract on top of a bindist somewhere
*else*).
2. Do the right thing with any symlinks in the src tree. Right now, we're
free of the buggers, but just in case.
I know that I said earlier that this should be unconditional behaviour,
but I thought about it a little more and concluded that the principle of least
surprise dictates that I make it an option.
