peter [Sun, 16 May 2004 22:11:50 +0000 (22:11 +0000)]
Kill the LAZYPMAP ifdefs. While they worked, they didn't do anything
to help the AMD cpus (which have a hardware tlb flush filter). I held
off to see what the 64 bit Intel cpus did, but it doesn't seem to help
much there either. Oh well, store it in the Attic.
njl [Sun, 16 May 2004 22:05:25 +0000 (22:05 +0000)]
Remove myself from da(4). MIHIRA Sanpei Yoshiro will be taking over quirk
maintenance given his work on USB. Also, the root cause of spamming da(4)
with NO_6_BYTE quirks was fixed last year and the extraneous quirks have
been removed. Please coordinate future quirk issues with sanpei@
joerg [Sun, 16 May 2004 21:11:46 +0000 (21:11 +0000)]
After successfully attaching an iicbus instance, instead of using a
NULL name in device_add_child(), explicitly name all of our known
child drivers in order to give them a chance to attach to us.
Otherwise, only the first one present would be probed and attached.
peter [Sun, 16 May 2004 20:44:41 +0000 (20:44 +0000)]
For consistency with i386, have pmap_kenter_temporary() take a vm_paddr_t
argument. It is actually the same type on amd64 (vm_paddr_t = vm_offset_t)
but this reduces the i386<->amd64 diffs a little.
peter [Sun, 16 May 2004 20:11:38 +0000 (20:11 +0000)]
Enable first part of kld's on amd64. This is known to not work right
yet, but building kld's is OK now and they can be loaded by kldload(2).
(but the machine will likely crash soon afterwards, a "minor" problem :-)
peter [Sun, 16 May 2004 20:00:28 +0000 (20:00 +0000)]
Make a small revision to the api between the elf linker core and the
elf_reloc() backends for two reasons. First, to support the possibility
of there being two elf linkers in the kernel (eg: amd64), and second, to
pass the relocbase explicitly (for relocating .o format kld files).
yar [Sun, 16 May 2004 19:29:33 +0000 (19:29 +0000)]
Add two new options to cron(8), -J and -j. They allow to specify
the maximum amount of time jitter for root and other users, respectively.
Before starting a job, cron(8) will sleep a random number of seconds,
from 0 to the amount specified. This can help to smooth down load spikes
when a lot of jobs are to start at the beginning of a particular minute
(e.g., the first minute of an hour.)
bde [Fri, 14 May 2004 20:51:42 +0000 (20:51 +0000)]
Fixed some common printf format errors. Don't assume that "struct foo *"
is "void *" (it isn't) or that the default promotion of pid_t is int.
Instead, assume that casting "struct foo *" to "void *" and printing the
result with %p is useful, and that all pid_t's are representable as longs.
Fixed some minor style bugs (mainly spelling errors in comments).
bde [Fri, 14 May 2004 19:52:35 +0000 (19:52 +0000)]
Style fixes:
Main ones: mostly use conditional expressions in ifdefs instead of a
mixture of conditional expressions and nested ifdefs.
Nearby ones:
- don't do less than echo the code in the comment about libc_r
- fixed some internal insertion sort errors and indentation errors.
joerg [Fri, 14 May 2004 18:46:16 +0000 (18:46 +0000)]
When I wrote this man page more than 5 years ago, I simply didn't
understand the true symmetric nature of Enigma, so my description of
``automatically detects that the input is encrypted'' was simply
wrong. Replace that by a more accurate description of why feeding the
ciphertext again into the engine will decrypt it.
njl [Fri, 14 May 2004 16:52:39 +0000 (16:52 +0000)]
Instead of scanning the entire lower 1 MB of RAM, only scan locations
where the RSD PTR can actually occur. According to section 5.2.2
of the ACPI spec, we only consider two regions for the base address:
1. EBDA (0x0 - 0x3FF)
2. High memory (0xE0000 - 0xFFFFF)
I don't know whether this fixes any actual problems but is more correct.
cognet [Fri, 14 May 2004 12:13:06 +0000 (12:13 +0000)]
Import the softfloat emulation library, needed for FreeBSD/arm right now.
It should become useless when gcc 3.4 will be imported, as libgcc from
gcc 3.4 contains this bits for arm.
cognet [Fri, 14 May 2004 11:46:45 +0000 (11:46 +0000)]
Import FreeBSD/arm kernel bits.
It only supports sa1110 (on simics) right now, but xscale support should come
soon.
Some of the initial work has been provided by :
Stephane Potvin <sepotvin at videotron.ca>
Most of this comes from NetBSD.
njl [Fri, 14 May 2004 04:17:56 +0000 (04:17 +0000)]
Add support for GPE being a package of { reference, gpe bit }.
Rework the ECDT probe to pass all the parameters in a temporary struct.
Note why we are mostly ok evaluating _GLK so early.
wpaul [Fri, 14 May 2004 03:57:17 +0000 (03:57 +0000)]
Fix a bug which I discovered recently while doing IPv6 testing at
Wind River. In the IPv4 output path, one of the tests in ip_output()
checks how many slots are actually available in the interface output
queue before attempting to send a packet. If, for example, we need
to transmit a packet of 32K bytes over an interface with an MTU of
1500, we know it's going to take about 21 fragments to do it. If
there's less than 21 slots left in the output queue, there's no point
in transmitting anything at all: IP does not do retransmission, so
sending only some of the fragments would just be a waste of bandwidth.
(In an extreme case, if you're sending a heavy stream of fragmented
packets, you might find yourself sending nothing by the first fragment
of all your packets.) So if ip_output() notices there's not enough
room in the output queue to send the frame, it just dumps the packet
and returns ENOBUFS to the app.
It turns out ip6_output() lacks this code. Consequently, this caused
the netperf UDPIPV6_STREAM test to produce very poor results with large
write sizes. This commit adds code to check the remaining space in the
output queue and junk fragmented packets if they're too big to be
sent, just like with IPv4. (I can't imagine anyone's running an NFS
server using UDP over IPv6, but if they are, this will likely make them
a lot happier. :)
jdp [Fri, 14 May 2004 01:29:21 +0000 (01:29 +0000)]
Fix a potential stack buffer overflow on systems whose ACPI OEMID
fills its field (6 characters). In that case the OEMID is not
null-terminated, and the sprintf that was used would copy up to the
next null byte, which could be pretty far away.