Greetings,
I'm interested in using mmap() with MAP_STACK. After writing a couple of
test programs and looking at vm_map_insert() and vm_map_stack(), it
appears that vm_map_stack() behaves as if MAP_FIXED were set.
Why is this? I would like to allocate stack space without having to
search for
Is this a library issue in libc_r?
I just wrote a quick test program using rfork(RFMEM|RFPROC) to
create a child thread.
I then had both the parent and the child attempt to open the pre-existing
file /tmp/blah with O_EXLOCK set. When I specify O_NONBLOCK, the child
open() fails immediately.
Date: Tue, 27 Feb 2001 15:15:33 + (GMT)
From: E.B. Dreger [EMAIL PROTECTED]
Is this a library issue in libc_r?
I just wrote a quick test program using rfork(RFMEM|RFPROC) to
create a child thread.
Correction:
RFTHREAD|RFPROC
Sorry... I have RFMEM on the brain
Date: Thu, 1 Mar 2001 12:44:39 -0500 (EST)
From: Peter Dufault [EMAIL PROTECTED]
This is a stupid question, basically it's how to debug something.
I have four cooperating p-threaded processes. One of them keeps getting
a SIGSEGV with the address 0x752f422f. I'm not sure if that address
Date: Fri, 2 Mar 2001 08:04:52 -0500 (EST)
From: Peter Dufault [EMAIL PROTECTED]
Any strings "/B/u" in your program? That would be stored as 0x752f422f.
If you're using assembly with using %ebp for stack frame (yay!), then make
certain %esp isn't getting corrupted.
(I meant
Greetings,
This is something of a repost of an earlier question, but in a different
vein. I should mention that I'm using 4.2-R.
Has anyone modified vm_map_stack(), in /usr/src/sys/vm/vm_map.c, to search
for the first suitable open block?
I tried modifying vm_map_stack() based on
Date: Sun, 04 Mar 2001 12:15:36 -0800
From: Farooq Mela [EMAIL PROTECTED]
It's not actually gcc-specific code, its more of a hint to the compiler
to not warn about something like, main() "falling off the end" when
there is actually an exit(0); at the end of it. If GCC knows exit()
never
Date: Sun, 04 Mar 2001 19:39:09 -0600
From: David A. Gobeille [EMAIL PROTECTED]
It would also be interesting to see the numbers for an Athlon/PIII
system with DDR, if anyone has such a machine.
Personally, I'd be [more] interested in a ServerWorks III HE core chipset
with four-way
Date: Thu, 8 Mar 2001 14:36:22 -0500 (EST)
From: [EMAIL PROTECTED]
Actually, the FA310TX comes in a couple of different forms. One with
the Tulip chipset and another with PNIC, I believe. It seems that they
changed chipsets partway into the life of the product code for the card.
IIRC,
Bear with me and allow me my delusions while I daydream...
What with FPGA technology as reasonable as it is, and the amount of hw/sw
talent on these lists, maybe people should band together and come up with
a NIC? Maybe have native mode + Tulip/PNIC clone compatibility mode.
Take a look at
Date: Thu, 5 Apr 2001 01:40:38 -0700
From: [EMAIL PROTECTED]
20) Multi-staged booting. You don't need to change your MBR when you
install a new kernel (or want to roll back to a different kernel). I
think I've only been stuck high and dry w/o a bootable system twice in
four years and
Greetings,
(If I'm OT, please feel free to tell me where to go... as in which list,
not that other place. ;-)
I have a couple of AHC cards. When I cold boot, I receive MPARERR
before "Waiting XXX seconds for SCSI bus to settle". Panic.
I've tried two different AHC boards and two different
Greetings all,
I'm no kernel hacker, and trying to think of useful little projects to
change that. ;-)
AFAIK, FreeBSD lacks support for TCP intercept. Is anyone already working
on this? Would it be of interest to anyone? My initial thoughts are that
it should be implemented in the same
Greetings all,
I just had a brainstorm...
I was thinking about database servers with several spindles in a RAID 5
array. Write performance is inherently disappointing -- which may or may
not be an issue.
Would it be worth the trouble to design an intermediate cache, whereby
data are quickly
Date: Mon, 28 May 2001 19:49:40 +0200
From: Christoph Sold [EMAIL PROTECTED]
My gut feel is that this would be more trouble than it's worth, would
not net any overall performance*reliability (expressed as a
product) gain, and that one might actually realize a p*r decrease.
IMHO it would
Date: Mon, 28 May 2001 17:54:24 +0100
From: Dominic Marks [EMAIL PROTECTED]
[ snip ]
disc caching. The idea of perhaps caching writes onto a RAID-0 system
I meant caching onto an arbitrary volume, probably using a simple
journalling filesystem. Personally, a RAID 1 volume would be my
Date: Mon, 28 May 2001 19:01:37 -0500 (CDT)
From: David Scheidt [EMAIL PROTECTED]
[ snip ]
If you're really interested in database performance, remember Spindles
is good. Spreading your IO load over as many seperate disks, on as
many independent IO channels as practical will improve
Date: Sat, 16 Jun 2001 11:56:45 -0700 (PDT)
From: Matt Dillon [EMAIL PROTECTED]
This is old. The guys running the tests blew it in so many ways
that you might as well have just rolled some dice. There's a slashdot
article on it too, and quite a few of the reader comments on these
bozos
Date: Sat, 16 Jun 2001 16:57:04 -0400 (EDT)
From: Albert D. Cahalan [EMAIL PROTECTED]
You mean they should just optimize for FreeBSD, or should they also
use completion ports on Win2K, /dev/poll on Solaris, and RT signals
on Linux? What is wrong with using the portable API on every OS?
If
On Sat, 16 Jun 2001, Albert D. Cahalan wrote:
I'd say the real winner was NT. It mostly kept up with Linux,
trashed FreeBSD and Solaris, and didn't need any tuning to do it.
FWIW, somebody pointed out (and I overlooked) that the test ran RSETs
instead of real mail messages. Excuse me, but
Quick question, hopefully not too basic for this list:
AIO vs. non-blocking IO vs. kernel queues
I'm familiar with (and *love*) kernel queues. Non-blocking IO is
straightforward. AIO seems simple enough.
My question is, from a performance standpoint, in what situations are
these techniques
Please pardon the cross-posting; I'd rather keep responses on whichever
list is more appropriate.
Why are bind(2), accept(2), kevent(2), etc. wrapped in libc_r?
I thought that the spl() calls prevented kernel recursion in the current
SMP system, and that a mutex handled reentrance in SMPng.
Date: Thu, 28 Jun 2001 21:28:56 -0500
From: Chris Costello [EMAIL PROTECTED]
Please pardon the cross-posting; I'd rather keep responses on whichever
list is more appropriate.
Currently, the pthreads implementation is done entirely in
userland. This means that a syscall which would
(on -hackers only, as this post is beyond the -smp charter)
Am I correct that libc_r does _not_ use multiple processes to create
threads? Grepping for fork in *.c files under /usr/src/lib/libc_r
leads me to believe that this is so...
That's correct. It's implemented using
(Warning: rather long message)
Date: Fri, 29 Jun 2001 00:50:30 -0700
From: Terry Lambert [EMAIL PROTECTED]
[ ... wrapped fd using functions in libc_r ... ]
[ fd locking, to prevent chopping feet from beneath ]
As-needed serialization to prevent breakage = proper behavior.
I should have
Date: Fri, 29 Jun 2001 18:33:52 +0300
From: Peter Pentchev [EMAIL PROTECTED]
The threads scheduler is in user space. It converts a
blobking call into a non-blocking call plus a context
switch. THus blocking _IS_ a problem.
Bad wording on my part again; perhaps a problem that I
(Cross-posting again... I'm willing to be larted with a herring if this is
unacceptable for the content presented.)
I was thinking about CPU affinity on SMP systems the following is
on-list brainstorming.
Take a two-way box running 10 httpd and 10 smtpd processes. Assuming
equal CPU time
Date: Fri, 29 Jun 2001 20:33:51 +0200
From: Bernd Walter [EMAIL PROTECTED]
I can't see how you make shure that on SMP systems all CPUs have
the same meaning from memory content.
Normaly you would use a mutex or similar before accessing a data range
from another thread which also enshures
(Personal CCs trimmed; back to Bernd and cross-posting -smp and -hackers)
Date: Fri, 29 Jun 2001 21:18:18 +0200
From: Bernd Walter [EMAIL PROTECTED]
Passing a token between threads. When a thread has the token, it may
assert a lock or a mutex on an object. Again, I subscribe to threads
Date: Fri, 29 Jun 2001 13:14:58 -0700
From: Matthew Rogers [EMAIL PROTECTED]
Why not just use First in line, Next processor available ? Then you
wouldn't care what processor did which task.
That was my question: Would the added complexity of CPU affinity
hinting be worth the reduction in
Date: Fri, 29 Jun 2001 21:44:43 -0500
From: Michael C . Wu [EMAIL PROTECTED]
The issue is a lot more complicated than what you think.
How so? I know that idleproc and the new ipending / threaded INTs
enter the picture... and, after seeing the HLT benchmark page, it
would appear that simply
Quick question(s):
1. Is AIO SMP-safe?
2. If not, how could one force coherency? (Read and rewrite locked
a word from each cache line?) Is it worth the effort, or should
one not use AIO across process boundaries?
I'm asking primarily about 4.x, unless anyone has good guesses of
how 5.x
Date: Sat, 30 Jun 2001 22:28:29 -0500
From: Alfred Perlstein [EMAIL PROTECTED]
Can you point to some specific PRs about this or crashdumps before
(or at least while) taking pot shots at the AIO implementation?
In the mean time, until somebody can substantiate that claim... is AIO SMP
safe?
Date: Sun, 1 Jul 2001 16:07:38 +0200
From: Bernd Walter [EMAIL PROTECTED]
In -currents NOTEs I found this:
# CPU_DISABLE_5X86_LSSER disables load store serialize (i.e. enables
# reorder). This option should not be used if you use memory mapped
# I/O device(s).
A good sign that it may
(Responding on-list so there's no flood of private responses. Considered
cross-posting to move the thread, but hoping it will just die on
-hackers.)
This topic would probably be better suited to freebsd-questions.
Date: Mon, 2 Jul 2001 09:00:15 -0700 (PDT)
From: Zac M. Speidel [EMAIL
Date: Tue, 03 Jul 2001 03:17:12 -0700
From: Terry Lambert [EMAIL PROTECTED]
who_has_lock++ ;
who_has_lock %= process_count ;
Your unsimplified assembly is not happy, and neither is
this. You want to use a LOCK CMPXCHG to implement your
mutexes; the LOCK prefix makes
Date: Thu, 5 Jul 2001 14:42:41 +0100
From: Leubner, Achim [EMAIL PROTECTED]
One of our testers has a boot problem with FreeBSD 4.3. He installed
the OS on a onboard SCSI drive. After installing the FreeBSD driver
of our ICP controllers and inserting a ICP card with one disk
attached, the
Date: Sat, 7 Jul 2001 15:51:51 -0700 (PDT)
From: Richard Hodges [EMAIL PROTECTED]
Sure, no argument there. Taking Wes' suggestion, maybe there is an
opportunity in the official distribution distinction. How about a
certificate of authenticity which costs the vendors $1 or $2 (or
I
Date: Sat, 07 Jul 2001 20:27:16 -0700
From: Jordan Hubbard [EMAIL PROTECTED]
such. It's often the case that actually tracking the effectiveness
of, say, $500 towards some theoretically worthy cause costs you
thousands of dollars in man-hours. Heck, If somebody tried to donate
a million
Date: Sun, 8 Jul 2001 12:29:40 EDT
From: [EMAIL PROTECTED]
If theres one thing that non-unix people have learned about unix
is that the cost of the product is insignificant compared to
the value of time in getting it to function. Its already free to
anyone who knows what they are doing. A
Date: Mon, 09 Jul 2001 12:09:08 -0700
From: Terry Lambert [EMAIL PROTECTED]
The most obvious way to do this, IMO, is to make FreeBSD
more accessiblle for use in commercial ventures, be it
embedded systems or as a backdrop for commercial software
using it as a platform on which some company
Date: Thu, 12 Jul 2001 21:09:44 -0400
From: Leo Bicknell [EMAIL PROTECTED]
http://www.iet.unipi.it/~luigi/sack.html
H. I don't yet know enough about kernel architecture to know
in advance how I'd fare trying to patch that into 4.x (I expect
the line number to be off, obviously), but I
Date: Fri, 13 Jul 2001 13:29:03 -0400
From: Leo Bicknell [EMAIL PROTECTED]
(The window autotuning was an interesting read...)
I think you're doing good work, but I'm concerned you're going
down a road that's going to take a very long time to get right.
It is not necessary to calculate the
Greetings all,
Any way to force the source address for an outbound SOCK_STREAM? I
know that one can do it for SOCK_DGRAM... but I've found no way to
do so for, say, a TCP connection.
Example:
+ dc0 has 192.168.0.1/24 as primary IP, 192.168.0.2/24 as alias
+ an outbound connection wishes to
Date: Fri, 5 Oct 2001 13:58:22 -0500
From: Alfred Perlstein [EMAIL PROTECTED]
[ snip ]
Where was the warning (so it can be removed)?
/usr/src/sys/i386/conf/LINT
Not sure about Alpha.
Eddy
---
Brotsman Dreger, Inc.
Greetings,
I'm currently STFWing, but would appreciate any pointers that
anyone might have. I wish to manually perform dynamic linking.
I'd like to load an executable, .so, or .o, and _manually_ handle
the symbol fixups. I looked at dlfcn.c, but found next to
nothing there. Next stop: kernel
Greetings all,
I wish to accomplish the following:
1. Program foo loads shared object bar using dlopen() and
dlsym()
2. bar needs certain symbols resolved, which foo intercepts.
For example, foo might wrap malloc() or open() to provide its
own behavior... much like subclassing window
Date: Tue, 2 Apr 2002 11:05:52 -0800
From: Alfred Perlstein [EMAIL PROTECTED]
From the dlopen manpage:
I reread that right before initial post.
If dlsym() is called with the special handle RTLD_NEXT, then the search
for the symbol is limited to the shared objects which were
Date: Tue, 2 Apr 2002 11:05:52 -0800
From: Alfred Perlstein [EMAIL PROTECTED]
From the dlopen manpage:
[ snip ]
Works great, less coding.
Looks like I just misunderstood the manpage and/or the workings
of the dynamic linker.
Time for me to have some fun. And, no, I'm not using wrappers
TL Date: Thu, 09 May 2002 10:34:33 -0700
TL From: Terry Lambert
TL Rice University has been doing that a lot with their
TL licenses, these days. I don't know if they would have done
TL this anyway, or if it's an overreaction on their part to
TL FreeBSD ignoring their patches.
Eclipse/BSD is a
Greetings all,
While writing CGIs in C, I'm getting a bit sick of escaping
quotes and line continuations in strings. Not a huge deal,
perhaps, but there must be a better way. Strings end up in .data
or .rodata in object files to be linked...
I'm about to whip up a utility that will take any
JM Date: Sun, 19 May 2002 21:41:05 -0700
JM From: Jonathan Mini
JM Take a look at file2c. You'll need to run the source through
JM the compiler first, but that is easy to do with make.
H. Definitely produces the desired results for the simple
case that I mentioned. In fact, more complex
52 matches
Mail list logo