On Mon, 7 Nov 2005, John Baldwin wrote:
> And even then it can't be used for any device interrupts since there aren't
> any I/O APICs. On a UP machine without I/O APICs, it's actually probably
> more optimal to just use irq0 and irq8 for clocks rather than the lapic timer
> anyway. The only
On Sun, 6 Nov 2005, John Baldwin wrote:
> We don't detect the local APIC via MSR's or the APIC bit in cpu_features, but
> rely on a working MP Table or MADT table to setup both the local APIC(s) and
> I/O APIC(s). Does your machine have a valid MP Table or an APIC table in its
> acpidump? Ma
hi,
FreeBSD 6.0 always uses local APIC for the clock.
But on my Pentium 3, 850 MHz machine, it doesnt detect local APIC and
falls back to using the motherboard clock for the clock interrupts.
I figured this out by printing the value of
"using_lapic_timer" variable in the sys/i386/isa/clock.c fi
there a way, I can call my function instantly from the
lapic_handle_timer, bcoz using the software interrupt thread, decreases
the accuracy of the scheduler i am using.
thanks
vaibhave
On Fri, 28 Oct 2005, Robert Watson wrote:
> On Thu, 27 Oct 2005, Vaibhave Agarwal wrote:
>
> &
In the file /usr/src/UPDATING, there is a following
statement
NOTE TO PEOPLE WHO THINK THAT FreeBSD 6.x IS SLOW:
FreeBSD 6.x has many debugging features turned on, in
both the kernel and userland. These features attempt to detect
incorrect use of system primitives, and enc
go to /usr/src/sys/netinet
For tcp hack:
tcp_input.c and tcp_output.c
for ip hack:
ip_input.c and ip_output.c
If you read the book:
The Implementation (TCP/IP Illustrated, Volume 2)
by Gary R. Wright, W. Richard Stevens
You will understand how each and every line of code works.
-vaibhave
hi,
Just a little background - I want to fire events at
10 microsecond granularity in FreeBSD, and it was suggested to use
Hz=10 in the kernel config file.
This is a good approach, only if I schedule events every 10 us ( which i
am not doing ). Due to this high frequency interrupts, I am i
> If you have a system newer than a 486SX, then npx interrupts shouldn't
> be used for anything except to probe that not using them works.
Can i disable the FPU, by commenting it out in
the kernel config file??
> It
> is barely possible that a bug in turning off npx interrupts after the
> probe
hi
I was writing some code at the ethernet layer,
mainly changing the if_ethersubr.c for my work.
I am not using any FP instructions in my code.
But while running, my kernel crashes and gives the following panic msg:
npx is the Floating point unit (fpu).
--
npxin
hi
I was in little confusion about how the network software interrupts
actually work.
Is there a priority level within the network software interrupts too??
If there are two or more network level interrupts ( same priority level
splnet) are pending, then does the scheduler schedules them in the
hi all
I am trying to configure Intel 82559 10/100 Mbps Ethernet card to generate
"Early Interrupts", so that I can receive interrupts from the driver, long
before the whole packet is transferred to the host memory.
I have set the Early Receive Interrupt field in the CSR register of the
card. But
hi
In ieee 802.3 CSMA/CD, is it possible that i could check a register value
set by hardware card or by some other way i could detect in the driver, as
soon as the card senses that there is a packet on the wire ( i.e. even
before the reception of the whole packet).
Anybody's help would be highly a
12 matches
Mail list logo
