Skip to contentSkip to conversation

interrupt latency k2.4 / i386?

DISCUSSION SA 3 participants 3 messages Feb 7, 2003

lore
Share & tools
lore

SA

author
#0 lore raw

Dear list,

What latency should I expect for hardware interrupts under k2.4 / i386 ?

ie how long should it take between the hardware signalling the interrupt and the interrupt handler being called?

I am wrting a driver which pace IO with interrupts, generating one interrupt for after every transfer is done. Looking at the hardware schematics the interrupts should occur virtually instantly after each transfer but the driver is waiting ~1ms/ interrupt.

I can use polling instead with busy waits but this seems a bit wasteful.

My interrupt is shared with my graphics card using the non-GPL nvidia driver - could this be responsible for the delay (any experience with this)?

cat /proc/interrupts
.....

 10:       3028          XT-PIC  eth0, VIA 82C686A
 11:    1117037          XT-PIC  nvidia, PI stage <-- my driver
 12:      14776          XT-PIC  usb-uhci, usb-uhci

.....

Thanks SA

#1 lore raw
Quoted 22 lines · SA

Dear list,

What latency should I expect for hardware interrupts under k2.4 / i386 ?

ie how long should it take between the hardware signalling the interrupt and the interrupt handler being called?

I am wrting a driver which pace IO with interrupts, generating one interrupt for after every transfer is done. Looking at the hardware schematics the interrupts should occur virtually instantly after each transfer but the driver is waiting ~1ms/ interrupt.

I can use polling instead with busy waits but this seems a bit wasteful.

My interrupt is shared with my graphics card using the non-GPL nvidia driver - could this be responsible for the delay (any experience with this)?

cat /proc/interrupts
.....

Quoted 7 lines
 10:       3028          XT-PIC  eth0, VIA 82C686A
 11:    1117037          XT-PIC  nvidia, PI stage <-- my driver
 12:      14776          XT-PIC  usb-uhci, usb-uhci

.....

Thanks SA
-

Well there are very great problems with a lot of drivers if you want good latency, but are sharing interrupts. Many drivers poll in the ISR! This means that they loop through their ISR code in some "interrupt mitigation scheme" that some near-do-well taught in web-page school. This means, that if you share an interrupt, you are probably screwed. You need to put your board in some slot where it will either be first on the IRQ chain, or not shared at all. Typical latencies are around 1.2 microseconds on a 400 MHz machine, better on faster, sometimes better with two CPUs and an IOAPIC.

Cheers,
Dick Johnson
Penguin : Linux version 2.4.18 on an i686 machine (797.90 BogoMips). Why is the government concerned about the lunatic fringe? Think about it.

#2 lore raw
In reply to SA
Quoted 9 lines · SA

Dear list,

What latency should I expect for hardware interrupts under k2.4 / i386 ?

ie how long should it take between the hardware signalling the interrupt and the interrupt handler being called?

I don't know the hardware part but it depends on how long interrupts are being disabled (and that depend on the code you use in the kernel) and that can be measured with akpm's timepeg+intlat patches

http://www.zip.com.au/~akpm/linux/#timepegs

I hope this helps

Ulisses

Quoted 26 lines

I am wrting a driver which pace IO with interrupts, generating one interrupt for after every transfer is done. Looking at the hardware schematics the interrupts should occur virtually instantly after each transfer but the driver is waiting ~1ms/ interrupt.

I can use polling instead with busy waits but this seems a bit wasteful.

My interrupt is shared with my graphics card using the non-GPL nvidia driver - could this be responsible for the delay (any experience with this)?

cat /proc/interrupts
.....

 10:       3028          XT-PIC  eth0, VIA 82C686A
 11:    1117037          XT-PIC  nvidia, PI stage <-- my driver
 12:      14776          XT-PIC  usb-uhci, usb-uhci

.....

Thanks SA
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 6 lines
                Debian GNU/Linux: a dream come true
-----------------------------------------------------------------------------
"Computers are useless. They can only give answers."            Pablo Picasso

--->	Visita http://www.valux.org/ para saber acerca de la	<---
--->	Asociación Valenciana de Usuarios de Linux		<---