[QUESTION] The cost of local interrupt enable/disable

Thread View: gmane.linux.kernel

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by Tejun Heo Sat, 17 Sep 2005 23:00

#307687

Author: Tejun Heo
Date: Sat, 17 Sep 2005 23:00

12 lines
363 bytes

  Hello, guys.

  I'm curious about the cost of local_irq_enable/disable()'s on various
architectures.  I found a freebsd discussion thread by googling which
says that each takes 3 cycles on i386 (very cheap), but for Pentium4,
people are talking in the order of several hundreds cycles.  Are these
correct?  How about other architectures?

  TIA.

--
tejun

Re: [QUESTION] The cost of local interrupt enable/disable

#307688

Author: Anton Blanchard
Date: Sun, 18 Sep 2005 00:32

12 lines
455 bytes

Hi,

>  I'm curious about the cost of local_irq_enable/disable()'s on various
> architectures.  I found a freebsd discussion thread by googling which
> says that each takes 3 cycles on i386 (very cheap), but for Pentium4,
> people are talking in the order of several hundreds cycles.  Are these
> correct?  How about other architectures?

It varies on ppc64, but its in the order of 10s of cycles. 40-50 cycles
is probably a decent estimation.

Anton

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads