Re: [Lse-tech] Re: RFC: patch to allow lock-free traversal of lists with insertion

Date: Wed Oct 10 2001 - 02:21:05 EST

Dipankar Sarma wrote:

>In article <> BALBIR SINGH wrote:
>>What about cases like the pci device list or any other such list. Sometimes
>>you do not care if somebody added something, while you were looking through
>>the list as long as you do not get illegal addresses or data.
>>Wouldn't this be very useful there? Most of these lists come up
>>at system startup and change rearly, but we look through them often.
>How often does the linux kernel need to go through the PCI device
>list ? Looking at only SCSI code, it seems that not very often.
>If that is the case in general, optimizing it for lockless
>lookup (assuming that you use RCU or something to support deletion),
>is probably an overkill.
Almost all pci drivers use the pci_find_slot or some functionality that
requires a scan of all the pci devices (to identify their device). I agree
that it is not very often.

>One example of where it is useful is maintenance of route information
>in either storage or network. Route information changes infrequently but
>needs to be looked up for every I/O and being able to do lockless
>lookup here is a good gain.
I have a question, can this kind of locking used in cases where an interrupt
context may be involved. For example looking through the list of timers, we
disable interrupts and grab a lock using spin_lock_irqsave(&timerlist_lock, flags)

Should we just use __cli() with the RCU or something similar? or can RCU
be used in such cases?



To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

This archive was generated by hypermail 2b29 : Mon Oct 15 2001 - 21:00:30 EST