Re: arca-57-against-2.1.131-ac11.diff.gz

Stephen Donnelly (sfd@cs.waikato.ac.nz)
Wed, 16 Dec 1998 11:08:44 +1300


Andrea Arcangeli wrote:
>
> I put out a new arca-57 patch. It would be nice if people with mm
> unbalancing problems or people that is used to print using a parallel port
> would try it.
>

> o jiffies wrap update
>

I tried the arca-59 patch, does it contain this too? It didn't fix my
problem, which may be related, as follows.

-----------------------------------------------------------------------
It seems that the kernel time is not monotonically increasing in recent
(i386) kernels.

2.1.126 introduced a bug that is still present in 2.1.131 and also the
AC11 and arca59 patches. Code that records the time during the top half
of an interrupt may get a timestamp that is exactly 10000 microseconds
back in
time from when it should be.

An easy way to see this is to use tcpdump on a reasonably busy network
(500 to 1000 packets per second), and examine the usec portion of the
packet timestamp. This is affected because the timestamp is recorded
during interrupt time (netif_rx in linux/net/core/dev.c).

See this URL for more details and graphs:

http://atm.cs.waikato.ac.nz/~sfd/bug/

Stephen.

-- 
-----------------------------------------------------------------------
    Stephen Donnelly (BCMS)             email: sfd@cs.waikato.ac.nz
    Waikato ATM Group          Lab G1.31 phone +64 7 838 4466 x6728
    Computer Science Department, University of Waikato, New Zealand
-----------------------------------------------------------------------

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/