Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Interrupt routing patch for testing

Hash: SHA1

On 8 May 2008, at 23:12, Joerg Sonnenberger wrote:

Hi all,
attached is a critical patch that potentially effects all x86 users
(beside Xen domU). It changes the way interrupt slots are allocated.

The old code could allocate two different interrupt vectors for the same
hardware interrupt when IOAPIC is used and the BIOS doesn't initialise
all devices. This typically happened for cbb0 on newer laptops.

With the patch, the code is simplified and I can actually start on
making the round robin scheduling work :-)

Please test the patch on as much systems as you can. If you run into
problems, e.g. system hangs, please provide me (offlist) with a dmesg of
a kernel build with -DINTRDEBUG and MPVERBOSE option w/ and w/o the

Hi there,

I just patched my current i386 running on an Athlon X2 and it works not problems so far, just what is curious is that the individual interrupt counters in vmstat -i and systat vm disapeart,
but that happened before the patch.

- --

Viele Grüße,
Lars Heidieker

- ------------------------------------

Mystische Erklärungen:
Die mystischen Erklärungen gelten für tief;
die Wahrheit ist, dass sie noch nicht einmal oberflächlich sind.

  -- Friedrich Nietzsche
  [ Die Fröhliche Wissenschaft Buch 3, 126 ]

Version: GnuPG v1.4.7 (Darwin)


Home | Main Index | Thread Index | Old Index