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

Re: Frequent "unaligned access", an init that dies, ...

On Wed, 22 Oct 1997, Bruce Perens wrote:

> Re-seat those SIMMS. It sounds like another unreliable RAM problem.

To fill in those just joining us, Lars was getting errors like this on his
DEC Alpha Linux system:

Oct 21 19:06:39 zeus kernel: kernel: unaligned trap at fffffc0000363c14: 
fffffc000443d036 28 2

According to my roommate, who runs a Red Hat Linux on his Alpha, this is a
bug at the program level.  Specifically, many programs try to do 8- or
16-bit memory accesses.  These often cause misaligned memory accesses on
the Alpha's 64-bit architecture.

The hardware generates an exception, which the kernel catches, and while it
spits out the warning, it does multiple memory accesses to get all the
required data and shuffles the registers around to produce what the program

This slows things down, but is not fatal.  Basically it's the product of
programmers not being prepared for a 64-bit system when building their data
structures, and possibly gcc not being intelligent enough in reorganizing
them for optimization.

In this specific case, the unaligned access is generated by the kernel
itself, though many other programs can and do cause them.

Oct 21 19:06:39 zeus kernel:
                ^^^^^^^^^^^ your kernel delivering the message

^^^^^^ the program that generated the exception (PID and process name if not

unaligned trap at fffffc0000363c14: 
                  ^^^^^^^^^^^^^^^^ program counter (?)

^^^^^^^^^^^^^^^^ address of attempted fetch (?)

28 2
^^^^ instruction(?)/system call in question

This breakdown is care of roommate, whose recollection may be inaccurate;
it would probably be wise to check with gurus if more exacting (and
confident) information is required.

G. Branden Robinson                 |
Purdue University                   |  If encryption is outlawed, only outlaws
branden@purdue.edu                  |  will @goH7OjBd7*dnfk=<q4fDj]Kz?.
http://www.ecn.purdue.edu/~branden/ |

TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . Trouble? 
e-mail to templin@bucknell.edu .

Reply to: