Re: [PATCH] Linux Kernel Markers 0.11 for 2.6.17

From: Mathieu Desnoyers
Date: Mon Sep 25 2006 - 19:33:56 EST


Hi,

* Frank Ch. Eigler (fche@xxxxxxxxxx) wrote:
> Hi -
>
> > [...]
> > - It _does not_ change the compiler optimisations.
>
> Like any similar mechanism, it does force the compiler to change its
> code generation, so one can't claim this too strongly.
>
Yes, memory dependencies are changed, you are right. I was principally talking
about the inline and unrolled loops optimisations.


> > [...] Comments are welcome,
>
> I'm still uneasy about the use of varargs. The current code now uses
> the formatting string as metadata to be matched (strcmp) between
> producer and consumer. A general tool that would use them would have
> to start parsing general printf directives.

If you want to generate probes automatically, yes.

> I believe they are not
> quite general enough either e.g. to describe a raw binary blob.
>
If you want to dump a raw binary blob, what about :
MARK(mysubsys_myevent, "char %p %u", blobptr, blobsize);
where %p is a pointer to an array of char and %u the length ?

My idea is to use the string to identify what is referred by a pointer, so it
can be casted into this type with some kind of coherency between the marker and
the probe.

> I realize they serve a useful purpose in abbreviating what otherwise
> one might have to do (like that multiplicity of STAP_MARK_* type/arity
> permutations). But maybe there is a better way.
>

I think that duplicating the number of marker macros could easily make
them unflexible and ugly. This is why I am trying to come with this generic
macro.

> Also, while regparm(0) may provide some comfort on x86, is there good
> reason to believe that the same trick works (and will continue to
> work) on non-x86 platforms to invoke a non-varargs callee with a
> varargs caller?
>

Good point, I will setup a va_args in the probe. When correctly used, however,
there is no need to use the format string : we can directly get the variables
from the var arg list if we know in advance what the string will be.

Mathieu

OpenPGP public key: http://krystal.dyndns.org:8080/key/compudj.gpg
Key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68

Attachment: signature.asc
Description: Digital signature