Re: [PATCH]Systemtap IA64 Runtime support

["Frank Ch. Eigler" <fche at redhat dot com>]

Hi -


> > (There seems to have been an oversight in attaching Red Hat copyright
> > notices to some of the runtime files.)
>
> [...]  The reason I copied the above was that some files had this
> GPL statement and some files didn't.  Also are you saying not keep
> even "Copyright (C) 2005 Intel Corporation"

Not at all - just that Red Hat should have already been there too.  If
you were so kind as to add it, that would be great.


> > > [...]
> > > +	// In IA64 platform function probe point is set at its
> > > +	// entry point rather than prologue end pointer, and in
> > > +	// IA64 prologue end point also is not computed in such way
> > 
> > I probably missed the discussion, but why exactly is this?  Are all
> > the parameters available at the debuginfo-identified location list
> > right at this entry address?
>
> On Ia64 platform, I found probe address of function entry translated by
> systemtap was one bundle more than the actual address. However, when
> queried from entrypc rather than prolouge-end of the function we got the
> right address translation. [...]

I guess it comes down testing it - whichever works should stay.  How
many of the pass-4 / pass-5 tests have you been able to run?  Also,
have you tested systemtap on a plain x86 after your changes, just to
make sure that the base port was not affected?

Do you already have physical CVS write privileges?  If so, you could
commit.


- FChE