Re: stap and staprun use cases

[fche at redhat dot com (Frank Ch. Eigler)]

hunt wrote:

> [...]  Doesn't matter much to me as long as we make the whole
> temporary directory hack invisible to the user.

One can make less drastic changes and still accomplish that.  

For example, stuff built by "-m module" could still be copied into the
cache directories.  (Some match validation logic -- comparing the
generated .c file to the stored .c one -- would need to be added, but
that's a good idea in any case.)  In general, "-m NAME" should not
imply anything other than overriding the hash generation of the module
name.

There is no need for "-C" (compile) unless the equivalent "-p4" is
deemed too wordy.

I don't have a strong opinion about the proposed "-i" (install), but
will note that stap's documented behavior to print the module file name
after "-p4" means that anyone can install a module with just a shell
script lines
          cp `stap -p4 ....` SOMEWHERE
or even   env SYSTEMTAP_DIR=SOMEWHERE stap -p4 ....
The possible automation of "SOMEWHERE" gives the "-i" idea some value.

Is the /lib/modules/`uname -r`/ path fairly universal amongst linux
distributions?


- FChE