Home / Linux / Another Episode of “Seems Perfectly Feasible and Then Dies”–Script to Simplify the Process of Changing System Call Tables
Another Episode of “Seems Perfectly Feasible and Then Dies”–Script to Simplify the Process of Changing System Call Tables
Another Episode of “Seems Perfectly Feasible and Then Dies”–Script to Simplify the Process of Changing System Call Tables

Another Episode of “Seems Perfectly Feasible and Then Dies”–Script to Simplify the Process of Changing System Call Tables

Another Episode of “Seems Perfectly Feasible and Then Dies”–Script to Simplify the Process of Changing System Call Tables

David Howells installed somewhat a little bit of paintings on a script,
./scripts/syscall-manage.pl, to simplify the complete procedure of converting the
machine name tables. With this script, it was once a easy subject to upload, take away,
rename or renumber any machine name you loved. The script additionally would get to the bottom of
git
conflicts, in the match that two repositories renumbered the machine calls in
conflicting techniques.

Why did David want to write this patch? Why were not machine calls already slightly
simple to handle? When you’re making a machine name, you upload it to a grasp record, and then
you upload it to the machine name “tables”, which is the place the working kernel appears to be like up
which kernel serve as corresponds to which machine name quantity. Kernel builders
want to be certain that machine calls are represented in all related spots in the supply
tree. Renaming, renumbering and making different adjustments to machine calls comes to a
lot of fiddly little main points. David’s script merely would do the whole lot
proper—finish
of tale no problemo hasta los angeles vista.

Arnd Bergmann remarked, “Ah, amusing. You had already threatened to upload that script in
the previous. The implementation of direction appears to be like effective, I used to be simply hoping lets
as a substitute do away with the want for it first.” But, bowing to necessity, Arnd introduced
some technical ideas for enhancements to the patch.

However, Linus Torvalds swooped in at this actual second, announcing:

Ugh, I hate it.

I am positive the script is a wide variety of artful and helpful, however I in reality suppose the
answer isn’t this type of helper script, however merely that we must paintings at no longer
having every structure upload new machine calls in my view in the first position.

IOW, we must take a look at having only one unified desk for brand new machine name numbers,
and goal for the per-architecture ones to be for “legacy numbering”.

Maybe that may not occur, however in the _hope_ that it occurs, I in reality would like
that individuals no longer paintings at making scripts for the present nasty state of affairs.

And the portcullis got here crashing down.

It’s attention-grabbing that, as a substitute of accepting this quite glaring development to
the current state of affairs, Linus would moderately depart it damaged and unsightly, in order that
somebody at some point someplace could be motivated to do the harder-yet-better repair. And,
it is all the extra attention-grabbing given how excessive the present drawback is. Without
if truth be told being damaged, the state of affairs calls for builders to installed an incredible
quantity of care and effort into one thing that David’s script may make trivial and
simple. Even for such an clearly “good” patch, Linus offers idea to the coverage
and cultural implications, and the long term motivations of people operating in
that area of code.

Note: in case you are discussed above and need to submit a reaction above the remark phase, ship a message along with your reaction textual content to [email protected]

Check Also

How to Watch TCP and UDP Ports in Real-time

How to Watch TCP and UDP Ports in Real-time

How to Watch TCP and UDP Ports in Real-time In tool phrases, particularly on the …

Leave a Reply

Your email address will not be published. Required fields are marked *

Categories

Recent Posts

Categories