linux-kernel - Re: [PATCH v2 3/3] augmented rbtree: rework the RB_DECLARE

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 2 Jul 2019 13:52:25 +0200
From:   Peter Zijlstra <peterz@...radead.org>
To:     Michel Lespinasse <walken@...gle.com>
Cc:     Davidlohr Bueso <dave@...olabs.net>,
        David Howells <dhowells@...hat.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/3] augmented rbtree: rework the RB_DECLARE_CALLBACKS
 macro definition

On Tue, Jul 02, 2019 at 12:58:19AM -0700, Michel Lespinasse wrote:
> - Change the definition of the RBCOMPUTE function. The propagate
>   callback repeatedly calls RBCOMPUTE as it moves from leaf to root.
>   it wants to stop recomputing once the augmented subtree information
>   doesn't change. This was previously checked using the == operator,
>   but that only works when the augmented subtree information is a
>   scalar field. This commit modifies the RBCOMPUTE function so that
>   it now sets the augmented subtree information instead of returning it,
>   and returns a boolean value indicating if the propagate callback
>   should stop.

I suppose that makes sense and saves a copy over adding RBEQUAL() like I
proposed earlier.

> - Reorder the RB_DECLARE_CALLBACKS macro arguments, following the
>   style of the INTERVAL_TREE_DEFINE macro, so that RBSTATIC and RBNAME
>   are passed last.

That's, IMO, a weird change. C has storage type and name first, why
would you want to put that last. If anything, change
INTERVAL_TREE_DEFINE().

Also; this is two changes and one patch; ISTR we have rules about those
things :-)

> The motivation for this change is that I want to introduce augmented rbtree
> uses where the augmented data for the subtree is a struct instead of a scalar.

I'm not seeing how this justifies the second thing.