Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 31 Jul 2023 19:35:15 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Thomas Gleixner <tglx@linutronix.de>
Cc:     axboe@kernel.dk, linux-kernel@vger.kernel.org, mingo@redhat.com,
        dvhart@infradead.org, dave@stgolabs.net, andrealmeid@igalia.com,
        Andrew Morton <akpm@linux-foundation.org>, urezki@gmail.com,
        hch@infradead.org, lstoakes@gmail.com,
        Arnd Bergmann <arnd@arndb.de>, linux-api@vger.kernel.org,
        linux-mm@kvack.org, linux-arch@vger.kernel.org,
        malteskarupke@web.de
Subject: Re: [PATCH v1 02/14] futex: Extend the FUTEX2 flags
Message-ID: <20230731173515.GP29590@hirez.programming.kicks-ass.net>
References: <20230721102237.268073801@infradead.org>
 <20230721105743.819362688@infradead.org>
 <87edkonjrk.ffs@tglx>
 <87v8e0m26q.ffs@tglx>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87v8e0m26q.ffs@tglx>
Precedence: bulk

On Mon, Jul 31, 2023 at 07:16:29PM +0200, Thomas Gleixner wrote:
> On Mon, Jul 31 2023 at 18:11, Thomas Gleixner wrote:
> > On Fri, Jul 21 2023 at 12:22, Peter Zijlstra wrote:
> >> -#define FUTEX2_MASK (FUTEX2_32 | FUTEX2_PRIVATE)
> >> +#define FUTEX2_MASK (FUTEX2_64 | FUTEX2_PRIVATE)
> >>  
> >>  /**
> >>   * futex_parse_waitv - Parse a waitv array from userspace
> >> @@ -207,7 +207,12 @@ static int futex_parse_waitv(struct fute
> >>  		if ((aux.flags & ~FUTEX2_MASK) || aux.__reserved)
> >>  			return -EINVAL;
> >
> > With the above aux.flags with FUTEX2_32 set will result in -EINVAL. I
> > don't think that's intentional.
> 
> Also if you allow 64bit wide futexes, how is that supposed to work with
> the existing code, which clearly expects a 32bit uval throughout the
> place?

Not allowed yet, these patches only allow 8,16,32. I still need to audit
the whole futex core and do 'u32 -> unsigned long' (and everything else
that follows from that), and only when that's done can the futex2
syscalls allow FUTEX2_64 on 64bit archs.

So for now, these patches:

  - add the FUTEX2_64 flag,
  - add 'unsigned long' interface such that
    64bit can potentiall use it,
  - explicitly disallow having it set.