Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 22 Apr 2021 13:48:39 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     David Malcolm <dmalcolm@redhat.com>
Cc:     Ard Biesheuvel <ardb@kernel.org>, linux-toolchains@vger.kernel.org,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Josh Poimboeuf <jpoimboe@redhat.com>,
        Jason Baron <jbaron@akamai.com>,
        "Steven Rostedt (VMware)" <rostedt@goodmis.org>,
        yuanzhaoxiong@baidu.com
Subject: Re: static_branch/jump_label vs branch merging
Message-ID: <YIFilwwzWiSDZU6b@hirez.programming.kicks-ass.net>
References: <5f78b7e2f9ae937271ef52ee9e999a91c2719da9.camel@redhat.com>
 <YHBCoijoopbsDn29@hirez.programming.kicks-ass.net>
 <YHBQPr8q0cx4iUfN@hirez.programming.kicks-ass.net>
 <YHBS70ZQ6gOpMk2K@hirez.programming.kicks-ass.net>
 <3c062f70ffef2dcd48a661f7c8162fb8fbaf6869.camel@redhat.com>
 <YHCfgHwDtT7m4ffq@hirez.programming.kicks-ass.net>
 <0a9da587b0330bafdf612c3d51285e144b0b9e46.camel@redhat.com>
 <YHC0dgwhYS9hKcRT@hirez.programming.kicks-ass.net>
 <5a07bde1a9fa9a056a19637399b0635252ddb303.camel@redhat.com>
 <20210409213949.GA33256@worktop.programming.kicks-ass.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210409213949.GA33256@worktop.programming.kicks-ass.net>
Precedence: bulk

On Fri, Apr 09, 2021 at 11:39:49PM +0200, Peter Zijlstra wrote:
> On Fri, Apr 09, 2021 at 05:07:15PM -0400, David Malcolm wrote:
> 
> > You've built a very specific thing out of asm-goto to fulfil the tough
> > requirements you outlined above - as well as the nops, there's a thing
> > in another section to contend with.
> > 
> > How to merge these asm-goto constructs?
> 
> By calling the function less, you emit less of them. Which then brings
> us back to the whole pure/const thing.
> 
> > Doing so feels very special-case to the kernel and not something that
> > other GCC users would find useful.
> 
> Doesn't it boil down to 'fixing' the pure/const vs asm-goto interaction?
> I could imagine that having that interaction fixed could allow other
> creative uses.

Here is another variant:

  https://lore.kernel.org/lkml/830177B0-45E0-4768-80AB-A99B85D3A52F@baidu.com/

Can we please have a __pure__ attribute that is prescriptive and not a
hint the compiler is free to ignore for $raisins ?