Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190424101934.51535-1-duyuyang@gmail.com> <20190424101934.51535-20-duyuyang@gmail.com>
 <20190425193247.GU12232@hirez.programming.kicks-ass.net> <CAHttsrY4jK2cayBE8zNCSJKDAkzLiBb40GVfQHpJi2YK1nEZaQ@mail.gmail.com>
 <20190430121148.GV2623@hirez.programming.kicks-ass.net> <20190507014712.GA14921@lerouge>
In-Reply-To: <20190507014712.GA14921@lerouge>
From:   Yuyang Du <duyuyang@gmail.com>
Date:   Tue, 7 May 2019 10:21:58 +0800
Message-ID: <CAHttsra_jACOSZpwnp0KtuAOFWtXt5AHf+RVWMvEbxWbieVw0w@mail.gmail.com>
Subject: Re: [PATCH 19/28] locking/lockdep: Optimize irq usage check when
 marking lock usage bit
To:     Frederic Weisbecker <frederic@kernel.org>
Cc:     Peter Zijlstra <peterz@infradead.org>, will.deacon@arm.com,
        Ingo Molnar <mingo@kernel.org>,
        Bart Van Assche <bvanassche@acm.org>, ming.lei@redhat.com,
        tglx@linutronix.de, LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, 7 May 2019 at 09:47, Frederic Weisbecker <frederic@kernel.org> wrote:
> > > But for usage checking, which vectors are does not really matter. So,
> > > the current size of the arrays and bitmaps are good enough. Right?
> >
> > Frederic? My understanding was that he really was going to split the
> > whole thing. The moment you allow masking individual soft vectors, you
> > get per-vector dependency chains.
>
> Right, so in my patchset there is indeed individual soft vectors masked
> so we indeed need per vector checks. For example a lock taken in HRTIMER
> softirq shouldn't be a problem if it is concurrently taken while BLOCK softirq
> is enabled. And for that we expand the usage_mask so that the 4 bits currently
> used for general SOFTIRQ are now multiplied by NR_SOFTIRQ (10) because we need to
> track the USED and ENABLED_IN bits for each of them.
>
> The end result is:
>
> 4 hard irq bits + 4 * 10 softirq bits + LOCK_USED bit = 45 bits.
>
> Not sure that answers the question as I'm a bit lost in the debate...

It was really I was lost: I didn't realize the enabling (or disabling)
is going to be fine-grained as well until I read this changelog:

Disabling the softirqs is currently an all-or-nothing operation: either
all softirqs are enabled or none of them. However we plan to introduce a
per vector granularity of this ability to improve latency response and
make each softirq vector interruptible by the others.

Sorry for the confusion I made :)