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;
Date:   Fri, 15 Jun 2018 12:46:09 -0400
From:   Steven Rostedt <rostedt@goodmis.org>
To:     Mikulas Patocka <mpatocka@redhat.com>
Cc:     Alan Stern <stern@rowland.harvard.edu>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ming Lei <ming.lei@redhat.com>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        USB list <linux-usb@vger.kernel.org>,
        Kernel development list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] usb: don't offload isochronous urb completions to
 ksoftirq
Message-ID: <20180615124609.7726924b@gandalf.local.home>
In-Reply-To: <alpine.LRH.2.02.1806151222200.5648@file01.intranet.prod.int.rdu2.redhat.com>
References: <Pine.LNX.4.44L0.1806131442230.1476-100000@iolanthe.rowland.org>
        <alpine.LRH.2.02.1806131521540.26196@file01.intranet.prod.int.rdu2.redhat.com>
        <20180613183145.17564d82@gandalf.local.home>
        <alpine.LRH.2.02.1806141809500.6435@file01.intranet.prod.int.rdu2.redhat.com>
        <20180614183546.7e04024c@gandalf.local.home>
        <alpine.LRH.2.02.1806151222200.5648@file01.intranet.prod.int.rdu2.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, 15 Jun 2018 12:41:10 -0400 (EDT)
Mikulas Patocka <mpatocka@redhat.com> wrote:

> BTW. when I subject the machine to a ping flood (ping -f), the 
> non-realtime kernel (with the patch to avoid offloading ehci urb 
> callbacks) performs better than the real-time kernel.
> 
> With the real-time kernel, all the networking work is done in the thread 
> "irq/12-eth0", that has (by default) priority -51, it consumes 30% CPU 
> time and causes sound skipping. I can avoid the skipping by lowering the 
> priority of "irq/12-eth0".

That's actually the correct thing to do. When dealing with RT
applications, one needs to modify the priority of the interrupts that
are required (or lower the ones that are not).

> 
> With non-realtime kernel, no such problem during ping flood exists.

But that was with a patched kernel?

-- Steve