linux-kernel - Re: [RFC PATCH] usb: hcd: complete URBs in threaded-IRQ context instead of tasklet

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:   Fri, 16 Feb 2018 16:46:41 -0500 (EST)
From:   Alan Stern <stern@...land.harvard.edu>
To:     Sebastian Andrzej Siewior <bigeasy@...utronix.de>
cc:     Mauro Carvalho Chehab <mchehab@...pensource.com>,
        Frederic Weisbecker <frederic@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        <linux-usb@...r.kernel.org>
Subject: Re: [RFC PATCH] usb: hcd: complete URBs in threaded-IRQ context
 instead of tasklet

On Fri, 16 Feb 2018, Sebastian Andrzej Siewior wrote:

> On 2018-02-16 13:29:01 [-0500], Alan Stern wrote:
> > We originally used tasklets because we didn't want to incur the delays 
> > associated with running in a process context.  It seems odd to be 
> > reversing that decision now.
> 
> The theaded interrupt runs SCHED_FIFO priority 50 by default. The only
> thing that can interrupt it are interrupts, a softirq (not ksoftirqd)
> and other tasks with a higher priority than 50.
> There should be no downside performance wise.

Maybe.  It would be nice to see some real measurements.

> > > The URBs from the root-hub never create an interrupt so I currently
> > > process them in a workqueue (I'm not sure if an URB-enqueue in the
> > > completion handler would break something).
> > 
> > It worked okay before we changed over to using tasklets.
> 
> Ah okay. I've seen that HCDs were no longer dropping their internal lock
> and I wasn't sure if such a change was also applied in RH-code.

Oh... my memory was faulty.  After going back and looking at the old
source, I see that before we switched to tasklets the code _did_ drop
the hcd_root_hub_lock when giving back root-hub URBs.  Now it doesn't.

(It's also worth noticing that the current code gives back root-hub 
URBs in a tasklet even for HCDs that don't support using tasklets for 
non-root-hub URBs.)

In any case, latency doesn't matter much for root-hub URBs.

Alan Stern