netdev - Re: [PATCH v3 19/20] RDMA: Add irdma Kconfig/Makefile and remove i40iw

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20200102180654.GB9282@ziepe.ca>
Date:   Thu, 2 Jan 2020 14:06:54 -0400
From:   Jason Gunthorpe <jgg@...pe.ca>
To:     "Saleem, Shiraz" <shiraz.saleem@...el.com>
Cc:     "Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>,
        "davem@...emloft.net" <davem@...emloft.net>,
        "gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "linux-rdma@...r.kernel.org" <linux-rdma@...r.kernel.org>,
        "nhorman@...hat.com" <nhorman@...hat.com>,
        "sassmann@...hat.com" <sassmann@...hat.com>,
        "parav@...lanox.com" <parav@...lanox.com>
Subject: Re: [PATCH v3 19/20] RDMA: Add irdma Kconfig/Makefile and remove
 i40iw

On Thu, Jan 02, 2020 at 05:50:45PM +0000, Saleem, Shiraz wrote:
> > Subject: Re: [PATCH v3 19/20] RDMA: Add irdma Kconfig/Makefile and remove
> > i40iw
> > 
> > On Thu, Jan 02, 2020 at 04:00:37PM +0000, Saleem, Shiraz wrote:
> > > > Subject: Re: [PATCH v3 19/20] RDMA: Add irdma Kconfig/Makefile and
> > > > remove i40iw
> > > >
> > > >
> > > > > >  - The whole cqp_compl_thread thing looks really weird
> > > > > What is the concern?
> > > >
> > > > It looks like an open coded work queue
> > > >
> > >
> > > The cqp_compl_thread is not a work queue in the sense that no work is
> > > queued to it. It is a thread that is signaled to check for and handle
> > > CQP completion events if present.
> > 
> > How is that not a work queue? The work is the signal to handle CQP completion
> > events.
> > 
> 
> Yes we could use the work as a signal. But this would mean,
> we allocate a work item, initialize it to an 'identical' value,
> queue it up and then free it.

You don't have to allocate a work item every time.

> Why is this better than using a single kthread that just
> wake ups to handle the CQP completion?

We'd have endless kthreads if people did this everytime they needed a
bit of async work.

Jason