linux-hardening - Re: [RFC 1/1] drivers/dma/*: replace tasklets with workqueue

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] [day] [month] [year] [list]

Message-ID: <CAK8P3a314SxhTNy6psYr_MzarOPmkjv1_KKGdeVUPpBjVkC5nA@mail.gmail.com>
Date:   Tue, 31 May 2022 21:27:35 +0200
From:   Arnd Bergmann <arnd@...db.de>
To:     Allen Pais <apais@...ux.microsoft.com>
Cc:     Arnd Bergmann <arnd@...db.de>, Vinod Koul <vkoul@...nel.org>,
        Linus Walleij <linus.walleij@...aro.org>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        olivier.dautricourt@...lia.com, Stefan Roese <sr@...x.de>,
        Kees Cook <keescook@...omium.org>,
        linux-hardening@...r.kernel.org,
        Ludovic Desroches <ludovic.desroches@...rochip.com>,
        Tudor Ambarus <tudor.ambarus@...rochip.com>,
        Florian Fainelli <f.fainelli@...il.com>,
        Ray Jui <rjui@...adcom.com>,
        Scott Branden <sbranden@...adcom.com>,
        bcm-kernel-feedback-list <bcm-kernel-feedback-list@...adcom.com>,
        Nicolas Saenz Julienne <nsaenz@...nel.org>,
        Paul Cercueil <paul@...pouillou.net>,
        Eugeniy.Paltsev@...opsys.com,
        Gustavo Pimentel <gustavo.pimentel@...opsys.com>,
        Viresh Kumar <vireshk@...nel.org>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Leo Li <leoyang.li@....com>, zw@...kernel.org,
        Zhou Wang <wangzhou1@...ilicon.com>,
        Shawn Guo <shawnguo@...nel.org>,
        Sascha Hauer <s.hauer@...gutronix.de>,
        Sean Wang <sean.wang@...iatek.com>,
        Matthias Brugger <matthias.bgg@...il.com>,
        Andreas Färber <afaerber@...e.de>,
        Manivannan Sadhasivam <mani@...nel.org>,
        Logan Gunthorpe <logang@...tatee.com>,
        Sanjay R Mehta <sanju.mehta@....com>,
        Daniel Mack <daniel@...que.org>,
        Haojian Zhuang <haojian.zhuang@...il.com>,
        Robert Jarzmik <robert.jarzmik@...e.fr>,
        Andy Gross <agross@...nel.org>,
        Bjorn Andersson <bjorn.andersson@...aro.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>,
        green.wan@...ive.com, Orson Zhai <orsonzhai@...il.com>,
        Baolin Wang <baolin.wang7@...il.com>,
        Lyra Zhang <zhang.lyra@...il.com>,
        Patrice CHOTARD <patrice.chotard@...s.st.com>,
        Chen-Yu Tsai <wens@...e.org>,
        Jernej Škrabec <jernej.skrabec@...il.com>,
        Samuel Holland <samuel@...lland.org>,
        dmaengine@...r.kernel.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [RFC 1/1] drivers/dma/*: replace tasklets with workqueue

On Tue, May 31, 2022 at 8:19 PM Allen Pais <apais@...ux.microsoft.com> wrote:
> >>> That is a good idea, lot of drivers are waiting for completion which can
> >>> be signalled from hardirq, this would also reduce the hops we have and
> >>> help improve latency a bit. On the downside, some controllers provide
> >>> error information, which would need to be dealt with.
> >>
> >>
> >>   I am not an expert in dma subsystem, but by using completion from
> >> Hardirq context be a concern? Especially with latency.
> >
> > I don't see how: to the task waiting for the completion, there should
> > be no difference, and for the irq handler sending it, it just avoids
> > a few cycles going into softirq context.
>
>   Thanks for clarification.
>
>   If I have understood it correctly, your suggestion is to move the current
> Callback mechanism out to dmaengine as a generic helper function
> And introduce completion in dma_async_tx_descriptor to handle what
> Tasklets currently do.

Right: around half the callbacks are a trivial version that does nothing
other than calling complete(), so these will benefit from skipping the
softirq, while the others still need the callback.

       Arnd