Date: Wed, 8 Jan 2020 12:15:40 +0100 From: Hans Petter Selasky <hps@selasky.org> To: Shichun.Ma@dell.com, freebsd-usb@freebsd.org Cc: Shunchao.Hu@dell.com Subject: Re: can not receive xfer interrupt after stop xfer is called intel XHCI Gemini Lake SOC Message-ID: <d0cb04b2-0dbe-748c-ead5-5a74f9ffe09d@selasky.org> In-Reply-To: <df8e2391-58a1-969f-d2ec-7101df6764aa@selasky.org> References: <1577408331523.24347@Dell.com> <acc5ccc7-d76c-d9ff-f9d5-f63ac40227d6@selasky.org> <1577411424906.21267@Dell.com> <db23ff3a-df66-e060-4409-5eccb214d3cf@selasky.org> <caff3f2cab964507b12df3841d443000@KULX13MDC130.APAC.DELL.COM> <df8e2391-58a1-969f-d2ec-7101df6764aa@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 2019-12-31 11:53, Hans Petter Selasky wrote: > On 2019-12-31 01:23, Shichun.Ma@dell.com wrote: >> Dell Customer Communication - Confidential >> >> Hi HPS, >> >> It's already in stopped status. I am also confusing on the root cause. >> I attached the test application and my patch for the xhci stop xfer >> workaround solution. >> The confusing points: >> 1. cancel xfer error can be reproduced on all CCID smart readers (I >> have tested three different model of readers); >> 2. keyboard has similar endpoint attribution, while I can't reproduce >> similar problem on the keyboard; >> > > Hi, > > I suspect it is the newer XHCI hardware which has some additional > checks. Can you verify if the XHCI controller in your computer accept > multiple configure_ep() commands? The state diagram in the XHCI > specification does not say you cannot do this, but I imagine this might > be causing it. > > The XHCI driver in FreeBSD configure one and one endpoint and not all at > the same time. > > Can you try the attached patch instead of yours? > Ping? --HPS
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?d0cb04b2-0dbe-748c-ead5-5a74f9ffe09d>