Date: Fri, 21 May 2004 17:45:14 +0100 From: Doug Rabson <dfr@nlsystems.com> To: Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp> Cc: freebsd-firewire@freebsd.org Subject: Re: async bug Message-ID: <1085157914.25509.10.camel@builder02.qubesoft.com> In-Reply-To: <87vfiphktk.wl@tora.nunu.org> References: <Pine.GSO.4.58.0405181432030.18744@mono.arc.nasa.gov> <87vfiphktk.wl@tora.nunu.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 2004-05-21 at 16:16, Hidetoshi Shimokawa wrote: > At Tue, 18 May 2004 14:33:57 -0700 (PDT), > Buzz Slye wrote: > > > > > > The code for the version of 2004/03/28 does not appear to work correctly > > for an asyncronus request of a single register. > > > > (i.e., try: fwcontrol -b 0) > > > > R. E. Slye > > NASA/Ames > > The ioctl(2) interface of /dev/fw0.0 is not well-tested and may > have many bugs. If you need only quad/block read/write request > (without LOCK transaction), /dev/fwmem interface should be > easier to use and well-tested. The souce code of dconschat(8) > is an example of using /dev/fwmem. I think that it might be simpler to break out the generic FW_ASYREQ ioctl into separate ones which do various different async requests. One problem with it now is that the ioctl argument is large because of the data array. The whole thing must be copied in and out even for single quad reads/writes.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1085157914.25509.10.camel>