Date: Sat, 30 Oct 2004 22:44:36 -0700 From: "Justin T. Gibbs" <gibbs@scsiguy.com> To: Bruce M Simpson <bms@spc.org> Cc: cvs-src@FreeBSD.org Subject: Re: cvs commit: src/sys/pci if_vr.c Message-ID: <C6D45A30E728D66460FBED53@caspian.scsiguy.com> In-Reply-To: <20041031044021.GA2123@empiric.icir.org> References: <20041031035037.59607.qmail@web21422.mail.yahoo.com> <E024D6F9B48B27EB9B2BF6CE@caspian.scsiguy.com> <20041031044021.GA2123@empiric.icir.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> On Sat, Oct 30, 2004 at 10:35:58PM -0700, Justin T. Gibbs wrote: >> > Shouldn't an API for adding or removing ISR's be responsible for >> > handling just such a race condition? >> >> bus_teardown_intr() already has these semantics, but you have >> to know to use it and that it should only be called without >> locks held since it may sleep waiting for another CPU to >> leave your interrupt handler. > > Unfortunately I don't have a vr(4) card to test the changes with, and > it looks like the copy of if_vr.c in my local tree uses bus_teardown_intr() > correctly; these changes will have to be tested and committed after 5.3R. > > BMS You should be able to get rid of the suspended check in the isr too. -- Justin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?C6D45A30E728D66460FBED53>