Date: Thu, 4 Jan 2001 22:52:56 +0100 (CET) From: =?ISO-8859-1?Q?G=E9rard_Roudier?= <groudier@club-internet.fr> To: Andrew Gallatin <gallatin@cs.duke.edu> Cc: freebsd-alpha@FreeBSD.ORG Subject: Re: bus_dmamap_load Message-ID: <Pine.LNX.4.10.10101042232100.405-100000@linux.local> In-Reply-To: <14932.59983.360316.243974@grasshopper.cs.duke.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 4 Jan 2001, Andrew Gallatin wrote: > G=E9rard Roudier writes: > > Btw, you missed the `de' driver that also performs bus_dmamap_load() o= n > > mbuf's as a possible candidate. > >=20 >=20 > Does it actually do this? I see all the code protected by: >=20 > =09#if defined(TULIP_BUS_DMA) >=20 > But I do not see TULIP_BUS_DMA defined anywhered. I think it was > inherited from NetBSD and the define was clobbered in Indeed. I missed that. But the sym driver only calls bus_dmamap_load() in 2 places: 1) Using an hardcoded value for buflen (1 PAGE) for internal data structures allocation. 2) Using `csio->dxfer_len' when preparing an IO for the DMA. Place (1) is unlikely to lead to a size problem, and (2) is provided by CAM or upper layer. If there is a driver bug in driver code path (2), then it should likely have already triggerred, in my opinion. G=E9rard. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.LNX.4.10.10101042232100.405-100000>