Date: Mon, 11 Jan 2016 12:43:48 -0800 From: Colin Percival <cperciva@freebsd.org> To: =?UTF-8?Q?Roger_Pau_Monn=c3=a9?= <royger@FreeBSD.org>, freebsd-xen@freebsd.org Cc: hps@freebsd.org, kib@freebsd.org Subject: Re: recent disk-related breakage Message-ID: <56941404.90805@freebsd.org> In-Reply-To: <5693E739.5030103@FreeBSD.org> References: <56934B8F.8050503@freebsd.org> <5693501F.9060008@freebsd.org> <5693E739.5030103@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 01/11/16 09:32, Roger Pau Monné wrote: > El 11/01/16 a les 7.47, Colin Percival ha escrit: >> I'm rather suspicious of r292255 here -- royger, hps, kib, can any of you >> comment on whether it would be responsible for making non-page-aligned I/Os >> no longer get split on page boundaries? The commit message is suggestive, >> but I don't know the code very well. > > This is caused by r291716 Agreed. Sorry for pointing at the wrong busdma change. > which re-introduced the broken mechanism > found in bus_dmamap_load_ma_triv. I've already notified ken about it, > and I have a patch that should solve it by reverting to the usage of > _bus_dmamap_load_ma in order to load an unmapped bio: > > https://lists.freebsd.org/pipermail/svn-src-all/2016-January/116629.html Ken replied shortly after the Europe -> North America "shift change", so I took the liberty of committing this fix (after doing further testing myself) as r293698. Thank you! -- Colin Percival Security Officer Emeritus, FreeBSD | The power to serve Founder, Tarsnap | www.tarsnap.com | Online backups for the truly paranoid
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?56941404.90805>