Date: Mon, 12 Sep 2011 11:53:00 +0800 From: Adrian Chadd <adrian@freebsd.org> To: Andreas Longwitz <longwitz@incore.de> Cc: freebsd-stable@freebsd.org, John Baldwin <jhb@freebsd.org> Subject: Re: UFS_DIRHASH panics on a dozen server within 30 hours Message-ID: <CAJ-Vmon4XBQ4vvJOVGSiOdgX1W78x8EUmcVqtMgFPJgnP-FfFw@mail.gmail.com> In-Reply-To: <4E6D34FE.70703@incore.de> References: <4E64933E.8030908@incore.de> <201109061104.43409.jhb@freebsd.org> <4E6D34FE.70703@incore.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On 12 September 2011 06:23, Andreas Longwitz <longwitz@incore.de> wrote: > > Yes. I found that megarc often wants a buffer of 12868 bytes, but the > controller sends always 25412 bytes back. Because this seems to be an > error in megarc I have submitted a patch for the existing PR ports/137938. > > Furthermore I saw some sporadic answers of the controller to megarc > ioctl's with much more data than the buffer size stated by megarc. > Therefore I still use the maximum size in my updated patch in kern/155658. Just a comment about that kernel PR: You're doing a 25k allocation (why not just go with an even 32k?) but you're not enforcing anywhere that if au_length > your #defined value, the malloc size is made larger. I know it's a nitpick, but if it were me doing a change, I'd use MAX(au_length, MAX_AMR_IOCTL) when doing the malloc. 2c, Adrian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmon4XBQ4vvJOVGSiOdgX1W78x8EUmcVqtMgFPJgnP-FfFw>