Date: Tue, 21 Nov 2017 13:54:28 -0500 From: Ryan Stone <rysto32@gmail.com> To: Kristof Provost <kristof@sigsegv.be> Cc: Catalin Salgau <csalgau@users.sourceforge.net>, freebsd-net <freebsd-net@freebsd.org> Subject: Re: BPF packet pagesize limit Message-ID: <CAFMmRNzw0SFBZKH5YNgnR2d7K0LyF=vD7dv9p5f_xquZ4BW69A@mail.gmail.com> In-Reply-To: <65E7BF88-EBEA-47DA-806B-5BFD6783F2B4@sigsegv.be> References: <966f384c-10b4-d018-efb1-68a7064c9521@users.sourceforge.net> <A2A39E3C-8A17-4C17-A52D-0EF72F809F99@sigsegv.be> <e52b9a29-ccc3-af47-9fdb-4e856bda4e49@users.sourceforge.net> <65E7BF88-EBEA-47DA-806B-5BFD6783F2B4@sigsegv.be>
next in thread | previous in thread | raw e-mail | index | archive | help
Do not allocate MJUM9BYTES clusters under any circumstances. Trying to allocate them when the system is under memory pressure is enormously expensive and stands a good chance of livelocking the system if you try to allocate too many too quickly, even when allocating with M_NOWAIT. Honestly, support for clusters > PAGE_SIZE should probably be retired from the kernel given the significant havoc they cause.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFMmRNzw0SFBZKH5YNgnR2d7K0LyF=vD7dv9p5f_xquZ4BW69A>