Date: Mon, 17 May 2010 11:28:47 -0700 From: "Matthew Fleming" <matthew.fleming@isilon.com> To: "Kostik Belousov" <kostikbel@gmail.com> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, Zachary Loafman <zml@freebsd.org>, src-committers@freebsd.org Subject: RE: svn commit: r208003 - in head/sys: kern sys Message-ID: <06D5F9F6F655AD4C92E28B662F7F853E021D4D6C@seaxch09.desktop.isilon.com> References: <201005122124.o4CLOk3b027904@svn.freebsd.org> <20100516050651.GZ83316@deviant.kiev.zoral.com.ua> <06D5F9F6F655AD4C92E28B662F7F853E021D4D66@seaxch09.desktop.isilon.com> <20100517162806.GM83316@deviant.kiev.zoral.com.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
From: Kostik Belousov [mailto:kostikbel@gmail.com] > Essentially, my argument is that whatever you do in VOP_ADVLOCKPURGE, > can be and should be done in VOP_RECLAIM. This would also cover the = v_data > issue. I disagree about the "should". I believe that if BSD wants to allow = overriding lf_advlock(9) with VOP_ADVLOCK it should fully support the = overriding. Alternatively, the call to lf_purgelocks(9) should be = pushed down into each filesystem's VOP_RECLAIM for consistency. Thanks, matthew
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?06D5F9F6F655AD4C92E28B662F7F853E021D4D6C>