Date: Mon, 26 Aug 2019 14:30:30 +0000 From: bugzilla-noreply@freebsd.org To: bugs@FreeBSD.org Subject: [Bug 240061] MADV_FREE rewinds time to before fork() Message-ID: <bug-240061-227-v4YCy5HFe7@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-240061-227@https.bugs.freebsd.org/bugzilla/> References: <bug-240061-227@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D240061 Mark Johnston <markj@FreeBSD.org> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |alc@FreeBSD.org, | |kib@FreeBSD.org, | |markj@FreeBSD.org --- Comment #1 from Mark Johnston <markj@FreeBSD.org> --- Out of curiosity, how did you come to notice this behaviour? I think the solution you proposed would solve the problem. We would want to ensure that the restrictions specified in vm_object_advice_applies() are checked before modifying the object chain. In particular, MADV_FREE only m= arks pages clean at the MI layer if the object is anonymous and is mapped into no more than one address space. For OBJT_VNODE mappings, MADV_FREE does not c= lear the MI dirty bits or requeue the page, but it will clear the dirty bit from PTEs in the specified range. --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-240061-227-v4YCy5HFe7>