Date: Fri, 16 Dec 2022 13:07:20 -0800 From: Rick Macklem <rick.macklem@gmail.com> To: Tomoaki AOKI <junchoon@dec.sakura.ne.jp> Cc: freebsd-current@freebsd.org Subject: Re: What to do about a few lines in vfs_domount() never executed? Message-ID: <CAM5tNy7RdfSMBOxRtkYrtPKfxsN8TZK0Cz4bCwVO_RSGOUQnHg@mail.gmail.com> In-Reply-To: <20221215021431.d190e55ee911f5e94799f953@dec.sakura.ne.jp> References: <CAM5tNy7DM__tj4CQ0L_-ugo5krmEvLJgbU-WFLs47MrRBuGaNQ@mail.gmail.com> <20221215021431.d190e55ee911f5e94799f953@dec.sakura.ne.jp>
index | next in thread | previous in thread | raw e-mail
[-- Attachment #1 --] Just to provide closure on this, I just committed a patch to main (that will be MFC'd in 2 weeks) that sets MNT_EXPORTED when the "export" option is specified to nmount(2). This restores the "only root can do exports" semantics that appear to have been the case pre-r158857. Thanks everyone for your comments, rick On Wed, Dec 14, 2022 at 9:15 AM Tomoaki AOKI <junchoon@dec.sakura.ne.jp> wrote: > Tracking the commits, it was originally introduced to > sys/kern/vfs_syscalls.c at r22521 [1][2] (Mon Feb 10 02:22:35 1997 by > dyson, submitted by hsu@freebsd.org) and later centralized into > sys/kern/vfs_mount.c at r99264 [2]. > > But according to the comment above the codes, maybe it would be > intended to block userland programs or ports FS modules setting > MNT_EXPORTED. > > If I'm not mis-understanding, it can be the case when > *vfs.usermount sysctl is non-zero, > *underlying FS (to be exported) allows it, and > *non-root user tries to mount the FS via NFS. > > > [1] > > https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?revision=22521&view=markup&pathrev=99264 > > [2] > > https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?r1=22520&r2=22521&pathrev=99264&; > > [3] > > https://cgit.freebsd.org/src/commit/sys/kern/vfs_mount.c?id=2b4edb69f1ef62fc38b02ac22b0a3ac09e43fa77 > > > On Tue, 13 Dec 2022 14:19:39 -0800 > Rick Macklem <rick.macklem@gmail.com> wrote: > > > Hi, > > > > While working on getting mountd/nfsd to run in a vnet > > prison, I came across the following lines near the > > beginning of vfs_domount() in sys/kern/vfs_mount.c: > > > > if (fsflags & MNT_EXPORTED) { > > error = priv_check(td, PRIV_VFS_MOUNT_EXPORTED); > > if (error) > > return (error); > > } > > > > #1 - Since MNT_EXPORTED is never set in fsflags, this code never > > gets executed. > > --> I am asking what to do with the above code, since that > > changes for the patch that allows mountd to run in a vnet > > prison. > > #2 - priv_check(td, PRIV_VFS_MOUNT_EXPORTED) always returns 0 > > because nothing in sys/kern/kern_priv.c checks > > PRIV_VFS_MOUNT_EXPORTED. > > > > I don't know what the original author's thinking was w.r.t. this. > > Setting exports already checks that the mount operation can be > > done by the requestor. > > > > So, what do you think should be done with the above code snippet? > > - Consider it cruft and delete it. > > - Try and figure out what PRIV_VFS_MOUNT_EXPORTED should check? > > - Leave it as is. After the patch that allows mountd to run in > > a vnet prison, MNT_EXPORTED will be set in fsflags, but the > > priv_check() call will just return 0. (A little overhead, > > but otherwise no semantics change.) > > > > rick > > > -- > Tomoaki AOKI <junchoon@dec.sakura.ne.jp> > > [-- Attachment #2 --] <div dir="ltr"><div class="gmail_default" style="font-family:monospace">Just to provide closure on this, I just committed a patch to</div><div class="gmail_default" style="font-family:monospace">main (that will be MFC'd in 2 weeks) that sets MNT_EXPORTED</div><div class="gmail_default" style="font-family:monospace">when the "export" option is specified to nmount(2).</div><div class="gmail_default" style="font-family:monospace"><br></div><div class="gmail_default" style="font-family:monospace">This restores the "only root can do exports" semantics that</div><div class="gmail_default" style="font-family:monospace">appear to have been the case pre-r158857.</div><div class="gmail_default" style="font-family:monospace"><br></div><div class="gmail_default" style="font-family:monospace">Thanks everyone for your comments, rick</div><div class="gmail_default" style="font-family:monospace"><br></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Dec 14, 2022 at 9:15 AM Tomoaki AOKI <<a href="mailto:junchoon@dec.sakura.ne.jp">junchoon@dec.sakura.ne.jp</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Tracking the commits, it was originally introduced to<br> sys/kern/vfs_syscalls.c at r22521 [1][2] (Mon Feb 10 02:22:35 1997 by<br> dyson, submitted by <a href="mailto:hsu@freebsd.org" target="_blank">hsu@freebsd.org</a>) and later centralized into<br> sys/kern/vfs_mount.c at r99264 [2].<br> <br> But according to the comment above the codes, maybe it would be<br> intended to block userland programs or ports FS modules setting <br> MNT_EXPORTED.<br> <br> If I'm not mis-understanding, it can be the case when<br> *vfs.usermount sysctl is non-zero,<br> *underlying FS (to be exported) allows it, and<br> *non-root user tries to mount the FS via NFS.<br> <br> <br> [1]<br> <a href="https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?revision=22521&view=markup&pathrev=99264" rel="noreferrer" target="_blank">https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?revision=22521&view=markup&pathrev=99264</a><br>; <br> [2]<br> <a href="https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?r1=22520&r2=22521&pathrev=99264&" rel="noreferrer" target="_blank">https://svnweb.freebsd.org/base/head/sys/kern/vfs_syscalls.c?r1=22520&r2=22521&pathrev=99264&</a><br>; <br> [3]<br> <a href="https://cgit.freebsd.org/src/commit/sys/kern/vfs_mount.c?id=2b4edb69f1ef62fc38b02ac22b0a3ac09e43fa77" rel="noreferrer" target="_blank">https://cgit.freebsd.org/src/commit/sys/kern/vfs_mount.c?id=2b4edb69f1ef62fc38b02ac22b0a3ac09e43fa77</a><br>; <br> <br> On Tue, 13 Dec 2022 14:19:39 -0800<br> Rick Macklem <<a href="mailto:rick.macklem@gmail.com" target="_blank">rick.macklem@gmail.com</a>> wrote:<br> <br> > Hi,<br> > <br> > While working on getting mountd/nfsd to run in a vnet<br> > prison, I came across the following lines near the<br> > beginning of vfs_domount() in sys/kern/vfs_mount.c:<br> > <br> > if (fsflags & MNT_EXPORTED) {<br> > error = priv_check(td, PRIV_VFS_MOUNT_EXPORTED);<br> > if (error)<br> > return (error);<br> > }<br> > <br> > #1 - Since MNT_EXPORTED is never set in fsflags, this code never<br> > gets executed.<br> > --> I am asking what to do with the above code, since that<br> > changes for the patch that allows mountd to run in a vnet<br> > prison.<br> > #2 - priv_check(td, PRIV_VFS_MOUNT_EXPORTED) always returns 0<br> > because nothing in sys/kern/kern_priv.c checks<br> > PRIV_VFS_MOUNT_EXPORTED.<br> > <br> > I don't know what the original author's thinking was w.r.t. this.<br> > Setting exports already checks that the mount operation can be<br> > done by the requestor.<br> > <br> > So, what do you think should be done with the above code snippet?<br> > - Consider it cruft and delete it.<br> > - Try and figure out what PRIV_VFS_MOUNT_EXPORTED should check?<br> > - Leave it as is. After the patch that allows mountd to run in<br> > a vnet prison, MNT_EXPORTED will be set in fsflags, but the<br> > priv_check() call will just return 0. (A little overhead,<br> > but otherwise no semantics change.)<br> > <br> > rick<br> <br> <br> -- <br> Tomoaki AOKI <<a href="mailto:junchoon@dec.sakura.ne.jp" target="_blank">junchoon@dec.sakura.ne.jp</a>><br> <br> </blockquote></div>help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAM5tNy7RdfSMBOxRtkYrtPKfxsN8TZK0Cz4bCwVO_RSGOUQnHg>