From owner-freebsd-audit Mon Nov 25 1:36:32 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2BA5E37B401 for ; Mon, 25 Nov 2002 01:36:30 -0800 (PST) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102]) by mx1.FreeBSD.org (Postfix) with ESMTP id B6F3B43E4A for ; Mon, 25 Nov 2002 01:36:25 -0800 (PST) (envelope-from doginou@cognet.ci0.org) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102] (may be forged)) by cognet.ci0.org (8.12.6/8.12.6) with ESMTP id gAP9WTFn010557 for ; Mon, 25 Nov 2002 10:32:29 +0100 (CET) (envelope-from doginou@cognet.ci0.org) Received: (from doginou@localhost) by cognet.ci0.org (8.12.6/8.12.6/Submit) id gAP9WSHk010556 for freebsd-audit@FreeBSD.org; Mon, 25 Nov 2002 10:32:28 +0100 (CET) Date: Mon, 25 Nov 2002 10:32:28 +0100 From: Olivier Houchard To: freebsd-audit@FreeBSD.org Subject: do_dup patch Message-ID: <20021125093228.GA10213@ci0.org> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="Nq2Wo0NMKNjxTN9z" Content-Disposition: inline User-Agent: Mutt/1.5.1i Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG --Nq2Wo0NMKNjxTN9z Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Hi all, This patch makes the "new" and "old" do_dup arguments unsigned int instead of int. I can't see any reason they would have to be int, and right now a call to dup or dup2 with an invalid negative fd, such as dup(-999999); will panic a -CURRENT box (the problem doesn't exist on -STABLE because the file descriptor validity is checked in dup() and dup2()). Is there anything wrong with committing this ? Any comments are welcome. Olivier --Nq2Wo0NMKNjxTN9z Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="kern_descrip.c.diff" Index: kern_descrip.c =================================================================== RCS file: /home/ncvs/src/sys/kern/kern_descrip.c,v retrieving revision 1.168 diff -u -p -r1.168 kern_descrip.c --- kern_descrip.c 27 Oct 2002 18:07:41 -0000 1.168 +++ kern_descrip.c 24 Nov 2002 19:52:55 -0000 @@ -101,7 +101,7 @@ static struct cdevsw fildesc_cdevsw = { /* How to treat 'new' parameter when allocating a fd for do_dup(). */ enum dup_type { DUP_VARIABLE, DUP_FIXED }; -static int do_dup(struct thread *td, enum dup_type type, int old, int new, +static int do_dup(struct thread *td, enum dup_type type, u_int old, u_int new, register_t *retval); static int badfo_readwrite(struct file *fp, struct uio *uio, struct ucred *active_cred, int flags, struct thread *td); @@ -171,8 +171,7 @@ dup2(td, uap) struct dup2_args *uap; { - return (do_dup(td, DUP_FIXED, (int)uap->from, (int)uap->to, - td->td_retval)); + return (do_dup(td, DUP_FIXED, uap->from, uap->to, td->td_retval)); } /* @@ -193,7 +192,7 @@ dup(td, uap) struct dup_args *uap; { - return (do_dup(td, DUP_VARIABLE, (int)uap->fd, 0, td->td_retval)); + return (do_dup(td, DUP_VARIABLE, uap->fd, 0, td->td_retval)); } /* @@ -452,7 +451,7 @@ done2: static int do_dup(td, type, old, new, retval) enum dup_type type; - int old, new; + u_int old, new; register_t *retval; struct thread *td; { --Nq2Wo0NMKNjxTN9z-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 2:19: 2 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0E0A237B401 for ; Mon, 25 Nov 2002 02:19:00 -0800 (PST) Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16]) by mx1.FreeBSD.org (Postfix) with ESMTP id DDDE043E4A for ; Mon, 25 Nov 2002 02:18:58 -0800 (PST) (envelope-from bde@zeta.org.au) Received: from bde.zeta.org.au (bde.zeta.org.au [203.2.228.102]) by mailman.zeta.org.au (8.9.3/8.8.7) with ESMTP id VAA32197; Mon, 25 Nov 2002 21:18:41 +1100 Date: Mon, 25 Nov 2002 21:32:05 +1100 (EST) From: Bruce Evans X-X-Sender: bde@gamplex.bde.org To: Olivier Houchard Cc: freebsd-audit@FreeBSD.ORG Subject: Re: do_dup patch In-Reply-To: <20021125093228.GA10213@ci0.org> Message-ID: <20021125210116.I56453-100000@gamplex.bde.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Mon, 25 Nov 2002, Olivier Houchard wrote: > This patch makes the "new" and "old" do_dup arguments unsigned int instead of > int. I can't see any reason they would have to be int, and right now a call to > dup or dup2 with an invalid negative fd, such as dup(-999999); will panic a > -CURRENT box (the problem doesn't exist on -STABLE because the file descriptor > validity is checked in dup() and dup2()). > Is there anything wrong with committing this ? Yes; it is sort of backwards. File descriptors are small positive (signed) ints, and the dup() and dup2() syscalls take file descriptors as parameters. However, dup() and dup2() are misdeclared as taking u_int's as their parameters. This used to be obfuscate/optimize away an explicit comparison with 0 in the bounds check for the parameters. E.g., in RELENG_4: % #ifndef _SYS_SYSPROTO_H_ % struct dup2_args { % u_int from; % u_int to; ^^^^^ Bogus type for bogus optimization. This is pseudo-code; see for the actual declaration. This misdeclaration needs something like normal 2's complement machines to work since we never actually convert the parameters from ints to u_ints; we just pass ints and pretend that they are u_ints. % }; % #endif % /* ARGSUSED */ % int % dup2(p, uap) % struct proc *p; % struct dup2_args *uap; % { % register struct filedesc *fdp = p->p_fd; % register u_int old = uap->from, new = uap->to; % int i, error; % % retry: % if (old >= fdp->fd_nfiles || ^^^^^^^^^^^^^^^^^^^^^ Bogus optimization. The upper and lower bounds are checked in a single comparison. The only reason to do this seems to be to hand-optimize for 20+-year old compilers on pdp11-like machines where the 2 comparisons can be collapes into a single machine instruction and the compiler doesn't do this automatically. % fdp->fd_ofiles[old] == NULL || % new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || % new >= maxfilesperproc) { % return (EBADF); % } In -current, this has rotted to: % int % dup2(td, uap) % struct thread *td; % struct dup2_args *uap; % { % % return (do_dup(td, DUP_FIXED, (int)uap->from, (int)uap->to, ^^^^^ ^^^^^ Bogus casts to break warnings about our type errors. % td->td_retval)); % } % ... % static int % do_dup(td, type, old, new, retval) % enum dup_type type; % int old, new; % register_t *retval; % struct thread *td; % { % register struct filedesc *fdp; % struct proc *p; % struct file *fp; % struct file *delfp; % int error, newfd; % % p = td->td_proc; % fdp = p->p_fd; % % /* % * Verify we have a valid descriptor to dup from and possibly to % * dup to. % */ % FILEDESC_LOCK(fdp); % if (old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || ^^^^^^^^^^^^^^^^^^^^^ Bug. The bogus optimization has been broken by changing the types. % new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || % new >= maxfilesperproc) { % FILEDESC_UNLOCK(fdp); % return (EBADF); % } Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 3:12:52 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0525B37B401 for ; Mon, 25 Nov 2002 03:12:49 -0800 (PST) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102]) by mx1.FreeBSD.org (Postfix) with ESMTP id AE43943E4A for ; Mon, 25 Nov 2002 03:12:47 -0800 (PST) (envelope-from mlfbsd@cognet.ci0.org) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102] (may be forged)) by cognet.ci0.org (8.12.6/8.12.6) with ESMTP id gAPB93Fn011233; Mon, 25 Nov 2002 12:09:03 +0100 (CET) (envelope-from mlfbsd@cognet.ci0.org) Received: (from mlfbsd@localhost) by cognet.ci0.org (8.12.6/8.12.6/Submit) id gAPB92aM011232; Mon, 25 Nov 2002 12:09:02 +0100 (CET) Date: Mon, 25 Nov 2002 12:09:02 +0100 From: Olivier Houchard To: Bruce Evans Cc: freebsd-audit@FreeBSD.ORG Subject: Re: do_dup patch Message-ID: <20021125110902.GA10961@ci0.org> References: <20021125093228.GA10213@ci0.org> <20021125210116.I56453-100000@gamplex.bde.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20021125210116.I56453-100000@gamplex.bde.org> User-Agent: Mutt/1.5.1i Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Mon, Nov 25, 2002 at 09:32:05PM +1100, Bruce Evans wrote: > On Mon, 25 Nov 2002, Olivier Houchard wrote: > > > This patch makes the "new" and "old" do_dup arguments unsigned int instead of > > int. I can't see any reason they would have to be int, and right now a call to > > dup or dup2 with an invalid negative fd, such as dup(-999999); will panic a > > -CURRENT box (the problem doesn't exist on -STABLE because the file descriptor > > validity is checked in dup() and dup2()). > > Is there anything wrong with committing this ? > > Yes; it is sort of backwards. > > File descriptors are small positive (signed) ints, and the dup() and dup2() > syscalls take file descriptors as parameters. However, dup() and dup2() > are misdeclared as taking u_int's as their parameters. This used to be > obfuscate/optimize away an explicit comparison with 0 in the bounds check > for the parameters. E.g., in RELENG_4: > > % #ifndef _SYS_SYSPROTO_H_ > % struct dup2_args { > % u_int from; > % u_int to; > ^^^^^ > Bogus type for bogus optimization. This is pseudo-code; see > for the actual declaration. This misdeclaration > needs something like normal 2's complement machines to work since > we never actually convert the parameters from ints to u_ints; > we just pass ints and pretend that they are u_ints. > > % }; > % #endif > % /* ARGSUSED */ > % int > % dup2(p, uap) > % struct proc *p; > % struct dup2_args *uap; > % { > % register struct filedesc *fdp = p->p_fd; > % register u_int old = uap->from, new = uap->to; > % int i, error; > % > % retry: > % if (old >= fdp->fd_nfiles || > ^^^^^^^^^^^^^^^^^^^^^ > > Bogus optimization. The upper and lower bounds are checked in a > single comparison. The only reason to do this seems to be to > hand-optimize for 20+-year old compilers on pdp11-like machines > where the 2 comparisons can be collapes into a single machine > instruction and the compiler doesn't do this automatically. > > % fdp->fd_ofiles[old] == NULL || > % new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > % new >= maxfilesperproc) { > % return (EBADF); > % } > > In -current, this has rotted to: > > % int > % dup2(td, uap) > % struct thread *td; > % struct dup2_args *uap; > % { > % > % return (do_dup(td, DUP_FIXED, (int)uap->from, (int)uap->to, > ^^^^^ ^^^^^ > > Bogus casts to break warnings about our type errors. > > % td->td_retval)); > > % } > % ... > % static int > % do_dup(td, type, old, new, retval) > % enum dup_type type; > % int old, new; > % register_t *retval; > % struct thread *td; > % { > % register struct filedesc *fdp; > % struct proc *p; > % struct file *fp; > % struct file *delfp; > % int error, newfd; > % > % p = td->td_proc; > % fdp = p->p_fd; > % > % /* > % * Verify we have a valid descriptor to dup from and possibly to > % * dup to. > % */ > % FILEDESC_LOCK(fdp); > % if (old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || > ^^^^^^^^^^^^^^^^^^^^^ > > Bug. The bogus optimization has been broken by changing the types. > > % new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > % new >= maxfilesperproc) { > % FILEDESC_UNLOCK(fdp); > % return (EBADF); > % } > > Bruce Right. So what would be the best ? Changing dup and dup2 prototypes in syscall.master to make new and old signed instead of unsigned and checking that new >= 0 && old >= 0 in do_dup ? Thanks, Olivier To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 3:54:38 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 73B8B37B401 for ; Mon, 25 Nov 2002 03:54:37 -0800 (PST) Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16]) by mx1.FreeBSD.org (Postfix) with ESMTP id 69CE243E3B for ; Mon, 25 Nov 2002 03:54:36 -0800 (PST) (envelope-from bde@zeta.org.au) Received: from bde.zeta.org.au (bde.zeta.org.au [203.2.228.102]) by mailman.zeta.org.au (8.9.3/8.8.7) with ESMTP id WAA07307; Mon, 25 Nov 2002 22:54:29 +1100 Date: Mon, 25 Nov 2002 23:07:53 +1100 (EST) From: Bruce Evans X-X-Sender: bde@gamplex.bde.org To: Olivier Houchard Cc: freebsd-audit@FreeBSD.ORG Subject: Re: do_dup patch In-Reply-To: <20021125110902.GA10961@ci0.org> Message-ID: <20021125225927.O56791-100000@gamplex.bde.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Mon, 25 Nov 2002, Olivier Houchard wrote: > On Mon, Nov 25, 2002 at 09:32:05PM +1100, Bruce Evans wrote: > > On Mon, 25 Nov 2002, Olivier Houchard wrote: > > > > > This patch makes the "new" and "old" do_dup arguments unsigned int instead of > > > int. I can't see any reason they would have to be int, and right now a call to > > > dup or dup2 with an invalid negative fd, such as dup(-999999); will panic a > > > -CURRENT box (the problem doesn't exist on -STABLE because the file descriptor > > > validity is checked in dup() and dup2()). > > > Is there anything wrong with committing this ? > > > > Yes; it is sort of backwards. e > > > [since file descriptors should be (represented by) (signed) ints] > So what would be the best ? > Changing dup and dup2 prototypes in syscall.master to make new and old > signed instead of unsigned and checking that new >= 0 && old >= 0 in > do_dup ? I just added the bounds checks. Cleaning up the prototypes can wait. (There are hundreds of other wrong prototypes their anyway, most involving use of "int" or "u_int" instead of foo_t or not using "const".) This has not been tested at runtime. %%% Index: kern_descrip.c =================================================================== RCS file: /home/ncvs/src/sys/kern/kern_descrip.c,v retrieving revision 1.168 diff -u -2 -r1.168 kern_descrip.c --- kern_descrip.c 27 Oct 2002 18:07:41 -0000 1.168 +++ kern_descrip.c 25 Nov 2002 11:56:27 -0000 @@ -471,6 +475,6 @@ */ FILEDESC_LOCK(fdp); - if (old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || - new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || + if (old < 0 || old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || + new < 0 || new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || new >= maxfilesperproc) { FILEDESC_UNLOCK(fdp); %%% Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 4:21:46 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 5A91937B401 for ; Mon, 25 Nov 2002 04:21:45 -0800 (PST) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102]) by mx1.FreeBSD.org (Postfix) with ESMTP id 06E6143E88 for ; Mon, 25 Nov 2002 04:21:44 -0800 (PST) (envelope-from doginou@cognet.ci0.org) Received: from cognet.ci0.org (cognet.ci0.org [80.65.224.102] (may be forged)) by cognet.ci0.org (8.12.6/8.12.6) with ESMTP id gAPCHsFn011752; Mon, 25 Nov 2002 13:17:54 +0100 (CET) (envelope-from doginou@cognet.ci0.org) Received: (from doginou@localhost) by cognet.ci0.org (8.12.6/8.12.6/Submit) id gAPCHl5x011751; Mon, 25 Nov 2002 13:17:47 +0100 (CET) Date: Mon, 25 Nov 2002 13:17:47 +0100 From: Olivier Houchard To: Bruce Evans Cc: freebsd-audit@FreeBSD.ORG Subject: Re: do_dup patch Message-ID: <20021125121747.GA11569@ci0.org> References: <20021125110902.GA10961@ci0.org> <20021125225927.O56791-100000@gamplex.bde.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20021125225927.O56791-100000@gamplex.bde.org> User-Agent: Mutt/1.5.1i Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Mon, Nov 25, 2002 at 11:07:53PM +1100, Bruce Evans wrote: > On Mon, 25 Nov 2002, Olivier Houchard wrote: > > I just added the bounds checks. Cleaning up the prototypes can wait. > (There are hundreds of other wrong prototypes their anyway, most involving > use of "int" or "u_int" instead of foo_t or not using "const".) > > This has not been tested at runtime. > > %%% > Index: kern_descrip.c > =================================================================== > RCS file: /home/ncvs/src/sys/kern/kern_descrip.c,v > retrieving revision 1.168 > diff -u -2 -r1.168 kern_descrip.c > --- kern_descrip.c 27 Oct 2002 18:07:41 -0000 1.168 > +++ kern_descrip.c 25 Nov 2002 11:56:27 -0000 > @@ -471,6 +475,6 @@ > */ > FILEDESC_LOCK(fdp); > - if (old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || > - new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > + if (old < 0 || old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || > + new < 0 || new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > new >= maxfilesperproc) { > FILEDESC_UNLOCK(fdp); > %%% I just tested the same change and it is ok. Shall I submit it to re@ or will you do it ? Olivier To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 4:57:20 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 725B637B401 for ; Mon, 25 Nov 2002 04:57:19 -0800 (PST) Received: from mailman.zeta.org.au (mailman.zeta.org.au [203.26.10.16]) by mx1.FreeBSD.org (Postfix) with ESMTP id 72F8143E91 for ; Mon, 25 Nov 2002 04:57:18 -0800 (PST) (envelope-from bde@zeta.org.au) Received: from bde.zeta.org.au (bde.zeta.org.au [203.2.228.102]) by mailman.zeta.org.au (8.9.3/8.8.7) with ESMTP id XAA11958; Mon, 25 Nov 2002 23:56:56 +1100 Date: Tue, 26 Nov 2002 00:10:20 +1100 (EST) From: Bruce Evans X-X-Sender: bde@gamplex.bde.org To: Olivier Houchard Cc: freebsd-audit@FreeBSD.ORG Subject: Re: do_dup patch In-Reply-To: <20021125121747.GA11569@ci0.org> Message-ID: <20021126000957.B430-100000@gamplex.bde.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Mon, 25 Nov 2002, Olivier Houchard wrote: > On Mon, Nov 25, 2002 at 11:07:53PM +1100, Bruce Evans wrote: > > %%% > > Index: kern_descrip.c > > =================================================================== > > RCS file: /home/ncvs/src/sys/kern/kern_descrip.c,v > > retrieving revision 1.168 > > diff -u -2 -r1.168 kern_descrip.c > > --- kern_descrip.c 27 Oct 2002 18:07:41 -0000 1.168 > > +++ kern_descrip.c 25 Nov 2002 11:56:27 -0000 > > @@ -471,6 +475,6 @@ > > */ > > FILEDESC_LOCK(fdp); > > - if (old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || > > - new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > > + if (old < 0 || old >= fdp->fd_nfiles || fdp->fd_ofiles[old] == NULL || > > + new < 0 || new >= p->p_rlimit[RLIMIT_NOFILE].rlim_cur || > > new >= maxfilesperproc) { > > FILEDESC_UNLOCK(fdp); > > %%% > > I just tested the same change and it is ok. > Shall I submit it to re@ or will you do it ? Please do it. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Mon Nov 25 11: 0:23 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 814E037B406 for ; Mon, 25 Nov 2002 11:00:22 -0800 (PST) Received: from freefall.freebsd.org (freefall.freebsd.org [216.136.204.21]) by mx1.FreeBSD.org (Postfix) with ESMTP id DBD1443EAA for ; Mon, 25 Nov 2002 11:00:21 -0800 (PST) (envelope-from owner-bugmaster@freebsd.org) Received: from freefall.freebsd.org (peter@localhost [127.0.0.1]) by freefall.freebsd.org (8.12.6/8.12.6) with ESMTP id gAPJ0Lx3081554 for ; Mon, 25 Nov 2002 11:00:21 -0800 (PST) (envelope-from owner-bugmaster@freebsd.org) Received: (from peter@localhost) by freefall.freebsd.org (8.12.6/8.12.6/Submit) id gAPJ0LlC081542 for audit@freebsd.org; Mon, 25 Nov 2002 11:00:21 -0800 (PST) Date: Mon, 25 Nov 2002 11:00:21 -0800 (PST) Message-Id: <200211251900.gAPJ0LlC081542@freefall.freebsd.org> X-Authentication-Warning: freefall.freebsd.org: peter set sender to owner-bugmaster@freebsd.org using -f From: FreeBSD bugmaster To: audit@FreeBSD.org Subject: Current problem reports assigned to you Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Current FreeBSD problem reports Critical problems Serious problems Non-critical problems S Submitted Tracker Resp. Description ------------------------------------------------------------------------------- a [1999/01/28] bin/9770 audit An openpty(3) auxiliary program 1 problem total. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Wed Nov 27 0:55:16 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EA2DC37B401 for ; Wed, 27 Nov 2002 00:55:15 -0800 (PST) Received: from axl.seasidesoftware.co.za (axl.seasidesoftware.co.za [196.31.7.201]) by mx1.FreeBSD.org (Postfix) with ESMTP id 470A943EA9 for ; Wed, 27 Nov 2002 00:55:13 -0800 (PST) (envelope-from sheldonh@starjuice.net) Received: from sheldonh by axl.seasidesoftware.co.za with local (Exim 4.10) id 18GxyM-000BWW-00; Wed, 27 Nov 2002 10:55:06 +0200 Date: Wed, 27 Nov 2002 10:55:06 +0200 From: Sheldon Hearn To: Poul-Henning Kamp Cc: audit@freebsd.org Subject: Re: ABIs and 5.x branch: freeze kernel module ABI at 5.0 or 5.1? Message-ID: <20021127085506.GC43828@starjuice.net> References: <2079.1038351585@critter.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <2079.1038351585@critter.freebsd.dk> User-Agent: Mutt/1.5.1i Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On (2002/11/26 23:59), Poul-Henning Kamp wrote: > It's very simple in my mind: we only freeze ABI's on -stable branches > (and we actually even violated that for 4-stable I belive). > > Whenever we branch a new -stable from -current, that's when we freeze > the ABI's for that branch. I agree, but I think that this attitude warrants a different name for pre-branch releases. Things have changed to the point where a large number of users don't understand the implications of point-naught releases. I'd go so far as to say that 5.0-RELEASE should be renamed to 5.0-PREVIEW and that 5.1-RELEASE should be renamed to 5.0-RELEASE. I really didn't want to send this message because I know people will fixate on the name to use, instead of the issue that pre-branch releases should be named differently from post-branch releases. Ciao, Sheldon. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Wed Nov 27 5:42:44 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 98AF637B401 for ; Wed, 27 Nov 2002 05:42:43 -0800 (PST) Received: from relay1.macomnet.ru (relay1.macomnet.ru [195.128.64.10]) by mx1.FreeBSD.org (Postfix) with ESMTP id 4E08843EAF for ; Wed, 27 Nov 2002 05:42:42 -0800 (PST) (envelope-from maxim@FreeBSD.org) Received: from news1.macomnet.ru (news1.macomnet.ru [195.128.64.14]) by relay1.macomnet.ru (8.11.6/8.11.6) with ESMTP id gARDgcs5084625; Wed, 27 Nov 2002 16:42:38 +0300 (MSK) Date: Wed, 27 Nov 2002 16:42:38 +0300 (MSK) From: Maxim Konovalov To: Harti Brandt Cc: freebsd-audit@FreeBSD.org Subject: Re: kern/32827: small SO_RCVTIMEO values are taken to be zero In-Reply-To: <20021122094351.D40230-100000@beagle.fokus.gmd.de> Message-ID: <20021127164222.D52081-100000@news1.macomnet.ru> References: <20021122094351.D40230-100000@beagle.fokus.gmd.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On 11:46+0300, Nov 22, 2002, Harti Brandt wrote: > > Hi, > > could someone please have a look at > > http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/32827 > > so that it can be committed? fixed in rev. 1.138 src/sys/kern/uipc_socket.c in -CURRENT. Thanks! -- Maxim Konovalov, maxim@FreeBSD.org To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message From owner-freebsd-audit Fri Nov 29 7:34:51 2002 Delivered-To: freebsd-audit@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 32E6937B401 for ; Fri, 29 Nov 2002 07:34:49 -0800 (PST) Received: from mail.gmx.net (mail.gmx.net [213.165.65.60]) by mx1.FreeBSD.org (Postfix) with SMTP id 238E343EA9 for ; Fri, 29 Nov 2002 07:34:48 -0800 (PST) (envelope-from csharp@gmx.net) Received: (qmail 18237 invoked by uid 0); 29 Nov 2002 15:34:46 -0000 Received: from pd9552198.dip.t-dialin.net (HELO dagon.cthulhu.net) (217.85.33.152) by mail.gmx.net (mp004-rz3) with SMTP; 29 Nov 2002 15:34:46 -0000 Received: by dagon.cthulhu.net (Postfix, from userid 1001) id DED2B6191; Fri, 29 Nov 2002 16:37:42 +0100 (CET) Date: Fri, 29 Nov 2002 16:37:42 +0100 From: Christopher Sharp To: freebsd-audit@freebsd.org Subject: patch for the make regression test Message-ID: <20021129153742.GA11955@gmx.net> Mail-Followup-To: freebsd-audit@freebsd.org Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="FL5UXtIhxfXey3p5" Content-Disposition: inline User-Agent: Mutt/1.5.1i Sender: owner-freebsd-audit@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG --FL5UXtIhxfXey3p5 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Hello, as I stated in an email I wrote to -current yesterday I'm not really happy with the output of the regression test for make which is run before every command in src. Attached you can find a patch to make it IMHO nicer and shorter. I would be really glad if something like this or this patch could get commited. Comments ? - Christopher --FL5UXtIhxfXey3p5 Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="make.Makefile.patch" --- Makefile.old Thu Nov 28 14:21:32 2002 +++ Makefile Fri Nov 29 16:30:57 2002 @@ -19,28 +19,29 @@ NIL= all: - @echo "Running test variables" + @echo "Doing regression tests for make:" + @echo -n "Running test variables... " @echo 1:${DATA1} 2:${DATA2} 3:${DATA3} 4:${DATA4} 5:${DATA5} | \ diff -u ${.CURDIR}/regress.variables.out - || ${MAKE} failure - @echo "PASS: Test variables detected no regression, output matches." - @echo "Running test targets" + @echo "ok" + @echo -n "Running test targets... " @${MAKE} double || ${MAKE} failure - @echo "PASS: Test targets detected no regression." - @echo "Running test sysvmatch" + @echo "ok" + @echo -n "Running test sysvmatch... " @${MAKE} sysvmatch || ${MAKE} failure - @echo "PASS: Test sysvmatch detected no regression." - @echo "Running test lhs_expn" + @echo "ok" + @echo -n "Running test lhs_expn... " @! ${MAKE} lhs_expn && true || ${MAKE} failure - @echo "PASS: Test lhs_expn detected no regression." - @echo "Running test notdef" + @echo "ok" + @echo -n "Running test notdef... " @${MAKE} notdef || ${MAKE} failure - @echo "PASS: Test notdef detected no regression." - @echo "Running test modifiers" + @echo "ok" + @echo -n "Running test modifiers... " @${MAKE} modifiers || ${MAKE} failure - @echo "PASS: Test modifiers detected no regression." - @echo "Running test funny_targets" + @echo "ok" + @echo -n "Running test funny_targets... " @${MAKE} funny_targets || ${MAKE} failure - @echo "PASS: Test funny_targets detected no regression." + @echo "ok" .if make(double) # Doubly-defined targets. make(1) will warn, but use the "right" one. If it @@ -91,5 +92,5 @@ .endif failure: - @echo "FAIL: Test failed: regression detected. See above." + @echo "failed (regression detected)" @false --FL5UXtIhxfXey3p5-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-audit" in the body of the message