From owner-svn-src-head@freebsd.org Fri Sep 29 22:13:27 2017 Return-Path: Delivered-To: svn-src-head@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id DC5F2E3610C; Fri, 29 Sep 2017 22:13:27 +0000 (UTC) (envelope-from ian@FreeBSD.org) Received: from repo.freebsd.org (repo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:0]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id AAB4C76E47; Fri, 29 Sep 2017 22:13:27 +0000 (UTC) (envelope-from ian@FreeBSD.org) Received: from repo.freebsd.org ([127.0.1.37]) by repo.freebsd.org (8.15.2/8.15.2) with ESMTP id v8TMDQLX081359; Fri, 29 Sep 2017 22:13:26 GMT (envelope-from ian@FreeBSD.org) Received: (from ian@localhost) by repo.freebsd.org (8.15.2/8.15.2/Submit) id v8TMDQLN081354; Fri, 29 Sep 2017 22:13:26 GMT (envelope-from ian@FreeBSD.org) Message-Id: <201709292213.v8TMDQLN081354@repo.freebsd.org> X-Authentication-Warning: repo.freebsd.org: ian set sender to ian@FreeBSD.org using -f From: Ian Lepore Date: Fri, 29 Sep 2017 22:13:26 +0000 (UTC) To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org Subject: svn commit: r324107 - in head: etc etc/defaults sbin/mdmfs X-SVN-Group: head X-SVN-Commit-Author: ian X-SVN-Commit-Paths: in head: etc etc/defaults sbin/mdmfs X-SVN-Commit-Revision: 324107 X-SVN-Commit-Repository: base MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-BeenThere: svn-src-head@freebsd.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: SVN commit messages for the src tree for head/-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 29 Sep 2017 22:13:28 -0000 Author: ian Date: Fri Sep 29 22:13:26 2017 New Revision: 324107 URL: https://svnweb.freebsd.org/changeset/base/324107 Log: Enhance mdmfs(8) to work with tmpfs(5). Existing scripts and associated config such as rc.initdiskless, rc.d/var, and others, use mdmfs to create memory filesystems. That program accepts a size argument which allows SI suffixes and treats an unsuffixed number as a count of 512 byte sectors. That makes it difficult to convert existing scripts to use tmpfs instead of mdmfs, because tmpfs treats unsuffixed numbers as a count of bytes. The script logic to deal with existing user config that might include suffixed and unsuffixed numbers is... unpleasant. Also, there is no g'tee that tmpfs will be available. It is sometimes configured out of small-resource embedded systems to save memory and flash storage space. These changes enhance mdmfs(8) so that it accepts two new values for the 'md-device' arg: 'tmpfs' and 'auto'. With tmpfs, the program always uses tmpfs(5) (and fails if it's not available). With 'auto' the program prefers tmpfs, but falls back to using md(4) if tmpfs isn't available. It also handles the -s argument so that the mdconfig interpetation of unsuffixed numbers applies when tmpfs is used as well, so that existing user config keeps working after a switch to tmpfs. A new rc setting, mfs_type, is added to etc/defaults/rc.conf to let users force the use of tmpfs or md; the default value is "auto". Differential Revision: https://reviews.freebsd.org/D12301 Modified: head/etc/defaults/rc.conf head/etc/rc.initdiskless head/etc/rc.subr head/sbin/mdmfs/mdmfs.8 head/sbin/mdmfs/mdmfs.c Modified: head/etc/defaults/rc.conf ============================================================================== --- head/etc/defaults/rc.conf Fri Sep 29 19:56:09 2017 (r324106) +++ head/etc/defaults/rc.conf Fri Sep 29 22:13:26 2017 (r324107) @@ -53,6 +53,7 @@ tmpmfs_flags="-S" # Extra mdmfs options for the mfs /t varmfs="AUTO" # Set to YES to always create an mfs /var, NO to never varsize="32m" # Size of mfs /var if created varmfs_flags="-S" # Extra mount options for the mfs /var +mfs_type="auto" # "md", "tmpfs", "auto" to prefer tmpfs with md as fallback populate_var="AUTO" # Set to YES to always (re)populate /var, NO to never cleanvar_enable="YES" # Clean the /var directory local_startup="/usr/local/etc/rc.d" # startup script dirs. Modified: head/etc/rc.initdiskless ============================================================================== --- head/etc/rc.initdiskless Fri Sep 29 19:56:09 2017 (r324106) +++ head/etc/rc.initdiskless Fri Sep 29 22:13:26 2017 (r324107) @@ -195,10 +195,10 @@ handle_remount() { # $1 = mount point to_umount="$b ${to_umount}" } -# Create a generic memory disk (using tmpfs) -# +# Create a generic memory disk. +# The 'auto' parameter will attempt to use tmpfs(5), falls back to md(4). mount_md() { - mount -t tmpfs -o size=$(($1 * 512)) tmpfs $2 + /sbin/mdmfs $flags -s $1 auto $2 } # Create the memory filesystem if it has not already been created Modified: head/etc/rc.subr ============================================================================== --- head/etc/rc.subr Fri Sep 29 19:56:09 2017 (r324106) +++ head/etc/rc.subr Fri Sep 29 22:13:26 2017 (r324107) @@ -1790,7 +1790,7 @@ mount_md() if [ -n "$3" ]; then flags="$3" fi - /sbin/mdmfs $flags -s $1 md $2 + /sbin/mdmfs $flags -s $1 ${mfs_type} $2 } # Code common to scripts that need to load a kernel module Modified: head/sbin/mdmfs/mdmfs.8 ============================================================================== --- head/sbin/mdmfs/mdmfs.8 Fri Sep 29 19:56:09 2017 (r324106) +++ head/sbin/mdmfs/mdmfs.8 Fri Sep 29 22:13:26 2017 (r324107) @@ -25,7 +25,7 @@ .\" .\" $FreeBSD$ .\" -.Dd September 9, 2016 +.Dd September 9, 2017 .Dt MDMFS 8 .Os .Sh NAME @@ -33,7 +33,9 @@ .Nm mount_mfs .Nd configure and mount an in-memory file system using the .Xr md 4 -driver +driver or the +.Xr tmpfs 5 +filesystem .Sh SYNOPSIS .Nm .Op Fl DLlMNnPStTUX @@ -63,9 +65,13 @@ utility is designed to be a work-alike and look-alike .Xr mount_mfs 8 . The end result is essentially the same, but is accomplished in a completely different way. -The +Based on +.Ar md-device , +the .Nm -utility configures an +utility either creates a +.Xr tmpfs 5 +filesystem, or it configures an .Xr md 4 disk using .Xr mdconfig 8 , @@ -81,6 +87,44 @@ compressed disk images, as long as the kernel supports All the command line options are passed to the appropriate program at the appropriate stage in order to achieve the desired effect. .Pp +When +.Ar md-device +is `auto', +.Nm +uses +.Xr tmpfs 5 +if it is present in the kernel or can be loaded as a module, +otherwise it falls back to using +.Xr md 4 +auto-unit as if `md' had been specified. +.Pp +When +.Ar md-device +is `tmpfs', +.Nm +mounts a +.Xr tmpfs 5 +filesystem, translating the +.Fl s +size option, if present, into a `-o size=' mount option. +Any +.Fl o +options on the command line are passed through to the +.Xr tmpfs 5 +mount. +Options specific to +.Xr mdconfig 8 +or +.Xr newfs 8 +are ignored. +.Pp +When +.Ar md-device +does not result in +.Xr tmpfs 5 +being used, then an +.Xr md 4 +device is configured instead. By default, .Nm creates a swap-based @@ -219,14 +263,10 @@ is .Em not specified. That is, -this will work for the default swap-backed -.Pq Dv MD_SWAP -disks, -and the optional -.Pq Fl M -.Xr malloc 9 -backed disks -.Pq Dv MD_MALLOC . +this will work when the backing storage is some form of +memory, as opposed to a fixed-size file. +The size may include the usual SI suffixes (k, m, g, t, p). +A number without a suffix is interpreted as a count of 512-byte sectors. .It Fl t Turn on the TRIM enable flag for .Xr newfs 8 . @@ -392,6 +432,7 @@ was given on the command line. .Sh SEE ALSO .Xr md 4 , .Xr fstab 5 , +.Xr tmpfs 5 , .Xr mdconfig 8 , .Xr mount 8 , .Xr newfs 8 Modified: head/sbin/mdmfs/mdmfs.c ============================================================================== --- head/sbin/mdmfs/mdmfs.c Fri Sep 29 19:56:09 2017 (r324106) +++ head/sbin/mdmfs/mdmfs.c Fri Sep 29 22:13:26 2017 (r324107) @@ -34,15 +34,19 @@ __FBSDID("$FreeBSD$"); #include +#include #include +#include #include #include #include #include #include +#include #include #include +#include #include #include #include @@ -78,7 +82,8 @@ static void debugprintf(const char *, ...) __printfli static void do_mdconfig_attach(const char *, const enum md_types); static void do_mdconfig_attach_au(const char *, const enum md_types); static void do_mdconfig_detach(void); -static void do_mount(const char *, const char *); +static void do_mount_md(const char *, const char *); +static void do_mount_tmpfs(const char *, const char *); static void do_mtptsetup(const char *, struct mtpt_info *); static void do_newfs(const char *); static void extract_ugid(const char *, struct mtpt_info *); @@ -89,14 +94,15 @@ int main(int argc, char **argv) { struct mtpt_info mi; /* Mountpoint info. */ + intmax_t mdsize; char *mdconfig_arg, *newfs_arg, /* Args to helper programs. */ *mount_arg; enum md_types mdtype; /* The type of our memory disk. */ - bool have_mdtype; + bool have_mdtype, mlmac; bool detach, softdep, autounit, newfs; - char *mtpoint, *unitstr; + const char *mtpoint, *size_arg, *unitstr; char *p; - int ch; + int ch, idx; void *set; unsigned long ul; @@ -105,6 +111,7 @@ main(int argc, char **argv) detach = true; softdep = true; autounit = false; + mlmac = false; newfs = true; have_mdtype = false; mdtype = MD_SWAP; @@ -119,6 +126,7 @@ main(int argc, char **argv) mdconfig_arg = strdup(""); newfs_arg = strdup(""); mount_arg = strdup(""); + size_arg = NULL; /* If we were started as mount_mfs or mfs, imply -C. */ if (strcmp(getprogname(), "mount_mfs") == 0 || @@ -175,6 +183,7 @@ main(int argc, char **argv) loudsubs = true; break; case 'l': + mlmac = true; argappend(&newfs_arg, "-l"); break; case 'M': @@ -213,7 +222,7 @@ main(int argc, char **argv) softdep = false; break; case 's': - argappend(&mdconfig_arg, "-s %s", optarg); + size_arg = optarg; break; case 't': argappend(&newfs_arg, "-t"); @@ -242,42 +251,107 @@ main(int argc, char **argv) if (argc < 2) usage(); - /* Derive 'unit' (global). */ - unitstr = argv[0]; - if (strncmp(unitstr, "/dev/", 5) == 0) - unitstr += 5; - if (strncmp(unitstr, mdname, mdnamelen) == 0) - unitstr += mdnamelen; - if (!isdigit(*unitstr)) { - autounit = true; - unit = -1; - mdsuffix = unitstr; - } else { - ul = strtoul(unitstr, &p, 10); - if (ul == ULONG_MAX) - errx(1, "bad device unit: %s", unitstr); - unit = ul; - mdsuffix = p; /* can be empty */ + /* + * Historically our size arg was passed directly to mdconfig, which + * treats a number without a suffix as a count of 512-byte sectors; + * tmpfs would treat it as a count of bytes. To get predictable + * behavior for 'auto' we document that the size always uses mdconfig + * rules. To make that work, decode the size here so it can be passed + * to either tmpfs or mdconfig as a count of bytes. + */ + if (size_arg != NULL) { + mdsize = (intmax_t)strtoumax(size_arg, &p, 0); + if (p == size_arg || (p[0] != 0 && p[1] != 0) || mdsize < 0) + errx(1, "invalid size '%s'", size_arg); + switch (*p) { + case 'p': + case 'P': + mdsize *= 1024; + case 't': + case 'T': + mdsize *= 1024; + case 'g': + case 'G': + mdsize *= 1024; + case 'm': + case 'M': + mdsize *= 1024; + case 'k': + case 'K': + mdsize *= 1024; + case 'b': + case 'B': + break; + case '\0': + mdsize *= 512; + break; + default: + errx(1, "invalid size suffix on '%s'", size_arg); + } } + /* + * Based on the command line 'md-device' either mount a tmpfs filesystem + * or configure the md device then format and mount a filesystem on it. + * If the device is 'auto' use tmpfs if it is available and there is no + * request for multilabel MAC (which tmpfs does not support). + */ + unitstr = argv[0]; mtpoint = argv[1]; - if (!have_mdtype) - mdtype = MD_SWAP; - if (softdep) - argappend(&newfs_arg, "-U"); - if (mdtype != MD_VNODE && !newfs) - errx(1, "-P requires a vnode-backed disk"); - /* Do the work. */ - if (detach && !autounit) - do_mdconfig_detach(); - if (autounit) - do_mdconfig_attach_au(mdconfig_arg, mdtype); - else - do_mdconfig_attach(mdconfig_arg, mdtype); - if (newfs) - do_newfs(newfs_arg); - do_mount(mount_arg, mtpoint); + if (strcmp(unitstr, "auto") == 0) { + if (mlmac) + idx = -1; /* Must use md for mlmac. */ + else if ((idx = modfind("tmpfs")) == -1) + idx = kldload("tmpfs"); + if (idx == -1) + unitstr = "md"; + else + unitstr = "tmpfs"; + } + + if (strcmp(unitstr, "tmpfs") == 0) { + if (size_arg != NULL && mdsize != 0) + argappend(&mount_arg, "-o size=%jd", mdsize); + do_mount_tmpfs(mount_arg, mtpoint); + } else { + if (size_arg != NULL) + argappend(&mdconfig_arg, "-s %jdB", mdsize); + if (strncmp(unitstr, "/dev/", 5) == 0) + unitstr += 5; + if (strncmp(unitstr, mdname, mdnamelen) == 0) + unitstr += mdnamelen; + if (!isdigit(*unitstr)) { + autounit = true; + unit = -1; + mdsuffix = unitstr; + } else { + ul = strtoul(unitstr, &p, 10); + if (ul == ULONG_MAX) + errx(1, "bad device unit: %s", unitstr); + unit = ul; + mdsuffix = p; /* can be empty */ + } + + if (!have_mdtype) + mdtype = MD_SWAP; + if (softdep) + argappend(&newfs_arg, "-U"); + if (mdtype != MD_VNODE && !newfs) + errx(1, "-P requires a vnode-backed disk"); + + /* Do the work. */ + if (detach && !autounit) + do_mdconfig_detach(); + if (autounit) + do_mdconfig_attach_au(mdconfig_arg, mdtype); + else + do_mdconfig_attach(mdconfig_arg, mdtype); + if (newfs) + do_newfs(newfs_arg); + do_mount_md(mount_arg, mtpoint); + } + do_mtptsetup(mtpoint, &mi); return (0); @@ -434,7 +508,7 @@ do_mdconfig_detach(void) * Mount the configured memory disk. */ static void -do_mount(const char *args, const char *mtpoint) +do_mount_md(const char *args, const char *mtpoint) { int rv; @@ -442,6 +516,19 @@ do_mount(const char *args, const char *mtpoint) mdname, unit, mdsuffix, mtpoint); if (rv) errx(1, "mount exited with error code %d", rv); +} + +/* + * Mount the configured tmpfs. + */ +static void +do_mount_tmpfs(const char *args, const char *mtpoint) +{ + int rv; + + rv = run(NULL, "%s -t tmpfs %s tmp %s", _PATH_MOUNT, args, mtpoint); + if (rv) + errx(1, "tmpfs mount exited with error code %d", rv); } /*