From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757391AbYJPVDW (ORCPT ); Thu, 16 Oct 2008 17:03:22 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754233AbYJPVDK (ORCPT ); Thu, 16 Oct 2008 17:03:10 -0400 Received: from fg-out-1718.google.com ([72.14.220.153]:26196 "EHLO fg-out-1718.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754232AbYJPVDJ (ORCPT ); Thu, 16 Oct 2008 17:03:09 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:reply-to:to:subject:cc:in-reply-to :mime-version:content-type:references; b=Kc8PsyRFXxZP1/nor9iw+0Gr+XJ/mFxziSxlCMqhEbQ13O2Qiz/JiLGs275JY4F5m5 ASyWJxbLHGTzwHwqa0bNmJ7t+pro0ouapz5BhFVmy5cjXXeLzGSPzsfuynzE6qVI6QrU HXnrgad+cf5GPRLzeoPX9FJ6ECGjEM8/IoCp4= Message-ID: Date: Thu, 16 Oct 2008 14:03:06 -0700 From: "Jeff Schroeder" Reply-To: jeffschroeder@computer.org To: "Bodo Eggert" <7eggert@gmx.de> Subject: Re: [PATCH] use tmpfs for rootfs v2 Cc: "Geert Uytterhoeven" , linux-kernel@vger.kernel.org In-Reply-To: MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_Part_131433_10735344.1224190986915" References: Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org ------=_Part_131433_10735344.1224190986915 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline On Thu, Oct 16, 2008 at 10:39 AM, Bodo Eggert <7eggert@gmx.de> wrote: > This patch allows chosing tmpfs instead of ramfs for the root filesystem. > Having tmpfs is usefull for running systems from RAM, because it does not > risk deadlocking the system and possibly allows swapping. Using tmpfs > increases the kernel by 10 bytes on x86_64. > > Having tmpfs as the root filesystem allows you to get rid of the now unused > ramfs and free some kernel memory. On my system, that's about 5198 bytes > compared to having a ramfs root. > > Off cause you can mount tmpfs and move around the data, but this is slower > (having to copy around the data), more error-prone and it uses more memory. > Besides that, I don't see the point in explicitely doing the wrong thing > and then having everybody fix it. > > > Signed-off-by: Bodo Eggert <7eggert@gmx.de> > > --- > Changes to v1: Include the change to the Makefile. > > Some numbers for my system: (from make and from size vmlinux) > > Setup is 10716 bytes (padded to 10752 bytes). > System is 2659 kB > text data bss dec hex filename > 5341667 874212 562992 6778871 676ff7 vmlinux > > having TMPFS_IS_ROOT: > Setup is 10716 bytes (padded to 10752 bytes). > System is 2659 kB > text data bss dec hex filename > 5341677 874212 562992 6778881 677001 vmlinux > +10 > > not having ramfs: > Setup is 10716 bytes (padded to 10752 bytes). > System is 2659 kB > text data bss dec hex filename > 5336733 873948 562992 6773673 675ba9 vmlinux > -4934 -264 > > > fs/Kconfig | 21 +++++++++++++++++++++ > fs/Makefile | 2 +- > fs/ramfs/inode.c | 39 +++++++++++++++++++++------------------ > mm/shmem.c | 26 ++++++++++++++++++++++++++ > 4 files changed, 69 insertions(+), 19 deletions(-) > > diff -pruNXdontdiff linux-2.6.27.2-numlock/fs/Kconfig linux-2.6.27.3-tmpfs/fs/Kconfig > --- linux-2.6.27.2-numlock/fs/Kconfig 2008-10-10 13:18:34.000000000 +0200 > +++ linux-2.6.27.3-tmpfs/fs/Kconfig 2008-10-11 15:42:21.000000000 +0200 > @@ -953,6 +953,27 @@ config TMPFS_POSIX_ACL > > If you don't know what Access Control Lists are, say N. > > +config TMPFS_ROOT > + bool "Use tmpfs instrad of ramfs for initramfs" > + depends on TMPFS && SHMEM > + default n > + help > + This replaces the ramfs used for unpacking the cpio images > + with tmpfs. > + > + If unsure, say N > + > +config RAMFS > + bool "Ramfs file system support" if TMPFS_ROOT > + default y > + ---help--- > + Ramfs is a file system which keeps all files in RAM. Unlike tmpfs, > + it cannot be swapped to disk, and it has the risk of deadlocking > + the system by using all memory. > + > + Ramfs is used for booting the system and unpacking the cpio if > + TMPFS_ROOT is not set. > + > config HUGETLBFS > bool "HugeTLB file system support" > depends on X86 || IA64 || PPC64 || SPARC64 || (SUPERH && MMU) || \ > diff -pruNXdontdiff linux-2.6.27.2-numlock/fs/ramfs/inode.c linux-2.6.27.3-tmpfs/fs/ramfs/inode.c > --- linux-2.6.27.2-numlock/fs/ramfs/inode.c 2008-07-13 23:51:29.000000000 +0200 > +++ linux-2.6.27.3-tmpfs/fs/ramfs/inode.c 2008-10-11 15:50:03.000000000 +0200 > @@ -190,6 +190,13 @@ int ramfs_get_sb(struct file_system_type > return get_sb_nodev(fs_type, flags, data, ramfs_fill_super, mnt); > } > > +static struct file_system_type ramfs_fs_type = { > + .name = "ramfs", > + .get_sb = ramfs_get_sb, > + .kill_sb = kill_litter_super, > +}; > + > +#ifndef CONFIG_TMPFS_ROOT > static int rootfs_get_sb(struct file_system_type *fs_type, > int flags, const char *dev_name, void *data, struct vfsmount *mnt) > { > @@ -197,30 +204,12 @@ static int rootfs_get_sb(struct file_sys > mnt); > } > > -static struct file_system_type ramfs_fs_type = { > - .name = "ramfs", > - .get_sb = ramfs_get_sb, > - .kill_sb = kill_litter_super, > -}; > static struct file_system_type rootfs_fs_type = { > .name = "rootfs", > .get_sb = rootfs_get_sb, > .kill_sb = kill_litter_super, > }; > > -static int __init init_ramfs_fs(void) > -{ > - return register_filesystem(&ramfs_fs_type); > -} > - > -static void __exit exit_ramfs_fs(void) > -{ > - unregister_filesystem(&ramfs_fs_type); > -} > - > -module_init(init_ramfs_fs) > -module_exit(exit_ramfs_fs) > - > int __init init_rootfs(void) > { > int err; > @@ -235,5 +224,19 @@ int __init init_rootfs(void) > > return err; > } > +#endif > + > +static int __init init_ramfs_fs(void) > +{ > + return register_filesystem(&ramfs_fs_type); > +} > + > +static void __exit exit_ramfs_fs(void) > +{ > + unregister_filesystem(&ramfs_fs_type); > +} > + > +module_init(init_ramfs_fs) > +module_exit(exit_ramfs_fs) > > MODULE_LICENSE("GPL"); > diff -pruNXdontdiff linux-2.6.27.2-numlock/mm/shmem.c linux-2.6.27.3-tmpfs/mm/shmem.c > --- linux-2.6.27.2-numlock/mm/shmem.c 2008-10-10 13:18:47.000000000 +0200 > +++ linux-2.6.27.3-tmpfs/mm/shmem.c 2008-10-11 16:05:52.000000000 +0200 > @@ -2366,6 +2366,10 @@ static void init_once(void *foo) > > static int init_inodecache(void) > { > +#ifdef CONFIG_TMPFS_ROOT > + if (shmem_inode_cachep) > + return 0; > +#endif > shmem_inode_cachep = kmem_cache_create("shmem_inode_cache", > sizeof(struct shmem_inode_info), > 0, SLAB_PANIC, init_once); > @@ -2583,6 +2587,28 @@ put_memory: > return ERR_PTR(error); > } > > +#ifdef CONFIG_TMPFS_ROOT > +static int rootfs_get_sb(struct file_system_type *fs_type, > + int flags, const char *dev_name, void *data, struct vfsmount *mnt) > +{ > + return get_sb_nodev(fs_type, flags, data, shmem_fill_super, mnt); > +} > + > + > + static struct file_system_type rootfs_fs_type = { > + .name = "rootfs", > + .get_sb = rootfs_get_sb, > + .kill_sb = kill_litter_super, > + }; > + > + int __init init_rootfs(void) > + { > + if (init_inodecache()) > + panic("Can't initialize shm inode cache"); > + return register_filesystem(&rootfs_fs_type); > + } > + #endif > + > /** > * shmem_zero_setup - setup a shared anonymous mapping > * @vma: the vma to be mmapped is prepared by do_mmap_pgoff > diff -pruNXdontdiff linux-2.6.27.2-numlock/fs/Makefile linux-2.6.27.3-tmpfs/fs/Makefile > --- linux-2.6.27.2-numlock/fs/Makefile 2008-10-16 19:25:20.000000000 +0200 > +++ linux-2.6.27.3-tmpfs/fs/Makefile 2008-10-16 19:24:00.000000000 +0200 > @@ -67,6 +67,7 @@ obj-$(CONFIG_PROFILING) += dcookies.o > obj-$(CONFIG_DLM) += dlm/ > > # Do not add any filesystems before this line > +obj-$(CONFIG_RAMFS) += ramfs/ > obj-$(CONFIG_REISERFS_FS) += reiserfs/ > obj-$(CONFIG_EXT3_FS) += ext3/ # Before ext2 so root fs can be ext3 > obj-$(CONFIG_EXT4DEV_FS) += ext4/ # Before ext2 so root fs can be ext4dev > @@ -74,7 +75,6 @@ obj-$(CONFIG_JBD) += jbd/ > obj-$(CONFIG_JBD2) += jbd2/ > obj-$(CONFIG_EXT2_FS) += ext2/ > obj-$(CONFIG_CRAMFS) += cramfs/ > -obj-y += ramfs/ > obj-$(CONFIG_HUGETLBFS) += hugetlbfs/ > obj-$(CONFIG_CODA_FS) += coda/ > obj-$(CONFIG_MINIX_FS) += minix/ Haven't gotten the hang of attaching patches inline using gmail yet. Here is a patch ontop of the previous one that fixes whitespace problems pointed out by checkpatch.pl. It is based on 2.7.27 3fa8749e584b55f1180411ab1b51117190bac1e5 It is also available here: http://www.digitalprognosis.com/opensource/patches/linux-2.6/0001-Cleanup-the-whitespace-to-be-checkpatch-compliant.patch -- Jeff Schroeder Don't drink and derive, alcohol and analysis don't mix. http://www.digitalprognosis.com ------=_Part_131433_10735344.1224190986915 Content-Type: text/x-diff; name=0001-Cleanup-the-whitespace-to-be-checkpatch-compliant.patch Content-Transfer-Encoding: base64 X-Attachment-Id: f_fmdve5wo0 Content-Disposition: attachment; filename=0001-Cleanup-the-whitespace-to-be-checkpatch-compliant.patch RnJvbSBlMWZlMGI4MTI5ZDE3NjljODM0YmZmNDI2MmJhYjM1NmU4ZDQ0NGJiIE1vbiBTZXAgMTcg MDA6MDA6MDAgMjAwMQpGcm9tOiBKZWZmIFNjaHJvZWRlciA8amVmZnNjaHJvZWRlckBjb21wdXRl ci5vcmc+CkRhdGU6IFRodSwgMTYgT2N0IDIwMDggMTM6NDc6MjAgLTA3MDAKU3ViamVjdDogW1BB VENIXSBDbGVhbnVwIHRoZSB3aGl0ZXNwYWNlIHRvIGJlIGNoZWNrcGF0Y2ggY29tcGxpYW50CgpB cHBsaWVzIG9udG9wIG9mOiBodHRwOi8vbGttbC5vcmcvbGttbC9kaWZmLzIwMDgvMTAvMTYvMzg4 LzEKU2lnbmVkLW9mZi1ieTogSmVmZiBTY2hyb2VkZXIgPGplZmZzY2hyb2VkZXJAY29tcHV0ZXIu b3JnPgotLS0KIG1tL3NobWVtLmMgfCAgIDMwICsrKysrKysrKysrKysrKy0tLS0tLS0tLS0tLS0t LQogMSBmaWxlcyBjaGFuZ2VkLCAxNSBpbnNlcnRpb25zKCspLCAxNSBkZWxldGlvbnMoLSkKCmRp ZmYgLS1naXQgYS9tbS9zaG1lbS5jIGIvbW0vc2htZW0uYwppbmRleCBhMWQyMDQyLi5kMWIzY2Q4 IDEwMDY0NAotLS0gYS9tbS9zaG1lbS5jCisrKyBiL21tL3NobWVtLmMKQEAgLTI1OTQsMjEgKzI1 OTQsMjEgQEAgc3RhdGljIGludCByb290ZnNfZ2V0X3NiKHN0cnVjdCBmaWxlX3N5c3RlbV90eXBl ICpmc190eXBlLAogCXJldHVybiBnZXRfc2Jfbm9kZXYoZnNfdHlwZSwgZmxhZ3MsIGRhdGEsIHNo bWVtX2ZpbGxfc3VwZXIsIG1udCk7CiB9CiAKLSAKLSBzdGF0aWMgc3RydWN0IGZpbGVfc3lzdGVt X3R5cGUgcm9vdGZzX2ZzX3R5cGUgPSB7Ci0gCS5uYW1lCQk9ICJyb290ZnMiLAotIAkuZ2V0X3Ni CQk9IHJvb3Rmc19nZXRfc2IsCi0gCS5raWxsX3NiCT0ga2lsbF9saXR0ZXJfc3VwZXIsCi0gfTsK LSAKLSBpbnQgX19pbml0IGluaXRfcm9vdGZzKHZvaWQpCi0gewotIAlpZiAoaW5pdF9pbm9kZWNh Y2hlKCkpCi0gCQlwYW5pYygiQ2FuJ3QgaW5pdGlhbGl6ZSBzaG0gaW5vZGUgY2FjaGUiKTsKLSAJ cmV0dXJuIHJlZ2lzdGVyX2ZpbGVzeXN0ZW0oJnJvb3Rmc19mc190eXBlKTsKLSB9Ci0gI2VuZGlm Ci0gCisKK3N0YXRpYyBzdHJ1Y3QgZmlsZV9zeXN0ZW1fdHlwZSByb290ZnNfZnNfdHlwZSA9IHsK KwkubmFtZQkJPSAicm9vdGZzIiwKKwkuZ2V0X3NiCQk9IHJvb3Rmc19nZXRfc2IsCisJLmtpbGxf c2IJPSBraWxsX2xpdHRlcl9zdXBlciwKK307CisKK2ludCBfX2luaXQgaW5pdF9yb290ZnModm9p ZCkKK3sKKwlpZiAoaW5pdF9pbm9kZWNhY2hlKCkpCisJCXBhbmljKCJDYW4ndCBpbml0aWFsaXpl IHNobSBpbm9kZSBjYWNoZSIpOworCXJldHVybiByZWdpc3Rlcl9maWxlc3lzdGVtKCZyb290ZnNf ZnNfdHlwZSk7Cit9CisjZW5kaWYKKwogLyoqCiAgKiBzaG1lbV96ZXJvX3NldHVwIC0gc2V0dXAg YSBzaGFyZWQgYW5vbnltb3VzIG1hcHBpbmcKICAqIEB2bWE6IHRoZSB2bWEgdG8gYmUgbW1hcHBl ZCBpcyBwcmVwYXJlZCBieSBkb19tbWFwX3Bnb2ZmCi0tIAoxLjUuNC4zCgo= ------=_Part_131433_10735344.1224190986915--